Residual Stress analysis

Informations: strains
Macro elastic strain tensor (I kind)
Crystal anisotropic strains (II kind)

C
Macro and micro stresses
Applied macro stresses

Fe

Cu

Residual Stress/Strain definition

!' : Macrostress
!'' : Microstress
!''' : r.m.s. Microstress

!'''
!''

!'
x

Experimental setting

Measurement of a high 2theta peak position for different tilting of the

sample
The sample can be tilted in omega or psi (chi)
Changing phi we will scan different direction for the stress in the sample
The simple behavior is a linear relationship between the measured dspacing and sin2psi . The slope is proportional to the macrostrain.

a(#1,#2,#3)

$(!,")

Strain measurement
cosN cos R sinN cosR & sinR

x L i ' ( ij x S j ,

( ij ' R ik N kj '

& sinN

cos N

cosN sinR

sinN sinR

cos R

; Hhkl2 x3 .

!"# $%&'#( !) ") #The

$%*full
+$)formula:
,"# -#.,/0( ! $%* ") $%* ,"#
L 3 $0# *#4%#* 1% 51&6 2 71," 0#(8#., ,/ ,"#
*10#.,1/%( 2) + e$%*
33 / eR N ' d RN / d 0 & 1
($98'# $%* ,"# :;0$< =#$96 !"# 801%.18'# /> ,"# !?(1%+"
9#,"/* 1( *#81.,#* 1% 51&6 + >/0 ,"0## *1>>#0#%, " $%&'#(6
2
2
2
%
&
' ,"#
% [e/>
%
N
N
R
N
e
]
sin
e
sin
e
sin2
eR N @%
e 33.$(#
cos
$
9$,#01$'
71,"/A,
80#>#00#*
.0<(,$''/&0$8"1.
33
12
22
11
/01#%,$,1/%() $% '1%#$0
0#'$,1/%("18
/=,$1%#*
=#,7##% ,"#
N % e23 sinN1(] sin2
R
[e13 cos
9#$(A0#* '$,,1.# 8'$%# *1(,$%.#( $%* (1%+"B ,"# 0#(1*A$';(,0#((
.$% =# *#01-#* >0/9 ,"# ('/8# /> ,"1( .A0-#6 5/0 ,"# .$(#
6 -$'A#
unknowns:
over-determined system, least-squares
/> $ =1$C1$' (,0#(( (,$,# 1% ,"# 8'$%# /> ,"# ./$,1%& D% 22 E % ++ E
% !) % 33 E FG) ,"# 0#'$,1/%("18 =#,7##% (,0$1%( $%* (,0#((#( 1(
&1-#% =< DH/<$% Isotropic
I J/"#%)planar
2KLMG stresses:
+
"#$
&!" ! "!"#$ #!' "$ % !F &(!F ! 2+%"#$
+ %! (1% " ' +%2 %! '

#2$

7"#0# !"#$D!) "G $0# ,"# #C8#019#%,$' '$,,1.# 8'$%# *1(,$%.#(

1% ,"# *10#.,1/% ."$0$.,#01N#* =< ,"# $%&'#( ! $%* ") !F 1( ,"#
"#$
(,0#((;>0## '$,,1.# 8$0$9#,#0) $%* 2+%"#$
+ $%* %2 $0# ,"# :;0$<
#'$(,1. ./%(,$%,(6 !"# :;0$< #'$(,1. ./%(,$%,( *#8#%* /% D"#$G

56780) !

\01%.18'# /> ,"# !

0#(8#.,1-#'< ,"# %
%/09$' ,/ ,"# *1>

!"!" #$%&' (

!"# X@UST
"$( =##% *#-#
./$,1%&( DT# U
*#%,;=#$9 $%&
,"# :;0$< =#$9

(22)

eviation, aE,
find

example of the procedures. Since the averages will be

taken over single-crystal compliances s'33u, the X-ray
elastic constants will result in the Reuss limit. It will be
assumed that theShear
orientation
distribution of the
stresses
crystallites can be idealized by some combination of

1
(23a)
(23b)

cept of al vs
ult, note that
components
). For q~=0,
(e22-/333) is
an then be
45 . From
e23 when ~0

nto account,
d from

tp=45

?
0.5

./

t--

tO= 90~"

o/

-~0.
-.5'

L..S"
o.

.5 ol

.5 o.

.5

sin2qu
Fig. 4. Lattice strain vs sin 2 ~Omeasured at the 211 reflection of ground
steel (D611e & Cohen, 1979). Measured values: ~b>0; ~O<0.
( - - ) Calculated from stress tensor (25). (---) Average strain at
(23a).

Non linear behavior

In many cases oscillation of d vs. sin2psi are observed; some possible

causes:
Textured sample -> the elastic tensor is anisotropic.
Plastic deformation: anisotropy of the plasticity behaviour and elastic
tensor results in anisotropy of the residual stresses/strains
Thermal expansion anisotropy
Shear stresses normal to the surface
Coherent and semicoherent interfaces (in thin film.)

Dolle in 1979 (J. Appl. Cryst., 12, 489) analyzed the problem in general and
was followed by other authors: Noyan and Nguyen for the plastic
deformation, Barral et al. for the texture connection.

Texture-Stress
Procedure:
Measurement of the texture ODF by traditional pole figures
Measurement of the d-spacing vs. sin2psi for high angle reflections
Computation of the effective macro-elastic tensor using single crystal elastic
constants and the ODF
Different theories can be used to average the elastic tensor over the ODF:
Voigt (stress compatibility)
Reuss (strain compatibility)
Hill (mean value between Reuss and Voigt)
Self Consistent, FEA.... (costly)
Geometrical mean

Analysis of the d-spacing vs. sin2psi using the averaged elastic tensor
Pro:
You control the entire process
Cons:
Lengthly procedure, two measurements, two analyses
Does not work (very difficult) for highly stressed or strongly textured materials

Traditional methods for the ZrO2 films

1.575
1.570

1.560
1.555
1.550
1.545

Voigt model (no texture)

Reuss model (texture)
0

0.2

0.4

0.6

sin2(!)

<C> vs. psi

0.8

210

1
Elastic modulus [GPa]

d113 []

1.565

200
190
180

ZS1
ZS2
ZS3

170
160

20

40
60
! [degrees]

80

ZrO2 thin films

Intensity [counts]

2000

1500

1000

500

0
25

35

45

55

2!

65

Macro residual stress on the ZrO2 serie

0
Whole pattern analysis

Residual stress [GPa]

-1

sin ! method
2

sin ! with texture

2

-2
Voigt model
F&L model

-3

Reuss model

-4
-5

0.3

0.5

0.7
Thickness [m]

1.2

Measuring the stress also by the curvature

2
d
a 11
"
s
! =K#
#
df 2

16
14

---- data
___ fit

12

[m]

10
8
6
!2y/!x2=-2.9272e-7 m-1

4
2
0
0

4000

8000
12000
scan length [m]

16000

20000

Comparison of results

method:

"11 = " 22

XRD: sin2!
curvature method
XRD
(220) plane (200) plane (113) plane Stoney's modified Ferrari and
formula
formula Lutterotti
-2.73
-1.66
-3.06
-1.36
-1.48
-2.92

[GPa]
" *c , i = j
" *c , i # j

[GPa]

0.631
0.043

Stress-texture for a zirconia film (WIMV)

Residual Stresses/Texture analysis

Voigt model + WIMV

ZrO2 film: results

Very high in plane residual stresses
(compression):
Reuss model: 3.6 GPa
Bulk Path GEO: 3.47(5) GPa
Curvature method: > 10 Gpa !?
Thickness: 320 Nanometer
Reconstructed pole figures

Experimental errors

Example: the CPT film shows big shift of the peaks increasing !.
The shift is not smaller at low 2theta angle.
In the fitting was perfectly reproduced by a beam 0.59 mm higher than the goniometer
center.
Using the Rietveld method peak shifts from low angle positions are also used normally ->
good sample positioning required, perfect alignment of the instrument also.