Residual Stress Measurement in Cold Worked Materials, Influence of Detector
Size on Stress Results, Simulation and Experiments
M. Belassel1, E. Bocher2, J. Pineault1.
1: Proto Manufacturing Ltd., Canada.
2: Renault Technocentre, France.
To enhance the fatigue resistance of mechanical components, different surface
treatment processes are often applied to put the near surface layer into compression.
Surface treatment processes are typically associated with deformation and
work-hardening of the material. When applying x-ray diffraction techniques to the
characterization of such surfaces, the work-hardening will cause the x-ray
diffraction peak width to increase. When peak widths reach high values, the peak
tail may extend beyond the active area or window of the multi-channel x-ray
detector, in which case the peak is truncated. Subsequent analytical treatment of
broad diffraction peaks is troublesome and advanced numerical methods are required
to accurately determine the peak position. Even though, when the detector window is
only 3 times or less the peak width, all techniques seem to have difficulties to
accurately determine the peak position.
The stress result is highly affected by the accuracy of the peak position determination. Different examples will be demonstrated to emphasize the effect of the detector width and the peak location technique on residual stress determination using simulation and experiments. After many studies conducted on this subject, we have concluded that the only possible solution when it is available is to increase the detector width. In this case, the x-ray diffraction peaks do not show any truncation and lead more stable and accurate results. In other cases, alternative x-ray diffraction peaks with reduced width and appropriate parameters can be used. As a result, residual stress can be determined with much improved repeatability and reproducibility.
SMARTS, the Engineering Neutron Diffractometer at LANSCE
B. Clausen, D.W. Brown, S. Kabra, T.A. Sisneros and M.A.M. Bourke. LANL,
Los Alamos, NM.
SMARTS (Spectrometer for Materials Research at Temperature and Stress) is the
engineering instrument at Los Alamos Neutron Science Center, LANSCE. It entered
the user program in August of 2002 and has since served more than 150 unique users
performing engineering related neutron diffraction measurements at LANSCE. Its
design maximizes capability and throughput for measurements of (a) residual
macroscopic strain in engineering components and (b) internal strains during
in-situ loading. The former is achieved using radial collimators to define the
beam along the beam direction, which together with the incident slits define a
cuboidal gauge volume that can be positioned at any given position and orientation
within a component. The latter is facilitated by the world leading ancillary
equipment at SMARTS, which enables in-situ measurements at uniaxial loads up
to +/- 250 kN, at temperatures between 90 and 2000K. Examples of both spatially
resolved and in-situ type measurements will be given.
Laser Shock Peening of Aero Engine Materials
Amrinder Gill, Yixing Zhao, S. R. Mannava and Vijay K. Vasudevan, University of Cincinnati, OH.
Laser shock peening (LSP) is a novel surface engineering process that generates deep compressive
residual stresses and microstructural changes and thereby dramatically improves fatigue strength,
service lifetimes and resistance to crack propagation of critical metal parts like aircraft engine
fan and compressor bladed. The present study was undertaken to develop a basic understanding of the
effects of LSP parameters on the residual stress distributions and microstructural changes in two
important aero engine alloys, IN718 and Ti-6Al-4V. Coupons of the alloys with and without a
sacrificial/ablative layer were LSP-treated using the GENIV system at GE Infrastructure Aviation.
Depth-resolved characterization of the macro residual strains and stresses and degree of cold work
was achieved using high-energy synchrotron x-ray diffraction. The near-surface and through-the-depth
changes in strain, texture and microstructure were studied using EBSD/OIM in an SEM and by TEM of
thin foils fabricated from specific locations using the focused ion beam method. Local property
changes were examined using microhardness and nanoindentation measurements. The results showing
the relationship between LSP processing parameters, microstructure, residual stress distributions
and hardness are presented and discussed.
Characteristics of Residual
Stress Profiles in Hard Turned versus Ground Surfaces With and Without a White Layer
A.W. Warren, Y.B. Guo, University of Alabama, Tuscaloosa, AL.
Hard turning and grinding are competitive processes for manufacturing various
mechanical products. Product performance is highly dependent on the process
induced residual stresses. However, there exists significant ambiguity
regarding the true residual stress profiles generated by hard turning and
grinding with and without the presence of a white layer. This study aims to
clarify the pressing issues via an extensive residual stress measurement for
five surface types: hard turned fresh (HTF), hard turned with a white layer
(HTWL), ground fresh (GF), ground with a white layer (GWL), and as heat
treated. X-ray diffraction with a Co source was used to calculate the residual
stress profiles in the near surface (˜100 µm). The x-ray diffraction data
revealed distinct differences in the residual stress profiles for the various
machining conditions. Specifically, the results have shown that (i) HTF surfaces
generate a “hook” shaped residual stress profile characterized by surface
compressive residual stress and maximum compressive residual stress in the
subsurface, while GF surfaces only generate maximum compressive residual stress
at the surface; (ii) HTWL surfaces generate a high tensile stress in the white
layer, but has highly compressive residual stress in the deeper subsurface than
the HTF surface; (iii) GWL surfaces only shift the residual stress to more
tensile but does not affect the basic shape of the profile; (iv) Tensile
residual stress in the HTWL surface is higher than that for the GWL one.
However, the residual stress for the ground white layer does not become
compressive and remains tensile in the subsurface; (v) Elliptical curve
fitting is necessary for measuring residual stress for the HTWL surface due
to the presence of shear stress induced severe ψ splitting in the x-ray data;
(vi) Residual stresses by grinding show more scattering than those by hard
turning; and (vii) Machining is the deterministic factor for the resulting
residual stress magnitudes and profiles compared with the minor influence of
initial residual stress by heat treatment.
Modeling of Stresses in Layered Systems
C. H. Hsueh, Oak Ridge National Laboratory, Oak Ridge, TN.
Advances in modeling the residual stresses generated in multi- and graded-layer
systems are summarized and the corresponding applications are addressed. (1) Unique
analytical models containing only three unknowns are developed to derive simple
closed-form solutions. The solutions are exact for locations away from the free-edges
of the system. Researchers can now predict how the properties and thickness of
multiple or graded layers influence the residual stress distribution in the system
using these solutions instead of relying on try-and-error or case-by-case computer
simulations. (2) For the first time, exact closed-form solutions for i. the
interfacial peeling moment resulting from localized stresses normal to the interface
and ii. the interfacial shear force resulting from localized shear stresses in the
edge region are derived for each interface in multilayers to respectively characterize
modes I and II edge delamination due to residual stresses. These solutions provide
guidelines for selecting the properties/thickness of each layer to mitigate residual
stresses-induced edge delamination in multilayers.
Other examples: (A) Contact-induced radial cracking in
ceramic coatings on compliant substrates is of interest. In this case, radial
cracks initiate at the coating/substrate interface beneath the contact where maximum
flexural tension occurs. The analytical expression is derived for ceramic bilayer
coatings on compliant substrates, which have significant applications in the
structure of dental crowns. The predicted critical loads for initiating radial
cracking are in agreement with existing measurements and finite element results,
and the solutions hence provide guidelines for selecting the properties/thickness
of the bilayer with the enhanced critical cracking load. (B) The effects of soft
adhesive interlayers on contact-induced radial cracking in brittle coatings on
supporting substrates are of interest for dental crowns, windshield, and other
laminated structure applications. An analytical model is derived to illustrate
how the properties/thickness of the adhesive interlayer influence the critical
load to initiate radial cracks in the brittle coating. (C) A combined
empirical-analytical method is developed to analyze the contact radius and indenter
displacement during Hertzian indentation on coating/substrate systems.
This allows one to consider the presence of substrate and to deconvolute elastic
properties of thin coating from the measured load-displacement relation.
(D) Cracking of brittle films on substrates is a major reliability problem in
microelectronic devices, protective coatings, and other thin film applications.
An analytical model is developed to predict the crack density in the brittle film
as a function of the applied strain for a film/substrate system subjected to
residual stresses and unidirectional loading. The solutions allow the film strength
(or fracture energy) to be characterized. (E) Closed-form solutions for the
elastic stress distributions in multilayered discs subjected to biaxial flexure
tests are derived for the first time. The solutions are verified by finite element
results, and the closed-form solutions hence provide a basis for evaluating the
biaxial strength of multilayers using biaxial flexure tests.
Evaluation of Residual Stress of Indentation Method
Hong Chul Hyun1, Jin Haeng Lee2 and Hyungyil Lee1.
1: Sogang University, Seoul, Korea.
2: University of Tennessee, TN.
By FE analyses of sharp indenters, this work then examine the relationships
between indentation parameters and residual stresses. We observe that hardness
is strongly dependent on the magnitude and sign of residual stress and material
properties, while prior indentation studies reported that hardness is hardly
affected by the residual stress. We select some dimensionless indentation
parameters which are free from the effects of material properties and
tip-rounding. With numerical regressions of the data obtained, we propose
indentation formulae for equi-biaxial residual stress evaluation. The proposed
indentation approach provides a substantial enhancement in accuracy compared
with the prior methods. We then extend the equi-biaxial indentation theory to
the evaluation the non-equi-biaxial residual stress state using 3D FE models
for conical indentation. We finally verified the proposed method for residual
stress evaluation via 4-point bending experiment of SS400 steel beam.
Laser-based Optical Strain Rosette for Residual Stress Measurement
Keyu Li. Oakland University, Rochester Hills, MI.
A laser based Optical Strain Rosette is developed to measure residual stresses.
The laser technique is based on laser reflection and interference fringe patterns.
Our technique development research has been published in journals and conferences.
We measure residual strains in x and y direction and shear stress - three
components. We obtain principal stresses (maximum and minimum stresses) and
principal direction of stresses. One location measurements are for several
increments of drilling into the materials surface. Residual stresses in each layer
of material removal are calculated from the measured strains by using incremental
finite element method and integral method.
There are incremental hole-drilling and ring-core cutting methods of material removal. Hole-drilling (diameter =0.8mm) depth to 0.5 mm and ring-core cutting (inner diameter= 2 mm) depth to 1.0 mm. Results of residual stress distribution with depth (mm) can be determined and plotted. The laser strain rosette method has advantages of non-contacting, high accuracy and extremely short gage length on the order of 100 micrometers compared with a resistance strain rosette. It is particularly useful at local areas with stress concentrations and sharp change of geometries.
Practical Analysis of Residual Stress Effects on Fatigue Crack Growth Using NASGRO
R. Craig McClung, Brian M. Gardner and Yi-Der Lee. Southwest Research Institute, San Antonio,
TX.
Practical analysis of fatigue crack growth (FCG) rates and lifetimes in stable
residual stress (RS) fields is possible with the NASGRO fracture mechanics
software. Calculations employ weight function (WF) stress intensity factor
(SIF) solutions and a superposition of applied and RS fields. NASGRO provides
an extensive library of advanced univariant and bivariant weight function SIF
solutions for embedded, corner, surface, and through cracks. Recent NASGRO GUI
modifications make it easy to input and superimpose residual and applied stress
gradient fields, including novel optimum point spacing methods to efficiently
combine gradients with different length scales. Residual stresses resulting
from elastic-plastic shakedown due to local yielding can also be calculated from
elastic stress distributions. Two specific applications are described briefly.
Application A: FCG in a Bivariant Residual Stress Field Created by Local Yielding Severe local yielding at a stress concentration in a beta-annealed Ti-6Al-4V cantilever bending coupon caused a pronounced bivariant RS to form, with the magnitude of the RS dependent on the maximum applied spectrum stress. ABAQUS elastic-plastic finite element analysis was used to calculate the applied and residual stress fields, and the crack plane (the plane with maximum cyclic normal stress) was identified analytically. A novel WF SIF solution for a corner crack at a chamfer with a fully bivariant stress distribution was developed and implemented in NASGRO to perform the FCG life calculations. A complex variable amplitude load spectrum was scaled to different maximum elastic stresses, and standard FCG properties (including load interaction model constants) were employed. Predicted FCG rates with and without RS effects were compared with actual FCG test data; no calibration was performed. The compressive RS fields were shown to cause substantial retardation of FCG rates, and the custom NASGRO predictions agreed closely with the test data.
Application B: FCG in a Relaxed Residual Stress Field Created by Surface Enhancement Surface enhancement methods such as shot peening (SP) or low plasticity burnishing (LPB) can introduce significant near-surface compressive RS fields, but these beneficial residual stresses can relax during thermal exposure or subsequent fatigue loading. FCG analysis can be used to predict the influence of the resulting stable RS fields on fatigue life. In this example, alpha-beta Ti-4Al-6V laboratory coupons were subjected to SP or LPB and then thermally exposed (425C/10 hrs) before RS profiles were measured. These RS profiles were inserted into a univariant WF surface crack SIF solution. Hypothesizing that the surface enhancement could have introduced microscopic damage that would initiate fatigue cracks quickly, FCG analyses with small initial crack sizes were used to calculate total fatigue life. Separate studies had shown that a simple El Haddad model could successfully correlate small-crack growth data with conventional long-crack growth data. The resulting fatigue life predictions were remarkably accurate. Small variations in the assumed initial crack size had relatively little impact on calculated life. Small shifts (±9 ksi) in the RS profiles, hypothetically arising from process variability or measurement uncertainty, had a somewhat larger impact on calculated life.
Measuring
Multiple Residual Stress Components using Multiple Cuts with the Contour Method
P. Pagliaro1, M. B. Prime2 and B. Zuccarello1.
1: Universitá degli Studi di Palermo, Palermo, Italy.
2: Los Alamos National Laboratory, Los Alamos, NM.
An extension of the contour method is proposed which allows the measurement
of multiple stress components by making multiple cuts. In the traditional
contour method, a specimen is carefully cut in two along a plane using wire
EDM. Residual stresses relax as free surface is created by the cut, and the
relaxed contours of the two opposing surfaces are measured. Then, a
straightforward Finite Element (FE) calculation, in which the opposite of
the measured contour is applied as displacement boundary conditions to the
cut surface in the FE model, reveals the original residual stresses normal
to the cut. In the proposed extension of the contour method, a second
(or more) transverse cut is made, and the contours of the new cut surfaces
are measured. Following the same procedure as above, it is possible to
obtain the stress normal to the second cut. These stresses, calculated on
the plane of the second cut, have been affected by the first cut. Then a
simple inverse reconstruction gives the original stresses on the plane of
the subsequent cuts, removing the effect of the first cut. In order to
validate the method, an experimental test on a quenched plate of HSLA steel
was carried out.
Repeatable Technique for Inducing Known Residual Stresses in Titanium Bars
Jahan Rasty, Texas Tech University, TX.
The objective of this research was to develop a repeatable technique for
inducing a known Residual Stress (RS) field in titanium bars. In addition,
it was sought to determine the accuracy of different RS measurement techniques
in predicting the magnitude and distribution of the RS field in titanium bars.
Above objectives were accomplished by plastic bending of titanium bars employing
a four-point bend fixture. The magnitude and distribution of the RS field in
bent bars was determined via elasto-plastic Finite Element Analysis (FEA) of the
four-point bending process. Bent titanium samples were then distributed to
various independent RS testing facilities, where the residual stress profile in
the samples was measured and reported back to us. The reported results were
generally in line with the residual stress profile obtained through FEA, in turn
verifying the repeatability of the method used for creation of samples with a
known residual stress profile. Further, the results indicate that all RS
measurement techniques involved in this round-robin study were capable of
measuring the residual stress profile accurately, despite some inconsistency
among labs employing the X-ray diffraction method.
Residual
Stress Measurement Using the Contour and the Sectioning Methods:
Application to MIG and FSW Joints
(POSTER)
V. Richter-Trummer, S. M. O. Tavares,
P. M. G. P. Moreira, P. M. S. T. de Castro.
IDMEC, Faculdade de Engenharia da Universidade do Porto, Porto, Portugal.
The longitudinal residual stress distribution in a 6082-T6 Al alloy MIG butt
welded thin plate was determined using the contour method. Although several
geometries havebeen successfully studied with the contour method before, the
authors are not aware of previous attempts to use this method in a long narrow
section as in the present case. Validation was done using experimental plate
surface measurements obtained using the established sectioning technique.
Additionally the results were compared with predictions obtained using the
CEGB-R6 procedure. The measured residual stress data was used in fatigue crack
growth simulations in order to compare the behavior of the analyzed plate with
and without residual stress. The detrimental effect of residual stresses on
fatigue life is shown. Finally the sectioning technique was also applied to a
friction stir welded (FSW) plate so that an approximate residual stress
distribution could be obtained. A qualitative comparison between the residual
stresses introduced by both welding methods was performed.