Documentos de Académico
Documentos de Profesional
Documentos de Cultura
Lecture Notes
Welding Technology 2
Welding Metallurgy
Table of Contents
Chapter
Subject
Page
1.
Weldability of Metals
2.
TTT - Diagrams
3.
Residual Stresses
4.
5.
21
31
44
6.
70
7.
89
8.
Welding of Aluminium
96
9.
Welding Defects
108
126
10.
1.
Weldability of Metals
1. Weldability of Metals
DIN 8580 and DIN 8595 classify welding into production technique main group 4 "Joining,
group 3.6 "Joining by welding, Figure 1.1.
Figure 1.1
Figure 1.2
1. Weldability of Metals
5
Material influence on weldability, i.e. welding
suitability, can be detailed for a better understanding in three subdefinitions, Figure
1.3.
material
characteristics
Figure 1.3
1. Weldability of Metals
A survey from quality point of view about the influence of the most important alloy elements
to some mechanical and metallurgical properties is shown in Figure 1.5.
Figure 1.5
Figure 1.6 depicts the decisive importance of the carbon content to suitability of
fusion welding of mild steels.
A guide number of flawless
fusion weldability is a carbon
content of C < 0,22 %. with
higher C contents, there is a
danger of hardening, and
welding becomes only possible by observing special
precautions (e.g. pre- and
post-weld heat treatment).
Figure 1.6
1. Weldability of Metals
In addition to material beha viour, weldability is also essentially determined through the design
of a component. The influence of the design is designated as welding safety, Figure 1.7.
Figure 1.7
and
post-weld
heat
before
welding
the
Figure 1.8
2.
TTT - Diagrams
2. TTT Diagrams
to
the
-Iron
face-centered
face-
Lattice constant
0.286 nm
at room temperature
br-eI-02-01.cdr
crystalline
structure,
ten-
Figure 2.1
etc..
In order to be able to understand the basic processes it is necessary to have a look at the
basic processes occuring in an idealized binary system. Figure 2.2 shows the state of a binary system with complete solubility in the liquid and solid state.
If the melting of the L1 alloy is cooling down, the first crystals of the composition c1 are formed
with reaching the temperature T1. These crystals are depicted as mixed crystal , since they
consist of a compound of the components A (80%) and of B (20%). Further, a melting with
the composition c0 is present at the temperature T1.
L1
L1
S
TsA
T2
T1
3
2
4
5
Temperature T
Li
So
TsB
Temperature T
- ss
A
(Ni)
br-eI-02-02.cdr
Figure 2.2
c1
c2
c0
c3
Concentration c
c4
B
(Cu)
Time t
2. TTT Diagrams
10
tribution of the components A and B in the solidified structure is homogeneous since concentration differences of the precipitated mixed crystals are balanced by diffusion processes.
The other basic case of complete solubility of two components in the liquid state and of complete insolubility in the solid state shows Figure 2.3 If two components are completely insoluble in the solid state, no mixed crystal will be formed of A and B. The two liquidus lines Li cut
in point e which is also designated as the eutectic point. The isotherm Te is the eutectic line.
If an alloy of free composition solidifies according to Figure 2.3, the eutectic line must be cut.
This is the temperature (Te) of the eutectic transformation:
S A+B (T = Te = const.).
This means that the melt at a constant temperature Te dissociates in A and B. If an alloy of
the composition L2 solidifies, a purely eutectic structure results. On account of the eutectic
reaction, the temperature of the alloy remains constant up to the completed transformation
(critical point) (Figure 2.2).
Eutectic
L1
L2
L1
TsA
1
TsB
So
Te
Li
Li
S+A
S+B
orientation
Temperature T
are
S
2
L2
structures
characteristic
between
the
3
4
B+E
c1
ce
Concentration c
Time t
br-eI-02-03.cdr
mation on transformation
behaviour in relevant specialist literature.
Figure 2.3
The definite use of the principles occurs in the iron-iron carbide diagram. Transformation behaviour of carbon containing iron in the equilibrium condition is described by the stable phase
diagram iron-graphite (Fe-C). In addition to the stable system Fe-C which is specific for an
equilibrium-close cooling, there is a metastable phase diagram iron cementite (Fe-Fe3C).
During a slow cooling, carbon precipitates as graphite in accord with the stable system Fe-C,
2. TTT Diagrams
11
while during accelerated cooling, what corresponds to technical conditions, carbon precipitates as cementite in agreement with the metastable system (Fe-Fe3C). Per definition, iron
carbide is designated as a structure constituent with cementite although its stoichiometric
composition is identical (Fe3C). By definition, cementite and graphite can be present in steel
together or the cementite
melt +
- solid solution
solid sol.
Fe3C
(cementite)
solid sol.
melt +
austenite
austenite + graphite
austenite + cementite
formation of cementite is
ledeburite
melt +
cementite
austenite
Temperature C
austenite +
ferrite
ferrite
ferrite + graphite
ferrite + cementite
perlite
stable equilibrium
metastable equilibrium
Mass % of Carbon
br-eI-02-04.cdr
melt +
graphite
melt
Figure 2.4
2. TTT Diagrams
12
The corners of the phase fields are designated by continuous roman capital letters.
As mentioned before, the system iron-iron carbide is a more important phase diagram for
technical use and also for welding techniques. The binary system iron-graphite can be stabilized by an addition of silicon so that a precipitation of graphite also occurs with increased
solidification velocity. Especially iron cast materials solidify due to their increased silicon contents according to the stable system. In the following, the most important terms and transformations should be explained more closely as a case of the metastable system.
The transformation mechanisms explained in the previous sections can be found in the binary system iron-iron carbide almost without exception. There is an eutectic transformation in
point C, a peritectic one in point I, and an eutectoidic transformation in point S. With a temperature of 1147C and a carbon concentration of 4.3 mass%, the eutectic phase called Ledeburite precipitates from cementite with 6,67% C and saturated -solid solutions with 2,06%
C. Alloys with less than 4,3 mass% C coming from primary austenite and Ledeburite are
called hypoeutectic, with more than 4,3 mass% C coming from primary austenite and Ledeburite are called hypereutectic.
If an alloy solidifies with less than 0,51 mass percent of carbon, a -solid solution is formed
below the solidus line A-B (-ferrite). In accordance with the peritectic transformation at
1493C, melt (0,51% C) and -ferrite (0,10% C) decompose to a -solid solution (austenite).
The transformation of the -solid solution takes place at lower temperatures. From -iron with
C-contents below 0.8% (hypoeutectoidic alloys), a low-carbon -iron (pre-eutectoidic ferrite)
and a fine-lamellar solid solution (perlite) precipitate with falling temperature, which consists
of -solid solution and cementite. With carbon contents above 0,8% (hypereutectoidic alloys)
secondary cementite and perlite are formed out of austenite. Below 723C, tertiary cementite
precipitates out of the -iron because of falling carbon solubility.
The most important distinguished feature of the three described phases is their lattice structure. - and -phases are cubic body-centered (CBC lattice) and -phase is cubic facecentered (CFC lattice), Figure 2.1.
2. TTT Diagrams
13
Different carbon solubility of solid solutions also results from lattice structures. The three above mentioned phases dissolve carbon interstitially, i.e. carbon is embedded between the
iron atoms. Therefore, this types of solid solutions are also named interstitial solid solution.
Although the cubic face-centred lattice of austenite has a higher packing density than the cubic body-centred lattice, the void is bigger to disperse the carbon atom. Hence, an about 100
times higher carbon solubility of austenite (max. 2,06% C) in comparison with the ferritic
phase (max. 0,02% C for -iron) is the result. However, diffusion speed in -iron is always at
least 100 times slower than in -iron because of the tighter packing of the -lattice.
Although - and -iron show the same lattice structure and properties, there is also a difference between these phases. While -iron develops of a direct decomposition of the melt (S
), -iron forms in the solid phase through an eutectoidic transformation of austenite (
+ Fe3C). For the transformation of non- and low-alloyed steels, is the transformation of ferrite of lower importance, although this -phase has a special importance for weldability of
high alloyed steels.
Unalloyed steels used in industry are multi-component systems of iron and carbon with alloying elements as manganese, chromium, nickel and silicon. Principally the equilibrium diagram Fe-C applies also to such multi-component systems. Figure 2.5 shows a schematic cut
through the three phase system Fe-M-C.
During precipitation, mixed carbides of the general composition M3C develop. In contrast to
the binary system Fe-C, is
the three phase system
Fe-M-C characterised by a
Ac3
three-phase field + +
M3C. The beginning of the
transformation of + M3C
to is marked by Aclb, the
end by Acle. The indices b
and e mean the beginning
br-eI-02-05.cdr
Figure 2.5
2. TTT Diagrams
14
The described equilibrium diagrams apply only to low heating and cooling rates. However,
higher heating and cooling rates are present during welding, consequently other structure
types develop in the heat
affected zone (HAZ) and in
the weld metal. The strucC
by
transformation
but
at
different
heating
br-eI-02-06.cdr
representation
transformation
of
the
processes
Figure 2.6
during isothermal austenitizing shows Figure
ASTM4; L=80m
ASTM11; L=7m
20m
ite means both, low carbon containing austenite is formed in areas, where ferrite was
present before transformation, and carbon-rich
austenite is formed in areas during transforma-
Temperature
tion, where carbon was present before transformation. During sufficiently long annealing
times, the concentration differences are balanced by diffusion, the border to a homogeneous austenite is passed. A growing of the
Time
br-er02-07.cdr
ISF 2002
Figure 2.7
2. TTT Diagrams
15
The influence of heating rate on austenitizing is shown in Figure 2.7. This diagram must only
be read along the sloping lines of the same heating rate. For better readability, a time pattern
was added to the pattern of the heating curves. To elucidate the grain coarsening during austenitizing, two microstructure photographs are shown, both with different grain size classes to
ASTM.
Figure 2.8 shows the relation between the TTA and the Fe-C diagram. It's obvious that the
Fe-C diagram is only valid for infinite long dwell times and that the TTA diagram applies only
for one individual alloy.
Figure 2.9 shows the dif-
Ac3
ferent
time-temperature
Ac1e
and
Ac1b
subsequent
cooling
down.
The heating period is composed of a continuous and
an isothermal section.
br-eI-02-08.cdr
Figure 2.8
Ac3
continuous
Ac1e
control
Ac1b
During
temperature
isothermal
control
br-eI-02-09.cdr
Figure 2.9
2. TTT Diagrams
16
the area of the homogeneous and stable austenite down to room temperature.
2. : During isothermal temperature control a quenching out of the area of the austenite is
carried out into the area of the metastable austenite (and/or into the area of martensite), followed by an isothermal holding until all transformation processes are completed. After transformation will be cooled down to room temperature.
Figure
2.10
shows
the
time-temperature diagram
of a isothermal transformation of the mild steel Ck 45.
Read such diagrams only
along the time-axis! Below
the Ac1b line in this figure,
there is the area of the metastable austenite, marked
with
an
A.
The
areas
lines which limit the area to the left mark the beginning of the formation of the respective
structure. The lines which limit the area to the right mark the completion of the formation of
the respective structure. Because the ferrite formation is followed by the perlite formation, the
completion of the ferrite formation is not determined, but the start of the perlite formation.
Transformations to ferrite and perlite, which are diffusion controlled, take place with elevated
temperatures, as diffusion is easier. Such structures have a lower hardness and strength, but
an increased toughness.
Diffusion is impeded under lower temperature, resulting in formation of bainitic and martensitic structures with hardness and strength values which are much higher than those of ferrite
and perlite. The proportion of the formed martensite does not depend on time. During
quenching to holding temperature, the corresponding share of martensite is spontanically
formed. The present rest austenite transforms to Bainite with sufficient holding time. The right
2. TTT Diagrams
17
detail of the figure shows the present structure components after completed transformation
and the resulting hardness at room temperature.
Figure 2.11 depicts the graphic representation of the TTT diagram, which is more important
for welding techniques. This is the TTT diagram for continuous cooling of the steel Ck 15.
The diagram must be read along the drawn cooling passes. The lines, which are limiting the
individual areas, also depict the beginning and the end of the respective transformation.
Close to the cooling curves, the amount of the formed structure is indicated in per cent, at the
end of each curve, there is the hardness value of the structure at room temperature.
approximately
processes
are
Figure 2.11
completely
marte nsitic
transformation
is
carried
behaves completely
2. TTT Diagrams
18
eutectoid ferrite formation cannot still be carried out (see also Figure 2.3).
The term of the figures 2.9 to 2.11 "austeniti zing temperature means the temperature,
where the workpiece transforms to an auste nitic microstructure in the course of a heat
treatment. Dont mix up this temperature with
the AC3 temperature, where above it there is
only pure auste nite. In addition you can see
that only martensite is formed from the austenite, provided that the cooling rate is sufficiently
high,
formation
of
any
other
Figure 2.14
Figure 2.15
2. TTT Diagrams
19
Figure 2.14 shows schematically how the TTT diagram is modified by the chemical composition of the steel.
The influence of an increased austenitizing temperature on transformation beha viour shows
Figure 2.15. Due to the higher hardening temperature, the grain size of the austenite is
higher (see Figure 2.6 and 2.7).
martensite
and
with
higher
austenitizing
proportion
is
increased
2. TTT Diagrams
20
Due to the strong influence of the austenitizing temperature to the transformation behaviour
of steel, the welding technique uses special diagrams, the so called Welding -TTT-diagrams.
They are recorded following the welding temperature cycle with both, higher austenitizing
temperatures (basically between 950 and 1350C) and shorter a usteniti zing times.
You find two examples in Figures 2.16 and 2.17.
Figure 2.18
3.
Residual Stresses
3. Residual Stresses
22
pressure
tension
grinding disk
3.2
details
the
pressure
weld
ISF 2002
br-eI-03-01e.cdr
Figure 3.1
stresses,
which
wear
production
e.g.
polyphase
systems,
non-metallic
inclusions,
grid defects
mechanical
thermal
chemical
e.g.
partial-plastic
deformation of
notched bars or
close to
inclusions,
fatigue strain
e.g. thermal
residual stresses
due to
operational
temperatur fields
e.g.
H-diffusion
under
electro-chemical
corrosion
nd
forming
deforming
separating
joining
plating
e.g. thermal
residual
stresses
residual stresses
due to
inhomogenuous
deformationanisotropy
residual
stresses due to
machining
residual
stresses due to
welding
layer residual
stresses
changing material
characteristics
induction hardening,
case hardening,
nitriding
br-eI-03-02e.cdr
st
type.
The
formation
of
residual
stresses in a transition-free
Figure 3.2
3. Residual Stresses
23
steel cylinder is shown in Figures 3.5. and 3.6. During water quenching of the homogeneous
heated cylinder, the edge of the cylinder cools down faster than the core. Not before 100
seconds have elapsed is the temperature across the cylinder's cross section again
s III
tension s
s II
sI
Residual stresses of the II. type are almost homogenuous across small
material areas (one grain or grain area). Internal forces and torques
related to residual stresses of the II. type are in an equilibrium across
a sufficient number of grains. When interfering with this equilibrium,
macroscopic dimension changes may develop.
x
0
grain boundaries
sE = s I + sII
sIII
<
<
<
sI
sII
sIII
ISF 2002
br-er03-03e.cdr
ISF 2002
br-er03-04e.cdr
Figure 3.3
Figure 3.4
Temperature
750
35 mm diameter
water cooling
500
250
MS
1 edge
2 50 % radius
3 core
1s
5s
15 s
800
1000
Temperature
0s
10 s
700
20 s
600
25 s
500
35 s
400
45 s
300
53 s
200
0 -2
10
68 s
10-1
10-0
101
102
Cooling time
103
s 104
100
280 s
0
17,5
14
10,5
0
3,5
3,5
10,5
Radius
ISF 2002
br-eI-03-05e.cdr
Temperature in a Cylinder
During Water Cooling
mm
17,5
3. Residual Stresses
24
tensile stress builds up at the edge areas while the core is exposed to pressure stress. Resulting volume differences between core and edge are balanced by elastic and plastic deformations. When cooling is completed, edge and core are on the same temperature level, the
plastically stretched edge now supports the unstressed core, so that pressurestresses are
present in the edge areas and tensile residual stresses in the core.
300
tension
pressure
N/mm
200
tension
tension
pressure
tension
pressure
100
-100
-200
B'
tension
pressure
br-er03-06e.cdr
-300
0
ISF 2002
400
600
br-er03-07e.cdr
ISF 2002
Figure 3.6
200
Figure 3.7
These changes are principally shown once again in Figure 3.7 with the 3-rod model. A warming of the middle rod causes at first an elastic expansion of the outer rods, the inner rod is
exposed to pressure stress (line A-B). Along the line B-C the rod is plastically deformed, because pressure stresses have exceeded the yielding point. At point C, the cooling of the rod
starts, it is exposed to tensile stress due to shrinking. Along the line D-E the rod is plastically
deformed due to the influence of the counter members beeing in tension. At the point E the
system has cooled down to its initial temperature. This point represents the remaining residual stress condition of this construction. If heating is stopped before point C is reached and
cooled down to the initial temperature, then stress increase in the centre rod will be in parallel
3. Residual Stresses
25
with the elastic areas. Starting with point B, the same residual stress condition is present as
in a case of heating up to a temperature above 600C.
Figure 3.8 divides the development of residual stresses in welded seams in three different
mechanisms.
Shrinking stresses: these are stresses formed through uniform cooling of the seam.
Caused by expansion restriction of the colder areas at the edge of the weld and base material , tensile stresses develop along and crosswise to the seam.
Quenching stresses: If cooling is not homogenous, the surface of the weld cools down
faster than the core areas. If the high-temperature limit of elasticity is exceeded due to buildup stress differences, pressure stresses will be present at the weld surface after cooling. In
contrast, the core shows tensile stresses in cold condition (see also Figure 3.6).
Transition stresses: Transitions in the ferrite and perlite stage cause normally only residual
stresses, because within this temperature range the yield strength of the steel is so low that
generated stresses can be undone by plastic deformations.
This is not the case with transitions in the Bainite and martensite stage. A transition of the
austenite causes an increase in volume (transition cfc in cbc, the cfc lattice has a higher density, additional volume increase through lat+y
-x
the edge areas happens earlier than the transition of the slower cooling core, plastic de-
-y
2. Quenching stresses
1. Shrinking stresses
-x
+x
-x
+s +y
-s -y
+x
4. Overlap options
of case 1., 2. and 3.
+s +y
+s +y
inhomogenuous
transformation
-x
+x
-x
+x
homogenuous
transformation
-s -y
br-er03-08.cdr
-s -y
ISF 2002
3. Residual Stresses
26
Seam
Stress
distribution sX
1. cut A-A
DT ~ 0
stress-free
2. cutt B-B
weldpool
B
area of plastic
deformations
pressure
3. cut C-C
M'
4. cut D-D
residual stresses
DT = 0
br-er03-09e.cdr
Figure 3.9
areas close to the seam, the weld metal
shows tensile stresses, the adjacent areas
compression stresses. In cut D-D is the temperature completely balanced, a residual
stress condition is recognised as shown in
the lower right figure.
31 15 mm
15 mm
material
S235JR (St 37)
103
a
a
1.
a = 100 mm
s = 800 N/mm
2.
a = 150 mm
s = 530 N/mm
3.
a = 200 mm
s = 400 N/mm
according to Hooke:
4.
a = 250 mm
s = 300 N/mm
= E
5.
a = 300 mm
s = 270 N/mm
br-er03-10e.cdr
ISF 2002
Shrinking Stresses in a
Firmly Clamped Plate
3. Residual Stresses
27
stant joint volume will shrinking and l always have the same value. Thus the elongation
depends only on the value a. The smaller the a is chosen, the higher are the resulting
stresses.
Effects of transition on cooling can be estimated from Figure 3.11. Here curves of temperature- and length-changes of ferritic and austenitic steels are drawn. It is clear that a ferritic
lattice has a higher volume than an austenitic lattice at the same temperature.
A steel which transforms from austenite to one of the ferrite types increases its volume at the
critical point. This sudden rise in volume can be up to 3% in the case of martensite formation.
Longitudinal expansion Dl
firm
clamping
force
sensor
el
el
thermo
couples
links
ste
nit
ic
ste
tic
rri
fe
ste
au
to calculator
1000
600
800
14
m
tra ild
ns ste
fo el
rm w
at ith
io
n
800
200
Temperature
Force
elektrode
400
600
heat affected
zone
400
force
0
Temperature [C]
200
temperature
-200
0 -1
10
100
101
102
103
104
105
Time
br-er03-12e.cdr
br-er03-11e.cdr
ISF 2002
ISF 2002
Force Measurement
During Cooling of a Weld
Longitudinal Expansion of
Various Steels
Figure 3.11
Figure 3.12
To record the effects of this behaviour on the stress condition of the weld, sample welds are
carried out in the test device outlined in Figure 3.12. Thermo couples measure the T-t curve
at the weld seam, a force sensor records the force which tries to bend the samples.
The lower picture shows the results of such a test.
The temperature behaviour at the fusionline as well as the force necessary to hold the sample over the time is plotted.
3. Residual Stresses
28
In the temperature range above 600C the force sensor registers a tensile force which is
caused by the shrinking of the austenite. Between 600 and 400C a large drop in force can
be seen, which is caused by the transition of the austenite. The repeated increase of the
force is based on further shrinking of the ferrite.
With the help of TTT diagrams
of base material and welding
steel
austenitic
consumable,
consumable
electrode
austenitic
austenitic
surface weld
surface weld
the
transition
sample shape
(V-groove, 60)
type of
welding
surface weld
pressure
tension
ISF 2002
br-eI-03-13e.cdr
Figure 3.13
which part by transition in the heat affected
542'
28'
151'
zone (HAZ).
These results can be used to determine the
longitudinal residual stresses transversal to
the joint, as shown in Figure 3.13. During
140
Angle change
%
100
80
60
40
20
f = 1
f = 3
f = 7
f = 13
a=7
a=9
br-er03-14e.cdr
ISF 2002
Figure 3.14
a = 12,5
to a StE 70, then there will be pressure residual stresses in the weld metal and tensile
residual stresses in the HAZ.
3. Residual Stresses
29
If parts to be welded are not fixed, the shrinking of the weld will cause an angular distortion of
the workpieces, Figure 3.14 . If the workpieces can shrink unrestricted in this way, the remaining residual stresses will be much lower than in case with firm clamping.
Methods to determine residual stresses can be divided
into
destructive,
plan
section
nona
as conditionally destructive,
workpiece
ISF 2002
br-eI-03-15e.cdr
are
Figure 3.15
then
measured by wire strain gauges. An essential advantage of the borehole method is the very
small material removal, the diameter of the borehole is only 1 to 5 mm, the bore depth is 1- to
2-times the borehole diameter.
The disadvantage here is that only surface elongations can be measured, thus the results are
limited residual stresses in the surface area of the workpiece.
wire expansion gauge
b(sb)
45
c(ec)
45
t0.21
a(sa)
D1.58
s1 (z)
s2 (z)
measurement point
ISF 2002
br-eI-03-16e.cdr
Figure 3.16
3. Residual Stresses
30
destructive
measurement
mechanical
deformations
A
E
A
E
A
E
A
E
A
E
optical
procedures
A
E
magnetic
thermal
processes
ultra sonic
breaking-up
bending deflection
ring groove
methods.
cam web
drilling out
turning off
causes
others
x - ray
mechanically - electrically
partial
optical
procedures
complete
bore hole
non-destructive
ring core
A
E
A
E
A
E
A
E
A
E
A
E
surfacetreatment
A - general application
E - further development desired
ISF 2002
br-eI-03-17e.cdr
Figure 3.17
assumption of
stress distribution
measured variable
residual stresses
cutting in layers
procedures
biaxial
any
of
bending deflection f
curves
reduced curves
sy
sz
tzy
tear f
cutting-in
uniaxial
locally different
linear, tensile residual
stresses on top, down
pressure stresses
drilling
e45
eT
eL
slitting
0.46f
tripleaxial
independent of
smple length sL, sT, sR
uniaxial
linear
symmetrically with
reference to rod axis
length change eL
circumference change eT
tear f
sL
sT
sR
ISF 2002
br-eI-03-18e.cdr
Figure 3.18
4.
Heat Treatment and
its Function During Welding
32
(process 1, 2, 7, and 8), influence directly the mechanical properties of the weld. In addition,
the chemical composition of the weld metal and adjacent base material are also influenced
by the processes 3 to 6.
elevated
areas
because
of
33
In Figure 4.4 you can see how much the formation of the individual structure zones and
the zones of unfavourable mechanical properties can be influenced.
Applying an electroslag one pass weld of a
200 mm thick plate, a HAZ of approximately
30 mm width is achieved. Using a three pass
technique, the HAZ is reduced to only 8 mm.
Figure 4.3
Figure 4.4
34
Figure 4.5
Figure 4.6
constant
stop-
temperature-time-
35
Figure 4.7 shows in the detail to the right a T-t course of coarse grain heat treatment of an
alloy containing 0,4 % C. A coarse grain heat treatment is applied to create a grain size as
large as possible to improve machining properties. In the case of welding, a coarse grain is
unwelcome, although unavoidable as a consequence of the welding cycle. You can learn
from Figure 4.7 that there are two methods of coarse grain heat treatment. The first way is to
austenite at a temperature close above A3 for a couple of hours followed by a slow cooling
process. The second method is very important to the welding process. Here a coarse grain is
formed at a temperature far above A 3 with relatively short periods.
Figure 4.8 shows schematically time-temperature
behaviour
in
TTT-
mally.
Mixed
or
isother-
types
like
The most important heat treatment methods can be divided into sections of annealing, hardening and tempering, and these single processes can be used individually or combined. The
normalising process is shown in Figure 4.9. It is used to achieve a homogeneous ferrite perlite structure. For this purpose, the steel is heat treated approximately 30C above Ac3
until homogeneous auste nite evolves. This condition is the starting point for the following
hardening and/or quenching and tempering treatment. In the case of hypereutectoid steels,
austenisation takes place above the A1 temperature. Heating-up should be fast to keep the
austenite grain as fine as possible (see TTA-diagram, chapter 2). Then air cooling follows,
leading normally to a transformation in the ferrite condition (see Figure 4.8, line 1; formation
of ferrite and perlite, normalised micro-structure).
36
To harden a material, austenisation and homogenisation is carried out also at
30C above AC3. Also in
this case one must watch
that the austenite grains
remain as small as possible. To ensure a complete
transformation to marte nsite, a subsequent quenc hing
follows
until
the
ing quenching must be high enough to cool down from the auste nite zone directly into the
martensite zone without any further phase transitions (curve 2 in Figure 4.8). Such quenching
processes build-up very high thermal stresses which may destroy the workpiece during hardening. Thus there are variations of this process, where perlite formation is suppressed, but
due to a smaller temperature gradient thermal stresses remain on an uncritical level (curves
3 and 4 in Figure 4.8). This
can be achieved in practice
for example- through stopping
water
quenc hing
Figure 4.10
37
Figure 4.11 shows the quenching and tempering procedure. A hardening is followed by another heat treatment below
Ac1. During this tempering
process, a break down of
C
austenite
900
A1
700
ferrite +
perlite
takes
place.
A3
Temperature
austenite
+ ferrite
Temperature
martensite
quenching
slow
cooling
500
300
0,4
0,8
C-Content
Time
br-eI-04-11.cdr
ness.
Figure 4.11
Figure 4.12 shows the procedure of soft-annealing. Here we aim to adjust a soft and suitable
micro-structure for machining. Such a structure is characterised by mostly globular formed
cementite particles, while the lamellar structure of the perlite is resolved (in Figure 4.12
marked by the circles, to the left: before, to the right: after soft-annealing). For hypoeutectic
steels, this spheroidizing of cementite is achieved by a heat treatment close below A1. With
these steels, a part of the cementite bonded carbon dissolves during heat treating close below A1, the remaining cementite lamellas transform with time into balls, and the bigger ones
grow at the expense of the
smaller ones (a transfor-
austenite
+ ferrite
oscillation annealing
+ / - 20 degrees around A 1
10 to 20C
below A1
A3
A1
Temperature
900
Temperature
austenite
700
ferrite +
perlite
or
500
condition).
300
0,4
0,8
C-Content
Time
cementite
Soft Annealing
grain boundaries.
Figure 4.12
38
spheroidizing
of
avoids
microstructure
Figure 4.13
Figure 4.13 shows the principle of a stress-relieve heat treatment. This heat treatment is
used to eliminate dislocations which were caused by welding, deforming, transformation etc.
to improve the toughness of a workpiece. Stress-relieving works only if present dislocations
are able to move, i.e. plastic structure deformations must be executable in the micro-range. A
temperature
the
increase
commonly
is
used
possible
be-
Figure 4.14
39
with
welding.
is
applied
before
40
level of hardening. If the material is hold at a temperature above MS during welding (curve 3),
then the martensite formation will be completely suppressed (see Figure 4.8, curve 4 and 5).
Figure 4.17
4.18. During a welding time
tS a drop of the working
temperature TA occurs. A
further air cooling is usually
carried out, however, the
cooling rate can also be
reduced by cove ring with
heat insulating materials.
constant
temperature.
working
This
is
41
achieved through further
warming during welding to
avoid a drop of the working
temperature. In Figure 4.19
is this case (dashed line,
TA needs not to be above
MS) as well as the special
case of isothermal welding
illustrated. During isothermal welding, the workpiece
is heated up to a working
temperature
Figure 4.19
above
MS
after welding until a transformation of the austenitised areas has been completed. The aim of
isothermal welding is to cool down in accordance with curve 3 in Figure 4.16 and in this way,
to suppress martensite formation.
42
welding
should
Figure 4.21
Figure 4.22
43
tion effects in the HAZ. The coarse grain zone with its unfavourable mechanical properties is
only present in the HAZ of the last layer. To achieve optimum mechanical values, welding is
not carried out to Figure 4.22. As a rule, the same welding conditions should be applied for all
passes and prescribed t8/5 times must be kept, welding of the next pass will not be carried
out before the previous pass has cooled down to a certain temperature (keeping the interpass temperature). In addition, the workpiece will not heat up to excessively high temperatures.
Figure 4.23 shows a nomogram where working te mperature and minimum and maximum
heat input for some steels can be interpreted, depending on carbon equivalent and wall thickness.
If e.g. the water quenched and tempered fine grain structural steel S690QL of 40 mm wall
thickness is welded, the following data can be found:
hatched
soaking
area,
area
called
must
be
Figure 4.23
5.
Welding Plain and
Low Alloy Steels
45
tiD
o
e
f
n
g
n
u
e
lti
E
e
d
fs
n
iB
rg
ra
h
m
e
c
d
tlS
a
n
d
e
r
ta
le
h
S
s
h
Z
s
a
ic
n
m
e
u
tilh
e
rg
S
cn
rh
i
t
o
s
n
m
e
g
tz
u
le
e
tS
sl
n
d
h
a
n
d
e
r
tilh
e
rg
S
le
-rg
ti
h
lu
t
Q
a
si
o
g
n
u
ltce
irh
E
S
sld
n
r
t
p
g
h
u
a
c
H
u
-g
n
ile
rs
k
a
s
tu
e
Q
a
lid
E
h
e
l-rg
it
r-e
0
d
5
c1
.
e
h
l
lh
sd
E
e
2
0
S
F
I
unalloyed steels
stainless steels
grades
in
accordance
with
their
stainless steels
other, alloyed steels
br-er05-01.cdr
Figure 5.1
In accordance with the chemical compoDetermined element
aluminium
boron
Bi
bismuth
Co cobalt
Cu copper
Cr
limit value
Mass fraction in %
chromium
La
lanthanides
(rated individually)
Mn manganese
Mo molybdenum
Nb niobium
Ni
nickel
Pb lead
Se selenium
Si
silicon
Te
tellurium
Ti
V
titanium
vanadium
tungsten
Zr zirconium
Others (with the exception
of carbon, phosphorus,
sulphur, nitrogen)
(Each)
a) If just the highest value has been determined for
mangenese, the limit value us 1,80% and the 70%-rule
does not apply.
br-er05-02.cdr
ISF 2004
in Figure 5.2.
Figure 5.2
46
As far as the main quality classes are concerned, the steels are classified in accordance with their main characteristics and main application properties into unalloyed,
stainless and other alloyed steels.
As regards unalloyed steels a distinction is made between unalloyed quality steels
and unalloyed high-grade steels.
Regarding unalloyed quality steels, prevailing demands apply, for example, to the
toughness, the grain size and/or the forming properties.
Unalloyed high-grade steels are characterised by a higher degree of purity than
unalloyed quality steels, particularly with regard to non-metal inclusions. A more
precise setting of the chemical composition and special diligence during the manufacturing and monitoring process guarantee better properties. In most cases these
steels are intended for tempering and surface hardening.
Stainless steels have a chromium mass fraction of at least 10,5 % and maximally
1,2 % of carbon. They are further classified in accordance with the nickel content and
the main characteristics: corrosion resistance, heat resistance and creep resistance.
Other alloyed steels are classified into alloyed quality steels and alloyed high-grade
steels.
Special demands are put on the alloyed quality steels, as, for example, to toughness,
grain size and/or forming properties. Those steels are generally not intended for
tempering or surface hardening.
The alloyed high-grade steels comprise steel grades which have improved properties
through precise setting of their chemical composition and also through special manufacturing and control conditions.
47
The European Standard DIN EN 10027-1 (September 1992) stipulates the rules for
the designation of the steels by means of code letters and identification numbers.
The code letters and identification numbers give information about the main application field, about the mechanical or physical properties or about the composition.
The code designations of the steels are divided into two groups. The code designations of the first group refer to the application and to the mechanical or physical
properties of the steels. The code designations of the second group refer to the
chemical composition of the steels.
l
E = Engineering steels
e.g. E295, E360
B = Reinforcing steels
e.g. B500A, B500B
Y = Prestressing steels
e.g. Y1770C, Y1230H
T = Black plate and tin plate and strips and also specially
chromium-plated plate and strip
e.g. TH550, TS550
br-er05-03.cdr
ISF 2004
Figure 5.3
48
An example of the code designation structure with reference to the usage and the
mechanical or physical properties for steels in structural steel engineering is explained in Figure 5.4.
Figure 5.4
49
For designating special features of the steel or the steel product, additional symbols
are added to the code designation. A distinction is made between symbols for special demands, symbols for the type of coating and symbols for the treatment condition. These additional symbols are stipulated in the ECISS-note IC 10 and depicted
in Figures 5.5 and 5.6.
Symbol1)2)
Coating
+A
+ AR
+ AS
+ AZ
+ CE
+ Cu
+ IC
+ OC
+S
+ SE
+T
+ TE
+Z
+ ZA
+ ZE
+ ZF
+ ZN
hot dipped
aluminium, cladded by rolling
coated with Al-Si alloy
coated with Al-Tn alloy (>50% Al)
electrolytically chromium-plated
copper-coated
inorganically coated
organically coated
hot-galvanised
electrolytically galvanised
upgraded by hot dipping with a lead-tin alloy
electrolytically coated with a lead-tin alloy
hot-galvised
coated with Al-Zn alloy (>50% Zn)
electrolytically galvanised
diffusion-annealed zinc coatings (galvannealed, with diffused Fe)
nickel-zinc coating (electrolytically)
1
2
) The symbols are separated from the preceding symbols by plus-signs (+)
) In order to avoid mix-ups with other symbols, the figure S may precede,
br-er-05-05.cdr
Figure 5.5
1 2
Symbol ) )
treatment condition
+A
+ AC
+C
softened
annealed for the production of globular carbides
work-hardened (e.g., by rolling and drawing), also a distinguishing
mark for cold-rolled narrow strips)
cold-rolled to a minimum tensile strength of nnn MPa/mm
cold-rolled
thermoformed/cold formed
slightly cold-drawn or slightly rerolled (skin passed)
quenched or hardened
treatment for capacity for cold shearing
solution annealed
untreated
+ Cnnn
+ CR
+ HC
+ LC
+Q
+S
+ ST
+U
1
2
) The symbols are separated from the preceding symbols by plus-signs (+)
) In order to avoid mix-ups with other symbols, the figure T may precede,
br-er-05-06.cdr
Figure 5.6
50
Figure 5.7 shows an example of the novel designation of a steel for structural steel
engineering which had formerly been labelled St37-2.
The steel St37-2 (DIN 17100) is, according to the new standard (DIN EN 10027-1),
designated as follows:
S235 J 2 G3
further property
(RR = normalised)
br-er-05-07.cdr
Figure 5.7
Steel
Stahl
S355J0
(St 52-3)
S500N
(StE500)
P295NH
(HIV)
S355J2G1W
(WTSt510-3)
S355G3S
(EH36)
Si
Mn
Cr
Al
Cu
Mo
Ni
Nb
0,20
0,55
1,60
0,040
0,040
0,009
0,035
0,030
0,30
0,020
0,20
0,020
0,1
0,05
0,22
0,21
0,26
0,35
0,05
0,05
0,15
0,035
0,40 0,80
0,25 0,5
0,30
0,65
0,02 0,12
0,18
0,05
0,6
Tensile
strength
Zugfestigkeit
RmRm
[N/mm]
Steel
Stahl
elongation
after fracture
Bruchdehnung
A A
[%]
impact
energy AVV
Kerbschlagarbeit
[J]
-20C
0C
S355J2G3
(St 52-3)
S500N
(StE500)
P295NH
(HIV)
S355J2G1W
(WTSt510-3)
S355G3S
(EH36)
510-680
355
20-22
27
31-47
610-780
500
16
460-550
285
>18
510-610
355
22
400-490
355
>22
27
21-39
49 (bei +20C)
76 (bei -10C)
ISF 2004
br-er-05-08.cdr
Figure 5.8
Figure 5.8 depicts the chemical composition and the mechanical parameters of different steel grades. The figure explains the influence of the chemical composition on
the mechanical properties.
51
The steel S355J2G2 represents the basic type of structural steels which are nowadays commonly used. Apart from a slightly increased Si content for desoxidisation it
this an unalloyed steel.
S500N is a typical fine-grained structural steel. A very fine-grained microstructure
with improved tensile strength values is provided by the addition of carbide forming
elements like Cr and Mo as well as by grain-refining elements like Nb and V.
The boiler steel P295NH is a heat-resistant steel which is applied up to a temperature
of 400C. This steel shows a relatively low strength but very good toughness values
which are caused by the increased Mn content of 0,6%.
S355J2G1W is a weather-resistant structural steel with mechanical properties similar
to S355J2G2. By adding Cr, Cu and Ni, formed oxide layers stick firmly to the workpiece surface. This oxide layer prevents further corrosion of the steel.
S355G3S belongs to the group of shipbuilding steels with properties similar to those
of usual structural steels. Due to special quality requirements of the classification
companies (in this case: impact energy) these steels are summarised under a special
group.
52
The steel grades are classified into four subgroups according to the chemical composition (Fig 5.9):
Unalloyed steels (except free-cutting steels) with a Mn content of < 1 %
Unalloyed steels with a medium Mn content > 1 %, unalloyed free-cutting
steels and alloyed steels (except high-speed steels) with individual alloying
element contents of less than 5 percent in weight
Alloyed steels (except high-speed steels), if, at least for one alloying element
the content is 5 percent in weight
High-speed steels
The unalloyed steels with Mn conUnalloyed steels (Mo content < 1%)
C45
0,45% Carbon
Carbon
10CrMo9-10
C=10/100=0,10%
Cr=9/4=2,25%
element
Mo=10/10=1%
factor
C, Ce, N, P, S
10
100
1000
Table 5.1
X10CrNi18-10
Legiert
C=10/100=0,1%
Cr=18%
Ni=10%
High-speed steels
HS 2-9-1-8
W=2%
Mo=9%
V=1%
Co=8%
br-er05-09.cdr
ISF 2004
Codes according
to the chemical composition
Figure 5.9
for the medium content of the respective alloying element, the content had been
multiplied by the factor as indicated in Fig. 5.9/Table 5.1 and rounded up to the next
whole number.
53
The alloyed steels are labelled with the code letter X, a number which again complies with the hundredfold of the mean value of the range stipulated for the carbon
content, the chemical symbols of the alloying elements, ordered according to decreasing contents of the elements and numbers which in sequence of the designating
alloying elements refer to their content.
High-speed steels are designated with the code letter HS and numbers which, in the
following sequence, indicate the contents of elements:: tungsten (W), molybdenum
(Mo), vanadium (V) and cobalt (Co).
XX
XX (XX)
Sequential number
The digits inside the brackets are intended
for possible future demands.
Steel group number (see Fig. 5.10)
Material main group number (1=steel)
Figure 5.10 specifies the material numbers for the material main group steel.
Figure 5.10
54
55
The influence of the austenite grain size on the transformation behaviour has been
explained in Chapter 2. Figure 5.11 shows the dependence between grain size of the
austenite which develops during the welding cycle, the distance from the fusion line
and the energy-per-unit length from the welding method. The higher the energy-peruntil
length,
the
13
Austenite grain size index according
to DIN 50601
11
12
of
36
the
creases.
HAZ
in-
Such
coarsened austen-
0,2
0,4
0,6
Distance of the fusion line
0,8
mm
1,0
ISF 2004
br-er-05-11.cdr
Figure 5.11
With fine-grained structural steels it is tried to suppress the grain growth with alloying
elements. Favourable are nitride and carbide forming alloys. They develop precipitations which suppress undesired grain growth. There is, however, a limitation due to
the solubility of these precipitations, starting with a certain temperature, as shown in
Figure 5.12. Steel 1 does not contain any precipitations and shows therefore a continuous grain growth related to temperature. Steel 2 contains AIN precipitations which
are stable up to a temperature of approx. 1100C, thus preventing a growth of the
austenite grain.
56
With
mm
1
8
6
10
8
6
-1
6 10-3
temperatures,
-2
Steel 1
Steel 2
Steel 3
Steel 4
12
900
1000
1100
1200
Austenitization temperature
1300
1400
Steel
%C
% Mn
% Al
%N
% Ti
0,21
1,16
0,004
0,010
0,17
1,35
0,047
0,017
0,18
1,43
0,004
0,024
0,067
0,19
1,34
0,060
0,018
0,140
ISF 2004
br-er05-12.cdr
specified
According
to
the
1
d
900
N/mm
800
Temperature in C:
700
-193
-185
600
-170
-100
300
-40
200
diameter d.
-180
-155
500
400
+20
0
5
6
-1/2
Grain size d
grain
for
is
mm
-1/2
10
Connection between
yield point and grain size
boundary
resistance K
measure
The
ISF 2004
br-er-05-13.cdr
material.
Hall-Petch-law:
the
above-mentioned
inversely
in
Rel = i + K
Figure 5.12
the
these
-4
10
-2
10
8
higher
Figure 5.13
the
influence of the grain size on the forming mechanisms. Apart from this increase of the
yield point, grain refinement also results in improved toughness values. As far as
57
structural steels are concerned, this means the improvement of the mechanical properties without any further alloying. Modern fine-grained structural steels show improved mechanical properties with, at the same time, decreased content of alloying
elements. As a consequence of this chemical composition the carbon equivalent
decreases, the weldability is improved and processing of the steel is easier.
The major advanSteel
type
Stahlsorte
S235JR
(St37-2)
S355J2G3
(St52-3)
S690Q
(StE690)
S890Q
(StE890)
S960Q
(StE960)
Ratio
Verhltnis
S235JR - S960Q
N/mm2
215
345
690
890
960
1:5
Plate
thickness
Blechdicke
mm
50
31
14,4
11
10
5:1
Yield
point
Streckgrenze
Weld cross-section
Nahtquerschnitt
mm2
870
370
100
60
50
17 : 1
Welding
wire 1.2
Schweidraht
1.2
mm
SG2
SG3
NiMoCr
X 90
X 96
Welding
wire costs
Schweidrahtkosten
Ratio
Verhltnis
2,4
3,2
3,3
1 : 3,3
Steel costs
Stahlkosten
Ratio
Verhltnis
1,2
1,9
2,3
2,4
1 : 2,4
Ratio
Verhltnis
5,3
2,3
1,5
1,16
5,3 : 1
Special
weld costs
Spez. Schweinahtkosten
Ratio
Verhltnis
12
5,1
1,8
1,18
12 : 1
Boundary condition:
Randbedingungen:
tages of microalloyed
fine-grained
structural steels in
comparison
with
conventional structural
5:1
steels
shown
welding
process = MAG
Schweiverfahren
= MAG
in
are
Figure
Deposition
rate = 3 kg=welding
wire/h, weld /shape
X -60 X - 60
Abschmelzleistung
3 kg Schweidraht
h, Nahtform
Costs
labour
and equipment == 60
30/h
Lohn-ofund
Maschinenkosten
DM / h
Special
costs = weld filler
materials + welding
Spez. weld
Schweinahtkosten
= Schweizusatzwerkstoffe
+ Schweien
considerably better
Berechnungsgrundlage
=szul = Re / 1.5
Calculation
base = szul = Re/1.5
ISF 2004
br-er-05-14.cdr
mechanical proper-
Figure 5.14
structural
steel in comparison
with unalloyed structural steel, substantial savings of material are possible. This
leads also to reduced joint cross-sections and, in total, to lower costs when making
welded steel constructions.
Based
on
steels
the
alloyed
unalloyed
classification
Figure
5.2,
of
Fig-
low-alloyed
mild steel
higher-carbon steel
Hardening
Underbead cracking
rimmed steel
to their problematic
cutting of
segregation
zones
processes
during
welding. When it
cold brittleness
(coarse-grained recrystallization
after critical treatment)
stress corrosion cracking
safety from brittle fracture
comes to unalloyed
high-alloyed
hardening
corrosion
tool steels
special properties are
resistant steels
achieved, for example:
Hardening,
special
properties
heat resistance,
are achieved
tempering resistant,
high-pressure hydrogen resistance,
toughness at low temperatures,
surface treeatment condition, etc.
ferritic
pearlitic-martensitic
austenitic
grain increase in
the weld interfaces
hardening
embrittlement
formation
of chromium
carbide
grain desintegration
stress corrosion
cracking hot cracks
(sigma phase
embrittlement)
br-er-05-15.cdr
Figure 5.15
58
casts, rimmed and semi-killed steels are causing problems. Killing means the removal of oxygen from the steel bath.
Figure 5.16 shows cross-sections of ingot blocks with different oxygen contents.
Rimming steels with increased oxygen content show, from the outside to the inside,
three different zones after solidification: 1.: a pronounced, very pure outer envelope,
2.: a typical blowhole formation (not critical, blowholes are forged together during
rolling), 3.: in the
centre
segregated
clearly
zone
where unfavourable
elements like sulphur and phosphorus are enriched.
0,025
0,012
0,003
semi-killed steel
rimmed steel
ISF 2004
br-er-05-16.cdr
Ingot cross-sections
after different casting methods
Figure 5.16
Figure 5.17 shows important points to be observed during welding such steels. Due
to their enrichment with alloy elements, the segregation zones are more transformation-inert than the
outer
envelope
In
to
cracking,
as,
hotin
and
sulphur
ISF 2004
br-er-05-17.cdr
enriched.
Figure 5.17
59
Therefore, touching such segregation zones during welding must be avoided by all
means.
In the case of lowalloy
steels,
the
Microstructures
Ferrite
80
Austenite
250
Perlite (granular)
200
welding
Perlite (lamellar)
300
observed.
Sorbite
350
Troostite
400
Cementite
600 - 650
hardness values of
Martensite
400 - 900
various microstruc-
problem
of
hardening
during
must
be
Fig-
ure 5.18
shows
ISF 2004
Br-er-05-18.cdr
HAZ
hardness
values
martensite
and
cementite. Hardness values of cementite are of minor importance for unalloyed and
low-alloy steels because its proportion in these steels remains low due to the low Ccontent.
However, hardening because of martensite formation is of greatest importance as the
martensite proportion in the microstructure depends mainly on the cooling time.
Figure 5.19 shows
the essential influHV
HRC
root cracking
presumable
400
41
1290
70
root cracking
possible
400 - 350
41 - 36
1290 - 1125
70 - 60
no root cracking
350
36
1125
60
280
28
900
30
content
in
joints.
Hardening through
martensite
forma-
with maximum
martensite
content
%
strength,
calculated at
max. hardness
N/mm2
maximum hardness
If too much martensite develops in the heat affected zone during welding (below or next to the weld),
a very hard zone will be formed which shows often cracks.
tion is not to be
ISF 2004
Br-er-05-19.cdr
carbon steels up to
about
0,22%,
Figure 5.19
60
because the critical cooling rate with these low C-contents is so high that it normally
wont be reached within the welding cycle. In general, such steels can be welded
without special problems (e.g., S. 235).
In addition to carIIW
C - qu. = C +
Mn Cr + Mo + V Cu + Ni
+
+
6
5
15
Stout
C - qu. = C +
Mo Ni Cu
Mn Cr + Mn
+
+
+
6
10
20 40
PCM = C +
Mannesmann
C - qu.PLS = C +
Hoesch
C - qu. = C +
C ET
Thyssen
Si Mn + Cu + Cr Ni Mo V
+
+
+
+ + 5B
30
20
60 15 10
site
formation
in
Si + Mn + Cu + Cr + Ni + Mo + V
20
Mn + Mo Cr + Cu Ni
= C+
+
+
10
20
40
stantial
it
comes to marten-
Si Mn + Cu Cr Ni Mo V
+
+
+
+
+
25
16
20 60 40 15
when
influence
on the transforma-
Br-er-05-20.cdr
tion behaviour of
Definition of C - Equivalent
steels
(see
Figure 5.20
appropriate just
to take the carbon content as a measure for the hardening tendency of such steels.
To estimate the weldability, several authors developed formulas for calculating the
so-called carbon equivalent, which include the contribution of the other alloy elements to hardening tendency, (Fig. 5.20). As these approximation formulas are empirically determined
as
for
0,35
Tp ==750
CET
- 150- 150
Tp
750
CET
delta Tp
HD HD0,35
- 100
delta
Tp= 62
= 62
- 100
80
200
the
delta Tp [C]
and
100
250
hardening tendency
Tp [ C]
150
100
d = 30
mm
d = 30
mm
HD HD
= 4= 4
1 kJ/mm
Q = Q1=kJ/mm
0
0,2
tions
like
thickness,
40
20
50
60
0,3
0,4
CET
= =0,33
%
CET
0,33 %
= 30mm
mm
d =d30
kJ/mm
Q =Q1= 1kJ/mm
0
0
0,5
60
heat
15
20
25
Wasserstoffgehalt
Hydrogen
contentHD
of des
theSchweigutes
weld metal [%]
Kohlenstoffquivalent
CET [%]
Carbon aquivalent
plate
10
40
delta TpTp
= 160
tanhtanh
(d/35)(d/35)
- 110 - 110
delta
= 160
delta Tp
CETCET
- 32)-Q32)
- 53Q
CET
+ 32
delta
Tp= (53
= (53
- 53
CET + 32
20
50
CET = 0,2 %
CET = 0,2 %
CET = 0,4 %
CET = 0,2 %
CET = 0,4 %
CET = 0,2 %
delta Tp [C]
delta Tp [C]
40
30
-20
-40
20
-60
of importance, the
10
CET
0,4
CET ==0,4
%%
HD =
2 2
HD
QQ== 11kJ/mm
kJ/mm
carbon
equivalent
cannot be a com-
20
40
60
80
100
-80
d =d50
= 50mm
mm
HDHD
= =8 8
-100
0
0,5
1,5
2,5
3,5
4,5
Wrmeeinbringen
Heat input Q [kJ/mm]
Plate thickness
Blechdicke d [mm]
0,35
Source:
Quelle: DIN EN 1011-2
ISF 2005
61
tion of the preheating temperature Tp, the formula as shown in Fig. 5.21 is used. The
effects of the chemical composition which is marked by the carbon equivalent CET,
the plate thickness d, the hydrogen content of the weld metal HD and the heat input Q are considered.
The essential factor
to martensite forma-
Temperature T
Tmax
time. As a measure
800
500
t8/5
t500
The
Time t
temperature
ISF 2004
br-er-05-22.cdr
Definition of t8/5
Figure 5.22
important structural transformations and that the time can be easily transferred to the
TTT diagrams.
Figure 5.23
shows
2000
time-
temperature
distri-
and
dwell
C
B
1500
Temperature T
measured
of
measurement
the
10mm
1000
B
500
0
0
and
50
100
150
200
250
300
Time t
ISF 2004
br-er-05-23.cdr
Temperature-time curves
in the adjacence of a weld
62
With the use of thinner plates with complete heating of the cross-section during welding, the heat conductivity is only carried out in parallel to the plate surface, this is the
two-dimensional heat dissipation.
With thicker plates, e.g. during welding of a blind bead, heat dissipation can also be
carried out in direction of plate thickness, heat dissipation is three-dimensional.
These two cases
K3
3 - dimensional:
t8 / 5 =
universal formula:
h
U I
1
1
2 p l v 500 - T0 800 - T0
) Uv I 5001- T
formulas given in
extended formula
For low-alloyed steel:
t8 / 5 = 0,67 - 5 10 - 4 T0
1
h N 3
800 - T0
2 - dimensional:
t8 / 5 =
universal formula:
extended formula
For low-alloyed steel:
provide a method
2
2
2
h2
1
1
U I 1
-
4 p l r c v d 2 500 - T0 800 - T0
of calculating the
2
2
2
2
1
1
U I 1
-
h N 2
t8 / 5 = 0,043 - 4,3 10 -5 T0
2
v d 500 - T0 800 - T0
d =
0,043 - 4,3 10 -5 T0
U I
h
0,67 - 5 10 - 4 T0
v
1
1
+
500 - T0 800 - T0
low-alloyed steels.
In the case of a
ISF 2004
br-er-05-24.cdr
three-dimensional
heat
dissipation,
t8/5 it independent
Figure 5.24
of plate thickness.
In the case of two-dimensional heat dissipation it is clear that t8/5 becomes the shorter
the thicker the plate thickness d is. Provided, the cooling times are equal, the plate
thickness can be calculated from these relations where a two-dimensional heat dissipation changes to a three-dimensional heat dissipation.
Figure 5.25 shows
welding methods
TIG-(He)-welding
welding method on
TIG-(Ar)-welding
MIG-(Ar)-welding
MAG-(CO2)- welding
heat
the
is
SA welding
input,
energy
which
transferred to the
base
material
depends
on
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
Br-er-05-25.cdr
the
Figure 5.25
63
groove
Type of weld
ge-
2-dimensional
heat dissipation
ometry is covered
weld factor
3-dimensional
heat dissipation
0,45 - 0,67
0,67
0,9
0,67
0,9
0,9
by seam factors
according
to
ISF 2004
br-er-05-26.cdr
easier calculation.
measure
Figure 5.26
the
cm
25
Plate thickness
TA=20C
40
TA=200C
60
80
100
150
30
40
3-dimensional
2
3-dimensional
60
100
2-dimensional
2-dimensional
0
0
10
20
30
40
50
10
20
30
40
Br-er-05-27.cdr
Figure 5.27
50
50
64
Fig. 5.28 shows the
possible range of
heat input depending on the electrode diameter. It is
clear that a relatively large working
range is available
for
arc
welding
procedures.
variation
of
A
the
energy-per-unit
Figure 5.28
length
can
be
carried out by alteration of the welding current, the welding voltage and the welding
speed.
Figure 5.29
65
50
40
30
T0 200C
150C
100C
20
20C
10
d = 7,5 mm
7
50
40
30
T0 200C
150C
100C
20
20C
10
d = 10 mm
7
50
40
30
T0 200C
150C
100C
20
20C
10
d = 15 mm
7
50
40
30
T0 200C
150C
100C
20
transition to
3-dimensional
heat flow
10
7
br-er05-30.cdr
7 8 9 10
20C
d = 20 mm
15 20
30
kJ/cm 50
Heat input E
ISF 2004
Figure 5.30
Transition thickness d
50
mm
40
aera of
3-dimensional
heat flow
30
T0
20
15
10
9
8
7
66
0 C
C 20 C
250
00
1
C
150
C
20
area of
2-dimensional
heat flow
7 8 9 10
15 20
30
kJ/cm 50
Heat input E
50
s
40
determined,
of
30
20
15
25
0
C
0
20
C
0
15
C
0
10
10
9
8
7
independent
7 8 9 10
C
20
15 20
30
Heat input E
br-er05-31.cdr
kJ/cm 50
ISF 2004
Dependence of
E, T0, t8/5 And d
Figure 5.31
The
relation
be35
V
30
welding is shown
in Fig. 5.32
and
Welding voltage
voltage
M21
M23
20
15
150
and
3,5
br-er-05-32.cdr
4,5
spray arc
mixed arc
short arc
welding current, or
200
250
Welding current
300
5,5
7,0
Wire feed
9,0
10,5
8,0
m/min
ISF 2004
C1
25
parameters. Welding
gas composition:
C1 100% CO2
M21 82% Ar + 18% CO2
M23 92% Ar + 8% O2
Figure 5.32
plate
67
F3 = 0,67
F2 = 0,67
t8/5 max = 30 s
t8/5 min = 6 s
Emax = 66 kJ/cm
Emin = 14 kJ/cm
60
fillet welds
T0= 150 C
kJ/cm
30s
50
53
40
20s
35
47
41
30
15s
25
Heat input E
MAG - weldind
Heat input E
SA - welding
t8/5
kJ/cm
25s
toughness affection
35
29
20
10s
15
two-dimensional
6s
cracking tendency
5
0
10
15
20 25
30
Plate thickness
and
three-
23
18
10
70
59
45
12
6
mm
0
40
br-er05-33.cdr
ISF 2004
Permissible E-Range
During SA - And MAG - Welding
Figure 5.33
Fig. 5.34 shows the same dependF3 = 0,9
F2 = 0,9
t8/5 max = 30 s
t8/5 min = 6 s
Emax = 49 kJ/cm
Emin = 10 kJ/cm
70
60
butt welds
T0= 150 C
kJ/cm
kJ/cm
59
50
toughness affection
45
30s
53
47
40
25s
35
30
20s
41
35
29
25
15s
20
23
15
10s
18
10
6s
12
cracking tendency
5
0
Heat input E
MAG - welding
preparation..
h'UP = 1
h'MAG = 0,85
dU max = 34 mm
dU min = 15 mm
Heat input E
SA - welding
br-er05-34.cdr
10
15
20 25
30
Plate thickness
mm
6
0
40
ISF 2004
Permissible E-Range
During SA - And MAG - Welding
Figure 5.34
68
The curve family in Fig. 5.35 shows the dependence of the heat input from the welding speed as well as the acceptable working range. The parameters of the curves 1
to 8 in the table
curve
25
kJ/cm
20
29
27
24
22
20
19
18
17
Heat input E
15
7
10
wor
king
ra
only
nge
related
diameter,
5
0
for
wire
feed,
10
15
20
25
30
35 40
45
Welding speed vS
50 cm/min 60
welding
voltage, etc.
ISF 2004
Figure 5.35
shows
Sheet
Nr. 0916).
In this example, a
24
22
20
19
18
17
59
53
47
41
30
20s
35
29
25
15s
20
10s
15
be-
27
25s
15 mm and a coolt8/5
29
35
10
time
30s
40
plate thickness of
ing
toughness affection
45
6s
cracking tendency
5
0
10
15
20 25
30
Plate thickness
mm
curve
kJ/cm
23
18
16
12 13
6
0
40
25
kJ/cm
20
heat input E
Reference
SA - welding
(according to DVS-
butt welds
T0= 150 C
50
Heat input E
70
60
kJ/cm
MAG - welding
a reading example
Heat input E
Figure 5.36
16
15
13
7
10
work
ing
rang
5
0
33
10
15
20
25
41
30
35 40
45
Welding speed vS
50 cm/min 60
ISF 2004
br-er-05-36.cdr
tween 10 and 20 s
are given. In this
case, the maximum
Figure 5.36
cooling time for MAG welding is 15 s. A solid wire with a diameter of 1.2 mm at 29V
and 300A is used.
The left diagram provides heat input values between 13 and 16 kJ/cm, based on the
given data. Using these values, the acceptable range of welding speeds can be
taken from the diagram on the right.
69
800
C
Temperature
700
500
400
Peak temperature
1000C
1400C
200
HV30=400
200
300
1400
Peak temperature
P
B
300
600
1000
Arc3
800
Arc1
plate thickness
40
F+P
F+B
B+M
1200
30
25
three-dimensional
two-dimensional
10
20
15
10 9 8
1000
t8/5
5 mm 4
2 3
10
20
energy-per-unit length
100
10
20
100 C
30
200
t8/5
preheating temperature
40
50 kJ/cm 70
bie5-37.cdr
ISF 2004
6.
Welding High Alloy Steels
71
Basically stainless steels are characterised by a chromium content of at least 12%. Figure
6.1 shows a classification
of
corrosion
corrosion-resistant steels
resistant
stainless
steels
and
perlitic
martensitic
semi-ferritic
ferritic
ferritic-austenitic
X40Cr13
X10Cr13
X8Cr13
X20CrNiSi25-4
austenitic
non-stabilized
stabilized
(austenite with
delta-ferrite)
X12CrNi18-8
(austenite without
delta-ferrite)
X8CrNiNb16-13
ISF 2002
br-er-06-01e.cdr
Figure 6.1
T
A4
T
d
A4
A4
g
A3
a(d)
fects
of
two
different
A3
A3
a
a
Alloy elements in %
Chromium
Vanadium
Molybdenum
Aluminium
Silicon
Alloy elements in %
Ferrite
Nickel
Manganese
Cobalt
developers
with
br-er-06-02e.cdr
Figure 6.2
72
tenite area, partly with downward equilibrium line according to Figure 6.2 (central figure).
With a certain content of the related element, there is a transformation-free, purely ferritic
steel.
An opposite effect provide austenite developers. In addition to carbon, the most typical member of this group is nickel.
Carbon
l
l
l
Chromium
l
Nickel
l
l
Effect
Increases the strength, supports development
of precipitants which reduce corrosion
resistance, increasing C content reduces
critical cooling rate
Works as ferrite developer, increases
oxidation- and corrosion-resistance
All types
Special types l
The table in Figure 6.3 summarises the effects of some selected elements on high alloy
steels.
1800
C
Figure 6.3
S+a
1400
ISF 2002
br-er06-03e.cdr
1600
Temperature
Element
1200
g
g+a
1000
800
d+a
d+a'
600
a'
400
200
0
Fe
10
20
30
40
50
60
70
80
90 % Cr
Chromium
ISF 2002
br-er06-04e.cdr
Binary System Fe - Cr
Figure 6.4
temperature range into a low-chromium -solid solution and a chromium-rich -solid solution.
Both, the development of the -phase and of the unary --decomposition cause a strong
73
embrittlement. With higher alloy steels, the diffusion speed is greatly reduced, therefore both
processes require a relatively long dwell time. In case of technical cooling, such embrittlement processes are suppressed by an increased cooling speed.
Nickel is a strong austenite developer, see Figure 6.5 Nickel and iron develop in this system
under elevated temperature a complete series of face-centred cubic solid solutions. Also in
1600
C
1400
Fe Ni3
S+g
decomposition
d+g
1200
Temperature
processes
1000
800
a+g
400
Fe Ni3
important
200
phases
which
0
Fe
20
10
30
50
40
60
70
80
90 % Ni
Nickel
br-er-06-05e.cdr
ISF 2002
Binary System Fe - Ni
Figure 6.5
60 % Fe
1600
C
1600
C
1500
S+g
S+d
1400
S+d+g
1400
S+g
S+d
body-centred
1300
1200
g
d+g
cubic
solid
1200
d
d+g
1100
1100
is developed by transfor-
1000
1000
900
900
800
800
700
d+s
d+
g+
Temperature
1300
1500
S+d+g
d+g+s
d+s
g+s
700
10
15
30
25
20
15
20 % Ni
10 % Cr
10
15
20
40
35
30
25
20
% Ni
15
% Cr
ISF 2002
br-er-06-06e.cdr
Figure 6.6
25
6.4.
74
During an ongoing cooling, the binary area ferrite + austenite passes through and a transformation into austenite takes place. If the coolls
st
ee
Au
s
te
n
c
te
ni
ti
3.
4.
Au
s
ar
M
2.
fe
rri
tic
ls
s
st
ee
l
s
te
n
si
tic
st
ee
l
Fe
rri
tic
1.
0.1
0.1
1.2
0.1
0.1
Si
max.
1.0
max.
1.0
max.
1.0
max.
1.0
Mn
max.
1.0
max.
1.5
max.
2.0
max.
2.0
Cr
15
18
12
18
17
26
24
28
Mo
up to
2.0
up to
1.2
up to
5.0
up to
2.0
Ni
1.0
2.5
7
26
4
7.5
up to
2.2
Cu
Nb
Ti
Al
ing (14% Ni, 16% Cr, left figure). Primary austenitic solidifying alloys are much more
susceptible to hot cracking than primary fer-
V
N
+
+
ISF 2002
br-er06-07e.cdr
these alloys.
Figure 6.7 shows some typical compositions
Figure 6.7
The diagram of Strau and Maurer in Figure 6.8 shows the influence on the microstructure
formation of steels with a C-content of 0,2%. The classification of high-alloy steels in Figure
6.1 is based on this dia-
28
%
24
Nickel
20
16
12
austen
it
ensite
e / mart
austenite / ferrite
austenite
/ martens
ite / ferrite
martensite / ferrite
6
10
12
14
Chromium
16
18
20
22
ISF 2002
br-er-06-08e.cdr
Maurer - Diagram
24 % 26
The
these
elements
is
of
de-
Figure 6.8
influence
and
nickel
75
equivalents. The Schaeffler diagram reflects additional alloy elements, Figure 6.9. It represents molten weld metal of high alloy steels and determines the developed microstructures
after cooling down from very high temperatures. The diagram was always prepared considering identical cooling conditions, the influence of different cooling speeds is here disregarded.
The areas 1 to 4 in this diagram limit the chemical compositions of steels, where specific defects may occur during welding.
Depending on the composition, purely ferritic chromium steels have a tendency to embrittlement by martensite and therefore to hot cracking (area 2) or to embrittlement due to strong
temperature,
diffusion-start
Figure
6.10
30
rit
28
26
0%
24
austenite
r
Fe
5%
10
22
%
%
0
A+F 2
20
40%
18
16
A +M
14
80%
12
10
8
100%
martensite
F
+
M
00
6
4
A+M+F
M+F
ferrite
8
10
12 14 16
18
20 22 24
26 28
30 32
34
36 38
40
considerably
br-er-06-09e.cdr
sigma embrittlement
between 500-900C
Figure 6.9
be considered as of limited
weldability.
6000
m
5000
ferrite
contents,
chromium
grain size
4000
3000
2000
1000
ferritic steel
con-
austenitic steel
200
400
600
800
1000
temperature
cooling speed.
br-er-06-10e.cdr
ISF 2002
Figure 6.10
1200
76
Finally, area 4 marks the strongly increased tendency to hot cracking in the austenite. Reason is, that critical elements responsible for hot cracking like e.g. sulphur and phosphorous
have only very limited solubility in the austenite. During welding, they enrich the melt residue,
promoting hot crack formation (see also chapter 9 - Welding Defects).
There is a Z-shaped area in the centre of the diagram which does not belong to any other
endangered area. This area of chemical composition represents the minimum risk of welding
defects, therefore such a composition should be adjusted in the weld metal. Especially when
welding austenitic steels one tries to aim at a low content of -ferrite, because it has a much
greater solubility of S and P, thus minimising the risk of hot cracking.
The Schaeffler diagram is not only used for determining the microstructure with known
chemical composition. It is also possible to estimate the developing microstructures when
welding different materials with or without filler metal. Figures 6.11 and 6.12 show two examples for a determination of the weld metal microstructures of so-called 'black and white' joints.
28
28
24
24
20
A+M
16
40
M
12
20%
123
: =1:1
80
A+F
Welding consumable
12
16
20
24
Chromium-equivalent
80
2
1 30%
100
%
A+M+F
M+F
F
0
28
32
36
12
16
20
24
28
32
Chromium-equivalent
Welding consumable
3
A+F
0
4
12
0
0
40
M+F
16
: =1:1
100
%
A+M+F
20
A+M
Nickel-equivalent
20
Nickel-equivalent
20
10
A
10
ISF 2002
br-er06-12e.cdr
br-er06-11e.cdr
Application Example of
Schaeffler - Diagram
Figure 6.12
ISF 2002
Application Example of
Schaeffler - Diagram
Figure 6.11
36
77
The ferrite content can only be measured with a relatively large dispersal, therefore DeLong
proposed to base a measurement procedure on standardized specimens. Such a system
makes it possible to measure comparable values which don't have to match the real ferrite
content. Based on these measurement values, the ferrite content is no longer given in percentage, but steels are grouped by ferrite numbers. In addition to ferrite numbers, DeLong
proposed a reworked Schaeffler diagram where the ferrite number can be determined by the
chemical composition, Figure 6.13. Moreover, DeLong has considered the influence of nitrogen as a strong austenite developer (effects are comparable with influence of carbon). Later
on, nitrogen was included into the nickel-equivalent of the Schaeffler diagram.
Nickel-equivalent = %Ni + 30 x %C + 30 x %N + 0,5 x %Mn
21
e
rit
19
r
fe
nu
austenite
18
r
be
m
0
2
20
d
re
su
ea
ym
all .-%
tic vol
e
n in
ag s
m nt 0%
ly te
er con
2%
rm
fo rrite
4%
Sc
e
f
ha
effl
6%%
er6
au
7, 2%
ste
nite
9, 7%
,
-m
art
10 ,3%
en
site
12 ,8%
-lin
13
e
17
16
15
14
13
12
11
10
16
8
10
12
14
16
18
austenite + ferrite
17
18
26
25
19
20
21
22
23
24
Chromium-equivalent = %Cr + %Mo + 1,5 x %Si + 0,5 x %Nb
27
ISF 2002
br-er-06-13e.cdr
De Long Diagram
Figure 6.13
welding process.
Figure
air
O
2Fe+++O+H2O 2Fe++++2OH-
OH
Fe+++
lytic
corrosion
under
water
O2
OH
H2O
2Fe++
cathode
anode
4e-
potential difference is a
precondition for the devel-
Fe(OH)3
2Fe 2Fe+++4e-
O2+2H2O+4e 4OH
iron
ISF 2002
br-er-06-14e.cdr
a cathode. To develop
Figure 6.14
78
such a local element, a different orientation of grains in the steel is sufficient. If a potential
difference under a drop of water is present, the chemically less noble part reacts as an anode, i.e. iron is oxidised here and is dissolved as Fe2+-ion together with an electron emission.
Caused by oxygen access through the air, a further oxidation to Fe3+ takes place. The cathodic, chemically nobler area develops OH- ions, absorbing oxygen and the electrons. Fe3+and OH--ions compose into the water-insoluble Fe(OH)3 which deposits as rust on the surface (note: the processes here described should serve as a principal explanation of electrochemical corrosion mechanisms, they are, at best, a fraction of all possible reactions).
If the steel is passivated by chromium, the corrosion protection is provided by the development of a very thin chromium oxide layer which separates the material from the corrosive
medium. Mechanical surface damages of this layer are completely cured in a very short time.
passive layer
active
dissolution
passive layer
gap
tensile stress
active dissolution
of the crack base
pitting corrosion
passive layer
activly dissolved
grain boundary
chromium
depleted zones
active
dissolution
of the gap
crevice corrosion
grain boundary
carbides
intergranular corrosion
incorrect
br-er06-15e.cdr
Figure 6.15
ISF 2002
br-er06-16e.cdr
correct
ISF 2002
Figure 6.16
The examples in Figure 6.15 are more critical, since a complete recovery of the passive layer
is not possible from various reasons.
79
br-er-06-17e.cdr
Figure 6.17
local break-up of the passive layer. Especially salts, preferably Clions, show this behaviour.
This local attack causes a dissolution of the material on the damaged points, a depression
develops. Corrosion products accumulate in this depression, and the access of oxygen to the
bottom of the hole is obstructed. However, oxygen is required to develop the passive layer,
therefore this layer cannot be completely cured and pitting occurs, Figure 6.17.
Stress-corrosion cracking occurs when the material displaces under stress and the passive
layer tears, Figure 6.18. Now the unprotected area is subjected to corrosion, metal is dissolved and the passive
layer redevelops (figures 13). The repeated displace1
ment
and
repassivation
offset;
passive layer;
10
11
metal surface;
dislocation
12
Stress
cracking
corrosion
takes
mainly
br-er-06-18e.cdr
Figure 6.18
not
follow
boundaries.
the
grain
80
Figure 6.19 shows the expansion-rate dependence of stress corrosion cracking. With very
low expansion-rates, a curing of the passive layer is fast enough to arrest the crack. With
very high expansion-rates, the failure of the specimen originates from a ductile fracture. In
the intermediate range, the material damage is due to stress corrosion cracking.
Figure 6.20 shows an example of crack propagation at transglobular stress corrosion cracking. A crack propagation speed is between 0,05 to 1 mm/h for steels with 18 - 20% Cr and 8 20% Ni. With view to welding it is important to know that already residual welding stresses
complete
cover layer
tough fracture
T=RT
SpRK
e2
e1
Elongation speed e
ISF 2002
br-er06-19E.cdr
br-er06-20e.cdr
Transgranular
Stress Corrosion Cracking
Figure 6.19
ISF 2002
Figure 6.20
The most important problem in the field of welding is intergranular corrosion (IC).
It is caused by precipitation of chromium carbides on grain boundaries.
Although a high solubility of carbon in the austenite can be expected, see Fe-C diagram, the
carbon content in high alloyed Cr-Ni steels is limited to approximately 0,02% at room temperature, Figure 6.21.
81
The reason is the very high affinity of chromium to carbon, which causes the precipita-
1200
C
1100
boundaries, Figure 6.22. Due to these precipitations, the austenite grid is depleted of
1000
900
800
600
0
0.05
0.25 % 0,3
be compensated by late diffusion. In the depleted areas along the grain boundaries (line
2 in Figure 6.22) the steel has become susceptible to corrosion.
br-er06-21e.cdr
ISF 2002
Carbon Solubility of
Austenitic Cr - Ni Steels
Figure 6.21
resis-
2
4
complete
resistance limit
disintegration.
br-er-06-22e.cdr
Figure 6.22
ISF 2002
82
mechanisms
br-er-06-23e.cdr
Grain Disintegration
process,
Figure 6.23
the
time
temperature,
the
3
Reciprocal of heat treatment temperature 1/T
certain
unsaturated austenite
oversaturated
austenite
regained by diffusion of
chromium.
br-er-06-24e.cdr
1 incubation time
2 regeneration of resistance limit
3 saturation limit for chromium carbide
ISF 2002
Figure 6.24
Figure 6.25 depicts characteristic precipitation curves of a ferritic and of an austenitic steel.
Due to the highly increased diffusion speed of carbon in ferrite, shifts the curve of carbon
precipitation of this steel markedly towards shorter time. Consequently the danger of intergranular corrosion is significantly higher with ferritic steel than with austenite.
83
As carbon is the element that triggers the intergranular corrosion, the intergranular corrosion
diagram is relevantly influenced by the c content, Figure 6.26.
By decreasing the carbon content of steel,
the start of carbide precipitation and/or the
start of intergranular corrosion are shifted
towards
lower temperatures
and
longer
quench
temperature
ELC-steels
(Extra-Low-Carbon)
18-8-Cr-Ni steel
Tempering temperature
so-called
cooling curve
of
the
shielding
gas,
Tempering time
the
ISF 2002
br-er06-25e.cdr
Figure 6.25
granular corrosion.
An often used method to
1000
C
900
800
700
Temperature
0.07%C
0.05%C
0.03%C
600
0.025%C
500
significantly
higher
than
10
104
103
105
Time
Influence of C-Content
on Intergranular Disintegration
10
ISF 2002
Figure 6.26
84
proportion of these alloy elements depend on the carbon content and is at least 5 times
higher with titanium and 10 times higher with niobium than that of carbon. Figure 6.28 shows
the effects of a stabilisation in the intergranular corrosion diagram. If both steels are subjected to the same heat treatment (1050C/W means heating to 1050C and subsequent water quenching), then the area of intergranular corrosion will shift due to stabilisation to
significantly longer times. Only with a much higher heat treatment temperature the intergranular corrosion accelerates again. The cause is the dissolution of titanium carbides at sufficiently high temperature. This carbide dissolution causes problems when welding stabilised
steels. During welding, a narrow area of the HAZ is heated above 1300C, carbides are dissolved. During the subsequent cooling and the high cooling rate, the carbon remains dissolved.
0.058 % C
0.53 % Nb
Nb/C = 9
C
600
0.030 % C
0.51 % Nb
Nb/C = 17
0.018 % C
0.57 % Nb
Nb/C = 32
M2
550
M1
500
S1
450
700
800
C
700
1050C
650
/W
600
550
500
450
0,3
400
0,2
0,5
2,5
10
50
25
100
250
1000
1
3
10
30
100
Time
W.-No.:4301 (0,06%)
X5CrNi18-10
C o m p o sitio n
S hie ld ing g a s
A r [% ]
C O2
O2
S 1
99
M 1
90
M 2
82
18
10000
unstabilized
650
1300C
/W
600
1050C
550
/W
500
450
0,3
ISF 2002
br-er06-27e.cdr
1000
800
C
700
1
3
W.-No.:4541
X5CrNiTi18-10
Figure 6.27
300
10
30
100
Time
300
1000
10000
stabilized
ISF 2002
br-er06-28e.cdr
Influence of Stabilization
on Intergranular Disintegration
Figure 6.28
If a subsequent stress relief treatment around 600C is carried out, carbide precipitations on
grain boundaries take place again. Due to the large surplus of chromium compared with niobium or titanium, a partial chromium carbide precipitation takes place, causing again inter-
85
granular susceptibility. As this susceptibility is limited to very narrow areas along the welded
joint, it was called knife-line attack because of its appearance. Figure 6.29.
In stabilised steels, the chromium carbide represents an unstable phase, and with a sufficiently long heat treatment to transform to NbC, the steel becomes stable again. The stronger
the steel is over-stabilised, the lower is the tendency to knife-line corrosion.
Nowadays the importance
of Nickel-Base-Alloys increases constantly. They
are ideal materials when it
comes
to
components
sistance, or combinations
Knife-Line Corrosion
Figure 6.29
ing of nickel-base-alloys.
Materials listed there are selected examples, the total number of available materials is many
times higher.
Group A consists of nickel
alloys. These alloys are
Alloy
Chem. composition
Alloy
Nickel 200
Ni 99.6, C 0.08
Nickel 212
Nickel 222
Incoloy 925
Ni 42.0, Fe 32.0, Cr 21.0, Mo 3.0, W 2.1, Cu 2.2, Al 0.3
Ni-Span-C 902 Y2O3 0.5, Ni 42.5, Fe 49.0, Cr 5.3, W 2.4, Al 0.5
Group A
Chem. Composition
characterized by moderate
Group D1
Group B
Group D2
Monel 400
Ni 66.5, Cu 31.5
Monel K-500
Monel 450
Ferry
Group C
Inconel 718
Ni 45.0, Cu 55.0
Inconel 600
Nimonic 105
Nimonic 75
Ni 80.0, Cr 19.5
Incoloy 903
Nimonic 86
Incoloy 909
Incoloy 800
Inco G-3
Incoloy 825
Inco C-276
Inco 330
Group E
Monel R-405
dition,
and
cold-drawn
ISF 2002
br-er-06-30e.cdr
material is recommended
Typical Classification of Ni-Base Alloys
Figure 6.30
86
Group B consists mainly of those nickel-copper alloys that can be hardened only by cold
working. The alloys in this group have higher strength and slightly lower toughness than
those in Group A. Cold-drawn or cold-drawn and stress-relieved material is recommended for
best machinability and smoothest finish.
Group C consists largely of nickel-chromium and nickel-iron-chromium alloys. These alloys
are quite similar to the austenitic stainless steels. They can be hardened only by cold working
and are machined most readily in the cold-drawn or cold-drawn and stress-relieved condition.
Group D consists primary of age-hardening alloys. It is divided into two subgroups:
D 1 Alloys in the non-aged condition.
D 2 Aged Group D-1 alloys plus several other alloys in all conditions.
The alloys in Group D are characterized by high strength and hardness, particularly when
aged. Material which has been solution annealed and quenched or rapidly air cooled is in the
softest condition and does machine easily. Because of softness, the non-aged condition is
necessary for trouble free drilling, tapping and all threading operations. Heavy machining of
the age-hardening alloys is best accomplished when they are in one of the following conditions:
1. Solution annealed
2. Hot worked and quenched or rapidly air cooled
Group E contains only one material: MONEL R-405. It was designed for mass production of
automatically machined screws.
Due to the high number of possible alloys with different properties, only one typical material
of group D2 is discussed here: Material No. 2.4669, also known as e.g. Inconel X-750.
The aluminium and titanium containing 2.4669 is age-hardening through the combination of
these elements with nickel during heat treatment: gamma-primary-phase (') develops which
is the intermetallic compound Ni3(Al, Ti).
During solution heat treatment of X-750 at 1150C, the number of flaws and dislocations in
the crystal is reduced and soluble carbides dissolve. To achieve best results, the material
87
should be in intensely worked condition before heat treatment to permit a fast and complete
recrystallisation. After solution heat treatment, the material should not be cold worked, since
this would generate new dislocations and affect negatively the fracture properties.
The creep rupture resistance of X-750 is due to an even distribution of the intercrystalline '
phase. However, fracture properties depend more on the microstructure of the grain boundaries. During an 840C stabilising heat treatment as part of the triple-heat treatment, the fine '
phase develops inside the grains and M23C6 precipitates onto the grain boundaries. Adjacent
to the grain boundary, there is a ' depleted zone. During precipitation hardening (700C/20
h) ' phase develops in these depleted zones. ' particles arrest the movement of dislocations, this leads to improved strength and creep resistance properties.
During the M23C6 transformation, carbon is stabilised to a high degree without leaving chromium depleted areas along the grain boundaries. This stabilisation improves the resistance
of this alloy against the attack of several corrosive media.
With a reduction of the precipitation temperature from 730 to 620C as required for some
special heat treatments additional ' phase is precipitated in smaller particles. This enhances the hardening effect and improves strength characteristics.
Further metallurgical discussions about X-750, can be taken from literature, especially with
view to the influence of heat treatment on fracture properties and corrosion behaviour.
The recommended processes for welding of X-750 are tungsten inert gas, plasma arc, electron beam, resistance, and pressure oxy arc welding.
During TIG welding of INCONEL X-750, INCONEL 718 is used as welding consumable. Joint
properties are almost 100% of base material at room temperature and about 80% at 700 820C. Figure 6.31 shows typical strength properties of a welded plate at a temperature
range between -423 and 1500F (-248 820C).
Before welding, X-750 should be in normalised or solution heat treated condition. However, it
is possible to weld it in a precipitation hardened condition, but after that neither the seam nor
the heat affected zone should be precipitation hardened or used in the temperature range of
precipitation hardening, because the base material may crack. If X-750 was precipitation
hardened and then welded, and if it is likely that the workpiece is used in the temperature
range of precipitation hardening, the weld should be normalised or once again precipitation
hardened. In any case it must be noted that heat stresses are minimised during assembly or
welding.
88
X-750 welds should be solution heat treated before a precipitation hardening. Heating-up
speed during welding must be from the start fast and even touching the temperature range of
precipitation hardening only as briefly as possible. The best way for fast heating-up is to insert the welded workpiece into a preheated furnace.
Sometimes a preheating before welding is advantageous if the component to be welded
has a poor accessibility, or the welding is complex, and especially if the assembly proves to
be too complicated for a post heat treatment. Two effective welding preparations are:
1. 1550F/16 h, air cooling
2. 1950F/1 h, furnace cooling with 25-100F/h up to 1200F, air
A repair welding of already fitted parts should be followed by a solution heat treatment (with a
fast heating-up through the temperature range of precipitation hardening) and a repeated
precipitation hardening.
A cleaning of intermediate layers must be
220
200
180
160
tensile strength
140
120
100
0.2% yield stress
80
Elongation, %
60
30
20
elongation in 1/2
10
0
20
elongation in 2
10
-423
800
1000
1200
1400
1600
ISF 2002
Mechanical Properties of
a Typical Ni-Base Alloy
600
br-er06-31e.cdr
400
Temperature, F
200
Figure 6.31
7.
Welding of Cast Materials
90
cast materials
vides a summary
of
the
metallic
cast materials
non-metallic
cast materials
plastics, gypsum and s.th.similar
different
iron-carboncast materials
non-iron-metal
cast materials
unalloyed
ferritic
referred
iron,
to
cast
cast
and
nodular graphite
cast iron
ferritic
perlitic
steel
lemellar graphite
cast iron
high alloyed
high Ccontent
austenitic
ferritic
not decarburized
decarburized
decarburized
annealed
malleable cast iron
malleable
cast
low alloyed
alloyed
perlitic
steel, as special
cast iron
malleable iron
cast steel
ferritic
perlitic
Cr-cast
iron
not decarburized
annealed
malleable cast iron
perlitic
ferritic
perlitic
ledeburitic
austenitic
graphite
other
elements
Si-cast
iron
Al-cast
iron
austenitic
ISF 2002
br-er-07-01e.cdr
materials,
Figure 7.1
no importance in
welding.
Figure 7.2 shows the designation of
Designation according to the material code (DIN EN 1560)
EN
GJ
L
F
150
Position 6:
2 34
standardised material
cast material
graphite structure (lamellar graphite)
microstructure (ferritic)
2
mechanical properties (Rm= 150 N/mm )
chemical composition (high alloyed)
optionally
the material code and the designation according to the material number. In Figure 7.2, examples of two
materials are specified.
EN
J
L
1
27
1
4,5,6
standardised material
cast material
graphite structure (lamellar graphite)
number for the main characteristic
material identification number
special requirement
ISF 2004
br-er07-02e.cdr
Designation of Materials
Figure 7.2
91
Figure 7.3 depicts a survey of the mechanical properties and the chemical compositions of several customary cast materials. As to its analysis and mechanical properties which are very different from other cast materials, cast steel constitutes an
exception to the rule.
In Figure 7.4 the stable and the metastable iron-carbon diagram are shown. The differences between
the cast material
are best explained
this way. Cast iron
with lamellar and
spheroidal
graph-
ite
carbon
has
contents
tween
of
2,8
beand
above
rials solidify fo llowing the stable system, i.e., the carbon is precipitated
in
the
form
of
graphite. Malleable
cast
iron
shows
similar C-contents,
the
solidification
from
the
molten
metal,
however,
follows
the
tastable
me-
system.
The C-contents of
cast steel, on the
Figure 7.4
92
Figure 7.5
Figure 7.6
93
improves castability. Besides the poor mechanical properties (elongation after fracture of approx. 1%), these chemical properties also impede welding with ordinary
means. It is not possible to carry out constructional welding with grey cast iron. Repair welds of grey cast iron are, in contrast, carried out more frequently as damaged
cast parts are not easily replaceable. For those repair welds, the cast parts must be
preheated (entirely or partly) to te mperatures of approx. 650C. Heating and cooling
must be done very slowly as the cast piece may be destroyed already by the thermal
stresses. The highly liquid weld metal also constitutes a problem, and thus the molten
pool must be supported by a carbon pile. Welding may be carried out with similar
filler material (materials of the same composition as the base). If grey cast iron is to
be welded without any preheating, the filler material must, as a rule, be dissimilar (of
different composition to the base metal). During this type of welding, there are always
strong structural changes in the region of the weld which lead to high hardening and
high residual stresses. For the minimisation of these structural changes, a highly ductile filler material is applied. The heat input into the base material should be as low as
possible.
Figure 7.7 depicts
the structural constitution of sphe roidal graphite cast
iron. The graphite
spheroidization
achieved
by
is
the
graphite,
the
considerably
lesser than this is
the case with grey cast iron, this type of cast iron is characterised by substa ntially
better mechanical parameters with a considerably higher elongation after fracture
and improved ductility. For this reason, the risk of material failure caused by weld
residual stresses or thermal stresses is considerably reduced for spheroidal graphite
94
cast iron. Frequently, nickel-based
alloys are used as filler material. Problems occur in the HAZ where, besides
the ledeburite eutectic alloy system,
also Ni-Fe-martensite is frequently
formed. Both structures lead to extreme hardening in the HAZ which
can
be
removed
only
Figure 7.8
Figure 7.9
by
time-
95
If annealing is carried out in neutral
atmosphere, the structure of Carburized Annealed Malleable Cast Iron
develops. Annealing in oxidising atmosphere leads to the decarburisation of the workpiece surfaces and
Decarburized Annealed Malleable
Cast Iron is developed, Figure 7.10.
Carburized
Annealed
Malleable
Figure 7.10
Figure 7.11
8.
Welding of Aluminium
8. Welding of Aluminium
97
Figure 8.1 compares basic physical properties
Property
Al
Fe
of steel and aluminium. Side by side with different mechanical behaviour, the following
Atomic weight
[g/Mol]
26.9
55.84
Specific weight
[g/cm]
2.7
7.87
fcc
bcc
Lattice
E-module
[N/mm]
71*10
210*10
R pO,2
PO,2
[N/mm]
ca. 10
ca. 100
R mm
[N/mm]
ca. 50
ca. 200
[J/(g*C)]
0.88
0.53
[C]
660
1539
[W/(cm*K)]
2.3
0.75
[nWm]
28-29
97
Expansion coeff.
[1/C]
24*10 -6
12*10 -6
Melting point
Heat conductivity
FeO
Oxydes
Al2O 3
[C]
2050
Fe 3O 4
Fe 2O 3
1400
1600
(1455)
ISF 2002
br-er08-01.cdr
Figure 8.2 compares some mechanical properties of steel with properties of some light
metals. The important advantages of light
Figure 8.1
shown in the right part of the figure. If a comparison should be based on an identical stiffness, then the aluminium supporting beam has a 1.44 times larger cross-section than the
steel beam, however only about 50% of its weight.
Figure 8.3 compares qualitatively the stress-strain diagram
of
Aluminium
and
cubic)-lattice
at
Aluminium
is
br-er-08-02.cdr
not
Figure 8.2
8. Welding of Aluminium
98
4
cm
2
200C
400
1000
1200
600
800
1500
-2
Steel
-4
Stress
8
cm aluminium
6
100C
200
Al-alloy
2
300
400 500
600
-2
-4
-6
-8
-18
Elongation
ISF 2002
br-er08-03.cdr
-16
-14
br-er08-04.cdr
Comparison of Stress-Elongation
Diagrams of Al and Steel
Figure 8.3
-12
-10
-8
-6
-4
-2
cm
ISF 2002
Figure 8.4
Figure 8.4 illustrates the effect of the considerably higher heat conductivity on the welding
process compared with steel. With aluminium, the temperature gradient around the welding
point is considerably smaller than with steel. Although the peak temperature during Al welding is about 900C below steel, the isothermal curves around the welding point have a clearly
larger extension. This is due to the considerably higher heat conductivity of aluminium compared with steel.
This special characteristic of Al requires a input heat volume during welding equivalent to
steel.
Figure 8.5 lists the most important alloy elements and their combinations for industrial use.
Due to their behaviour during heat treatment can Al-alloys be divided into the groups hardenable and non-hardenable (naturally hard) alloys.
8. Welding of Aluminium
99
Al Cu Mg
ing consumables.
Al Mg Si
Cu
Al Zn Mg
Al Zn Mg Cu
678
hardenable alloys
Al
Zn
Al Si Cu
the seam.
Si
Al Mg
Al Mg Mn
Mn
ISF 2002
br-er08-05.cdr
678
non-hardenable alloys
Figure 8.5
second group is explained by the figures 8.7 und 8.8. Example: If an alloy containing about
4.2% Cu, which is stable at room temperature, is heat treated at 500C, then, after a sufficiently long time, there will be only a single phase structure present. All alloy elements were
dissolved, Figure 8.8 between point P and Q.
When quenched to room
Al - alloys
Al99,5
AlCuMg1
AlMgSi0,5
AlSi5
AlMg3
AlMg2Mn0,8
AlMn1
Typical use
electrical engineering
mechanical engineering, food
industries
architecture, electrical
engineering, anodizing quality
architecture, anodizing quality
architecture, apparatus-, vehicle-,
shipbuilding engineering, furniture
industry
apparatus-, vehicle-, shipbuilding
engineering
apparatus-, vehicle-engineering,
food industry
W elding consumable
SG-Al 99,5Ti;
SG-Al 99,5
SG-AlMg4,5Mn
SG-AlMg5; SG-AlMg4,5Mn;
SG-AlSi5
SG-AlSi5
SG-AlMg3;
SG-AlMg4,5Mn
SG-AlMg5; SG-AlMg3;
SG-AlMg4,5Mn
of equilibrium. If such a
structure is subjected to an
SG-AlMn1;SG-Al99,5T
elevated
temperature,
ISF 2002
br-er-08-06.cdr
Figure 8.6
precipitation of a second
phase takes place in ac-
8. Welding of Aluminium
100
cordance with the binary system, the crystal tries to get back into thermodynamical equilibrium.
Depending on the level of
stable condition
repeated hardening
quenching
regeneration
warm ageing
ageing at slightly
increased temperature
coherent
precipitations,
cold aged
condition
temperature
rise
longer warm
ageing
partly coherent
and incoherent
precipitations,
softening
from
the
longer warm
ageing
stable incoherent
equilibrium phase
stable condition
ISF 2002
br-er-08-07.cdr
deviating
Ageing Mechanism
same
lattice
structure),
partly
coherent
particles
Figure 8.7
tained),
and
incoherent
particles (lattice structure completely different from the matrix), Figure 8.7. Coherent particles
formed at room temperature can be transformed into incoherent particles by increase of temperature (i.e. enabling diffusion).
The precipitations cause a restriction to the
700
liquid
600
copper containing
aluminium solid solution
500
Temperature
400
300
aluminium solid solution
and copper aluminide
(Al2Cu)
ure 8.7) a maximum of second phase has precipitated after elapse of a certain time.
Consequently a prolonged stop at this tem-
200
100
copper content of
AlCuMg
mass-%
Copper
br-er08-08.cdr
space).
ISF 2002
Figure 8.8
8. Welding of Aluminium
101
After a very long heat ageing a stable condition is reached again with relatively large precipitations of the second phase in the matrix.
In Figure 8.7 is this stable final condition iden-
500
P
quenching
Temperature
400
300
200
heat ageing
100
10
12
14
Time
br-er08-09.cdr
ISF 2002
Figure 8.9
Figure
8.10
shows
the
380
diffusion
condi-
320
0.2% yield stress s0.2 in N/mm
1 in dependence of time.
260
120C
200
RT
140
80
10-1
age
hardening.
quenched
100
101
10
10
Ageing time in h
ISF 2002
br-er-08-10.cdr
The
Figure 8.10
lier. The curve of hot ageing shows clearly the begin of strength loss when held at a too long
stoppage time. This figure shows another specialty of the process of ageing. During ageing, a
8. Welding of Aluminium
102
second phase is precipitated from a single-phase structure. To initiate this process, the structure must contain nuclei of the second phase. However, a certain time is required to develop
such nuclei. Only after formation of nuclei can the increase in strength start. The period up to
this point is called incubation time.
500
110
N/mm
Tensile strength sB
135
400
150
180
300
190
205
230
260C
200
110
N/mm
400
0.2% yield stress s0.2
300
135
200
180
190
205C
150
230
260
190
180
205
150
135
20
10
110C
260
230
1
day
30
min
10
year under a temperature of 110C. An increase of the ageing temperature shortens the
30
%
-2
10
-1
10
10
Ageing time
1
week
10
1
1
month year
10
h 10
br-er08-11.cdr
ISF 2002
Figure 8.11
leads to a lowering of the
400
N/mm
Tensile strength Rm
300
AlMg5
AlMg3
tion is counter-proportional
200
100
strength
increase
Al99,5
companied by an embrit0
0
30
70
Figure 8.12
Strain
ISF 2002
br-er-08-12.cdr
8. Welding of Aluminium
103
Figure 8.12 shows a method of how to increase the strength of non-hardenable alloys. As no
precipitations are present to reduce the movement of dislocations, such alloys can only be
strengthened by cold working.
Figure 8.12 illustrates two essential mechanisms of strength increase of such alloys. On
300
N/mm
250
Rm or Rp0,2
200
of the lattice.
150
0,7
100
0,5
50
HV30
0,4
Rp0,2/Rm
0,6
0,3
0,2
0
80
60 40 20 0 20 40
Distance from Seam Centre
60 mm 100
br-er08-13.cdr
ISF 2002
Non-Hardenable Al Alloy
of a welding process.
Figure 8.13
400
Figure
8.14
illustrates
the
90 days RT
N/mm
Rm
350
21 days RT
Rp0,2
250
90 days RT
Stress
As a consequence of the
1 day RT
300
21 days RT
200
1 day RT
150
50
80
br-er-08-14.cdr
40
20
20
60
0
40
Distance from seam centre
Hardenable Al Alloy
60
Figure 8.14
80
100
mm
140
ISF 2002
8. Welding of Aluminium
104
Figure 8.15 shows another
problematic nature of Alwelding. Due to the high
thermal expansion of aluminium, high tensions develop during solidification
of the weld pool in the
course of the welding cycle. If the welded alloy indicates
ISF 2002
br-er-08-15.cdr
high
melting
100
%
2
60
1
40
X
X
20
Mg
Cracking susceptibility
influences
content
80
Si
X
X
100
300
Preheat temperature
400
500
Alloy content
1: AlMgMn 2: AlMg 2,5 3: AlMg 3,5
ISF 2002
br-er-08-16.cdr
200
Figure 8.16
To avoid hot cracking, partly very different preheat temperatures are recommended for the
alloys. Zschtge proposed a calculation method which compares the heat conductivity conditions of the Al alloy with those of a carbon steel with 0.2% C. The formula is shown in Figure
8. Welding of Aluminium
105
8.17, together with the related calculation result. These results are only to be regarded as
approximate, the individual application is subject to the information of the manufacturer.
strong
porosity
of
the
TS
Tvorw.
lAl-Leg.
in C
in C
in J/cm*s*K
660
by
hydrogen,
600
Recommended preheat temperature
745
l Al-Leg.;
C
500
400
300
200
100
TVorw. = TS -
Al Zn Mg Cu 0,5 Al Zn Mg Cu 1,5
problem
Al Si 5
Al Cu Mg 1
Al R Mg 2
Al Cu Mg 0,5
Al Mn Al Mg 2
Al Cu Mg 2
Al Mg 3
Al Mg 3 Si Al Mg Mn
major
Al 99,98R Al99,9
Al99,8
Al 99,7 Al 99,5
Al 99
Al R Mg0,5
Al Mg Si 0,5 Al Mg Si 0,8 Al Mg Si 1
E Al Mg Si 1
Al Mg 1
Another
br-er-08-17.cdr
aluminium
abruptly
on
changes
the
Figure 8.17
phase
ISF 2002
br-er-08-18.cdr
Figure 8.18
8. Welding of Aluminium
106
To suppress such pore
irregular
wire
electrode
feed
formation it is therefore
humid air
(nitrogen, oxygen, water)
poor
current transition
VS
humid air
feuchte Luftpores
Poren
solid weld metal
base material
Al.
H2
H2
festes
Schweigut
Grundwerkstoff
ISF 2002
br-er-08-19.cdr
Figure 8.19
parallel gap
weld pool
overlap
opening gap
weld pool
ISF 2002
br-er-08-20.cdr
8. Welding of Aluminium
wedge
br-er08-21.cdr
107
flame
ISF 2002
Figure 8.21
9.
Welding Defects
9. Welding Defects
109
Figures 9.1 to 9.4 give a rough survey about the classification of welding defects to DIN
8524. This standard does not classify existing welding defects according to their origin but
only to their appearance.
Figure 9.1
Figure 9.3
Figure 9.2
9. Welding Defects
110
A distinction of arising defects by their origin is shown in Figure 9.5. The development of the
most important welding defects is explained in the following paragraphs.
or
insufficiently
the upper part, arc characteristic lines of MAG welding are shown using CO2
and mixed gas. The welding voltage depends on
welding current and is selected according to the
joint type. With present
tension, the welding current is fixed by the wire
feed
speed
(thus
also
ure.
Melting rate (resulting from selected welding parameters) and welding speed define the heat
input. As it can be changed within certain limits, melting rate and welding speed do not limit
each other, but a working range is created (lower part of the figure). If the heat input is too
low, i.e. too high welding speed, a definite melting of flanks cannot be ensured. Due to the
9. Welding Defects
111
poor power, lack of fusion is the result. With too high heat input, i.e. too low welding speed,
the weld pool gets too large and starts to flow away in the area in front of the arc. This effect
prevents a melting of the base metal. The arc is not directed into the base metal, but onto the
weld pool, and flanks are not entirely molten. Thus lack of fusion may occur in such areas.
Figure 9.6
Figure 9.7
Figure 9.7 shows the influence of torch position on the development of weak fusion. The upper part of the figure explains the terms neutral, positive and negative torch angle. Compared
with a neutral position, the seam gets wider with a positive inclination together with a slight
reduction of penetration depth. A negative inclination leads to narrower beads. The second
part of the figure shows the torch orientation transverse to welding direction with multi-pass
welding. To avoid weak fusion between layers, the torch orientation is of great importance, as
it provides a reliable melting and a proper fusion of the layers. The third figure illustrates the
influence of torch orientation during welding of a fillet weld.
With a false torch orientation, the perpendicular flank is insufficiently molten, a lack of fusion
occurs. When welding an I-groove in two layers, it must be ensured that the plate is com-
9. Welding Defects
112
pletely fused. A false torch orientation may lead to lack of fusion between the layers, as
shown in the lower figure.
Figure 9.8 shows the influence of the torch orientation during MSG welding of
a rotating workpiece. As
an example, the upper figure shows the desired
torch orientation for usual
welding speeds. This orientation depends on parameters
like
workpiece
diameter
and
thickness,
groove
shape,
melting
Figure 9.8
Figure 9.9
metal: the mechanical and the metallurgical pore formation. Figure 9.9 lists causes of a mechanical pore formation as well as possibilities to avoid them. To over-weld a cavity (lack of
9. Welding Defects
113
fusion, gaps, overlaps etc.) of a previous layer can be regarded as a typical case of a mechanical pore formation. The welding heat during welding causes a strong expansion of the
gasses contained in the cavity and consequently a development of a gas bubble in the liquid
weld metal. If the solidification is carried out so fast that this gas bubble cannot raise to the
surface of the weld pool, the pore will be caught in the weld metal.
a) X-ray photograph
b) Surface cross-section
c) Transverse section
br-er09-10.cdr
ISF 2002
Figure 9.10
Figure 9.11
Figure 9.10 shows a X-ray
photograph
of
pore
Figure 9.12
9. Welding Defects
114
Figure 9.11 summarises causes of and measures to avoid a metallurgical pore formation.
Reason of this pore formation is the considerably increased solubility of the molten metal
compared with the solid state.
During solidification, the transition of liquid to
solid condition causes a leapwise reduction of
gas solubility of the steel. As a result, solved
gasses are driven out of the crystal and are ena) X-rax photograph
front.
With
slow
growth
of
the
c) transverse section
br-er09-13.cdr
ISF 2002
Figure 9.13 shows a X-ray photograph, a surface and a transverse section of a seam with
Figure 9.13
metallurgical pores. The
evenly distributed pores
across the seam and the
accumulation of pores in
the upper part of the seam
(transverse
section)
are
typical.
during
MAG
9. Welding Defects
115
is mainly caused by hydrogen and nitrogen. Oxygen is bonded in a harmless way when using
universal electrodes which are alloyed with Si and Mn.
ISF 2002
br-er09-15.cdr
Classification of Cracks
to DIN 8524 Part 3
Figure 9.15
9. Welding Defects
116
Figure 9.15 classifies cracks to DIN 8524, part 3. In contrast to part 1 and 2 of this sta ndard,
are cracks not only classified by their appearance, but also by their development.
Figure
9.16
allocates
during
the
cracks)
(cold cracks).
Figure 9.16
A model of remelting development and solidification cracks is shown in Figure 9.17. The
upper part illustrates solidification conditions
in a simple case of a binary system, under
the provision that a complete concentration
balance takes place in the melt ahead of the
solidification front, but no diffusion takes
place in the crystalline solid. When a melt of
a composition C0 cools down, a crystalline
solid is formed when the liquidus line is
reached. Its concentration can be taken from
the solidus line. In the course of the ongoing
solidification, the rest of molten metal is enriched with alloy elements in accordance with
the liquidus line. As defined in the beginning,
no diffusion of alloy elements in the already
solidified crystal takes place, thus the crystals
are enriched with alloy elements much slower
Figure 9.17
and
0020
9. Welding Defects
117
of
shrinking
9. Welding Defects
118
remaining
melt
Figure 9.19
to
the
unsuitable
groove geometry.
Figure 9.20
clearly see that an existing crack develops through the follo wing layers during over-welding.
Figure 9.22 classifies cold cracks depending on their position in the weld metal area. Such a
classification does not provide an explanation for the origin of the cracks.
9. Welding Defects
Figure 9.21
119
Figure 9.22
Figure 9.23 shows a summary of the three main causes of cold crack formation and their
main influences. As explained in previous chapters, the resulting welding microstructure depends on both, the composition of base and filler materials and of the cooling speed of the
joint. An unsatisfactory structure composition promotes very much the formation of cold
cracks (hardening by martensite).
Another cause for increased cold crack susceptibility is a higher hydrogen content. The hydrogen content is very much influenced by the condition of the welding filler material (humidity of electrodes or flux, lubricating grease on welding wire etc.) and by humidity on the
groove edges.
The cooling speed is also important because it determines the remaining time for hydrogen
effusion out of the bead, respectively how much hydrogen remains in the weld. A measure is
t8/1 because only below 100C a hydrogen e ffusion stops.
9. Welding Defects
Figure 9.23
120
Figure 9.24
A crack initiation is effected by stresses. Depending on material condition and the two already mentioned influencing factors, even residual stresses in the workpiece may actuate a
crack. Or a crack occurs only when superimpose of residual stresses on outer stress.
Figure 9.24 shows typical cold cracks in a workpiece. An increased hydrogen content in the
weld metal leads to an increased cold crack tendency. Mechanisms of hydrogen cracking
were not completely understood until today. However, a spontaneous occurrence is typical of
hydrogen cracking. Such cracks do not appear directly after welding but hours or even days
after cooling. The weld metal hydrogen content depends on humidity of the electrode coating
(manual metal arc welding) and of flux (submerged arc welding).
9. Welding Defects
121
apply
to
all
electrode
types.
The
can be noticed.
Figure 9.26
9. Welding Defects
122
very damp conditions can be rebaked to reach acceptable values of water content in the
coating.
Figure 9.27 shows the influence of cooling speed
and also the preheat temperature on hydrogen content of the weld metal. The
values of a high hygroscopic
cellulose-coated
weld metal. Reason is that hydrogen can still effuse all the way down to room temperature,
but diffusion speed increases sharply with temperature. The longer the steel takes to cool,
the more time is available for hydrogen to effuse out of the weld metal even in higher quantities.
The table in Figure 9.28
shows an assessment of
the quantity of diffusible
hydrogen in weld metal
according to DIN 8529.
9. Welding Defects
123
A cold crack development
can
be
followed-up
by
component.
Figure 9.29
vides a qualitative assessment of the crack size. The observation is carried out without applying an external tension, i.e. cracks develop only caused by the internal residual stress condition. Figure 9.32 shows that most cracks occur relatively short after welding. At first this is
due to the cooling process. Ho wever, after completed cooling a multitude of deve loping
sounds can be registered. It is remarkable that the intensity of late occurring pulses is especially high. This behaviour is typical for hydrogen induced crack fo rmation.
Figure 9.31 shows a characteristic occurrence of lamellar cracks (also called lamellar tearing). This crack type occurs
typically during stressing a
plate across its thickness
(perpendicular to rolling direction). The upper picture
shows joint types which are
very much at risk to formation of such cracks. The two
lower
pictures
show
the
9. Welding Defects
124
the casting process. With follo wing production steps, such segregations are stretched in the
rolling direction. Zones enriched and depleted of alloy elements are now close together.
These concentration diffe rences influence the
transformation behaviour of the individual
zones. During cooling, zones with enriched
alloy elements develop a different microstructure than depleted zones. This effect which
can be well recognised in Figure 9.31, is
called structure banding. In practice, this formation can be hardly avoided. Banding in
plates is the reason for worst mechanical
properties perpendicular to rolling direction.
This is caused by a different mechanical behaviour of different microstructures.
When stressing lengthwise and transverse to
rolling
direction,
the
individual structure
br-er09-31.cdr
ISF 2002
strength is provided.
Such support cannot be obtained perpendicular to rolling direction, thus the strength of the
Figure 9.31
workpiece is that of the
weaker microstructure areas. Consequently, a lamellar crack propagates
through
weaker
micro-
9. Welding Defects
125
a considerable shrinking. A welded construction which greatly impedes shrinking of this joint,
may generate stresses perpendicular to the plane of magnitude above the tensile strength.
This can cause lamellar tearing.
Precipitation cracks occur mainly during stress relief heat treatment of welded components.
They occur in the coarse grain zone close to fusion line. As this type of cracks occurs often
during post weld heat treatment of cladded materials, is it also called undercladding crack,
Figure 9.33.
Especially susceptible are steels which contain alloy elements with a precipitation hardening
effect (carbide developer like Ti, Nb, V). During welding such steels, carbides are dissolved in
an area close to the fusion line. During the following cooling, the carbide developers are not
completely re-precipitated.
Figure 9.33
10.
Testing of Welded Joints
Ls
127
The basic test for determination of material
a
S S
S S
in test area
in test area
b1
L0
Lc
Lt
total length
head width
Lt
b1
plates
tubes
1 2
Lc
parallel length ) )
radius of throat
r
) for pressure welding and beam welding, L S = 0.
2
) for some other metallic materials (e.g.aluminium, copper and their alloys)
__ L c L S +100 may be required
mens are used. Figure 10.1 shows both standard specimen shapes for that test. A
specimen is ruptured by a test machine while
the actual force and the elongation of the
d1
L0 = measurement length
(L0 = k S0 with k = 5,65)
Lt = total length
S0 = initial cross-section within
test length
br-er10-01.cdr
ment values, tension and strain are calculated. If is plotted over , the drawn diagram
LO
LC
Lt
d = specimen diameter
d1 = head diameter depending
on clamping device
LC = test length = L0 + d/2
r = 2 mm
Figure 10.1
point.
The most important characteristic values
Rm
ReH
Rel
sf
A.
To determine the deformability of a weld, a
e
ALud
Ag
A
Rm
RP0,2
RP0,01
sf
0,2 %
0,01 %
Ag
A
br-er10-02.cdr
ISF 2002
Figure 10.2
128
test result is the bending angle and the diameter of the used former. A bending angle of 180
is reached, if the specimen is pressed through the supporting rollers without development of
a crack. In Figure 10.3 specimen shapes of this test are shown. Depending on the direction
the weld is bent, one distinguishes (from top to bottom) transverse, side, and longitudinal
bending specimen. The tension side of all three speci-
section A-B
tension side
former
supporting roller
r
a
B
Lt
section A-B
a
Lf
l
Lt
tension side
the
Side
bending
thickness.
plate
is
tension side
specimens
B
Lt
bending specimen
l
Lt
d
D
a
r
b
Lt
br-er10-03.cdr
is fixed at 10 mm.
A determination of the toughness of a material or welded joint is carried out with the notched
bar impact test. A cuboid specimen with a V-notch is placed on a support and then hit by a
pendulum ram of the im55
10
10
r = 0,25
45
J
Charpy impact energy
40
45
40
average values
maximaum values
minimum values
35
D im e nsio ns
leng th
width
hight
notc h angle
No m inal s ize
55
mm
10
mm
10
mm
45
To leranc e
0,6 0
mm
0,1 1
mm
0,0 6
mm
2
8
0 ,2 5
mm
mm
0,0 6
0,0 2 5
mm
mm
2 7,5
mm
0,4 2
mm
30
25
20
15
90
10
-80
-60
br-er10-04.cdr
-40 -20
0
Temperature
20 C 40
Figure 10.4
Figure
10.4
shape
(Iso-V-
129
Three specimens are tested at each test tem-
Designation
VWS a/b
Dicke
RL
VWS a/b
(fusion
weld)
Weld centre
Designation
RL
Dicke
VHT 0/b
VWT a/b
VWT 0/b
VHT a/b
VWT a/b
a
RL
RL
VHT a/b
a
RL
takes place.
V = Charpy-V notch
W = notch in weld metal; reference line is centre line of weld
H = notch in heat affected zone; reference line is fusion line or bonding zone
(notch should be in heat affected zone)
S = notched area parallel to surface
T = notch through thickness
a = distance of notch centre from reference line (if a is on centre line of weld, a = 0 and
should be marked)
b = distance between top side of welded joint and nearest surface of the specimen
(if b is on the weld surface, then b = 0 and should be marked)
br-er10-05.cdr
Figure 10.5
a
RL
tion temperature cannot be carried out. Following DIN 50 115, three definitions of the
transition temperature are useful, i.e. to fix T
to:
1.) a temperature where the level of impact values is half of the level of the high range,
2.) a temperature, where the fracture area of the specimen shows still 50% of tough fracture
behaviour
3.) a temperature with an impact energy value of 27 J.
Figure 10.5 illustrates a specimen position and notch position related to the weld according to
DIN EN 875. By modifying the notch position, the impact energy of the individual areas like
HAZ, fusion line, weld metal, and base metal can be determined in a relatively accurate way.
Figure 10.6 presents the influence of various alloy elements on the AV-T - curve. Three basically different influences can be seen. Increasing manganese contents increase the impact
values in the area of the high level and move the transition temperature to lower values. The
values of the low levels remain unchanged, thus the steepness of the drop becomes clearer
with increasing Mn-content. Carbon acts exactly in the opposite way. An increasing carbon
content increases the transition temperature and lowers the values of the high level, the steel
becomes more brittle. Nickel decreases slightly the values of the high level, but increases the
130
values of the low level with increasing con-
specimen position:
core longitudinal
specimen shape:
ISO V
300
2% Mn
1% Mn
200
0,5% Mn
100
0% Mn
27
200
J
100
27
13% Ni
8,5%
5% 3,5%
2% Ni
0% Ni
200
0,1% C
100
0,4% C
0,8% C
27
-150
-100
-50
0
Temperature
50
C 100
ISF 2002
br-er10-06.cdr
ergy values together with the highest transition temperature. The development of finegrain structural steels resulted in a clear im-
Figure 10.6
addition, the application of such steels could be extended to a considerably lower temperature range.
With the example of the steels St E 355 and St E 690 it is clearly visible that an increase of
strength goes mostly hand in hand with a decrease of the impact energy level. Another improvement showed the application of a thermomechanical treatment (controlled rolling during
heat treatment). The appliJ
sulted in an increase of
300
S355N
S690N
200
S235J2G3
S355J2G3
100
27
-150
-100
-50
Temperature
50
Figure 10.7
100
131
1,2h 0,25
0,55h 0,25
P
a
CT - specimen
L
h
1,25h 0,13
2,1h
2,1h
SENB -specimen
3PB
specimen width b
bearing distance S = 4h
F,U
crack initiation
U
F
UE,aE
U
measured.
UO
V
br-er10-08.cdr
ISF 2002
of such a test is shown. If the force F is plotted across the widening V, the drawn curve does
not indicate precisely the crack initiation. Analogous to the stress-strain diagram, a decrease
of force is caused by a reduction of the stressed cross-section. If the voltage drop is plotted
over the force, then the start of crack initiation can be determined with suitable accuracy, and
the crack propagation can
F
be observed.
Brinell
(standardised
to
d1
br-er-10-09.cdr
Figure 10.9
132
to the surface of the tested workpiece. The diameter of the resulting impression is measured
and is a magnitude of hardness. The hardness value is calculated from test load, ball diameter, and diameter of rim of the impression (you find the formulas in the standards). The hardness information contains in addition to the hardness magnitude the ball diameter in mm,
applied load in kp and time of influence of the test load in s. This information is not required
for a ball diameter of 10 mm, a test load of 3000 kp (29420 N), and a time of influence of 10
to 15 s. This hardness test method may be
0,200 mm
10
8,9
10
3
10
specimen surface
130
30
0
hardness
scale
6
hardness
scale
100
reference
level for
measurement
4
5
3
8
130
30
0
specimen surface
0,200 mm
100
0
4 5
3
8
0,200 mm
0,200 mm
8,9
reference
level for
measurement
7
10
Terms
Abbreviation
F0
test preload
F1
test load
t0
penetration depth in mm under test preload F0. This defines the reference level
for measurement of tb.
t1
tb
10
HRC
HRA
br-er10-10.cdr
ISF 2002
Hardness Test to
Rockwell
HRB
HRF
e =
Figure 10.10
Hardness testing to Vickers is almost universally applicable. It covers the entire range of materials (from 3 VHN for lead up to 1500 VHN for hard metal). In addition, a hardness test can
be carried out in the micro-range or with thin layers.
Figure 10.10 illustrates a hardness test to Rockwell. In DIN 50103 are various methods standardised which are based on the same principle.
With this method, the penetration depth of a penetrator is measured.
133
At first, the penetrator is put on the workpiece by application of a pre-test load. The purpose
is to get a firm contact between workpiece and penetrator and to compensate for possible
play of the device.
Then the test load is applied in a shock-free way (at least four times the pre-force) and held
for a certain time. Afterwards it is released to reach minor load. The remaining penetration
depth is characteristic for the hardness. If the display instrument is suitably scaled, the hardness value can be read-out directly.
All hardness test methods to Rockwell use a ball (diameter 1.5875 mm, equiv. to 1/16 Inch)
or a diamond sphero-conical penetrator (cone angle 120) as the penetrating body. There are
differences in size of pre- and test load, so different test methods are scaled for different
hardness ranges. The most commonly used scale methods are Rockwell B and C. The most
considerable advantage of these test methods compared with Vickers and Brinell are the low
time duration and a possible fully-automatic measurement value recognition. The disadvantage is the reduced accuracy in contrast to the other methods. Measured hardness numbers
are only comparable under identical conditions and with the same test method. A comparison
of hardness values which were determined
with different methods can only be carried out
for similar materials. A conversion of hardness
values of different methods can be carried out
piston
reference bar
workpiece and whose hardness is then determined in a test machine. If a workpiece on-site
is to be tested, a dynamical hardness test
specimen
br-er10-11.cdr
ISF 2002
Poldi - Hammer
134
hardness test using a Poldi-Hammer. With this (out of date) method, the measurement is carried out by a comparison of the workpiece hardness with a calibration piece. For this purpose
a calibration bar of exactly determined hardness is inserted into the unit, which is held by a
spring force play-free between a piston and a penetrator (steel ball, 10 mm diameter). The
unit is put on the workpiece to be tested. By a hammerblow to the piston, the penetrator
penetrates the workpiece and the calibration pin simultaneously. The size of both impressions is measured and with the known hardness of the calibration bar the hardness of the
workpiece can be determined. However, there are many sources of errors with this method
which may influence the test result, e.g. an inclined resting of the unit on the surface or a
hammerblow which is not in line with the device axis. The major source of errors is the
measurement of the ball impression on the workpiece. On one hand, the edge of the impression is often unsharp because of the great ball diameter, on the other hand the measurement
of the impression using magnifying glasses is subjected to serious errors.
Figure 10.12 shows a modern measurement method which works with ultrasound and combines a high flexibility with easy handling and high accuracy. Here a test tip is pressed manually against a workpiece. If a defined test load is passed, a spring mechanism inside the test
tip is triggered and the measurement starts.
The measurement principle is based on a
Test force
5.0
kp
4.0
2.0
penetration depth the damping of the ultrasonic oscillation changes and consequently
the frequency. This change is measured by
the device. The damping of the ultrasonic os-
br-er10-12.cdr
ISF 2002
Figure 10.12
135
Measurements can also be carried out in confined spaces. This measurement method is not
pulsation range
(compression)
Application
m = a
m > a
m < a
compression -
+ tension
Description
m = 0
m < a
m = a
m > a
yet standardised.
time
pulsation range
(tension)
alternating range
surface with
developer shows
the crack by coloring
Whler line
II
A workpiece is placed
between the poles of
a magnet or solenoid.
Defective parts disturb
the power flux. Iron
particles are collected.
III
Stress
failure line
0
1
10
102
103
104
105
106
Fatigue strength (endurance) number lg N
107
br-er10-13.cdr
ISF 2002
br-er10-14.cdr
ISF 2002
Figure 10.13
Figure 10.14
To test a workpiece under oscillating stress, the fatigue test is standardised in DIN 50100.
Mostly a fatigue strength is determined by the Whler procedure. Here some specimens
(normally 6 to 10) are exposed to an oscillating stress and the number of endured oscillations
until rupture is determined (endurance number, number of cycles to failure). Depending on
where the specimen is to be stressed in the range of pulsating tensile stresses, alternating
stresses, or pulsating compressive stresses, the mean stress (or sub stress) of a specimen
group is kept constant and the stress amplitude (or upper stress) is varied from specimen to
specimen, Figure 10.13. In this way, the stress amplitude can be determined with a given
medium stress (prestress) which can persist for infinite time without damage (in the test: 107
times). Test results are presented in fatigue strength diagrams (see also DIN 50 100). As an
example the extended Whler diagram is shown in Figure 10.13. The upper line, the Whler
line, indicates after how many cycles the specimen ruptures under tension amplitude a. The
Description
136
Application
W ire diameter
Tolerated
deviation
mm
3,2
2,5
2
1,6
1,25
1
0,8
0,63
0,5
0,4
0,32
0,25
0,2
0,16
0,125
0,1
W ire number
mm
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
0,03
0,02
0,01
0,005
Abbreviation
FE 1/7
1 to 7
FE 6/12
FE 10/16
CU 1/7
CU 10/16
AL 1/7
AL 6/12
workpiece
AL 10/16
ISF 2002
W ire length
mm
50
6 to 12
50 or 25
10 to 16
50 or 25
1 to 7
CU 6/12
radiation source
W ire number to
Table 1
6 to 12
10 to 16
Material groups
to be tested
mild
steel
iron materials
copper
aluminium
aluminium
and its alloys
50
50
50 or 25
1 to 7
50
6 to 12
50
10 to 16
W ire material
50 or 25
br-er10-16.cdr
ISF 2002
Figure 10.16
Figure 10.15
Application
Mainly for defects with an orientation
transverse to sound input direction.
Test
methods
described
above
require
br-er10-17.cdr
ISF 2002
Figure 10.17
137
position in the plate plane, but not about the position within the thickness depth. A clear advantage is the good documentation ability of defects.
An information about the
depth of the defect is provided by testing the workpiece with ultrasound. The
principle is shown in Figures
10.17
and
10.18
(principle of a sonar).
The
of
original
br-er10-19.cdr
Figure 10.19
display
138
This method provides not only a perpendicular sound test, but also inaccessible regions can
be tested with the use of so called angle testing heads, Figure 10.19.
30
40
The oblique and rough defect
from 20 to 30 mm provides a
wide echo of 20 to 30 mm.
Starting with SKW 4, an unbroken echo sequence follows.
The inclination of the reflector
is recornised by a change of
the 1st echo when shifting the
test head.
br-er10-20.cdr
ISF 2002
br-er10-21.cdr
ISF 2002
Figure .10.20
Figure 10.21
Figures 10.20 and 10.21
show
macro section
base material
schematically
the
ferrite
+ perlite
bainite
2,5 mm
ferrite
+ perlite
fusion line
weld metal
cast structure
Steel: S355N
(T StE 355)
bainite
because
the
br-er10-22.cdr
Figure 10.22
139
etching with an acid makes the microstructure
visible. The reason is that depending on
structure and orientation, the individual grains
react very differently to the acid attack thus
100
25
Fe
% Fe
% Cr
80
Cr
20
15
20
10
10
Ni
6
4
60
40
% Ni
microstructure is possible.
2
0
0
200
mm
100
100
ISF 2002
Figure 10.23
focused electron beam of high energy, its atoms are excited to radiate X-rays. There is a
simple relation between the wave length of this radiation and the atomic number of the
chemical elements. As the intensity of the radiation depends on the concentration of the elements, the chemical composition of the solid body can be concluded from a survey of the
emitted
X-ray
qualitatively
and
spectrum
quantita-
50
50
20 20
20 20
50
0
10
50
1. weld
2. weld
weld
axis of
bending former
weld
Agents:
- electrolytic copper in the form of chips (min. 50 g/l test solution)
- 100 ml H2SO4 diluted with 1 l water and then
.
110 g CuSO 5 H2O are added
Test:
The specimens remain for 15 h in the boiling test solution.
Then the specimens are bent across a former up to an angle
of 90 and finally examined for grain failure under a
6 to 10 times magnification.
Strau - Test
axis of
bending former
140
solid body can be determined. Figure 10.23 presents the distribution of Ni, Cr, and Fe in the
transition zone of an austenitic plating in a ferritic base metal. The upper part shows the related microsection which belongs to the analysed part. This microanalysis was carried out
along a straight line between two impressions of a Vickers hardness test. The impressions
are also used as a mark to identify precisely the area to be analysed.
The so called Strau test is
standardised in DIN 50
12
web
80
measurement points
tack welds
base plate
weld1
40
40
20
a a
20
aa
12
weld2
120
80
aa
Figure 10.25
Figure 10.25 presents a specimen shape for testing the crack susceptibility of welding consumables. For this test, weld number 1 is welded first. The 2. weld is welded not later than 20
s in reversed direction after completion of the first weld. Throat thickness of weld 2 must be
20% below of weld 1. After
cooling down, the beads
are examined for cracks. If
tensioning bolt
hexagon nut
min. M12 DIN 934
guidance plates
a
tensioning plate
specimen
base body
br-er-10-26.cdr
Figure 10.26
141
surface and root cracks together with information about position, orientation, number, and
length. The welding consumable is regarded as 'non-crack-susceptible' if the welds of this
test are free from cracks.
Figure 10.26 presents two proposals for self-stressing specimens for plate tests regarding
their hot crack tendency. Such tests are not yet standardised to DIN.
thermo couple
electrode
cross-section
groove shape
60
60
welding direction
weld
metal
support plate
Wd./2
H
Wd.
implant
Hc
Wd./2
2
load
temperature in C
specimen shape
load in N
Tmax
start
end crater
150
crack coefficient
C=
x 100 (in %)
800
500
2 3
4 5
sections
60
anchor weld
80
test weld
150
100
60
anchor weld
br-er10-27.cdr
t8/5
ISF 2002
rupture time
br-er10-28.cdr
Tekken Test
Figure 10.27
time in s
ISF 2002
Implant Test
Figure 10.28
There are various tests to examine a cold crack tendency of welded joints. The most important ones are the self-stressing Tekken test and the Implant test where the stress comes from
an external source.
In the Tekken test which is standardised in Japan, two plates are coupled with anchor joints
at the ends as a step in joint preparation see Figure 10.27. Then a test bead is welded along
the centre line. After storing the specimen for 48 hours, it is examined for surface cracks. For
a more precise examination, various transverse sections are planned. The value to be determined is the minimum working temperature at which cracks no longer occur. The specimen shape simulates the conditions during welding of a root pass.
142
The most commonly used cold crack test is the Implant test, Figure 10.28. A cylindrical body
(Implant) is inserted into the bore hole of a support plate and fixed by a surface bead. After
the bead has cooled down to 150C the implant is exposed to a constant load. The time is
measured until a rupture or a crack occurs (depending on test criterion 'rupture' or 'crack').
Varying the load provides the possibility to determine the stress which can be born for 16
hours without appearance of a crack or rupture. If a stress is specified to be of the size of the
yield point as a requirement, a preheat temperature can be determined by varying the working temperature to the point at which cracks no longer appear.
As explained in chapter 'cold cracks' the hydrogen content plays an important role for cold
crack development. Figure 10.29 shows results of trials where the cold crack behaviour was
examined using the Tekken and Implant test. Variables of these tests were hydrogen content
of the weld metal and preheat temperature. The variation of the hydrogen content of the weld
metal was carried out by different exposure to humidity (or rebaking) of the used stick electrodes. Based on the hydrogen content, the preheat temperature was increased test by test.
Consequently, the curves of Figure 10.29 represent the limit curves for the related test.
Specimens above these
heat input: 12 kJ/cm
basic coated stick electrode
plate and support plate thickness: 38 mm
C
Implant-Test
150
Tekken-Test
100
50
150
Preheat temperature
fractured
starting cracks
crack-free
20
100
temperature
50
starting cracks
crack-free
20
0
10
20
30
ml/
40
100 g
increased
10
20
30
ml/
40
100 g
preheat
considerably
development
be-
Figure 10.29
If both graphs are compared it becomes obvious that the tests produce slightly different findings, i.e. with identical
hydrogen content, the determined preheat temperatures required for the avoidance of cracking, differ by about 20C.
143
Figure 10.30 illustrates a method to measure the diffusible hydrogen content in welds which
is standardised in DIN 8572. Figure a) shows the burette filled with mercury before a specimen is inserted. The coupons are inserted into the opened burette and drawn with a magnet
through the mercury to the capillary side (density of steel is lower than that of mercury, coupons surface). Then the burette is closed and evacuated. The hydrogen, which effuses of the
coupons but does not diffuse through the mercury, collects in the capillary. The samples remain in the evacuated burette 72 hours for degassing. To determine the hydrogen volume
the burette is ventilated and the coupons are removed from the capillary side. The volume of
the effused hydrogen can be read out from the capillary; the height difference of the two mercury menisci, the air pressure, and the temperature
provide the data to calculate
the
norm
volume
to pump
hydrogen
under reduced pressure
under
VT
air pressure B
evacuated
standard
conditions.
This
capillary side
meniskus1
meniskus2
mercury
coupons
a) starting condition
the
hydrogen
b) during degassing
br-er-10-30.cdr
Figure 10.30