Correction of errors introduced by hole eccentricity in residual stress

measurement by the hole-drilling strain-gage method

Paul Barsanescu
a,1
, Petru Carlescu
b,
*
a
Technical University Gh. Asachi, Blvd. Mangeron 59, Iasi 700050, Romania
b
University of Agricultural Sciences and Veterinary Medicine, Aleea Sadoveanu 3, Iasi 700490, Romania
a r t i c l e i n f o
Article history:
Received 29 April 2008
Accepted 7 September 2008
Available online 18 September 2008
Keywords:
Residual stress measurement
Accuracy
a b s t r a c t
The accuracy of the hole-drilling strain-gage technique, the most widely used method for
residual stress measurement, is directly related to: (1) the eccentricity between the hole
center and rosette center, respectively, and (2) the accuracy of corrections for integration
tendency of the strain gages. Although the drilling tool is optically aligned, the eccentricity
always cannot be zero and it is a source of error, affecting the accuracy of this technique. An
equation for correction of the data affected by hole eccentricity has been proposed by A.
Ajovalasit, but it is not precise enough, because it is founded on the wrong hypothesis that
in the strain gage area there is a uni-axial state of stress, instead bi-axial one. The paper
presents a new equation for data correction, which take account of the plane (bi-axial)
state of stress in the strain gages area. After the hole-drilling and data acquisition, the hole
eccentricity will be measured, using an optical microscope. Knowing the eccentricity, the
data will be corrected using the proposed equation. This equation is more precise, because
it is founded on a realistic hypothesis (bi-axial state of stress in the strain gages area).
2008 Elsevier Ltd. All rights reserved.
1. Introduction
A predominant factor contributing to structural failure
of machine parts may be the residual stresses. Hole drilling
is the most widely used method for residual stress mea-
surement [1,2]. ASTM E837-01 is the accepted standard
for residual stress measurement by the hole-drilling meth-
od [3]. A detailed explanation of this ASTM standard is pre-
sented in the Technical Note TN 503 by the Measurement
Group [4]. The accuracy of this technique is directly related
to the precision with which the hole is drilled through the
center of strain-gage rosette. In practice, the required
alignment precision is to within 0.025 mm [3]. In order
to obtain a so small eccentricity, the drilling tool is opti-
cally aligned so that its drilling axis is precisely positioned
over the target at the center of the strain-gage rosette. But,
although the tool is optically aligned, the eccentricity can-
not be zero and eccentricity of the hole is a source of error,
affecting the accuracy of this technique [57]. An equation
for correction of the data affected by hole eccentricity has
been proposed by Ajovalasit [8], but it is not precise en-
ough, because it is founded on the wrong hypothesis: in
the strain-gage area there is a uni-axial state of stress, in-
stead bi-axial one.
The paper presents a new analytical method for exper-
imental data correction, in order to diminish the errors
introduced by the hole eccentricity and to improve the
accuracy of the hole-drilling technique. This method take
account of the plane (bi-axial) state of stress in the
strain-gages area. The case of the through hole is studied.
2. Description of the actual methods
Rendler and Vigness [5] observed for the rst time that
the accuracy of the method. . . will be directly related to
the operators ability to position the milling cutter
0263-2241/$ - see front matter 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.measurement.2008.09.002
* Corresponding author. Tel.: +40 232 275070; fax: +40 232 260650.
E-mail addresses: paulbarsanescu@yahoo.com(P. Barsanescu), pcarlescu
2004@yahoo.com (P. Carlescu).
1
Tel./fax: +40 232 232337.
Measurement 42 (2009) 474477
Contents lists available at ScienceDirect
Measurement
j our nal homepage: www. el sevi er. com/ l ocat e/ measur ement
precisely in the center of the strain gage rosette. More re-
cent works have quantied the error in calculated stress
due to eccentricity of the hole [9,10]. In the conditions of
a non-uniform stresses distribution around the hole, the
integration tendency of the strain gages is an important
source of error [9]. This error can be corrected using the
average theorem from the integral calculus.
The output of the strain gage is proportional with the
average strain on the grid area [11].
In the non-uniform axial state of stress, the W. Soette
and R. Van Crombrugge method [11] can be used in order
to correct the experimental data affected by errors. The
output of the strain gage will be

e
1
x
2
x
1
_
x
2
x
1
e
x
x dx 1
where x
2
x
1
= GL is the length of the strain gage grid; e
x
(x)
is the strain in the direction of the strain grid laments.
In the non-uniform bi-axial state of stress, the R.G. Boi-
ten and Ten Cate method [11] can be used in order to cor-
rect the experimental data

e
1
S
_
S
e
x
x; y dS 2
where S is the strain gage grid surface.
In the paper [8], Ajovalasit presented an analytical
method that can be used in order to correct the experimen-
tal data affected by the hole eccentricity. Because the
mathematical problem is complex one, he used Eq. (1) in-
stead of Eq. (2). Using this way, we must do only a usual
integral, instead a double one but, of course, the precision
of obtained equations is affected. The present paper use Eq.
(2) in order to obtain a better formula for experimental
data correction (affected by the hole eccentricity).
3. Presentation of a new method
In Fig. 1, a strain gage grid near a hole with eccentricity
is presented. The wrong position of hole has the center in
O
0
, instead of O. In Figs. 2 and 3, the gage grid reported at
its own system of axis is presented. The average value of
the strain gage output can be determined using the R.G.
Boiten and Ten Cate method

e
1
S
_
x
2
x
1
dx
_
y
2
y
1
e
x
x; y dy 3

e
1
S
_
x
2
x
1
dx
_
0
y
1
e
x
x; y dy
_
x
2
x
1
dx
_
y
2
0
e
x
x; y dy
_ _
4
For a symmetric function (as e
x
in the area near a hole in a
plate in tensile), we ca write
_
0
y
1
e
x
x; y dy
_
y
1
0
e
x
x; y dy 5
From Eqs. (4) and (5) we can write

e
1
S
S
00
S
00
_
x
2
x
1
dx
_
y
1
0
e
x
x; y dy
S
0
S
0
_
x
2
x
1
dx
_
y
2
0
e
x
x; y dy
_ _
6
and

e
1
S
S
00

e
00
S
0

e
0
_ _
7
where the grid surface is
S S
0
S
00
8
The main residual stresses are calculated for a plate in a
plane state of stress, with a through hole in it, function of Fig. 1. Strain gage grid near an eccentric hole through a plate.
Fig. 2. Strain gage grid in xy system of reference.
Fig. 3. Strain gage grid in x
00
y
00
system of reference.
P. Barsanescu, P. Carlescu/ Measurement 42 (2009) 474477 475

a and

b coefcients, which take account of integration ten-
dency of strain gages. The two coefcients are well known
from the literature [10] and are determined for a centric
hole (with the center in O). For this particular situation,
we have (Figs. 2 and 3)
y
1
y
2

GW
2
9
S
0
S
00

S
2
10

e
0


e
00


e 11
In order to calculate the residual stresses for a through cen-
tered hole,

a and

b coefcients are used (Fig. 4):

a
2
wr
2
r
1

1
r
2
h
1
h
2
12

b
21m
wr
2
r
1

1
r
2

21m
1m
h
1
h
2
sin2h
1
sin2h
2

1
r
2
sin2h
1
cos
2
h
1
2r
2
1

sin2h
2
cos
2
h
2
2r
2
2
_ _
_
13
where
w
2GW
D
r
1

2R
1
D
r
2

2R
2
D
r
D
D
0

D
2R
0
D R
1
R
2
14
Inorder tocorrect theerrors introducedbyhole eccentricity.
For an eccentric hole, the correction can be made simply
using Eq. (7) and coefcients

a
e
;

b
e
, instead of

a and

b. The

a
e
;

b
e
coefcients are calculated for two strain gages near a
centric hole, having the grid width 2(GW/2 + DGW) and
2(GW/2 DGW), respectively

a
e

1
GW
GW
2
DGW
_ _

a
0

GW
2
DGW
_ _

a
00
_ _
15

b
e

1
GW
GW
2
DGW
_ _

b
0

GW
2
DGW
_ _

b
00
_ _
16
where

a
0
;

b
0
and

a
00
;

b
00
are determined using Eqs. (12) and
(13), but applied for S
0
and S
00
, respectively, surfaces.
Developing the equations we obtain

a
e

R
2
0
GW GL
Dh
0
Dh
00
17

b
e
21 m

a
e

R
2
0
1 m
GW GL
_
sin2h
0
1
sin2h
0
2
_ _
sin2h
00
1
sin2h
00
2
_ _

R
2
0
2
1
R
2
1e
sin2h
0
1
cos
2
h
0
1
sin2h
00
1
cos
2
h
00
1
_ _
_

1
R
2
2e
sin2h
0
2
cos
2
h
0
2
sin2h
00
2
cos
2
h
00
2
_ _
__
18
where
Dh
0
h
0
1
h
0
2
Dh
00
h
00
1
h
00
2
19
The method presented above it is a very easy way to cor-
rect the errors introduced by eccentricity. It is in the same
time more accurate, because Eq. (2), good for a plane stress
of state, was used in this demonstration, instead of Eq. (1),
good for axial state of stress and used by the actual meth-
ods for error correction.
4. Conclusions
The hole eccentricity of is an important source of error,
which affect the accuracy of residual stress measurement
by the hole-drilling method. Although in practice the opti-
cal alignment is used, the eccentricity cannot be zero.
A very easy and more accurate method for errors intro-
duced by eccentricity correction has been presented in this
paper. It takes account of the plane (bi-axial) state of stress
in the strain gages area reduces the case of the eccentric
hole at the well known case of the centric hole.
Acknowledgements
The researches presented in this paper are partially sup-
ported by Romanian Ministry of Education and Research,
Contract No. 49/2006 (CEEX Rolight) and Contract No.
7016/2007 (PN II Modis).
References
[1] F.A. Kandil, J.A. Lord, P. Mummery, G.A. Webster, UK needs for
residual stress measurement, N.P.L. Report, C.M.M.T., vol. 7,
December 2001, p. 177.
[2] J. Lu (Ed.), Handbook of Measurement Residual Stresses, SEM-CETIM,
Fairmont Press, 1996.
[3] ASTM E837-01, Standard Test Method for Determining Residual
Stresses by the Hole-Drilling Strain Gage Method, Annual ASTM Book
of Standards. Available from: <www.astm.org>.
[4] Measurements Group, Technical Note TN 503, Measurement of
Residual Stresses by the Hole-Drilling Strain Gage Method. Available
from: <www.vishay.com/brands/measurements_group/guide/tn/
tn503/>. Fig. 4. Centered hole with a strain gage grid.
476 P. Barsanescu, P. Carlescu/ Measurement 42 (2009) 474477
[5] N.J. Rendler, I. Vigness, Hole drilling strain-gage method of
measuring residual stresses, in: Proceedings of the SESA XXIII, vol.
2, 1966, pp. 577586.
[6] J.P. Sandifer, G.E. Bowie, Residual stress by blind-hole method
with off-center hole, Experimental Mechanics 18 (1978) 173
179.
[7] H.C. Wang, The alignment of the hole-drilling method, Experimental
Mechanics 17 (1979) 2327.
[8] A. Ajovalasit, Measurement of residual stresses by hole drilling
method: inuence of hole eccentricity, Journal of Strain Analysis 14
(4) (1979) 171178.
[9] E. Procter, E.M. Beaney, Recent developments in center-hole
technique for residual-stress measurement, Experimental
Techniques 6 (1982) 1015.
[10] S. Redner, C.C. Perry, Factors affecting the accuracy of residual stress
measurement using the blind-hole drilling method, in: Proceedings
of the 7th International Conference on Experimental Stress Analysis,
Haifa, Technion, 1982.
[11] L. Pet, P. Barsanescu, O. Ciobanu, Analytical and Numerical Methods
Used to Correct the Errors Introduced by Integrating Tendency of the
Strain Gages, Bulletin of Technical Institute of Iasi, Tom XXXIX
(XLIII), Fasc. 14, Sect. V, 1993, pp. 113124.
P. Barsanescu, P. Carlescu/ Measurement 42 (2009) 474477 477