Development of a three-axis hybrid mesh isolator using the pseudoelasticity of a shape

memory alloy

This article has been downloaded from IOPscience. Please scroll down to see the full text article.

2011 Smart Mater. Struct. 20 075017

(http://iopscience.iop.org/0964-1726/20/7/075017)

Download details:

IP Address: 143.248.69.144

The article was downloaded on 23/06/2011 at 01:35

Please note that terms and conditions apply.

View the table of contents for this issue, or go to the journal homepage for more

Home Search Collections Journals About Contact us My IOPscience

IOP PUBLISHING SMART MATERIALS AND STRUCTURES

Smart Mater. Struct. 20 (2011) 075017 (12pp) doi:10.1088/0964-1726/20/7/075017

Development of a three-axis hybrid meshisolator using the pseudoelasticity of ashape memory alloySe-Hyun Youn1, Young-Soon Jang1 and Jae-Hung Han2,3

1 Structures and Materials Department, Korea Aerospace Research Institute, 45 Eoeun-dong,Yuseoung-gu, Daejeon 305-333, Korea2 Department of Aerospace Engineering, Korea Advanced Institute of Science andTechnology, 291 Daehak-ro, Yuseong-gu, Daejeon 305-701, Korea

E-mail: jaehunghan@kaist.ac.kr

Received 2 December 2010Published 22 June 2011Online at stacks.iop.org/SMS/20/075017

AbstractLaunch vehicles and satellites experience severe vibration and pyroshock loads during flightphases. In particular, intense pyroshock, which is generated by the actuation of separationdevices, can cause malfunctions in the electronic components in launch vehicles and satellites,potentially resulting in catastrophic failure during flight. This paper introduces a new three-axishybrid mesh isolator using the pseudoelasticity of a shape memory alloy wire that wasmanufactured and tested to attenuate pyroshock and vibration transmitted to the electroniccomponents. To characterize the isolation capability, quasi-static loading tests were performed;the test results showed that the pseudoelastic effect of the shape memory alloy wire significantlyabsorbs energy due to the stress-induced phase transformation. The ground pyroshock testresults showed a remarkable pyroshock load attenuation of the hybrid mesh isolator in allfrequency ranges. The dynamic characteristics and vibration isolation performances of the meshisolators were also verified by random vibration tests. The healthiness of the hybrid meshisolator was also studied under a harsh vibration loading level, and the results confirmed itswide applicability without degradation of the isolation capability.

(Some figures in this article are in colour only in the electronic version)

1. Introduction

Generally, launch vehicles and satellites experience severevibration loadings during flight phases such as lift-off, transientflight, maximum dynamic pressure conditions and main enginecutoff. Moreover, transient structural vibrations, knownas pyroshocks, are also generated by the actuation of theordnance devices which are used for several separation eventsincluding stage separation, fairing separation, and satelliteseparation. Pyroshock is typically characterized as a shortduration transient response with a high amplitude effective upto a very high frequency range.

These severe vibration loads as well as high magnitude,high frequency pyroshocks can cause malfunctions in theelectronic components of a launch vehicle or satellite, resulting

3 Author to whom any correspondence should be addressed.

in catastrophic flight failure [1]. In order to attenuate theseshock and vibration loads, rubber-based isolators are widelyemployed in the aerospace industry. However, the applicationof these rubber-based isolators can amplify low frequencyvibrations due to the low stiffness of the isolating material,and severe vibration loads can cause breakage of the isolatoritself [2, 3].

The hybrid mesh isolator presented in this paper canprovide a solution to these problems because it has higherstiffness than the elastomeric materials and offers largerhysteresis loops, providing excellent isolation performance.

Shape memory alloys (SMAs) have two special wellknown effects: the shape memory effect and pseudoelasticity(or superelasticity). These special effects originate from thephase transformation between martensite and austenite. Theshape memory effect describes the restoration of the original

0964-1726/11/075017+12$33.00 © 2011 IOP Publishing Ltd Printed in the UK & the USA1

Smart Mater. Struct. 20 (2011) 075017 S-H Youn et al

shape of a deformed material by heating, which results fromthe phase transformation from martensite to austenite. Thepseudoelastic behavior is a material effect in which the originalshape restores itself upon unloading after a large deformation.Pseudoelastic SMA alloys can be considerably deformedwithout being plastically deformed. This pseudoelasticity ofSMA is caused by a stress-induced phase transformation [4, 5].

The load–displacement relationship of a pseudoelasticSMA during the loading–unloading cycle appears as anonlinear hysteresis curve, which results from energyabsorption in the phase transformation from austenite tomartensite and vice versa [6, 7]. The larger the hysteresisloop area, the more energy is dissipated in one cycle. Thepseudoelasticity of SMAs has been applied to various passivevibration isolation studies [7–11]. However, most previousworks on the passive use of SMAs have been confined to lowfrequency vibration attenuation and have not been extendedto high frequency pyroshock. The present authors proposed anew SMA mesh washer isolator, the excellent shock isolationperformance of which was experimentally demonstrated [12].However, usage of a washer type isolator in a real assemblyprocess is somewhat limited because several washers arerequired for each fastening point; thus, an integrated isolatoris desirable. Another issue is that a washer type isolator iseffective only for the axial direction.

This paper introduces a new type of three-axis shockisolator called a ‘hybrid mesh isolator’, made out of 50% metalwire and 50% pseudoelastic SMA wire. The damping of thegeneral mesh isolator was primarily obtained from the frictionof the constituting wires as well as from the inherent dampingof the wire. Moreover, the pseudoelastic behavior of the SMAwire dramatically enhanced the damping capacity of hybridmesh isolators.

The present hybrid mesh isolator shows a much moreefficient isolation performance than common rubber-basedisolators due to the additional energy absorption that resultsfrom the large hysteresis of the pseudoelastic behavior.

In this paper, two types of mesh isolator (MI) arepresented: one is made out of 50% pseudoelastic alloy wire(SE508) and 50% stainless steel wire (STS310S), referred toin this paper as the hybrid-MI, and the other isolator is madeout of full stainless steel wire, called the STS-MI.

To identify the isolation capacity of each MI, quasi-staticloading tests were performed and the isolation capabilitiesof the MIs were compared in terms of the dissipated energyand specific damping capacity. Ground pyroshock tests wereperformed using a shock test machine with a 4 kg dummymass and the isolation performance was characterized inboth time and frequency domains. To verify the dynamiccharacteristics and vibration isolation performances of eachMI, random vibration tests were also performed. In addition,the healthiness of the hybrid mesh isolator was studied under aharsh vibration loading level and the results confirmed its wideapplicability without degradation of the isolation capability.During these tests, the isolation performances of each MI and acommercial rubber-based isolator, called AM007-13, were alsocompared.

2. Hybrid mesh isolator

2.1. Thermomechanical characteristics of SMA wire

To ensure better isolation performance, the SMA should havedefinite pseudoelastic effects at room temperature. Therefore,SE508 pseudoelastic SMA wire produced by Nitinol Devices& Components (NDC) was selected to manufacture the hybrid-MI. The datasheet from NDC reports that the austenite finishtemperature (Af) value of SE508 is in the range of 5–18 ◦C [13].

To study the thermomechanical characteristics of the SMAwires, a universal testing machine (INSTRON® UTM5583)with a thermal chamber was used to investigate the stress–strain relationship. The test temperatures of the thermalchamber can be set from −30 to 100 ◦C. At each fixedtemperature, a tension load is applied until 6% strain and isthen unloaded to the initial state. After that, the tension load isreapplied immediately until the wire fractures.

The tensional stress–strain test results of SE508 SMAwires in the temperature range from −30 to 30 ◦C are shownin figure 1. When the tension loading was applied, a plateauregion appeared due to the stress-induced phase transformationfrom austenite to martensite, and another low plateau regionwas seen during the unloading stage due to the phasetransformation from martensite to austenite in temperatureconditions of over −15 ◦C. This reveals a hysteretic stress–strain relationship, and the enclosed area indicates the energyabsorbed by the pseudoelastic SMA wire during the phasetransform from austenite to martensite. Table 1 compares thebasic physical properties of the SE508 and STS310S wiresused in the manufacture of the mesh isolators [12].

2.2. Manufacture of the MI

Two types of MIs, hybrid-MI and STS-MI, were manufacturedin this study and the manufacturing procedures for the generalMIs are introduced in figures 2 and 3.

The density and shape of the MI products were determinedby the size of the knitted mesh element and the design of thepressing mold. The manufacturing method for hybrid-MI isalmost the same as that of STS-MI except that the two wires(SE508 and STS310S) were knitted simultaneously.

The MI was designed to suppress multi-direction externalloads, and the manufactured and designed shapes of the MIare shown in figure 4. Figure 5 presents a detailed half-cuttingview that shows the wires are woven in a very complicatedmanner with many spaces between the wires.

The physical characteristics of each MI are also presentedin table 2. In this table, the volume fraction indicates theratio of wire volume to the total outside volume. For afair comparison of the isolation performance of each MI, thevolume fraction was set to the same value (0.38) in this study.

In addition, an AM007-13 rubber-based isolator manufac-tured by Lord Corporation, shown in figure 6, was also testedto compare the isolation performance; the datasheet states thatthe dynamic spring rate of AM007-13 is set to 426 N mm−1

in both the axial and lateral directions [14]. This isolatorhas been widely used to suppress the shock and vibration

2

Smart Mater. Struct. 20 (2011) 075017 S-H Youn et al

Figure 1. Stress–strain relationship of the SE508 pseudoelastic wire with temperature variation.

Knitting1Manufacturing wire mesh using knitting

machine with a continuous wire

2

3

Cut & Fold overWire mesh is cut and folded over

before inserting to the press machine

Insert

Pressing

Insert folded wire mesh to the designed press mold

Pressing folded wire mesh with a designed pressure

Inspection

Figure 2. Manufacturing diagram of the mesh isolator.

loads transmitted to the electronic components in many launchprograms.

3. Isolating capacity of mesh isolators underquasi-static loading

3.1. Quasi-static test setup

To apply a mesh isolator between electronic components (ordummy mass) and equipment structure (or test bay), the MIassembly was constructed as shown in figure 7. The upperfixture and upper washer were used to connect the MI with theelectronic component, and the lower fixture and lower coverwere used to fix the MI to the equipment structure. Throughthe loading path from the equipment bay to the electronic

Table 1. Mechanical properties of the wires.

Material

Wirediameter(mm)

Density(g cm−3)

Tensilestrength(MPa)

Modulus ofelasticity(GPa)

Strain(%)

SE508 0.2 6.5 143042 (austenite)

14.418(martensite)STS310S 0.2 7.8 650 199 40

Table 2. Physical characteristics of mesh isolators.

Material

Volumepercentage(%)

Mass(g)

Total mass(g)

Volumefraction

STS-MI STS310S 100 34.7 34.7 0.38

Hybrid-MISE508 50 14.5

31.9 0.38STS310S 50 17.4

component, the shock and vibrations could be attenuated bythe MI.

Quasi-static loading tests for the MI assemblies inthe axial and lateral directions were performed using theINSTRON® MTS810 test machine as shown in figure 8. Twotypes of MI and a commercial isolator, AM007-13, were usedfor the tests and a quasi-static load of 0.3 Hz was applied usingthe displacement control method.

3.2. Quasi-static test results

Figure 9 presents and compares the quasi-static loadingtest results in the axial and lateral directions. Theload–displacement relationships of the STS-MI show thatsome progressive plastic deformation occurred in the highdisplacement loading condition; this resulted from the gradualplastic deformation inside the stainless steel wire with theincreased loading. However, these tendencies are barelyvisible in the hybrid-MI because the SE508 pseudoelasticwire dissipates much of its external energy, and the rubber

3

Smart Mater. Struct. 20 (2011) 075017 S-H Youn et al

insert

Folded Mesh

Figure 3. Manufacturing processes of the mesh isolators.

material of the commercial isolator did not show local plasticdeformations either.

Figure 10 compares the quasi-static test results of theseisolators in both directions with a ±0.5 kN peak loading.

From these test results, the internal dissipated energy andspecific damping capacity (SDC, �) were calculated and arecompared in table 3. The internal dissipated energy can beobtained by calculating the hysteretic looped area, and the

4

Smart Mater. Struct. 20 (2011) 075017 S-H Youn et al

Figure 4. Manufactured and designed shape of the mesh isolator.

Figure 5. Half-cutting view of the mesh isolator.

Insert

AM007-13 conventional IsolatorInsert

Figure 6. A conventional elastomeric isolator (AM007-13).

Mesh Isolator

Lower Fixture

Lower Cover

Lower Cover

bolting with text object

bolting with bay

Test object

Upeer Fixture

Upper Washer

Mesh Isolator

Lower Fixture

Test Bay

Figure 7. Assembly of the mesh isolator used in the tests.

specific damping capacity is defined as the ratio of internaldissipated energy to external total energy [15]. Under thesame loading conditions, the hybrid-MI shows outstandingisolation capability due to the characteristics of the meshisolator itself, as well as the pseudoelasticity of the SE508

SMA wire compared with other isolators. This enhanceddamping ability is obtained by substituting 50% of the stainlesssteel wire with the pseudoelastic SMA wire; therefore, it isbelieved that the pseudoelasticity of the SMA wire greatlyimproves the isolation capacity of the hybrid-MI.

5

Smart Mater. Struct. 20 (2011) 075017 S-H Youn et al

Lateral Direction Axial Direction

Figure 8. Quasi-static loading test of the mesh isolator in both directions.

Table 3. Comparison of the dissipated energy and specific dampingcapacity of each isolator.

Type of WI

Hybrid STS-II AM007

Dissipatedenergy (N m)

Axial 0.43 0.10 0.43Lateral 0.98 0.09 0.42

SDC �Axial 0.78 0.49 0.62Lateral 0.83 0.54 0.64

4. Ground pyroshock isolation tests

4.1. Pyroshock test setup

To verify the pyroshock isolation performance of each isolator,ground pyroshock tests were performed using a shock testmachine that generates a pyroshock by impacting a resonantplate using a coil spring and an impact hammer as shown infigure 11 [12, 16].

In this paper, a 4 kg dummy mass that imitates anelectronic component and four MI assemblies were used in thepyroshock tests. AM007-13 was also tested for comparison.To measure the applied pyroshock and attenuated accelerationresults, two PCB 350B03 shock accelerometers and a PULSEdata acquisition system were used, and the locations of theaccelerometers are shown in figure 12. Channel 1 was locatedon the shock test fixture plate and was used as the referencesignal to verify the loading conditions. Channel 2 was locatedon the dummy mass to measure the results attenuated by theisolators.

The characteristics of the damage potential of a pyroshockin the frequency domain are compared using the shockresponse spectrum (SRS) [17].

4.2. Pyroshock isolation test results

The pyroshock isolation tests were performed for two typesof MI and the AM007-13 isolator using the ground shock

Table 4. Comparison of the peak acceleration results for eachisolator.

Type of WI

Hard mount Hybrid STS-II AM007

Axial peakacceleration (g)

Max. 2713 87 200 162Min. −1552 −20 −57 −70

Lateral peakacceleration (g)

Max. 506 34 78 60Min. −343 −11 −32 −26

test machine. For a precise comparison of the results, a non-isolated test condition, called a ‘hard mount’, was verified inadvance.

The attenuated shock acceleration results in the timedomain of Channel 2 are shown in figure 13 in the axialand lateral directions. In addition, the peak accelerationcomparison results are summarized in table 4. For bothloading directions, the hybrid-MI showed the best isolationperformance and peak reduction results as expected fromthe quasi-static test results. The AM007-13 isolator alsoshowed a better isolation capability compared with that of theSTS-MI.

The calculated SRS results for each isolator are presentedin figure 14 and these results demonstrate again that the hybrid-MI showed outstanding isolation performance in all frequencyranges and the AM007-13 isolator demonstrated the next bestresults.

These outstanding isolation abilities in all frequencyranges can be attributed to 50% of pseudoelastic SMA wireshaving the stress-induced phase transformation.

5. Random vibration tests

5.1. Random vibration test setup

Random vibration tests were performed using an LDS V8shaker to study the dynamic characteristics and vibrationisolation performances for the isolators. Two PCB 353B03

6

Smart Mater. Struct. 20 (2011) 075017 S-H Youn et al

Figure 9. Comparison of the quasi-static test results in both directions.

accelerometers were used for the measurement. The locationof each accelerometer is presented in figure 15: Channel 1 wasused for the shaker control and reference signal and Channel2 was located on the dummy mass to measure the attenuatedacceleration signal. The vibration isolation performance

was evaluated by comparing accelerations in Channel 1 andChannel 2. The random vibration level was 12grms overall witha uniform intensity between 20 and 2000 Hz.

To verify the healthiness of each isolator according tothe vibration level, the frequency variation under the 1grms

7

Smart Mater. Struct. 20 (2011) 075017 S-H Youn et al

Figure 10. Comparison of the quasi-static test results under the sameloading condition.

vibration level condition was reconfirmed after applyingvibration loadings of 6, 12 and 20grms for 3 min.

5.2. Random vibration test results

Table 5 presents the dynamic characteristics of the test systemswith the isolators in both directions. The transmissibility ofeach isolator was calculated using the Channel 1 and Channel

Test object

Impact Hammer

Coil spring

Linear Drive System

Resonant Plate

Figure 11. Ground pyroshock test machine.

2 signals; this is presented in figure 16. Both the hybrid-MIand AM007-13 show good vibration isolation performancescompared with the STS-MI and these similar results wereconfirmed by the time domain acceleration data shown infigure 17. As in the previous tests, the superior performanceof the hybrid-MI is due to its pseudoelastic effects; thus, thehybrid-MI absorbs more vibration energy than the STS-MI.

The healthiness of each isolator with the increments in thevibration loading levels was evaluated by the variation of thenatural frequencies under the 1grms vibration level conditionafter experiencing vibration loads of 6, 12 and 20grms levelsfor 3 min (see tables 6 and 7). In the hybrid-MI and AM007-13 test cases, the natural frequencies under a 1grms vibrationlevel changed within 3–5% after imposing 20grms vibrationloads compared with that of the initial 1grms vibration level.However, for the STS-MI, a frequency change of as muchas 17% was observed. This result reveals that the STS-MI suffered from plastic deformation and that the isolationperformances are degraded in the severe vibration loadingcondition. However, the hybrid-MI and AM007-13 isolatorsdid not experience such degradations, showing that the hybrid-MI is widely applicable without degradation in its isolationcapability.

6. Conclusion

Launch vehicles and satellites experience severe vibrationand pyroshock loads during flight phases, and these severeloads can cause malfunctions in their electronic components.

Axial Ground Pyroshock Test Lateral

Ch.1 (reference signal) 4Kg Dummy mas4Kg Dummy mas

Ch.2 (isolated signal)Ch.2 (isolated signal)

Figure 12. Locations of shock accelerometers.

8

Smart Mater. Struct. 20 (2011) 075017 S-H Youn et al

Figure 13. Comparison of the pyroshock test results in the time domain.

Table 5. Dynamic characteristics of each isolation system (overall 12grms).

Hybrid STS-II AM007

Direction Freq. (Hz) grms TR Freq. (Hz) grms TR Freq. (Hz) grms TR

Axial 207.50 9.06 2.86 226.25 10.35 3.74 172.50 7.83 2.67Lateral 196.25 7.30 2.32 316.25 13.21 3.89 190.00 8.63 2.86

9

Smart Mater. Struct. 20 (2011) 075017 S-H Youn et al

Figure 14. Comparison of pyroshock test results in the frequencydomain (SRS).

To avoid these types of failures, a new type of isolator isproposed in this study, called a ‘hybrid mesh isolator’. Thehybrid mesh isolator is manufactured by mixing pseudoelasticSMA wire and general stainless steel wire in order tosuppress severe vibration and pyroshock loads. The goodisolation performance of this isolator results from the energydissipation characteristics of the mesh isolator itself and thepseudoelastic behavior of the SMA due to the stress-inducedphase transformation.

In this paper, two types of mesh isolator were prepared:one fabricated from 50% pseudoelastic alloy wire (SE508) and

Figure 16. Dynamic characteristics of each isolator (frequencydomain).

50% stainless steel wire (STS310S) and named the hybrid-MI;the other made from a full stainless steel wire and named STS-MI.

Quasi-static loading tests were performed to verify theisolation capacity of the hybrid mesh isolator and, throughthese tests, the dissipated energy and specific dampingcapacity were calculated. The test results of the hysteretic

Axial

4Kg Dummy Mass 4Kg Dummy Mass

Mesh IsolatorAssemblies

Random Vibration Test Lateral

Mesh IsolatorAssemblies

Figure 15. Locations of vibration accelerometers.

10

Smart Mater. Struct. 20 (2011) 075017 S-H Youn et al

Figure 17. Attenuation performances of each isolator (time domain).

load–displacement relationship showed that the hybrid meshisolator has a high capability for dissipating vibration andexternal energy. The ground pyroshock tests were performedusing a shock test machine with a 4 kg dummy mass; theisolation performance was verified in the time and frequencydomains. The test results showed that the hybrid meshisolator has remarkable shock attenuation in all frequency

ranges. Through random vibration tests, the dynamiccharacteristics and vibration isolation performances were alsostudied, including the verification of the health of the hybridmesh isolator under the harsh vibration loading levels.

The series of verification test results showed outstandingisolation performances and it is concluded that the proposedhybrid mesh isolator can be effectively employed to isolate

11

Smart Mater. Struct. 20 (2011) 075017 S-H Youn et al

Table 6. Comparison of the natural frequencies under the 1grms vibration level after increments in the random vibration loads (axial axis).

6grms 12grms 20grms

Isolator type Test level Initial 1grms Main 1grms Main 1grms Main 1grms

STS-IIFreq. (Hz) 456.25 297.50 428.75 226.25 406.25 192.50 378.75% 100 94 89 83

Hybrid-MIFreq. (Hz) 355.00 278.75 343.75 207.50 341.25 157.50 340.00% 100 97 96 96

AM007-13Freq. (Hz) 307.50 227.50 302.50 172.50 297.50 143.75 297.50% 100 98 97 97

Table 7. Comparison of the natural frequencies under the 1grms vibration level after increments in the random vibration loads (lateral axis).

6grms 12grms 20grms

Isolator type Test Initial 1grms Main 1grms Main 1grms Main 1grms

STS-MIFreq. (Hz) 497.50 403.75 487.50 316.25 480.00 237.50 442.50% 100 98 96 89

Hybrid-MIFreq. (Hz) 330.00 255.00 323.75 196.25 320.00 148.75 312.50% 100 98 97 95

AM007-13Freq. (Hz) 325.00 253.95 318.75 190.00 313.75 125.00 307.50% 100 98 97 95

severe pyroshock and vibration transmitted to the electroniccomponents in launch vehicles and satellites. Theseoutstanding isolation performances were obtained from thepseudoelastic characteristics of the applied SMA wire.

Acknowledgments

This research was supported by the NSL (National Space Lab)program through the National Research Foundation of Koreafunded by the Ministry of Education, Science and Technology(grant number 2009-0091934).

References

[1] Moening C J 1985 Pyrotechnic shock flight failures Proc.Aerosp. Test. Semin. 8 95–109

[2] Mercer C A and Rees P L 1971 An optimum shock isolatorJ. Sound Vib. 18 511–20

[3] Racca R H 1996 Theory of Shock Isolation in Shock andVibration Handbook 4th edn, ed C M Harris (New York:McGraw-Hill) chapter 31

[4] Otsuka K and Wayman C M 1998 Shape Memory Materials(Cambridge: Cambridge University Press)

[5] Duerig T W, Melton K N, Stockel D and Wayman C M 1990Engineering Aspects of Shape Memory Alloys (Butterworth:Heinemann)

[6] Machado L G and Lagoudas D C 2006 Nonlinear dynamics ofa SMA passive vibration damping device Proc. 13th SPIEInt. Symp. Smart Structures and Materials (San Diego, CA);Proc. SPIE 6169 61690X

[7] Lagoudas D C, Khan M M, Mayes J J and Henderson B K 2004Pseudoelastic SMA spring elements for passive vibrationisolation: part II—simulations and experimental correlationsJ. Intell. Mater. Syst. Struct. 15 443–70

[8] Yiu Y C and Regelbrugge M E 1995 Shape-memory alloyisolators for vibration suppression in space applications 36thAIAA/ASME/ASCE/AHS/ASC Structures, StructuralDynamics and Materials Conf. pp 3390–8AIAA-95-1120-CP

[9] Yan X J and Nie J X 2008 Study of a new application form ofshape memory alloy superelasticity Smart Mater. Struct.12 N14–23

[10] Roh J H, Lee I and Han J H 2008 Damping characteristics ofSMA films and their application for passive vibrationisolation Int. J. Appl. Electromag. 27 225–41

[11] Saadat S, Noori M and Davoodi H 2001 Using NiTi SMAtendons for vibration control of coastal structures SmartMater. Struct. 10 695–704

[12] Youn S H, Jang Y S and Han J H 2010 Compressed meshwasher isolators using the pseudoelasticity of SMA forpyroshock attenuation J. Intell. Mater. Syst. Struct.21 407–21

[13] http://www.nitinol.com[14] http://www.lord.com[15] Predki W and Kloenne M 2002 Damping of superelastic

NiTi-alloys under torsional loading Materials Science Forum394–395 115–8

[16] Barrett P, Cornford S L and Larson C N 1997 Pyroshocktesting using large resonant plates and srs theory using timeand frequency domain method Proc. Aerosp. Test. Semin. 17101–10

[17] Mulville D R 1999 Pyroshock Test Criteria NASA TechnicalStandard NASA-STD-7003

12