279

Application of the incremental hole drilling method for the

measurement of residual stress distribution in shot-peened
components.
J. LU and J.F. FLAVENOT
SERVICE FATIGUE RUPTURE
CEntre Technique des Industries Mecaniques (CETIM),
Senlis, France.
Introduction
Shot-peening is now a very well known manufacturing operation,
the industrial applications of which are many and varied. The
residual compression stresses generated in the surface layers of
mechanical parts make this process a very useful tool for improv-
ing the resistance of manufactured parts to fatigue failure, cor-
rosion fatigue or stress corrosion.
Various methods have been developed to measure residual stresses
due to shot-peening. These include : the X-ray diffraction method
(1-3), the incremental hole drilling method (4-6) and the bending
deflection method (7). This document describes recent develop-
ments in the incremental hole drilling method and its applica-
tions in the field of shot-peening, together with the various
measurement technique problems involved in this specific treat-
ment.
Principle of the incremental hole drilling method (8)
The incremental hole drilling method is based upon the principle
of the relief of residual stresses, following the removal of ma-
terial by step-by-step drilling.
For each drilling depth, Z, the surface strains Ei(Z) are mea-
sured using three gauges in a rosette, centred upon the hole
concerned. When drilling is completed, the principal residual
stresses 0'1 (Z) and 0'2 (Z) can be calculated, using equations
the strains ( Ei(Z)) and the calibration co-
(A in and Bin ).
To determine the principal residual stresses (0' 1hi and 0' 2hi ,
with 0'1hi > CT2hi ) and their directions with respect to any
reference axis, for each depth hi, three independent strain
measurements are required. The equation for the radial strain
corresponding to principal residual stresses is :
Equ. 1
Coefficients A in and Bin depend upon the hole diameter, the po-
sitions of the strain gauges, the position of the layer i, the
depth of the hole and the elastic constant of the material. OJ
is the angle between the first strain gauge and the maximum
principal residual stress.
In the new approach, for a given drilling increment, surface
strain changes due to the removal of the previous layer must also
be considered.
Equ. 3
280
strains due to the actual removed layer are
k k n-1 k
En =E
mn
- .! E in k =1,2,3 Equ. 2
k 1=1
E mn is the total strain measured on the surface when the nth
increment is removed. E i ~ corresponds to that part of the total
strain related to the itli layer when the nth increment is re-
moved, and is calculated as follows
1
in =A
in
(olhi + 2hi) + Bin (olhi - 02hi) cos 2e
i
2
in A
in
(olhi + 2hi) + Bin (olhi 02hi) cos 2 (e
i
+ ,)
3
in =A
in
(olhi + 2hi) + Bin (olhi - 02hi) cos 2 (e
i
+ $)
~ and 0/ are the angles between the other two strain gauges,
respectively, and the first strain gauge.
For a 45 rosette, the three unknowns On, (Jlhn and (J 2hn
be calculated using the relationship :
In this case : lfJ=225, 0/=90
can thus
Equ. 4
The calibration coefficients Ain and Bin are calculated using the
finite element method which is modelled digitally, as described
in ref. (6).
Selection of strain rosette used for measurement
In the case of prestress shot-peening, the layer affected by
plastic deformation is relatively thin, of the order of 0.1 rom to
1 rom. optimum gauge configuration is therefore important, to
obtain maximum measurement sensitivity.
Following a series of calculations using the finite element
method (8) (fig. 1), it will be seen that sensitivity is good for
hole depths up to 50 % of hole diameter. Sensitivity decreases
significantly when depth exceeds 50 %. The curves also show that,
in the case bf uniform shear stress, loading ( (J1 = - (J2 )
decreases more slowly than in the case of loading which is axi-
SYmmetrical ( (J 1 =(J2 ). The latter case must therefore be used
as a reference, to define the "significant" measurement area. It
will be seen that the maximum measurable depth is 0.7 d, While
optimum hole depth is considered to be around 0.5 d.
281
~ .100
1<0
130
, 20
110
100
00
or
10
60
SO
H
30
20
10
Fig. 1 : Sensitivity of str{lin measurement in depth for an
axisYmmetrical and uniform shear stress loading.
For shot-peening, the measurement area is generally less than
1 mm. Hole diameter should thus be around 2 mm. Small strain
gauge rosettes are therefore recommended (e.g. VISHAY EA-XX-031
RE-120, EA-XX-062 RE-120, HBM RY61). The best compromise would
seem to be offered by the VISHAY EA-XX-062 RE. Method sensitivity
is around 20 M P a / ~ d for steels and 7 M P a / ~ d for aluminium alloys,
for 20 ~ m of material removed. Smaller rosettes (VISHAY EA-XX-
031-RE) pose a problem of drilling depth and centring accuracy,
but the measured signal accuracy is almost doubled for a given
thickness of material removed.
Drilling technique
As shot-peening residual stresses only cover a very shallow
layer, and the stress gradient is extremely high, the drilling
technique must guarantee extreme accuracy in the depth of hole
drilled (around 1 pm), and must create neither strain nor over-
heating. Also, the bottom of the hole must be flat, to respect
the geometry used in the modelling of the calibration coeffi-
cients. '];'wo drilling systems are recommended, to satisfy the
above requirements (see (6 :
1 - Drilling using a high-speed, compressed air turbine 6),
(9), (10.
2 - Drilling using a fixed-bed milling machine, with a solid
carbide cutter 6.
In the case of shot-peening, i.e. for drilling diameters of
0.8 mm to 1. 6 mm, the optimum system drilling speed is between
4500 rpm and 10000 rpm, depending upon the material and the type
of gauge rosette selected.
282
Applications
Table 1 shows the shot-peening conditions applied to the
materials tested.
Almen
Material N' Shot Coverage
Intensity
Glass beads
Aluminium 7075 la 6 - SA 200 %
d. 0.4 mm
alloy
7075 lb Sl70 10 - 12A 100 %
MN350-10 2 SI70 12 - 14A 100 %
Cast
,
MN700-2 3 SI70 12 - 14A 100 %
Iron
GS7oo-2 4 SI70 12 - 14A 100 %
Table 1
Figure 2 shows the results obtained on the shot-peened 7075 alu-
minium alloy of test pieces 1a and lb.
For this material, it will be seen that the maximum residual
stress value does not vary significantly with the shot-peening
conditions. However, the plastically deformed depth changes,
depending upon the treatment. In this case, the diameter and
nature of the shot are extremely important. An increase in shot
diameter deepens the residual stress and increases the plasti-
cally deformed depth. This result is interesting as, in this type
of material, it has been found that increasing the plastically
deformed layer extends the fatigue life of the part (11).
aR IN/mm
2
)
100
Depth Imm)
<> 1a
1b
- 100
- 300
- 200

Fig. 2 : Comparison of residual stresses obtained on shot-peened
aluminium 7075, with two different treatments.
Residual stresses were measured for shot-peened cast iron compo-
nents (test pieces 2 , 3 and 4), and the results are shown in
figure 3.
283
Resldu81 ,tre.1 (MPA)
- 200
_ 700
, ~
Depth lmm)
MN 350.2
MN 700.2
'" FGS 700.2
shot ell. 0.6 mm
,Imen I. 0.2-0.25 mm
~ ~ Residual stresses obtained on test pieces 2, 3 and 4.
Analysis of the above results leads to the following conclusions:
As in the case of steels, residual stresses generated by shot-
peening increase with the mechanical characteristics of the
material concerned.
Maximum compression stress values obtained for the various
test pieces (2, 3 and 4) are similar to the basic material
tensile strength value for MN 350-10 cast iron, and around
80 % to 85 % of the basic material tensile strength for MN
700-2 and GS 700-2 cast irons. These stress values are thus
very high in relation to the mechanical characteristics of the
cast irons tested. In the case of MN 350-10 cast iron, the
high deformation capacity of this grade (A = 16 % to 18 %)
probably provokes the very high residual stresses in relation
to its tensile strength. It should also be pointed out that
the mechanical properties of cast iron, in compression, are
frequently far superior to its properties in traction.
In terms of appearance and in terms of the residual stress
generated by shot-peening, the behaviour of nodular cast irons
is not very different from that of steels. The value and
distribution of these stresses depend mainly upon the mechan-
ical characteristics (hardness, elastic limit, deformation
capacity) and the shot-peening conditions.
The preceding text describes the analysis of parts which were
simply shot-peened. In industry, another, more efficient treat-
ment is frequently used. This involves subjecting a part to
traction or bending, before the shot-peening operation (strain
peening). Reference (12) describes an example of this technique,
tested using the incremental hole method. In the case of a leaf
spring, the transversal and longitudinal stress results obtained
on parts, shot-peened with and without prestressing, were
compared under various conditions. It was found that the
transversal stresses did not vary significantly with the various
treatments. Conversely, the residual longitudinal compression
stresses (in the direction of the prestress) and the depth of the
area under compression both increased considerably. The value of
the increase varied directly with the prestressing level.
2811
It was also found that the maximum stresses were always close to
the surface. This is probably due to a bending prestress, which
results in maximum stress on the surface. Using this treatment,
the fatigue life of parts could be significantly increased.
To improve the fatigue life of welded parts, finishing treatments
are being increasingly developed, especially since the introduc-
tion of high elastic limit steels, and shot-peening is one of the
most widely used methods. Figure 4 shows the results obtained on
fatigue test pieces which were welded and then shot-peened along
the weld toe, under various conditions (13). In this case, a spe-
cial VISHAY rosette (CEA-XX-062 UM-120) was used, to get as close
as possible to the weld. Residual shot-peening stresses (compres-
sion) were observed on the surface, with welding stresses (trac-
tion) in the sub-layer.
JOO
100
100
-1M
-1eo
"
, '
7 , ~ 9 10
- .;;-;;;:; 11/10..1
r SHOT PEENING
I. CONDITIONS
~ f ~ 1 A'I 5130
1.+
~ f 55 A II ..".. 5130
o f 55 A ;0'0 5330
f 10 C.J 5J30
Fig. 4 :. Shot-peening residual stress profiles, measured by the
incremental hole drilling method, for E 36.4 steel, in the direc-
tion perpendicular to the weld toe (crT).
The duplex austenite-delta-ferrite steels are becoming increas-
ingly used in modern industry. Figure S shows the results ob-
tained on shot-peened URANUS 4S steel, using the incremental hole
method and the X-ray diffraction method (the X-ray diffraction
tests were conducted at the ENSAM-PARIS). The stresses measured
using the former method were macroscopic and those measured using
the latter method were residual stresses measured in both phases
(Q.' and Y). It will be seen that the two methods give similar re-
sults in the case of shot-peening as, in this case, the residual
stresses present in each material phase are not significantly
different. Other comparisons between the hole drilling and X-ray
methods also show comparable results for the two methods, which
are based upon very different principles (6) (14).
285
RESIDUAL STRESS
(MPel
(mm)
X-RAY {o
METHOD
HOLE {
DRILLING
METHOD
100
-100
-1000
0.4 0.5 0.6
0.1 0.2 0.3 .
Depth
Phase Y l.AkQ:.
Mn
}
AXIAL
Phase Q: (.Aka Cr I
AXIAL
TANGENTIAL
Fig. 5: comparison between the hole drilling methode and X-ray
diffraction method, on URANUS 45 steel. Conditions
cast steel shot, dia. 1 mm., Almen Intensity 19 - 21 A.
Conclusion
The above results indicate that the incremental hole drilling
method is suitable for measurement of the residual stress profile
in depth, in the case of shot-peening. The accuracy provided by
the latest developments in this technique not only allows the
maximum residual stress to be measured, but also the compression
stress layer to be defined, which has a considerable influence
upon the effectiveness of prestress shot-peening. Using this
method, a specific treatment can rapidly be optimised, as a
stress profile only requires two hours of testing. The field of
application of the method, for this type of treatment, was dem-
onstrated by applying the method to various materials (steels,
light alloys, cast irons, two-phase steels).
The specific measurement problems posed by shot-peened parts have
been discussed, such as optimum gauge rosette selection, drilling
conditions, etc
The incremental hole drilling method is one measurement method
which complements other existing methods. In addition, it offers
advantages for coarse grain materials or those with significant
sUb-layer texture, which pose problems when using the X-ray dif-
fraction method (stainless steel, nickel-based alloy, for example
IN 100).
Recent developments in the incremental hole drilling method, both
theoretical and experimental, will probably result in the early
development of industrial applications of this method, due to its
simplicity and rapidity.
References :
(1) G. Maeder, J.L. Lebrun, A. Diament: "1st International
Conference on shot-peening", Paris, (1981).
(2) L. Castex: "Groupement fran9ais pour l'analyse des
contraintes par diffractometrie X", (1984).
(3) A. Saint Etienne, F. Lecroisey, B. Miege: "CETIM Technical
Memo No. 33", (1978).
(4) 5.5. Birley, A. Owens, D. Clarke: "The application of
residual measurement in welded aluminium alloy sections",
(1978) .
286
(5) J. LU, A. Niku-Lari, J.F. Flavenot: "5th International
Congress on Exp. Mech.", SESA, (1984) 678.
(6) J. LU, A. Niku-Lari, J.F. Flavenot: "Revue de Metallurgie"
SFM, 2 (1985) 69.
(7) J .F. Flavenot, A. Niku-Lari: "CETIM Technical Memo No. 37"
(1977) .
(8) J. LU, A. Niku-Lari, J.F. Flavenot: "Materiaux et
Technique", 12 (1985) 709.
(9) M.T. Flaman: "Exp. Mech." 1 (1982) 26.
(10) M.T. Flaman, J.A. Harring: "Exp. Tech." 1 (1986) 34
(11) A. Niku-Lari, G. Gillereau: "2nd International Conference
on shot-peening", (1984) 102.
(12) J. LU, J. Flavenot: "CETIM Information", 98 (1986) 60.
(13) C. Bouhelier: "Les contraintes residuelles dans les
constructions soudees", CETIM, (1986).
(14) J. LU, A. Niku-Lari, J.F. Flavenot: "The effects of
fabrication related stresses on product manufacture and
performance", Paper 10, The Welding Institute, (1985).