Documentos de Académico
Documentos de Profesional
Documentos de Cultura
ENGINEERING DESIGN
HANDBOOK
EXPLOSIONS IN AIR
PART ONE
AMC PAMPHLET
No. 706-181 15 July 1974
TABLE OF CONTENTS
Paragraph Page
LIST OF TABLES xv
PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii
Characteristics : . 1-3
AMCP 706-181
Paragraph Page
References . 1-24
ii
AMCP 706-181
Paragraph Page
a Shock . 2-21
References . 2-22
References . 3-24
iii
AMCP 706-181
Paragraph Page
of State . 4-13
Material) . 4-27
References . 4-35
Conditions . 5-13
References . 5-19
iv
AMCP 706-181
Paragraph Page
6-0
List of Symbols . 6-1
6-1
General . 6-2
6-2
Sources of Compiled Data on Air Blast . 6-2
6-3
Generation of Tables and Graphs of Air Blast Wave
Properties . 6-3
6-3.2
ImpUlses and Durations . 6-9
6-3.3
Time Constant and Initial Decay Rate . 6-11
6-3.4
Oblique Reflection Data . 6-12
6-3.5
Conversion Factors . 6-15
6-4
Example Calculations . 6-17
References . 6-21
7-0
List of Symbols . 7-1
7-1
General . 7-1
7-2
Pressure Transducers . 7-1
7-2.1
Side-on Gages . 7-1
7-2.1.1
BRL Side-on Gages . 7-2
7-2.1.2
Southwest Research Institute Side-on Gages . 7-2
7-2.1.3
Atlantic Research Corporation Side-on Gages . 7-4
7-2.1.4
British Side-on Gages . 7-4
7-2.1.5
Other Side-on Gages . 7-6
7-2.2
Reflected Pressure Gages . 7-6
7-2.3
Miniature Pressure Gages . 7-7
7-2.3.1
BRL Miniature Transducers . 7-8
7-2.3.2
Langley Research Center Miniature Transducers . 7-8
7-2.3.3
Other Minature Transducers . 7-10
7-3
Arrival-time Gages and Zero-time Markers . 7-14
7-4
Total Head Gages and Drag Gages . 7-17
7-4.1
Total Head Gages . 7-17
7-4.2
Drag Gages . 7-17
7-4.2.1
Drag Gage of Johnson and Ewing . 7-18
7-4.2.2
NOL Drag Force Gages . 7-18
7-4.2.3
SRI Drag Probes . 7-18
7-4.2.4
BRL Biaxial Drag Gage . 7-19
7-5
Density Gage . 7-19
7-6
Impulse Transducers . 7-20
AMCP7.181
Paragraph Page
References . 7-33
Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-15
vi
AMa' 706-181
Paragraph Page
8-4
Calibration Techniques .
8-20
References .
8-25
9-1
General . 9-1
9-2
Motion Picture Equipment . 9-1
9-2.1
Low-speed Framing Cameras . 9-1
9-2.2
High-speed Framing Cameras . 9-1
9-2.2.1
Rotating Prism Cameras
. 9-1
9-2.2.2
Rotating Drum Cameras
. 9-2
9-2.2.3
Rotating Mirror Cameras
. 9-3
9-2.2.4
Image Dissector Cameras
. 9-4
9-3
Streak Photography Equipment . 9-6
9-4
Still Photography Equipment . 9-7
9-4.1
Conventional Cameras . 9-7
9-4.2
Fast Shutter Cameras . 9-7
9-4.3
Image Converter Cameras . 9-9
9-5
Shadowgraph and Schlieren Equipment . 9-13
9-5.1
Shadowgraph Equipment. . 9-13
9-5.2
Schlieren Equipment . 9-14
9-6
Techniques in Photography of Air Blast Waves . 9-15
References . 9-22
10-0
List of Symbols . 10-1
10-1
General ' . 10-1
10-2
Reduction of Film and Paper Traces . 10-1
10-3
Reduction of Magnetic Tape Data . 10-9
10-4
Reduction of Data from Self-recording Gages . 10-10
10-5
Reduction of Data from Motion or Still
Photographs . 10-10
10-6
Other Data Reduction . 10-12
References . 10-13
BIBLIOGRAPHY . B-1
INDEX . I-I
vii
AMCP 706-181
LIST OF ILLUSTRATIONS
viii
AMCP 706181
2-7 r
Variation of Shock Strength vs Dimensionless
Shock Radius Rs/R o for Spherical Blast Waves,
'Y = 1.4 . 2-15
2-8 Spherical Blast Wave . 2-18
2-9 Cylindrical Blast Wave . 2-18
2-10 Plane Blast Wave . 2-19
3-1 Hopkinson Blast Wave Scaling . 3-4
3-2 Pressure-distance Curves for Ground-burst Blast of
Bare Charges . 3-4
3-3 Experimental Positive Impulses vs Distance Curves
(on ground) from Various Sources . 3-5
3-4 Comparisons of Peak Particle Velocities for Surface
Burst TNT Charges . 3-6
3-5 Peak Overpressure vs Sachs' Scaled Distance . 3-14
3-6 Hopkinson Scaled Impulse vs Scaled Distance . 3-15
3-7 Sachs' Scaled Impulse vs Scaled Distance . 3-16
3-8 pvsR for Various Values of~, High-p Range . 3-17
3-9 Comparison of Predicted and Measured Reflected
Impulse I r -Sea Level Conditions . 3-18
3-10 Comparison of Predicted and Measured Reflected
Impulse I, - Reduced Pressure Ambient Conditions 3-19
3-11 "Replica" Scaling of Response of Structures to
Blast Loading . 3-21
3-12 "Replica" Scaling of Elastic Response of Aluminum
Cantilevers to Air Blast Waves . 3-22
3-13 "Replica" Scaling of Permanent Deformation of
Aluminum Cantilevers Under Air Blast Loading . 3-22
3-14 Peak Overpressure Ratio vs Scaled Distance . 3-24
4-1 Peak Excess Pressure Ratio vs Distance in Charge
Radii for Pentolite at a Loading Density of
1.65 g/cm 3 . . . . . . . . . . . . . . . . . . .
4-5
4-2 Initial Singularity in Method of Characteristics .... 4-6
4-3 Schematic of Region of Numerical Solution for
Method of Characteristics . 4-8
4-4 Physical Plane Showing the Main Shock, Contact
Surface, and Second Shock . 4-9
4-5 Physical Plane (A-r) Showing Shock Front, Contact
Surface, and Second Shock for Cases With Equal
Initial Energy and Equal Initial Mass . 4-10
4-6 Pressure as a Function of Lagrange Position R o for
the Point-source Solution at Times Indicated .... 4-12
4-7 A Comparison of the PAF Detached Bow Wave
Positions (Dashed Lines) After Impact With Those
Observed in a Shock Tube Experiment Involving
a Mach 1.35 Flow Past a Wedge . 4-22
4-8 The Steady-state Detached Shock Front Position
in a Mach 1.58 Flow Past a Blunt Cylinder. . 4-23
ix
AMCP 706-181
x
AMCP 706-181
5-17 Measured Arrival Times for Flat Top I, II, and III
Compared With Prediction . 5-10
5-18 Measured Positive Duration for Flat Top I, II, and
III Compared With Prediction . 5-11
5-19 Measured Overpressure for Flat Top I, II, and III
Compared With Prediction . 5-11
5-20 Measured Positive Overpressure Impulse for Flat
Top I, II and III Compared With Prediction . 5-11
5-21 Measured Dynamic Pressure for Flat Top I, II, and
III Compared With Prediction . 5-11
5-22 Paths of Triple Point . 5-12
5-23 Typical Time Histories in Mach Reflection Region .. 5-12
5-24 Triple Point Loci Over Reflecting Surfaces of Hard-
packed Dirt and Dry Sand . 5-13
5-25 Typical Complex Shock Waves Observed in
Reflection Studies . 5-13
5-26 Normally Reflected Peak Overpressure vs Scaled
Distance . 5-14
5-27 Scaled Normally Reflected Positive Impulse vs Scaled
Distance . 5-14
5-28 Geometrically Scaled Reflected Impulse vs Scaled
Distance at Different Atmospheric Pressures ..... 5-15
5-29 Normally Reflected Positive Impulse as a Function
of Scaled Distance (R!W 1/3) and Ambient Pressure
Po . 5-16
5-30 Normally Reflected Pressure-Time History, Scaled
Distance = 0.10 ft!lbM3, 0.1 mm Hg (approx.
210,000-ft altitude) . 5-17
5-31 Phenomenon of Blast Wave Coalescence for Two
Charges Detonated With Time Delay . 5-17
5-32 Scaled Delays Between Shock Fronts from Sequential
Explosions . 5-18
6-1 Compiled Shock-front Parameters for Incident Air
Blast Waves . 6-9
6-2 Compiled Shock-front Parameters for Normally
Reflected Air Blast Waves . 6-11
6-3 Compiled Impulses and Durations . 6-13
6-4 Geometry for Regular Reflection . 6-14
6-5 Reflected Overpressure Ratio as a Function of Angle
of Incidence for Various Side-on Overpressures ... 6-14
6-6 Typical Reflected Overpressure vs Horizontal Dis
tance for Selected Heights of Burst, I Ibm
Pentolite at Sea Level . 6-17
6-7 Typical Dynamic Pressure vs Distance for Selected
Heights of Burst, I Ibm Pentolite at Sea Level. .... 6-17
7-1 Schematic of BRL Piezoelectric Side-on Blast
Gage . 7-3
xi
AMCP 706-181
7-2
SwRI Side-on Blast Gage . 7-3
7-3
Atlantic Research Corp. Pencil Blast Gage, Type
LC-13 .
7-4
7-4
The British H3 Side-on Gage .
7-5
7-5
The British H3B Blast Gage .
7-5
7-6
The British H3C Blast Gage .
7-6
7-7
Side-on Blast Gage Using Small, Flush-diaphragm
Transducers . 7-6
7-8
Reflected Pressure Gage of Granath and Coulter . 7-7
7-9
Exploded View of Half-inch Gage of Granath and
Coulter . 7-8
7-10
Sectional View of Gage of Baker and Ewing . 7-9
7-11
Sectional View of NASA Miniature Transducer of
Morton and Patterson . 7-10
7-12
Atlantic Research Corp. Miniature Pressure
Transducers . 7-11
7-13
Kistler Model 603A Quartz Miniature Pressure
Transducer . 7-12
7-14
Internal Schematic of Kistler Model 603A Pressure
Transducer Showing Scheme for Acceleration
Compensation . 7-12
7-15
Basic Single Coil Variable Impedance Pressure
Transducer, Kaman Nuclear . 7-12
7-16
Shaevitz-Bytrex Miniature Pressure Transducers . 7-13
7-17
Dynisco Pressure Transducer . 7-13
7-18
British AWRE MQ20 Pressure Transducer . 7-16
7-19
Early Type of Blast Switch . 7-16
7-20
BRL Arrival-time Gage of Watson and Wilson . 7-16
7-21
Cross Sections of Typical BRL Total Head Gages .. 7-17
7-22
Comparison of Pressure-Time and Drag-Time Traces
for 0 deg and 90 deg to Flow Direction, Gage
of Johnson and Ewing . 7-18
7-23
SRI Total Drag Probe Assembly, Schematic Cross
Section .
7-19
7-24
Assembly Drawing of BRL Biaxial Drag Gage .
7-19
7-25
Diagram of Density Gage of Dewey and Anson .
7-19
7-26
Record from Density Gage of Dewey and Anson .
7-20
7-27
Permanent Tip Deflection of 0.05 I-in. 6061
McMurtry .
7-24
7-29
Squirt Blast Pressure Gage of Palmer and
Muirhead . 7-25
8-1
Block Diagram of CRT Oscilloscope Recording
System . 8-1
8-2
BRL Four-channel Recording Equipment . 8-2
xii
AMCP 706-181
System . 8-7
System . 8-8
Instrumentation . 8-8
Recording . 8-9
Current . 8-12
Camera . 9-5
xiii
AMCP7~181
xiv
AMCP7~181
LIST OF TABLES
2-1
Blast Wave Energy Parameter B for Some Values
of Ratios of the Specific Heats . 2-16
3-1
List of Physical Parameters for Hopkinson
Blast Scaling . 3-8
3-2
Sach's Scaling Parameters . 3-11
3-3
Blast Scaling Laws Proposed by Wecken . 3-19
3-4
Additional Parameters in Wecken's Analysis 3-19
3-5
Dimensionless Products Corresponding to Wecken's
Scaling . 3-20
3-6
Primary Buckingham 11' Terms. Blast Loading and
Response of High-speed Structure . 3-23
4-1
Coefficients of Partial Derivatives in
Kirkwood-Brinkley Method . 4-5
4-2
Comparison of Detonation Velocities D Calculated
for LSZK Substance With Detonation Velocities
Determined at Bruceton . 4-16
4-3
Input Data for Flow Past a Wedge-PAF Method . 4-22
4-4
A Comparison of Methods for Calculating Time
Dependent Fluid Dynamics . 4-34
6-1
Explosive Properties . 6-4
6-2
Sachs' Scaled Nondimensional Blast
Parameters . 6-5
6-3
Scaled Shock-front Parameters for Incident
Blast Waves . 6-8
6-4
Scaled Shock-front Parameters for Reflected
Blast Waves . 6-10
6-5
Scaled Impulses and Durations of Overpressure 6-12
6-6
Time Constant and Initial Decay Rate of P's 6-14
6-7
Typical Compiled Data for Strong,
Obliquely Reflected Shocks . 6-15
6-8
Limit of Regular Reflection ex extreme vs Shock
Strength . 6-16
6-9
Conversion Factors for Scaled Blast Wave
Properties . 6-18
7-1
Summary of Properties of AWRE Foulness Pattern
Standard Piezoelectric Transducers . 7-15
7-2
Characteristics of Side-on Pressure Transducers . 7-26
7-3
Characteristics of Flush-mounted Pressure
Transducers . 7-27
xv
AMCP 706-181
xvi
AMCP 706-181
PREFACE
The Engineering Design Handbooks fall into two basic categories, those
approved for release and sale, and those classified for security reasons. The
Army Materiel Command policy is to release these Engineering Design
Handbooks to other DOD activities and their contractors and other
Government agencies in accordance with current Army Regulation 70-31,
dated 9 September 1966. It will be noted that the majority of these
Handbooks can be obtained from the National Technical Information
Service (NTIS). Procedures for acquiring these Handbooks follow:
XVll
]
AMCP 706-181
Commander
Letterkenny Army Depot
ATTN: AMXLE-ATD
Chambersburg, PA 17201
Commander
US Army Materiel Command
ATTN: AMCRO-TV
5001 Eisenhower Avenue
Alexandria, VA 22333
xviii
AMCP 706-181
CHAPTER 1
GENERAL PHENOMENOLOGY
s = slipstream locus
A = path of triple point
Sf = reflecting surface locus
a,b,c.f,g,h = constants
T = triple point
c, C' = charge center; image center
= positive phase duration, nega
E = total explosive energy tive phase duration
I-I
AMCP 706-181
1-2
AMCP 706-181
where
fa + T+
J
fa
[p(t) - Po ] dt (1-1) ta < t ~ ta +T+
*Where the symbols designating peak pressures, durations, *Superscript numbers refer to References at the end of each
and impulses appear without superscript plus or minus signs chapter.
later in this handbook, the plus sign indicating positive **In the foUowing equations f can be set equal to zero or
phase will be implied. any other convenient number~
1-3
AMCP 706-181
will accurately fit many gage records over to match time histories of positive phase
most of the positive phase. With this form one overpressure which he predicted from theoret
also can match the amplitude and the P; ical calculations of blast waves generated from
initial decay rate or the amplitude and the a point source. The most complex formula to
positive impulse* with experimental results. date which has been proposed for fitting
Eq. 1-4 is undoubtedly a better representation positive phase time history data is also due to
than the purely linear decay predicted by Eq. Brode4 . This equation, involving five param
1-3. eters, is
1-4
AMCP 706-181
acteristics. The only proposed functional shocks of small amplitude occurring at various
form for this phase which the author could times after t a. These are caused by the
locate is one due to Brode 3 , given by the successive implosion toward the center of
equation rarefaction waves from the contact surface
p(t) = Po - P; [(tIT-) (1-9)
between explosion products and air.* Sec
ondary and tertiary shocks of this nature,
(I - tiT -) e -4 t/TJ sometimes facetiously called "pete" and
"repete", have indeed been observed, as can
where be seen in Fig. 1-2. These later waves have
little effect on any of the characteristics of
t a + T+ < t < t a + T+ + T the positive phase of the blast wave with the
exception of positive duration T+. This param
This form is based on Brode's point-source eter can be changed quite markedly if a
theoretical solution. secondary shock happens to arrive just prior
to the initial decay reaching Po On the other
1-3.2.2 PARTICLE VELOCITY AND hand, secondary and repeated shocks can
OTHER PARAMETERS markedly affect the negative phase, causing it
to be abruptly tenninated, or markedly reduc
The blast front in its passage through the ing the negative impulse I; or amplitude Ps-'
air not only increases the pressure, but also The only reasonably complete discussion of
increases density p and temperature 0, and secondary shocks appears to be that of
accelerate&' .the air particles to produce a Rudlin 6 who points out differences in scaled
particle vielQCity, u in;ttte' direction of travel. If arrival times ani(} overpressures' for secondary
we were to plot time histories of these shocks with type of explosive source and
physical quantities, they would be similar to presence or absence of a ground reflecting
Fig. I-I with the exception that the durations plane.
would not necessarily be the same as for
pressure-time history. 1-4 "NONIDEAL" BLAST WAVES
For any finite explosion source our ideal *These later shocks for explosions in free air should not be
confused with reflected shocks occurring when reflecting
blast wave also can exhibit numerous repeated boundaries are present.
1-5
AMCP 706181
1-6
AMCP 706-181
;~ ..J _ _ ~-----
R = 30 ft
_"._,._J'''v,,~~~ ~~ -_
R = 38 ft
~
FRAGMENT \ R = 52 ft ~REFLECT ION FR OM
SHOCK ~ ""'-~ PRIMARY TARGET
PATIERNS ~ _ N
R = 80 ft
of measured blast wave properties agree quite shock front are, as indicated, those for the
well with those of ideal waves. In other free-air shock wave discussed previously in
words, the characteristics of the ideal waves this chapter. On the right, the front R is
discussed earlier are also the characteristics of shown immediately after reflection from the
stable blast waves. wall. It is moving away from the wall with a
velocity Ur into the flow field and com
1-5 REFLECTION AND DIFFRACTION OF pressed region associated with the incident
BLAST WAVES wave. In the reflection process, the incident
particle velocity Us is arrested (u = 0), and the
So far we have considered the properties of pressure, density, and temperature of the
air blast waves as they propagate freely reflected wave are all increased above the
through the air. On encountering any solid or values in the incident wave. The overpressure
dense object, these waves are seriously modi at the wall surface usually is termed the
fied, as they reflect from this object and "reflected overpressure", and is designated
diffract around it. Let us now discuss these p,.. * For very weak shocks, Ps <<Po, acoustic
two phenomena. approximations are valid, and the reflected
overpressure is twice the incident overpres
1-5.1 REFLECTION OF A PLANE WAVE sure, Pr = 2Ps' For stronger incident shocks,
the enhancement of reflected pressure is
1-5.1.1 TYPES OF REFLECTION increased. An upper limit often cited in the
literature lO is Pr = 8Ps' This limit constitutes
1-5.1.1.1 NORMAL REFLECTION a popular misconception and is probably
considerably in error, since it is based on the
The simplest case of reflection is that of assumption that the air behaves as a perfect
normal reflection of a plane shock wave from gas even at the high pressures and tempera
a plane, rigid wall. This phenomenon is shown tures extant under strong shock conditions.
schematically in Fig. 1-5. On the left, the Doering and Burkhardt 1 1 and Shear and
incident wave I is shown just prior to impinge McCane l2 have shown that this ratio can be
ment on the wall. It is moving at velocity U much greater (perhaps 20 or more) if real gas
into still air whose ambient conditions are
designated by the symbols with subscript *Superscript plus signs for positive phase are implied in this
zero. The conditions immediately behind the discussion.
1-7
AMCP 706-181
R " 670 ft
200
20
R = 795 ft
10
0
0 100 200
1:'
8
R = 940 ft
V'l
Cl..
0
u.J
~
0 100 200
=::l
~ I?
u.J
~
a.. R = 1050 ft
.~
~
u.J
:>
n 0
fJ 100
200
12
~J R ,. 1200 ft
0
0 100 200 300
12
6 R = 1350 ft
0
0 100 200 300
TIME, msec
Figure 1-4. Typical Nonideal Pressure Traces Showing Precursors
I Po + ps :r 9s Us
the air molecules are accounted for.
J j ) j 9; /; ; ; ; 7 ;;;;;;;;;);j
Figure 1-5. Normal Reflection of a Plane The next case, in order of complexity to
Shock from a Rigid Wall the "nonnal" case of reflection is that of
1-8
AMCP7~181
so-called regular oblique reflection of a plane the reflected shock is greater than it is for
shock wave from a rigid, plane wall. This head-on reflection. This is given by cx' = (l/2)
phenomenon is illustrated in Fig. 1-6. The COS -I ('Y - 1)'2. For air (approximated as an
incident shock travels into still air (Region ideal gas with 'Y = 1.40), cx' = 39.23. However,
Q)) at velocity U, with the incident shock it is only for weak or moderate shocks, P2 /PI
front making an angle of incidence cxl with < 7.02 in ideal air, that this can occur before
respect to the wall. The properties behind this regular reflection is forbidden.
front (Region @) are those for a free-air
shock. On making contact with the wall, the 3. For a given strength of incident shock,
flow behind the incident shock is turned, there is some value of the angle of incidence
because the component normal to the wall such that for cxl = cx min the strength of the
must be zero, and the shock is reflected from reflected shock, Pr/Po' is a minimum.
the wall at a reflection angle OR not neces
sarily equal to cxl" The symbols in Region 4. The angle of reflection cx R is an increas
indicate the reflected shock properties which ing monotonic function of the angle of
are the conditions for that region. A pressure incidence cxl"
transducer flush-mounted in the wall would
record only the ambient and reflected wave These properties of oblique shocks - refer
pressures (direct jump from conditions of ring respectively to items 1 through 4 for
Region CD
to those of Region ) as the reflected shocks - differ quite markedly from
wave pattern traveled along the wall, while corresponding properties of acoustic waves,
one mounted at a short distance from the wall which are:
would record the ambient pressure, then the
incident wave pressure, and finally the re 1. Regular reflection occurs forO <cxl ~90
flected wave pressure. Some interesting prop deg
erties of this regularly reflected shock are:] 3
2. P,. = 2~ for all values of cx
1. For a given strength of incident
shock, there is some critical angle of incidence 3. P,. = 2Ps for all values of cx
cx l erit such that the type of reflection de
scribed previously cannot occur for '7 > 4. '7 =cxR for all values of cxR
cx'm" In the limit of vanishing shock
strength, cx l ait = 90 deg; and in the limit of 1-5.1.1.3 MACH REFLECTION
infinite shock strength, CXlcrit = Sin-] l/'y =
39.97 deg for air with 'Y = 1.4 (see Fig. 54, The next type of reflection, in order of
Ref. 35).
complexity, is Mach reflection of a plane
shock wave that is obliquely incident on a
2. For each gaseous medium, there is some
angle cx' such that for cx l > cx' the strength of plane, rigid wall. As noted in the preceding
discussion of regular oblique reflection, there
is some critical angle of incidence
dependent on shock strength - greater than
R at which regular reflection cannot occur.
Ernst Mach showed] 3, in 1877, that the
Q)
CD incident and reflected shocks would coalesce
Po + Pr to form a third shock. Because of the ge
Pro Sr. ur
ometry of the shock fronts, they are termed
( U 5 in aR 0;.-1/7'....,....."...,...,-h-.,....;.l4-,.....,.L,.:.,... ......._
1-9
AMCP 706-181
1-10
AMCP 706-181
Sf
In Fig. 1-9 the geometry of the Mach Fig. 1-10 shows two successive stages of
reflection process can be seen in more detail. this reflection process. In the first stage the
By comparison with Fig. 1-7, one can see that incident wave 11 is just tangent to the
incident and reflected shocks are both curved, surface. The excess pressure over that of the
and that the path of the triple point is no atmosphere at the reflecting surface is just
longer a straight line. Although the Mach stern double (for very weak shock waves) that of
is shown as a vertical straight line in Fig. 1-9, the incident wave where it is not in contact
this is not always the case in reality. with the surface. At a later stage, the incident
wave is represented at 12 and the reflected
1-5.1.2.2 WEAK SHOCK WAVES wave at R 2 , which is imagined to arise from
the image source ct. Again, the pressure at the
Very weak shock waves, Le., those of line of contact of 12 R2 , and the surface Sf is
nearly acoustic strength, are reflected from just double that at 12 where it is not in
plane surfaces in such a way that a geomet contact with the surface. The angles at which
rical construction of the wave system can be the shocks 12 and R 2 meet the surface Sf are
made very simply. Consider a point source of equal, and no Mach stern is formed. For most
the shock C (Fig. 1-10) and, some distance practical cases of interest in air blast tech
from it, a plane reflecting surface Sf. The nology, shocks are too strong for this acoustic
incident wave I, striking the surface, will be approximation to be applicable, and this
reflected from it in such a way that the simplified geometry cannot be used.
reflected wave R may be considered to arise
from a second image source C' which is on the
opposite side of the reflecting surface, on a 1-5.2 DIFFRACTION OF A PLANE WAVE
line perpendicular to Sf through the true
source, and at a distance from Sf equal to the When a blast wave encounters a solid object
distance of C from the surface. of finite extent, very complicated processes
I-II
AMCP7~181
/
C /
I
/
R , /
/ ~
Y
he /'0 I erit /" P
/" M
d
Y = height of Mach stem do = horizontal distance for start of formation of
he = height of charge above reflecting surface Mach stem
d = any horizontal distance> do
Figure 1-9. Geometry of Mach Reflection
/ \
I
I
I
"\ , \ \
\
(in fact, the discussion that follows is based
on the results of such experiments). The
I (lC reader is referred to Refs. 15 through 23 for
I I comprehensive studies of the diffraction pro
\ I I
\
\
\.
\
"- ..... __ / /
/
/
/
cess.
"- ,/
" """---- ./
The diffraction process can perhaps best be
Figure 110. Reflection of Weak Shock Waves illustrated by describing the sequence of
1-12
AMCP 706-181
events occurring when a plane blast wave this value in the vicinity of the vortex. Ahead
traveling over a rigid plane encounters a rigid, of the incident shock, the pressure on the top
thick wall protruding from the plane, as face and on the rear face of the wall is still Po.
illustrated in Fig. 1-ll(A). The geometry is
shown in the first sketch of this figure, with As the incident shock front passes beyond
the blast front being normally incident on the the rear face of the wall, it diffracts around
front face of the wall, and the pressure on all this face, as shown in Fig. I-II (C). A second
faces on the wall being at ambient pressure vortex is formed at the upper righthand
Po' As the incident wave I first encounters the comer of the wall. At the instant shown in
wall, reflection of the portion of the wave Fig. I-II (C), the reflected wave from the
striking the front face of the wall occurs; the front face of the wall has been completely
reflected wave R moves to the left, and the attenuated by the rarefaction wave, and the
pressure on the front face jumps to Po + p,.. pressure on the front face is Po + q, where q is
Above the wall, the incident waves continue dynamic pressure. On the top face the pres
on relatively undisturbed. sure is still nearly equal to Po + Ps . Behind the
diffracted incident wave on the rear face,
As the reflected wave moves to the left pressure is somewhat less than Po + ~. Ahead
away from the front face of the wall, a of the front I, the pressure is Po. The
rarefaction front moves down the front face, maximum back wall pressure develops slowly
as shown in Fig. I-II (B). A vortex is shed as a result of vortex phenomena and the time
from the upper lefthand comer at the wall. A required for the back wall to be enveloped by
vortex is a region of air spinning about an axis the blast. In the final stage, the incident wave
at a high speed. Low overpressures exist at its has passed beyond the wall, the diffraction
center because of the Venturi effect. At the process is over, and the wall is immersed in
instant depicted in Fig. I-II (B), the lower the flow field behind the front. For a long
portion of the fron t face still feels the duration blast wave, pressures are nearly
reflected pressure Po + P" while the upper those which would be measured in steady
portion feels a lower pressure quite near the state wind tunnel experiments.
pressure Po + Ps in the incident wave. The
portion of the top face behind the incident 1-5.2.2 THREE-DIMENSIONAL BLOCK
shock I is subjected to pressure Po + Ps ' with
the pressure perhaps somewhat reduced below For a three-dimensional block structure,
the phenomena described in par. 1-5.2.1 also
occur along the sides of the block, so that the
VORTEX preceding discussion also applies to diffrac
RAREFACTION tion about the sides of such structures. This
WAVE
process is illustrated in Fig. 1-12. Fig. 1-13
DIFFRACTED gives recorded pressure-time histories for the
SHOCK
FRONT (8) front, top, and back faces of a model three
dimensional block structure, as recorded in
U DIFFRACTED
U shock tube at BRLJ J. The pressure-time
SHOCK history for the front face of the block shows
FRONT
reflected pressure (initially P,.) and the effect
of the rarefaction wave produced at the front
IC) face which causes rapid reduction in reflected
Figure 1-11. Diffraction of a Shock Front Over pressure. The pressure recorded on the top
a Wall 1 6 face of the block shows an initial peak of
side-on pressure Ps and a less rapid pressure
(Reprinted by permission of C. H. Norris, R. J. Han
sen, M. J. Holley, Jr., J. M. Biggs, S. Namyet, and decay. The pressure recorded at the back face
J. K. Minami, Structural Design for Dynamic Loads, of the block shows a slow rise time of
McGraw-Hili Book Co., N. Y., 1959.) pressure with no real "shocking-up". Detailed
1-13
AMCP 706-181
20
BACK FACE
FRONT FACE
VI
-.
c.
C.
0
15
I
C.
10
/"
DIRECTION
.....
a<:
~
Vl
Vl
OF SHOCK FRONT FACE u.J
c::: 5
TRAVEL a..
c:::
.....
>
0
00 1 2
TIME t. mSllC
10 8
VI
TOP FACE
--
c.
c.
0
6
I
C.
5 u.J 4
g;
V'l
BACK FACE
~ 2
c:::
o L-- ----L ....J-____ a..
c::: 0
o 1 2 ~ 0 1 2
TI ME t, mS8C TIME t. msec
Figure 1-13. Pressures on a Three-dimensional Block Structure During Diffraction 33
1-14
AMCP 7Q6.181
I
CA)
FORMATI ON OF
VORTICES VORTEX GROWTH
Figure 1-14. Tracings of Shadowgraphs Showing the Interaction of a Shock Front With a Cylinder
extends upstream, slanting down to meet the stems, M and M', have moved some way
surface. In the present case, however, the downstream of the cylinder. A Mach stem M "
increased flow near the cylinder surface has joins the free air parts of M and M' with the
caused the foot of the slipstream to move diffracted parts PD and P ~ which terminate
nearer to the foot of M. The slipstream, on the cylinder surface. The growth of the
therefore, presents a curved appearance. In vortices is apparent in this figure. In Figs.
Fig. 1-14(C), the feet of the Mach stems have 1-15(A) and (B), the foot of PD has moved
reflected from each other and are now moving further round the cylinder upstream. Notice
on a second circuit of the cylinder. The that the point of flow separation has followed
slipstreams have been swept nearer the rear of this shock. In Fig. 1-15(C), the vortices, V I
the cylinder and now intersect with the and V2 , are breaking away from the cylinder;
diffracted parts of the Mach stems X and Y. while in Fig. 1-15(D), the vortices are being
The commencement of two vortices is indi swept downstream, and the point of flow
cated at V I and V2. These are probably separation has moved toward the rear of the
induced by the back pressure behind the cylinder again.
shocks X and Y interacting with the boundary
layer flow at the surface of the cylinder. The The phenomenology described indicates
shaded portion is due to a localized region of the complexity of the diffraction process for
supersonic flow. In Fig. 1-14(D), the Mach even the very simple case of passage of a plane
1-15
AMCP 706-181
FLOW
Muzzle blast waves from guns are axisym source than do waves from line sources.
metric but not spherically symmetric. They Lindberg and Firth 24 have compared theoret
usually consist of a single shock ,front (see ically predicted variations in peak overpres
Fig. 1-16), but one which has highly direc sure with distance for spherical, infinite cylin
tional properties near the muzzle. The general drical (line), and infinite plane blast sources
characteristics at any point in the blast field (see Fig. 1-17). The scaled distance parameter
are nearly similar to those of spherical R/r0 in their plot is based on characteristic
sources, but the difference is that the muzzle dimension ro which is defined as
blast field characteristics are a function of
two spatial cylindrical coordinates, (r, z) ro =E/ [Po L(3-V)] l/V (l-ll)
rather than one spherical coordinate R. Diver
gence is more nearly spherical than for line where
charges, but it is definitely a function of the
two cylindrical coordinates rand z. v = 1,2,3, respectively, for plane, cylindri
cal, and spherical blasts
1-6.1.1.3 LARGE PLANE CHARGE E = total explosive energy.
1-17
AMCP 706-181
...... -...::
'" ....... ,
...... ........ .......
.......
........ "~I'-. t'-..."
"'"' t": ~~
..........
t'-..." ~
~ I'..:
~ r-.....
~
""" ~ t': ~ /00
t": JJ
...... ......
....... "
r-.............
,, .......
"
.......
~
....... , ............
............
Id -5
10
1-19
AMCP 706-181
3000
2000
I.LJ
-
0
:::>
I
~
<
1()()()
be made based on empirical observations of through the blast source location. The equa
pressures and damage to light structures or tions must then be solved for each separate
glass. Perkins, et al. 2S have done this, and azimuth and for a number of sound rays. All
they have correlated blast strengths with of the calculations must be performed quite
various velocity gradients. They report their rapidly so that the meteorological conditions
estimates in terms of factors which multiply do not change appreciably in the meantime. If
the blast overpressures predicted by the much firing is anticipated at a test site, it may
homogeneous atmospheric case (Fig. 1-20). prove advisible to automate as much of this
The multiplication factors for various types of procedure as possible. Perkins, et aI. 2 S report
gradients are illustrated in that figure. the use of an analog computer for solution of
the equations involved, and the acquisition of
1-6.2.2.2 PRACTICE a library of velocity gradients and correspond
ing focus predictions at Aberdeen Proving
In practice, the application of Berning's Ground over a period of several years. They
relatively simple theory for the prediction of found that prediction of focus conditions
focusing can be quite tedious and time con often could be made with sufficient accuracy
suming. One must obtain data on the tem by comparing the existing current velocity
perature and wind velocity and direction, as a gradients with those in the library. Using this
function of altitude up to at least 10,000 ft. procedure, and allowing testing to occur only
These data must be translated into sound on days when no focus was predicted in
velocity gradients along a number of azimuths inhabited areas, they were able to reduce
1-20
AMCP 706-181
I
1 SINGLE NEGATIVE GRADIENT
~~VELOCITY
0
4
SURFACE WITH POS ITIVE
GRADIENT ABOVE
I
I
SURFACE WITH STRONG
POS ITIVE GRAD lENT ABOVE lL 35
markedly the incidence of complaints and typical vertical sound velocity profIles and
claims of damage from these areas as a result resulting distortions and focusing of ray
of the firing of large weapons and detonation paths2 9.
of large explosive charges.
1-6.3 VARIATION OF TYPES OF ENERGY
Procedures similar to those developed by SOURCE
Berning2 6 and Perkins, et al. 2 5 have been
employed at a number of other test sites to Most of the available experimental data and
control the incidence of focusing effects from analyses of blast waves in air are limited to a
large blast sources. An excellent general dis few types of chemical explosives (usually
cussion of the overall problem of effects of either TNT or Pentolite) and to nuclear
long-range blast focusing and of the difficul explosives. What are the effects for other
ties in obtaining accurate estimates of these explosives or other types of blast sources?
effects is given by Reed 2 9, who was con
cerned with side effects of cratering with 1-6.3.1 CHEMICAL EXPLOSIVES
nuclear explosives. Reed also includes an
extensive bibliography on the subject in his During World War II, the British and
report. He uses the adjective "caustic" to Americans conducted many tests with differ
indicate exceptional disturbances at long ent types of chemical explosives in attempts
ranges. Figs. 1-21 and 1-22 are illustrations of to optimize blast damaging effects. They
1-21
AMCP 706-181
WAVE
proximately equate blast waves from different
FRONT
POSITIONS explosives by using a conversion to an "equiv
alent weight" or "equivalent energy" of some
standard explosive which is usually TNT. This
conversion is a multiplier which would cause
the blast data for the other explosive to
coalesce with TNT data, as nearly as possible.
This coefficient for Pentolite is usually as
sumed to be about 1.1 X TNT. Note that it is
not equivalent to the coefficients given in
Table 1-1, or their inverses. While not exact,
this method offers a way of estimating blast
for explosives where limited data exist. Some
times the conversion coefficient can be esti
mated from ratios of the heats of detonation
for a new explosive to that of TNT, with
these heats of detonation being measured in a
bomb calorimeter. Note from Table 1-1, that,
although blast parameters are measurably
different for different chemical explosives,
800 1000 1200 DISTANCE
VELOCITY, ft/see
the entire range of differences is not great. A
maximum coefficient of 1.5 covers the entire
Figure 1-21. Shock Wave Distortion by
range of peak pressures and impulses, for all
Layered Atmospheric Temperature
the explosives compared on an equal volume
and Wind Structure 29
basis. One, therefore, should view with cau
tion claims for vastly superior chemical explo
found that a number of explosives generated sives for generating air blast.
blast waves of measurably different peak
pressure and impulse from those of a "stan
dard" explosive of their choosing. In par '-6.3.2 NUCLEAR EXPLOSIVES
ticular, those explosives which contained sig
nificant quantities of aluminum powder exhib Nuclear explosives differ from chemical
ited those differences.* Kennedy 13 has sum explosives in their ability to produce air blast
marized much of this early work, and he has waves primarily because of the sheer magni
noted that the general characteristics of blast tude of their total energy release, and because
waves from different chemical explosives are of the difference in energy density between
quite similar-their relative peak overpressures the two types of explosive sources. Close to a
and impulses being essentially independent of nuclear weapon burst, the blast overpressure
distance from the charge. Results, of measure will be greater than that from a scaled
ments over a number of distances, therefore, chemical explosion of the same effective blast
could be averaged and quoted as ratios yield. At greater distances, the blast waves
applicable over all distances. Such ratios for a will be quite similar30 In comparing blasts
number of explosives, relative to Composition from TNT and nuclear explosives at sea level,
B are shown in Table 1-1 13 . Comparisons for the total yield from the nuclear explosive is
o~her explosives are given in AMCP often related to the detonation energy of
706-1 82(S), Explosions in Air, Part Two (U).
TNT by an effectiveness factor of 0.5 i.e.,
Y total (nuc) = 0 . 5 Y total (TNT). In the calcu
Because the ratios for peak pressure and
impulse do not differ greatly, one can ap lations of properties of blast waves from
nuclear explosions, the nuclear explosive
*The reaction of aluminum with the "oxidizer" present in
chemical explosives is more energetic than the decom
sources often are assumed to be point sources
position of the chemical explosive by itself. because of their high energy density.
1-22
AMCP 706-181
SILENT ZONE
SOUND
VELOCITY RANGE
1-23
AMCP 706-181
~\, \
10' helium at a pressure of 3500 psia. The NASA
~\. Manned Spacecraft Center was concerned
\ \
about the possible blast effects of rupture of
\" ~ 52 \
\ '"'"-'"' \\\\~\\
this vessel during vibration testing in their
laboratory. What is the TNT equivalent of
\ \5, . \ burst of this vessel?
\ 10 4 \
\.
\10 2
\\
,\
,.~\
.~\\ Input values for Eq. 1-1 2 are
10 50
K( );TNT(----)
v = 7 X 104 in. 3
Figure 1-23. Variation of Peak Overpressure 'Y = 5/3 (Helium is a monatomic gas.)
Ratios P With Shock Radius A-
for Various Explosions3 s
From Eq. 1-12,
where
p = initial absolute pressure E= 14.7 X 70,000 J
f(3500)_(3500)3/S . . .
v = total volume (3 - ~ 14.7 14.7
5
1)
REFERENCES
1-24
AMCP 706-181
4. H. L. Brode, The B last Wave in Air Vol. I: Explosions and Explosives in Air,
Resulting from a High 1emperature, High Chapter 2, Part II, p. 84, 1946.
Pressure Sphere of Air, RM-1825-AEC,
The Rand Corp., Santa Monica, Cali 14. E. Mach and J. Sommer, "Uber die
fornia, December 1956. Fortpflanzungsgeschwindigkeit von Ex
p1osionsschallwellen", A kademie der Wis
5. John M. Dewey, "The Air Velocity in senschafter, Sitzangberichte der Wiener,
Blast Waves from TNT Explosions", Proc. Vol. 74,1877.
of the Royal Soc. A, 279, 366-385
(1964). 15. V. J. Bishop, R. D. Rowe, The Inter
action of a Long Duration Friedlander
6. L. Rudlin, Private communication regard Shaped Blast Wave with an Infinitely
ing second and secondary airshocks from Long Right Circular Cylinder. Incident
HE explosions. Blast Wave 20.7 psi, Positive Duration 50
ms and a 16 em Diameter Cylinder,
7. R. J. Larson and W. Olson, Measurements AWRE Report No. 0-38/67, Atomic
of Air Blast Effects from Simulated Weapons Research Establishment, A1der
Nuclear Reactor Core Excursions, BRL waston, Berkshire, England, April 1967.
Memorandum Report No. 1102, Aber
deen Proving Ground, Md., September 16. C. H. Norris, R. J. Hansen, M. J. Holley,
1957. Jr., J. M. Briggs, S. Namyet, and J. K.
Minami, Structural Design for Dynamic
8. E. J. Bryant, N. H. Ethridge, and J. H. Loads, McGraw-Hill Book Co., New
Keefer, Operation TEAPOT, Project York, 1959, pp. 249-252.
1.14b, Measurements of Air-Blast Phe
nomena with Self-Recording Gages, 17. W. Bleakney, Shock Loading of Rec
WT-1155,Ju1y 1965 (AD-617170). tangular Structures, Tech. Rept. 11-11,
Princeton Univ. Dept. of Physics, Prince
9. John M. Dewey, "Precursor Shocks Pro ton, N. J., January 10, 1952.
duced by a Large Yield Chemical Exp1o
sion", Nature, 205, p. 1306. 18. W. Bleakney, Rectangular Block, Dif
fraction of a Shock Wave Around an
10. S. G1asstone (Ed.), The Effects of Nu Obstacle, Princeton Univ. Dept. of
clear Weapons, U. S. Government Print Physics, Princeton, N. J., December 17,
ing Office, Washington, D. C., p. 124, 1949.
1962.
19. W. Bleakney, The Diffraction of Shock
11. W. Doering and G. Burkhardt, Contribu Waves Around Obstacles and the Tran
tions to the Theory of Detonation, TR sient Loading of Structures, Tech. Rept.
No. F-TS-1227-1A, Wright-Patterson 11-3, Princeton Univ. Dept. of Physics,
AFB, Ohio, p. 306, May 1949. Princeton, N. J., March 16, 1950.
1-25
AMCP 706-181
tion of Shock Waves Around Cylindrical 28. E. A. Milne, Phil. Mag. and Jour. of
and Rectangular Obstacles, University of Science, 42, 96-114 (1921).
Michigan, Ann Arbor, Mich., June 21,
1950. 29. J. W. Reed, Acoustic Wave Effects
Project: Airblast Prediction Techniques,
22. G. Uhlenbeck, Diffraction of Shock SC-M-69-332, Sandia Laboratories,
Waves Around Various Obstacles. Univer Albuquerque, N. Mex., May 1969.
sity of Michigan, Engineering Research
Institute, Ann Arbor, Mich., March 21, 30. C. N. Kingery, Parametric Analysis of
1950. Sub-Kiloton Nuclear and High Explosive
Blast, BRL Report No. 1393, February
23. W. Bleakney, D. R. White and W. C. 1968 (AD-833 698).
Griffith, "Measurements of Diffraction of
Shock Waves and Resulting Loading of 31. R. J. Larson and W. Olson, Measurements
Structures", Jour. of Appl. Mech. 17, I, of Air Blast Effects from Simulated
439-445 (March 1950). Nuclear Reactor Core Excursions, BRL
Memorandum Report No. 1102, Sep
tember 1957.
24. H. E. Lindberg and R. D. Firth, Struc
tural Response of SPINE Vehicles, Vol.
32. W. E. Baker, S. Silverman, and T. D.
II, Simulation of Transient Surface Loads
Dunham, Study of Explosions in the
by Explosive Blast Waves, Tech. Rept.
NASA-MSC Vibration and Acoustic Test
AFWL-TR 66-163, Vol. II Air Force
Facility (VATF), Final Report on Con
Weapons Lab., Kirtland AFB, N. Mex.,
tract NAS 9-7749, March 1968.
May 1967.
33. G. Beichler, Records of Air Shock Load
25. B. Perkins, Jr., P. H. Lorrain and W. H. ing on a Three-Dimensional Model, BRL
Townsend, Forecasting the Focus of Air Mechanical Note No. 842, APG Mary
Blasts due to Meteorological Conditions land, Jan. 1954.
in the Lower Atmosphere, BRL Report
No. 1118, Aberdeen Proving Ground, 34. R. J. Janus and C. N. Kingery, Air Blast
Maryland, October 1960. Loading on a Three-Dimensional Model
of a Gabled Shelter, AFSWP 1054, Jan.
26. W. W. Berning, Investigations of the 1958.
Propagation of Blast Waves Over Rela
tively Large Distances and the Damaging 35. R. Courant and K. O. Friedrichs, Super
Possibilities of such Propagation, BRL sonic Flow And Shock Waves, Inter
Report No. 675, Aberdeen Proving science Publishers, New York, N.Y., 1948.
Ground, Maryland, November 1948.
36. AMCP 706-1 82(S), Engineering Design
27. S. Fujiwhara, Bull. Cent. Meteor. Obs. Handbook, Explosions in Air, Part Two
Japan 2, I, 1912;and 2, 2,1916. (U).
1-26
AMCP 706-181
CHAPTER 2
2-1
AMCP 106-181
= a parameter in Sakurai's ap
proximation In studying air blast theory, one's first step
should be to discard any notion that acoustic
= a function of shock Mach theory is adequate to describe air blast in all
number but a limited class of problems. In fact, one
should include a state of mind in which one
= RJR o = dimensionless shock considers an acoustic wave as a degenerate
radius shock wave, rather than the state of mind
which considers a shock wave as a strong
= density, density at location i acoustic wave. One should become ac
customed to thinking of waves which move
= initial density faster than sound (sometimes much faster);
which have finite (sometimes large) pressure,
Po = ambient density density, and temperature changes and finite
particle velocities associated with them; and
r = time in Lagrangian system which have fronts across which changes in
these various parameters are so rapid that
r* dimensionless time they are usually described by discontinuous
"jump". None of these notions are consistent
x relative angle with acoustic theory.
2-2
AMCP 706-181
2-2 BASIC EQUATIONS and the dot here denotes differentiation with
respect to time in a Lagrangian system.
2-2.1 COORDINATE SYSTEMS Constants of integration can be taken as
parameters aj
In writing fluid dynamic equations in
volving large motions and changes in prop 2-2.2 FORMS OF EQUATIONS
erties, one can employ one of two different
forms of these equations, which are named 2-2.2.1 LAGRANGIAN
for Lagrange and Euler, respectively. The
equations in Lagrangian form are fixed con In Lagrangian form, considerations of con
ceptually to particles in the fluid, and de servation of momentum of moving fluid
scribe motion of particles as functions of time elements lead to the set of equations
and other parameters characterizing each indi
vidual particle. In a Cartesian coordinate
+ (2-3)
system,
xj=a j + 8. (a.,a2,a3,t),i =1,2,3 (2-1)*
where
p = absolute pressure*
Xj = Eulerian coordinates
aj = Lagrangian coordinates of the particle From considerations of conservation of mass,
at some specified initial time one obtains the equation
t = time
where
In the Eulerian form, attention is directed to a(X 1 ,X2,X 3 )
J = (2-5)
points fixed in an inertial frame of reference a(al ,a2 ,a3)
and to what happens at these points in course
of time t. A rigidly mounted side-<>n blast is the Jacobian.
gage would record variation of pressure in an
Eulerian system fixed with respect to the In shock theory, viscosity and heat conduc
ground, for example. Motion in the Eulerian tion effects are usually assumed negligible
system is described by giving velocity com everywhere but in shock fronts. This is
ponents u j as a function of x j and t. Trans expressed by the equation
formation from Eulerian to Lagrangian co
ordinates is effected by solving the system of DS (2-6)
ordinary differential equations - = 0
Dt
t =T p = !(p,S) (2-7)
*Subscript j will indicate three components throughout this
discussion. *Gravity and other body forces are assumed negligible.
2-3
AMCP 706-181
(2-9)
+~ +i
(conservation of energy)
(conservation of mass)
Here, subscripts 1 and 2 denote one side or
the other of the discontinuity.
as
at + .~
3
Uj
(as)
ax. =0 (2-10) An alternative form of Eqs. 2-14 for an
I = 1 I inertial system with the discontinuity moving
with velocity U is given by equations
(adiabatic change of state)
where
PI +PI (U I -U)2 == P2
P = rep, s) (2-11) + P2 (U2 - U)2
(2-15)
is the form of the equation of state.
(U 2 - U)+P2 U2
2-4
AMCP 706181
here: "They are so certainly correct in their distance r from the origin of coordinates. All
application to real gases that the only possible flows are radial, with a single velocity com
explanation of any set of measurements u.
ponent The fundamental equations are
which appears to contradict them is that the
au + u (au)+ (~) = 0
observations are wrong." at ar 1.p\ar (2-19)
The general equations for air blast trans ap + u (ap) + p(au) + 2up = 0 (2-20)
mission which have been previously given are at ar ar r
very difficult to solve, either analytically or
(conservation of mass)
numerically, in arbitrary three-dimensional
cases. Most of the available solutions are
as (as) =0 (2-21)
limited to one of the three "one-dimensional" ifi+uar
cases, i.e., cases in which the shock and flow
fields can be described in terms of a single (adiabatic changes of state)
spatial variable and time. The equations for
these three special cases are presented in a where
nurr_ber of standard works in fluid dynam
ics' ,2, but will be repeated here for complete r = (xi + xi + xn Y2
ness.
In this case again, an equation of state is
2-2.4.1 LINEAR FLOW needed to complete the set of equations.
ap (a p) (au) =0
-+u-+p- (2-17)
au + u(au\, ~(ap)=
at ax ax 0 (2-22)
at ari p ar
(conservation of mass)
(conservation of momentum)
(2-18)
as + ufOS)
,ax =0
ap + u (a p
at <fi ar) + p (au)
ar +rup = o (2-23)
(adiabatic changes of state)
(conservation of mass)
The equation of state, Eq. 2-11, is needed
to complete the set of equations. as + u (as)
at ar = 0 (2-24)
2-5
AMCP 706-181
2-2.4.4 APPLICATION
(2-2Sa)
Of these three cases, the one most applica
ble to blast waves in air is the second , for density
spherically symmetric. This case applies to
either a spherical source far from any reflec (2-25b)
ting surface or to a hemispherical source
located on a perfectly rigid reflector, both of and for radial velocity
which approximate a number of real blast
sources. The first (plane wave) case is quite (2-25c)
useful for prediction of performance of shock
tubes, but shock tubes are not a topic of
discussion in this handbook.
In Eqs. 2-25, R is the radius of the shock
2-3 ANALYTIC SOLUTIONS TO EQUA wave forming the outer edge of the disturb
ance and is a function of time, , is any radial
TIONS
distance from the blast center, and 1'/ = ,IR.
Eqs. 2-1 through 2-15 constitute all of the Also, Po and Po are the pressure and density,
general equations that usually are used in air respectively, of the ambient air. Eq. 2-19 can
blast theory. To solve specific problems, one now be satisfied. This problem is solved most
must add initial and boundary conditions and conveniently by the inverse method of mak
ing appropriate choices for velocity of propa
must choose specific equations of state. Once
equations of state are chosen, Eqs. 2-14 or gation and shock radius. If the velocities U of
propagation and shock radius R are given by
2-15 also yield a number of additional inter
relations between shock front properties. The U = dR = A R - 3/2 (2-26a)
remainder of this chapter will be devoted to dt
specific problems and to those analytical
solutions which are available in the literature. R=(~ At) 2/5 (2-26b)
Some specific or partial analytic solutions then from Eq. 2-19, the conservation of
are avaliable for one-dimensional cases. Sev momentum constraint,
eral of these solutions are used to generate
initial conditions for computer code nu
merical calculations. We will now discuss in
-A (~ cI> +1'/cI>') + cI>cI>'+:: :' = 0
some detail a few of these solutions. (2-27)
2-6
AMCP 706-181
For constant specific heats the ent ropy 6 is By eliminationofl/J' and ljJ' from Eq. 2-38
given by using Eqs. 2-36 and 2-37, one obtains the
equation for calculating!, when!, l/J, 'IJ, and 1/
'YR
S = - g Qn
- - R
('Y- I )
(T)
To g
~~ I ~
where
t n p -') - n (P P0 O -') ] (2-30)
(2-39)
UI 2[(PI/Po) - I]
l/J = cf>/A :. cf> = A l/J (2-34)
= --'---....:....-_---- (2-42)
U 'Y-I + ('Y+l)pdpo
3! + 1/t' + ~ ! (-1/+l/J)
!!.!..:::: 'Y+ 1
(2-38)
-l/Jf' = 0 Po - 'Y -I (2-43a)
2-7
AMCP 706-181
Iw~'~'dn
(2-43c)
(2-54)
It is noted that an equally important
analogous solution for a cylindrical blast wave where, for most secondary explosives,
was derived by S. C. Lin 7.
AD ~ 4.5 (km/sec)/ (g/cm 3 ) (2-55)
Sakurai 21 has extended G.I. Taylor's solu
tion by expanding an expression for the shock
However, the "suitable" general equation
radius in an infinite power series of terms in
of state assumed by Landau and Stanyukovich
M. He considers plane, cylindrical, and
s~hencal waves by including a multiplier of takes the form
the form (Ro/R s ) + I , where a = 0, I, 2, in
Ci
P = 4>(v) + f(v)T (2-56)
the expansion. The basic approach is similar
to Taylor's, but higher order approximations where v is the specific volume, instead of
can be made and plane and cylindrical shocks the ideal gas form assumed by Taylor. Other
are handled in addition to the spherical equations of state will be given later in Chapter
shocks of Taylor's solution. 4.
If the forces of repulsion and attraction
2-3.2 INITIAL CONDITIONS FOR SOLU between molecules are represented by
TIONS
-n -m
4>(v) = av - bv (2-57)
2-3.2.1 INITIAL ISOTHERMAL SPHERI
CAL DETONATION FRONT and since, for highly heated explosion prod
ucts, the forces of attraction can be neglected,
A modification to initial conditions is given this equation of state reduces to
by Brode 8. However, the complex physical
chemical processes in an explosion are such (2-58)
that accurate theoretical predictions are a
very difficult task, see Oppenheim 9 . Brode's
initial condition was either a point source or *More complete discussions of equations of state appear in
an isothermal sphere. Some theoretical esti- Chapter 4.
2-9
AMCP 706-181
but B (v) is a slowly varying function and These conditions are applied throughout the
approaches Rg as v goes to 00. extremely narrow initially disturbed region.
There are then eight equations with eight
unknowns, e l , e 2 VI. V 2. PI' P2. U I U2' and
five parameters ee'Pe' Po. VOl eo' If ee' Pe *
2-3.2.2 OTHER INITIAL CONDITIONS are measured and, since Po, Vo. eo, (or Pe , Po,
To, cv )' are known for undisturbed atmo
If near-field measurements are available, spheric conditions, the initial conditions can
other initial conditions can be used to cal be completely determined.
culate the far-field properties. For instance,
one may specify either P and e TOT at the The wave speeds are
charge surface or P and ap/ar at some radial
distance r.I (Shear and Wright l I).
e2 =ee+(P2+P e )(V e -V2)/2 (2-63) *In place of measuring p , one may also measure U = D (Eq.
2-59 with D = U for an ~stimate) and add one more Eq. 2-69
(2-64) to solve for the above eight unknowns plus p e'
2-10
AMCP 706-181
tion" will occur! 3 as shown in Fig. 2-1 with INCI DENT REFLECTED
shock-fixed coordinates. For a straight inci
dent shock in such cases, there will be a SHOCK SHOCK
curved reflected shock plus a normal shock
stem near the wall. We have noted before that
the point of intersection is called the triple
point T, and that there will be a contact
surface, called the "slipstream", trailing
ul-......
_
........... _
SLIPSTREAM
similarly, Mach reflection occurs as shown in Figure 2-1. Mach Shock Reflection! 3
Fig. 2-2 with space-fixed coordinates. Whit
ham 34 formulated a two-dimensional theory*
for shock dynamics over convex or concave (2-72)
walls. He used a set of curvilinear coordinates,
(ex, (3) one parallel and the other perpendicular provided thatA s is a decreasing function of
to the shock front. They are correlated with M.s For a small change of area dA, the corre
the time-distance coordinates. The curves sponding change in Mach number is 3 3
with coordinates, ex = 00 t and (3 = constant,
are "rays", which are the discrete trajectories
dA s = - 2Ms dMs
orthogonal to the moving (curved) incident
~ -(M-"'""2- --I""")-K""""""'(M=-=s"""') (2-73a)
shock. Since the distance measured along a s
ray (3 = constant between the shock positions
at t, and t+ dt is U dt, it is equal to Ms (ex, (3) dex. where
The function Ms =Ms (ex,(3) is the local Mach
number of the shock referred to the sound
velocity of the undisturbed medium. Let the (2-73b)
corresponding distance across the "quasi
streamtube" bounded by (3 and (3 + ~(3 be
A(ex,f3) d(3. The geometric relations that are
then satisfied are and
ae _
a(3 -
I
M
(aAaex'
s)
= (-y-l) M/ + 2
s (2-73c)
(2-70) 2
2'"1 Ms -(-y-l)
M... I (aM~
= __
aex As a(3)
and and K(Ms ) is a slowly varying function of Ms
(decreasing from 0.5 at Ms = I to 0.3941 at
~a ~,(~,)]
2-11
AMCP 706-181
(2-76)
(2-77)
0.50
where
0.45
0.40
and where the "trace" of the triple point in
O. 35 +-~-r---r-.......---r--,----r~----,- the (ex, (3) plane (ex being ao t), is traveling like a
1 5 10 wave front in a one-dimensional flow diagram.
Ms The speed of the triple point CT = ilj3/ilex is
measured along this trace. Since PR is an
Figure 2-3. Graph of Chester's Function element of the trace, Eq. 2-77 yields an
K (Ms) (Ref. 14) approximate formula for CT(t).
2-12
AMCP 706-181
M
so that w (t+t1t) calculated from the "quasi
stream tube" bounded by f3 and f3 + t1f3 agrees
with that from the "quasi-stream tube"
bounded by f3 and O.
...... (0
The advantage of the method lies in a quick
, <l estimate in some cases of the locus of the
'<t:.O
\ triple point and the conditions just behind the
B \ Mach shock. A graphical procedure is required
but the assumption of negligible regular re
flection possibly limits greatly its practical
TltI
application. The locus of the triple point for
diffraction of incident shock of strength 2.81
over a cylinder is in good agreement with
experiments of Bryson and Gross 15 , Fig. 2-6;
however, there was no direct check on the
shock Mach number and the overpressures.
2-13
AMCP 706-181
4
EXP. TR IPLE POINTS
ylD
o 2
xlD
4 6
2-14
AMCP 706-181
the shock velocity is determined from the For a weak oblique blast wave encounter
energy integral. The solution is in good ing a small band along a plane wall, a theory is
agreement with known numerical and analy presented by Srivastava and Chopra I 8. It is
tical solutions for low shock strength as assumed that the relative outflow from the
shown in Fig. 2-7. reflected shock is supersonic. The solution is
5,....--...,.....-.....,...----r--~-...,.....-.....,..- __r__-_r_-'"""T""-___,
GOLDSTINE-VON NEUMANN)
EXACT NUMER ICAl SOlUTlON/ I
3
! tlOSHIMA
h /
J /~I/
2 , / ,,'
/.~,,//'
, A~-"/' / /
~~
~ "",/ " /
SECOND ORDER
~"
&~ ,," Rs~3 .. c: (A 1+ A2C: + A3C: 2 )
R
1 ~,. PERTURBATION o
ryV- SOLUTIONS FIRST ORDER
c: .. 11M2s
Figure 2-7. The Variation of Shock Strength ~ vs Dimensionless Shock
Radius R/R o for Spherical Blast Waves ('Y= 1.4) (Ref. 19)
2-15
AMCP 706-181
TABLE 2-1
BLAST WAVE ENERGY PARAMETER B
FOR SOME VALUES OF RATIOS OF THE SPECIFIC HEATS 20
References 1= 1= 1= 1= 1= 1= 1= 1= 1= "1=
5/3 7/5 1/3 6/5 5/3 7/5 6/5 5/3 7/5 6/5
3. lin 7 3.85
5. Rouse26 3.965
6. Sedov27 3.11 5.32 6.94 10.9 2.20 4.00 8.16 0.675 1.22 2.45
*Jones in his Appendix C also gives the numerical procedure to calculate B for a given
constant 1 for all three cases.
R o
(n + 2)/2
where n = 3, 2, 1, respectively, for spherical, cy For Ps/Po < 10, Sakurai's second approxi
lindrical, and plane shocks; and r* is dimension mation to Eq. 282 for the cylindrical (n = 2)
less time. is2l
2-16
AMCP 706-181
In Eq. 2-86a
r* = a
J
[(1 + bJ ~ dJ ) eJ - 1] (2-89) In the weak limit, one may assume
To check the strong blast limit, Eq. 2-82, one Ps -qJ (2-94)
has as ~ -+ 0 - proportional to ~
Po
(2-90) where
*o./Us = l/Ms =dT*/d~ and Eq. 2-86a is therefore a differen *This limit for weak spherical shock can also be derived from
tial equation which can be integrated. Eqs. (41) and (34) of Whitham 2 4.
2-17
AMCP 706-181
1, lKXll-----+------lr ----t--;t---i 10
(2-96)
Ps ' aIm
~Xl~
sec
The strong shock limit (~-+ 0) yields lool------l----/'---'\o-t---~
10\-----+---cf-----t--~-1 0.1
The weak shock limit (~-+ 00) yields
1\. aIm
~
Po ::: 100 f-----.Jf------'lr---+-=:,---------j
<; ~l +hJ ~.) (2.-1)/2.' _ ~
(2-101)
10 I---f--+-::-. "S4"G""R--A;;;IN-t------>.~
where CJ is 0.611,0.555, and 0.500 and hJ is 2SGRAIN
v AFTER VLASES AND JONES
5.89,4.80, and 4.00 for spherical, cylindrical, (1965)
and plane blast, respectively.
1 L....-_ _---J L....-_ _---J 0.01
O. 01 O. 1 1. 0 10. 0
Results from Eqs. 2-89 and 2-101 have REDUCED RAD IUS R/Ro
been compared with Brode's numerical solu
tions 8 for spherical shocks, Fig. 2-8. The The solid curves are calculated from Eqs. 2-89 and
trajectory is in excellent agreement while the 2-101 n = 2 for T and P" respectively. The data
overpressure did not agree too well near the points on the T curve are from the measurements of
weak limit, as may be expected; but, in the Vlases and Jones 22 On the P, curve the data points
range shown, the worst error is less than 22%. represent the work of Plooster.
Results from Eqs. 2-89 and 2-101 for cylin Figure 2-9. Cylindrical Blast Wave
2-18
AMCP 706-181
2-19
AMCP 706-181
By taking total derivatives and by manip Eq. 2-109 can be further reduced using the
ulating the basic equations, Theilheimer Rankine-Hugoniot Eq. 2-15, and an explicit
arrived at four equations equation of state for air. Theilheimer makes
this reduction using the perfect gas law with
ratio of specific heat 'Y = 104. When this is
p (au + u au ) + ~ = 0
at ar , ar done, Eq. 2-111 becomes
- Y2
1 = _a o (6P s +7) [7(Ps +I)
(2-108) e 7 3Ps R
!!:P.-
dR
= ~+_I
ar u at!
(a p\ P. 7 7 2 (2-112)
+ ~~ (3P; + 2P +~ +7)J s
du = au + -.l (au)
dR ar u at
These equations were combined to yield a This equatio!! is explicit in dimensionless
single equation for ap, overpressure P s' sh~k front radius R, and the
at
first derivative of ~ with respect to sho~
front radius R. So, if the "shock line" of Ps
ap 2pua2
uJ) + ~ ~a:::u)]
exposition or development of other equa
X (a' + U (U- tions. We will attempt to list here those
equations most likely to be used by a blast
analyst, under brief descriptive subheadings.
(2-111) The numbers used in earlier paragraphs in this
chapter will be retained.
2-20
AMCP 706-181
Momentum
~; + u (~;)= 0 (2-21)
au; 3 au; 1 ap
at + ~
j = 1
u. -
J ax.J + -
p
-a
xi = 0 State
Mass
3
ap + ~ =0 (2-9) 2-4.4 TAYLOR'S SIMILARITY SOLUTION
at i =1
Energy
Energy
E = 5.36 P O A 2 (2-50)
as
~~ +
i
~= 1 u/\ax)\.= 0 (2-10) Maximum Pressure
Pmax = P 1 = 0.155 R -3 E
State
Shock Velocity
p =[(p, S) (2-11)
Arrival Time
(2-15)
Shock Radius
~ 1536
R = 2. Eo 1/2 Po -1/2) JUS
t (2-52)
[2
Mass EE
at = U {2Pua2 dp r:
~ (U-u) + dR e2 +u(U-u~
)J
+du I
- a 2 pU [a 2 _(U-u)2r 1
dR
(2-20) (2-109)
2-21
AMCP 706-181
X (~ 2
+ ~ + _2 ~U (2-112)
X {~ [a 2 _ (U-U)2] } -1 (2-111) 3P s 2Ps Ps + 7
REFERENCES
2-22
AMCP 706-181
16. G. Rudinger, Wave Diagrams for Non 27. L. I. Sedov, Similarity and Dimensional
steady Flow in Ducts, D. Van Nostrand Methods in Mechanics, p. 210, Academic
Co., Inc., New York, 1955. Press, New York 1959.
17. G. G. Bach and J. H. S. Lee, "An
Analytic Solution for Blast Waves", 28. D. L. Jones, "Strong Blast Waves in
AIAA Journal, 8,2 (1970). Spherical, Cylindrical and Plane Shocks",
Phys. Fluids, 4, 1183 (1961). (Also see
18. R. S. Srivastava and M. G. Chopra, Erratum, Phys. Fluids,S, 637, May
"Diffraction of Blast Wave for the 1962).
Oblique Case", Jour. Fluid Mechanics,
40, Part 4,821-831 (1970). 29. N. Gerber and J. M. Bartos, Tables of
Cylindrical Blast Functions, BRL Memo
19. D. L. Jones, "Intermediate Strength Blast randum Report 1376, Aberdeen Proving
Wave", Phys. of Fluids, 11,8 (1968). Ground, Maryland, 1961.
2-23/2-24
AMCP 706-181
CHAPTER 3
BLAST SCALING
3-1
AMCP 706-181
p = density
= stresses or
R
1/3 Z = ---m (3-1 b)
T = tiE =scaled time E
3-2
AMCP 70~181
The use of E instead of W is preferred, for the The implications of Hopkinson scaling can
reasons that follow. perhaps be best described by the example
illustrated in Fig. 3-1. An observer located a
In much of the reported work on air blast distance R from the center of an explosive
technology, W is given in pounds weight or source of characteristic dimension d will be
pounds mass of the explosive, or in "TNT subjected to a blast wave with an amplitude
equivalent" (kilotons or megatons), which is (peak overpressure) P, a duration T, and a
common in reporting of blast data from characteristic pressure-time history p(t}. The
nuclear weapon tests. But, any study of the positive impulse I in the blast wave is defmed
physics of generation of blast waves demon by
strates that the important parameters of the ta + T
explosive source are its total energy E and its
energy density, Le., energy per unit volume or
I = f ta
p(t}dt (3-2)
3-3
AMCP 706-181
R-------i
--~
~KdL_ _ KR ~_
~- KTa
Figure 3-1. Hopkinson Blast Wave Scaling
~
at some scaled distances by factors of as much
as two. B
\\ 1\
1'\ l\1\ KA
Another example of early published work
reporting Hopkinson-scaled blast wave data is
that of Stoner and Bleakney 5, which demon '\\
\\ \ \
strated Hopkinson scaling for a limited range
\\ \\
of distances and source energies. The validity \\ \\
of Hopkinson scaling of peak particle velocity \~
has been well demonstrated for a much larger
range of explosive charge weights by Jo1m '\ .'\
2
Dewey 6, who measured such velocities in
~\
blast waves from TNT explosions. Fig. 3-4
\
shows the close agreement that Dewey found I
\.
in his scaled data for various charge 4 6 8 10 20 40
SCALED DISTANCE RfW II'. ftllb'" III
In reporting experimental data in this handbook, dimen Figure 3-2. Pressure-distance Curve for
sions used in the original references usually will be retained. Ground-burst Blast of Bare Charges4
3-4
AMCP7~181
1/3 113
Ps = 10 psi
T = tiE or t/W (3-3) I
(scaled time)
ZI = R I I (W~13 ) = 9 ft/l-Ib m )1/3
1/3 1/3
= I/E or I/W = 9 ft/lb~13
(scaled impulse)
3-5
AMCP 706-181
Symbol Wt of TNT, Ib
A 60
+ 10,000
0
40,000
200,000
1.0
~
U
g 0.8
LU
>
LU
....J
U
t
o::
c..
~0.6
LU
c..
o
LU
....J
u
tn
0.4
14
The line shows the particle velocity calculated from the shock velOCity.
Figure 3-4. Comparisons of Peak Particle Velocities for Surface Burst TNT Charges 6
3-6
AMCP 706-181
3-7
AMCP 70&181
TABLE 31
r Size of source L
,
r. Shape of source, geometry of experi
ment
U Shock velocity Ln
t Time T
blast sources
7T 3 ~ Au =A
a o (3-6)
7T4 = U/u 7T8 =tu/r
3-8
AMCP 706-181
The model law, Eq. 3-5, is quite general and 3-3 occurs if one applies to Eq. 3-9 the
until additional restrictions are imposed. relations Ao = 1, X; =AE and the cube root
First, recall that Hopkinson scaling is limited of this latterO expression, AT =A~I3. Then Eq.
to model and prototype experiments con 3-9 becomes
ducted under identical atmospheric condi
1/3 1/3
tions. Scale factors for Po and ao, relating the P =f(ri' R/E ,t/E , ... ) (3-10)
model and prototype, are, therefore, unity for
this type of scaling, Le.,
We have now seen how dimensionless prod
>. =A0 =1 (3-7) ucts uniquely determine functional forms for
'Po 0 certain dimensional groups, given specific
restrictions on a model-prototype com
This restriction drastically reduces the gen parison.
erality of Eqs. 3-6, as follows:
3-2.2 SACHS' SCALING
11'1 ~ geometric similarity of
energy
tao pA/3 )
11'4' 71's, 11'7 ~ /l..p = 1, equivalence of
E 113
density
time and
are stated to be unique functions of
space scaling
(RpA 13 /E I I3).
Eq. 3-8 is, indeed, Hopkinson's law. If one Sample Calculation
drops from Eq. 3-5 all terms which are
identically satisfied by the assumptions*, Eq. An experiment conducted under sea level
3-5 can then be rewritten as atmospheric conditions yields a prediction of
blast parameters under any other ambient
pr 3
T = f(ri'R/r. tao/r ...) (3-9) conditions. A TNT sphere weighing 2 lbm is
detonated in free air at sea level ambient
which states that a scaled pressure parameter conditions given by Po = 14.7 psi and ao =
is a function of scaled geometry and time. 1100 ft/sec. At a distance R = 9 ft from the
Had this form been used in reporting scaled center of the charge, the measured overpres
blast data, then the Hopkinson-scaled data sure is Ps = 10 psi, positive duration is T = 1.8
would have been dimensionless. Introduction msec, and positive impulse is ~ = 9 psi-msec.
of dimensions in the usual sense of Eqs. 3-1 We wish to know how this measurement
scales according to Sachs' law at an altitude of
*By "identically satisfied". we mean that all scale factors on 40,000 ft where Po = 2.73 psi and ao = 968
both sides of one of the Eq. 3-8 are unity. ft/sec, with the same explosive source.
3-9
AMCP 706-181
= 2.8 X 10 6 ft-Ib f
Then, = (1.8 msec) X fto -3 sec) X(.100 !:-\
\ msec \ sec)
R =
R p~/3
X (12 inJ X (14.7 ~br )113
E
1/3
\ ft7 m?
=
9 ft X ~4.7 ..!Q,
. 2
X 144 in~
In.
(2.8 X 10 6 ft-Ib f )
ft2
1/3
r X [~8X 10' ftolb, X 12 ~:rJ
T = 0.230
9 X 12.8
= 140.9
These values scale to the specified altitude
conditions as follows:
R = 0.821
in:
~ ft ~
113
Ib
R(ft) X 2.73~ X 144 2
Note that this parameter is rendered dimen In.
/
sionless by suitable choice of units. Similarly, 0.821 1/3
the three other groups in Sachs' law can be 140.9 (ft-lb f )
made dimensionless, as follows:
Is =
0.680 = Ps (psi){ 2.73 psi
X (968
\
~)
sec
X (12 in.)
ft
3-10
AMCP 7~181
O.230X322
T = = 4.57 msec O<p+O< +cxE
10 -3 X 968 X 12 X 1.397 1T = (M) P
3-2.2.1 ASSUMPTIONS
3-11
AMCP 706-181
(PR 3
o.
p a.f
P (h) R (h)
kp = - ,
p(O) kR = R(o)
(3-21)
p(h)
ka a'oh)
kp = -0 - ' = where 0 0 is the absolute temperature, Po is
(3-17) the ambient pressure, and 'Y is the ratio of
P'o) a'o0)
specific heats of the ambient air. When we
E (h) substitute Eqs. 3-17 and 3-21 into Eq. 3-20,
kE = -
E (0)
the scaling factors for pressure and distance
become
p
'Y
[P~)
Po
(h)
J
(h)
(h) Po
0
kp = =
p (0) Po (0)
0 'Y po(O)
Po (0)
p(h)a~h) 2 R(h) 3 =
~
1/3
(3-18)
E(h) ~
E (0) ]
(3-22)
P (0)
=
(3-19) (0)
Po
3-12
AMCP 706-181
~ j
1l3
P (h)
P 0(0) X were recorded. The two primary blast param
_0_ X R (h) R (0)
= E (0) )
eters reported were peak overpressure P and
E (h) (
positive impulse I. A number of repeat tests
(3-23) were conducted for any given condition and
distance. Fig. 3-5 shows their data for peak
Making use of the fact that the time scaling overpressure which has been scaled according
factor is defined as to Sachs' law. Their data for impulse with
Hopkinson scaling applied are shown in Fig.
3-6, and with Sachs' scaling applied, in Fig.
t (h) 3-7. One can see that, within the limits of
kt = (3-24) scaled distance covered by Dewey and Sper
t (0)
razza! 0, Sachs' law is indeed verified by their
tests. *
and that the definition of positive impulse is
given by Eq. 3-2, we obtain the equation Experiments performed at scaled distances
close to explosive sources and under very low
ambient pressure conditions (simulating high
(3-25)
altitudes above sea-level) by Jack and Ar
mendt l 3 showed that the entire character of
the blast wave changes at such distances and
By substitution of Eqs. 3-17, 3-20, and 3-21
ambient conditions and that Sachs' scaling for
into Eq. 3-25, one obtains Sachs' general
pressures does not apply. The reason for this
scaling law for positive impulse
is that the assumption that air behaves like a
perfect gas is untrue for tests close to the
blast source, and Sachs' scaling is based on
this assumption. An anomaly observed by
both Olson, et a1. 1 2 and Jack and Armendt! 3
is that this law apparently does apply for the
reflected impulse parameter, even very close
(3-26) to the explosive source. We note here that this
[E
a0
(0)1/3
(0)
p
o
(0)213
l'Ol . agreement is strictly fortuitous, and an ex
planation is given later in par. 3-2.3.2.
3-2.2.4 APPLICATION
3-13
AMCP 706-181
50
I Q..V1
L&.J
0:::
::::>
VI
VI
L&.J
20
~
Q..
~
L&.J
~
~ 10
~
Q..
C
0
L&.J
....J
<C
<..> 5
VI
2t-----t-----+-----+---+--"r--....,.------t
5 10 20 50
PO"3_R_. (atm)"3 ft/lb m" 3
WI / 3
Figure 3-5. Peak Overpressure vs Sachs' Scaled Distance l 0
law, in addition to the assumption in Sachs' scaling as the special case when there are no
law that air behaves as a perfect gas, are the changes in ambient pressure and temperature
assumptions that gravity and viscosity effects conditions between model and prototype
are negligible. Sachs' law includes Hopkinson experiments.
3-14
AMCP 706-181
20 --------,.---"""'1"":"A=-:TM":":O=-=S-=PH::":E=R::1C:-:TE::M:;P:::ER:-;A~T;-;;UR~E~: A;-;P;;;P;:RO~X;-;-IMA;;;:;TEnLvY::'+20;n-r,C
EXPLOS IVE CHARGES: 1116 - 1/21b 50150 PENTOLITE
SPHERES m
KEY: <D 1/10 atm
10
I atm
e 1/3
atm
0 2/3 atm
2/3 atm O I atm
W
E
1/10 atm
I
.Q
Cl
-
U.I
-'
u
V'l
('\"\
~ 2
Il--------+----~------="'_.J__----;;;;;:---t----t-------j
L - L - - - - l-.......L-...J- ---.L-.J20L---~---:S..!:-0----I:-:OO=--.......
0.5 2
S I0
SCALED DI STANCE Z. ft/lb 1/3
m
Figure 3-6. Hopkinson Scaled Impulse lIS Scaled Distance lo
3-2.3 OTHER SCALING LAWS FOR BLAST eters as a, E, Po, Po, R, t, 'Y; and fonn the
PARAMETERS fundamental dimensionless sets:
3-2.3.1 ADDITIONAL BLAST SOURCE
PARAMETER
5~---------j-~IJ'e-""'""'!;::-t-'ol"-t---l---r---------I
Q
~
<
U If)
VI .....
N
o
Q..
If)
.....
21-----------I~----+-------+--~---~
5 10 20 50
SCALED DI STANCE Z Po1/3, fHatm)1I3/1blT\ 113
Figure 3-7. Sachs' Scaled Impulse vs Scaled Distance l 0
The shock strength pIp can then be ex- where P is the absolute pressure, provided
pressed as 0 model and prototype experiments are con
p = -L = II (A,~) (3-28) ducted in atmospheres having the same 'Y. We
Po can see that this scaling extends Sachs' scaling
3-16
AMCP 706-181
(3-30)
Lutzky and Leht0 1 6 computed the shock
strength from a one-dimensional hydro
dynamic computer code, for various values of
(whereM T is the total mass of the explosive
~ over a wide range of X. For large enough X,
source plus the air engulfed by the shock
the shock strength was shown to be indepen
front at radius R)
dent of t i.e., Sachs' scaling applies in its
original form (see Fig. 3-8).
with the data of Olson, et aLl 2 and Jack and
Armendt 1 3, (shown in Figs. 3-9 and 3-10).
Because this expression does agree well with
f-\--::; experiment, one can generate from it a
ji. 0.158 ii- 3
\ limited scaling law, applicable only for reflect
~ e .01 ed impulse in strong shocks. If one assumes
~ e. I- 0123 that the mass of the air MA engulfed by the
~ shock front is much less than the mass of the
10
4 ~ explosive ME' i.e.,
Ie:..
0 e. 02
~I'
.... ..\
~
..... \l (3-31 )
0:
::::l
Vl e= ~\ .03
Vl
.....
0: ,,-' and one uses the fact that M1:: ex: E, then one
5
~ 10 can manipulate Eq. 3-30 to form the scaling
law
e 05 r-.. (3-32)
3-17
AMCP 706-181
DISTANCE R , ft
0.5 1.0
1.5 2.0 2.5
1000
800
0
~ 600 PREDICTED (EQ. 3-30)
EI
'Vi - - MEASURED (OLSON IT Al. 12)
c.. 400
~
LLJ
FOR l-Iq..CHARGE
VI
....J
::::>
Q..
200 ~
:E
Cl
LLJ
l
e.>
LLJ
....J 100
LL.
LLJ
c:: 80
60
40
0.4 0.6 0.8 1.0 1.2 1.4 1,6
DISTANCE R , ft
Figure 3-9. Comparison of Predicted and Measured Reflected Impulse Ir - Sea Level Conditions
3-18
800
TABLE 33
600
l8
BLAST SCALING LAWS PROPOSED BY WECKEN
400
1-Ib",CHARGE.
Po 0.0105 BAR
/ Parameter Dimensions Law 1 ~ ~ ~
200
M r L >.. 1 1
If hI
~ T >.. 1 1
100 >..3 1 >.. 1
M
...:.....
Vl
80
m
0 0 1 >..2 1 >..
:::> 60 u L T- I 1 >.. 1 1
D.. ~
~ Me 3 1 >.. 1
40
~ / I/S-Ib",CHARGE. p 1
>..2
@ ~ Po 0.0007 BAR p ML -I T- 2 1 >.. 1
'-' ~ b LT-'2 ~:I >..2 1 1
$ ~
K ML T- 2 >..2 >..2 >.. 1
"" 20 w Me T- 2 >..3 >..2 >.. 1
e L 2T- 2 1 >..2 1 1
L 2 T- 2 0- 1 1 1 1
>.. -I
C.
10 ("'T'O 1 1 1 >..
M
S
- - PREDICTED (EO.3-3O)
0.5 LO L5 ~O 2.5
DISTANCE R. It TABLE 3-4
Parameter Dimensions
other laws. In his presentation he fIrst lists the a (ambient sound veloc
0
signifIcant physical parameters, and then he ity) L r - l
immediately drops all but one of any group U (shock wave velocity) L r- t
which have identical dimensions (analogous to H (enthalpy) ML 2r- 2
the technique of using nondimensional dis
c p (specific heat at con-
tance ratios, pressure ratios, etc.). He lists L2 r- 2 (r l
stant pressure)
remaining parameters in a table with their
dimensions in an M, L, T. 0 system (0 being s (specific entropy) L2 r- 2 (r l
temperature), and enumerates four separate
"( (ratio of specific heats)
laws, simply by stating powers to which a
basic scale factor must be raised to satisfy the
particular law. The scale factor Amayor may Wecken speaks of the four laws in Table
not be the conventional geometric scale factor 3-3 as laws of similitude of length, velocity,
for any. particular law. These laws are sum density, and molar mass, respectively. That is,
marized in Table 3-3. In this table, parameters in each of these laws, one of the quantities r,
not included in other laws discussed in this u, P. M is varied (i.e., is assigned the scaling
chapter are mass of gas m, temperature 0, parameter A), while the remainder are held
acceleration b, force K, specific energy e, constant (i.e., are assigned a scaling parameter
specifIc heat at constant volume )" and molar of unity). The law of length (Law 1) is
mass of gas M. Other parameters included as Hopkinson's law. The remaining three laws
ratios of those in. Table 3-3, or already are special ones differing from any previous
nondimensional, are given in Table 3-4. laws known by the authors. They all imply
3-19
AMCP 706-181
model and prototype testing to the same would be incomplete without some indication
geometrical scale, rather than a reduced scale of response scaling, and (2) the response
for the model tests. The second law requires model laws may prove quite useful to the
"model" testing at greatly reduced tempera reader.
ture (X e = X2 ) from the prototype to achieve
the necessary reduced velocity scaling (Xu =
X). The third law requires testing at reduced 3-3.1 "REPLICA" SCALING
pressure and energy to achieve results at
reduced density. The fourth law requires the Model laws for elastic and plastic structural
use of a "model" gas with a different c (but response to blast loading are of much more
with the same ,) at different temper;tures recent origin than the majority of the blast
from the prototype. Whether Laws 2 through loading laws discussed previously in this
4 in Table 3-3 are useful scaling laws is chapter, and they usually receive scant at
doubtful. The paper by Wecken is an interest tention in any blast scaling discussion. As far
ing one, however, and well worth reading for as we can determine, the first statement of a
its other aspects, since it includes discussions blast scaling law which included structural or
so.~id material response is the law discussed by
of past theoretical work and of the history of
blast scaling. For completeness, we include Doering and Burkhardt20 . Their similarity
here in Table 3-5 a list of dimensionless theorem relates to the transmission of strong
parameters from which Wecken's analysis can shocks into a solid from an explosive source
be derived, even though he omitted this located in a fluid of any type, and essentially
is an extension of Hopkinson's size-scaling law
intermediate step in his paper.
to include shock properties in the solid. Their
proof was based on the scaling of linear
3-3 SCALING LAWS FOR INTERACTION
hydrodynamic equations, and on considera
WITH STRUCTURES
tions of boundary conditions at the interface
between fluid and solid media. They noted
Originally, we planned to limit the scope of
that viscosity, strain-rate, and gravity effects
this handbook to air blast phenomenology
must be neglected, in both fluid and solid
and specifically to exclude any detailed treat
media, for their law to be applicable.
ment of response of structures to air blast
loading. We will now depart somewhat from
this philosophy by discussing scaling laws for In a much more systematic treatment, H.
interaction of blast waves with structures, N. Brown 21 considered the interaction of
because we feel that (I) a modeling discussion blast waves with elastic structures and in
ferred the same model law as that of DOering
and Burkhardt from equations of motion. He
limited his treatment to equations for small
TABLE 3-5 deformations and strains and assumed that
pressures, stresses, and densities should be
DIMENSIONLESS PRODUCTS CORRESPONDING
unchanged between model and prototype.
TO WECKEN'S SCALING
Baker, et al. 22 , later extended Brown's anal
ut/r 1T 9 = CV M ysis to show that the same law could be used
1T1
2
for prediction of large elastic and plastic
1T2 pu /p 1TIO = U/u deformations of structures, and conducted a
3
1T3 pr /W 1T 1 1 = a0 /u series of experiments on the blast response of
1T4 bt/u 1T12 = H/W aluminum alloy cantilevers which verified the
1T K/pr 2
1T 13 s/c v law. For want of a better name, this response
5
2 scaling law is usually termed "replica scaling",
1T6 mu /W 1T 14 R/r
since geometrical similarity must be main
1T7 et 2 /r2 1T IS 'Y = C/Cp tained, and material and fluid media prop
2 erties must be identical in model and proto
1Ts C/} /u
3-20
AMCP 706-181
type experiments. The law is shown conceptu 3-3.2 SCALING FOR IMPULSIVE LOAD
ally in Fig. 3-11, and the experimental ver ING
ification for appropriately scaled response
parameters from the paper by Baker, et al. 22 Although the replica response law can be
given in Figs. 3-12 and 3-13. In Fig. 3-13, quite useful, it also imposes severe restrictions
maximum permanent deflection 8 divided by on model testing. Other laws have been
original beam length L is the nondimensional generated to allow greater flexibility in such
measure of beam response. As in Hopkinson's testing. Nevill 23 proposed a limited model
scaling, all quantities with dimensions of law for structural response which applies only
pressure and velocity are unchanged in replica for those structures whose characteristic
response scaling. So, all stress components at response times are long enough for the blast
scaled locations in the structure are un loading to be considered impulsive. In his law,
changed. All response times (such as natural scale factors for length, time, and density are
vibration periods and times for transmission independent, so that these scales can be
of elastic or plastic waves) and displacements selected arbitrarily within limits imposed by
are scaled by the same fa~tor as the length available materials and fabrication techniques.
scale factor K. Strains, being dimensionless, Thus, he introduced the concept of "dis
are identical at homologous times in the similar material modeling", wherein structural
model and prototype. response to impulsive loading can be modeled
x(t )
~ oIS.PLACEMENT
.. '\~
\ Xm
,
I
I
\\X~~--;
Il~~
___ _--1I+A~-;- t
N e~
,, lIe ----
STRAIN
, I
t.. ,J
~~--~~--R -----------t-1
t
01 SPLACEMENT
x
- -- - - - - - - ., I't::+--+
STRAIN
e
r------KR - - - - . . ,
3-21
AMCP 706-181
z 30001----,a;rw-----1--------+------I--------1
<
l=
V') .L
I
~ 2(XX)1---------.:lo'<ll:h---+-------.j-------+------~
~
X
<
~
o
1(xx) 1-------+----~}---"'""""""-+-------+--------1
OL..------.....L..------.l...-------L..---------'
10 15 20 25 30 3 l 3
SCALED DISTANCE RIW ' / , ft/lbm/
Figure 3-12, "Replica" Scaling of Elastic Response of Aluminum Cantilevers to Air Blast Waves
1. 0 ......---.L.---.~--6r--6----.,:---------r--O-l-/8--Ib..l-CH-A-RG-E-,-BE-A-M-A-..........
~. ~~-
.~ '\ J---+---f-------+-----+-~
o~ O. 6 t-----+---~-'\-\-_+_
~ 0
~0.41---~~~~-~~-~~~o~~~~--~~-~~~~~~
i ~
~ ~ ~"
O. 2 1---~~-+-~~-~---+----jD.ftJk-I~~~----,I------t----I------I
@-~j
tn~
oL_---l_ _-...L_ _L_---lL_--.l-::::. ~~~,.::::::=~~i)~
4 5 6 7 8 9 10 11
3
SCALED 01 STANCE RfWl/ , ft/lbm
1/3
accurately by the use of construction mate structural materials for the model which
rials in the model structure which may be differ in material characteristics from the
quite different from those in the prototype prototype and for model experiments which
structure. He also reports experimental ver are conducted under ambient conditions dif
ification of this law in a limited series of tests fering from those of the prototype. The
of simple structures. In this law, geometric neglect of gravity and strain-rate effects is also
similarity is maintained, and strains in the inherent in this law. The law implies geo
model are unchanged from those in the metrical similarity, and reduces to replica
prototype. The stress-strain curve for the response modeling when ambient conditions
model material must be similar to that of the are assumed identical to model and proto
prototype, but the elastic and plastic moduli type.
can be quite different. The most important
dimensionless parameter in this modeling law 3-4 LIMITATIONS OF SCALING LAWS
is
In discussing the various blast scaling laws
in this chapter, we have attempted to note the
assumptions made in derivation of the laws
1T
I
= II [L (p
sf
E)
sf
1/2 ]
(3-36)
and, therefore, some of the limitations of
11 9 Gil ui
3-23
AMCP 706-181
i
I I
I i
ILl 100
details of loading and response where such II:: I "- \
I I
'" "
,
~ \ :
effects are important. Examples of this limita Vloo Itm ... ~ 1'\ [\
tion include the inability of these laws to
II)
ILl
II::
Po - i 1i
predict the trajectories of missiles generated
Q.
II::
ILl
I'~ !
11
by blast loading, the rise of a fireball Or ~
10
i i
Z
o \\ I
i I
heated air in the atmosphere, strains or
stresses in heavily rate-dependent solid mate ~ " \ T I
iii , i
rials such as viscoelastics, etc.
1\' I
I ~
Hopkinson's scaling, and the corresponding i I~
replica response scaling for structures, applies !K
Scaled Distance 2 5
REFERENCES
3-24
AMCP 706-181
6. John M. Dewey, "The Air Velocity in mission, Rev. Ed., April 1962, pp.
Blast Waves from TNT Explosions", Proc. 128-129.
of the Royal Soc., A, 279, 1964, pp.
366-385. 16. M. Lutzky and D. Lehto, On the Scaling
of Pressures from Nuclear Explosions
7. C. N. Kingery, Air Blast Parameters with Some Observations on the Validity
versus Distance for Hemispherical TNT of the Point-Source Solution, NOLTR
Surface Bursts, BRL Report 1344, Aber 65-74, U.S. Naval Ordnance Laboratory,
deen Proving Ground, September 1966. White Oak, Md., September 1965.
9. J. Sperrazza, "Modeling of Air Blast", 18. F. Wecken, "Les Lois de Similitude dans
Use of Models and Scaling in Shock and 1es Explosions a Symetric Spherique",
Vibration (W. E. Baker, Ed.), ASME, Memorial de Q'Artillerie Francaise, Sci
New York, November 1963, pp. 65-78. ences et Technique de Q' Armement, 35,
136, pp. 438-459 (1961).
10. Jane M. Dewey and J. Sperrazza, The
Effect of Atmospheric Pressure and Tem
perature on Air Shock, BRL Report No. 19. C. Cranz, Lehrbuch der Ballistik, 2, Ber
721, Aberdeen Proving Ground, Md., lin, 1926, p. 181.
1950.
20. W. Doering and G. Burkhardt, "Beitrage
zur Theorie der Detonation", Technische
11. V. Ericsson and K. Edin, "On Complete
Akademie der Luftwaffe, Berlin-Gatow,
Blast Scaling", Jour. of the Physics of
April 1944, (translated from the German
Fluids, 3, 5, pp. 893-895 (September
as Contributions to the Theory of De
October 1960).
tonation, Tech. Report No.
F-TS-1227-1 A, by Brown Univ., May
12. W. C. Olson, Jr., J. D. Patterson, II, and
1949).
J. S. Williams, The Effect of Atmospheric
Pressure on the Reflected Impulse From
Blast Waves, BRL Memo Report No. 21. H. N. Brown, Effects of Scaling on the
1421, Aberdeen Proving Ground, Md., Interaction Between Shock Waves and
January 1960. Structures, Appendix I to BRL Report
No. 1011, Aberdeen Proving Ground,
13. W. H. Jack, Jr., and B. F. Armendt, Jr., Md., March 1957.
Measurements of Normally Reflected
Shock Parameters From Explosive 22. W. E. Baker, W. O. Ewing, Jr., and J. W.
Charges Under Simulated High Altitude Hanna, Laws for Large Elastic Response
Conditions, BRL Report No. 1280, Aber and Permanent Deformation of Model
deen Proving Ground, Md., April 1965. Structures Subjected to Blast Loading,
BRL Report No. 1060, Aberdeen Proving
14. H. L. Brode, "Numerical Solutions of Ground, Md., December 1958.
Spherical Blast Waves", Jour. of Appl.
Physics, 26, 6, pp. 766-775 (June 1955). 23. G. E. Nevill, Jr., Similitude Studies of
Re-Entry Vehicle Response to Impulsive
15. S. G1asstone, Ed., The Effects of Nuclear Loading, AFWL TDR 63-1, Vol. 1, Kirt
Weapons, U.S. Atomic Energy Com- land Air Force Base, New Mexico, 1963.
3-25
AMCP 706-181
24. W. E. Baker, P. S. Westine, and S. 25. R. E. Shear, Incident and Reflected Blast
Silverman, Feasibility Study on Simu Pressures for Pentolite. BRL Report No.
lating the Structural Response of High 1262, Aberdeen Proving Ground, Md.,
Altitude Missiles to Blast Loading, Final September 1964.
Tech. Rept., Contract No. DA-18
001-AMC-794(X), Southwest Research
Institute, Texas, January 1966.
3-26
AMCP 706-181
CHAPTER 4
COMPUTATIONAL METHODS
4-1
AMCP 706-181
X, T = dimensionless radius, time, which are imposed by both the nature of the
etc. etc., in Brode's method governing equations and the capabilities of
the computing machines. In this chapter, we
Yj = specific partial derivatives will discuss computational methods geared to
digital computers.
a, B = constant in WUNDY code
As in most facets of air blast technology,
= ratio of specific heats; a poly little work in computation of blast wave
tropic index associated with properties was accomplished prior to World
intermolecular forces War II, and the initial impetus for such work
was the result of research conducted during
Ll t, Ll x, = an increment in time, space, World War II. The best known of the early
etc. etc. efforts are those of Kirkwood and Brink
ley!,2 and von Neumann and Richtmyer 3 ,
= characteristic length in Brode's the latter paper being a classic one on which
method many following computer programs have been
based. More recent investigators who have
= constant ratio of specific heats contributed most heavily to advancements in
computational methods have been Shear at
p = density BRL, Lutzky and co-workers at NOL, Brode
at Rand Corp., and Chou and co-workers at
= initialdensity in explosives or Drexel University. Specific references to con
high pressure sphere tributions of these and other individuals will
be given with descriptions of their different
= density of ambient air methods of computation.
4-2
AMCP 70&181
(4-4)
(4-11 )
Eqs. 2-19 and 2-20 are, in Lagrangian coordi
nates,
Eq. 4-11 agrees with Eq. 4.28 of Ref. 5, while
+ 1 (op ) =0
Eq. 3 of Shear and Wright 4 is an identity
contrary to their statement derived as a third
P oa L (4-5)
T relation from Eq. 4-10.
2
2pu ::: 0 A fourth relation, along with Eqs. 4-8,4-9,
( oP) + P ( -OU) -pr- + r
OT a oaL P a 2 and 4-11, is required to solve these equations
L o L
for OUI/OT, oUt/oRs, OPI/ OT, and oPI/oRs.
(4-6) This relation is supplied by equating the
remaining energy E of the energy source at
where time t s to the work that remains to be done
on air by the blast wave as time goes to
infinity.
(4-7)
E(Rs ) = I s
r 2 (ts) [PI (ts) - Po] UI (ts)dts
and c is the speed of sound. Just behind the
shock front, (subscript 1) Eqs. 4-2,4-5,4-6, 00
(4-12)
+ ~(~;)::: 0
(4-8)**
o S 2
p u 2 dR S
R sOt
4-3
AMCP 106-181
4
An empirical equation describing the energy
time curves is Ij =1 A IJ.. YI =hi for i = 1,2,3,4 (4-19)
Let
(4-20)
then Eqs. 4-1 2 and 4-13 give
as the solution of the conservation of mass
E (R s ) = R;p 1 (ts)U 1 (ts)V KB (Rs)UKB
and momentum equations for a shock wave
(4-14) moving into an undisturbed medium. This
expresses U and Ul as functions of pressure
and density only. Eqs. 4-20 may be derived
where
from Eqs. 2-15 where P2 = Po, U2 = Uo = O.
Eq. 2-63 on solution of Eqs. 4-20 yields
!o I
00
=
,2 (t) [PI (f)-po] Ul(t)
From the equation of state,
R; [PI (ts)-Po]U 1 (ts)
c 2 = (op) (4-22)
(4-16) 1 op s
An approximate formula that depends on the The coefficients in Eq. 4-19 (Table 4-1) are all
shape of decay curve was derived by Kirk
functions of PI' R s' and E, since Po is a
wood and Brinkley for blast in air as
constant and Ul and U can be found as in Eq.
4-20. Inversion of Eq. 4-19 yields
4 -1
Y ,' = 2; A..II b.I , i = 1,2,3,4 (4-23)
j =1
Eqs. 4-13 and 4-14 yield
- 1
where A.. IS the ij-element of the inverse
II
matrix of A.
4-4
AMCP 706-181
TABLE 4-1
IN KIRKWOODBRINKLEY METHOD
A ij
All 1 A 31 PoU
A I2 0 A 32 2 -2udRs
POU
A I3 0 A 33 [(POUI dU/dpI! - 1] o
A I4 1/po A 34 [(POUI dU/dPI 1-1] U - [R: PI UI vKB (Rsl]/E(Rsl
A 21 0 A 41 1/ul
A 22 PI/PO A 42 0
A 23 1!(Plc?1 A 43 1/PI
A 24 0 A 44 0
Eq. 4-12 with respect to R s gives made by other than U. S. investigators is that
reported by Granstrom ll . He uses a semi
empirical approach. The pressure-time history
(4-25) at a fixed distance r is measured; or, instead,
if the shock front and general shape of the
In principle, Eqs. 4-24 and 4-25 can be solved blast wave are measured, the physical quan
simultaneously - given E and PI, or given PI
and api laRs - at the charge surface R, or at
a fixed value of R s . Details may be best found x - EXPERIMENTAL
in the original paper by Kirkwood and Brink - CALCULATED: INITIAL CONDITIONS
leyl. Some misprints are suspected in the ATR/R I '1
---- CALCULATED: INITIAL CONDITIONS
group of equations quoted by Shear and AT R/R 1 20
Wright 4 . IlL'" 100 !--------'>.<k--+
4-5
AMCP 706-181
2'"(/('"( - 1)
4-2.3 METHOD OF CHARACTERISTICS plpo = (clco) (4-30)
Chou 9 ,10,11 and his associates extended where the tilde means convergence between
Hartree's method of characteristics 1 2,13, for two points.
a given time interval, to a spherical explosion
with a shock front, while constant entropy is Eqs. 4-29 are obtained from Eqs. 4-27 with
assumed as in the previous investigations by the perfect gas law assumption. The waves as
Unwin 1 4 and Fox and Ralston I 5 . singularities, starting from the origin t 1 = 0, ,
= '0, are shown in Fig. 4-2. In Fig. 4-2, '0 is
In Rudinger'sl 6 terminology for one-dim the initial radius of the sphere of high
ensional unsteady flow, there are Riemann'sP pressure gas.
waves and Q waves, and entropy S waves for
which Rudinger l 6 uses a table for the conditions
behind the shock, which is not suitable for
dx computer calculation. For the right-traveling
dt = U + c, U- c, U, respectively.
shock (Fig. 4-2), at t = M, Chou, et al. I 1 ,
(4-26) derived from the Rankine-Hugoniot equations
and the constant surface conditions P3 = P4,
U3 = U4 that
In Lagrangian coordinates the respective prop
erties of the P, Q, and S waves are (Rudinger,
Ref. 16, p. 37)
HEAD OF
~ +P = _ !!. (cu ~ + r) \ RAREFACTION WAVE
, 1
(4-27) to" "1.,..'l8:,l,9~1,*0~~",rr-+--T----'f,
n
Q = - -, (cu~ - r)
The well-known Riemann variables are de Figure 4-2. Initial Singularity in Method
fined as of Characteristics
4-6
AMCP 706-181
4-7
AMCP 706-181
In addition, the likelihood of late-stage Richtmyer 3 for plane shock waves. Instead of
equivalence with equal initial energy released approximating the thin but continuous shock
Eo, but different pressure and density ratios layer by a mathematical discontinuity, it is
(Pe/Po, Pe/Po), is demonstrated by the four approximated by a layer of the order of the
examples shown in Fig. 4-5. Here, the quan thickness of one mesh size. von Neumann and
tities Pe, Pe are the initial values of pressure Richtmyer's work 3 is very well known, and
and density in the explosives or high pressure only the fictitious viscosity term and stability
sphere; Po, Po are those ahead of the main criterion will be summarized.
shock (Point 1), while the total energy Eo is
given by Fictitious viscosity coefficient,
TIME t D
TH IRD STAGE
FIRST STAGE
o~-------:;:~-------_--L...._----------
ro RAD IUS r
Figure 4-3. Schematic of Region of Numerical Solution for Method of Characteristics
4-8
AMCP 706-181
, 4
CONTACT ,,,
SURFACE
0.10 ,I
,
I
,,
,
I
0.08 I
I
Cit
SECOND SHOCK ,/
T=-- _. 0.06 I
I
E I
I
I
I
I
I
0.04 I
I
I
MAIN SHOCK
I ""
0.02 ,,/ ""
I
, /"
.-.
0
0 O. 1 0.2 0.3 0.4 0.5 0.6
A= __x.
E
Stability criteria for plane shock are: Brodel!l was apparently the first to apply
von Neumann and Richtmyer's fictitious vis
cosity technique to spherical blast waves. The
So At outside the shock, integration process consists of the stepwise
L = AaL
~l
solution of difference equations which ap
proximate the differential equations of mo
in the tion of the gas. The practical conditions to be
Sot At ~ 'Y
Y2 /(2c ) shock
Lt = VN satisfied as stated by Brode are that: (l) the
Ax region, differencing scheme must be stable, (2) it
must offer reasonable desired results, (3) it
must conserve numerical significance, and (4)
where when put in the form of coded instructions
for a high-speed computer, must be fast
enough to reach desired solutions with a rea
sonable expenditure of machine time.
4-9
AMCP 706-181
Pe Pe I
I
'TTO; flo=- I
Po Po
0.10 ~ 50 0.58
-0-- 100 1. 16
~ 200 2.32
0.08 -6
500 5.80
Y2 = 1.4
Cit
T=-- 0.06 SECOND SHOCK
SHOCK FRONT
E
COALESCE
COALESCE AT
LATE TIME
0.04
0.02
0
0 O. 1 0.2 0.3 0.4 0.5 0.6
A=~
E
Figure 4-5. Physical Plane (A-T) Showing Shock Front, Contact Surface, and Second Shock for Cases
With Equal Initial Energy and Equal Initial Mass
B = scaling length
The governing equations are then
Etot = total blast energy
af __1_ or
E int = specific internal energy ax - pf 2
Rs = shock radius ap =
aT" -P
- (2T U
+ au/ax)
af/ ax (mass) (4-39)
and the subtracted term represents the part of
the total energy ambient pressure ratio dueto au ,2 a In _)
preshock internal energy of the compressed of" - - 'Y axV'+q
(momen- (4-40)
sphere. tum)
4-10
AMCP 706-181
(energy) (4~41 )
where Q indicates the Qth node, etc.
a;:
U = aT (velocity) (4-42)
(-n+% + -n+V2)
2 \U
t
where an ideal gas has been assumed for Q uQ_ 1
which the internal energy is
;:n+l +,n + ,n+l +,n
Q Q Q-l Q-l
E int =p
p
(1' - 1)
(po)
Po
(4-43) u n +% _ li n +%
J
Q Q -1
+
and the entropy is given by Eq 2-30. Here q is ,n+l +,n _ ,n+l - -n
r
Q Q Q- 1 Q- 1
nondimensional fictitious viscous pressures.
-qn+%
91' (1' + I) (M )3B _p n + 1
q= Q -y, = 2 31T Q-%
negative.
(4-50)
The difference equations are approximated
by
p- n +1
Q-%
= ~('Y+
--p
1'-1
I -n+Y2
Q-%
-p-n
Q-%
J-n
P
Q-%
+2 (if n + 1 - Pn ) if n + yJ
J(4-5l)
'-I
Q-% Q-% Q-%
X~'Y+1'- II
-- P
-n
Q- %
- P
-n +
Q - Y,
;:n+l
~
=;:n
Q
+ u-n+%
Q
A
~ r- (4-46) .1T~.1x/[r2 (PP)JYz (4-52)
max
4-11
AMCP 706-181
200----------..-------,
24
20
.0043 p
100 .0053
.0067 10
008!01l6
oL..-_ _......J.._......u~.u....L__l_........L.:.;;.0.;;,;;,155
o
L -_ _....L.._ _.......L ....l.-_---'
3.0 .1386
.1620 1.4
If .2011
.2479 P
2.0 .2948
.3709
.4979 1.0 E~~~~><:::.-!-
0.8
0.6 L -_ _---L. ---J. ......
4-12
AMCP 706-181
= u~
J
+ AlJ
provided that q n + V. or (q n + 1 + qn )/2 is
(4-54) used. However, as q is artificial, qn is usually
used. Simple algebraic manipulation of Eq.
-q
4-60 yields
(momentum) 'Y + 1) P n + 1 v n + 1 _P n + 1 vn = 2 en
( 'Y - 1
I
from which Eq. 4-57 follows at the point j - Y2.
"+~
x~ + !
J
= x~J + ill'!! + %
In Eqs. 4-54 through 4-61
X u (4-55)
x = distance from origin
J (velocity)
u = velocity of interface
v n +!
J-%
= Vn+!,I.:
J_%/m j _%
}
(4-56) v = specific volume of zone
4-13
AMCP 706-181
q = artificial viscosity in zone* the elastic energy and the elastic part of the
pressure are predominant. Their theory has
p =p + q been described and expanded by Zeldovich
and Kompaneets 26 so that it is referred to as
g = acceleration of gravity component the LSZK equation of state, which is
(~ -i)
A'! = area of interface j at time n; = 471'
J Cv
X (X. n )2 for sphere 1 r
J B (4-62)
p=
V.J nIl = volume of zone j - Y2, 4
= "3 71'
[(X.J n )3 v"Y + v
- 72
- (x; _ Y2 )3] for sphere eE = B
+ Cv T (4-63)
("Y - 1)
t::.. tIl = time step = (6t;+ Y2+ 6t;-Y2)12 ('Y - 1) V
J
m = mass of zone
eE = energy density (per unit mass)
v = specific volume
Appropriate modifications have to be made
for other equations of state. Other details
such as initialization, rezoning, summary rou
T = temperature
tines, equation for 'Y , will not be given here. and B, Cv ' Cv ' and 'Yare constants defined
as follows/ 'Y is a dimensionless constant
Although calculations of the air shock serving as a polytropic index connected with
motion produced by a spherical TNT explo
the intermolecular forces, Cy is the specific
sion with the reaction considered gaseous heat at constant volume, Cv 1 is a specific heat
have given satisfactory agreement with experi associated with the appropriate lattice vibra
mental results, the experimental motion of tions, and B is a dimensional constant. The
the explosive interface and of the second
elastic part of the pressure is B Iv"Y andB 1[< 'Y-I)
shock have not agreed with theoretical cal
v ("Y-l Bis the elastic part of the energy. Elim
culations. Initial attempts to improve the
theory were made by Lutzky 24 using the inating T from Eqs. 4-62 and 4-63 one has
Landau-Stanyukovich equation of state that
was derived by drawing an analogy between B
v"Y
~ - Cd~ -1)J (4-64)
the state of the detonation products of a
condensed explosive and the crystal lattice of
the solid. It is well known that the energy of a where
solid body has a two-fold origin; it is made up
of an elastic energy arising from the binding ~ =
forces between the atoms and molecules, and
a thermal energy associated with oscillation of
the atoms or molecules about their positions Or, one may use
of stable equilibrium. Landau and Stanyu
kovich 2s have attempted to describe the B CvT B
p = + = - (l + y)
behavior of the detonation products by con v"Y ~v v"Y
sidering them as a solid with the property that
eE B + CvT
("Y-l)
('Y-l)v
*Lehto and Lutzky gave no expressions for q; however. in
the KO-code there were two expressions, one linear and one
B~
~+ ~~'Y-I)J
quadratic. They are 1 'Po COU\(dX) and ~ c'
, q ~4c'K 0 v d.V q KO =
v ("Y-1) (4-65)
!:..2.IOU) (dx)' with C KO = 2. Presumably some expres
y 'ox
sion similar to Eq. 4-44 was used. where
4-14
AMCP 706-181
~(~{aY+
1
y = VI
--
-y-l
that
_2 [~ (~): a)YJ~(7
(1+ -I)
-~
P = Kp + Bp"f (4-66)
2 (el - Q)
where PI = if Po, eo ~ 0
Vo - VI
K = constant of integration
0 p~) Po ~
~ VOP/{[l + ~ +(~) J
= 0 - B , for example.
(4-67)
From Eqs. 4-64 and 4-66
ci
= exKp 1/c< + _1__ B ph-I)
[~+(~) y]}
(-y - 1) (4-68)
X
from Eqs. 4-65 and 4-68,
ex K p 1/c< (4-69)
T =
Cv
dp
dp
(4-70)
1
-y-l where all quantities behind the detonation
wave are related implicitly to the specific
volume V o of the undetonated explosives
ahead of the shock, and Q is the chemical
energy released per unit mass of explosives.
4-15
AMCP 706-181
(Bruceton)
4-16
AMCP 706-181
rij =
I~ iA
F
-+ij
=-F
-+ji
r = ~j - ~i
(4-76)
-+ ij
One has 1
-2 (u. + uJ
-+1 -+/
F ..
~I/
+ g ..
-+ 1/ = f if' t. ij + g-+ 1/" = 12 ~ /.~~I -+
F ij' (u-+ i + -+
U
j
) =o (4-78)
where
in which the contribution for each pair of
(4-75) particles vanishes since F .. =-F ..' By defini
-+ 1/ -+/1
tion K j = mj!:!j lj,j /2 therefore, from Eq.
In order to correspond to fluid mechanics, 4-72
4-17
AMCP 706-181
dK duo
i = ~I .
m.1 ~I
U. -
dt
==u' }:J. F
~ ~ii
(4-79)
dt I
dt I -+ii ~i ~i
(4-83)
(
U.
-+1
n + Y,
-u
-+1
n +%)
Therefore, using Eq. 4-74
dE. dr .. where
/ . ~l/ .
- dt == - (12)
!;
L!fr
I 1/-+ ij
-dt- + (12)
/; r,l.I ",'/'
(n
<>,.
n+% + 1 ) /2
U
~
== U
~
+ un (4-84)
~
(U.
~I
u ,), where ~I/
- ~/ r .. == r i/ i. i/' gives
Kn + 1 _ K~
. n + Y,
/ /
= ~: F if U j (4-85)
dEj .' dr..1/ di~ ij~ t!:.t
== - (Y2) L /I f .. r
- 1/ -+1/
...
[ ~i - +rij-
dt dt dt
I
==
1 / 2 I .... ij
(4-86)
~U n + Y,
f dr..
_,_/ + r . ~ij
dr) ~i
+ U n.+Y2)
~ /
== - (Y2) L
i
j
ij (dt ~I/ dt
* TItis
is used in preference to a forward formula based on the
(forward formula) (4-81 ) argument of a stability requirement, Harlow and Meisner' 9
4-18
AMCP 706181
and a maximum number of neighbors N* are the most distant particle would be dropped as
predetermined for each problem, as described a neighbor.
later. Among all of the particles in the system
which lie within the search radius of the jth In order to achieve the greatest possible
particle, the N* closest particles are tenta speed with this method, it is imperative to
tively chosen to be neighbors; any of these choose R as small as accuracy will permit. The
which do not similarly find j as one of their customary choice is about one and a half
N* closest neighbors are then subtracted times the anticipated particle separation in
from the list. Good results have been attained the least compressed region.
by choosing N* to be twice the number of the
physical dimensions of the problem. An alter If one starts at the lower left corner of the
native method of achieving neighbor recip cell mesh and works from left to right and
rocity is to add rather than subtract neigh upward through the cells, it is possible to
bors, but the subtractive method has the restrict somewhat the number of cells which
distinct programming advantage that a lower must be searched for any given particle. For
upper limit can be put on the storage require instance, if the jth particle lies in cell (k, Q)
ments for inter-particle relations. Tests indi then it is necessary only to search through
cate that the two methods produce equally cells (k, Q), (k, Q + I), (k + I, Q- I),(k + I, Q),
good results. However, physically, due to the and (k + l, Q + I) for neighbors of j. If j had
selection of N*, adding may be more realistic additional neighbors in other cells bordering
than subtracting. (k, Q) this fact would have already been
determined in searching for neighbors of
The success of the PAF method depends to those particles.
such a large extent upon the proper statistical
averaging of inter-particle fluctuations that it After these tentative neighbor determina
seems necessary to search for neighbors every tions have been made, a second pass is made
time cycle. It would be preferable to avoid through the particles dropping neighbors as
this because it is by far the most time-con necessary for reciprocity.
suming phase of the calculation, requiring
50% or more of the calculation time. Experi 4-4.2.2 FORCES
ments "are" being performed to see under
what circumstances the neighbor search can The formulation is not yet complete until
be conducted less often. In addition, an the forces are defined.
attempt has been made by Daly, et aL, to
speed up this part of the calculation as 4-4.2.2.1 NONDISSIPATIVE
described in the paragraphs which follow.
The nondissipative part of the force func
The first step is to overlay the computa tiontij should be associated with the equation
tional systems with a guide of square cells, of state of the fluid through an integral
each cell having a side of length R, the search equation which in dimensional Cartesian
radius. The particles are then classified ac space has the form 3 0
cording to the cells in which they fall and the
actual search for neighbors begins. For any
2
given particle j, the distances to all other 1rn fOO 2
PD (mn, eE ) == -2 xD a
particles within its own or a neighboring cell
are computed and compared with R. I( any o (4-87)
such distance'ij is less than R, then i is listed
as a neighbor of i and j is listed as a neighbor
X (nx;)t(x D , eE)dx D
4-19
AMCP 106-181
density of neighbors as defined in par. 4-4.2.1 From Eqs. 4-87 and 4-90 it can be seen that a
and is restricted in form by the equation term \Ilk in the series of integrals is given by
1T 1a (~)d~ =N*
""
o
(4-88) 'Irk = ~
2
B k (e E ) f
""
x~aD + 2)
o
2
where the change in variable ~ = nxD
(4-89)
a = ol)(~ - I) (4-93)
The nondissipative part of the force function Substituting Eqs. 4-93 and 4-94 into Eq. 4-87
takes a similar form yields
PD (p, eE )
""
= !:l.!f f xb [(x D' eE )
o
(4-90)
x (-.!!L _l)k
p
o D
x 2
X l) (n x D
2
- l)dx
D
= N*n%
-4-
1"" 0 ~
'h (4-95)
X l) ( ~ - 1) d~
(4-91)
= N*n'h [(y'l/n,e E )
from Eq. 4-89, then a must take a form so 4
that the rest of the integrals vanish at p = po. In plane coordinates, at ~. = 1,
4-20
AMCP 706-181
4-4.2.2.2 DISSIPATIVE
(4-98)
compression; thus g..
-+1/
= for (r.
-+/
- i J
-+1
(!! i - !! j) = i ij' (.!! i - !! j) < 0, otherwise,
Eqs. 4-96 and 4-98 yield
g
-+ ij
=mw r [r .. (u. -
-+ ij -+ 1/ -+ I
uJl
-+/ :J
(4-102)
4(y-l)meE
(4-99)
N* 'ij where w is a constant with dimensions T -1.
4-88. For a "stiffened" gas equation of state and Meixner29 was inferior to Eq. 4-102
because it did not rigorously conserve angular
momentum. Further, the new alternative
(4-100) form gives a measure of the compression
between particles i and j; th us, it is consistent
with the one-dimensional stability analysis
which assumes that all compression is head
The corresponding result from Eqs. 4-96, on. Eq. 4-102 is also less likely to lead to
4-97, and 4-100 is, analogously, inter-penetration for a large value of w. Both
forms, however, have the drawback of being
4(-y-l)me E independent of the inter-particle spacing.
Therefore, further modifications used employ
N* 'ij either
(4-101)
+ lei+ ejl )'12
g =mw -
( 2m
-+ij (4-103)
X [r ij (u i - u j )1J
--> -+ -+
,-+ij
4-21
AMCP 706-181
or TABLE 43
m w ' u [ ; (u - u
o' )J
o -+ I] -+ i -+ i
INPUT DATA FOR FLOW PAST A WEDGE
g .. = r o.
PAF METHOD
-+1] r.. -+ I]
I]
and closely related to it for many other m 1.861 x 10- 5 g (particle mass)
materials. At 0.1 fJsee
,
w 0.07 - (using Eq. 4102)
The discussions on boundary conditions *here v is y. component of velocity.
and the use of images for rigid wall reflection
will not be detailed here.
Griffith's value of Ms = 1.35, however, may
be subject to an experimental error as much
4-4.3 TEST CASES as 5%; thus, the qualitative agreement be
tween the two curves is as much as could be
Let subscripts 1 and 0 be used for input expected and demonstrates that PAF calcu
data related to shocked air and ambient air, lates early bow wave development correctly
respectively. (at least qualitatively).
lines while the experiments are solid lines. 0.5 1.0 1.5
The shapes at the first observation time (18 DI STANCE X. em
p.sec) are somewhat different but essentially Figure 4-7. A Comparison of the PAF De
the same at the second time, (46 p.sec) except tached Bow Wave Positions (Dashed Lines)
for the reflection on the PAF curve which was After Impact With Those Observed in a
caused by interaction with the top reflected Shock Tube Experiment Involving a Mach
boundary. 1.35 Flow Past a Wedge 23
AMCP 706-181
1. 00 ,---...,-------,----,---------,-----,
4-4.3.2 FLOW PAST A BLUNT CYLINDER
I
et:
The detached shock front observed experi Z
<.:>
V"}
'"
V"}
cone, the steady-state pressure along the o 0.2 0.4 0.6 0.8 1.0
obstacle face and the final bow wave shape SCALED LENGTH ALONG CONE X'S
were compared with those observed experi
Figure 4-9. A Comparison of Steady-state
mentally by Marschner 32 in Figs. 4-9 and PAF Pressures (the Dots) Along the Cone
4-10. The nondimensional free stream condi Face With Experimental Values Observed in
tions are: U o =: 1.41, V o =: 0; Co =: 1.0; .!lzo =: a Mach 1.41 Flow Past a 75-deg Cone
0.1, .!l '0 =: 0.1; .!It =: 0.0 I; w =: 1.0 (using Eq.
4-103). * In Fig. 4-9, X is measured as the
EXPERIMENT distance along the cone force from the nose,
30 \
and S is the short length of the cone face to
: : : :_.~
''y
.V ..
.'-.--r.--;- the shoulder. For PAF prediction, Ps is the
theoretical stagnation pressure which develops
..... :::::~.: :. :' ',.:,
r-:-. -;-- .t-:. ")1: , ..'. .....
-J
20 .'...j ,- -<~j
..... I ' "
~.< .
to... '. ..
......
u
z
EXPERIMENT
c:::
-J
UJ
:::: j .' .~:
'1',',
:-:' . :". :-:-.' .' . .. I ....
~
Vl
C>
~~
-'
et:
-
V')~
C>
~
Z
UJV_' ) -'
et:
20 ~
o
~
Z
4-23
AMCP 706-181
fluctuations that arise as a result of the pressure continuity across an interface yields
differencing technique. This dissipative term
is of the form of a "true" viscosity in the p .. O) ==
I]
t i ) (p(i), e(.1)) =
\ I] I]
pP)
I]
sense that it is proportional to the velocity
gradient. However, artificial viscosity (of (4;105)
linear form) must be introduced for stability == t 2) (p~?) , e~.2)
~ I] I]
in the low speed region. Artificial viscosity
might be eliminated, according to Daly3 1 , by where
proper choice of the parameters of the system
to obtain a tolerable maximum error, which is
bounded due to the transport mechanism.
Details in the paragraphs which follow are p~~) = /(lI.) (
a(~)
M}/)
LU..::ly
)e(II.)}
ij
4-24
AMCP 706-181
(4-109)
+/
(2) ~(2)
Pij , e~2)J } (4-112)
- P..
I, }
,,1
1:J
The assumption here is that the volume
fraction of material I in cell ij is wherePij = Pij + qij' Based on experience, the
effects of Pij and qij are best treated separa
M(1) tely. Cell boundary velocities are averaged
= _~ __'.:. . }__ from adjacent cells. The reason that Pij and
M~.l) + R M~.2) (4-110) qij are treated in different fashion is that Pij is
I} M I}
basically a cell-centered quantity, while the
(2) fictitious viscous pressure q ij depends on
= I - a I}..
velocity differences, and is basically a cell
boundary quantity (in the PIC method). The
(where subscript 0 means ambient condition transport terms are again neglected in the
and R M is the ratio of the initial densities, * finite-difference energy equation
p~2) I p~l) and that the compression of each of
the two materials is in the same ratio as their ~
..
P,} oeij ) _{
~ __ p ..
(u j + y"j -uj - Y"j
initial compressions. It is easily seen from Eq. ut I} ~x
4-IIOthatifp(02) =O,thena(~'> I}
-+- I, andif
p(ol) = 0, a(~'> -+- O.
I}
+ V;,j+::V;,j_~)
*In Harlow's paper,38 the dot material is materiall and the
x material is material 2; however, his R is wrong in the
limiting case of p(1) -+ 0 and p( 1) -+ O. equation continued on page 4-26
o 0
4-25
AMCP 706-181
[(qU)i+%,j -(qu)i_%,j]
'Eij
" = Eij +!J.t (aaEti/') (4-116)
(where the tilde means tentative new values)
but, instead, with
[(qV)i,j + V2 - (qv)'i,j -V2]
u = (u + u)j2, v = (v + V) /2 (4-117)
~y
the proper form to use is
+ Uii (qi '~'i~qi_~'/) ~E"
(4-119)
(~y) (~t)
~ uij - --:M=-=-ij-
e I!" +
(4-120)
~ vij
If the cell is mixed, then at least three
M ij
To define the specific internal energy for a (l) The materials could be treated as
mixed cell, set though each had been compressed or ex
panded adiabatically through the same pres
ae.,)
(a/! I (aE.,) sure change.
Pij ~ (~x) (.!ly) a~ (4-115 )
(2) Each could be given the same change in
where Eij is the total internal energy of the total internal energy.
cell,
For rigorous energy conservation, one may (3) Each could be given the same change in
not use the similar approximation specific internal energy.
4-26
AMCP 706-181
The first and second were proven to be and from these the final specific energies for
satisfactory in several trials, while the third the cycle are computed from the equation
inhibited the flow of energy across an inter
face in a test problem.
H'
e~ (Q) 1
= - - - 2 [(U ' )2 + (V ' )2],
}
(4-126)
M' (Q)
Q = 1,2,
4-5.2.2 PHASE 2 OF CALCULATION (THE
TRANSPORT OF MATERIAL)
4-5.2.3 PHASE 3 OF CALCULATION
First, there is a calculation of cell-wise (FUNCTIONALS OF MOTION)
energies and momenta from quantities ob
tained in Phase 1 through the equations To allow immediate entry into Phase 1 of
the next cycle, various functionals for each
material - such as the total kinetic energy
H (Q) = M (Q) { ,(Q) ++ [(u (Q))2
and internal energy, the components of total
momentum, positions of centers of mass,
+ (jj(Q)fJ}, Q= 1,2 (4-121 ) entropy, and numerous other quantities - are
computed. In some cases, total boundary
fluxes may be used as a check on the changes
x = ~M(Q)u (4-122) of these quantities. For rigid walls the total
Q
energy should be rigorously conserved within
bounds of round-off-error. Likewise, some of
(4-123) the boundary forces can be used to check
changes in momentum components. Such
checks were found valuable to indicate ma
chine or coding error.
Next, the particles are moved, the coordi
nates of each mass point become
4-5.3 OTHER BOUNDARY CONDITIONS
is attained by setting
4-27
AMCP 706-181
U;+l,i
= -u1.1
.. (4-129) Several examples are given even though
they do not directly relate to air blast. The
iI.I+n11 , = 0
I result of a shock in nitrogen passing around a
90-deg corner was given by Harlow, et al. *35
These properties are assigned such that the and is reprod uced in Fig. 4-11 for t =
interpolated velocity at the wall between i 12.593**, M s = 1.008; and in Fig. 4-12 for t =
and i + 1 is zero. Calculation of the boundary 6.329, M s = 1.588. The rarefaction front in
value of q.l+n,1
IL uses appropriate reflected both cases is not in too good agreement with
experiment, thus suggesting that refinements
quantities. may be required for accurate prediction of
flow behind the shock. Some discrepancy in
In the velocity weighting procedure, a the shock front is also seen, which would
rectangle of cell size is imagined to be located possibly worsen as time increases. As will be
about each particle, the particle being at the commented in the FLIC method, Phase 1 of
center. Such a rectangle then overlaps four the PIC method may also need improvement.
adjacent cells and the effective velocity for
moving the particle is taken as the weighted Harlow states that the strong advantage of
average of the four cell-wise tilde velocities, the PIC method is its applicability to flow
the weighting being proportional to the over with large distortions or in which voids may
lap areas. If the surrounding rectangle lies open or close. Its disadvantages are:
partly in an empty cell, then that cell may be 1. Lack of rotational and translational
assumed to have the same velocity as does the invariance; (not a serious disadvantage in most
cell with the particle. If it lies partly outside cases)
the walls of the computation region, assumed
rigid, then the fictitious outside cells may be 2. Lack of resolution of the fine detail of a
given either reflected velocity or the same large system
velocity as in the adjacent interior cells. In the
former case, (partly in an open cell) no 3. A relatively great consumption of com
particle will be lost but may lead to "bound puter storage space (both the Lagrangian and
ary catastrophe" discussed by Harlow, et the Eulerian meshes require storage) which
al. 3S In the latter case (partly outside), it is also leads to somewhat greater computation
necessary to reflect the particle back in; the time. Computation must be made for both
particle then carries a change in momentum as
though entering from a cell with reflected *Harlow, et al. 3 5 use fust and second ordinary viscosity
coefficient while Harlow 38 , uses
velocity, and the boundary catastrophe is ou.
avoided. q .. =
If
(+ p I u.I I !:.xi)-'
oX
i
**The unit was not given, nor is t nondimensionalized;
4-28
AMCP 706-181
Solid and long-dashed lines represent, respectively, the momentum, and energy through a Eulerian
computed and observed positions of shock and rarefaction mesh of cells. While the use of these particles
fronts. Short-dashed line is a theoretical Prandtl-Meyer
streamline. facilitated the calculation of multi-fluid prob
lems, it also resulted in nonphysical fluctua
:-::::-;-:::: :-:-:::::-:::-:-:::~"7.:"~~":~::::::::"7:~":::":":":~W:":- :~~:
I
--:-:~:T::::: tions of the fluid quantities. The PIC method
~! \!\ \~j \~ [[! [[ [[:: [LU~~~\~HL:?:++~:::Ai;:L~::::::::
.,., , .. , y , ............................. . . . . . .
placed great demands on computer memory
t.t
t
, . y ,." .. , .
. capacity and calculation time due to the use
of the dual coordinate system. The FLIC
method uses concepts similar to those of the
PIC method but eliminates the computation
of the motion of particles, and it is a
"Eulerian differencing method". The scheme
to be described is suitable for problems with
symmetry about an axis or a plane.
','
J "
. '0
" ....
' '" .... 4-6.1 COMPUTING MESH
t ,',
, Gentry, et al. 3 3 use a mesh composed of
unifonn rectangles to demonstrate the FLIC
. : '
... method. To be more precise, for the plane
. ,' .. : : :
" " ','.
:. . .
................
case and the axisymmetric case, these ele-
. .. . .. , .. , .
~:..~~:-~;-~~~:~:;.~~-~~~:::::::::::::::::::::::::::::::::::::::
. . . . . . . . . . . . . . . . . . . . . . . . . . . :..
..............................
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
_~ ' __0_- _ _ ' . ' ' _ ~._ ' _ - , ' : '. ~ ...--! __ __ '..._ ..:._.__ ~._.:--=- ..~_.!...-
Solid and long-dashed lines represent, respectively. the
Figure 4-11. Configuration of Mass Points
computed and observed positions of shock and rarefaction
fronts. Short-dashed line is a theoretical Prandtl-Meyer
at Time t = 12.593 for the Calculation
streamline.
for Nitrogen With Ms = 1.008
scheme by itself would have lost accuracy Figure 4-12. Configuration of Mass Points
when fluid distortion was large. The PIC at Time t = 6.329 for the Calculation
method used fluid particles to transport mass, for Nitrogen With M s = 1.588
4-29
AMCP 706-181
ments are rectangles and toruses, respectively. By using the intermediate values of phys
The latter is to be discussed. ical quantities calculated in the donor cells,
the transport of mass, momentum, and energy
from a donor cell can be calculated.
s; + Y2 ~ 21T (i + 1) (.1 r) (.1 z), area of Energy conservation is achieved by the two
contact of step per cycle scheme and with proper choice
cell ik and of time entering the difference Eq. 4-131.
cell i + 1,
k in r-di
rection 4-6.2.1 STEP 1
The basic state variables are density p; velo Let subscript i,k be used for quantities in
cities u and w along r, z, respectively; and cell i, k. Then pI!I, k is first calculated for each
specific internal energyeE (or ei,k)' Pressure cell using the equation of state Eq. 130 and
and sound speed c are determined through the nnw' h
Pi,k' ei,k' It
eq uation of state which is, for a polytropic
gas,
n
Pi,k + 'Iz
,) } (4-132)
(4-130)
Wi,k = W n
(.1t) <k
4-6.2 THE DIFFERENCE EQUATIONS i,k P7.k(.1z)
4-30
AMCP 706-181
n n n
En B e
x {2 ~i [S~ + V2 (P7 + 1, k P7,k )
q i,k + J;h = Ci,k+% Pi ,k+V2 + %
X (W~k
I,
- W~k
I,
+ 1) (4-134)
+ sri . V2
(pn
i,k
_ 1,1
Pi - 1, k
) ]
if
(4-133)
n n
W i.k > W i,k + 1
, and
e.I,k = e I,k
~
- n -sr
X U -n)
I uU'k
I, k 72 I,
Otherwise
n (sr'+lU,
- n lk + srI,U'
-n )
+ t q+u
I 72,
k I I +, I, k
1 n r _n r
-Y2q.1-72,
l"k(S.Uk+S.
I I, I -
lU,1 -lk)
,
if
n >
-n
-w i,k - % (p 7, k + q 7k - % ) ] U i,k
n
Ui+1,k ,an d
(q~,k+V2 -q7,k-V)})
q i + %, k = 0,
where <,
k and E t k are unity for a full cell. where En is fraction of cell filled, and is unity
for a full cell, and C is the local speed of
For partial cells, these quantities will be de
fined later in this paragraph. sound. K e is a constant determining the
maximum value of the Mach number at a cell
This step is a first approximation, neglect interface for which the fictitious viscosity will
ing convective terms in the Eulerian equa be applied. Be is also a constant which
tions, to get first estimates of u, W, eE . In the determines the magnitude of the viscous
next step the effect of convective terms is pressure term, which should be large enough
calculated by taking mass, momentum, and to insure stability but small enough to avoid
energy transport into the cell into considera obscuring important details of the solution.
tion. A fictitious viscosity term is introduced Generally, Be need not exceed a value of 0.5.
in Eqs. 4-132 "to enhance the stability
properties of the difference equations in 4-6.2.2 STEP 2
regions where the fluid velocity is small
compared to the local sound speed",33 and The transport effects are calculated in this
this term is given by step. The mass flows from cell to cell are
4-31
AMCP 706-181
taken as being directly proportional to the where r,I, k is unity for a full cell.
density of the donor cell, which precludes the
possibility of developing negative density in a To determine the momentum and energy
cell, and which results in good stability transport, if j = 1,2,3,4 be the left, lower,
properties for the far subsonic regions of the right, and upper boundaries of the ikth cell,
mesh. This is known as "donor cell, mass flow respectively, with an axis z pointing to the
differencing" method. right and axis r pointing upward, and if aT. k
(j) is defined such that "
Let t1M I,~k +.,71 be the mass flowing across
the area S: during time l:1t, and t1M7 + Yl ,k be T. k (j)
"
= I, if fluid flow into cell
i,k across side j
that across the area S;+Yz' respectively, then
the mass flow from the "right" side* of the (4-139)
cell i,k is given by = 0, if fluid flows out
of cell i, k across side j
n _ _Z z n - n
ifw.I, k + I'
71
< 0 (flow in)
(4-136)
A ~An - r Sr n - At
~'l I'+"
71, k - a I'+ 1L
TJ., k I'+ 1L
n. P I,. k u,I +"71 k ~ ,
J
-
- a'i + V. ,k s'i + Yz Pin + 1 ,k U- i + Yz ,k ~t, A
(4-138)
*z-axis is in the "horizontal" direction pointing "to the *The introduction of Ti; is to conform with the "donor cell
right", mass flow differencing" idea,
4-32
AMCP 706-181
These two steps complete a cycle. A min = min lBG > 1, - 'Y BG
c /).t
/).x < A min (4-142) Figure 4- 13. Measured and Calculated
Shock Positions at the Time of First
Shock Reflection M 1 = Ms = 3. 15
where
4-33
AMCP 70&181
TABLE 4-4
PAF L-1, L-2, L-3, L-4, E-2 None of above, but may need
plus handle cavitation bigger and faster machine,
easily. since calculations are somewhat
time consuming and require a
large amount of storage.
aL and E, respectively. refer to the Lagrangian and Eulerian topics described in the table.
4-34
AMCP 706-181
space dimension. For complex situations such method of characteristics. The statement of
as strong Mach reflection, the usefulness and von Neumann and Richtmyer could be true if
accuracy of all these methods remain to be appropriate mesh and artificial viscosity con
shown. However, if one shares the optimism stants were used and if needed refinements
of von Neumann and Richtmyer 3 that the and/or modifications are introduced. Never
method of artificial viscosity would be able to theless, the method of artificial viscosity
treat "all shocks correctly and automatically would be unable to obtain more refined
whenever and wherever they may arise", this information such as the effect of true vis
approach seems to be more powerful than the cosity and energy dissipation.
REFERENCES
4-35
AMCP 706-181
16. G. Rudinger, Wave Diagrams for Non 26. Ia. B. Zeldovich and A. S. Kompaneets,
Steady Flow in Ducts, D. Van Nostrand Theory of Detonation, Academic Press,
Co., Inc., 1955. Chapter 14,1960.
4-36
AMCP 706-181
4-37/4-38
AMCP 706-181
CHAPTER 5
charge
W = mass of explosive
h = height of charge
x = distance in charge radii
H = height of gage
y = scaled height of Mach stem
= height of triple point
z = scaled distance
5-1
AMCP 706-181
instead be made from tables and graphs in lead to confusion, particularly if pounds mass
Chapter 6. of an explosive are assumed incorrectly to be
pounds force. Such confusion can be elimi
5-2 UNITS AND DIMENSIONS FOR BLAST nated by use of true energy units of force
DATA times length, rather than mass of an explosive
which simply happens to be proportional to
Blast data are reported by various investiga energy.
tors in a variety of different units and
combinations of units, and dimensions are Most blast experimentalists are familiar
given in the English or metric system, and with Hopkinson's scaling law of Chapter 3
combinations of units, and dimensions are (even though they may not know it by that
given in the English or metric system, and name) and report their data in Hopkinson
sometimes in a combination of these systems. scaled units. These scaled units will, therefore,
We will not attempt in this chapter to convert appear often in this chapter. Some investiga
reported data to a single system of measure tors simply state that their data are scaled to
ment, but will instead present data in the one pound (mass) of TNT, or one kiloton of
most commonly used system in the U.S. and nuclear yield, and do not enter the units
U.K. is the English system, and the most associated with the Hopkinson scaling. We do
commonly used units are an unfortunate not condone this practice-we merely report
mixture of English units. For explosions of it.
chemical sources, these units are:
5-3 "FREE-AIR" MEASUREMENTS
Pressures, Ps' Pro Po, etc.: pounds force per
square inch (psi or Ib f /in. 2 ) Many small-charge blast experiments have
been conducted with the ex plosive charges
Times, tao T, etc.: milliseconds (msec) and the blast transducers far enough from the
nearest reflecting surface (usually the ground)
Impulses, Is' Ir : pounds force times milli that the entire time history of the blast wave
seconds per square inch (psi-msec or lb f can be recorded prior to arrival of any
msec/in. 2 ) reflected waves. Such measurements are
usually termed "free-air" measurements, and
Distances, R: feet form much of the empirical data base for
prediction of air blast parameters and for
Velocities, U. u: feet per second comparison with theory.
Blast energy, W: pounds mass of TNT or The original free air blast measurements
other explosive. were made by various groups in the U.S. and
the U.K. during World War II. In the earliest
For nuclear explosives, similar units are work, the investigators failed to recognize
usually used, with the exception that blast that shape of the explosive charge could have
energy is usually given in units of "kilotons of a significant effect on the blast wave and that
TNT (kT)", and times are sometimes quoted the geometry of the blast transducers could
in seconds. By "kilotons of TNT" one does affect pressure by altering flow behind the
not mean that the nuclear explosive source shock front l . In spite of these failings, results
has a mass equal to the specified mass of of a number of experiments were averaged
TNT. Instead, he implies an energy released and are reported in Ref. 1 as smooth curves.
by the nuclear explosive which produces a These curves for peak overpressure and posi
blast wave which matches in intensities and tive impulse from TNT are reproduced here as
durations, over some range of distances, the Figs. 5-1 and 5-2. No detailed data are given
blast wave from the specified mass of TNT. in Ref. 1, so the curves in Figs. 5-1 and 5-2
The use of these mixed units can sometimes should be used only as indications of the
5-2
AMCP 706-181
30 10
20 1\ ~ 8 ...... -
1\ ~
""
iii
a.
1\
\.
,
"-
r\
\
'\
'"
'\ 1
5 6 8 10 20 30
\ f\.
SCALED DISTANCE R/W 1/3
, ft/lb m1/3
ozI 2
\
Figure 5-2. Logarithmic Plot of Positive
o Cast TNT l
en
1
5 6 8 10 20 30
l/3
SCALED DISTANCE R/W , ft/lb m1/3
They also obtained empirical fits to their Scaled Distance for Four Types of
5-3
AMCP 706-181
"The pressure-distance relations for the "bible" for air blast technologists. It compiles
four principal charge types are given by the measurements of peak overpressure, positive
following fitted equations, in which Ps impulse and positive durations from bare,
represents excess peak pressure in atmo spherical Pentolite which were taken from
spheres, and the distance, scaled according 1945 to 1960. Spreads in each set of reported
to charge weight, is given by the nondi data are given, both graphically and numeri
mensional variable Z =: R/(pT)1/3 where R cally. Empirical fits are made to the data for
is the distance from a charge of volume T side-on peak overpressure and positive im
and specific gravity p: pulse. Goodman also compiled data for nor
mally-reflected blast waves in Ref. 4. Figs. 54
Y2-lb rectangular blocks, TNT, through 5-6 are graphs reproduced from this
P sQ =: 13.50/Z -769.9/Z2 + 36280/Z 3 ; reference. The second report, by Lutzky at
NOL 5 , is primarily a comparison of theory
8-lb cylinders, Pentolite, with experiment, but also reports free-air
~A =: 1O.49/Z - 135.5/Z2 + 21070/Z 3 ; blast parameters which are seldom reported,
such as first and second shock arrival times
4-lb cylinders, TNT, and the time history of motion of the contact
PSB =: 1 1. 34/Z - 185.9/Z2 + 1921O/Z3 ; surface (the boundary between the hot explo
sion products and the surrounding air). Such
3.75-lb spheres, Pentolite, data are shown in Fig. 5-7, taken from Ref. 5.
Psc =: 8.63/Z + 295.1/Z 2 + 7823/Z 3
We know that the British have also con
These equations are valid for values of Z ducted free-air blast experiments, but we
between approximately 18 and 110. The
indicated probable error of the fitted
curves is of the order of one percent for
intermediate distances, increasing to from \
\-:- NORMAllY REFICTD
two to seven percent at the extremes of the \ FROM SIDE-ON
distance range covered." \\\
\
ffi 20
ducible data when detonated in small quan ~
tities. On the other hand, the more familiar !3
explosive TNT proved difficult to detonate ~ 10
8
reproducibly in quantities of one pound mass 6
and less 3 To avoid effects of charge shape,
4
cast spheres were used exclusively. Rather
than list the numerous reported and unreport
ed efforts of these two laboratories, we will 2
*The many BRL and NOL reports summarized here are listed Figure 5-4(AJ. Side-on and Normally
in the Bibliography. Reflected Pressure vs Scaled Distance 4
54
AMCP 706-181
4lXXl ~(l 20
\
\
\
2lXXl
IlXXl \ 10
8 TEN TIMES NORMAllY REFLfCTED
i,
t SMOOTHED BY EYE
e 600 6
400
4
200
I e ~ jO,...!- SIDE-oN
i 100
o'
8 '0
0
SMOOTHED BY EYE
-fo 60
40
~i 1
.8
i
20
10
~
~
.6
; 6
4
Q
~
~
.4
.2
2
11L.........~Z-'--74...l..6~1'=-O~ZO::-'-40..,.,.......~':'-'-'-11ll~iDl
Figure 5-5. Side-on and Normally Reflected Figure 5-6. Side-on and Normally Reflected
5-5
AMCP 70&181
11r----r--I-r--Ir-T-,...'T'"'I"""T'r---,.........--,-----r--T-~..,....,..,
20 r - - - r - - 1 - r - III-'I"'T"I"'T"I'T'"
II
EXPERINlNTAL POINTS
o POTIER AND JARVIS(MAIN SHOCK)
10 6 WEI BULUMAIN SHOCK)
"
8 ~ 0 FISHER(NAVORD 2980)(SECOND SHOC.K)I-~- -~=
~~-~A:'-..1~-h-.:l-~ill~;~r-
6 0 POTIER AND JARVIS (SECOND SHOCK) _1_
_~-Zr-~b~]'-
UJ
U - \l RUDLIN (MAIN AND SECOND SHOCK) -- ~ - O~:-v~/--t-+-+++I
z
4- X RUDLIN (CONTACT SURFACE)-- _ _~ ~ ~ 'V / f-
I-
Vl
Cl
e- j- j I! __~~~~r,_ - - '- ' )~~~~~D -
AJI~RSHp~vV - - --- ~v:a:1TI
-oJ
Cl
2 -- 1- MAIN:
0:::
I -- 1 - - t :
, j I!P
-- - - - - -- - ---- SURFACE -+-+-+-1--1
-~- - - - - V ' J- ------ ----- - - - 1- ---
.8 ~. - - - IJ
. 6 ~-=~--=- 7 / _-
-....-:
__ i
-
- -
- - --
~_ _= - - f
-1- ----- -- ---+-+-+--1-+-+-1
----1----+--+-+--++-++-1
r--.~--.-- ._- - --. - ---- ---- - - --+-+-+-t-~-+~
reflected wave merges with the incident wave available real estate or, for nuclear weapons,
so rapidly that a single, strengthened blast treaties which prohibit testing. So, there is a
wave is formed. The characteristics of this considerable body of experimental data for
single wave are often almost identical with the large-scale ground bursts of conventional and
characteristics of blast waves in free-air ex nuclear weapons, as well as data for small
periments, with the exception that the blast scale experiments. An advantage of the large
source appears to have greater energy than for scale experiments is that it is often possible to
free-air tests. The proportion of energy re use transducers, recording equipment and
flected from the ground is a function of how measurement techniques with relatively low
perfect a reflector it is, i.e., how little energy frequency response or time resolution, and
is imparted to the ground in cratering, ground therefore obtain measurements which cannot
shock, etc. If the ground were a perfectly be made during very small scale tests. An
rigid surface, then the equivalent "free-air" offsetting disadvantage of the large-scale tests
energy driving the air blast wave would be E' is the great cost in money, time, and man
= 2E. The other extreme case is that of a power.
perfect absorber, for which E' = E. All actual
tests will have equivalent free-air energies Measurements of air blast from ground
lying between these limits. burst explosive sources date from the same
initial efforts in World War II as do free-air
There have been very few large scale blast data. In the early test results reported in
free-air experiments because of practical Ref. I, it is noted that there is more scatter in
limits on height above ground at which the the data than for free-air measurements. Data
experiments must be conducted to avoid reported are peak overpressures and positive
ground reflection effects. For ground burst impulses, and blast sources range from 8 lb of
experiments, this limit is removed, and the bare explosives such as Composition Band
scale of the blast test is limited only by TNT to bombs with steel casings containing
5-6
AMCP 706-181
Typical curves from Ref. 1 are reprod uced 40 A BOMBS, ALL SIZES
I;>'" K. ~
I
bomb casings by use of suitable conversion U IO-Ib",BARE CHARGES
~
'8,
TWICE CHARGE WEIGHT
factors. This early work is more of interest for N T IN FREE AIR (THEORY)
t'\.1\
its historical value rather than for its ac ~ 10
l"lo.
"' A
curacy, because it has been largely superseded LU
Vl
8 ~,
i?
:;:
6 ""........."-
..........
~'I\..
ment and at a more leisurely pace. ~~
~ 4
Vl ~~~
It is apparent that a number of measure o
~~ ~
0..
o
ments of blast wave properties have been ~
B
made during nuclear tests, but until recently, <5 2
Vl
~
these data were classified. Some are now
available in Refs. 6, 7, and 8. Measurements
1
reported are times of shock arrival, peak 246810 20 40
overpressures, positive impulses, and positive SCALED DISTANCE R/WI/3 , ft/lb~3
ranging in weight from 5 tons to 500 tons and Distance Curves (on ground) from Various
,
60 \
U lO-lb BARE CHARGES
test data are shown for various blast param
lWlCE CHARGE WEIGHT
u/~ T
IN FREE AIR eters in Figs. 5-10 through 5-13. Note that
40
(THEORYl
the blast yield of the nuclear devices is
~
';;;
CL
assumed to be only half the blast yield of
0..:" 20
B TNT in these comparisons. Kingery also
\~ notes 7 that multi-kiloton nuclear tests also
1'\1\1\1\
A
LU
a::
=>
~
LU
10
K produce essentially the same scaled data at
scaled distances greater than 200 ft/(kT) 1/ 3.
g: 8
\.
a::
LU
e; 6 1\\ \ \
\\ \\ The majority of the experimental data
'"
~ ~ \\ from large explosive sources detonated on the
0.. 4
\~ ground have been accumulated during recent
years for TNT hemispheres ranging up to 500
\ \\tons in weight at the Suffield Experiment
2
\~ Station in Canada. There are a number of
~
reports for results from individual tests, such
4 6 8 10 20
SCALED DISTANCE R/W
1/3
40
,ft/lb
1/3
m
'"
as Ref. 10. The smooth curves in Figs. 5-10
through 5-13 are fits to Hopkinson-scaled
data from a number of these tests. In addition
Figure 5-8. Pressure-distance Curves (Ex to the usual blast parameters, other para
perimental and Theoretical) for Ground meters which have been seldom measured in
2 other tests were carefully measured during
Burst Blast of Bare Charges
5-7
AMCP 706-181
1ooo~-------+---r--------i
a.
~IU;AR ] NUCLEAR
LITTLE FELLER I DATA
o
L1TILE FELLER Il SCALED TO
JOHNIE BOY lkT
SMALL BOY
------j-.j
..
10
E
.P
~ 100
;::
..,-'
~
..,""""
"
r$>
. ~
-
o
{;
..
TNT SCALED TO 1/2 'T
SUGAR
JOHNIE BOY
SMALL BOY
NUCUAR
UTILE FElLER D VALUES
SCALED TO
I kT
.
I -c----j 1.0
10
" \OO=------:1:-:':OOOc:-----"---..J..,.,..------IJ 0.1
100 1000 10000
GROUND RANGE R. fl GROUND RANGE R, It
Figure 510. Scaled Arrival Time vs Figure 5-11. Scaled Peak Overpressure vs
Ground Range 7 Ground Range 7
1000 .-------,-------~------~----~-----,
o
u
w
VI
E
1--'
x
Z
o
I-
a:::
::l
f X
r
100 ~--_+__--_+.I'-----'-lll---"----\-----------'----------1
.
__ ,
-~--.
x./
o /
tj:J
I.LJ
> SUGAR
,,
/
X FIG NUCLEAR
'.
I-
VI
o
I
, o LITTLE FELLER I DATA
0..
,I
I
o LITTLE FELLER II SCALED TO
/
6. JOHNIE BOY 1 kT
--/. SMALL BOY
10 ~~ ~__+--------
-HE, STANDARD TNT SCALED TO l/2kT
---- 500-TON SNOWBALL
STANDARD TNT SCALED TO l/2kT
4L------l-------..I......-------....I.----------I
40 100 1000 10000 100000
GROUND RANGE R, ft~
5-8
AMCP 706-181
10000 x
X FIG
6
~ SUGAR
E NUCLEAR
I
VI 0 LITTLE FELLER I DATA
c..
.... II x 0 LITTLE FELLER IT SCALED TO
VI
6 6 JOHNIE BOY 1 kT
I.LJ
V"l 4 0 SMALL BOY
-I
=>
c... - HE, STANDARD TNT
:E 1000 SCALED TO 1/2 kT
I.LJ
0:::
=>
V"l
V"l
I.LJ X
0::: x
c...
0:::
I.LJ
>
0
I.LJ
>
J- 100
V"l
0
c...
many of the Canadian experiments. (These are insufficient in number to establish the
tests were also the first to demonstrate the variation of this parameter with scaled dis
"wavy" shape of the curves of scaled impulse tance. The final set of large-scale ground-burst
and scaled duration at relatively small scaled tests which we will note here were conducted
distances.) These include time histories of at Nevada Proving Ground with 20-ton spher
particle velocity and density. John M. ical TNT charges half-buried in the ground 1 3
Dewey l 1 reports data for particle velocity for The purpose of these tests (Code name Flat
a range of TNT charges from 30 to 200,000 Top I, II and III) was to obtain airblast data
lb. Typical data from Ref. 11 taken from high in the high overpressure region of 10 to
speed motion picture photography of smoke 10,000 psi, and to compare with previous
trails displaced by flow behind the shock data from the Canadian tests. Three tests were
front, are shown in Figs. 5-14 through 5-16. conducted and data collected on arrival times,
Dewey has also made an empirical fit (see Eq. side-on overpressures and impulses, and
1-10, Ref. 11) to an equation for time history dynamic pressures. Overpressures were
of decay of velocity in a blast wave. (Dewey's slightly lower than predicted for Ps > 10 psi,
parameter S is proportional to (W/ po) 1/3.) presumably because the charge was half
Anson and Dewey 12 also report some meas buried. In the same range, durations were
urements of time history of density, but they longer than predicted. Typical plots of data
5-9
AMCP 706-181
I!!
r
~
..
1&1
2
i=
6
SECONDARY SHOCK
II"
\)\f
I."ac.'(\ /'
'~
,
:
t
~
~
J
I
:
I
. t
~ 0
1&1
G .J
i ~
~
g 12
g 5 6 7 8 9 10 11
~
(.)
SCALED RADIAL DISTANCE, filS
;::
Ill:
~ Figure 5-16. x-t Diagram from Particle
><
:s 0.6
Velocity and Shock Front Datal 1
'"
0.4
5-5 MEASUREMENTS OF MACH WAVES
AND OTHER OBLIQUELY REFLECT
ED WAVES
SCALED DISTANCE, filS
The formation of Mach waves for explosive
Figure 5-14. Comparisons of Peak Particle
sources located in finite distance above a
Velocities for Surface Burst TNT Charges
reflecting surface is discussed in Chapter I.
of Various Weights from 60 Ibm to
The initial impetus for experimental studies
20,OOOlb m 11
of these waves and waves in the region of
regular reflection came from suggestions early
in World War II that blast damage from
bombs and warheads could be increased by
'are shown in Figs. 5-17 through 5-21. In this
report, the authors note that a reflection
factor of 1.7 gave reasonably good correlation
with free-air data, but that pressures were
higher over a rock surface than over alluvium.
I)
0.8
~
c
=>
.c
u
~
m_~~
0.4
~
=
u
0
-'
U-l
0.2 G
PREDICTD
MEASURED FLAT TOP I
> II Gl MEASURED FLAT TOP II
I. MEASURED FLAT TOP m
0 0.5 1.0 1.5 2.0 2.5 V 1 .L...L
I L..-...4f:...-L..l-u..u.J.L----L.-l.. II.J.,LJ 1 I...I.U.J
111I.l.l.-11_.l...-1J-..I11...L.L III
SCALED TIME cis. msec 10 100 um 10, 1m
GROUND RANGE R, ft
Figure 5-15. Comparison of the Time Varition
of Velocity at a Specific Scaled Distance Figure 5-17. Measured Arrival Times for
from Surface Burst TNT Charges from 60 Flat Top I, II, and III Compared With
Ib m to 200,000 Ibm 1 1 Prediction 1 3
5-10
AMCP 706-181
llXXl lO,lXXl
a
~~
I--.-'"" ...~ llXXl
~
/
"7
:IE
...... ..
iA; "-
.
PREDICTED
PREDICTED
(;J MEASURED FLAT TOP I
!-'
0 MEASURED FLAT TOP II
(;J MEASURED FLAT TOP I
l'\b
11> MEASURED FLAT TOP ill 0 MEASURED FLAT TOP II
&I MEASURED FLAT TOP m t'\
1 11I1111 II I1II
10
I I I1I1I11 I I I
10 100 llXXl ID,IXXl 10 100 llXXl ID,IXXl
GROUND RANGE R, It GROUND RANGE R, It
bursting in air rather than on the ground I . It findings were aimed primarily at determining
was first thought that such an increase in these optimum heights for use by weapon
damage would be due to a reduction in the designers. Loci of the triple point were
screening of one building by another. However, determined experimentally, as shown in Fig.
early experiments reported in Ref. I showed 5-22.
that screening had little effect; instead, peak
pressures and impulses both increased over Few additional experiments have been per
large areas as charges were elevated off the formed since the early work summarized in
ground, up to some optimum height of burst. Ref. I. The author has been able to locate
Later tests conducted to follow up these
llXXl
10,000 Io
\
'\
'"CL
r~
'(
-~,
0.
1000 I'\ I ...
-.; ...C1 100
'"
I
::>
II>
II>
~ "'l
. u
~ 100
: 1\ i co 10 1\
-
(;J
PREDICTED
MEASURED FLAT TOP I
! PREDICTED
""~ <:>
11>
MEASURED FLAT TO~
MEASURED FLAT TOP
~
I -'II 0
0
MEASURED FLAT TOP I
MEASURED FLAT TOP II
I I'
10
10 100
I I I I TTl '--T
1000
0.01
10,000
1 I II
10 100 llXXl lO,lXXl
GROUND RANGE R, It
GROUND RANGE R, It
Prediction 13
With Prediction 1 3
5-11
AMCP 706-181
Figure 5-22. Paths of Triple Point l When an air blast wave is reflected nor
mally from a large, rigid surface, then the
only two references reporting such experi pressures and impulses are considerably en
ments. The first of these is an investigation by hanced. The physics of normal reflection is
Bryant, et al. l 4 of triple-point loci for Pento discussed in Chapter I. Because the normally
lite spheres detonated over hard packed dirt reflected parameters represent upper bounds
and dry sand surfaces. Typical gage records to blast loading of structures, a number of
recorded during this program are shown in investigators have measured them. Most of
Fig. 5-23. These authors reported that the these experiments have been conducted using
incident and reflected shocks coalesced more small Pentolite spheres.
rapidly than indicated in Fig. 5-22. Plots of
triple-point loci from Ref. 14 are reproduced The first comprehensive set of measure
here as Fig. 5-24. The second reference ments was accomplished by Hoffman and
consists of more recent experiments by Mills l 6 , with piezoelectric gages of their own
Schleuter, et al. l 5, wherein the primary pur design (see Chapter 7) which were flush
pose of the tests was to determine the mounted in a reinforced concrete wall. Re
detailed time histories of the blast waves in flected peak pressures and impulses were
the region of regular reflection at various reported in Ref. 16, over the range in which
scaled heights over an essentially rigid (con the authors felt that gage response was reli
crete) surface. These authors also used Pento able (up to about 1500 psi overpressure). The
lite spheres. Typical time histories of data curves in Figs. 5-4(A) and 5-4(B) for normally
obtained by these investigators are shown in reflected pressure are based on this reference.
Next, Johnson, et a1. 17 devised a simple
mechanical means for measurement of im
INCIDENT
pulse in normally reflected blast waves (see
WAVE Chapter 7) which allowed measurement of
this specific blast parameter with excellent
INCIDENT accuracy to very small scaled distances. These
WAVE measurements, conducted for a range of
charge weights, agreed with Hopkinson's scal
ing law and considerably extended the range
of the previous measurements. In 1961, Olson
and Wenig I 8 employed a different experi
INCIDNT mental technique to extend the range of
WAVE
measurements of reflected pressure-time his
tories. Their technique consisted of simulta
neously detonating two equal charges placed
Figure 5-23. Typical Time Histories in Mach eq uidistant from a suitable oriented side-on
Reflection Region 14 blast transducer, and measuring the time
5-12
AMCP 706-181
- DRY SAND
60 - HARD GROUND
THE NUMBERS ON THE CURVES
REFER TO CHARGE HEIGHTS
I"'l
50
:::E
..c
E40
10
O~~~~~~~~~~~~~~
o 10 20 30 40 50 60 70 80 90 100 110
inJlb~3
I 3
HOR IZONTAl DISTANCE d IW / ,
Figure 5-24. Triple Point Loci Over Reflecting Surfaces of Hard-packed Dirt and Dry Sand! 4
~
" PETE"
P 19.10
5-13
AMCP 706-181
30,000
CURVES:
20,000 1\ - KINGERY AND PANNILL 1\
1\ -- BRL M-1273 & BRL-l1J92
f'i
1\
1\ HOI[,
10,000 VERTICAL HEIGHT <' EACH
8,000 LINE REPRESENTS STANDARD
DEVIATION <' THE INDIVIDUAL
'~ 6,000 l\ NOTE:
~
\ VERTICAL HEIGHT OF
\ EACH LINE REPRESENTS
I STANDARD DEVIATION
',THE INDIVIDUAL
"'~
2,000 " Itl
~
f'
"\
1000 Ir"
800 l'
\.
~
600
400 ~
W
~ 0.2 0.4 0.6 0.8 1
SCAlED DISTANCE
2 4
z. RIW 113 ftllb';:'
6 8 10
\\
\
200 Figure 5-27. Scaled Normally Reflected
I
I, I
\ Positive Impulse lIS Scaled Distance I 9
100 ,
I
,
80 have been made in altitude simulating cham
60
1\1 bers under conditions of reduced pressure and
temperature or both. Because such chambers
40
0.4 0.6 1.0 2.0 4.0 6.0 10.0 are necessarily limited in size, most of these
SCALED DISTANCE Z" RIW '13 , ftfl~5 measurements have been made using relatively
small explosive charges. The first series of
Figure 5-26. Normally Reflected Peak Over tests under simulated altitude conditions is
pressure lIS Scaled Distance 1 9 reported by Dewey and Sperrazza 2 1. In these
experiments, side-on pressures and impulses
were measured for a limited range of spherical
change in ambient pressure and temperature Pentolite charge weights and a limited range
from sea level conditions. Probable magni of scaled distances under conditions simula
tudes for these effects were predicted by ting several altitudes up to 55,000 ft. These
Sachs 20 when he generated his scaling law for measurements confIrmed the adequacy of
effect of ambient conditions on air blast (see Sachs' scaling law over the range of altitudes
Chapter 3 for a more complete discussion and and scaled distances possible in the test
derivation of Sachs' Law). These predictions chamber.
predated any measurements under altitude
conditions. Kennedyl reports that limited After the work of Dewey and Sperrazza 2 I ,
series of tests were conducted late in World most of the tests under simulated altitude
War II with side-on peak pressures and im conditions involved measurement of normally
pulses being measured at altitudes of 650, reflected blast parameters. Olson, et al. 22,
6600, and 14,000 ft above sea level. These measured reflected impulse for bare, spherical
results did not disagree with predictions based Pentolite in an altitude-simulating chamber in
on scaling sea-level data using Sachs' law. which ambient pressure could be varied (but
There are apparently no other reported meas not temperature), using the "flying-plug"
urements of blast parameters measured at technique described in Chapter 7. Ambient
high altitudes. Instead, most measurements pressures as low as 8 mm Hg, simulating an
5-14
AMCP 706-181
altitude of 100,000 ft, were utilized. Tests Sachs' law was strictly fortuitous, and de
were conducted at quite small scaled dis pendent on the functional relationship of
tances, for a limited range of small explosive reflected impulse with distance close to the
charge weights. Results for these tests are explosive source. They also noted that Hop
shown graphically in Fig. 5-28. These inves kinson's Law is more appropriate for this
tigators noted that the data appeared to be particular blast parameter close to explosive
consistent with Sachs' scaling law, even sources. The most recent set of measurements
though a basic assumption in Sachs' law is is that of Jack and Armendt 24 , wherein
violated for such strong shock waves. Dewey, parameters of reflected blast waves were
et a1. 23 , followed this work with additional measured with pressure transducers flush
experiments using the same technique, to mounted in the wall of a test chamber. These
investigate further the apparent agreement of authors report peak reflected overpressures,
reflected -impulse data with Sachs' law. Their reflected impulses, arrival times, and positive
results are shown graphically in Fig. 5-29. durations at ambient pressures down to 0.1
They discovered that ambient pressure had no mm Hg (approximately equivalent to
effect on impulse at "small enough" scaled 210,000-ft altitude). Pentolite spheres of
distances. They noted that agreement with nominal 1/8-lb weight were used, with a range
KEY
==
o SEA LEVEL
700 t-------CT\---~ I 200 mm HG (32,000 ftl
~ E
-,g o 40 mm HG (66,000 ftl
.:::::::
u
51 500 r---~
o 8 mm HG (l00, 000 ttl
i
E I
NOTE:
I i
iii
Q.
400 - - -------+---\c+ VERTICAL DISTANCE ON EACH GEOMETRIC
i FIGURE REPRESENTS !-crLlMITS (STA~IDARD
300 I
1- -- --~1-- -------
DEVIATION OF THE INDIVIDUAU
---~------+~------
200 ~- ------1
o l
I.LJ
--'
~
U
trl 100 +---. j... -----I
~0 I
--- - - - - - - - - f
-1 -+-----1
50 ~-----::-J':":"""'--::-~-:-l-::----~=----~=-----~:------'
CD
0.50 0.75 1. 0 1. 5 2.0 2.5
SCALED DISTANCE R/W 113 ft/lb 1/3
, m
Figure 5-28. Geometrically Scaled Reflected Impulse vs Scaled
Distance at Different Atmospheric Pressures 2 2
5-15
AMCP 7Q6.181
~
.......
1000
E 1/3
--
.0
U
Q,)
VI
500
R/W =0.5 0--- 0" . '. __ .... - ...----- ...... 0-.. _ ..-- ...... _ ... ~-_..--.... -.-.- .. --.~-o
E
VI
0 0.75-,-..-_._-" --_....0 --_-0 --.------CL .o..-~
~
.......
200
~....
1.0 ----O-P - - - - - . ' . -----.--q-.-----__. ._o.-__
"U---~
I...LJ
Vl
--J
100 _---._--0
:::l 1.5 - - - 0 - -... .....
O----o---D------O-~
a..
2:
50 2.0
a
I...LJ
--J
u
Vl 25
3.0 0--_.
of scaled distances similar to that reported in ting chambers, with the exception of a few
Refs. 22 and 23. These investigators noted data points for very large scaled distances.
that the entire time history of the reflected
blast wave changed character below about 8 5-8 MEASUREMENTS FOR SEQUENTIAL
mm Hg ambient pressure, with the usual sharp EXPLOSIONS
shock front being almost indistinguishable
below about 0.5 mm Hg. A typical time When high explosives are stored in large
history at 0.1 mm Hg is shown in Fig. 5-30. quantities, an important consideration in
Impulses obtained by integrating the pressure determining safe distances from these storage
traces agreed well with data from Refs. 22 areas is the physics of interaction of the blast
and 23, although the experimental scatter was waves from closely-spaced, sequentially
greater than for the moving plug method. initiated piles of explosive. Accidental detona
Reflected pressures in excess of 4500 psi are tion of one storage pile often will cause
reported in this reference. These authors also detonation of an adjacent pile with a signifi
noted that Sachs' scaling law cannot be used cant time lag between the two detonations.
for scaling all blast parameters close to explo Such questions as, "Do the two blast waves
sive sources. coalesce?" "Where and at what time do they
coalesce?" are important in determining pos
Indicative of the small amount of data sible blast loading of structures.
obtained under actual high altitude condi
tions, and of the difficulties of obtaining such To obtain answers to these questions, some
data, are Refs. 8 and 25. These references add blast measurements have been made on both
little to the data obtained in altitude-simula- small and large scale. All have involved deton
5-16
AMCP 706-181
_._---+--- - - ~----+-------+--
I! I
I
!
I'~'
-
l-i---+------+-----~---+----+------t
1-----+--
TIME
Figure 5-30. Normally Reflected Pressure- Time History, Scaled
Distance = O. 10 ft/lb ~ 3 0.1 mm Hg (approx. 210,000-ft altitude]2 4
5-17
AMCP 706-181
explosive charges of varying weight ratios and ment and techniques. Accuracy of measure
delay times. Time histories of the blast waves ment of course, is, affected by many factors
were measured along two orthogonal gage such as repeatability and linearity of trans
lines. Delays were varied over a large enough ducers, amplifiers, and recorders; changes in
range to determine the limits for scaled time gage sensitivity with changing ambient condi
for which the waves did not coalesce. Shock tions; spurious response of transducers to
coalescence was found to be a strong function thermal pulses, electromagnetic pulses, etc.;
of orientation as well as delay time, as can be cable noise due to triboelectric effect; suf
seen from Fig. 5-32. In this figure, "axial" ficient high and/or low frequency response of
refers to the gage line orthogonal to the line recording system; gage size and shape effects;
of centers of the two charges, and "lateral" reading errors in data reduction. If factors
refers to the gage line containing the line of such as those cited have not been considered
centers. This excellent report 2 !l should be by an experimental investigator, then it is
considered the basic reference in any study of quite possible that totally meaningless data
sequential ex plosions. can be collected and reported, and a discus
sion of possible accuracy in measuring blast
parameters is as meaningless as the data. We
5-9 ACCURACY OF MEASUREMENT OF
must also differentiate between accuracy and
BLAST PARAMETERS
precision, with the former term indicating
correspondence between measurements and
We will close this chapter with a brief
physical reality, and the latter term indicating
discussion of the accuracy with which one can the number of significant figures (or "least
expect to measure various air blast wave count") of any given measurement. Good
parameters, using present experimental equip precision is not necessarily a sign of good
accuracy. In discussing accuracy, we then
3 presume that one will employ as nearly an
SH~ri NO. 4. 11 msec/l~OELAY AXIAL
optimum system as he can for conducting his
blast experiments. In presenting estimates of
2 accuracy, we employ the common statistical
15 measure of la, or one standard deviation
14
,.., of the mean, expressed as a percent of the
;:;e
:!:! 9 mean.
-g
E 13
L.U'
Many of the measurements reported in
:E 0
I-
figures in this chapter show their accuracy
4
z
0 20 4.11
graphically by scatter in individual data
i= LATERAL points, or by symbols or bars indicating
c::
c.. 3.65
15 -_.. _ - - - - --
standard deviations of individual sets of meas
L.U
3 - -
V1
14 3.17
urements. By surveying these and other data
Cl
L.U
.....J reports, by personal experience, and by dis
u
V1
2
cussion with active investigators in blast ex
perimentation we have ascribed approximate
error bounds for various blast parameters.
5-18
AMCP 706-181
graphs or from precise timing marks im pressure returns to ambient therefore cannot
pressed on high-speed film, and detector be measured with good accuracy. For blast
locations relative to a fixed blast source and parameters such as dynamic pressure which
to each other can also be surveyed quite cannot be measured by a single transducer but
accurately. Such parameters as time of shock must be synthesized by subtracting signals
arrival and time of travel of a shock front over either electrically or by data reduction from
a known short baseline (and consequently separate transducers, the accuracy is some
shock velocity) then can be measured in the what reduced. Peak dynamic overpressures are
field over a large range of blast wave strengths probably not accurate to better than 10%;
to about 2%. Somewhat better accuracy dynamic pressure impulses, to better than
may be possible in controlled "laboratory" 20%. So few reliable measurements have been
experiments in blast chambers. made of duration of dynamic overpressure
that no accuracy can be quoted. Similarly,
The accuracy of various parameters ex not enough direct measurements have been
tracted from recordings of pressure-time his made of time histories of density to quote a
tories of either face-on or side-on blast pres probable accuracy.
sure transducers varies. Face-on peak over
pressures P, can be measured, up to levels of Finally, one of the simplest of measure
several thousand psi, with an accuracy of ment techniques for a blast parameter yield
about 5%. This same accuracy can be data with excellent accuracy and repeat
quoted for side-on peak overpressures Ps ability. The "flying plug" technique l5 for
below shock strengths at which flow effects measurement of reflected impulse I, has been
around the transducer head are important. No shown to be capable of an accuracy of about
specific upper bound for overpressures can be 2%.
cited here because the aerodynamic shape of
the transducer is all-important. For gages The numerical accuracies stated in pre
mounted flush in a level surface, however, the ceding paragraphs should by no means be
5% accuracy is possible up to several considered as "gospel truth". They represent
thousand psi. For impulses integrated from the author's opinion of what one can expect
both reflected and side-on pressure-time his to achieve in blast experiments, with four to
tories, I, and Is, accuracy is poorer, say about five replications of any measurement. Much
10%, than for peak overpressures. Lastly, larger spreads in test results probably indicate
one can expect even poorer accuracy in that something is amiss, and that one should
duration of positive overpressure T, say carefully review his entire procedure. Sig
20%. This last parameter is inaccurate because nificantly smaller spreads are probably a sign of
the approximately exponential decay of pres either quite careful work or of enough replica
sure approaches ambient pressure nearly tion of tests to insure small statistical scatter
asymptotically, and the exact time at which because oflarge sample size.
REFERENCES
5-19
AMCP 706-181
12. W. A. Anson and J. M. Dewey, Density 21. J. Dewey and J. Sperrazza, The Effect of
Measurements in the Blast Wave from a Atmospheric Pressure and Temperature
Surface Burst SOO-ton TNT Hemispher on Air Shock, BRL Report No. 721 , May
ical Charge, Suffield Technical Paper No. 1950.
305, S.E.S., Ralston, Alberta, Canada,
Aug. 1965. 22. W. C. Olson, J. D. Patterson II, and J. S.
Williams, The Effect of Atmospheric
13. R. E. Reisler, L. Giglio-Tos, and R. C. Pressure on the Reflected Impulse from
Kellner, Ferris Wheel Series, Flat Top Air Blast Waves, BRL Memorandum Re
Event, Project Officers Report - Project port No. 1241, Jan. 1960.
5-20
AMCP 706-181
25. J. G. Schmidt, Interim Report, Phase I 28. T. A. Zaker, Blast Pressures from Sequen
and II, Project BANSHEE, BRL Memo tial Explosions, Final Report J6166,
randum Report No. 1413, DASA Report IITRI, October 1969.
5-21/5-22
AMCP 706-181
CHAPTER 6
-------------------- ~61
AMCP 706-181
6-2
AMCP 706-181
6-3
AMCP 706-181
Some data in the literature are presented in of compiled data and theory, as necessary.
terms of charge radii, and many more in terms The basic blast source will be assumed to be
of charge weight W. We will use energy E bare, spherical Pentolite and all parameters
exclusively. For conversion to charge radius will be presented for free-air bursts under
or weight, one needs to know explosive standard ambient conditions.
density and specific energy release per unit
weight, mass, or volume. Densities, energies of Once one of the front parameters for
incident waves has been determined as a
detonation per unit weight and volume, and
function of scaled distance, all other param
radii of spheres weighing one pound are given
eters can be computed using the Rankine
in Table 6-1 for a few common explosives.
Hugoniot conditions and a simple integration
Similar values for many other condensed scheme. For completeness, the basic equa
chemical explosives can be calculated easily tions are given as Eq. 6-2 under the assump
from data in Ref. 10. tion of zero particle velocity ahead of the
shock front (uo = 0), followed by forms
In the graphs and tables in this chapter, derived from Eq. 6-2, for air behaving as a
quantities with a bar - i.e., ji, Ii, etc. perfect gas with ratio of specific heats equal
designate the Sachs-scaled dimensionless to'Y (see Ref. 11):
values, for which no dimensions need be
given. Quantities without bars are dimension 6-3.1.1 RANKINE-HUGONIOT EQUATIONS.
al, and dimensions will usually be given in the
English system of Ibf' in., and sec. (If the These equations hold for a unit mass of mate
subscript is omitted in units for pounds, rial.
pound force is implied.) The nondimensional
parameters that will be presented in this Mass:
chapter are listed and defined in Table 6-2. Ps (U-us)=Po U
TABLE 6-1
EXPLOSIVE PROPERTIES
Weight Volume
Specific Specific Radius r
Energy, Energy, of 1-lb
Specific Density PE , EIW, E/V, Sphere,
Explosive Gravity Ibf sec 2 /in.4 in.-Ibf/lb m in.-lbf /in. 3 in.
--
Pentolite (50/50) 1.66 1.551 x10-4 20.50xl06 1.230xl0 6 1.584
4 6 6
TNT 1.60 1.496xl0- 18.13xl0 1.048xl0 1.604
4 6 6
RDX 1.65 1.542xl0- 21.5xl0 1.283xl0 1.588
6
Comp B (60/40) 1.69 1.580 xl 0-4 20.8 x 10 6
1.271 X 10 1.575
4 6 6
HBX-l 1.69 1.580 x 10- 15.42 X 10 0.944 X 10 1.575
6-4
~MCP 706-181
("( - I)(Ps + 2)
(6-3)
es - 1 =
(6-:7)
("( + I) Ps + 2"(
"( [("( + 1) Ps + 2,,(]
TABLE 6-2
*'s and " are specific impulses, having dimensions of FTIL 2, rather than true impulses with dimensions of FT.
6-5
AMCP 706-181
U = 1 +
(:: ')
0.0494
Ps
(6-15)
the asymptote of Eq. 6-12. The very limited
data for R > 3 agree much better with the
Lehto and Larson curve.
6-6
AMCP 706181
6-12 thro ugh 6-16 define the acoustic wa ve. reflected shocks. For values of P, < 3.5,
equations for normally reflected shock front
As noted previously, all other front param parameters in a perfect gas will hold. These
eters for the incident wave can be deter are ls :
mined once Ps is defined. For values of R
from 0.01423 to 1.1, the parameters Us. U, (is
('Y + I) p S
2
(6-21 )
Scaled arrival time for the shock front, based
on a zero of shock arrival at the explosive
surface, can be obtained directly from Ref. I
out to R = 0.244. For larger scaled distances, [('Y-I)Ps + 'Y] [(3'Y- 1)Ps +2'Y
this parameter is obtained by integrating the 8, =
'Y [('Y + I) Ps + 2 'Y]
incremental relationship
(6-22)
(6-19)
6-7
0\ l>
I TABLE 6-3 SCALED SHOCK-FRONT PARAMETERS FOR INCIDENT BLAST WAVES ~
00 (")
-a
R p
Us U ta Ps Q Os
.....
s
~
...
0.01423 819 23.2 25.2 0 12.18 4570 39.9 ...
CO
10- 2 2
R
J. 5 6 7
10-1 2 J. 5 67 102 JR. 5671039104
104 9
8
7
6
1012 Q ~
6
5 5
J
n03
7
6
5
102 98 91cf
8
7
8 67
5 5
101 98 910
8
1
67 7
5 56
100
10-2
2 J
R 56710 1 2 J 56 7 100 J 567 102 2 ~. 56 7103 lcfJ
This chart is reproduced to a larger scale located in the rear of the handbook.
Figure 6-1. Compiled Shock-front Parameters for Incident Air Blast Waves
6-9
AMCP 706-181
R Pr pr er
0.0538 1840
0.070 1110
0.080 860 37.8 20.7
0.100 585 33.2 16.8
0.150 277 24.4 12.1
0.200 146 18.1 7.46
0.250 80.3 13.5 5.15
0.300 37.7 10.0 3.71
0.400 15.3 6.10 2.42
0.500 9.40 4.16 1.90
0.600 6.05 3.14 1.65
0.800 2.63 2.12 1.39
1.00 1.31 1.66 1.26
1.50 0.580 1.32 1.13
2.00 0.358 1.22 1.088
2.50 0.250 1.16 1.0612
3.00 0.188 1.12 1.0594
4.00 0.126 1.087 1.0344
5.00 9.48-2* 1.0664 1.0268
6.00 7.65-2 1.0532 1.0214
8.00 5.36-2 1.0392 1.0149
10.0 4.01-2 1.0282 1.0113
20.0 1.76-2 1.0124 1.00496
30.0 1.10-2 1.00774 1.00310
40.0 7.88-3 1.00558 1.00224
50.0 6.12-3 1.00434 1.00174
60.0 4.96-3 1.00354 1.00142
80.0 3.58-3 1.00206 1.000825
100 2.80-3 1.00165 1.000660
500 4.86-4 1.000330 1.000132
1000 2.31-4 1.000165 1.0000660
6-10
AMCP 706-181
R R 102
10-2 2 3 4 5 67
10- 1 2 3 4 56 7
10 3 4 5 6 7
101 2 3 4 5 6 7
1048
7
6
5
1()4 Pr
4
-
Pr Pr
23<'.1567 23 4 567 3 4 5 6 7 2 3 4 5 6 7 2 3 4 5 6 7 100
R 10- 1 10 R 101 102 R 103
This chart is reproduced to a larger scale located in the rear of the handbook.
Figure 6-2. Compiled Shock - front Parameters for Normallv Reflected Air Blast Waves
BRL (see Chapter 5). These data are gathered therefore, use the curves and tabular data for
in Ref. 13, and this reference has been used to Ts to estimate T, for R > 0.7.
6-11
AMCP 706-181
The empirical Friedlander equation of Chap differentiating Eq. 6-29, and is given by
ter I,
- P (b + I)
s
(6-31)
(6-29)
Values of b and initial decay rate are given in
Table 6-6, together with intermediate quan
tities used in the calculation. The parameter b
is also plotted graphically in Fig. 6-3.
is integra~ed over the posit~e phase, 0 <
(t-Ta ) < T s ' to obtain impulse [sand, the decay If the reader desires time-distance curves
constant b is determined from the resulting for blast parameters, it is suggested that he
transcendental equation obtain Ref. 6.
P
- T
sb s
[ 1
~
(I ebl }6-30)
6-3.4 OBLIQUE REFLECTION DATA
0.0500 1.84-2
6-12
AMCP 706-181
"-
"
10-3 "- 's R10- 3
R
7
6 ~I r X10- 5 "" "" 7
6
5 "-
"" "- "" 5
through the air and waves normally reflected conditions 4 ,18 . These data can, however, be
from an infini te rigid wall. The in termedia te easily converted to the Sachs-scaled param
case of oblique reflection of waves from eters of Ta ble 6- 2.
spherical blast sources located at various
heights of burst above a rigid, plane surface is Oa ta for overpressure ratios for relatively
also of considerable interest (see Chapter I). weak shocks at all angles of incidence up to
It is much more difficult to present compiled grazing (XI = 90 deg) are reproduced graphic
blast parameters in concise form for this case ally here in Fig. 6-5 from a set of curves in
because an additional geometric parameter, Ref. 4. Although the measurements on which
heigh t of burst H or angle of incidence (XI' is these curves are based are not described in
required to define the geometry of the shock Ref. 4, it is likely that they date from World
reflection. Fig. 6-4 shows this geometry in the War II or shortly thereafter, and consist of
region of regular reflection, i.e., for (x, small tests using small TNT or Pentolite charges.
enough that the Mach stem has not been Recent work at BRL indicates that these old
formed (see Chapter I). Because the reflecting curves may still be reasonably accurate over
surface of most interest is the ground, data the indicated range of shock strengths.
for oblique reflections seldom are presented
in Sachs-scaled parameters, but instead are Kingery and PanilJ I S have generated a
given for either Hopkinson-scaled or unscaled comprehensive set of tables and graphs for
quantities for standard sea level ambient prediction of a number of shock-front param
6-13
AMCP 706-181
SLANT RANGE R
INCIDENT ~ 0"
n'
~ ,
SHOCK WAVE _____ o t36
O '0 20
cr
ANGLE OF INC I OrnCE 01
ANGLE
OF REFLECTION O2
Function of Angle of Incidence for
TABLE 6-6
*Oigits preceded by minus (plus) sign indicate negative (positive) powers of 10.
6-14
AMCP 706-181
form is also given as a function of shock Ref. 18 if he wishes to predict front param
strength. The data in Ref. 18 are all given in eters for obliquely reflected strong shocks.
dimensional units for one pound of spherical
Pentolite at standard sea level ambient condi 6-3.5 CONVERSION FACTORS
tions. They are much too voluminous to
reproduce here, but some typical data are The nondimensional presentation of blast
shown in Tables 6-7 and 6-8, and Figs. 6-6 parameters used here has the great advantage
and 6-7. It is suggested that the reader obtain of being independent of systems of units. It
TABLE 6-7
Peak
Peak
Horizontal Overpressure Incident Reflected
Dynamic
Distance Incident Reflected Pressure Angle Angle
d, ft Ps ' psi P" psi Q, psi (X" deg (X2, deg
6-15
AMCP 706-181
TABLE 6-8
Peak Side-
on Over- Slant Height of Horizontal
aextreme
pressure Range Burst Distance
6-16
AMCP 706-181
10 , - - - - , - - - - - - , - - - r - - - - - , - - - - , - - - - - - ,
H 2.0 It
H ' 2.5 It
2.25
.~
.~
ci
~ 5
'"::>
'"
'"
~
'"
0.. ~
U
:<
3.25 ~
'" 3
""
:5
3.50 0..
35
3.75
does, however, have the disadvantage of dif From Table 6-1, weight specific energy for
fering from the presentation in most other Pentolite is
sources of blast properties, with the exception
of Brode 2 ,3. Some conversion factors are
listed in Table 6-9 for those who wish to E!W = 20.50 X 10 6 in.-Ibf/lb m
convert the nondimensional parameters of
Table 6-2 to dimensional ones, or vice versa. So,
The standard atmospheric conditions given
previously are assumed in calculating these E = W X 20.50 X 10 6
conversion factors.
= I X 20.50 X 10 6
To demonstrate the use of figures and Standard sea level ambient conditions are:
tables in this chapter, we will present here
some example calculations. 14.70Ib f /in. 2
Po
Example 1. We wish to predict blast param Po = 1.146 X 10- 7 Ib f-sec 2 /in. 4
eters in free air at 10ft from the center of a
I-Ibm Pentolite sphere detonated under sea ao = 13,400 in./sec
level ambient conditions, R = 120 in., W = 1
Ibm' eo 5l9R
6- \ 7
AMCP 706-181
TABLE 6-9
R~,r
P Pr
ft
(r _
atm
Ibm
Goodman l , Mills,
et a1. 6 , Baker &
Schuman s
0.1075
-
R
- -
-Pos -
Po I
atm Goodman l , Mills, 1 Ps 'Pr
I et al 6
-
X charge radii Goodman! 0.01428 R
- -
~:,) (;", )
Goodman! , Mills, 0.120 T, ta
msec et al 6 , Baker &
- -113 Schuman s
Ibm
I / (p 213 wI 13 )
so' - -
psi-msec Goodman! , Mills, 4.92 X 10- 3
I /(p 2/3 W1/3)
Is' I r
r 0 atm 2/3 Ib 1/3 et a 1. 6
m
- -
Ps 'Pr psi Lehto & Larson 1 2, 0.0681 Ps , Pr
Goodman, etc.
-
R meters Lehto & Larson 12 0.353 R
- -
R - Brode 2 ,3 1.13 R
- - - -
T, ta - Brode 2 ,3 1.18 T, t a
Scaled distance R = Rpo 1/3/113 (Table 6-2) write down all of the side-on scaled param
eters we can find. These are:
IJ3
120 X 14.70
R= I.n. (lbI!3)
---,f:........,,....
20.50
1/3
X J0 2
.
10.
2/3 Ps 0.450 Q = 7.20 X ]0- 2
U
s
0.276 8s 1.12
X
1 )
= 1.20 X 2.452 =1.072 *
( in. 113 Ib;13 2.74 X 10 U = 1.17 Is
= 2.85 X 10 - 2
6-J 8
AMCP 706181
=: 2.33 msec
Us = Us XQo =: 0.276 X 13,400 in./sec
=: 15,700 in./sec
Example 2. We wish to predict properties
of a blast wave normally reflected from a rigid
- 113
t a X E 0.520 X 2.74 X 10 2 wall located R = 15 ft from the center of a W
ta =: = = 10 Ibm RDX sphere, in an altitude chamber
a Xp 1/3 1.34 X 10 4 X 2.45
o 0
which is at sea level temperature, but at
reduced pressure and density which are III 0
= 4.33 X 10- 3 sec =: 4.33 msec
of sea level values.
So,
Q = Q X Po = 7.20XI0- 2 X 14.70 psi
E = W X 21.5 X 10 6
=: 1.06 psi
=: lOX 21.5 X 10 6
es =: es
X eo = 1.12X 519R=581oR
2.15 X 10 5 in.-Ib f
2/3
X EI/3
Is X Po Ambient conditions are:
is
Qo
2/3
2.85 X 10- 2 X (14.70) X 2.74 X 10 2 Po = 0.1 X 1.146 X 10-7 = 1.146 X
=
1.34 X 10 4
X Q
o =: 13,400 in./sec*
(in ./sec)
6-19
AMCP 706-181
- 1/3
Scaled distance R Rp01l3/E1/3
Tr X E 8.20 X 10- 2 X 600
IS X 12 X (1.47) 113
1.34 X 10 4 X 1.137
(0.215 X 10 9 ) 1/3
~ = 0.258
er = er X eo = 3.03 X 519 R
= 47.97 - 0.34
2/3
0.258 X (1.47) X 600 = 47.63 deg
1.34 X 10 4
The angle of incidence (XI = 50 deg and (X ex
6-20
AMCP 706-181
REFERENCES
8. J. R. Ruetenik and S. D. Lewis, Com 17. A. J. Hoffman and S. N. Mills, Jr., Air
putation of Blast Properties for Spherical Blast Measurements About Explosive
TNT or Pentolite from Measured Pressure Charges at Side-On and Normal inci
Histories, AFFDL-TR-66-47, October dence, BRL Report No. 988, Aberdeen
1966, AD-807 756. Proving Ground, Maryland, July 1956.
6-21/6-22
AMCP 706-181
CHAPTER 7
7-1
AMCP 706-181
slender, streamlined housing. They are all under their supervision. Some of the early
directional, in the sense that they will read difficulties in attempting to use gages that
properly the free field blast properties only were not streamlined are mentioned by
when properly oriented with respect to the Stoner and Bleakneyl. The design which has
direction of shock travel. No suitable omnidi evolved at BRL is shown schematically in Fig.
rectional blast transducer of this type has as 7-1. Gages of this general design often are
yet been developed. called colloquially "pancake" or "lollipop"
gages because of the flat disc shape of the
Many of the different types of side-on blast housing. The sensing elements consist of
gage employ either natural or synthetic piezo stacks of even numbers of piezoelectric discs
electric materials as sensing elemen ts. The placed in a cavity in the housing as shown,
natural crystals are usually either tourmaline and interleaved with metal foil discs and tabs.
or quartz, while the synthetic materials may The individual piezoelectric discs are alter
be barium titanate, lead zirconate, lead meta nated in polarity in the stack, with all tabs of
niobate, or other materials of compositions one polarity being connected to an insulated
which are considered as proprietary by their electrical lead in the stem, and the others
manufacturers. Natural crystals only can be being grounded to the metal housing. Elec
used in the form of flat plates or discs cut trical connection is made to a coaxial con
along certain crystal planes, while the syn nector at the end of the stem.
thetic materials can be made and polarized in
a wide variety of different geometries. The The elements must be made of a material
synthetic materials are usually much more that is sensitive to hydrostatic pressure for
sensitive than the natural, but are apt to this gage to function properly -usually either
exhibit appreciable changes in sensitivity with tourmaline or a synthetic pizeoelectric
change in ambient temperature, while the ceram ic*. The head of this gage is made in a
natural materials do not. Advantages to use of variety of sizes, depending on desired sensi
piezoelectric pressure sensing elements are (I) tivity and scale of experiment, but the ratio
the elements are self-generating and very of diameter to thickness always is kept greater
linear over extremely wide ranges in applied than 10: I to minimize flow effects. In use it
pressure, (2) very high frequency response is normally is mounted at the end of a long tube
possible, and (3) most of the piezoelectric with the stem parallel to the direction of
materials are quite strong mechanically and travel of the blast front. It can be employed,
therefore can survive high pressures and much however, for measurement of any blast wave
rough handling. Disadvantages are (1) they do whose direction of propagation lies in the
not respond to static pressure and so are plane of the gage head, and has been used to
difficult to calibrate, (2) they are brittle, (3) measure the characteristics of two blast waves
they are without exception also pyroelectric striking head-on 2 . As far as we know, there is
so that one must guard against direct heating no presently available commercial counterpart
of the elements during passage of a blast to this type of gage.
wave, (4) they are sensitive to acceleration,
and (5) they only can be employed properly 7-2.1.2 SOUTHWEST RESEARCH INSTI
with recording equipment of very high input TUTE SIDE-ON GAGES
impedance to minimize leakage of the elec
trical charge generated under application of A gage similar in basic design to the BRL
pressure. gage has been developed at Southwest Re
search Institute (SwRI). The basic com
ponents of this gage are shown in Fig. 7-2(A),
7-2.1.1 BRL SIDE-ON GAGES
and completed gages in Fig. 7-2(B). The
sensing elemen t is a two-crystal piezoelectric
Much of the development of side-on blast
pressure gages in the U. S. has been done at
the Ballistic Research Laboratories (BRL) or *Quartz is insensitive to hydrostatic pressure.
7-2
AMCP 706-181
CRYSTALS
GAGE
HOUS ING CAVITY FILLED WITH
SILICONE GREASE
_.
-
-
(A) GAGE COMPONENTS
7-3
AMCP 706-181
stack of lead zirconate, 0.060-in. thick, and housing by mounting on corprene rings, and is
0.125-in. in diameter. The gage components covered by an ou ter neoprene shea th that
are assembled in a mold, and the housing cast provides electrical and heat insulation. Elec
from an epoxy resin. The sensing element is trical connections are made to an integral
isolated mechanically from the housing by a coaxial connector at the rear of the housing.
thin layer of silicone rubber painted on before In use, this type of transducer is mounted in
casting. It is considerably smaller than the much the same manner as the BRL "pancake"
smallest BRL gage, being intended for use in gage, with the exception that its axis always
quite miniature-scale experiments. Possible must be closely aligned with the direction of
advantages over the BRL gage in addition to blast wave travel if it is to record side-on
the small size are a su perior aerody nam ic pressures.
shape with very high diameter-to-thickness
ratio of the "pancake" head and smooth 7-2.1.4 BRITISH SIDE-ON GAGES
transition of head into stem, fast rise-time
because of small sensor diameter, and relative At several British laboratories, side-on
ease of manu facture. These particular gages piezoelectric gages have been developed and
are designed for mounting in the end of a used for air blast measurements. Several ver
0.5-in. diameter tube. In spite of their small sions of a more or less standard basic design
size, they have good sensitivity and can be have been evolved at Royal Armanent Re
used to measure low overpressures with long search and Development Establishment
lines in the field. (RARDE). The basic design is the standard
H3 gage illustrated in Fig. 7-4. The sensitive
7-2.1.3 ATLANTIC RESEARCH CORPORA element consists of twelve X-cut quartz
TION SIDE-ON GAGES crystals sandwiched between a pair of one
inch diameter pistons that serve to com
Cylindrical side-on blast gages ("pencil" municate the pressure to the crystal pile. The
gages) employing synthetic piezoelectric ele whole pile is suspended between neoprene
ments are made commercially by Atlantic rubber diaphragms that are clamped around
Research Corp. These gages originally were their periphery to the gage body and provide
developed under contract to BRL. One is a radially compliant mounting. This feature
shown in Fig. 7-3. The sensing element is a helps to reduce the response of the crystal
small hollow cylinder of barium titanate or pile to axial acceleration forces when the gage
lead zirconate which has been polarized radi is struck by a blast wave. The gage housing is
ally. This element is shock isolated from the made of stainless steel, with a molded plastic
rear section. The gage has a nominal sensi
tivity of approximately 100 *pC/psi, but the
actual sensitivity of individual gages is deter
mined, in the la bora tory, by a q uasi-sta tic
calibration system. These calibrations are re
pea ted at regular intervals and in practice the
sensitivity constant of a well-made gage, used
normally, changes very little even over a
period of years. The undam ped na tural fre
quency of the crystal pile has been calculated
to be about 200 kHz. Field experiments have
shown that only if the gage is subjected to
high shock strengths do marked oscillations
appear in the decay curve and that these die
Figure 7-3. A tlantic Research Corp. Pencil
au t very rapidly. The measured frequency of
Blast Gage, Type LC-13
(Courtesy of Atlantic Research Corp.) 'pC ~ picocoulomb
7-4
AMCP 706-181
the damped oscillations is approximately 130 overpressures indicated by the H3B gages
kHz, and it is concluded that the crystal pile were compared with those given by a com
is very nearly critically damped. pletely independent system, and excellent
agreement was obtained.
In use at high overpressures CPs> 3), the
rear portion of the H3 gage was found to be A third version of the basic H3 gage is
too weak. A stronger design was evolved, designated H3C, and is illustrated in Fig. 7-6.
using the same hatchet-shaped front position, The sensing element is identical to that in the
but a streamlined brass rear section that was other two designs, but the housing is similar
threaded to mate with a 1.25-in. O.D. tube. to the BRL pancake design. The disc-shaped
This gage, designated H3B, is shown in Fig. head of the gage is made of aluminum alloy,
7-5. The British report use of this gage in and has an aspect ratio of greater than 12/ I.
free-field and Mach stem measurements of The head is attached to a streamlined brass
pressures up to and exceeding Ps = 20. The rear section that is threaded to mate with
H3B gage has been calibrated under the 1.25-in. O. D. tubing, as is the H3B gage. The
conditions of use, i.e., using explosively gen British report that the use of this gage is
erated blast waves of widely differing inten similar to use of the H3B, but with omnidirec
sities and hence now velocities. The peak tional properties in one plane.
.,.---
7-5
AMCP 706-181
{British Crown copyright reserved. Reproduced with the permission of the Controller, Her Brittanic Majesty's
7-2.1.5 OTH ER SIDE-OI\J GAGES here that some types may be preferable to
piezoelectric gages for long duration blast
Many types of miniature pressure trans waves, because they have DC response and are
ducers can also be employed as side-on gages, less sensitive to temperature or acceleration.
provided they are small enough to be
mounted in one or both sides of a pancake 7-2.2 REFLECTED PRESSURE GAGES
head or other suitable streamlined housing.
Ruetenik and Lewis J report of the use of In the measurement of blast pressures
small, commercial transducers in this manner.
The design, shown schematically in Fig. 7-7, is
nearly iden tical in geometry to the BRL --- -- ---1
side-on gages. Pierce and Manning4 also have
used very small flush-diaphragm gages of
NASA design in a similar manner. Whiteside 7
discussed the use of a number of different
SHAEVITZ-BYTREX ....-
types of piezoelectric elements made in the TRANSDUCERS
British Atomic Weapons Research Establish
ment (AWRE), Foulness, in side-on trans
ducers. We will defer until later in this chapter
the discussion of the characteristics of these
and other miniature transducers which could
o <> D.Sin.
be used in suitable housings as side-on gages,
because their use as blast transducers is not Figure 7-7. Side-on Blast Gage Using Small
limited to this application. We merely note Flush-diaphragm Transducers
7-6
AMCP 706-181
reflected at normal or oblique incidence from time histories for reflected pressures up to
a rigid plane surface, flow and diffraction about 1500 psi. Granath and Coulter6 later
effects are no longer important, provided that improved somewhat on this design, develop
the pressure transducer is capable of being ing the gage shown in Fig. 7-8. This gage has
mounted with its sensing element or dia been used successfully for reflected pressure
phragm flush with the reflecting surface. measurements up to 4500 psi. It is also sold
Many more types of transducers thus can be commercially.
satisfactorily used for measurement of re Finally, the British at RARDE have de
flected pressures than can be used for side-on veloped a gage for reflected pressure measure
pressures. We will confine our discussion here ments using the basic quartz sensing element
to those gages that either have been designed of their H3 side-on gages. No data are
specifically for reflected pressure measure available regarding the upper limit of pressure
ments, or are too large to be classified as at which it is considered usable.
minia ture transducers.
The majority of the data reported to date 7-2.3 MINIATURE PRESSURE GAGES
on reflected blast waves from small explosive
charges were obtained using a reflected pres Many types of miniature pressure trans
sure gage designed by Hoffman and Mills at ducers have been found to be suitable as air
BRL s . This gage used tounnaline or synthetic blast gages, either for mounting in small
piezoelectric sensing elements mounted in a models or structures to measure details of
cavity in (he face of a massive metal housing. diffracted pressure loading, for mounting in a
The length of the gage housing effectively large flat surface to measure reflected pres
determined the lower limit on response time, sures, or for mounting in streamlined housings
i.e., several times greater than the transit time to record side-on blast wave pressures. Aside
of an elastic dilatational wave along the length from their small size, the transducers that
of the housing. This gage produced acceptable perform well under blast loading all appear to
7-7
AMCP 706-181
have several characteristics in common. They ducers that have been developed at Govern
all have a sensing surface that can be mounted ment laboratories. Several types of gage em
flush with the surface of some housing or ploying piezoelectric elements have been de
structure. This surface may be a diaphragm veloped at BRL. Two types, designed and
whose displacement under applied pressure is made by Granath and Coulter 6 for mounting
sensed, a diaphragm or piston transmitting in small models, are shown in sectional view
pressure to an interior piezoelectric sensor, or in Fig. 7-9. The smaller of these gages is 0.31
a surface containing one or more ports leading in. in diameter and 0.31 in. long. Somewhat
to a very small interior chamber containing an similar gages were designed by Baker and
interior sensor. They all employ quite simple Ewing 9 for flush-mounting in the surfaces of
and rugged construction. Most have been airfoils subjected to blast loading. These latter
developed specifically for measurement of gages, about 0.5 in. in diameter and 0.5 in.
blast pressures, and are designed to minimize long, were designed to minimize acceleration
spurious signals from acceleration and heat. and transient temperature sensitivity. One
Some also are designed to minimize effects of type is shown in sectional view in Fig. 7-10.
nuclear radiation on output or sensitivity. From sinusoidal vibration tests, the signals
Those transducers which are of complex generated under accelerations of one "g"
in ternal construction, particularly those em normal to the gage face were determined to
ploying unbonded wire strain gages, have be less than that which would be generated by
generally proven to be quite unsuitable for air 0.004 psi blast pressure, over a wide range of
blast measurements. excitation frequencies. The frequency re
sponse of these gages is dependent on the
In paragraphs that follow we will limit our diameter of the piezoelectric discs that con
discussion to those transducers that have been stitute the sensing element and the velocity of
tested thoroughly and have been proven the shock front passing over the gage, rather
adequate as air blast transducers. than on the inherent frequency response of
the sensing element. For weak shock fronts
7-2.3.1. BRL MII\IIATURE TRANSDUCERS
moving at near sonic veloci ty, rise times are in
the order of 10-12 microsec. The lower limi t
Let us first discuss the miniature trans on frequency is dependent on the input
impedance of the associated amplifiers, as is
true for any capacitive device.
7-8
AMCP 706181
CERAMIC DISCS
EPOXY
RESIN
SOLDERED
/' . BRASS
CONNECT IONS
PIN
TEFLOI\J
INSULATOR
TEFLON
TLI BII\JG
~METAL CON
NECTOR BODY
COPPER
WIRE
STAINLESS STEEL
HOUS I NG
7-9
AMCP 706-181
DIAPHRAGM
3/16 in.
PARAPLEX
T
0.1 in.
_l~~~~,--JJ
REFERENCE
TUBING
Figure 7-". Sectional View of NASA Miniature Transducer of Morton and Patterson I I
slightly less for more sensitive gages. As for ture sensitivity of 0.0 I percent of full scale
the earlier design, this gage exhibits quite low per one degree Fahrenheit are claimed by the
sensitivity to temperature and acceleration. manufacturer, rendering this transducer much
less sensitive to these effects than are most
7-2.3.3 OTHER MII\IIATURE TRANS other piezoelectric transducers. Response
DUCERS time to reflected shock waves has been
demonstrated to be about one microsecond.
Several commercial firms manufacture min
iature pressure transducers that employ piezo The Kaman Nuclear Division of Kaman
electric materials as sensors. Atlantic Research Aircraft Corp. manufactures a line of variable
Corp. produces several types. Photographs of reluctance blast pressure transd ucers which
several of these gages are reproduced in Fig. act on an eddy-current loss principle. Mag
7-12. The smallest (Type LD-80) is 0.14 in. in netic flux lines, emanating from a coil, pass
diameter and 0.5 in. long. All employ syn into a diaphragm and produce eddy currents
thetic piezoelectric materials as sensors. The in the conductive diaphragm circuit. As the
manufacturer does not report sensitivity to diaphragm is brought closer to the coil, more
temperature or accelerations. The Type flux lines are intercepted and the losses
LD-80 transducer is claimed to be usable for become greater; and, as the diaphragm is
shock pressures as high as 10,000 psi and to moved away from the coil, the losses become
have face-on rise times of less than one less. When the coil is connected as an arm of a
microsecond. conventional AC bridge circuit, the coil im
pedance wjJl change with applied pressure;
The Kistler Instrument Corp. makes several and this in turn will result in a change of the
types of pressure transducers employing AC output signal from the bridge circuit. The
quartz as the piezoelectric sensing element. electrical output of the bridge can be made
One type in particular, Model 603A, has been linear with respect to the applied pressure by
employed as a miniature blast pressure trans proper selection of the diaphragm thickness,
ducer. This gage, shown in Fig. 7-13, is 0.22 active diaphragm area, diaphragm material,
in. in diameter by 0.45 in. long. A novel and diaphragm-to-coil spacing. The basic ele
feature is the inclusion of additional quartz ment is shown in Fig. 7-15. The smallest gage
discs to compensate for acceleration, as is somewhat larger than the other miniature
shown schematically in Fig. 7-14. Accelera gages discussed previously, but all of the
tion sensitivity of 0.001 psijg and tempera Kaman Nuclear gages are shielded from nu
7-10
AMCP 706-181
clear radiation effects, as well as being de Nuclear gages all have response, when used
signed to minimize temperature and accelera with suitable carrier systems.
tion sensitivity. Acceleration sensitivity of the
Model K-1200 gage is stated as 0.01 percent One of the few types of miniature strain
of full scale per "g", depending on pressure gage pressure transducer that has proven
range and output sensitivity. The Kaman useful for blast wave measurements has been
7-11
AMCP 706-181
QUARTZ CRYSTALS
J - FLAT COIL 6 - ELECTRICAL TERMINAL
7-12
AMCP 706-181
ducer, and a ceramic element such as barium deformation of a diaphragm by strain gages
titanate or lead zircon ate in the NOL trans bonded to a cylindrical tube that supports the
ducer. Levine l 2 notes that one must be quite diaphragm. A schematic of one of these, the
careful in the design and construction of such Dynisco PT 76, is shown in Fig. 7-17. Strain
a transducer to isolate completely the ac elements are bonded to a thin cylinder that
celerometer-element from direct pressure and has one end secured to the case and the other
from the effects of pressure on the element attached to the diaphragm. The small mass
housing. and minute deflection resolution result in
very-high-frequ ency response characteristics,
Two types of commercial miniature trans bu t the bonded strain-wire gages have low
ducers of very similar internal design have sensitivity output about 2 to 4 m VjV full
been employed on nuclear or long-duration scale. Although the design and assem bly of
conventional air blast tests 8 . Both sense the the gages reportedly make them insensitive to
OUTER BAFFLE
SHELL
INNER BAFFLE
STRA IN TUBE HOLD DOWN
PSC CO~INECTOR
BODY STRA IN TUBE HOLDER
T 102 WM-6P-F2
DEFLECT ION
TUBE
PISTON
DIAPHRAGM WITH
ZERO BALANCE COMPENSATION
ALUMINUM OXIDE
COMPENSATION HEADER STRAIN TUBE
COATING
o 0.5 1.0
SCALE: IN.
Figure 7-17. Dynisco Pressure TransducerS
7-li
AMCP 706-181
vibration and accelerations no q uan titative of the blast front over known, accurately
values are given by Rowland~. Bonding the measured base lines, or the time of arrival of
strain gages to the tu be ra ther than directly to the front at a number of stations at various
the diaphragm delays the effects of thermal distances from the blast source. To measure
transients, and the two passive arms of the time-of-arrival, one must detect a "zero
bridge circuit are used for temperature com time", or time of start of release of blast
pensation which further reduces thermal ef energy. Simple devices also are employed for
fects. this purpose.
7-14
AMCP 706-181
TABLE 7-1
Sensitivity
..-t
mY/psi
Transducer Maximum
Transducer {/J with
Diameter, Pressure, p.. Applications
inches
Type
psi
...... Cable
u Supplied
p..
6-1/2
FQ Ilc
Pressure
Insert
300 100
Q Omnidirectional
Baffle B12
or
( 0
~
Unid irectional
Baffle B2
Side-<>n
--~
Face-<>n
~e:>
Total Head Pressure Measurements
Using Bafl1e BS 5
reflections after shock arrival, since one is nector, and a dust cap for this connector, as
interested only in generating a large signal on basic parts. The sensing element is a single
arrival of the front. Watson and Wilson 13 piezoelectric disc, mounted on a brass pedes
describe a transducer of this type which is in tal and "potted" in epoxy resin.
current use at BRL. It is shown in cross
section in Fig. 7-20. The gage uses a com There are also several time-of-arrival gages
mercial ly available feed-through coaxial con- available commercially. Atlantic Research
7-15
AMCP 706-181
MX-913/U
REAR BODY
1 IN.
COAXIAL SOCKET
Transducer
Watson and Wilson I 3
7-16
AMCP 706-181
q = Pu 2
/2 (7-1 )
where
\ GAGE HOUS ING
p = density in blast wave \ ELECTRICAL
\ LEADS
\ .
U = particle speed
tion pressure Ph
BRL Total Head Gages
(7-2)
I 7-4.2 DRAG GAGES
where
Drag or dynamic pressure gages are de
Ps = peak side-on pressure signed to measure only the dynamic pressure
q. They consist of bodies of some regular
shape for which steady-state wind-tunnel drag
data exist, mounted on or incorporated in a
7-4.1 TOTAL HEAD GAGES sensing system that measures total drag force
on the body, or one or more components of
The total head gages resemble pitot tubes this force. Drag pressures are then inferred
that are used to measure this same parameter from the outputs of the sensors, using the
in steady or slowly-varying flow fields. They conventional aerodynamic drag formula
consist of axisymmetric housings with, blunt
noses, and sensing elements located at fltagna (7-3)
tion points for flow around the housin~s (see
Fig. 7-21). For gages intended to be used for
I
where
relatively weak blast waves where peak par
ticle speed Us is considerably less than Isound D = drag force
speed, the nose is hemispherical. (Fig.
7-2l(A)). For stronger shocks where ~s ap CD = a drag coefficient which is presumed
proaches or exceeds sound speed, a trupcated to be known for the particular shape
conical nose shown in Fig. 7-21 (B) is used. of body
These gages have been made only in Irather
large sizes for use in field tests o( large S = projected area of the body normal to
chemical or nuclear blast sources. the flow direction
AMCP 70&181
For these gages to function properly, the drag conventional explosive charges (Rowland
body must be small compared to the length of 1967). Some of these are described in the
the blast wave, and the lowest vibration paragraphs that follow.
period of the body-sensor system must be
much shorter than the blast wave duration.
7-4.2.2 NOL DRAG FORCE GAGES
7-4.2.1 DRAG GAGE OF JOHNSON AND
EWING NOL developed and used three-component
force gages in a number of nuclear tests.
An example of a drag gage is discussed by These gages measured the blast-wave-induced
Johnson and Ewing l 4 . Their gage consists of forces on a small target in three mutually
a cantilever beam of circular tubular cross perpendicular axes. The targets were spheres,
section which protrudes above a base plate. cubes, cylinders, and parallelepipeds. Vari
Bending strains are sensed in the tube near its able-inductance sensing elements contained
base, in two orthogonal planes. The averages within the target responded to the excitation
of each strain-time history recorded as a blast produced by the blast wave. The targets were
wave sweeps past the gage are measures of spring-mounted on sets of springs located in
two orthogonal components of drag pressure. three orthogonal directions. Each axis of the
The gage was intended for use in blast fields force gage had its own natural frequency. The
from surface-burst or low-height-of-burst frequencies were limited by the mass of the
sources where the exact direction of travel of moving parts of the gage and the spring
the shock front is not known precisely. Tubes constants required to allow this mass to move
with fundamental frequencies of 500 Hz and only as far as necessary to generate the
1000 Hz were tested, for use with blast waves required electric signal. These frequencies
of 10 msec or greater duration. A typical ranged from 85 Hz to 550 Hz. This relatively
trace recorded in a shock tube is shown in low frequency response prohibits the use of
Fig. 7-22. these gages for measuring short-duration dif
fraction forces; hence, their usefulness was
Various other types of drag gages have been limited to the long-duration drag phase of the
made for nuclear field tests or tests with large shock wave interaction.
7-18
AMCP 706-181
75 DENSITY GAGE
PROJECT ION OF
SOLID MOUNTING
CYLINDER
A blast wave density gage has been devel
oped by Dewey and Anson' 5 for use in
large-scale field experiments. The general con
Figure 723. SRI Total Drag Probe Assembly
figuration is shown in Fig. 7-25. It consists
Schematic Cross Section 8
essentially of two 4-ft X 2-ft X 3.5-in.
aluminum sections, each rigidly bolted to a
ing from 50 to 500 psi. The natural frequency 0.5-in. steel base plate. The leading and top
of these gages varied from 4 kHz to 5.5 kHz. edges of the sections are bevelled to knife
edges and the inner surfaces are plane so that,
7-4.2.4 BRL BIAXIAL DRAG GAGE when the gage is aligned with the charge
center, the blast wave receives little distortion
BRL has constructed a biaxial drag gage for as it passes between the sections. One section
measuring the magnitude and direction of contains a l3-source and the other a detector
dynamic blast pressure. The sensing element is unit consisting of a phosphor scintillator, a
a load cell that senses forces in two cross axes photomultiplier, and an amplifier. The spac
in a target area that behaves approximately ing between the two sections may be varied.
like a section of a cylinder of infinite length. When a blast wave passes between the two
Fig. 7-24 is an assembly drawing of the sections, the increase of air density causes a
completed probe. The entire surface of the greater absorption of l3-particles producing a
TOP VIEW
--DRAG CUP
BIAXIAL LOAD CELL
- - CENTER SECTION 2
6-3'4 in.
3
6
BASE CONE 4
L
1-112 in.
I
l~ I
MOUNTING BASE
I-SOLID CAST ALUMINUM SECTIONS; 2-13-S0URCE AND LOCKING
L3in. d -J
ASSEMBLY; 3-DETECTOR; 4-BAmRIES; 5-SIDE COVER PLATES;
6-POWER SUPPLY SWITCHES AND OUTPUT CONNECTORS; 7-STEEl
BASE PLATt.
Figure 7-24. Assembly Drawing of BRL Figure 7-25. Diagram of Density Gage
Biaxial Drag Gage 8 of Dewey and Anson' 5
.! 7-19
AMCP 706-181
change of the detector output. This signal is more useful in shock tube studies than in field
amplified and transmitted via several thou experiments.
sand feet of coaxial cable to an instrumenta
tion bunker where it is recorded as a fre 7-6 IMPULSE TRANSDUCERS
quency-modulated signal on magnetic tape.
7-6.1 FREE PLUG TRANSDUCER
Calibration of the gage can be achieved
easily by inserting foils of aluminum or Mylar In measuring the time histories of pressure
of known mass per unit area between the in reflected air blast waves with piezoelectric
source and the detector, and recording the transducers, investigators at BRL found that
output of the detector. In practice, a disc the limits for satisfactory function of these
with inserts of materials of different areal gages were in the range of several thousand
density is rotated through the beam path to psi. To determine at least some blast param
provide multistep calibration. eter accurately at high overpressure levels,
they then developed a free plug transducer for
A typical trace of density in a blast wave measurement of reflected impulse I 6. This
from a large chemical energy source is shown device consists simply of a cylindrical plug
in Fig. 7-26. The "noisiness" is inherent in that is lightly held in a hole in a large, rigid
this type of gage, because the output of the plate and is accelerated by a normally-reflect
scintillator detector consists of a series of ed blast wave. Measurement of the plug
discrete bursts rather than a continuous sig velocity after blast wave passage, and knowl
nal. edge of its area presented to the blast front
and its mass allows determination of the
This ingenious gage is, as far as the authors reflected impulse from the impulse-momen
are aware, the only successful one for contin tum theorem. In use, the plug velocity is
uous analog recording of time histories of measured either by photographing its flight
density in blast waves. Some optical methods against a scale background with an accurately
exist, as discussed in Chapter 9, but they are timed motion picture camera, or by some
7-20
AMCP 706-181
other means of accurately measuring time of mass and spring strength are adjusted so that
travel over a known base line. The "flying the natural period of the mass is about four
plug" has been used under simulated altitude times the positive duration of the blast. Under
conditions 1 7 as well as sea level conditions. these conditions, the maximum compression
This simple method is quite accurate and of the spring is a measure of the positive
precise, provided one adjusts the plug mass so impulse. Again, no data are given, so the
that little motion occurs during the total usefulness of such an impulse transducer is
duration of the blast wave. Measurements are questionable.
made routinely at very small scaled distances
(down to 6 in. from the center of I-Ib 7-7 VARIOUS MECHANICAL GAGES
explosive charges of Pentolite), where piezo
electric transducers have been either erratic or Throughout the history of air blast testing,
useless. there have been sporadic efforts to replace the
complex instrumentation usually required to
7-6.2 SLIDING PISTON GAGE measure blast parameters with simple mechan
ical gages requiring either no or very unsophis
In theory, a modification of the technique ticated recording equipment. One must sacri
described in par. 7-6.1 for measuring reflected fice exact knowledge of the complete time
impulse should prove adaptable for measure history of pressure, etc., with such simple
ment of side-on impulse. One such device is devices, and be content with estimates of
described by Kennedy 1 l\ as being used at peak overpressure alone, or peak drag pres
Underwater Explosions Research Laboratory sure, or drag impulse, or perhaps only an
during World War II. The gage was described effective equivalent explosive charge energy.
as having a freely-sliding piston, and being Balancing this disadvantage is the extreme
provided with a rotating drum carrying re simplicity and relative cheapness of mechan
cording paper on which a stylus attached to ical devices-one can easily emplace dozens or
the piston writes. The resulting record is a hundreds of properly calibrated mechanical
plot of the integral of impulse versus time. gages during a field test.
Thus, the impulse at any time is proportional
to the slope of the curve at that time, and the 77.1 DEFORMATION GAGES
positive impulse is proportional to the maxi
mum (positive) slope of the curve. The gage The simplest possible gages are those per
records the negative impulse as well. In manently deformed by the blast wave. Many
practice, this device apparently was much less such devices have been used by both U. S. and
accurate than integration of time histories British investigators. Ref. 18 includes descrip
from side-on pressure transducers, since no tions of a number of such devices used prior
appreciable amount of data appears to have to and during World War II. Examples of this
been generated with it. Other attempts at group are two types of gage used by the
BRL for measurement of side-on impulse Research Department, Woolwich, England, up
using relatively simple transducers that would to the beginning of World War II, both of
mechanically integrate the pressure-time his which gave an empirical estimate of blast.
tory also has proven abortive. They were the "foil gage" and the "cylinder
gage". In both the deformation of copper
76.3 SPR ING PISTON GAGE discs under the action of the blast wave was
measured.
For measurement of blast intensities from
charges of moderate size for which the posi The foil gage consisted of thin annealed
tive durations encountered are not extremely copper discs rigidly clamped round their
long, Kennedylll claims that a spring piston periphery over holes in a steel plate. The gages
gage is capable of precise measurement of were calibrated by clamping simple discs over
positive impulse. For this purpose, the piston the open end of a cylinder connected to an air
i 7-21
AMCP 706-181
supply and measuring the deflection as the air Calibration was achieved in the same manner
pressure was increased progressively. In the as for the British "flags". Typical calibration
field the discs were exposed face-on to the curves from Ref. 20 are shown in Fig. 7-27.
blast, and from their deformation the so
called "equivalent static pressure" in the Exposing a series of stiff vertical wires,
blast was measured by reference to the mounted as cantilevers, has also been tried by
calibration curve. the British. Under blast loading a given wire
will be bent through an angle which at any
The larger and much thicker discs of the particular distance is a function of the drag
cylinder gage were clamped on to form the force experienced and therefore of the weight
opposite airtight ends of a steel cylinder some or energy of explosive detonated. This func
6 in. in diameter and I0 in. long. These gages tion can be determined by experiment. The
were placed with the axes of the cylinders at circular cross section of the wires makes gages
right angles to the direction of the blast wave; of this type omnidirectional, thus they can be
the discs were therefore exposed approxi used to indicate asymmetry of the blast wave.
mately side-on to the blast. Their deformation They also have been used in atomic weapon
was used as a comparative measure of the tests to measure the dynamic pressure in the
blast from different charges. blast wave.
Another simple blast meter of the deforma Another use of very simple gages occurred
tion type used in England consists of a during an early atomic weapon test at I3ikini.
number of aluminum strips, of different Sir William Penney, who was present as an
thicknesses, clamped at their centers to a steel observer, deployed around the test site a large
post to form a series of double cantilever number of empty gasoline tins; these tins
beams. These blast "flags" have been used to deformed to various degrees by the blast
determine the high-explosive equivalent of wave. He was able to estimate the peak
propellant explosions. They are calibrated by pressure to which the tins had been subjected
exposing sample strips to the blast from by measuring the change in internal volume
known weights of explosive and determining which each had sustained. In this way a
the relation, for each thickness, among the considerable amount of data was gained at a
charge weight, the distance from the charge to trial in which some of the more sophisticated
the cantilever in question, and the angle of methods of measurement failed. One-gallon
deflection of the cantilever. The deflection of empty varnish cans were later used in the
a given strip decreases rapidly with distance same manner by other investigators 2 I to
from a charge and the device is also reported compare the relative blast effectiveness of
to be subject to a scale effect; as the charge conventional explosives. The advent of atomic
weight was increased a disproportionate in weapons in fact renewed interest in mechan
crease in deflection was observed. Wind velo ical gages. There were two main reasons for
city can also affect the deflection of the strip. this. Firstly, the duration of an atomic blast
The calibration requires care, and the use of wave was so long that the use of mechanical
blast "flags" must be restricted to the range systems for accurate measurements became
of charge weights for which they are cali feasible, despite their inherent low frequency
brated. Their main virtue. as with many response. Secondly, the electromagnetic radia
simple mechanical gages, is that their com tion emitted by nuclear devices interfered
parative cheapness permits them to be used in with the use of piezoelectric systems.
sufficiently large numbers for the results to be
treated statistically and so to yield significant 7-7.2 PEAK PRESSURE GAGES
answers. A U. S. variant of this same type of
blast gage} 9 employs single aluminum canti Peak-pressure gages have been devised to
lever beams clamped in simple vises that are operate on the principle that a thin dia
mounted on relatively massive base plates. phragm, stretched over a hole in a rigid plate,
7-22
AMCP 70~181
will rupture at a certain pressure when the or equal to, the pressure required to break the
diaphragm is subjected to a blast wave. If diaphragm over the smallest hole broken. In
several such diaphragms are provided, cover theory, the pressure in thus bracketed quite
ing holes of various sizes, the pressure re closely, simply by having a sufficient number
quired to rupture a diaphragm over a given of holes of graduated sizes.
hole will depend on the hole size. Hence,
given a calibration of the device, the peak The first device of this type was apparently
pressure of a blast wave is established as less a "paper blast meter". It consisted of two
than that required to break the diaphragm of boards clamped together with a sheet of paper
the largest hole unbroken, and greater than, held tightly between them. Holes of about ten
I I I I I I I I 1
I I I I I
l/8lb m l/2lb rn l Ibm31b m8 Ib
I ~I r 38 Ibm 216 Ibm 1000 Ibm 10,000 Ibm
~ ~ ~
, ,
,
000. b
10 \
I
\ ~ \ \ ~
I
1\ 0\ ~ ~ \ g ~
f C'\
\ \ \
o~ ~
I
1 \ B
b
\
I I
1
~
\
\
c
f- o~
z
o
f
\,
P
f-
U
f-
u..J
C
--I
u
u..J
Cl
0... )
f-
0.1 f-
l-
f-
f-
I I I I I I 1 I I I 1
0.01
I 10 100 1000
CHARGE DISTANCE, ft
Figure 7-27. Permanent Tip Deflection
of 0.051-in. 6061 Aluminum Alloy Beam
vs Distance for Spherical
Pentolite or TNT
7-23
AMCP 7~181
different sizes were bored through both Suffield Experiment Station, Canada. One is
boards, in register. The gage was mounted termed a "surface tension gage", and employs
with the plane of the diaphragm face-on to the principle that the pressure required to
the wave. break a surface film of a liquid over an orifice
is directly proportional to the surface tension
A later modification of this gage was the of the liquid and inversely proportional to the
foilmeter, or "Bikini gage", which consisted diameter of the orifice. As reported in Ref. 23
of a wooden or metal box with one open end and shown in Fig. 7-29, the gage developed on
over which was clamped an assembly similar this principle consisted of a can filled with a
to the paper blast meter but with aluminum very dilute acid to a level above that of a
foil instead of paper. Foil was used because it series of orifices of different diameters. On
is much less sensitive than paper to changes in application of pressure to the surface of the
atmospheric conditions such as temperature liquid, surface films over some orifices would
and humidity. The box gage could be oriented break, ejecting the liquid. Detection was
either face-on or side-on to the direction of recorded by litmus paper located beneath
propagation of the blast, since the box pre each orifice. Thus, the peak pressure could be
vented blast from acting on the reverse side of bracketed in the same manner as for paper
the diaphragm. The great advantage of this blast meters or Bikini gages. The authors
type of peak-pressure gage was its simplicity. claim a response time of 3 msec and accuracy
The operation and the interpretation of re to within 0.0 1 psi in the range of 0.0 15 to
sults were simple, and no elaborate machine 0.15 psi, using orifices ranging from 8 to 60
work was involved. Its great limitation was mils diameter. They also state that a modified
that the precision of results was not high, and version using mercury as a liquid and catching
the limits within which the pressure could be trays beneath each orifice could be used to
bracketed with a reasonable number of holes measure peak pressures in the range 0.1 to 1.2
were rather wide. Such gages have been used psi, with a response time of 10 msec. The
on both conventional and nuclear explosion second type of gage was termed a "squirt
tests. gage". It is shown schematically in Fig. 7-29,
and described in Ref. 24. This gage functions
Two types of simple peak pressure gages on the principles that the velocity of stream
that employ liquids have been developed at line flow in a tube is directly proportional to
PRESSURE
APPLICATION
WATER
LEVEL
LITMUS
(A) Schematic of Surface Tension (B) Use of Litmus Paper to Record Ejection
Pressure Gage of Acidified liqUid
Figure 7-28. Surface Tension Blast
Pressure Gage of Muirhead and McMurtry 2 3
7-24
AMCP 706-181
the pressure causing the flow, and that the simple instrumentation of low frequency re
distance which a horizontally-ejected jet sponse. Peak bending strains of slender canti
travels before falling to earth is a unique lever beams have been recorded in this man
function of its velocity. In practice, the device ner, on direct writing oscillograph systems
consisted of an ink reservoir, a nozzle ar with frequency response limited to about 100
ranged for horizontal ejection of ink, and a Hz, and correlated with blast source en
"catching" material located on a horizontal ergy 19, 20 Because the beams only are de
plane below the nozzle. Everything but the formed elastically, they can be reused for
upper surface of the ink reservoir was pro many tests, rather than being replaced as must
tected from the overpressure in a blast wave. all devices that are permanently deformed.
A typical calibration curve is shown in Fig. For this simple device, it is possible to predict
7-29. The authors claim a response time of analytically the dynamic response under blast
less than 4 msec. They also note that the loading20 so that the specific blast parameters
device is temperature-dependent, because of that affect the gage response can be iden
change in viscosity of the ink. Both of the tified.
devices described are limited in use to
measurement of peak pressures for relatively One or more of the types of mechanical
long-duration blast waves, i.e., those whose gages described here can prove to be quite
positive phases are several times as long as the useful, particularly in large-scale field tests,
quoted gage response times. but should be used always with caution and
the knowledge that they yield only partial
The ranges of some of the mechanical and sometimes misleading information about
devices can be extended by use of relatively the blast wave characteristics.
50
40
INK "CATCHING
MATERIAL"
E
u
UJ
U
Z
<C
30
20
-rrrtT
t . :,
,-: I
I
I-
-
t.n fi
0
10
,.
0
.
INK RESERVOIR
0 20 40 60
AIR-TIGHT CONTAINER
PRESSURE, kN/m 2
(A) DIAGRAM OF THE SQUIRT GAGE (8) COMPOS ITE RECORD OF TEN GAGE READ INGS
7-25
AMCP 706-181
TABLE 7-2
CHARACTERISTICS OF SIOEON PRESSURE TRANSDUCERS
Aspect I
Ratio Sensing Temp. Gage Gage
Sensing Type of Overall Element Sensi Pressure High Freq* Accel. Sensi Resis Capaci
Trans Prin of Hous Length, Length, tivity Range, Cutoff, Sensi tivity. tance, tance,
ducer ciple Element Shape ing in. in. pC/psi psi kHz tivity %fF ohm pF
-
Susque piezo lead pencil 0.055 16.0 0.188 20.0 0.1500 250 'Unknown 0.1 10' (l 150
hanna electric metanio
Instr. bate
Model
ST-7
Celesce piezo' lead pencil 0.072 5.20 0.125 610 0.1500 >120 Unknown Un 10" 1750
(Atlan electric zirconate known
tic titanate
Res.
Corp.)
LC-13
Celesca piezo lead pencil 0.063 10.0 0.25 3150 0.01-500 >67 Unknown 0.25 2.5 x 4500
(Atlan electric zirconate 109
tic titanate
Res.
Corp.)
LC-3~,
SRL piezo tour pancake 0.10 18 0.25 Varies 5-500 12-70 Unknown Varies 10" Varies
Pan electric maline, 0_20 1.00 with
cake lead type
zirconate, of ele
etc. ment
SwRI piezo lead pancake 0.063 4.2 0.125 20 0.0150 120 Unknown Un' 10" 350
Pan electric zirconate known
cake
British
H3
piezo quartz
electric
hatchet 0.082 12 1.00 100 0.1300 60 Unknown UnI
known
10 10 Unknown
*The low frequency cut off for all gages in this table is a function of the input characteristics of the first amplifier seen by the gage, be
cause all gages are capactive devices. '
7-26
TABLE 7-3
o
Sus piezo' lead 0.210 20.0 0.1-500 * 250 Un 150 10 10 1%/10 C Unknown
que- electric metan pC/psi known
hanna iobate
Instr.
Model
ST-2
* 10
Sus piezo lead 0.375 100.0 0.1-100 100 Un Un 10 Unknown Unknown
que- electric zircon- pC/psi known known
hanna ate
Instr.
Model
ST-3
Sus-
que
hanna
Instr.
Mo piezo' tour 0.210 0.05 10-2000 * 1500 Un 10 10 10 Unknown Unknown
del electric maline known
ST-4,
2000
I
IV
00 TABLE 7-3 (Cont.)
"V
.....
Dia. of Gage Gage ~
...
Sensing Sensing
Trans- Sensing Ele- Surface, Sensi-
Pressure Low Freq. High Freq. Accel.
Range, Cut-off, Cut-off, Sensi
Capaci
tance,
Resis
tance,
Temp.
Sensi
Operating
Temp. Range,
...
OIl
ducer Principle ment in. tivity psi Hz kHz tivity pF ohm tivity of
Celesco piezo lead 0.375 35 t - 3000 * 350 Un 240 5 x 108 Unknown -40to 225
(Attan electric zircon- pC/psi known
tic ate
Res. tita-
Corp.) nate
Model
LD-25
Celesco piezo lead 0.375 1200 - 200 * 50 0.012 4700 2.5 x 10 9 0.26%/oF - 40 to 225
(Atlan electric zircon- pC/psi 0.019
tic ate psi/g
Res. tita-
Corp.) nate
Model
LC-60
& LC
65
Celesco piezo lead 0.208 120 -300 * 150 0.013 800 2.5 x 0.14%/ - 40 to
(Atlan electric zircon- pC/psi 0.013 10 9 F 225
tic ate psi/g
Res. tita-
Corp.) nate
Model
LC-70
& LC
71
*Function of amplifiers
Kistler piezo crystal 0.25 1.0 10-3000 * 130 0.01 5 10 14 0.01%/ -450 to
Instr. electric line pC/psi psi/g o F 500
Corp. quartz
Model
601A
Kistler piezo crystal 0.22 0.35 - 3,000 * 400 0.001 20 10 13 0.01%/ - 450 to
Instr. electric line pC/psi psi/g o F 500
Corp. quartz
Model
603A
--
-...J
~
I
N co
1.0
~
-.J
I
W
o ."
TAB L E 7-3 (Cont.)
......
Dia. of
......~
00
Sensing Gage Temp.
Sensing Sur Pressure Low Freq. High Freq. Accel. Capaci Resis Temp. Operating
Trans Sensing Ele face, Sensi Range, Cut-off, Cutoff, Sensi tance, tance Sensi Temp. Range,
psi Hz kHz tivity pF ohm tivity of
ducer Principle ment in. tivity
CEC strain dia 0.500 0.33 to 0-100 0 8 Poor - 350 0.01%/ -65to 165
Model gage phragm 0.0133 to of
Type & mV/psi 0-5000
4-311 strain
gage
CEC strain dia 0.625 2.56 to 0-26 to 0 6-15 Poor - 350 0.01%/ -65 to 250
Model gage phragm 0.0133 0-5000 oF
Type & strain mV/
4-313 gage psi
for
max
pres
sure
Schae strain dia 0.075 5 mV/ 0-2 to 0 60 0.075 to - 900 0.03%/ + 30 to 180
vitz- gage phragm psi 0-25 0.006% oF
~
Bytrex and g
Mod. strain "tI
...'"~
HFD _u
gage ... . .~ .. .- . ....
. --',".-'~.-, .. ,._.,_._-~"
-.l
...
I
W 00
.,~
-J
W
N
.....
~
...
...
CD
Schae strain dia 0.125 10 mV/ 0-2 to 0 20 0.0525 - 900 0.03%/ + 30 to 180
vitz gage phragm psi 0-25 to F
Bytrex and 0.0042%/
Model strain g
HFK gage
Schae strain dia 0.250 200mV 0-2000 to 0 100 0.0003 - 450 0.01%/ - 65 to 300
vitz gage phragm for to to O.OOOl%/g F
Bytrex and max. 0-20,000
Model strain pres-
HFM gage sure
AMCP 706-181
REFERENCES
7-33
AMCP 706-181
7-34
AMCP 706-181
CHAPTER 8
INSTRUMENTATION SYSTEMS
SEQUENCE
TIMER
I
I-------------~------------------I
I ELECTR ICAL
I CALI BRATOR
I
CRT
I
OSCILLOSCOPE
I
I Jt=e:m::::::::J-"
I BLAST PREAMPLI FI ER!
I WAVE AMPLIFIER
I COAXIAL
I CABLE TIMING
L OSCILLOGRAPH CAMERA
-------------------------------~
Figure 8-1. Block Diagram of CRT Oscillo
scope Recording System
---------------------- I R-l
AMCP 706-181
Equipment
A CRT recording system that has been in replaced by another CRT system that is
use for a number of years at the Ballistic assembled in eight-channel units. This system
Research Laboratories (BRL) consists of four was designed by B. Soroka and G. T. Wat
channel units incorporating all of the ele son 2 It is built around commercially available
ments shown in Fig 8-2. Frequency response oscilloscopes and drum cameras and is shown
of the system is somewhat dependent on in Figs. 8-3 and 8-4, and in block diagram in
amplifier gain, but is at least flat from 0 - 100 Fig. 8-5. Frequency response for the system is
kHz. Film speed of the 35-mm camera is 0.03 Hz to 250 kHz. The drum camera can be
adjustable up to a maximum of 100 ft/sec, operated at film speeds ranging from 0.024 to
providing a resolution of up to 1.2 in./msec. 2.5 in./msec, recording all eight channels on a
All circuits are switched on and off in proper 12-cm wide film strip. The system also can be
sequence, including imposition of four-step used as an eight-channel, single-sweep record
calibrations prior to charge detonation, by the ing system using Polaroid cameras. It is much
sequence timer. The system is designed for more versatile than the earlier BRL system
use with piezoelectric transducers and cannot because it can accept signals from a variety of
be used with other types. The only commerci types of transducers-including piezoelectric,
ally available component in the system is the potentiometer, thermocouples, and strain
oscillograph camera. gages. As in the older system, an integral part
is a sequence timer that automatically se
Although a number of the units of the type quences all events once the timer is started. It
described are still in use, they are being is much more compact per channel, occupy
8-2
AMCP 706-181
8-3
AMCP 706-181
ing less space than the basic four-channel unit The British system includes all of the other
of the older system. peripheral equipment for internal calibration,
sequencing, preamplifying, etc., described in
the BRL system.
8-2.1.2 THE CEC TYPE 5-140 CRT SYS
TEM
8-2.1.4 THE DENVER RESEARCH INSTI
Another multichannel system that still is TUTE CRT SYSTEM
used widely in blast recording is the Con
solidated Electrodynamics Corporation (CEC) Denver Research Institute operates a multi
Type 5-140. Unfortunately, this versatile and channel CRT system for air blast recording,
relatively compact system is no longer man which was built entirely by personnel at that
ufactured. It is shown in Fig. 8-6. The heart Institute. The equipment and its design and
of the system consists of eight dual-beam operation are described in Ref. 3. It is built
oscilloscopes that are arranged below an into a semi-trailer which houses the twelve
optical system and camera magazine, as channel system, a dark room, and work area.
shown in the left of the figure (the oscillo Frequency response is 0 to 80 kHz, and
scopes are opened for adjustment-in opera recording is on drum cameras with mm speeds
tion, their faces are out of sight below the from 1 to 5 m/sec. Sequence timers, calibra
optical system). The optical system refocusses tion units, amplifiers, etc., are included in the
8-4
AMCP7~181
SHORTING I RL-4 I
RELAY ~
TEKTRONIX RM561 SCOPE
GAGE
..
CATHO E
FOLLONER
~
-
"" 2B63
AMPLIFIER
0 2B6I6: TIME
l..o
r-.
RL-I
RL-3
EXPLOSIVE
- ""- BASE ~
~RGE
TIME OF
~ EVENT
TRANSDUCER
Ito TIMING LIGHT
I""
UNIT
TIM~
CALI BRATE
CAPACITOR
UNIT LIGHTS
..
~
..
CALI BRATORI FIDUCIAL
....
UNIT
MARKER
~ UNIT N.O.
6 M~CRO
.. I SHUTTER
SLIDE) S ITCt!..
I N0 C
~ONTACTS ~ ..
- SLIDEI
~ FLASH +25 I
UNIT V J
/I .,. RUM CONTACTS
DRUM CAMERA
SOUTHERN INSTRUM~~TS
MODEL M906
, !
lfI SEQUENCE
TIMER SHORT RELEASE "' PRE-TIMER BY-PASS
-25 ~II U
N.O. .. i
F'R'Nck V ~~
. .
IDELAY
.. ... ~
~RT ;
,
- UNIT ~
I~P~~~?T~ PRE -TIMER
CONTROL UN IT
BRL Recorder 2
AMCP 706-181
system in much the same manner as in the As in other systems described here, timing
BRL and British systems. The system is circuitry, calibration, etc., are bullt in as part
intended for use exclusively with piezoelectric of the system.
transducers.
8-2.1.6 OTHER CRT SYSTEMS
8-2.1.5 THE LANGLEY RESEARCH CEN Other agencies, such as the U.S. Naval
TER CRT SYSTEM Ordnance Laboratory and the Naval Ship
Research and Development Center, are known
Morton and Patterson 4 at Langley Re to have and use multichannel CRT systems
search Center of NASA designed and built a for blast recording, but the author could find
fourteen-channel CRT air blast recording no explicit descriptions of their systems.
system that was installed in an instrumenta
tion shelter at Wallops Island, Virginia. This 8-2.2 MAGNETIC TAPE SYSTEMS
system uses seven commercial dual-channel
oscilloscopes, but in all other respects was Blast recording systems built around mag
designed and built by NASA personnel. It is netic tape recorders are now as popular or
shown in Fig. 8-7. The system was designed more popular than CRT systems. The heart of
specifically for use with the NASA miniature such systems are multichannel, instrument
gages described in Chapter 7. A carrier ampli grade tape recorders made by several different
fier system with 120 kHz carrier provides manufacturers. The recorders are usually
system response flat from 0 to 20 kHz, with either 7-ehannel or l4-channel units, and
recording over long cable lengths. Traces are commonly employ FM signal electronics.
recorded on seven homemade drum cameras. Such systems with a frequency response of
, :i
8-6
AMCP 706-181
0-10 kHz probably were first employed for The two systems differ in minor ways, with
blast recording during nuclear tests shortly the primary difference being that the Sw RI
after World War II. Similar recorders using FM system (Fig. 8-8) employs a single input
electronics with twice this frequency range amplifier while the IITRI system (Fig. 8-9)
have been available for some years, and uses two stages of such amplification. Both
currently three manufacturers can supply systems are basically fourteen channel and
such recorders with response of about have an ancillary capability for several chan
0-400 kHz. nels of CRT recording. Eventual readout is on
photographic paper in the galvanometer oscil
The basic elements of such systems are lograph, played back at much lower tape
included in block diagrams for two magnetic recorder speed to avoid limitation of fre
tape recorder systems discussed in Refs. 6 and quency response due to the lower response of
7, reproduced here as Figs. 8-8 and 8-9, the galvanometer oscillograph and to expand
respectively. The system of Fig. 8-8 installed the time scale.
in an instrument trailer is shown in Fig. 8-10.
The basic elements of these systems are: The primary advantage of systems of this
type for recording blast data are (I) the data
(1) Transducer and cable are stored on the tape and therefore can be
retrieved at any time, (2) good time correla
(2) Zero time circuit tion between channels is always possible, (3)
the system employs commercially available
(3) Input amplifier or amplifiers components almost entirely, and (4) it can be
used for recording other types of transient
(4) Tape recorder with FM electronics data simultaneously or with some minor
changes. "Bad" data can often be recovered
(5) Output amplifiers by playback through suitable band-pass fIl
ters. Disadvantages are somewhat poorer fre
(6) Galvanometer oscillograph. quency response than CRT systems and dif
8-7
AMCP 706-181
2 CHANNELS
I ST-4 ~ SCOPE
&
CAMERA
TIME 0
MARK CIRe.
KISTLER
KISTLER 566
603A ,601 H CHARGE AMPLIFIER
PRESSURE TRANSDUCER
SANBORN IlTRI AMPEX CP-IOO
8875 A MONITOR MAGNETIC TAPE
D. C. AMPLIFIER UNIT REC /REP.
DYMEC
2401 B
PHOTOCON INTEGRATING DIGTAL VOLTMETER
PHOTO CON DG 605 HEWLETT-PACKARD
752A DYNAGAGE 200A TEKTRONIX 502
PRESSURE TRANSDUCER
OSCILLATOR OSCILLOSCOPE
oi------,r-- Y
IONIZATION PROBE
o..--E C. E.C.
5-124 OSCI LLOGRAPH
(7-363 GALVONOMETERS)
SANBORN -8875A
G
DC AMPLIFIERS :
Figure 8-10. Magnetic Tape Recorder System for Air Blast Recording
i
(Courtesy of South lIVest Research Institute) !
I
ficulty of incorporation of automatic elec graphs are used for analog recording of
trical calibration in a firing sequence without magnetic tape systems, using low playback
construction of special circuitry. speeds to overcome their relatively low fre
quency response. For recording of blast data
The two examples of ground-based mag from nuclear or very large chemical explo
netic oscillograph systems described here are sions, the frequency response can be ade
indicative of how such systems are usually quate, and systems based on direct recording
arranged. on galvanometer oscillographs are used ..
8-9
AMCP 706-181
D will record static and dynamic outputs blast-recording systems discussed earlier in
between 0 and 600 Hz. It uses an amplitude this chapter have short total recording times
modulated, suppressed-carrier signal, with the and can be used only if the time of explosion
amplified gage signal transmitted to an oscillo is known and controlled within milliseconds.
graphic recorder. The system may be used Even magnetic tape recorders and oscillograph
with any type of two- or four-arm bridge recorders, which have relatively long record
transducers operating on the resistance change ing times, cannot be used efficiently if they
Qr variable reluctance principle. A signal of must run for many minutes to record an event
1 mV will cause a full-scale deflection. At lasting milliseconds.
tenuators enable the system to operate with
input signals in the range of + 1 V to -1 V. The A recent (1970) addition to instrumenta
system includes an oscillator power supply for tion systems designed to solve this problem is
sensor excitation with an output of 10 Vat 3 a line of transient recorders manufactured by
kHz, an attenuator to vary the input signal Biomation. These devices accept signals from
levels, an amplifier to boost low-signal levels, transducers when initiated by external trig
and a phase-sensitive demodulator to provide gers, or when triggered by the transient signal
correct polarity to the signal output. Under itself, with no loss of the initial portion of a
the condition of zero stress on the sensor, the record. They store the signal digitally in an
output signal amplitude is zero. The signal is internal memory bank. The signal can be
amplified, transmitted, and admitted to the played back later in analog form on a CRT
demodulator, where the carrier is decoded oscilloscope, tape recorder, or oscillograph
and the proper sign and magnitude given to recorder; or in digital form to a digital
the signal. The basic "building blocks" of this recorder or directly to a computer for proces
system are groups of four carrier amplifiers sing. Amplitudes and time scales can be
and an oscillator-power supply capable of adjusted at will during playback. Recording
energizing up to twelve amplifiers. A compact times can be adjusted over wide ranges, 20 sec
twelve-channel unit can be mounted on a to 5 hr, depending on the expected event and
single shock-mount base. the particular recorder model. These recorders
have proven to be quite versatile and useful in
blast recording, and undoubtedly will be used
The second CEC carrier amplifier system is more widely in the future.
their Type 1-127, sometimes called "System
E". The CEC System E, like the System D,
operates on the AM-suppressed carrier princi
ple and functions in essentially the same way
8-2.5 INSTRUMENTATION PROBLEMS
as the System D; however, the System E uses
ASSOCIATED WITH NUCLEAR
a carrier frequency of 20,000 Hz, with a
BLAST TESTS
bandpass of 0 to 3000 Hz, which permits
recording of a much higher frequency from
Ground-based systems have been used to
the gage. The basic building block for this
record air blast data during many nuclear
system contains four channels of carrier am
weapon tests, and these systems are basically
plifiers, and the necessary power supply and
the same as those discussed previously in this
oscillator in a single compact unit.
chapter. There are some special problems
associated with such testing which are caused
by nuclear radiation and electromagnetic
8-2.4 TRANSIENT RECORDERS pulse effects associated with the nuclear
detonation, and by the necessity of using long
For some blast experimentation, such as recording cables, remote-operated recorders,
"cook-off' tests of ordnance items, the time etc. Many of these problems are reviewed in
at which the explosion occurs may vary by Ref. 9, and much of the discussion that
minutes from test to test. Many of the follows is taken from that reference.
8-10
AMCP70&181
Most transient effects result from the gen The gamma rays that enter the ground (or
eration of ion pairs in the system by the ocean) from a detonation slightly above the
incident radiation. These ion pairs ultimately surface are absorbed in a very short distance,
cause either photocurrents in transistors or a few meters at the most. Thus, over most of
diodes, or leakage currents in dielectrics. the distances where there are sizable Compton
*This EM signal is not unique to a nuclear detonation, but
currents in the air, there are none in the
can also be observed in large chemical explosions. ground. We thus have a hemispherical distri
, 8-11
AMCP 706-181
bution of Compton currents in the air. How mechanism generates very intense high-fre
ever, the ground is usually a better conductor quency EM fields and becomes increasingly
than the air (except very near the burst), so important as the burst altitude is increased.
that the current of conduction electrons,
instead of flowing radially inwards, will flow 8-2.5.2.4 EMP INTERACTION WITH
partly to and in the ground (Fig. 8-11). Thus SYSTEMS
current loops are formed, with Compton
electrons flowing outward in the air, and The EMP interaction with systems is singu
conduction electrons returning in the air and lar among nuclear weapons effects in that the
ground. These current loops give rise to a interaction is often with the configuration of
magnetic field, which is largest at the surface the entire system and not necessarily with any
of the ground, and which runs clockwise subsystem by itself. The complete system
azimuthally around the burst point. The forms an antenna that responds as a whole to
electric field is tilted near the ground so as to the EMP. Damage may occur at the gage, in
be roughly perpendicular to the ground, and the cable, or at the recording site.
is directed upwards so as to drive conduction
electrons into the ground.
8-2.5.2.4.1 GAGES
8-2.5.2.3 FREE AIR BURST The major problem results from transducer
inductance coils being short circuited. Dam
The previous paragraph discussed the fields age has not been significant with balanced
produced by the gamma-ray-induced Comp
reluctance gages. The most serious trouble has
ton recoil electrons, neglecting the effect of
been permanent grounding of one circuit by
the magnetic field of the earth. In all cases the
flashover, causing disturbances on other
asymmetries (ground, air, and bomb) were in
traces.
the gamma-ray flux, production of Compton
electrons, and ionization. The net electron
8-2.5.2.4.2 INDUCTION OF CURRENTS
motion was radial, and thus the source for the
INTO CABLES
EM fields was a pulse, a radial current
expanding with light speed from the burst The influence of the electric and magnetic
point. fields near the surface of the ground on
electrical conductors depends on the
In the presence of the geomagnetic field, configuration of the conductor. The manner
the Compton recoil electrons are deflected in which the conductor is coupled to the
from their initially radial directions. The electric field is affected by the presence or
current pulse then contains transverse as well absence of insulation, the type of insulation,
as radial components. Thus, even with com and the quality of contact between the
plete symmetry of gamma-ray flux and elec conductor and the soil. The effectiveness of
tron production, there are sources for mag shielded cables depends on these factors and
netic and nonradial electric fields. In fact, this the manner in which the shield is terminated.
In addition, the implications of a signal
induced on a conductor are determined
LOW-ENERGY~
ELECTRONS ~'f. __ COMPTON RECOIL largely by the sensitivity of the system served
~ _-----,::c- ELECTRON by the conductor. Thus, for example, a given
-~~---- / pulse may cause serious malfunction if it is
B I
BURST I
GROUND PLANE
induced in a circuit designed for low-level
signals, whereas the same pulse induced in a
power circuit would be of no consequence. A
more complete discussion of effects in insula
Figure 8-11. Influence of Ground on Return ted, bare, and shielded conductors is given in
Conduction Current 9 Ref. 9.
8-12
AMCP 706-181
i 8-13
AMCP 706-181
by systems based on small, rugged tape transport. The recording heads also have
recorders, which will be discussed next. provision for playback. A summary of the
recorder characteristics listed by the manu
8-3.2 MAGI\! ETIC TAPE RECORDER facturer is given in Table 8-1. Response time
SYSTEMS (5%) has been measured in the laboratory
using a square wave input and found to be
Prior to discussion of complete systems, let 100 J.J.sec, using a standard filter; with a
us first discuss the compact, multichannel gaussian filter the response was about 70 J.J.sec,
tape recorders that make these systems pos but the output was noisy.
sible. Two units have been used extensively in
blast recording, the Leach MTR-1200 and the In general, all data are recorded on wide
Genisco Data 10-110. band FM channels, which provide better
signal resolution than analog recording and
8-3.2.1 THE LEACH MTR-1200 RECORD also respond to DC signals. To reduce the
ER number of recorder channels required, data
can be multiplexed and recorded on analog
The Leach MTR-1200 recorder is a com channels but with some sacrifice in frequency
pact, rugged piece of equipment specifically response. However, the high power limit on
designed for use under extreme environmental frequency response for the analog channels
conditions such as those encountered in usually precludes blast recording for all but
rocket sled testing. It is a self-contained unit very short duration waves.
capable of recording fourteen channels of
wide band FM and/or direct analog informa 8-3.2.2 THE GENISCO DATA 10-110 RE
tion. The data signals are recorded on I-in. CORDER
magnetic instrumentation tape. Any of the
fourteen data channels may be used to record The Genisco Data 10-110 was designed for
self-generated 100-kHz time reference signals use in adverse environmental conditions. The
that in tum may be used for wow and flutter 10-110 is small (7 X lOX 12.5 in.), light
compensation. Two seven-track record heads weight (28lb), and portable. The system uses
and one erase head are provided on tape a unique Cobelt tape drive and transport
Total Harmonic
Distortion Less than 3% during analog recording
Reference Oscillator
Frequency 100 Hz 50 Hz
8-14
AMCP 706-181
8-15
AMCP 70&181
CAliBRATE
CALI BRATE
n-. SHAEVITZ-BYTREX
~ AMPLIFIER L.. . . .,;,S. . ;. U......
PP.. .; ;,LY---I y-
PO'NER t--~ ...SHAEVITZ-BYTREX L r
AMPLI FIER
L 22-32 VDC
BATTERY
SHAEVITZ
BYTREX
SHAEVITZ
PRESSURE
BYTREX
TRl\NSDUCER
PRESSURE
TRANSDUCER
for recording; and (2) an analog magnetic tape The DAQ-PAC also contains a programmer,
recording system for recording in both direct electromagnetic pulse (EMP) input circuit
and FM format per IRIG specification protection, and internal power supply. Upon
106-60. activation by external control signals, the
programmer automatically directs the DAQ
PAC through a series of operations including
All components of the system are plug-in pre-test warm-up and calibration, shorting
modules, so that a wide variety of transducers input lines for EMP protection, data measure
can be used to obtain magnetic tape recording ments recording, and post-test calibration.
without additional circuitry. After recording, Thus, the DAQ-PAC is a complete instrumen
the tape is recovered and played back on any tation system and requires no external sup
standard IRIG magnetic tape playback sys port other than the initial activation. The high
tem. A total of twelve data channels plus two shock and nuclear radiation resistance has
channels for flutter compensation and time been obtained by an all solid-state design and
reference data are provided. the careful selection of components.
8-16
AMCP 706-181
Temperature
The DAQ-PAC uses a modular construction been refined and further developed, and have
achieve a broad flexibility for measurement been used for field blast measurements for
purposes. A combination of modules or cards conventional explosive tests as well as nuclear
is provided which allows measurement of low tests.
frequency phenomena (DC-600 Hz) using a
3-kHz carrier for transducer excitation, or The design of these gages is described well
wide band linear (200-200,000 Hz) proces in Appendix A, Ref. 12, which will be
sing for applications requiring high frequency paraphrased here. The basic component of the
response. Wide band recording over the range system is a pressure-sensing capsule consisting
of DC to 20,000 Hz is also available as an of two concentrically convoluted diaphragms,
optional feature for those applications where nested together to reduce volume, and silver
both relatively high frequency response and soldered together around their periphery. In
DC levels must be recorded (see Table 8-3). brief, these capsules operate by an increase in
Tape speeds of 3.7S to 60 in. per sec can be outside air pressure entering through a small
provided, depending on the frequency re inlet hole, causing expansion of the dia
sponse requirements. The corresponding re phragms. A light spring stylus soldered to the
cording times range from 60 to 4 min, center of the free diaphragm transmits this
respectively. motion and produces a scratch on a coated
glass recording blank. The amplitude of this
8-3.3 SELFRECORDING GAGES scratch is proportional to the movement of
the diaphragm, which in turn is proportional
The last type of portable blast recording to the applied pressure. A sapphire-tipped
system that will be discussed differs markedly phonograph needle, with a O.S-mil radius tip
from the two previous systems. These latter soldered to the stylus arm, is used to insure a
systems, termed "self-recording gages", were very fine scratch. Ten ranges of capsules, from
developed originally by BRL in the 1950's for o to 1 psi to 0 to 400 psi, are in general use in
recording time histories of blast pressure from BRL self-recording gages. The basic specifica
nuclear explosions in air. They have since tions are:
8-17
AM~ 706-181
GENERAL CHARACTERISTICS
lined housings to record dynamic pressure or was restricted to a single axis by a Teflon
total pressure. The gage also incorporates a bushing in the support member. Refinements
simple system for remote actuation. Data followed rapidly. For example, a stainless
recorded with these gages are shown in Ref. steel flange was added to the assembly to
12. facilitate sealing the sensor to the gage hou
sing with an O-ring. Later, alignment pins
The design of the original BRL self-record were placed in the flange to orient the stylus
ing gages fonned the basis for an improved in proper relation to the recorder when
design by Bendix Corp., under contract to installed in the gage housing. Still later, the
BRL. The development of the improved separate support and flange were replaced
design is described by Wells l 3. This reference with a single stainless steel support disc
also contains a good history of improvements incorporating an a-ring groove, alignment
in the original BRL design for the 1953-1963 pins, and a jewel bearing to guide the stylus.
period that is presented here.
Some typical performance specifications
8-3.3.1 BLAST PRESSURE SENSORS for a series of 1963 model blast pressure
sensors were:
The basic element in blast pressure sensors
has been the metal diaphragm. The diaphragm Pressure range: 0-1000 psi (in steps such as
has been defined as a pressure responsive 0-1; 0-2; 0-5; ....0-100; etc.)
element that is movable in a direction sub
stantially perpendicular to its flexible surface. Motion for each range: 0.015 to 0.031 in.
The sensor of an early blast pressure gage with a mean value of about 0.022 in.
consisted of a capsule formed by two nested
corrugated diaphragm discs joined by a brazed Response time: Imsec or less except in the
circumferential seam. A mounting block, lowest pressure range
brazed to the center of one of the dia
phragms, provided support for the sensor and, Linearity: 0.5 to 5%
by means of a through hole, connected the
small sensing volume of the capsule with the Hysteresis: up to 1.1 %
atmosphere. A stylus was mounted on a flat
spring brazed to the center of the opposite Another aspect of sensor development con
diaphragm. Thus, the stylus motion was con cerns the techniques used to provide damping.
trolled directly by the motion of the dia The damper used in the early gages consisted
phragm in response to changes in sensed of an outer screen over a tabular cavity
pressure. While this sensor performed fairly leading to the sensing volume of the capsule.
well, the principal faults were the relatively Later, interchangeable aperture plates made it
slow rate of response and the tendency for possible to provide multiple holes of various
the unguided stylus to wander from a straight diameters for optimum damping in specific
line motion. During 1960 and 1961, shock environments.
tube tests at BRL indicated that single-dia
phragm sensors would provide higher rates of 8-3.3.2 TIME BASE
response to pressure changes. The motion of
the original single diaphragms (1-5/8 in. dia A reliable time base is an essential require
meter) was about one-half that of the capsule ment for analyzing the recorded data obtain
design; bu t, in general, the response time was ed from the self-recording instruments. At
less than 1 msec. In 1962, a new design blast least two methods have been used to measure
sensor based on the single diaphragm was time intervals on the recordings. One method,
introduced. This design used a similar 1.25-in. used on the earlier gages, employed a constant
O.D. Ni-Span C diaphragm, spot welded to a speed electric motor to drive a disc on which
formed Ni-Span C support. The stylus motion the test data were recorded. The time base
8-19
AMCP 706-181
was governed chronometrically to about 5%. initiation signal. The negator-spring powered
Power for the motor was supplied by self-con recorders accelerate more rapidly then electric
tained batteries. A second method, a time motor driven turntable models. Nevertheless,
marker, was developed for metal tape record the following special techniques have been
ing systems where there is an inherent varia developed to further reduce the spring-drawn
tion in the linear speed of the tape due to the recorder acceleration time to very short inter
change in radius as the tape unwinds from a vals:
constant speed spool. The early time markers
consisted of a solenoid-operated oscillator. (1) Closure of the relay (by any of the
The stylus scribed an oscillating timing trace initiating methods previously given) com
on the moving recording tape simultaneously pletes a circuit to a solenoid that releases a
with the test data traces. The nominal fre kick spring and accelerates the recorder to
quency of the timer trace was 50 Hz. The operating speed in an average of 18.4 msec.
principal faults in this time marker were its
low frequency and its sensitivity to shock. (2) When the solenoid in (1) was replaced
by an explosive piston actuator and fired
8-3.3.3 INITIATION METHODS electrically by the relay closure, acceleration
time was reduced to an estimated 10 to IS
Since the self-recording instruments have a msec.
limited operating period (about 20 sec), initia
tion of the recorder must precede arrival of (3) The acceleration time of the recorder
the event to be measured by as short a time was reduced to 5 msec or less by means of a
interval as possible, consistent with accelera rack and pinion starter, driven by an explosive
ting the recording medium (tape or disc) to piston actuator, fired directly from the initia
normal speed. Anum ber of methods have tion line.
been used by BRL to accomplish initiation,
such as: Based on the BRL developments up to
1963, requirements for an improved system
(l) The visible light output from the were developed and used as guidelines by
monitored detonation operated a self-record Bendix Corp. in their contract. These require
ing instrument photocell that closed a relay ments are listed in Table 8-4. Bendix develop
and started the recorder motor. ed a number of prototype units based on
these requirements. These units were sub
(2) The thermal radiation from a large-scale jected to a number of laboratory tests, as well
detonation melted low temperature solder in as field testing with a SaO-ton TNT blast
a thermal link on the self-recording instru source l 4. Based on their work, an improved
ment. The link supported a spring-loaded self-recording system emerged with capabili
plunger that, when released by the melting ties as listed in Table 8-5. Whether the
solder, closed a switch that started the record improved system has been produced in suf
er motor. ficient quantity for wide field use is not
known.
(3) Closure of an electrical switch in an
external circuit attached by wire to the 8-4 CALIBRATION TECHNIQUES
self-recording instrument closed a relay that
started the recorder motor. In air blast experimentation, some type of
calibration of transducers and/or recording
8-3.3.4 ACCELERATION METHODS systems usually is employed. Most blast test
ing involves measurement of times and pres
Closely associated with initiation methods sures. Calibration of time bases for oscillo
are means for rapidly accelerating the record scope sweeps, timing marks for moving film
ing medium to normal speed following the records, etc., is accomplished in a variety of
8-20
AMCP 706-181
4 in. diameter
6 in. length
6 Ib weight max
Error band: 5%
Hysteresis: 1%
0
Temperature range: _65 to + 165F operational
Radiation: 10 5 neutrons/cm 2
101 1 rad/sec
incremental steps
80 to 2000 Hz at 20g
Thermal pulse: 300 cal/cm 2 total dose seen at entrance to present port
Shock: 500g for 11 msec min, 1000g desired
ways. Usually, timing is compared with signals ing equipment can be calibrated with sub
from secondary standards such as crystal-con microsecond accuracy. No more specific
trolled oscillators or timing mark generators, methods will be discussed here, because the
counter chronographs, etc. Periodically, these methods are very dependent on the type of
secondary standards can be checked against recording system in use.
standard time signals broadcast by the Na
tional Bureau of Standards over Station Calibration of blast pressure transducers is
WWV. With care, time bases for blast record- accomplished in a number of ways, which are
8-21
AMCP 706-181
Sensor: Interchangeable assembly with G-ring seal, alignment dowel pins and integral reference stylus
Linearity: 5% max
Hysteresis: 1% max
Natural Frequency: Greater than 1 kHz except in ranges below 0-2 psi
Recorder: Negator-spring powered with separate metal recording tape. Nominal tape speed of 3
in./sec and 20 sec min running time. Start-up time with explosive piston actuator;
0.5 msec
Recording Tape: magnetic stainless steel, 3/8 in. width x 0.001 thick x 60 in. maximum
length
Time Base: Fluidic type time marker, nominal frequency 475 Hz 1% at constant temperature and
nominal 20 psi gas supply pressure
Initiation: External line (electrical) or gamma radiation of 1.5 x 104 radlsec or more
Physical Size: Gage: 4.75 in. diameter (flange) x 4.5 in. length. Fluid supply and Regulator: 4-11/16 x
2-11/16 x 1-314 in.
Mounting: Flange
Weight: 4.61b
Environmental:
0
Temperature: (Timer only) _65 to + 165F
Vibration: 10 to 80 Hz, 0.06-in. double amplitude and 80 to 2000 Hz at 209
Acceleration: 75 g tested
Shock: 100g, 10 msec. At shock levels of 300 to 500g, the pressure trace is subject to an error of 3
to 7%.
8-22
AMCP 706-181
transducers, a fixed resistor usually is parallel reservoir) a small chamber containing the
ed with the resistive gage element to generate transducer, or rapidly dump pressure from the
a voltage change. For piezoelectric trans small chamber. A schematic of one such
ducers, a step in electrical charge is generated system used for laboratory calibration 2 5 is
by switching into the gage circuit a known shown in Fig. 8-13. This particular system
capacitance applied to a standard voltage. includes a means of superimposing an elec
Often, electrical calibration will be accom trical calibration step on the pressure calibra
plished in an automatic sequence in blast tion record. Recording is accomplished by a
recording equipment prior to charge detona CRT oscilloscope. By multiple exposure of
tion. Calibration steps of several different film in an oscillograph record camera and
amplitudes usually are impressed on records. adjustment of initial oscilloscope sweep posi
in this manner. The accuracy of purely tion, multiple calibrations can be obtained
electrical calibrations of this type is totally easily on one piece of film, as shown in an
dependent on separate static or dynamic inset in Fig. 8-13. A similar system for field
calibration of the transducers, so that the step calibration is shown in Fig. 8-14. Here, an
changes in voltage can be correlated with Atlantic Model LC-33 transducer is shown in
equivalent pressure changes. For the electrical the calibration chamber. This particular sys
calibration method to be useful, the pressure tem is light and portable, and can be taken
transducers must be stable and preferably into the field to calibrate an entire blast
linear in their outputs. pressure recording system. Quasi-static pneu
matic systems can be used for calibration up
Static pressure calibration can be accom to several hundred psi.
plished with any means of applying static
pressure to a transducer and a number of
Dynamic calibration of pressure trans
methods of recording output. The applied
ducers has been accomplished in several ways.
pressure is measured by manometer, precision
Rathke 26 reports a clever adaptation of a
bourdon gage, or any of a number of other
simple drop test device for applying a known
means. Transducer output can be measured
half-sine pulse at very high pressure ampli
by digital or analog voltmeter, galvanometer
tudes (up to 20,000 psi). The pressure cham
oscillograph, etc. With suitable equipment for
ber and contained liquid deform elastically
accurate measurement of pressure and trans
under impact of the drop weight, acting
ducer output, those types of blast pressure
amazingly like a linear spring. By varying drop
transducers that have DC response often can
be calibrated accurately by static methods. In height and mass of the impacting weight, the
a static pressure calibration, one, of course, amplitude and duration of the pressure pulse
obtains no information on shock response of easily can be varied over wide ranges. A
a transducer. This method is useless for schematic of Rathke's apparatus and typicai
calibration of piezoelectric transducers be pressure traces are shown in Fig. 8-15. This
cause they have no DC response. system is suitable for calibration of all types
of flush-mounted blast pressure transducers.
Rathke used a common calibration technique
By quasi-static pressure calibration we here of comparison with a reference or
imply calibration by application of a pressure standard transducer, which presumably had
increase or decrease to a transducer in a been calibrated separately by some other
relatively long time (i.e., a number of millisec means.
onds) but a time short enough for adequate
response of a transducer and associated re
cording system having a finite electrical time The best method of dynamic calibration of
constant. Calibration systems of this type, air blast transducers is essentially a "boot
commonly used for piezoelectric transducers, strap" method. One subjects the transducer to
are pneumatic and usually employ solenoid a blast or shock wave itself and, by indepen
valves to rapidly fill (from a much larger dently measuring shock velocity or equiva
. 8-23
ELECTROM CRT
PIEZO
GAGE
f-------""-----j#----jlf----
-1 I- Ie msec
-, CAL.
STE P
" LOW-PRESSURE
TWO-WAY, SOLENOI D
ACTUATED VALVE RESERVOIR
FROM HIGH-PRESSURE
HELIUM CYLINDER
t
WITH REGULATOR
Figure 8-13. Schematic Diagram of Quasi
static Gage Calibration Apparatus
8-24
AMCP 706-181
L II
J L
STRA IN GAGE TRAN SDUCER
1 msec;
entire transducer and recording system is
calibrated under conditions best simulating
actual use. Gage "ringing" and other spurious
( E i TYPICAL PRESSURE TRACES
response will be evident here, when they will
Figure 8-15. Dynamic Pressure Calibrator of not be apparent under static or quasi-static
Rathke I 6 calibration.
REFERENCES
8-25
AMCP 706-181
Rohm & Haas Co., Huntsville, Ala., Oct. 13. H. S. Wells, Development and Test of
1967, (AD-822 133). Prototype Miniature, Rugged Self
Recording Air-Blast Instrumentation, Re
9. R. H. Rowland, Blast and Shock Measureport No. E.I.R. 700, The Bendix Corp.,
ments, State-of-the-Art Review, DASA Baltimore, Md., Nov. 1966.
1986, DASj AC Special Report 45, Aug.
1967, (AD-824 160).
14. D. P. LeFevre, Evaluation of New Self
Recording Air Blast Instrumentation:
10. R. L. Jarvis, et al., A Program to Induce
Project 1:36 Operation SNOWBALL,
High Blast-Induced Airloads and Struc
BRL Memorandum Report No. 1815,
tural Response of Lifting Surfaces, ASD
Jan. 1967.
TDR-63-764, Vol. II, July 1964,(AD-605
407).
15. W. E. Baker and W. O. Ewing, Jr.,
II. I. O. Wolf, et al., An Experimental Miniature Piezoelectric Gages for Measur
Investigation of Blast-Induced Airloads ing Transient Pressures on Airfoils, BRL
and Response of Lifting Surfaces, Memorandum Report No. 1329, March
AFFDL-TR-64-176, March 1965, (AD 1966.
465 721).
16. T. S. Rathke, A System for Dynamic
12. E. J. Bryant, N. H. Ethridge and J. H. Calibration of Pressure Transducers, Inst.
Keefer, Measurements of Air-Blast Phe Soc. of Am. Preprint No.
nomena with Self-Recording Gages, PI9-2-PHYMMID-67, 22nd ISA Confer
WT-1155 Operation TEAPOT - Project ence and Exhibit, Chicago, III., Sept.
1.14b, Feb.-May 1955, (AD-617 170). 1967.
8-26
AMCP 706-181
CHAPTER 9
9-1 GENERAL
9-1
AMCP 706-181
j
cameras available at the moment use optical L~ I
-~~~~~~n~--~5 ~
models of camera have been made on this
principle. One of the first was the Eastman
High Speed Camera, using 16-mm film and
capable of framing rates up to 3500 fps. For
many years, various models of the Fastax . . .-.- 0D
camera-accepting 8-mm, 16-mm, and 35-mm
film and capable of rates up to about 10,000
fps-were considered to be the best available
cameras of this type. The mechanisms for
these cameras are often considerably simpler
-~----ll~
- ----~
- ----r
[]
:1 I
I
D
~ DD
than for intermittent cameras, as can be seen
~~~~~~ER D~
from Fig. 9-4. These early rotating-prism
cameras did not operate at a controlled
framing rate, but instead the film would
~
continuously accelerate during operation.
Some auxiliary method of impressing accurate DD
timing marks on the edge of the film was,
therefore, mandatory for accurate determina
SHUTTER
tion of framing rate and was incorporated in The diagram shows four stages in the revolution of
each camera. For use in large-scale field tests the glass block, the dotted lines indicating the refrac
where good time correlation between differ tion of the beam of light from the lens.
ent cameras is required, coded real-time sig Figure 9-3. Rotating Plane Prism Used in the
nals often are impressed on the films. Newer Eastman High Speed Cameral
designs of rotating-prism cameras are exempli
fied by the Red Lake Laboratories Hycam
cameras, capable of rates up to 44,000 fps
with 16-mm film. These cameras employ a short acceleration period. The control allows
controlled drive mechanism that fixes the their use as low-speed framing cameras, if
framing rate at a preselected value after a desired. The optical system employed in the
Hycam cameras is shown in Fig. 9-5. Most of
the rotating-prism cameras employ built-in or
auxili;lry timing mechanisms to allow remote
actuation and/or synchronization with rapid
processes.
9-2
AMCP 706-181
SUPPLY
SPOOL
LENS
-----
-t~~~ STR I PPER
EIGHT-SIDED
ROTATING PR ISM FILM
TAKE UP
SPOOL
9-3
AMCP 706-181
RELAY LENS
SECOND FIELD
LENS
SECOND PR ISM
/ SEGMENTED SHUTTER
in Fig. 9-7. Here an objective lens and a about 25 total frames. To project their films
diaphragm are placed in front of the main lens as motion pictures, they must be rephoto
as shown. The objective lens forms an image graphed frame-by-frame and run repetitively.
of the object, and this image is then focused
by the main lens on the rotating mirror. The 9-2.2.4 IMAGE DISSECTOR CAMERAS
diaphragm is placed in the plane of the image
formed by the objective lens and serves to The final type of motion picture camera
limit the angle of the marginal rays of light in which can be used for blast wave photography
the system thereafter, so as to prevent more is termed an image dissector camera. The
than two of the secondary lenses from being system works fundamentally by means of a
illuminated at any instant. grid that splits the picture up into a large
number of parallel strips. If the grid has
The first commercially available cameras of relatively narrow transparent spaces and
the type discussed were made by the Beck opaque bands, any single picture taken
man and Whitley Co., and were very large and through it, stationary, will consist of a number
heavy instruments suitable only for labora of narrow bands of image with unexposed
tory use. This company and later the Cordin material between. Obviously, by moving the
Co. have since made lighter and more portable grid sideways through the width of one of the
models. Typical Cordin cameras are shown in spaces in it, another picture can be taken on
Fig. 9-10. These cameras record at framing this unexposed portion, and so on until the
rates as high as 5 X 106 fps, but are limited to whole of the surface of the plate has been
9-4
AMCP 706-181
ROTATING
MIRROR
IMAGE OF
OBJECT OBJECT
f'M IN LENS
Figure 9-6. Essential Features of a Rotating-mirror Framing Camera 3
(Reprinted from High-speed Photography by R. F. Saxe, copyrighted 1966 by the Focal Press, London and New
York. Used by permission of the pubUsher.)
used. After processing, the result will appear it is possible to obtain thirty pictures in
to the eye to be a mere jumble until the grid succession by means of a movement of the
is placed over the picture, when once again grid of 0.015 in. At first sight it might be
any of the pictures recorded can be selected. supposed that this would lead to a very
unsatisfactory picture from the point of view
The number of pictures which can be taken of continuity and definition. However, each
depends upon the width of the transparent picture, when the grid is in position, seems
spaces in the grid and naturally there is a limit very nearly continuous, in spite of the fact
to the amount of unused space, relative to the that twenty-nine thirtieths of it is missing. In
amount of picture space, at which the picture the simplest form of this device, the grid is
becomes too "diluted" to be observed prop placed in the focal plane of the lens, almost in
erly. In practice, it is possible to use a series contact with the stationary film. When the
of clear slits 0.0005 in. wide cut at intervals operation is to be photographed, the grid is
of 0.015 in. in an opaque plate. Thus each moved through 0.015 in. as quickly as pos
line will occupy only one-thirtieth of the total sible and during that period of time a streak
area on to which it can be photographed, and containing the elements of thirty pictures is
\-_-=:::::::::I.~ __--=-:?7"~~~~r--ROTATING
TO
MIRROR
OBJECT
9-5
AMCP 706-181
made. The resulting picture either can be framing cameras record. They are, of course,
analyzed subsequently into thirty separate simpler in design and construction than the
exposures, or the grid can be moved at framing cameras.
uniform slow speed, thus producing a flicker
less motion picture. It should be emphasized The streak camera is used to obtain a
that, under these circumstances, the image continuous monitoring of the rate and direc
produced on the film is not actually a series tion of the propagation of light along a line
of separate pictures but is a smear which is defined by a slit in the camera. No informa
analyzed subsequently by the grid; the resolv tion is recorded about the behavior at points
ing power of the system is determined, other than those on the line selected. It
therefore, by the characteristics of the grid. follows, therefore, that this method mainly is
One camera which has been built on this used where circular symmetry is expected and
principle has been described by Sultanoff4; the line then is selected as a diameter.
this camera uses a focal plane shutter with
slits 0.0001 in. wide moving at 10,000 in. per The essential features of a streak camera
sec and thus producing 100,000,000 pictures are shown in Fig. 9-10. An image of the
per sec. It is difficult to move an actual grid object to be studied is formed by the first lens
across the film at such a speed.An image of it, in the plane of the slit, which is adjusted to
therefore, is moved by reflection from a let through only that part of the image which
rotating mirror, the grid itself being placed in is to be studied. The light that passes through
the optical system in such a place that the the slit is focused by the second lens, via the
objective forms an image on to it. The image mirror, on to the film. The mirror is mounted
is then reprojected, by means of a second on a shaft perpendicular to the plane of the
lens, through a rotating mirror on to the film diagram and the film is constrained to form
plane (Fig. 9-8). an arc of a circle with the mirror shaft as its
center.
9-6
AMCP 706-181
FILM
-
OBJECT
~'\ SUT , /
~
rROT~T1NG
MIRROR
SECOND
FIRST LENS
IMAGE OF OBJECT LENS
Figure 9-9. Essential Features of a Streak
Camera
(Reprinted from High-speed Photography by R. F. Saxe, copyrighted 1966 by the Focal Press, London and New
A test arrangement used at BRL for simul of conventional still camera with a reasonably
taneous streak and still photography of blast fast shutter speed (e.g., .,;;; 1/200 sec) can be
waves from small spherical Pentolite charges is used. Fig. 9-13 is an example of a photograph
shown in Fig. 9-11. The tests are conducted in taken with such a camera in which the shock
a blast chamber, with the optical instrumenta front is clearly visible. There is no point in
tion observing the experiment through a listing or suggesting suitable cameras of this
viewing port. The shock wave is backlit with type, because there are very many on the
an exploding wire light source and a Fresnel market.
lens collimator. An objective lens focuses the
light on the slit in a streak camera, and 9-4.2 FAST SHUTTER CAMERAS
simultaneously on the focal plane in a Kerr
cell still camera (see par. 9-4 for a description We are more concerned here with descrip
of such cameras) through an angled half-silver tion of the special scientific devices which
ed mirror. Examples of streak camera records have much shorter exposure times than that
obtained with this test arrangement are shown given for conventional cameras. Most of these
in Fig. 9-12. The traces of the shock front and are true cameras with very fast special shut
the front surface of the explosive products are ters, but some are not. One type of fast
clearly visible in these records. shutter, developed by Edgerton, employs the
Faraday magneto-optic effect. A cylindrical
slug of polarized glass is surrounded by an
9-4 STILL PHOTOGRAPHY EQUIPMENT electrical coil. On discharge of a heavy current
pulse through the coil, the strong transient
9-4.1 CONVENTIONAL CAMERAS magnetic field depolarizes the glass slug, and
"opens" the shutter. By proper choice of
Many types of still cameras have been used electrical circuit parameters, the depolariza
for photography of air blast waves. For tion can be made to last less than a microsec
photographs of large chemical and nuclear ond. A second type of fast shutter is the Kerr
explosions taken at some distance, any type cell, which relies for its action on the birefrin
9-7
AMCP 706-181
PENTOLITE
I==r=
_________ EXPLOD ING WI RE
x
'(FRESNEL LENS
SPHERE COLLIMATOR
R-----J
!
POSITION
MARKS ON [VIEWING PORT
FILM
BLAST CHAMBER
I NSTRUMENT ROOM
I
E3>-- OBJECTIVE LENS
~ ~i ~PARTIALLY
SILVERED MIRROR
I
M""'-""""' MOD"
STREAK CAMERA .~ ~
9-8
AMCP 706-181
9-4.3 IMAGE CONVERTER CAMERAS image tube is shown in Fig. 9-15. It consists
of a photo-cathode, an electron lens, and a
The final type of still "camera" used in fluorescent screen placed within an evacuated
shock wave photography is not in a strict glass envelope. The photo-cathode and fluo
sense a camera at all, but is instead an rescent screen are on opposite end faces as
adaptation of a cathode ray tube. Devices shown. If an optical image is formed on the
using such tubes are termed image tubes or photo-eathode by means of an optical system
image converter tubes. A schematic of an (on the left hand side in Fig. 9-15) electrons
9-9
AMCP 706-181
will be emitted from the photo~cathode from reproducing the original optical image formed
the illuminated regions of the image. Unless on the photo-cathode. The optical quality of
some limitation occurs, at any instant the the final image obtained will depend to some
current density of the photo-electrons emit extent on the design of the image tube and to
ted from a point on the photo-cathode will be some extent on its method of operation.
directly proportional to the light intensity of
the image on that point of the photo-cathode A schematic of a complete camera employ
at the same instant of time. If a potential ing an image tube, manufactured by STL, is
difference is established between the fluores shown in Fig. 9-16. The STL Image-converter
cent screen and the photo-cathode, these . Camera consists of an objective lens, image
photo-electrons will be accelerated away from converter tube, rear lens, film holder, and
the photo-cathode and towards the fluores plug-in unit and power supply control console
cent screen, being focused by the electron packaged into a single instrument. The objec
lens to form an electron image resembling the tive lens focuses the light radiated from the
optical image on the photo-cathode at the event under study on the photo-cathode of a
fluorescent screen. The bombardment of the curved cathode image-eonverter tube. The
fluorescent screen by energetic electrons photo-cathode transfonns the photon image
causes the screen to emit light, th:ereby into an electron image thereby permitting
9-10
AMCP 70&-181
I
I
I
ELECTRON LENS I
I
I
t
FLUORESCENT
SCREEN
PHOTOCATHODE
Figure 9-15. Image Tube Diagrammatic 3
(Reprinted from High-Speed Photography by R. F. Saxe, copyrighted 1966 by the Focal Press, London and New
York. Used by permission of the publisher.)
9-11
AMCP 706-181
FILM
- __-J~~::-- [-- -- J
V1--=---
---=--=:-.JL
[
- ---
ADJUSTABLE APERTURE -:
ADJ. APERTURE &
SHUTTER
MECH. SHUTTER
ACTUATOR
9-12
AMCP 706-181
particular camera records two frames with the 9-5 SHADOWGRAPH AND SCHLIEREN
same minimum exposure time as the STL EQUIPMENT
camera. A series of backlit photographs of
weak shock waves taken with a Cordin Image Some very simple equipment can be used
converter Camera is shown in Fig. 9-18. to obtain single pictures of air blast waves.
This equipment also is used widely in wind
tunnel and shock tube photography.
9-13
AMCP 706-181
x~~==---\~7 SCREEN
(Reprinted from High-Speed Photography by R. F. Saxe, copyrighted 1966 by the Focal Press, London and New
region of density change is between it and a are due to the second spatial derivative of the
screen S. In the absence of the region of density, and also that the resolution is to a
density change, and with an isotropically large extent dependent on the distance ratio
emitting source L, the illumination on the of source-to-object to source-to-screen.
screen would be sensibly uniform, varying
only as cosO, where 0 is the angle subtended
at the point source by the line joining the 9-5.2 SCHLIEREN EQUIPMENT
point of observation on the screen and the
foot of the normal from the source to the If a density variation exists such that there
screen. is a refractive index gradient in a direction
normal to that of the light rays, the light rays
However, if the region of density change is will be deflected. This deflection may be
interposed, the evenness of illumination on observed by means of a Schlieren system. A
the screen is upset as shown in Fig. 9-19, and Schlieren apparatus is shown in Fig. 9-20. The
a shadowgraph of the region of varying light source L is placed at the focus of a
density is obtained on the screen. concave mirror, and the reflected rays form a
parallel beam of light that illuminates the
If a permanent record is required, a photo "working section". This parallel beam of light
graphic emulsion can be substituted for the is then refocused by another concave mirror
screen and a record made either by transiently to a point focus at P. A lens Q, placed behind
opening a shutter in front of the emulsion, or the focus point P, images in a plane p in the
by transiently illuminating the system by working section on the screen S. If in the
flashing the point source under otherwise plane p, a small section exhibits a refractive
dark conditions. index gradient different from the rest of the
plane, the light rays will be deflected and
The disadvantages of the direct shadow focused at P. However, the lens 12 still will
graph method are that the observed effects form an image of the region ab on the same
9-14
AMCP 706-181
---
al
--
I
pi
5
Figure 9-20. Schlieren System Diagrammatic
(Reprinted from High-speed Photography by R. F. Saxe, copyrighted 1966 by the Focal Press, London and New
York. Used by permission of the publisher.)
part of screen S as when the rays were sures the total angular deflection suffered by
undeflected. a ray of light in crossing between the mirrors,
and the system thus integrates the effects
One method for rendering detectable the experienced along this path length. The meth
movement of the focus point from P to p' od gives no indication of the behavior of the
is to use a knife edge. A knife edge is rays of light from a particular region in the
introduced at the focal plane P so as partially length between the mirrors. It is quite pos
to cut off the light passing through the focus sible that the light rays may suffer the same
P. If now the light from the region ab is total deflection and yet have been subjected
deflected to P', the corresponding region on to quite different conditions.
screen S will be more brightly illuminated
than was the case when the rays were unde The Schlieren method can be combined
flected. Conversely, if the region ab is such with a streak camera to obtain density gradi
that the deflection is in the opposite sense, ent information in one direction as a function
the corresponding region on screen S will be of time.
less brightly illuminated. A Schlieren picture
of an exploding, pressurized glass sphere 9-6 TECHNIQUES IN PHOTOGRAPHY OF
obtained at the University of Toronto is AIR BLAST WAVES
shown in Fig. 9-21.
The techniques applied by various investi
One of the disadvantages of the Schlieren gators in acquiring air blast data by photo
method is the difficulty of obtaining quan graphic means have been as varied as the
titative data regarding the values and positions equipment available to them. The primary
of the gradients that give rise to the observed data obtainable from either motion picture or
picture. Normally, the Schlieren method rnea still photographs are the shape and position of
9-15
AMCP 706-181
the blast front at either a single known time bers-only "sample" the position of the shock
or a series of accurately known successive in specific directions. Third, a very great
times. Positions of particles behind the shock amount of data may be obtained photograph
front have been traced by observing displace ically with relatively little effort. An optical
ments of smoke trails from rockets. Obvious system cannot completely replace pressure
ly, motion pictures provide more data than do transducers, however, because the latter pro
stills, and allow estimation of velocities by vide additional information on the pres
frame-by-frame data reduction methods. A sure-time history at specific points. A photo
good description of motion picture photog optical system can be used to show the
raphy techniques employing cameras such as contour of a shock at small time intervals and,
the Fastax and Hycam rotating prism types is from velocities estimated from these con
given in Ref. 5, and advantages and disad tours, other shock front properties obtained
vantages are compared for other instrumenta via the Rankine-Hugo nio t equations (see
tion techniques. This reference is paraphrased Chapter 2).
in the ensuing discussion.
At the shock front-where a very rapid and
The advantages of an optical system in almost discontinuous rise in pressure, density,
measuring the position of a shock front are and temperature occurs-a similar change also
well known. The most important is that an occurs in the refractive index of the gas.
optical system does not disturb the blast wave Accordingly, a ray of light passing through
that is to be measured. Second, an optical the region immediately behind the shock
system detects the contour of a shock, where front generated by an explosion is deflected
as pressure transducers or other devices that towards the high-pressure region. This prin
indicate the position-time relation of the ciple is used widely in shadowgraph and
shock-unless used in prohibitively large num Schlieren techniques and is used also to
9-16
AMCP 706-181
measure the peak overpressure from nuclear The equipment to obtain the position-time
explosions. In this last case, rocket trails are relation of the shock profile from chemical
established in a region behind the explosion explosions consists of a striped screen or
and photographed through the expanding backdrop (at times supplemented by smoke
shock with cine cameras. Those rays of light trails), one or more high-speed cameras, a
from elements of rocket trails to the camera source of illumination, and a timing system.
which pass through the region immediately
behind the shock front are refracted, and the The striped backdrop consists of a pattern
photographs of the rocket trails show appar of alternate black and white stripes, of a
ent breaks corresponding to the position of width suitable for the field of view concerned;
the shock front. The technique described here and painted on wood, canvas, or metal.
is an extension of this idea to chemical Usually the lines are sloped at 45 deg or 60
explosions and consists merely of photograph deg towards the vertical plane through the
ing the moving shock against a screen or camera and ground zero. The backdrop gen
backdrop painted with alternate black and erally is placed at an expected overpressure
white stripes. Here the position of the shock level of 4 psi; at this level the blast brings the
profile is clearly detectable by the distortion backdrop down, but causes no appreciable
of the regular pattern of stripes introduced by damage to it. Markers are placed in the
the presence of the blast front. The major backdrop plane or in a plane in front of it
differences between the application to nuclear perpendicular to the camera-ground-zero line
explosions and to chemical explosions are in in order to obtain distance measurements
the illumination available and the speed of the from the films, independent of viewing or
cameras required. In the nuclear case, the projection lens conditions. Films are analyzed
illumination provided by the fireball reaches a by projection at a magnification of about 20
value of several hundred suns, the field of onto a horizontal plotting table in order to
view is measured in thousands of feet, and the obtain frame-by-frame observations of the
time of travel of the shock to low-pressure progress of the shocks.
levels is measured in seconds. In the chemical
case, the illumination provided by the fireball This system of measurement has been
is negligible compared with that of the sun, applied to tracing triple-point loci, and to
except for a brief interval after the detona deriving overpressures in the Mach wave and
tion; the field of view may be as small as a the incident wave from over a hundred TNT
few feet; and the time of travel of the shock charges consisting of air-burst spheres of 8
to the pressure level of lowest interest may be Ibm to 1000 Ibm and ground burst spheres
only a few milliseconds. Accordingly, cameras and hemispheres of from 8 Ibm to 5 tons in
having a much higher framing rate and lower
exposure interval are required, and either weight. Observations show that the striped
supplementary illumination or very sensitive backdrop permits delineation of the blast
film must be used. For nuclear tests, inter contour quite readily down to 1 psi, which is
mittent movie cameras with framing rates as the lowest pressure to which it has been
low as 100 fps have proven satisfactory; for applied. The second shock from TNT also is
tests with small blast sources such as 8-lb TNT frequently discernible. An example of the
spheres, rotating-prism cameras with framing shock wave in the Mach reflection region
rates of at least 3000 fps are required. from Ref. 5 is shown in Fig. 9-22.
9-17
AMCP 706-181
Figure 9-22. Views of Shock Wave from 8-lbm TNT Spheres Detonated 8-ft Above ConcreteS
Figure 9-23. View of Shock Wave from 5-ton TNT Ground-burst Hemispheres
9-18
AMCP 706-181
plane behind the source as described by Still photography of air blast waves as a
Groves s . source of blast data suffers in comparison
with either motion picture photography or
Streak camera photography of air blast streak photography because only a single
waves can provide time-distance plots of picture is obtained. But, because of the wide
shock front motion along a chosen line. It has variety of types of equipment available and
been used only for small scale tests in blast the ease and low cost of use of much of this
chambers or in the field. For very strong equipment, it is used widely for photography
shocks close enough to explosive sources that of air blast waves from all possible energy
the shock front is self-luminous, direct streak sources.
photography will record the time history of
shock front motion. Sultanoff and McVey' The simplest equipment for still shock
used this technique in obtaining shock veloc wave photography, the shadowgraph, often
ity measurements close to Pentolite spheres has been used on a laboratory scale to
detonated in air. The use of the streak camera photograph shock fronts. The equipment
can be extended to much weaker shock waves usually requires that the test be conducted in
by backlighting techniques, with spark dis a darkroom. For a blast source such as a
charges, exploding wires, gas-filled flash tubes, chemical explosive that emits considerable
or argon bombs being used as the intense light light, the film must be shielded from this light
sources for backlighting. Glass 8 also has ap by a mask. An excellent example of shadow
plied backlighting to a Schlieren streak system graph photography is given in Fig. 9-25, due
to obtain time histories along a line of to Glass8 . The shadowgraph technique has
shock-front motion, plus density gradients. been adapted by Edgerton to blast wave
One of Glass's streak records is reproduced photography in daylight (see Ref. 2, pp.
here as Fig. 9-24. 427-428). Two of the principal disadvantages
of the conventional shadow method of shock
wave photography are the requirements for
(l) complete darkness during the test, and (2)
a large film size approximating that of the
area to be studied. The technique used by Dr.
Edgerton utilizes a large sheet of ScotcWite
sheeting, either No. 244, Signal Silver, for use
in the dark; or No. 234, Black "C", for
daylight operation. Scotchlite sheeting manu
factured by the Minnesota Mining and Manu
facturing Corporation, has a high degree of
reflective brilliance by returning light directly
back to its source with an efficiency twenty
to over two hundred times that of a white
painted surface. In the Edgerton technique,
the Scotchlite sheeting is used as a backdrop
behind the explosive phenomenon to be
photographed. A single light source close to
the camera lens provides the necessary il
lumination. Light from the source that strikes
the screen normal to its surface is reflected
back directly into the camera lens. The size of
the area covered by this method is limited
Figure 9-24. Schlieren Streak Record of the
9-19
AMCP 706-181
photographs at a 20-ft distance from Signal other than snapping the shutter at the right
Silver Scotchlite sheeting. time and employing a short enough exposure
time to "stop" the shock front. If interrupted
Kerr cell and magneto-optic shutter cam backgrounds have been set up, such as in Fig.
eras both require intense light sources for 9-22, a conventional still camera will "see"
sufficient film exposure. They can be used in the shock fronts as well as a movie camera.
both indoor blast chambers and for field Fig. 9-13 is indicative of good quality still
photography within the range of strong photography of air blast waves.
shocks where they can detect shock fronts.
Self-luminous shocks can be photographed
directly, and somewhat less intense shocks by The coverage in this chapter of photog
backlighting techniques. An example of an raphy of air blast waves as a source of blast
intense, self-luminous shock taken with a data is by no means exhaustive. The author
magneto-optic shutter is shown in Fig. 9-26. instead has attempted to cover, with a few
As is common in such photography, this examples, the kinds of equipment and tech
figure is a double exposure, with the back niques for use of this equipment which have
ground being pre-exposed before the charge been employed by successful experimental
detonation. A typical system for backlit ists. Data obtained by such investigators are
photography employing a Kerr cell shutter is included in much of the work reported in
shown in Fig. 9-27 9 Chapter 5, and are scattered throughout the
air blast literature. Shock photography does
No special techniques are needed for use of offer by far the most spectacular and graphic
conventional cameras in still photography coverage of air blast phenomenology.
9-20
AMCP 706-181
KERR CELL
DELAY
VACUUM
POWER
GENERATOR
CHAMBER
SUPPLY
ENCLOSED SPARK GAP
MIRR\
PARABOLIC
TRIGGER
GLASS
PORTS
VACUUM
SEALS
PULSE
SPARK GAP
AMPLIFIER
POWER
SUPPLY
9-21
AM~7~'8'
REFERENCES
9-22
AMCP 706-181
CHAPTER 10
10-1
AMCP 706-181
is a record from a side-on gage, and Fig. similar system I on 5-in. wide paper. Fig. 10-3
10-1 (B) from a face-on gage. Because dual shows a typical record from an eight-channel
and four-beam scopes are now in common recorder4 . As many as sixteen data traces may
usage, records of this type may include as be recorded on a single record up to 12 in.
many as four traces, recorded simultaneously. wide by certain types of blast recorders.
More common in blast measurement are In addition to being of various widths and
multichannel traces recorded on moving film being either transparent or opaque, the blast
or paper. In Fig. 1O-2(A) we see traces from a records may contain either much or little
four-channel system impressed on 35-mm additional data. Any time history worth
fIlm 3 , while in Fig. 1O-2(B) are traces from a reducing must of course have an accurate time
..., 25 psi -- (\
\ r\. -
t 1 I I I 1 I I I 1 I I I I I I I 1 I I I I 1 I I I I 1 1 I I I I I
" "'"
u.J
e:::: -
=>
V')
V')
r\.
u.J
e::::
c....
~ I---... --.... ,..,
- -I-- t-- ---~
- f--
-
-
t-
t- l-J
-
TIME, 2OOJ.L5eCI division
(A ) SIDE-ON PRESSURE-TIME TRACE
~ 50 JlS~
L~ ARRIVAL TIME --~J
(B ) FACE-ON PRESSURE-TIME TRACE
10-2
AMCP 706-181
CHANNEL 3 ......
CHANNEL 4
R=30ft
R = 38 ft
10-3
AMCP 706-181
10-4
AMCP 706-181
1.0
~
I .10
l.LJ
1/.,.-20
aO. 75
Vl
I J ./
~
c.. 1//
Vl
l.LJ
II /1 /...-
a::: '1/ ./ .50
l.LJ
I" /'" _ ~------
2: O. 50
-
v_.,.- -
I 'L,/
:5
l.LJ
a:::
~
/ f i l,
..... / / / I
0.25 / /' / / '
/ / ", / / I
o
-0.5
--_...-.-
_/
V./
-0.25
",
/
/
/
/
/
/
J
I
I
o 0.25 0.5
REDUCED TIME {tl&)
Figure 10-4. Calculated Response of a Gage of Finite Diameter to Linearly Decaying Pressures
flow effects. Accurate corrections for errors several intermediate calculations. In these
in pressure due to flow effects can be deter calculations pressure as a function of time is
mined only for a particular gage geometry by obtained from the individual total head and
carefully conducted shock tube or wind tun side-on records. The difference between the
nel tests. Ruetenik and Lewis 7 report calibra measured and the true free stream total head
tions and a correction method for a pancake pressure in the absence of the probe has been
side-on gage, and Goodman ll reports the
results of an extensive wind tunnel investiga
8 F: INDICATED PEAK OVERPRESSURE
tion on flow effects around side-on gages of
IcC E: TAKEN AS TRUE PEAK OVERPRESSURE
several different geometries. Correction for
flow effects about even well-streamlined gages f, ,
are indicated by these authors to be as great
as 10-20 percent for shocks which are strong I "8'
enough for flows to be of sonic or greater I ~
Mach number. For relatively weak shocks
(flow of small Mach number), flow effect I :
corrections are negligible provided the gage I :
I
housing is well streamlined. 1
I
I
I H A BG; EXPERIMENTAL RECORD
IAE-EF
10-2.4 REDUCTION OF DYNAMIC I
I
PRESSURE DATA
H A IE iF
A special case in data reduction occurs in ---~i~~;~~~;~i-------------
determination of an important blast wave
parameter, i.e., the dynamic pressure, from
measured pressures. This parameter is usually TIME
not measured directly, but must instead be
inferred from separate measurements of Figure 10-5. Method of Extrapolation of
"total head" and side-on pressure, with Experimental Records
10-5
AMCP 7~181
determined by wind tunnel tests. These cor correction is applied to the data by use of a
rection factors, which are a function of the Mach number versus percent of error curve
Mach number M of the particle flow behind obtained from wind tunnel calibration data
the shock wave, must be applied to the for the probe used.
uncorrected data. The Mach number of the
flow is found by the use of the following two Step 3. The first step then is repeated
equations: using the corrected total head pressure to
calculate a new Mach number. This process is
'Y repeated until sufficient accuracy of the Mach
--
Pt
Ps
= rt(\I)M'] ,1 forM < 1
(10-1)
number is obtained.
[~: l~ Mj'
'Y -1 free-stream side-on pressure finally is used in
Eq. 10-3 to calculate the dynamic pressure as
Pp
= for M > 1 (10-2) a function of time.
~~~'
Ps 'Y- l
--- Typical dynamic pressure-time histories ob
'Y+ l tained from the calculations are plotted in
Fig. 10-6, along with total head and side-on
where pressure records.
Pp ::: total head pressure M >1 At a given time after shock arrival, Pt ::: 25
psi and Ps ::: 17 psi. From Eq. 10-1,
Ps = free stream side-on overpressure
2/7
10-6
AMCP 70~181
.- 60
~50
....i 40
0=
30
:::>
TOTAL HEAD
~ 20
~ 10
c..o-l-.......,..-.,...::::;:==;==ar----,r--..:=h:::::::=,--,.-.,..-=I==~r:_r:~
o 10 20 30 40 50 60 70 80 90 100 110 120 130 1 0 150
:~~~
g 20
~ 10
SIDE-ON
~ 0 -T"'I -::::S::=::I=:=r::::Jf':r- J
c.. 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
VI
CI.
~f~~ DYNAMIC
0+--=;::::::::I;:::::~ill'====;I;:=:=:::;I;::::===;i;::::===;r====;Ir===r====;Ir-O=;Ir===;ir===;Ir===r----,i
VI
~ 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
c.. TIME, msec
Figure 10-6. Recorded Side-on and Total Head Pressure- Time Histories and Calculated
Dynamic Pressure- Time History
Had M been greater than one, recalculation methods for reading and smoothing pressure
using Eq. 10-2 would have been required time data which allow correction for gage size
before substitution in Eq. 10-3. and frequency response errors in peak pres
sure, and better estimation of positive phase
10-2.5 DETERMINATION OF POSITIVE duration than by direct reading of records.
PHASE DURATIOI\l His procedure consists of replotting the stan
dard linear records such as shown in Figs.
In addition to correction of peak pressure 10-1 through 10-3 on semi-logarithmic plots,
for gage size error, a recurring problem in and fitting straight lines to these plots to
blast data reduction is that of accurate deter obtain estimates of peak pressure and dura
mination of the duration of the positive phase tion. A linear plot such as Fig. 10-7 is
of a pressure-time history. Relatively large transformed into a semi-logarithmic plot
variations in this blast parameter are almost (pressure on the logarithmic scale) as in Fig.
unavoidable because of the more or less 10-8 to obtain an estimate of the true peak
exponential character of the pressure decay pressure and initial decay rate. Because the
and consequent nearly horizontal slopes of initial decay rate is nearly exponential, a
blast pressure records on return to ambient straight line can be accurately fitted to the
pressure. Ethridge9 has proposed graphical initial portion of the record. For estimating
10-7
AMCP 706-181
40
30
In
Co. 20 \
\
u.f
c::::
::>
~
~1O
c::::
"'
c....
~
~ .
o ~
- ... -
-10 ....
' L' ..1.' ..1.' .J.' .J.' ..J' ,'--_ _.....L.'_ _--!'
_.
positive phase duration, time is plotted loga 40
rithmically as in Fig. 10-9, and a curve fitted ~m
to the new plot. "\
10 \.
~
Although no limiting exponential is ap 11'1
proachedat the end of the positive phase, Fig.
10-9 does allow more accurate determination '\u~P. pme-et
of the ~end of the positive phase and the final II
-~
slope of the positive phase, as is apparent in In
'I
thefigUl'e. Ethridge also claims that the L
semi-logarithmic plots are useful for manually "'
developing a smoothed waveform.
'\
data under the assumption that the traces are 1"'
"clean" and easily read. But, unfortunately,
in blast research, one often must attempt to
recover data from records of poor quality.
The particular test may be an expensive
-t 20 40 60
l1ME 1 . _
80 100 120 140
10-8
AMCP 706-181
16
~
\
12
~
Vi
~
...
Q.
0::::
:::>
8 '" "
~~
...
tn
tn ~~ 'L.
4
'~
0::::
c.. t = toe- BP
......
0 ~ ~I
~\J"
j~ ~
~
I
~"!'-/
-4 L...-_..J...._ _ ...I..._---L_...J.........J.---J---J~I.....J.I L....-_ _ .L.-_~_l...-....l-_'_I......L..!........
' I
etc. How does one reduce data from such voltage calibration steps, entering calibration
poor quality records, or decide that these data data, printing out numerical results, and
should be discarded? Unfortunately, we can re-plotting the records on known pressure and
not answer this question for you. We can only time scales. This process requires quite elabo
warn you that you will, at some time, need to rate and sophisticated equipment and a some
find the answer for yourself if you are what complex digital computer program, and
involved in blast testing. usually is not employed in reduction of blast
data, or is employed only after careful scru
10-3 REDUCTION OF MAGNETIC TAPE
DATA
tiny of oscillograph traces recorded on play
back of the tapes. No one conversant with
Magnetic tape systems used in recording blast measurement technology apparently is
blast or other dynamic data usually employ willing to entrust his data to machines for
either seven- or fourteen-channel recorders, reduction without first examining the time
with one channel being reserved for a timing histories. The usual procedure is simply to
signal. So, either six or thirteen data channels playback the magnetic tape and record on
are recorded simultaneously. The raw data magnetic oscillographs, with calibration steps
consist of the magnetic tapes themselves. on the same records. The data are then in the
Conceivably, completely automated systems form of paper photographic traces, very simi
could be employed directly to reduce the lar to those obtained from multichannel
magnetic tape data-reading the voltages gen oscilloscope blast recording systems. Data
erated by the tapes as they are played back at reduction then can be accomplished by the
discrete time intervals, comparing these with methods described earlier in the chapter.
10-9
AMCP7~'8'
Magnetic tape records offer one important outputs are time (msec), pressure (psi), and
advantage over direct recording on film or impulse (psi-msec), which are punched on
paper. One can, by playing back through IBM cards. These cards are used for plotting
suitable filters, recover data from "ringing" and tabulating the results.
transducers or recording systems which other
wise would be lost or seriously reduced in 10-5 REDUCTION OF DATA FROM MO
usefulness. Also, rather low frequency re TION OR STI LL PHOTOGRAPHS
sponse oscillographs can be used to reproduce
tape data by playing back at tape drive speeds Motion or short duration still photographs
somewhat less than those used to record the often are taken of blast experiments for
data originally. documentary or publicity purposes. (One can
tell at a glance that an engineer or scientist
10-4 REDUCTION OF DATA FROM SELF has participated in nuclear or large-scale con
RECORDING GAGES yentional explosive tests by the beautiful
color photographs of fireballs or mushroom
The data from self-recording gages consist clouds which adorn his office walls.) Such
of lines that have been scribed by styli on the distant overall views are even useful at times
surface of rotating polished metal or silvered in detecting anomalies in large-scale explo
glass discs, or of translating polished metal sions. To be useful for obtaining data on blast
tapes. Deflections of the styli (and cor wave parameters, cameras and background
responding trace amplitudes) are of the order must be carefully arranged, as described in
of 0.020 to 0.060 in. The records mayor may Chapter 9. Usually, only the shock front can
not have a superimposed fiducial mark indi be observed (see Fig. 10-10), so one can only
cating some common zero time, or timing obtain those properties that can be inferred
marks on a separate trace to give the time from successive positions of this front at
base. Amplitude calibration is almost never known time intervals.
included on the record, but instead is deter
mined by a separate laboratory test. Although A method for obtaining peak air blast
it would be desirable to show typical records pressures employing photographic techniques
here, this cannot be done because they do not requires observing the passage of the shock
reproduce well. wave on an interrupted background. This
technique is dependent upon the principle of
Because the traces from these gages are of light refraction. Light waves passing obliquely
small amplitude, they must be magnified from one medium to another, in this case
considerably to be read. At BRU 0 records from undisturbed air to the compressed re
from self-recording gages are read with the aid gion immediately behind the shock wave,
of a toolmaker's microscope modified to use undergo an abrupt change in direction. The
magnetic reading heads. Output signals from "bending" of light rays by the shock wave
the heads are fed into suitable conditioning causes an apparent displacement of the back
equipment that punches the x- and y-co ground against which the shock is viewed. The
ordinates of each point read into an IBM card. data consist of a series of photo'graphs or
These cards, representing readings taken at frames, similar to those in Fig. 10-10.
short intervals throughout the span of the
record, together with cards representing cali Peak overpressures can be inferred from
bration steps and time interval information, shock velocities compu ted from such data by
are used as input to a digital computer. The use of the Rankine-Hugoniot equation which
pressure values are calculated from a straight expresses pressure as a function of shock
line interpolation between the various calibra velocity.
tion steps. A time calibration is applied to the
readings, and at the same time the impulse is ps = 2'Y 1M; - I) Po (04)
summed as the cards are processed. The 'Y + I \"
10-10
AMCP 706-181
INTERRUPTED BACKGROUND
R..
( b-R~
Figure 10-10. Photo-optical Records of Shock
Front Profile 1 1
a+J) fori~22
(10-5)*
8. = Tan
-I (R.- '-b)
, a+d for i ~ 23
where
= h sin (8 - 8j ), i ~ 22
Ps peak side-on overpressure } (10-6)
= h sin (8 + 8), i ~ 23
I
Fig. I0-11 illustrates the geometry of camera,
r. r
charge, and background. In this figure, R j is , (1-1)
10-11
AMCP7.181
The velocity component of the prevailing ing the time taken for the plug to travel a
winds parallel to the interrupted background known distance.
at the time of the test should be stripped
from all velocity measurements before ap
plication of these data to the Rankine
I = i(f -gt/2) (10-9)
Hugoniot equation. This is necessary as the
wind effectively increases or decreases the where
velocity of the shock front, depending upon
the direction of the wind vector in relation to I = impulse, psi-msec
shock wave propagation. A wind vector tan
gent to the shock front will have no effect on m = mass of plug, Ib-msec 2 /in.
shock velocity measurements.
A = area of plug, in. 2
10-6 OTHER DATA REDUCTION x= distance plug travels between top
and bottom plate, in.
Data obtained from air blast transducers
such as the mechanical transducers described g = acceleration due to gravity, in./msec 2
in Chapter 7 usually can be reduced quite
simply, once the devices are "calibrated". t = time of travel, msec
Only limited information, such as effective
energy yield of the explosive usually is For the optical measurements, where the
obtained, rather than estimates or measures of time origin is not known but the time interval
specific blast wave parameters. The relatively over a predetermined distance is known, ve
simple measures of damage (such as per locity at distance Xi is given in terms of a
manent tip deflection of a metal cantilever known time interval by
beam) can be used to estimate effective
energy yield from calibration curves such as (10-10)
shown in Chapter 7, provided only that the
distance from explosive energy source to the
mechanical gage is known. where
10-12
AMCP 706-181
REFERENCES
10-13/10-14
AMCP 706-181
BIBLIOGRAPHY
R-l
AMCP 706-181
P. C. Chou, R. R. Karpp, and S. L. Huang, C. W. Davis, Jr., The AFWL ROC-VTS Com
"Numerical Calculation of Blast Waves by the puter Code, AFWL-TR-65-74, July 1965.
Method of Characteristics", AIAA Journal,S,
4, 618-623 (April 1967). V. D. D. Dawson, Pressure-Gauge Design for
the Measurement of Pressures in Shocktube
W. H. Chu and W. E. Baker, Performance Wind Tunnels, Shocktubes, and Guns.
Analysis of a Helium-Air Shock Tube, Final Navweps Report 7326, U. S. Naval Ord. Lab.,
January 1961.
Report, SwRI Project 02-1974, Southwest
Research Institute, San Antonio, Texas, De
cember 1966. R. DeBar, One Dimensional Eulerian Hydro
dy namics Difference Equations, UCRL
12238-T, Lawrence Radiation Lab., Liver
R. H. Cole, Underwater Explosions, Dover more, Calif., October 1965.
Publications, Inc., N.Y., 1965.
J. H. deLeeuw, I. I. Glass and L. E. Heuck
M. Collins, Use of the HYDRO I Code in the roth, A High-Speed Multi-Source Spark
Analysi; of Transient Axisymmetric Shock Camera, UTIA Tech. Note No. 26, 1960.
Hydrodynamics Problems, Tech. Report
3251, Picatinny Arsenal, July 1965. Jane M. Dewey, O. T. Johnson, and J. D.
Patterson II, Mechanical Impulse Measure
M. A. Cook, The Science of High Explosives, ments Close to Explosive Charges BRL Re
Reinhold, N.Y., 1958. port.
R. Courant and K. O. Friedrichs, Supersonic Jane Dewey and J. Sperrazza, The Effect of
Flow and Shock Waves, Interscience Pub., A tmospheric Pressure and Temperature on
Inc., N.Y., 1948. Air Shock, Ballistic Research Labs. Report
B-3
AMCP 7~181
No. 721, Aberdeen Proving Ground, Md., 10780, U. S. Naval Ordnance Laboratory,
1950. January 23, 1950.
John M. Dewey, "The Air Velocity in Blast J. W. Fitzgerald, Design Considerations for
Waves from TNT Explosion", Proc. of the Pencil Type Air Blast Gauges, Contract No.
Royal Soc., A 279,366-385 (1964). DA-36-034-QRD-1860, Prepared for BRL by
Atlantic Research Corporation, Alexandria,
John M. Dewey, "Precursor Shocks Produced Va., November 1,1955.
by a Large Field Chemical Explosion", Na
ture, 205, 1306. P. D. Flynn, Elastic Response of Simple
Structures to Pulse Loading, BRL Memo
John M. Dewey and W. A. Anson, "A Blast Report No. 525, Aberdeen ~roving Ground,
Wave Density Gage Using Beta-Radiation", J. Md., November 1950.
Sci. Instrum. 46,568-572 (1963).
P. Fox and A. Ralston, "On the Solution of
W. Do'ering and G. Burkhardt, Contributions Equations for Spherical Waves of Finite Am
to the Theory of Detonation, TR No. plitude", J. Math. and Phys., 36, 4 (Jan.
F-TS-1227-IA, Wright-Patterson AFB, Ohio, 1958 ).
May 1949, p. 307.
S. Fujiwhara, Bull. Cent. Meteor, Obs. Japan,
R. E. Duff, and R. N. Hollyer, The Effect of 2, I (1912), and 2, 2 (1916).
Wall Boundary Layer on the Diffraction of
Shock Waves Around Cylindrical and Rectan W. R. Gage and C. L. Mader, Three-Dimen
gular Obstacles, Univ. of Michigan, Ann sional Cartesian Particle-In-Cell Calculations,
Arbor, Michigan, June 21, 1950. LA-3422, January 1966.
V. Ericsson and K. Edin, "On Complete S. K. Garg and J. Siekman, "Approximate
Blast Scaling", Jour. of the Physics of Fluids, Solution of the Propagation of an Axisym
3, 5, 893-895 (Sep - Oct 1960). metric Blast Wave Generated by a Finite
Spherical Charge", ZAMP, 17, I, 108-121
N. H. Ethridge, A Procedure for Reading and (1966).
Smoothing Pressure-Time Data from H. E.
and Nuclear Explosions, BRL Memo Report R. A. Gentry, R. E. Martin, and B. J. Daly,
1691, Aberdeen Proving Ground, Md., Sep "An Eulerian Differencing Method for Un
tember 1965. steady Compressible Flow Problems", J. of
Computational Physics, 1,87-118 (1966).
W. O. Ewing, Jr., and J. W. Hanna, A
Cantilever for Measuring Air Blast, BRL Tech.
N. Gerber and J. M. Bartos, Tables of Cylin
Note No. 1139, August 1957.
drical Blast Functions, BRL Memo Report
W. J. Fader, Amplifier Frequency Response 1376, Aberdeen Proving Ground, Md., 1961.
Requirements for Recording Short Duration ,
Air Blast Pressure Pulses, BRL Memo Report L. Giglio-Tos and R. E. Reisler, Air Blast
No. 569, October 1951. Studies of Large Ammonium Nitrate/Fuel Oil
Explosions, BRL Memo Report No. 2057,
E. Fisher, Air Blast Resulting From the August 1970.
Detonation of Small TNT Charges, Report
NAVORD 2890, U. S. Naval Ordnance Labo I. I. Glass, Aerodynamics of Blasts, UTIA
ratory, July 27,1953. Review No. 17, Institute of Aerophysics,
Univ. of Toronto, September 1960.
E. Fisher, Spherical Cast TNT Charges: Air
Blast Measurements on, Report NOLM I. I. Glass, "Spherical Flows and Shock
B-4
AMCP 706-181
1. 1. Glass and J. H. Hall, "Shock Sphere - An W. Griffith, "Shock Tube Studies of Tran
Apparatus for Generating Spherical Flows", sonic Flow Over Wedge ProfIles", J. Aero.
Jour. Appl. Phys., 28,424 (1957). Sci., 19,249 (1949).
1. I. Glass and L. E. Heuckroth, "Head-on T. K. Groves, "A Photo-Optical System of
Collision of Spherical Shock Waves", Phys. of Recording Shock ProfIles from Chemical Ex
Fluids, 2 , 5, 542 (1959). plosives", Shock, Vibration, and Associated
Environments, Bulletin No. 28, Part III, 59-67
S. Glasstone, Ed., The Effects of Nuclear (Sept. 1960).
Weapons, U. S. Atomic Energy Commission,
Rev. Ed., April 1962. F. H. Harlow, "Hydrodynamic Problems In
volving Large Fluid Dynamics", J. Assn.
Compo Machinery, 4, 2 (April 1957).
H. Goldstein and A. Hoffman, Preliminary
Face-On Air Blast Measurements BRL Tech
F. H. Harlow, "The Particle-In-eell Method
nical Note No. 788, Aberdeen Proving
for Numerical Solution of Problems in Fluid
Ground, Md., April 1953.
Dynamics", Symposia in Appl. Math., Proc.,
XV (1963).
H. Goldstein and J. von Neumann, Blast Wave
Calculation, Collected Work of J. von Neu
F. H. Harlow, Theory of Correspondence
mann, Vol. VI, Pergamon Press, Oxford,
Between Fluid Dynamics and Particle and
England, 1963, pp. 386-412 (also McMillan,
Force Models, Report LA 2806, Los Alamos
N.Y.).
Scientific Lab., November 1962.
H. J. Goodman, Aerodynamic and Frequency
F. H. Harlow, et al., Two-Dimensional Hydro
Dependent Errors in an Air Blast Gauge, BRL
dynamic Calculations, LA-230 I, Los Alamos
Report No. 1345, October 1966.
Scientific Lab., Los Alamos, New Mexico,
September 1959.
H. J. Goodman, Compiled Free-Air Blast Data
on Bare Spherical Pentolite, BRL Report No.
F. H. Harlow and B. D. Meixner, The Particle
1092, Feb. 1960.
and-Force Computing Method in Fluid Dy
namics, Report LAMS-2567, Los Alamos
H. J. Goodman and R. E. Shear, Pressure,
Scientific Lab., June 1961.
Density and Internal Energy of Pentolite
Explosion Products, BRL Report No. 1212.
D. R. Hartree, Some Practical Methods of
Using Method of Characteristics in the Cal
J. W. Goresh and R. G. Dunn, Tables of Blast culation of Non-Steady Compressible Flows,
Wave Parameters-i. Spherical Explosions, Report LA-HU-I, Los Alamos, 1952.
ARL 69-0011, Aero Research Lab., Office of
Aerospace Res., USAF, Jan. 1969. D. Hicks, Hydrocode Test Problems, AWL-TR
67-127, Air Force Weapons Laboratory, Feb
B. A. Granath and G. A. Coulter, BRL Shock ruary 1968.
Tube Piezo-Electric Blast Gauges, BRL Tech
nical Note No. 1478, Aberdeen Proving
Ground, Md., August 1962. D. Hicks and R. Pelzl, Comparison Between a
von Neumann-Richtmyer Hydrocode
S. A. Granstrom, "Loading Characteristics of (AFWL's PUFF) and a Lax-Wendroff Hydro
Air Blasts from Detonating Charges", Acta code, AFWL-TR-68-112, Air Force Weapons
I'Qlytechnica, 196, (1956); also, Trans. of the Laboratory, October 1968.
B-5
AMCP 706-181
J. Hilsenrath and M. Klein, Tables of Thermo Speed Photography, The Macmillan Co., N.Y.
dynamic Properties of Air in Chemical Equi 1962.
librium Including 2nd Virial Corrections from
1500 0 K to 15,OOOoK, TDR-63-161, Arnold W. H. Jack, Jr., Measurements of Normally
Engineering Dev. Center, 1963. Reflected Shock Waves from Explosive
Charges, BRL Memo Report No. 1499, July
J. Hilsenrath, et al., Thermodynamic Prop
1963.
erties of Highly Ionized Air, Air Force Special
Weapons Center, AFSWC-TR-56-35, April W. H: Jack, Jr., and B. F. Armendt, Jr.,
1957. Measurements of Normally Reflected Shock
Parameters from Explosive Charges Under
D. V. Ho, Notes on Shock Propagation in a Simulated High Altitude Conditions, BRL
Non-Uniform Duct, Tech. Report No.4, Report No. 1280, Aberdeen Proving Ground,
Contract DA-19-Q20-0RD-5126, Div. of Md., April 1965.
Appl. Math, Brown Univ., Providence, R. I.,
Nov. ~60. K. J. Jarvis, The Measurement of Air Blast,
Armament Research and Development Estab
A. J. Hoffman and S. N. Mills, Jr., Air Blast lishment, Ft. Halstead, Kent, England (un
Measurements About Explosive Charges at dated).
Side-On and Normal Incidence, BRL Report
No. 988, Aberdeen Proving Ground, Md., July R. L. Jarvis, et al., A Program to Induce High
1956. Blast-Induced Airloads and Structural Re
sponse of Lifting Surfaces, ASD-TDR-63-764,
M. Holt, Basic Developments in Fluid Dy Vol. II, July 1964.
namics, Vol. I, Academic Press, N.Y., 1965.
E. G. Johnson, Propellant Hazards Research
M. Holt, "The Initial Behaviour of a Spherical Facility, Technical Report S-152, Rohm &
Explosion", Proc. Roy. Soc. (London), A234, Haas Co., Huntsville, Ala., Oct. 1967.
89 (1956).
O. T. Johnson and W. O. Ewing, Jr., An
B. Hopkinson, British Ordnance Board Min Omni-Directional Gauge for Measuring the
utes 13565, 1915. Dynamic Pressure Behind a Shock Front,
BRL Memo Report No. 1324, March 1962.
S. L. Huang and P. C. Chou, Calculations of
Expanding Shock Waves and Late-Stage O. T. Johnson, J. D. Patterson, II, and W. C.
Equivalence, Report 125-12, Contract DA Olson, A Simple Mechanical Method for
I 8-00 I-AMC-8K(x), Drexel Institute of Tech., Measuring the Reflected Impulse of Air Blast
Philadelphia, Penn., April 1968. Waves, BRL Memo Report No. 1088, July
1957, Also, Proc. of 3rd U. S. Mat. Congo of
S. L. Huang and P. C. Chou., Solution of Blast Appl. Mech., ASME, N. Y., 203-207 (June
Waves by a Constant Time Scheme in the 1958).
Method of Characteristics, Report No. 125-9,
Contract No. DA-18001-AMC-876(X), Drexel W. E. Johnson, OIL-a Continuous Two
Institute of Tech., August 1966. Dimensional Eulerian Hydrodynamic Code,
GAMD-5580, October 1964.
H. Hugoniot, "M~moire sur la propagation du
mouvement dans les Corps et specialement D. L. Jones, "Blast Waves and Scaling Laws",
dans les gaz parfaits", J. de l'ecole poly tech. Physics of Fluids, 13, 5; 1398-1399 (May
Paris, 57 (1887), and 58 (1889). 1970).
W. G. Hyzer, Engineering and Scientific High- D. L. Jones, Energy Parameter B for Strong
6-6
AMCP 706-181
Blast Waves, National Bureau Std. Tech. Note A. R. Kiwan, Self Similar Flows Outside an
155, 1962. Expanding Sphere, BRL Report No. 1495,
September 1970.
D. L. Jones, "Intermediate Strength Blast
Wave", The Physics of Fluids, 11, 1664-1667 H. G. Kolsky, A Method for the Numerical
(Aug. 8, 1968). Solution of Transient Hydro-dynamic Shock
Problems in Two Space Dimensions, Report
D. L. Jones, "Strong Blast Waves in Spherical, LA-1867, Los Alamos Scientific Lab., March
Cylindrical and Plane Shocks", Phys. Fluids. 1955.
4, 1183 (Sept. 1961). Also see Erratum, Phys.
Fluids, 5,637 (May 1962). L. D. Landau and K. P. Stanyukovich, "On
the Study of Detonation in Condensed Explo
D. M. Jones, P. M. Martin, and C. K. sives", Doklady Akad.. Nauk SSSR, 46,399
Thornhill, "A Note on the Pseudo-Stationary (1945).
Flow Behind A Strong Shock Diffracted or Re
flected at a Comer", Proc. Royal Soc., A209 R. J. Larson and W. Olson, Measurements of
238 (1951). Air Blast Effects from Simulated Nuclear
Reactor Core Excursions, BRL Memo Report
G. A. Jones, High Speed Photography, John No. 1102, Aberdeen Proving Ground, Md.,
Wiley and Sons, Inc., N. Y., 1952. September 1957.
B-7
AMCP70~181
S. C. Lin, "Cylindrical Shock Waves Produced E. Mach and J. Sommer, "Uber die Fortpflanz
by an Instantaneous Energy Release", 1. ungsgeschwindigkeit von Explosion
Appl. Phys., 25, 1 (Jan. 1954). sscha1lwellen", Akademie de Wissenschafter,
Sitzangberichte der Wiener, 74 (1877).
H. E. Lindberg and R. D. Firth, Structural
Response of SPINE Vehicles, Vol. II, Simula
R. C. Makino, The Kirkwood-Brinkley Theory
tion of Transient Surface Loads by Explosive
of the Propagation of Spherical Shock Waves,
Blast Waves, Tech. Report AFWL-TR-66-l63,
and Its Comparison with Experiment, BRL
Vol. II, Air Force Weapons Lab., Kirtland
Report No. 750, Aberdeen Proving Ground,
AFB, New Mexico, May 1967.
Maryland, April 1951.
B-8
AMCP 706-181
R. W. Morton and J. L. Patterson,A Transient T. Orlow, et al., A Computer Program for the
Pressure Measurement System for Blast Effect Analysis of Transient Axially Symmetric Ex
Research, Paper 148-LA 61-1, Presented at plosion and Shock Dynamics Problems,
16th Annual ISA Instrument-Automation USNOL Report NAVWEPS 7265,1960.
Conf., Los Angeles, California, September
1961. K. Oshima, Blast Waves Produced by Explo
sive Waves, Report 358, Aeronautical Re
J. C. Muirhead and W. M. McMurtry, "Surface
search Institute, Univ. of Tokyo, Tokyo,
Tension Gauges for the Measurement of Low
Transient Pressures", Review of Sci. Inst., 33, Japan, July 1960.
12,1473,1474 (Dec. 1962).
D. C. Pack, "The Reflection and Diffraction
J. A. McFadden, Initial Behavior of a Spher of Shock Waves", J. Fluid Mech., 18, 549
ical Blast, NAVORD Report No. 2378, 1952. (1964).
B-9
AMCP 706-181
B. Perkins, Jr., P. H. Lorrain and W. H. ball, BRL Memo Report No. 818, December
Townsend, Forecasting the Focus of Air Blasts 1966.
due to Meteorological Conditions in the Lower
Atmosphere, BRL Report No. 1118, Aber R. E. Reisler, 1. Giglio-Tos, and R. C. Kellner,
deen Proving Ground, Md., October 1960. Ferris Wheel Series, Flat Top Event, Project
Officers Report - Project 1.1, A irblast
R. Piacesi, Numerical Hydrodynamic Calcula Phenomena, POR-3001, October 1966.
tion of the Flow of the Detonation Products
from a Point-Initiated Explosive Cylinder, R. E. Reisler, 1. Giglio-Tos, G. D. Tee1, and
NOLTR 66-150, U. S. Naval Ordnance Lab., D. P. LeFevre, Air Blast Parameters from
White Oak, Md., January 1967. Summer and Winter 20-Ton TNT Explosions,
Operation Distant Plain, Events 3 and 5. BRL
R. Piacesi, et aI., A Computer Analysis of the Memo Report No. 1894, November 1967.
Two Stage Hypervelocity Model Launchers,
NOLTR-62-87, U. S. Naval Ordnance Lab., R. E. Reis1er, N. H. Ethridge, D. P. LeFevre,
Feb. 1963. and 1. Giglio-Tos, Air Blast Measurements
From the Detonation of an Explosive Gas
H. B. Pierce and J. C. Manning, Experimental Contained in HemisNherical Balloon (Opera
Investigation of Blast Loading on an Airfoil in tion Distant Plain Event 2A), Ballistic Re
Mach Number 0.7 Airflow With Initial Angle search Laboratories, Aberdeen Proving
of-Attack Change of 20, NASA TN D-1603, Ground, Maryland, July 1971.
February 1963.
M. Rich. A Method for Eulerian Fluid
Dynamics Report No. LAMS-2826, Los
K. D. Pyatt, HECTlC-a Two-Dimensional
Alamos Scientific Lab., 1963.
Interaction Code, AFWL-TR-65-90, Vol. II,
Addendum, October 1965.
P. I. Richards, Sharp Shock Lagrangian
Hydrodynamics Code, Report No. TO-B
A. Ralson and H. S. Wilf, Mathematical 66-74, BRL, Aberdeen Proving Ground, Md.,
Methods for Digital Computers, John Wiley & October 1966.
Sons, N. Y., 1960.
E. H. Rogers, Two-Numerical Methods for
T. S. Rathke, A System for Dynamic Calibra Calculating Axisymmetric Flow, BRL MR
tion ofPressure Transducers Inst. Soc. of Am. 1745, Ballistic Research Lab., Aberdeen Prov
Preprint No. PI9-2-PHYMMID-67, 22nd ISA ing Ground, Md., May 1966.
Conference and Exhibit, Chicago, Ill., Sept.
1967. M. H. Rogers, "Similarity Flows Behind
Strong Shock Waves", Quart. J. Mech. Appl.
R. E. Reis1er, "The Mechanical Self-Recording 8, 337-367 (1960).
Pressure-Time Gage - A Useful Instrument for
the Acquisition of Air Blast Data from Nu J. Rosciszewski, "Calculations of the Motion
clear and Large HE Detonations", Shock, of Non-Uniform Shock Waves", J. Fluid
Vibration and Associated Environments, Bul Mech., 8,337-367 (1960).
letin No. 28, Part III, 99-112 (Sept. 1960).
C. A. Rouse, "Theoretical Analysis of the
R. E. Reister, J. H. Keefer, and L Giglio-Tos, Hydrodynamic Flow in Exploding Wire Phe
Basic Air Blast Measurements from a SOO-Ton nomena", Exploding Wires, 227, Plenum Press,
TNT Detonation Project 1.1 Operation Snow N.Y. (1959)..
B-10
AMCP 706-181
L. Rudlin , On the Origin of Shockwaves from L. I. Sedov, Similarity and Dimensional Meth
Spherical Condensed Explosions in A ir. Part I, ods in Mechanics, Academic Press, N. Y.,
Results of Photographic Observations of Pen 1959, p. 210.
toWe Hemispheres at Ambient Conditions, .
R. E. Shear, Detonation Properties of Pento
NOLTR 62-182, DASA-1360, U. S. Naval
lite, BRL Report No. 1159, December 1961.
Ord. Lab., August 1963.
R. E. Shear, Incident and Reflected Blast
J. R. Ruetenik and S. D. Lewis, Computation Pressures for Pentolite, BRL Report No.
of Blast Properties for Spherical TNT or 1262, Aberdeen Proving Ground, Md., Sep
Pentolite from Measured Pressure Histories, tember 1964.
AFFDL-TR-66-47, October 1966.
R. E. Shear and B. D. Day, Tables of
Thermodynamic and Shock Front Parameters
J. R. Ruetenik and S. D. Lewis, Pressure for Air, BRL Memo Report No. 1206, May
Probe and System for Measuring Large Blast 1959.
Waves, Tech. Report AFFDL-TR-65-35, A. F.
Flight Dynamics Lab., Wright-Patterson AFB, R. E. Shear and R. C. Makino, Unsteady
Ohio, June 1965. Spherical Flow Behind a Known Shock Line,
BRL Report No. 1154, BRL, Aberdeen Prov
ing Ground, Md., November 1961.
R. G. Sachs, The Calibration of Paper Blast
Meters, BRL Report No. 472, Aberdeen R. E. Shear and P. McCane, Normally Re
Proving Ground, Md., June 1944. flected Shock Front Parameters, BRL Memo
Report No. 1273, Aberdeen Proving Ground,
R. G. Sachs, The Dependence of Blast on Md., May 1960.
Ambient Pressure and Temperature, BRL
Report No. 466, Aberdeen Proving Ground,
Md., May 1944. R. E. Shear and E. Q. Wright, Calculated Peak
Pressure-Distance Curves for PentoWe and
TNT, BRL Memo Report No. 1423, August
A..Sakurai, "On the Propagation of Structures
1962.
of the Blast Wave, Part I", J. of Phys. Soc. of
Japan, 8,5,662-669 (1953), also "Part II",
Jour. Phys. Soc. of Japan, 9, 2, 256-266 B. Soroka and G. T. Watson, An Eight-Chan
(1954). nel High-Performance Oscillograph Recording
B-ll
AMCP 706-181
System, BRL Memo Report No. 1765, May G. I. Taylor, "The Formation of a Blast Wave
1966. by a Very Intense Explosion: II The Atomic
Explosion of 1945", Proc. Royal Soc., 201,
B. Soroka and J. Wenig, High Impedance 175-186 (1950).
Cathode Followers for Piezoelectric Gage
Recording in Air Blast Research, BRL Memo G. I. Taylor, "The Formation of a Blast Wave
Report No. 1474, April 1963. by a Very Intense Explosion: I Theoretical
Discussion", Proc. Royal Soc., 201, 159-174
J. Sperrazza, "Modeling of Air Blast", in Use (1950).
of Models and Scaling in Shock and Vibra
tion, W. E. Baker, Ed., ASME, N. Y., Nov. G. I. Taylor, "The Propagation and Decay of
1963, 65-78. Blast Waves", Paper 20 in The Scientific
Papers of Sir Geoffrey Ingram Taylor, III,
R. S. Srivastava, and M. G. Chopra, "Diffrac Cambridge Univ. Press, 221-235 (1963).
tion of Blast Wave for the Oblique Case", 1.
Fluid Mech., 40, Part 4, 821-831 (1970). G. I. Taylor, "The Propagation of Blast Waves
Over the Ground", Paper 29 in The Scientific
K. P. Stanukovich, Unsteady Motion of Con Papers of Sir Geoffrey Ingram Taylor, III,
tinuous Media, Pergamon Press, N.Y., 1960. Cambridge Univ. Press, 274-276 (1963).
&12
AMCP 706-181
F. Wecken, "Les Lois de Similitude Dans Les M. Wilkins, et al., A Computer Program for
Calculating One-Dimensional Hydrodynamic
Explosions A Syme trie SpMrique", Memorial
De L 'Artillerie Francaise Sciences et Tech Flow-Ko Code, UCRL 6919, Univ. of Calif.
niques de l'Armement, 35,136 (1961). Radiation Lab., July 1962.
B-I3
ANK;P 706-181
B-14
AMCP 706-181
INDEX
I-I
AMCP 706-181
c D-E
Calibration techniques
Diffraction (See: Blast wave, diffraction)
Cameras Equations
drum, 9-2
one-dimensional blast waves
Dynafax, 9-2
cylindrically symmetric flow, 2-5
Eastman, 9-2
linear flow, 2-5
Hycam, 9-2, 94
time histories of blast parameters, 1-3
intermittent, 9-1
through 1-5
F
image-dissector cameras, 94, 9-5, 9-6
6-11,6-13
Computational methods
4-33
Gages
Granstrom method, 4-5, 4-6
arrival-time
through 4-5
piezoelectric, 7-15, 7-16
4-8
mechanical
through 4-24
cantilever beam, 7-22 through 7-23
4-29
squirt, 7-24, 7-25
4-8,4-9
Galvanometer oscillograph instrumentation
through 10-5
methods)
1-2
AMCP7~181
Hopkinson's scaling law (See: Blast wave, Methods of characteristics (See: Computation
scaling) al methods)
through 8-13
impulse, 7-20
Lutzky and Lehto scaling (See: Blast wave, drag, 7-17 through 7-19
1-3
AHCP 706-181
(AMCRD-TV)
JOSEPH W. PEZDIRTZ
~~,_.
Major General, USA
Chief of Staff
d
JOHN LYCAS
Colonel, GS
Chief, HQ Admin Mgt Ofc
DISTRIBUTION:
Special
Available to All: activit1 , DOD ..enci.l, and Govera-,nt aaenei frca Letterk.nny Amy Depot f Chambersburg, PA 17201.
Uncllified documents available to contractor. aad untv.rlitie. fraa National Technical Infol"lUtion Service OITIS),
Dapart.-nt of Co..erce, Sprinafield, VA 22151.
No. No.
AiicP 706 Aiep 706
100 De.lgn Guidance for Producib11itY 201 *Helicopter Ealline.rina, Plrt One, Praliainary
104 Value Engineerina Deltan
106 Eleaentl of A~nt Enainearing, 'art Ona, 202 *Helicopter Engineering, Part Tva, Detail Deaign
Sourcel of Iner,)' 203 Helicoptar Enainearing, Part Three. Qualification
107 Elements of ArIU_nt Insineering, Part Tvo, Allurance
Ballisticl 204 Helicopter Perfor'll8nce Testing
108 !leaents of Araament Insineering. 'art Three, 205 *Tiaina Sylteu and Coaponents
Weapon 5y.teas and CCJ8POnent8 210 Fuael
109 l'abl.s of the Cuaulative B1110111_1 21l(C) Fu&el, Proxiaity, Electrical, Part One (U)
Probabilities 212 (S) FUIU, Proxiaity, Electrical, Part Two (0)
110 Expert.ental Statiatlc8, Secttcm 1, Basic 213(S) rUleS, Proxi.ity. Electrical, Part Three (U)
Concepta and Analysis of Meaaurement Data 214(S) Fuzes, Proxi.ity, Electrical, Part Four (U)
111 Experi....nt.l Statistics, Section 2, Analysis 215(C) Fuzea, Proxiaity, Electrical, Part Five (U)
of Enumerative and ClaSBiUcatory Data 235 IWrdenina Weapon SYlteas Asainst IF Energy
112 Experi.ental StatiaticI, Section 3. Plaonina 238 .RecoUlels lifle Weapon Systnl
and Analylis of COliparative Experi_atl 239 *5.11 ArM Weapon SYlt~s
113 Experiaental Statistics, Section 4, Special 2401&) Grenades (U)
Topics 242 Deaign for Control of Projectile Flight Char
114 Experi.eatal Statistica, Sectioa 5, Tables acteristici (Rep 'Laces -246)
115 **Envirol'llll!ntal Series, Part One, Basic Ea 244 ~n1tion, Section 1. Artillery ~nition-
vironaental Concepts Ceneral. with Table of Conteatl, Glollary.
116 *Enviroldlental Series, Part Tva, Natural and Index for Series
Envirom-ental Factors 245(C) ~nition, Section 2. Delign for Tenlinal
117 *Envirol'llll!ntal Series, Part Three, Induced Effects (U)
Inviromaental Factors 246 +~nition, Section 3, Desiga for Control of
118 *EnvirolWlental Series, Part Four, Life Cycle PU8ht CbsrecteriaUc. (Repl4ced by -242)
Environments 247 ~nition, Section 4. Destan for Projection
119 *Environmental Series, Part Pive, Gloseary 248 +~nition, Section 5, Inlpection Aspectl of
of Environaental Tenu Artillery ~nit1on Design
120 Criteria for Environaental Control of 249 ~nition, Section 6, Kaaufacture of Metallic
Mobile Sy.t.....
COIIpOaents of Artillery ~unition
121 Packaging and Pack Enaineerina
250 Gual-~eneral
123 Hydraulic rluids
251 Muzzle Devices
124 *Reliable Military Electroaics
252 **Gun Tubel
125 Electrical Wire and Cable
253 *Breech Mechani_ Delign
.easures (U)
270 **Propellant Actuated Devices
127 Infrared Hilitary Systeas, Part One
280 Delian of Aerodynamically Stabilized Free
128(S) Infrared Military Systeas, Part Two (U)
Rockets
129 'E1ectroma8netic COIIIpaUbility (ElK:)
28l(SRD) Weapon SYltem Effectiveness (0)
130 Design for Air Tranaport and Airdrop of
282 +Propulaion and Propellant. (Repl4ced by -285)
Hateriel
283 AerodynMlics
136 **ServotlechaniSll8, Section I, Theory 291 Surface-to-Air Hislilel, Part One. System
293
294(S)
+ Surface-to-Air Miss11es, Psrt Three, Coaputers
Surface-to-Air Mis.iles, Part Four, Miss11e
140 Trajectories, Differential Effects, aad
A.......nt (U)
Data for Projectiles
295(S) Surface-to-Air Misl1les, Part Five, Counter
150 Interior Balliltics of Guns
.asures (U)
160(S) Elements of Terainal Ballistics, Part One,
162(SBD) Elements of Terminal Ballistica, Part Three, 327 Fire Control Systems-~eneral
Application to Mi8s11e and Space Targets(O) 329 Fire Control Computina Syste1ls
170(S) Armor and Its Applications (U) 335(SRD) *Design Engineers l Nuclear Effects Manual (DENIM),
175 Solid Propellants, Part One Volume I, Munitions and Weapon Syst-.s (U)
176(C) +Solid Propellants, Part Two (U) 336(SRD) *Design Engineers' Nuclear Iffects Manual (DEMEM) ,
177 Properties of Explosives of Military Volmae 11. Electronic Systems and Logistical
Interest Sy.t..... (U)
178(C) .Properties of Explosives of Military
337 (SBD) *oesign Ingineerl I Nuclear Effects Kanual (DENEM) ,
Interest, Section 2 (U) (Replaced by
Volume III, Nuclear Environment (U)
-177)
338(SRD) *Design Engineers l Nuclear Effects Hanual (DEMEM).
179 Explosive Trains
Volldle IV, !fuclear Effects (U)
185 Military Pyrotechnics, Part One, Theory and 343 Top Carriages
188 Military Pyrotechnics, Part 'our, Design of 355 The Automotive Assembly
189 Military Pyrotechnics, Part Five, l'ibliography 357 Automotive Bodies and Hulls