Ultrasonic Shot Peening - MTU Aero Engines
Ultrasonic Shot Peening - MTU Aero Engines
Ultrasonic Shot Peening - MTU Aero Engines
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
<strong>Ultrasonic</strong> <strong>Shot</strong> <strong>Peening</strong> (USP) on Ti-6Al-4V and Ti-6Al-2Sn-4Zr-6Mo<br />
<strong>Aero</strong> Engine Components<br />
Paper_ICSP11.doc<br />
Inga Stoll 1 , Dietmar Helm 1 , Holger Polanetzki 1 , Lothar Wagner 2<br />
1 <strong>MTU</strong> <strong>Aero</strong> <strong>Engines</strong> GmbH, Munich, Germany<br />
2 Department of Materials science and engineering, Technical University of Clausthal,<br />
Clausthal, Germany<br />
Abstract<br />
Bladed disks (Blisks) have been developed and applied in the compressor sections of modern<br />
aero-engines to reduce the component weight and improve efficiency. Overtime, the blisk<br />
design has been further optimized with respect to weight reduction resulting in thinner<br />
geometries in the web area of the disk. Consequently, precautions have to be taken into<br />
account to avoid stress induced distortion in thin walled areas during the peening process.<br />
<strong>MTU</strong> has therefore implemented ultrasonic shot peening as an alternative to the conventional<br />
shot peening process. For safety reasons detailed knowledge about the effect of this process on<br />
the properties of the blisk material, especially the fatigue behavior is a necessity.<br />
Most known is the conventional shot peening process, where the peening media is accelerated<br />
by pressurized air. A new method of this mechanical strengthening process is shot peening<br />
using bearing balls accelerated by an ultrasonic vibrating plate. The bearing balls are<br />
surrounded by a closed chamber that fits the geometry of the part to be peened. Only a very<br />
small amount of bearing ball peening media is needed. To understand the mechanics of this<br />
rarely researched process, extensive experiments were performed at <strong>MTU</strong>.<br />
The effect of <strong>Ultrasonic</strong> <strong>Shot</strong> <strong>Peening</strong> compared to conventional shot peening on the near<br />
surface area – surface roughness, residual stress - of Ti-6Al-4V (Ti64) and Ti-6Al-2Sn-4Zr-6Mo<br />
(Ti6246) was investigated using common methods. The resulting fatigue behavior including<br />
fracture characterization for both materials was studied in detail comparing both mechanical<br />
strengthening processes.<br />
Keywords<br />
<strong>Ultrasonic</strong> <strong>Shot</strong> <strong>Peening</strong>, Ti-6Al-4V, Ti-6Al-2Sn-4Zr-6Mo, residual stress, surface roughness,<br />
fatigue behavior, fracture characterization, BLISK<br />
Introduction<br />
In the compressor section of jet engines at operating temperatures
shot peening are able to shift the crack initiation from the surface to sub-surface regions which<br />
leads to significantly higher fatigue life of the component [4, 5]. <strong>Shot</strong> peening has been used for<br />
many years as the standard process in the final surface treatment of rotating and static engine<br />
components [6].<br />
For specific reasons (weight reduction, elimination of assembly costs) in modern aero-engines,<br />
blisks (bladed disks) made from Titanium alloys are commonly used in the compressor section.<br />
blisks are integral components of rotor blades and disks (see Figure 1).<br />
Paper_ICSP11.doc<br />
Figure 1. Picture of a BLISK part [<strong>MTU</strong> <strong>Aero</strong> Engine].<br />
These components are either produced by milling from the solid or by linear friction welding [7].<br />
For further weight and cost savings, the BLISK should be manufactured exhibiting thinner wall<br />
thickness in the disk area. During the conventional shot peening process of such a blisk<br />
distortion may occur due to the directionally focused beam of peening media and the reduced<br />
wall thickness of the part. <strong>MTU</strong> has therefore implemented ultrasonic shot peening as an<br />
alternative to the conventional shot peening process. In addition to some other benefits of<br />
ultrasonic shot peening, lower distortion of thin-walled parts are expected as a result of the<br />
randomly distributed peening media during the ultrasonic shot peening process.<br />
<strong>Ultrasonic</strong> <strong>Shot</strong> <strong>Peening</strong><br />
<strong>Ultrasonic</strong> waves are generated by means of<br />
a piezoelectric excited oscillator and<br />
transmitted by an acoustic amplifier on the<br />
sonotrode. The acceleration of the peening<br />
balls occurs by kinetic energy transfer during<br />
collision of the shot with the continuously<br />
vibrating sonotrode. In case of ultrasonic shot<br />
peening the peening media consists of<br />
perfectly spherical bearing balls having a<br />
diameter several times larger than that of<br />
conventional shot peening media. <strong>Peening</strong><br />
media and the component sealed off in the<br />
peening chamber. Due to interflection of<br />
peening media on the component, chamber<br />
walls and the sonotrode, randomly distributed<br />
collisions between the peening media and<br />
the part, e.g. the blisk can be realized.<br />
Figure 2 shows the operating principle of a<br />
ultrasonic shot peening facility.
Experimental Methods<br />
To measure the effects of shot peening using different peening processes on the near surface<br />
zone, two different experimental methods were employed: Roughness measurement and<br />
residual stress measurement by means of X-ray diffractometry.<br />
Fatigue tests were carried out using smooth LCF and HCF tests under axial loads well as LCF<br />
rotating bending tests for Ti6246.<br />
The surface of all samples were modified by conventional or ultrasonic shot peening with a low<br />
Almen intensity (0,10mmA) and a high Almen intensity (0,23mmA for Ti64 / 0,2mmA for Ti6246)<br />
for LCF tests. HCF tests representing the blade area of a blisk were conventionally shot peened<br />
using glass beads as peening media or ultrasonic shot peened using bearing balls with Almen<br />
intensities of 0,18N or 0,27N. The coverage was 100% in all cases.<br />
Experimental Results<br />
Roughness<br />
Comparing the surface roughnesses of ultrasonic shot peened and conventionally shot peened<br />
specimens peened with the same Almen intensity and coverage, it is obvious that conventional<br />
shot peening results in surface roughness values much higher than ultrasonic shot peening.<br />
This behavior is observed on both Ti64 (Figure 3) and Ti6246 (Figure 4). For Ti64, the absolute<br />
value of surface roughness is higher.<br />
Figure 3. Roughness measurements Ti64. Figure 4. Roughness measurements Ti6246.<br />
[µm]<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
conventional 0,1A<br />
Ra [µm]<br />
Rmax [µm]<br />
Rz [µm]<br />
USP 0,1A<br />
Paper_ICSP11.doc<br />
Ti-64<br />
conventional 0,23A<br />
USP 0,23A<br />
[µm]<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
conventional 0,1A<br />
Ra [µm]<br />
Rmax [µm]<br />
Rz [µm]<br />
USP 0,1A<br />
Ti-6246<br />
conventional 0,2A<br />
Residual stress measurements<br />
For both alloys, the resulting compressive residual stresses at the surfaces for conventional<br />
shot peened specimens are higher than for ultrasonic shot peened ones. For a given alloy, both<br />
methods result in a similar location of the subsurface compressive residual stress maximum as<br />
well as comparable penetration depths. For Ti6246, the penetration depth is lower than for Ti64<br />
as shown in Figure 5 and Figure 6<br />
Figure 5. Residual stress measurements Ti64.<br />
residual stress [MPa]<br />
200<br />
0<br />
-200<br />
-400<br />
-600<br />
-800<br />
-1000<br />
Ti64<br />
conventional peened, 0,23mmA<br />
USP, 0,23mmA<br />
0 20 40 60 80 100 120 140 160 180 200 220<br />
distance from surface [µm]<br />
USP 0,2A<br />
Figure 6. Residual stress measurements.<br />
residual stress [MPa]<br />
200<br />
0<br />
-200<br />
-400<br />
-600<br />
-800<br />
-1000<br />
Ti6246<br />
conventional peened,<br />
0,2mmA<br />
USP, 0,2mmA<br />
0 20 40 60 80 100 120 140 160 180 200 220<br />
distance from surface [µm]
LCF Fatigue Life<br />
For Ti64, ultrasonic shot peening and<br />
conventional shot peening lead to higher LCF<br />
life compared to the unpeened condition<br />
indicated by the fatigue results at 0.9% strain<br />
range. The peening benefit of ultrasonic shot<br />
peened specimens is lower than for<br />
conventional shot peened specimens if both<br />
are tested at high strain ranges. For lower<br />
strain ranges, the resulting LCF life is<br />
comparable (see Figure 7). SEM<br />
investigations show surface failures for the<br />
unpeend (Figure 8) and the ultrasonic shot<br />
peened specimens at high strain ranges<br />
(Figure 9). At lower strain ranges,<br />
conventional shot peening (Figure10) and<br />
Paper_ICSP11.doc<br />
strain range [%]<br />
1,1<br />
1<br />
0,9<br />
0,8<br />
0,7<br />
0,6<br />
Figure 7. LCF test results Ti64.<br />
Ti64 axial LCF tests at RT, R=0<br />
<strong>Ultrasonic</strong> shot peened 0,23 mmA<br />
conventionell 100% 0,23 mmA<br />
unpeened<br />
0,5<br />
1,E+04 1,E+05 1,E+06 1,E+07<br />
cycles to failure<br />
ultrasonic shot peening (Figure 11) result in a shift of the crack initiation site from the surface to<br />
subsurface regions.<br />
R-16746 0609A03084<br />
Figure 8:<br />
unpeened<br />
Figure 9: USP<br />
(high strain rate)<br />
R-16488 0605A01309<br />
For Ti6246, conventional shot peening leads<br />
to LCF lives under rotating bending<br />
significantly higher than that of the unpeened<br />
condition. <strong>Ultrasonic</strong> shot peening still gives a<br />
life benefit compared to the unpeened<br />
condition but exhibits significantly lower LCF<br />
lives than conventional shot peening (Figure<br />
12). Crack intitiation sites at the surface were<br />
detected by SEM investigations for all<br />
unpeened specimens (Figure 13). For<br />
ultrasonic shot peening, crack inititation sites<br />
were observed at both locations, surface and<br />
subsurface (Figure 14, 15). Conventional shot<br />
Figure 10: conv.<br />
shot peened<br />
R-16383-16385 0603A02261<br />
Stress range [MPa]<br />
Figure 11: USP<br />
(low strain rate)<br />
R-16552 0606A01201<br />
Figure 12. LCF rotating bending at RT<br />
Test results Ti6246.<br />
900<br />
800<br />
700<br />
Ti6246 - rotating beending tests at RT, R=-1<br />
unpeened<br />
conventionell peened; 100%; 0.2A<br />
Ulrasonic shot peened; 100%; 0.2A<br />
600<br />
1,E+03 1,E+04 1,E+05 1,E+06 1,E+07 1,E+08<br />
cycles to failure<br />
peening (Figure 16) generally shifted the crack initiation to subsurface regions.
R-18114 0901A00907<br />
Figure 13:<br />
unpeened<br />
Paper_ICSP11.doc<br />
Figure 14: USP<br />
surface failure<br />
R-18288 0904A00636<br />
Figure 15: USP<br />
subsurface failure<br />
R-18288 0904A00646<br />
Figure 16: conv. shot<br />
peened<br />
R-18244 0903A02651<br />
HCF Fatigue Life – Comparison of glass peening and ultrasonic shot peening<br />
<strong>Shot</strong> peening of Ti64 gives similar HCF life for Figure 17. HCF Test results Ti64.<br />
conventional glass peening and <strong>Ultrasonic</strong><br />
RT; 600MPa, R=0<br />
shot peening. In case of <strong>Ultrasonic</strong> shot<br />
<strong>Ultrasonic</strong> shot peened 0,27N<br />
peening with a high Almen intensity value, the<br />
<strong>Ultrasonic</strong> shot peened 0,18N<br />
conventional glass peened 0.27N<br />
scatter can be reduced.<br />
conventional glass peened 0,18N<br />
Compared to the unpeened surface condition,<br />
unpeened<br />
all tested peening processes and Almen<br />
intensities show an increase of the HCF life<br />
about one magnitude.<br />
SEM investigations indicate surface failures<br />
for the unpeened condition and subsurface<br />
failures for all peened specimens (Figure 18,<br />
19, 20).<br />
R-16691 0608A02861<br />
Figure 18:<br />
unpeened<br />
peening parameter<br />
Figure 19: conv.<br />
glass peened<br />
R-16383-16385 0603A02279<br />
<strong>Shot</strong> peening of Ti6246 also leads to similar<br />
resulting HCF life for conventional glass<br />
peening and <strong>Ultrasonic</strong> shot peening.<br />
Compared to the unpeened surface condition<br />
specimens peened with the higher Almen<br />
intensity show a slightly improved HCF life<br />
whereas peening with low Almen intensity<br />
leads to HCF life which is more or less<br />
comparable to unpeened surfaces (Figure<br />
21).<br />
SEM investigations indicate surface failures<br />
for the unpeened condition and subsurface<br />
failures for all peened specimens (Figure 22,<br />
23, 24).<br />
peening parameter<br />
1,E+04 1,E+05 1,E+06 1,E+07<br />
cycles<br />
Figure 20: <strong>Ultrasonic</strong><br />
shot peened<br />
R-16525 0605A02960<br />
Figure 21. HCF Test results Ti6246.<br />
<strong>Ultrasonic</strong> shot peened 0.27N<br />
<strong>Ultrasonic</strong> shot peened 0.18N<br />
conventionell glass 0.27N<br />
conventionall glass 0.18N<br />
unpeened<br />
RT, 800MPa, R=0<br />
1,E+05 1,E+06 1,E+07<br />
cycles to failure
R-17907 0810A00057<br />
Paper_ICSP11.doc<br />
Figure 22:<br />
unpeened<br />
Figure 23: conv.<br />
glass peened<br />
R-18069 0812A01165<br />
Figure 24: <strong>Ultrasonic</strong><br />
shot peened<br />
R-18069 0812A01171<br />
Discussion and Conclusions<br />
Conventional shot peening leads to higher surface roughness and higher resulting compressive<br />
residual stresses at the surface than <strong>Ultrasonic</strong> shot peening using same Almen intensity. This<br />
characteristic can be shown for Ti64 as well as for Ti6246. The location of the compressive<br />
residual stress maximum under the surface as well as the penetration depth is comparable for<br />
both methods.<br />
The lower surface roughness and penetration depth for Ti6246 compared to Ti64 can be<br />
explained by the lower deformation resistance of Ti64.<br />
The resulting LCF life of peened surfaces depends on the pening process. Conventional shot<br />
peening leads to higher LCF life than for the unpeened condition of the same alloy for all tested<br />
almen intensities. SEM investigation show that conventional shot peening is able to shift the<br />
crack initiation from the surface to the subsurface region. Therefore crack initiation occurs under<br />
vacuum giving a substantial contribution to life improvement. Additionally, life extension due to<br />
crack growth under vacuum must be considered. For ultrasonic shot peened Ti64 specimens<br />
this behavior takes place only at low strain rates, while at higher strain rates crack initiation<br />
occurs again at the surface results in lower LCF life. Ti6246 specimens with ultrasonic shot<br />
peened surface condition exhibits lower LCF life compared to specimens with conventional<br />
peened surface.<br />
The resulting HCF life for peened specimens is comparable for both investigated peening<br />
processes. The life benefit of peened Ti64 specimens is nearly one magnitude. Ti6246<br />
specimens hardly show a life improvement for both shot peening methods. This difference to<br />
Ti64 may be explained by less sensitivity of Ti6246 to crack initiation at laboratory air.<br />
References<br />
[1] Boyer, R. R., An Overview on the Use of Titanium in the <strong>Aero</strong>space Industry, Materials<br />
Science and Engineering A; 213, 1996, p. 103-114.<br />
[2] Williams, J. C., Titanium Alloys: Production, Behaviour and Application, High Performance<br />
Materials in <strong>Aero</strong>space. London: Chapman & Hall, 1995, p. 85-134<br />
[3] König, G. Life Enhancement of <strong>Aero</strong> Engine Components by <strong>Shot</strong> <strong>Peening</strong>: Opportunities<br />
and Risks, Wagner, L. (Ed.) <strong>Shot</strong> <strong>Peening</strong>. Weinheim: Wiley-VCH 2003, p. 13-22.<br />
[4] Wagner, L. Der Einfluss von Kugelstrahlen auf das Ermüdungsverhalten hochfester<br />
Titanlegierungen. Bochum, Ruhr-Universität Bochum, Dissertation, 1981.<br />
[5] Wagner, L. Mechanical Surface Treatments on Titanium Alloys: Fundamental Mechanisms,<br />
1996 TMS meeting Symposium on Surface Performance of Titanium, TMS, 1997, p. 199-<br />
215<br />
[6] Carek, G.A <strong>Shot</strong> <strong>Peening</strong> for Ti-6Al-4V Alloy Compressor Blades, NASA Technical Paper<br />
2711, 1987, p. 1-7<br />
[7] Wilhelm, H.; Schneefeld, D.; Helm, D. Lineares Reibschweißen von Verdichterläufern aus<br />
Titan-Legierungen, Moderne Werkstoffe im Luft- und Raumfahrzeug. Düsseldorf: DVS<br />
Verlag, 2000, p. 42-47.
ISCSP agrees to publish:<br />
Paper_ICSP11.doc<br />
International Scientific Committee for <strong>Shot</strong> <strong>Peening</strong><br />
Manuscript Title: __________________________________________________<br />
Names/affiliation of all Authors:<br />
__________________________ ______________________________<br />
__________________________ ______________________________<br />
__________________________ ______________________________<br />
__________________________ ______________________________<br />
in the conference proceedings of the 11 th International Conference on <strong>Shot</strong> <strong>Peening</strong> (ICSP-11)<br />
and its affiliated web site http://www.shotpeening.org.<br />
On acceptance for publication the undersigned author(s) (“Author”) of the Article assigns<br />
exclusively to the International Scientific Committee on <strong>Shot</strong> <strong>Peening</strong>, ISCSP, worldwide<br />
copyright in the Article for all media and formats in all material published as part of the Article,<br />
which expression includes but is not limited to the text, abstract, tables, figures and graphs, but<br />
excludes any supplementary material.<br />
If any of the Named Authors are Government employees, on acceptance for publication the<br />
Author shall grant ISCSP a royalty free exclusive license for copyright for all media and formats.<br />
In signing this Agreement the Authors represent and warrants that the Article is the original work<br />
of the Named Authors, that it has not been published previously in any form (other than in a<br />
research thesis or dissertation which fact has been notified to ISCSP in writing), that all Named<br />
Authors have participated sufficiently in the conception and writing of the Article, have received<br />
the final version of the Article, agree to its submission and take responsibility for it, and<br />
submission has been approved as necessary by the authorities at the establishment where the<br />
research was carried out.<br />
The principal Author warrants that he/she signs this Agreement as authorized agent for and on<br />
behalf of all Named Authors and has the full power to enter into this Agreement and to make the<br />
grants it contains, that the Article has not been and will not be submitted to another publisher,<br />
that it does not infringe any third party rights, that it contains nothing libelous or unlawful, that all<br />
factual statements are to the best of the Author’s knowledge true or based on valid research<br />
conducted according to accepted norms, and that all required permissions have been obtained<br />
in writing.<br />
______________________________ ___________________<br />
Author’s signature Date