I SandiaNationallaboratories

al Isolatorneter, the

Issued by Sandia National Laboratories, operated for the United StatesDepartment of Ener~ by Sandia Corporation.

NOTICE: This report was prepared as an account of work sponsored by anagency of the United States Government. Neither the United States Government,nor any agency thereof, nor any of their employees, nor any of their contractors,subcontractors, or their employees, make any warran~, express or implied, orassume any legal liability or responsibility for the accuracy, completeness, orusefulness of any information, apparatus, product, or process disclosed, orrepresent that its use would not infringe privately owned rights. Reference hereinto any specific commercial product, process, or service by trade name,trademark, manufacturer, or otherwise, does not necessarily constitute or implyits endorsement, recommendation, or favoring by the United States Government,any agency thereof, or any of their contractors or subcontractors. The views andopinions expressed herein do not necessarily state or reflect those of the United

States Government, any agency thereof, or any of their contractors.

Printed in the United States of America. This report has been reproduced directlyfrom the best available copy.

Available to DOE and DOE contractors fromU.S. Department of EnergyOffice of Scientific and Technical InformationP.O. BOX 62Oak Ridge, TN 37831

Telephone: (865)576-840 1Facsimile: (865)576-5728E-Mail: reports@adonis. osti.gov

Online ordering http:/ /www.doe.gov/bridge

Available to the public fromU.S. Department of CommerceNational Technical Information Service5285 Port Royal RdSpringfield, VA 22161

Telephone: (800)553-6847 .

Facsimile: (703)605-6900E-Mail: orders@,ntis.fedworld. gov

Online order http:/ /www.ntis.gov/ordering. htm

.

.

,

.

DISCLAIMER

Portions of this document may be illegiblein electronic image products. Images areproduced from the best available originaldocument.

SAND 2000-1528Unlimited ReleasePrinted July 2000

.

High Shock, High Frequency Characteristics of aMechanical Isolator for a Piezoresistive Accelerometer,

the ENDEVCO 7270AM6*

Vesta I. Bateman, Fred A. Brown, and Michael A. NusserEngineering Sciences CenterSandia National Laboratories

l?. O. BOX 5800Albuquerque, NM 87185-0553

Abstract

A mechanical isolator has been developed for a piezoresistive accelerometer.The purpose of the isolator is to mitigate high frequency shocks before they reachthe accelerometer because the high frequency shocks may cause theaccelerometer to resonate. Since the accelerometer is undamped, it often breakswhen it resonates. The mechanical isolator was developed in response to impacttest requirements for a variety of structures at Sandia National Laboratories(SNL). An Extended Technical Assistance Program (ETAP) with theaccelerometer manufacturer has resulted in a commercial mechanically isolatedaccelerometer that is available to the general public, the ENDEVCO 7270AM6*,for three shock acceleration ranges of 6,000 g, 20,000 g, and 60,000 g. The in-axisresponse shown in this report has acceptable frequency domain performancefrom DC to 10 kHz & 10’XO)over a temperature range of -65°F to +185”F.Comparisons with other isolated accelerometers show that the ENDEVCO7270AM6 has ten times the bandwidth of any other commercial isolator.ENDEVCO 7270AM6 cross-axis response is shown in this report. Finally,pyroshock and ballistic shock measurements, performed by internationaloumnizations, show extended applicability of the ENDEVCO 7270AM6.

~ *Reference to a commercial product implies no endorsement by SNL or theDepartment of Energy or lack of suitable substitute.

3

Contents

AbstractIntroductionSelection of the Elastic Material for the Mechanical IsolatorDevelopment of the ENDEVCO 7270AM6 Mechanical DesignHopkinson Bar ConfigurationsHopkinson Bar Analysis for In-Axis ResponseHopkinson Bar Analysis for Cross-Axis ResponseIn-Axis Isolated Accelerometer PerformanceCross-Axis Isolated Accelerometer PerformanceUncertainty AnalysisConclusionsAcknowledgementsReferences

Appendix

A Frequency Response Functions for the ENDEVCO 7270AM6

Frequency Response Functions at Ambient, 70”F as Measuredwith a Titanium Hopkinson Bar (DC-1O kHz Non-DispersiveBandwidth).Frequency Response Functions at -65*F as Measured with aTitanium Hopkinson Bar (DC-1O kHz Non-DispersiveBandwidth).Frequency Response Functions at+185°F as Measured with aTitanium Hopkinson Bar (DC-1O kHz Non-DispersiveBandwidth).Frequency Response Functions for the ENDEVCO 7270AM6Mounted on a 2.0 in. Diameter Beryllium Hopkinson Bar(Ambient Temperature, DC-30 kHz Non-DispersiveBandwidth).

B ENDEVCO 7270AM6 in a Ballistic Environment -International Explosive Tests in Germany

Letter of Commendation from U.S. Army Aberdeen Test CenterExplosive Test Descriptions Performed in Lichtenau, Germany.Instrumentation Descriptions for Tests Performed in Lichtenau,Germany.Acceleration Comparison 50 gram Test, 80 kHz Filter.Acceleration Comparison for 50 gram Test, 30 kHz Filter.Wideband Velocity Comparison 50 gram Test, 30 kHz Filter.

4

37

1013172123252833333434

37

39

41

43

45

46

474848

494950

Appendix (cont.)

Acceleration Comparison for 50 gram Test, 30 kHz Filter.Wideband Shock Spectra Comparison 50 gram Test.Wideband Velocity Comparison 1000 gram Test, 80 kHz Filter.Wideband Shock Spectra Comparison 1000 gram Test.

C Frequency Response Functions for the Bobkat Isolator and theENDEVCO 7255

Bobkat Isolator Frequency Response Functions at 5000 gAmplitude and Ambient (70”F) (DC-1O kHz Non-DispersiveBandwidth).Bobkat Isolator Frequency Response Functions at 10,000 gAmplitude and Ambient (70”F) (DC-1O kHz Non-DispersiveBandwidth).Bobkat Isolator Frequency Response Functions at 15,000 gAmplitude and Ambient (70”F) (DC-1O kHz Non-DispersiveBandwidth).ENDEVCO 7255 Frequency Response Functions at Ambient(70”F) (DC-1O kHz Non-Dispersive Bandwidth).

D TNO Report 94-CMC-R1366, Mechanical Isolators

E ENDEVCO 7270AM6 Performance Specification

Figures

12

3

456

7

89

SIapdown Test Impact for a Nuclear Transportation Cask.Shaker Calibration Comparison for SNL Prototype MechanicalIsolator and a ComrnerciaI Isolator, Bruel & Kjaer ModelUA559.Hopkinson Bar Configuration for In-Axis IsolatedAccelerometer Characterizations.Simple Fixture to Make Polysulfide Rubber Plaques.Exploded View of Initial ENDEVCO 7270AM6.Unacceptable Response of Initial SNL Mechanical IsolatorDesign.Exploded View of Improved Mechanical Isolator withAcceptable Performance to 60,000 g for DC-10 kHz.Beryllium Hopkinson Bar Strain Anomalies.Split Beryllium Hopkinson Bar Configuration for Cross-AxisInput.

50515152

53

55

57

59

61

62

84

711

11

121415

16

1819

Figures (cont.)

10

11

121314

15

16

17

18

19

20

Two Flats on the Beryllium Insert for Static Compression andMechanical Compressive Shock Cross-Axis Experiments (2 in.Diameter).Theoretical Prediction of Axial and Lateral Accelerations forthe Beryllium Hopkinson Bar.Axial Strain on the Beryllium Insert.Lateral Strain on the BerylIium Insert.Consistent Hopkinson Bar Time History Data Used forFrequency Response Function Calculations (40,000 g atAmbient Temperature),Strain Response on the Beryllium ‘Insert (2 in. Diameter) Usedfor the Static Compression Experiments.ENDEVCO 7270AM6 Response on the Beryllium Insert (2 in.Diameter) Used for the Static Compression Experiments.Compressive Acceleration Shock Measured on the BerylliumInsert.ENDEVCO 7270AM6 Compressive Shock ResponsesMeasured on the on the Beryllium Insert (OOOrientation).ENDEVCO 7270AM6 Compressive Shock ResponsesMeasured on the on the Beryllium Insert (45° Orientation).ENDEVCO 7270AM6 Compressive Shock ResponsesMeasured on the on the Beryllium Insert (55° Orientation).

Tables

I

11

111

IV

v

In-Axis ENDEVCO 7270AM6 Accelerometer Experiments.

Cross-Axis ENDEVCO 7270AM6 AccelerometerEnvironments.

Static Base Strain Results.

Static Beam Results

Dynamic Beam Strain Results

20

23

242426

29

29

31

31

32

32

9

9

30

30

30

.

.

6

High Shock, High Frequency Characteristics of a

●

✎

✌

✎

Mechanical Isolator for a Piezoresistive Accelerometer,the ENDEVCO 7270AM6*

Introduction

Figure 1: Slapdown Test Impact for a NuclearTransportation Cask.

7

mechanical isolator for a piezoresistive accelerometer in response to impact testand pyroshock measurement requirements for various SNL structures. Themechanical isolator development was initiated by the impact test of a nucleartransportation cask in an end-on orientation. All the accelerometers resonated orbroke during the impact. The resonant response of the accelerometers that didnot break was so large that the data was clipped and rendered useless. Nouseable accelerometer responses were recorded for the test. Although thephysical damage to the structure could be discerned from the impact, the moresophisticated analyses of the structural response as measured by theaccelerometers could not be performed because the accelerometer data were lost.The accelerometers were subsequently isolated with a mechanical isolator designfor the slapdown impact shown in Figure 1, and data were successfully obtained.The development of the mechanical isolator continued for other programs atSNL, and the characteristics of mechanical isolator have been reportedpreviously for frequencies of DC-10 kHz and shock magnitudes of up to 15,000 g[1,2,3,4]. When technology transfer funds became available to transfer SNLtechnology to private industry, a proposal was accepted for an ExtendedTechnical Assistance Program (ETAP) with the accelerometer manufacturer,ENDEVCO*, to transfer the mechanical isolator technology.

The ETAP for the mechanical isolator consists of four phases. The first phase wasto determine if the rubber material in the isolator was the most appropriateelastic material. The second phase was to determine how the elastic materialwould be obtained by the commercial manufacturer. For the third phase, theshock magnitudes applied to the isolator were increased so that the upper limitof the shock magnitudes for the mechanical isolator could be determined. TheSNL Mechanical Shock Laboratory has been developing test and data analysiscapabilities to extend our understanding of the mechanical isolator assembly inthe high frequency, high shock environments where measurements are beingmade during impact and pyroshock tests. In-axis results for up to 70,000 g and afrequency bandwidth of DC-30 kHz with a Hopkinson bar test configuration aredescribed [5]. The last phase of the ETAP is the testing and evaluation of themechanical isolator prototypes made by ENDEVCO. The piezoresistiveaccelerometer is now available as the ENDEVCO 7270AM6 for three shockacceleration ranges of 6,000 g, 20,000 g, and 60,000 g.

Further characterization of the ENDEVCO 7270AM6 has been directed by otherSNL programs so that the high frequency in-axis (DC-30 kHz) and the cross-axisresponse have been characterized as shown in Tables I and 11 [6]. These wide-bandwidth characterizations are needed for many applications because moresophisticated analvses are being Performed with the field data. Also,“Reference to a commercial product implies no endorsement by SNL or theDepartment of Energy or lack of suitable substitute.

8

.

.

requirements have been written to qualify components for frequency ranges of10 kHz. For example, recent Army research has found that armored vehiclecomponents can be damaged by the high frequency content of ballistic shock[7,8,9]. To enhance survivability of the new generation of combat vehicles, theArmy has speafied a minimum frequency range of DC to 10 kHz for the designand qualification test of components. Since there is no capability to calibrate orcharacterize accelerometers with shock inputs for frequencies above 10 kHz, theSNL Mechanical Shock Laboratory has been given the task of characterizingaccelerometers for the conditions shown in Tables I and II. Additionally, resultsfrom SNL shock environments other than impact and international explosivetests have been made available to the authors for publication by AberdeenProving Grounds.

Table I: In-Axis ENDEVCO 7270AM6

Low Amplitude(up h 20,000 g)

Low Frequency(de-lo kHz)

High Frequency(dc-30 ~z)

Titanium

Beryllium

Accelerometer Experiments.

High Amplitude(up to 70,000g)

Titanium

Beryllium

Table II: Cross-Axis ENDEVCO 7270AM6 AccelerometerExperiments.

Environment Experimental Configuration

Static Compression Beryllium Cylinder

Static Strain Steel Beam

Dynamic strain Steel Beam

Compressive Mechanical Shock Beryllium Cylinder in a Split HopkinsonBar Configuration

9

Selection of the Elastic Material for theMechanical Isolator

The material used in the mechanical isoIator is polysulfide rubber which is soldas an adhesive in semi-solid form. This material was used for historical reasonsbecause early mechanical isolator designs used a layer of polysulfide rubberbetween two plates. The accelerometer was then mounted on the top plate withscrews in the usual manner, and the assembly was mounted on the structurewith a 10-32 stud in the bottom plate. Polysulfide rubber has been made at SNLfor many years in small plaques with a thickness of 0.010 in. However,ENDEVCO wanted a commercial source for the polysudfilde material, but nocommercial suppIier of this rubber in 0.010 in. thick plaques could be found.Consequently, several alternative materials were considered for the mechanicalisolator. These materials were chosen using the criteria: Shore A durometer inthe range of 50-90; temperature range of -65°F and +185”F, and easy formabilityinto 0.010 thick plaques. The materials chosen were: Silicone (GE RTV 511),Adiprene, Adhesive Film (3M), and Ethylene Propylene (EPDM) rubber. Twomechanical isolators were assembled with each material candidate. InitiaIly, aIlmechanical isolator assemblies were calibrated with both a shaker calibrationand a dropball calibration. From previous experience, the shaker calibration at30-50 g’s must have less than ~5% deviation over a bandwidth of DC-10 kHz inorder for the isolator to be a candidate for high shock levels. A comparison ofshaker calibrations for the ENDEVCO 7270AM6 isolator and a commercialisolator, Bruel & Kjaer Model 0559, is shown in Figure 2 and indicates thesuperior performance of the ENDEVCO 7270AM6 isolator. For the shakercalibration, an ENDEVCO 7270A-6K was mounted on the Bruel & Kjaer Model0559 isolator to maximize the performance of the Bruel & Kjaer Model 0559.Heavier accelerometers, greater than 1.5 grams, will lower the resonantfrequency of the Bruel & Kjaer Model 0559. Shaker calibrations were alsoperformed at -65°F and +185°F with similar results.

Additionally, the dropball calibration should be flat (33%) for shock levels up tothe maximum peak magnitude of the dropball apparatus, 20,000 g. If thematerial performed adequately during these calibrations, then the evaluationcontinued with Hopkinson bar testing at ambient temperatures. The Hopkinsonbar test configuration is shown in Figure 3. This titanium Hopkinson barconfiguration for in-axis characterizations has a 0.75 in. diameter and a DC-10,000 Hz non-dispersive bandwidth. This apparatus and the associated dataanalyses performed in the SNL Mechanical Shock Laboratory are described insubsequent sections of this report [1-6, 10]. The advantage of Hopkinson bartesting is that shorter pulse durations (and consequently high frequency content)can be applied to the test item and the data is available for analysis. Only two ofthese materials, silicone and adiprene, performed adequately at ambient

10

.

40

30

20

10

0

-10

-20

-30o 2000 4000 6000 8000 1000(3

FREQUEPJCY (Hz)

Figure 2: Shaker Calibration Comparison of SNL MechanicalIsolator and a Commercial Isolator, Bruel & Kjaer

Model UA559.

Fiopkhson BarCertifi@d Laser DoppIex

Air Gun \Vzbronwter

Figure 3: Hopkinson Bar Configuration for In-Axis IsolatedAccelerometer Characterizations.

11

temperatures at shock amplitudes up to 15,000 g. Adequate performance hastwo criteria: the isolator peak acceleration amplitude must not differsignificantly from the peak acceleration amplitude derived from a referencemeasurement and the accelerometer must not show resonant response. If theisolator and its elastic material have good agreement with the accelerationderived from the strain gages at these low shock amplitudes, then the materiaIhad adequate stiffness. If the accelerometer did not resonate, then the materialhad adequate shock mitigation with its impedance mismatch characteristics. Thenext evacuation was on the Hopkinson bar at -65°F, and neither materialperformed satisfactorily at this cold temperature. Consequently, the search foran another elastic material was abandoned, and polysulfide rubber remained theelastic material in the mechanical isolator.

Polysulfide rubber is purchased as a liquid and is not available commercially inthe 10 mil thick plaques required for the isolator. In order for the commercialmanufacturer to have an adequate supply of the polysulfide rubber, a simplecircular fixture was constructed to make rubber plaques with a four inchdiameter. As shown in Figure 4, this fixture consists of a circular housing and acircular piston that fits in the housing. The piston has three O-rings on it thatallow a vacuum to be drawn on the area below the piston using a fitting in theside of the housing. The piston is forced down onto stops at the top of thehousing with an arbor press. Six bolts are put in place to hoId the piston in thecorrect position during the 24 hours required for the polysulfide rubber tosoIidify. The rubber is cured for an additional 48 hours after it is removed fromthe fixture.

PSR(.01 O THICK)WITHTEFLON SHEETUSED AS MDLD RELEASE

I \ I—

=————&

PISTON:

Y3

O–RING SEALS

a+. I /= ’0’’””””

.

Figure 4: Simple Fixture to Make I?olysulfide Rubber Plaques.

12

Development of the ENDEVCO 7270AM6Mechanical Design

As part of the ETAP, tweIve accelerometers were assembled in the mechanicalisolator shown in Figure 5 as an exploded view. Since the accelerometer isavailable in five acceleration ranges, all ranges (except the lowest range of 2,000g) were used: two with 6,000 g range; six with 20,000 g range, two with 60,000 grange, and two with 200,000 grange. The mechanical isolator has never beenused with the 2,000 g accelerometer because the resonance of the isolator (=30-50kHz) tends to excite the resonance of the accelerometer at 90 kHz. Testing wasplanned for each accelerometer up to its maximum acceleration range. Testing ofthe 6,000 g accelerometers in the mechanical isolators was conducted at 5,000 g toconfirm previous results. Testing continued with the 20,000 g accelerometers at10,000 g to confirm previous results and at 20,000 g. Although the isolatorperformed acceptably at 10,000 g, the performance at 20,000 g was clearlyunacceptable as shown in Figure 6. The acceleration time history shows anonlinear response at the end of the Hopkinson bar pulse that was not evident atlower acceleration levels. The Fourier transform in Figure 6 shows several peaksthat are not usually present and a high peak for the accelerometer resonance atalmost 400 kHz. For acceptable isolator performance in the Fourier transform,the accelerometer resonance should have a magnitude that is at least twodecades below the low frequency asymptote. The observation of thisunacceptable performance led to a redesign of the isolator with six major changesas shown in Figure 7. First, the mounting stud was changed from a 10-32 to a1/4-28 to increase the stiffness of the stud and to allow a larger torque value tobe used (75 in-lb instead of 25 in-lb). The thickness of the isolator top wasincreased by 1/16 in. and the width of the slot for the accelerometer wasdecreased from 0.375 in. to 0.305 in. to raise the frequency of any modes in theplate-like top. Four 2-56 screws replaced the previous two 3-48 screws to clampthe top tighter to the bottom. This screw configuration aIIowed a decrease in thediameter from 0.745 in. to 0.675 in. Finally, the brass pins that were originallypress-fit into the isolator bottom were replaced with two 3-48 screws that haveshrink tubing over the threads to prevent metal-to-metal contact between thescrews and the accelerometer. The elimination of the brass pins simplifies themachining and assembly considerably because the press fit requires tighttolerances and an extra operation for assembly. With these design changes, theENDEVCO 7270A in the mechanical isolator respond correctly at accelerationlevels up to 70,000 g without significantly increasing the weight of the isolator(9.5 grams as compared to the originaI 7.5 grams). Above 70,000 g, the nonlinearresponse appears and is similar to that shown in Figure 6. The nonlinearresponse is amplitude dependent and maybe caused by small gaps opening inthe isolator assembly. A finite element analysis has been made of the new

13

II—I

—I

- —I II II d I

I I

II ‘ :~ ‘1

348SOCKEI’HEADCAPSCREW

ACC~EROMETER

R1-lBBERPAD,BOTTOM

SHRINKTUBING

BRASSPINS

ISOL4TER,BOTTOM

k+—————74’ ——————MMOUNTINGSURFACE

Figure 5: Exploded View of InitiaI SNL Mechanical Isolator

.

Design.

14

20.0

1s.0

10.0

S.o

(Lo

-s.0

-10.0

-16.0

-20.0

10’

10°

10-’

N

:10-’u

10-’

10-’

~.#$s

..... ..... .... .. .

HOPKINSON BAR INPUTISOLATOR 2Cikg t? 19815 g,52w j

. . . . . .. . . . .

. . . ... . . . . ...”...

120.0 122.5 126.0 127.5 1s0.0 132.6 195.0 137.5 140.0 142.5 $45.0 147.5 160.0

TIME (S) ●lo-’

a) Acceleration Time History

I-i”i!,> ,,,, ,,1, .,,

, I I I I 11! ,-’1 I I I I I 111 \ I I I 1 Ii; I I t [ I I ! I!, 1 ! ,1

I’M

2, 1 1 [ ,? *,! ! I ,,, ,! 11 I ,,, ,, },, I II

,, I 1 1 I & f 1. I I ! ( 1 ! 4 , I 1 1 t , I

5’10’

Figure 6:

d 10’ 104FREQUENCY INHZ

b) Fourier Transform

Unacceptable Response of Initial SNL MechanicalIsolator Design.

15

/ 348 SOCKET HEAD CAP SCREW (2)

/

Iik 2-56SOCKET HEAD CAP SCREW (4)

/

ISOLATER, TOP

1 I

I

I

* I SHRINK TUBING

RUBBER PAD

~f=-.

OM

SURFACE

Figure 7: Exploded View of Improved SNL MechanicalIsolator with Acceptable Performance to

60,000 g for DC-10 kHz.

16

.

isolator design, and no resonance was identified in the frequency range of 30-50kHz where resonance response has been observed experirnentalIy for theENDEVCO 7270AM6. All resonance frequencies identified by finite elementanalyses are above 100 kHz. The characterization of the ENDEVCO 7270AM6for the conditions shown in Tables I and II and results from shock environments otherthan impact comprise the remainder of this report.

Hopkinson Bar Configurations

The Hopkinson bar configuration for characterizing accelerometers for in-axisinput is shown in Figure 3. Both titanium and beryllium materials were used forthe ENDEVCO 7270AM6 characterization. In-axis input in this configuration isan acceleration that is normal to the mounting surface or parallel to the sensingaxis and the integral mounting stud. A maximum shock magnitude of 200,000 gis used for the titanium bar to insure that the 6 AL, 4V titanium alloy (6%aluminum and 4% vanadium) does not yield. In this study, the titaniumHopkinson bar has a 0.75 in. diameter, 72 in. Iength, and a non-dispersivebandwidth of DC-10 kHz. Strain gages are used as the reference measurementfor the titanium Hopkinson bar and have an uncertainty of +6% [11]. Thistitanium capability has been used to characterize the ENDEVCO 7270AM6 atacceleration levels up to 200,000 g [5].

A maximum shock magnitude of 70,000 g is used for the beryIIium bar to insurethat the beryllium (99% pure) does not yield because it is a health hazard inparticle form. In this study, the beryIIium Hopkinson bars have a 50 in. length,0.75 and 2.0 in. diameters, and a non-dispersive bandwidth of DC-50 kHz. ALaser DoppIer Velocimeter (LDV) with high frequency (up to 1.5 MHz) and highvelocity (10 m/s) capability was purchased from a commercial source and hasbeen certified by the Primary Electrical Standards Department at SNL. For the1000 mm/s/V range (positive velocity), the total uncertainty with approximatelya 95 YO confidence level for the velocity is +5 YO. When the LDV is used over 90’XOof its range, this LDV has a +2-3% uncertainty for all specified frequencies andvelocities. The uncertainty decreases for decreasing velocity scales. The LDVprovides a reference velocity measurement for velocities up to 10 m/s and forfrequencies up to 1.5 MHz. The LDV provides information in a bandwidth thatis not available from strain gages that are generally considered to have abandwidth of no greater than dc-40 kHz. Since these piezoresistiveaccelerometers have resonance at these high frequencies of 100’s of kHz, the LDVis a useful diagnostic tool. Ref. [12] gives the details of the certification process.Strain gages cannot be used as the reference measurement for the beryllium barbecause its response exhibits two anomalies that are shown in Figure 8: the highrate of damping and the non-return to zero of the stress time history.

17

.*. . I. . ...-1--.. i. l’.+ ..1.. 1.- : .p“~

-, J* . d !* “’$.

Tim, (,.,).> . ,, . a

‘/In? (m,)

TitariimnBar Beryllium Bar

Figure & Beryllium Hopkinson Bar Strain Anomalies.

A configuration that uses a quartz crystal to directly measure the accelerationapplied to an accelerometer mounted on a flyaway device at the end of a bar wasinvestigated [13] for use with the beryllium bar. However, it was found that theHopkinson bar material for this configuration is limited to aluminum that cIoselymatches the impedance of the quartz crystal.

Cross-axis sensitivity of the piezoresistive accelerometers has been studied with theberyllium split Hopkinson bar configurations shown in Figure 9. An in-axisresponse is the response of an accelerometer whose sensitive axis is in the directionof the shock. An out-of-axis or cross-axis response is the response of anaccelerometer whose sensitive axis is@ in the direction of the shock but isperpendicular to the to the direction of the shock. With beryllium’s Poisson’s ratioof 0.07, the cross-axis performance is obtained with negligible motion in theaccelerometer’s sensitive axis. No special preparations of the beryllium FIopkinsonbar interfaces with the inserts are made other than insuring that the surfaces are flatand polished. Strain gages are used as the reference measurement for the cross-axisHopkinson bar configuration and have an uncertainty of +6% [11]. A picture of theberyllium insert with instrumentation is in Figure 10.

Another configuration for cross-axis characterizations with the accelerometersmounted normal to the stress wave at the end of the bar was investigated [4].However, this approach was abandoned because the resonance at the end of the baris about 120 kHz and caused the accelerometers to resonate and break. The splitHopkinson bar configuration avoids the resonance at the end of the bar and iseasily modified to accommodate different mounting requirements for a variety ofsensors.

careful alignment of the bars with projectiles and/or inserts is required. AllHopkinson bars used for the results shown in this paper are freely supported. A

18

low pressure air gun is used to fire a 3 inch Iong hardened tool steel (fortitanium) or magnesium (for beryllium) projectile at the end of the bar. Thisimpact creates a stress pulse that propagates toward the opposite end of theHopkinson bar. Regulating the air gun pressure that determines the impactspeed controls the amplitude of the velocity pulse. Placing a number of indexcards on the impact surface controls the shape (approximately a hdf sine) andduration of the velocity pulse.

All accelerometers in this study were calibrated in the SNL CalibrationLaboratory using two methods: 1) shaker calibration and 2) dropball calibration.The two methods are traceable to the National Institute of Standards andTechnology, NIST, formerly National Bureau of Standards (NBS) as describedelsewhere [14].

BerylliumHopkinson Bars50 in. length each

Air Gun

\

,

‘7Projectile ‘

Axial & Radial ‘Insert withStrainGages Accelerometers

Figure 9: Split Beryllium Hopkinson Bar Configuration for Cross-Axis Input (2.0 in. Diameter).

19

Figure 10: Two Flats on the Beryllium Insert for StaticCompression and Mechanical Compressive ShockCross-Axis Experiments (2 in. Diameter).

20

Hopkinson Bar Analysis for In-Axis Response

The theory of stress wave propagation in a Hopkinson bar is well documented inthe literature [15, 16]. The results of this theory are summarized as follows:

A Hopkinson bar is defined as a perfectly elastic,homogeneous bar of constant cross-section.

A stress wave will propagate in a Hopkinson bar as aone-dimensional elastic wave without attenuation ordistortion if the wavelength, 1, is large relative to thediameter, D, or 10D ~ 2.

For a one-dimensional stress wave propagating in aFIopkinson bar, the motion of a free end of the bar as aresult of this wave is:

v = 2c& (1)

or,*=2c@

dt

where,

(2)

(3)

and v and a are the velocity and acceleration,respectively, of the end of the bar, c is the wavepropagation speed in the bar, E is the modulus ofelasticity, p is the density for the Hopkinson barmaterial, and s is the strain measured in the bar at alocation that is not affected by reflections during themeasurement interval.

The motion of an accelerometer mounted on the end of the bar is governed byequations (1) and (2) if the mechanical impedance of the accelerometer is muchless than that of the bar or if the thickness of the accelerometer is much Iess thanthe wavelength. The requirement on the strain gage is that the gage length (g.1.)be much less than the wavelength or 2 z 10 g.1.

Both time domain calculations, as a percent difference from the referencemeasurement (LDV or strain gages), and frequency domain calculations, asfrequency response functions, are made with the Hopkinson bar data.

21

The magnitude, phase and coherence of frequency response functions werecalculated so that a quantitative evaluation could be made of the frequencyresponse for the ENDEVCO 7270AM6. The reference acceleration data,calculated from either strain gages or the LDV, and the accelerometer responsedata were used to calculate a frf, H(jo), using the equations below [17_J.

HI+ H2H(jcv) = ~ (4)

where,5

~G XYH, (j~) . ~

~Gn=l xx

(5)

and

(6)

and where GXYis the cross-spectrum between the reference acceleration, x, andthe accelerometer response, y; GY~is the cross-spectrum between theaccelerometer response, y, and the reference acceleration, x; Gu is the auto-spectrum of the accelerometer response, y; and GXXis the auto-spectrum of thereference acceleration, x. The frf, HI is biased by the error on the referenceacceleration, and the frf, H2 is biased by the error on the accelerometer response.The Hopkinson bar data for these frf calculations have noise on both thereference acceleration and the accelerometer response, so the average of the twofrf’s in (4) is used. The summations are performed for the ensemble of fivereference accelerations and their corresponding accelerometer responses. Thecoherence, yLxY(jo),was also calculated for an ensemble of five data sets accordingto the equation [17J

(7)

as a measure of the linearity between the reference acceleration and theaccelerometer response and of the noise in these data.

22

Hopkinson Bar Analysis for Cross-Axis Response

The axial motion, a, at a location in the Hopkinson bar other than the free end is[18]

..3dt

(8)

wheres is the axial strain. The radial motion, y, is

y = i-&r (9)

where r is the radius of the Hopkinson bar ands, is the radial strain. Since therelationship between axial and radial strain is

&r= jlE (lo)

where y is Poisson’s ratio, then the fi.md expression for radial acceleration, ay, is

~y=.irda——colt”

(11)

A prediction of the axial and radial acceleration for the beryllium Hopkinson barusing the equations above is shown in Figure 11. Strain measurements co-located with the ENDEVCO 7270AM6 on insert of Fig. 10 in the split Hopkinsonbar configuration are shown in Figures 12 and 13 and indicate a Poisson’s ratio ofabout 0.07.

IT

5.6 i

I

/“”””’\\ ~_Axial Acceleration

0.6

/

‘\ [-----1 CKiXRadial Accekration

[‘>

g 0.4 -j-‘>..:

=

!! 0“: -L.s= =----’-’’’”’\\ “’’’’... ~~ ““’”””‘“-‘“ ““’=-----. /’

~ “0”2/

\ \ / //~,. /\ \ .,.. ~. ,,

/“z -0.4 -J-

/

‘ .. ,. . . .?’(”~.

-0.6 ‘.\ /“//“.,,

-0.81,

.\ ./”\. /“x.....--_#_::’”-,..--,.,..-,,”._-_,..-Y : .—+— ....__,.+.,.,.=....”... .. .... .......” -—--------

0 g U& ~ : :.> :2 g,--, g .?q ;

& w G $ Ii JN m

w &U-)

uw 1.

Lus o

Figure 11: Theoretical Prediction of Axial and Lateral Accelerationsfor the Beryllium Hopkinson Bar.

23

400

300

z WIzccwmafxC.2E 100

0

-ioo

5

0

-5

-lo

-;s

-20

Figure 12: Axial Strain on the Beryllium Insert.

1

1000 1100 1200 1303 ~+oo !5Mj cRcs~c

Figure 13: Lateral Strain on the

24

Beryllium Insert.

00

In-Axis Isolated Accelerometer Performance

The in-axis titanium Hopkinson bar testing was conducted for the finalmechanical isolator design at shock levels of 20,000 g, 40,000 g, and 60,000 g for aduration of about 40 ps and for temperatures of -65”F, ambient (70”F), and+185”F. Figure 14 shows the consistent time history data that is required for thefrequency response function analyses of these data. Appendix A shows thefrequency response functions for temperatures of -65”F, ambient (70”F), and+185”F, respectively,with a non-dispersive bandwidth of DC to 10 kHz. The

frequency response function magnitudes show Iess than four percent deviationfor the three temperatures and the three shock leveIs. For these same conditions,the frequency response function phase changes are minimal.

The frequency response function coherence plots confirm the quality of the data,and the coherence is acceptable to 20 kHz. The large deviations in magnitudeand phase that occur above 10 kHz, are due to the loss of coherence at higherfrequencies. The frequency at which the Ioss of coherence starts to effect thefrequency response function is not known, but the large excursions in thesefunctions above 10 kHz are not due to the ENDEVCO 7270AM6. Rather, theexcursions occur because of differences between small numbers at thesefrequencies in the autospectrum for the Hopkinson bar acceleration and thecross-spectium for the Hopkinson bar acceleration and the isolator response.These computational anomalies have been noted and discussed previously [1-6,10-12], and the frequency response functions are useable to at least 10 kHz andabove.

Even though additional testing was conducted with acceptable results up to70,000 g at ambient temperature, the acceleration Ievel of 60,000 g is themaximum recommended upper limit for this isolator design over thetemperature range of -65oF to +185”F. The original intention was to reach levelsof 200,000 g, but the 200,000 g acceleration level does not appear to be achievablewith this design approach. In order to reach 200,000 g, an increase in the isolatorweight and consequently, a decrease of the useable frequency range would berequired.

In an attempt to characterize the resonance for the ENDEVCO 7270AM6, it wasmounted on the end of a 2.0 in. diameter beryllium Hopkinson bar with a 1A-28threaded hole in the in-axis orientation as shown in Figure 3. A low amplitudeshock pulse of -6,000 g with a 15 ps duration was created, and the frequencyresponse functions (magnitude and coherence) were calculated. A highermagnitude shock pulse was not used because the duration of the shock pulse at15 ps causes the mechanical assembly to resonate. Although this configurationwas used successfully to characterize the 1.5 gram ENDEVCO 7270A for abandwidth of 50 kHz, the coherence is not acceptable past 30 kHz for the

25

6.0

4.0

3.0

2.0

1.0

0.0

-1.0

-2.0

-s.5

-4.00.0 2.s 6.0 7.5 10.0 32.6 $6.0 17.6 20.0

The (See] *10-4

S.o

4-0

3.0

2.0

1.0

0.0

-1.0

-2.<

-3.C

-4.(

a) Hopkinson Bar Acceleration

,0.0 2.s 5.0 7.5 10.0 12.5 15.0 17.6 Zoo

Time (see) “lo-’

b) ENDEVCO 7270AM6 Response

Figure 14: Consistent Hopkinson Bar Time History Data Used forFrequency Response Function Calculations

(40,000 g at Ambient Temperature).

26

ENDEVCO 7270AM6 that has a higher weight of 8.4 grams. The correspondingfrequency response functions are shown in Appendix A.

Although the first (lowest) ENDEVCO 7270AM6 resonance was not identified inthese beryllium Hopkinson tests, a resonance has been observed in otherexperiments. Experiments conducted in Germany subjected the ENDEVCO7270AM6, mounted on a steel plate, to shock from 50 grams of C-4. TheENDEVCO 7270AM6 exhibited a resonance at 48 kHz. The ENDEVCO 7270AM6was aiso subjected to explosive shock from 1000 grams of C-4 in the sameconfiguration. The explosive tests are described in Appendix B that showsexcellent performance and survival for the ENDEVCO 7270AM6 in ballisticshock environments.

The performance of two isolated accelerometers, the Bobkat Isolator for theENDEVCO 7270A [19] and the ENDEVCO 7255 are presented in the form of frf’sin Appendix C. The frf’s were calculated from Hopkinson bar in-axis tests usingthe configurations and equations described above. The tests were conducted onsteel for the Bobkat and titanium for the ENDEVCO 7255 Hopkinson bars withpeak amplitudes shown on the plots. Both isolators have resonance below10kHz. The resonance changes value with shock ma~”tude, so both isolatorsare non-linear for the frequency range of DC-10kHz. Their useable frequencyrange is a maximum of DC to 1 kHz. Both the Bobkat Isolator for the ENDEVCO7270A [19] and the ENDEVCO 7255 are availabIe cornrnercially.

TNO Building and Construction Research in The Netherlands performed anunsolicited, independent evaluation of the ENDEVCO 7270AM6. TNOevaluated the in-axis response for four isolators for the ENDEVCO 7270A, one ofwhich was the ENDEVCO 7270AM6. TNO wants to measure structural responseto underwater explosions that have high frequency content. The conclusion oftheir evaluation is that the ENDEVCO 7270AM6 performed the best for theirapplications. The evaluation report is in Appendix D. Finally, the performancespecification for the ENDEVCO 7270AM6 is in Appendix E.

27

Cross-Axis Isolated Accelerometer Performance

The characterization of the cross-axis response of the ENDEVCO 7270AM6accelerometer continued in the four environments of static compressionexperiments up to 82 pe; static strain on a beam of 250 ps; dynamic beam strainexperiments at 250 ps (as per ISA-RI? 37.2, Paragraph 6.6); and mechanicalcompressive shock with a 2 in. diameter split Hopkinson bar configuration. Theorientation of the sensors for the static compression and mechanical compressiveshock cross-axis experiments are shown in Figure 10. Radial strainmeasurements verified the performance of the beryllium. Figure 15 shows thein-axis and Poisson strain as a function of applied load, and Figure 16 shows the7270 AM6-20K response (nominal 10 pV/g sensitivity) as a function of strain.For the static and dynamic base strain experiments, a steel beam as specified inISA-RF’ 37.2, Paragraph 6.6 was used that creates maximum surface strain at thefixed end of a cantilever beam. The transducers mounted at this location aresubjected to base strain. The base strain is one-dimensional surface strainbecause the beam has a very large radius of curvature that minimizes the motionat the transducer and the centrifugal acceleration, roz. The results of the staticcompression, static beam, and dynamic beam strain are shown in Tables III-V,respectively. The ENDEVCO 7270A-60K (flat package) and ENDEVCO7270AM4-200K (can package) [6] results are included for comparison. Theresponse of the ENDEVCO 7270A and ENDEVCO 7270AM4 accelerometers inthese three environments is a base strain response and is consistent with themanufacturer’s specifications of <0.5 mv output for a strain of 250 pe. Theresponse of the ENDEVCO 7270AM6 is much lower in these three environmentsbecause the ENDEVCO 7270A in the isolator mechanical case is essentiallydecoupled from the mounting surface.

The compressive acceleration shock applied to the ENDEVCO 7270AM6 isderived from the strain gages co-located with the ENDEVCO 7270AM6 as shownin Figure 10. The time history compressive acceleration shock has a magnitudeof approximately 25,000 gas shown in Fig. 17. The ENDEVCO 7270AM6response to compressive shock measured on the insert is A 750 g or less for three,randomly chosen, orientations of 0°,45°, and 55° as shown in Figs. 18-20. Thecross-axis responses are 370 (30,000 Hz bandwidth) of the compressiveperpendicular shock magnitude and are very consistent for the three orientationsof the isolator perpendicular to the compressive shock. The frequency content ofthe ENDEVCO 7270AM6 cross-axis response is predominantly 20,000 Hz and40,000 Hz. These two frequencies have not been observed for any other testconditions and are assumed to be rocking motion of the ENDEVCO 7270AM6, inthe absence of other information. Since these frequencies are significantlybeyond the specified bandwidth of DC-1OKHZ for the ENDEVCO 7270AM6 andthe magnitude is a very small percentage of the compressive shock, the cross-axisresponse is considered minimal.

28

Applied Force (kip)

15

I

-lo

5

0

-5

.10

-15

0

15: Strain Response on the Beryllium Insert (2 in. Diameter)Used for the Static Compression Experiments.

sOtid-ZNDEV[

1

Dash -!ENDEVI. ...... ...................... .... ......

. ... . .. .. ...... ...... ...... ... .------ -----

........................... ... ..

-10

3

7270,4 fIj6-20K ~ 5 Degrees (S/N D!+l 1F) ~

17270A$6-20K @j 95 Degr4es (S/N ~71WF) ~... ................. .. . . . ..... ........ . .-<.. ..... ......... ... ..... .. ........... ............ .......

t ,30 40 50 60 70 E

In-Axis Strain (micro-strain)

..............

.. ... .

——..............

16: ENDEVCO 7270AM6 Response on the Beryllium Insert(2 in. Diameter) Used for the Static CompressionExperiments.

29

Table III: Static Base Strain Results.

Accelerometer Orientation Torque Strain Sensitivity Voltage outputType (in-lb) (g/p&) at 250 ~ (rev)

7270AM4-200K 80° 75 0.5816 0.187

7270AM4-200K 25° 75 0.0612 0.019

7270AM6-20K 5° 75 0.0188 0.045

I 7270AM6-20K I 95° I 75 I 0.0188 I 0.045 I7270A-60K I In Line \ 9 I 0.1531 I 0.064t7270A-60K 45° 9 0.4184 0.166

7270A-60K 90° 9 0.3061 0.124

Table W Static Beam Results.

Accelerometer Orientation Torque Strain Sensitivity Voltage outputType (in-lb) (g/ye) At 250 ps (rev)

7270AM4-200K In Line* 75 3.2260 1.0907270AM4-200K 15° 75 0.2368 0.080

I ,

7270AM6-60K 15° 75 0.0070 0.0067270A-60K In Line 9 0.4232 0.1757270A-60K 45° 9 0.6772 0.280

7270A-60K 90° 9 0.0943 0.039

* 0.009 in shim

Table V Dynamic Beam Strain Results.

Accelerometer Orientation Torque Strain Sensitivity VoItage outputType (in-lb) (g/w) At 250 w (rev)

7270AM4-200K In Line* 75 3.1370 1.0607270AM4-200K 15° 75 0.5387 0.182

7270AM4-200K I 15° I 30 I 0.1717 I 0.058

7270AM6-60K 15° 75 0.0147 0.012

7270A-60K In Line 9 0.3918 0.162

7270A-60K I 90° 191 0.0435 I 0.018I , , ,

7270A-60K 45° 9 0.5804 0.240

* 0.009 in shim

30

0

gC3

&00

xmmo(nmc-)

G00

g00

z00

——.===

T..

.

.

....,

..

. ..

. ....

—

------

—

—

—---—

. —

—

—

—-

<

<

.

! -—.—(

.._.

_—

2

—.-’,

----->

— -.

< :

,-- ----J

.. ~

c :

—

---

—

—-—---

—

—

—

.

—

.—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

1500

1000

500

0

-500

-1000

.. . .. .. . .................. . . .. .. . . . . ._$dRlkllKi[/./.;.........

...................................................................................1 i

o 500 1000 i 500 2000 2500MICROSEC

Figure 19: ENDEVCO 7270AM6 Compressive Shock ResponsesMeasured on the Beryllium Insert (45 Degree Orientation).

1000

500

u)

C.9o

zol-.cEL3

-s00dc-l‘.2G

-1000

-15000 500 1000 1500 2000 2500

MI CROSEC

Figure 20: ENDEVCO 7270AM6 Compressive Shock ResponsesMeasured on the Beryllium Insert (55 Degree Orientation).

32

Uncertainty Analysis

The uncertainty in these measurements and results are attributed to: uncertaintyin the sensors, the reference measurement (strain gages or LDV), the dataacquisition system, and accelerometer response due to variation in mountingtorque. The sensor and data acquisition uncertainty is monitored on a continualbasis in the SNL Mechanical Shock Laboratory as required by the SNLSpecification 9958003 [13]. These requirements include the performance of boththe hardware (sensors, amplifiers, digitizers etc.) and the IMPAX software thatcontrols the data acquisition system through a computer [13,14,15]. The 9958003specification allows an accuracy of ~10% for amplitude, ~5 % for duration, and+8% for rise and fall time for any measured pulse greater than 50 ps in duration.—The current data acquisition system and software meet these requirementswithin ~0.5 ‘XO,and documentation of these results is maintained in theMechanical Shock Laboratory. Consequently, the uncertainty in thesemeasurements is the uncertainty in the accelerometer calibration, ~5 % [14], theuncertainty of the reference measurement 6 % [11, 12], and the uncertainty in thetorque Wench calibration, ~5% [6]. These three uncertainties are consideredrandom, so they maybe combined in an uncertainty analysis with a 95%confidence level as [24-25].

‘T J= W:+w; +w: (1)

where: wT = totaI uncertain~,WS= accelerometer calibration uncertainty, 5%,w~f=uncertainty of the reference measurement, 670, andWt = torque wrench calibration uncertainty, 35 %.

The vaIue of the total uncertainty, wT, is ~ 8% and is typical for themeasurements made in the SNL Mechanical Shock Laboratory.

Conclusions

A mechanical isolator for a piezoresistive accelerometer, the ENDEVCO7270AM6, has been developed at SNL, and has been transferred to ENDEVCO.The isolator has acceptable performance for shock levels of 20,000 g, 40,000 g,and 60,000 g and for temperatures of -65°F, ambient (70°F), and +185°F with afrequency bandwidth of DC to 10 kHz. The performance was determined by themagnitude, phase and coherence of frequency response functions calculatedfrom in-axis Hopkinson bar data and cross-axis response in four environments.This mechanical isolator has ten times the bandwidth of any other commercialisolator as shown in Figure 2 and Appendix C. The mechanical isolatortechnology, that was deveIoped in response to SNL impact and pyroshock test

33

requirements shows an extended capability in some explosively generatedballistic and underwater shock environments as shown in Appendices B and D.The performance specifications for the ENDEVCO 7270AM6 are in Appendix E.

Acknowledgements

The authors would like to acknowledge the expertise of the SNL personnel in theOrganic MateriaIs Deparbnent, Mathew Donnelly and David Zamora, whodeveloped a process and constructed the fixture for making the polysulfiderubber plaques. The authors would also like to thank Thomas J. Baca whoprovided a starting point, and R. Glenn Bell III, Neil T. Davie, and Patrick L.Walter for their contributions during the development of the isolator.

References

1.

2.

3.

4

5.

Bateman, V. I., R. G. Bell, and N. T. Davie, “Evaluation of Shock IsolationTechniques for a Piezoresistive Accelerometer,” Proceedings of the 6@Shock and Vibration Symposium, David Taylor Research Center,Portsmouth, VA, November 1989.

Bateman V. I., R. G. Bell, F. A. Brown, N. T. Davie, and M. A. Nusser,“Evaluation of Uniaxial and Triaxial Shock Isolation Techniques for aPiezoresistive Accelerometer,” Proceedings of the 61st Shock and Vibration

Symposium, Vol. IV, October 1990, pp. 161-170.

Bateman V. I., N. T. Davie, F. A. Brown, “Characteristics of a PiezoresistiveAccelerometer in High Frequency, High Shock Environments,” Proceedings

of the 4@ lES ATM, Chicago IL, May 1994.

Bateman V. I., F. A. Brown, and N. T. Davie, “Isolation of a PiezoresistiveAccelerometer Used in High Acceleration Tests, F!roceedings of the 17fiTransducer Workshop, Vehicular Instrumentation/Transducer Committeeof RCCTG, San Diego, CA, June 1993 pp. 46-64.

Bateman V. I., F. A. Brown, N. T. Davie, and M. A. Nusser, “High Shock,High Frequency Characteristics of a Mechanical Isolator for a PiezoresistiveAccelerometer, Proceedings of the 18fi Transducer Workshop, VehicularInstrumentation/Transducer Committee of RCCTG, Colorado Springs, CO,June 1995.

34

References (cont.)

6. Bateman V. I and F. A. Brown, “The Use of a Beryllium Hopkinson Bar toCharacterize In-Axis and Cross-Axis Accelerometer Response in ShockEnvironments,” Proceedings of the 44A International lMrurnentation

Symposium, Reno Nevada, May 3-7,1998, pp. 540-549.

7. Walton, Scott W., “New Ballistic Shock Protection Requirement for ArmoredCombat Vehicles,” 6(N Shock and Vibration Symposium, Vol. 1, November1989.

8. Walton, Scott W., “The Significance of Shock Content Above 10kHz onEquipment,” 6.W Shock and Vibration Symposium, Vol. 1, November 1992.

9. Walton, Scott W., “Methodology Investigation Final Report of Correlation ofComponent Damage to Ballistic Shock II,” U.S. Army Combat SystemsActivity, Aberdeen Proving Ground, MD, TECOM Report No. 7-CO-M91-CSD-004, Report No. CSTA-7241.

10. Bateman V. I and F. A. Brown, “The Use of a Beryllium Hopkinson Bar toCharacterize In-Axis and Cross-Axis Accelerometer Response In ShockEnvironments,” SAND97-2862, National TechnicaI Information Service, 5285Port Royal Road, Springfield, VA 22162, January 1999.

11. Bateman V. I, W. B. Leisher, F. A. Brown, and N. T. Davie, “Calibration of aHopkinson Bar With a Transfer Standard,” Shock and Vibration Journal,Vol. 1, No. 2, pp. 145-152 (1993).

12. Bateman V. I, B. D. Hansche, and O. M. Solomon, “Use of a Laser DopplerVibrometer for High Frequency Accelerometer Characterizations,”Proceedings of the 66~ Shock aqd Vibration Symposium, Vol. I, Biloxi, MS,November 1995.

13. Togami, T. C., V. L Bateman, and F. A. Brown, “Evaluation of a HopkinsonBar Flyaway Technique for High Amplitude Shock AccelerometerCalibration,” Proceedings of the 68A Shock and Vibration Symposium, Vol.I, Hunt Valley, MD, November 1997.

14. Bateman, V. I. And O. M. SoIomon, Jr., “Characterization of AccelerometerMountings in Shock Environments,” Proceedings of the 14fi TransducerWorkshop, Vehicular Instrumentation/Transducer Committee of RCCTG,Colorado Springs, CO, June 1987.

35

References (cont.)

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

Davies, R., “A Critical Study of the Hopkinson Pressure Bar,” PhilosophicalTransactions, Series A, Royal Society of London, Vol. 240, pp. 352-375,January 8,1948.

Kolsky, H., Stress Waves in Solids, Oxford University Press, 1953.

Bendat, J. S. and A.G. Piersol, Random Data, 2“d Edition, John Wiley andSons, 1986, pp.164-185.

Bateman V. I, F. A. Brown, and N. T. Davie, “The Use of a BerylliumHopkinson Bar to Characterize a Piezoresistive Accelerometer in ShockEnvironments,” Journal of the Institute of Environmental Sciences, Vol.XXXIX, No. 6, November/December 1996 pp. 33-39.

Walton, W. Scott, “Pyroshock Evaluation of Ballistic Shock Techniques,”Proceedings of the 62fi Shock and Vibration Symposium, Vol. III, pp. 422-431, Springfield, VA, October 1991.

Ulibarri, Davie, and Kuehnle, “Mechanical Shock Test Instrumentation,”Sandia NationaI Laboratories Specification 9958003-4,1992, pp. 1-19.Bateman, V. I., “Data Acquisition System Hardware Fidelity CheckProcedures,” February 14,1997, pp. 1-21.

Bateman, V. I., “Software Management Plan for Software SupportingProduction Lot Acceptance Testing,” January 11,1996, pp, 1-5.

Bateman, V. I., “Software Quality Requirements for Area I MechanicalShock Laboratory,” Issue A, January 11,1996, pp. 1-10.

Doebelin, E. O., “Measurement Systems Application and Design,” McGrawHill Book Company, New York, N. Y., 1983, pp.57-60.

Abernathy, R. B. “Measurement Uncertainty Handbook,” InstrumentSociety of America, Research Triangle Park, N.C., 1980.

36

Appendix A

ENDEVCO 7270AM6

Frequency Response Functions

37

rER

:NT

DIFFE

:PicE

PHRcJ

E

:~

EEs

103 .l~+ 1!35

FREQUENC”i [HZ)

a) Frequency Response Function MagnitudeI 1 ! i 1 I 1II 1 I I I 1I1I 1 I I 1 1 i t I

::: :;::: :::: :::: :::: :~::::; ;::: :::!: ::: ::;: :::

::: ::::; : ::: .: ::: :::: ::::.:: .. :.: ::::: ::: ::: :;::. .::::: ::!; :::

2D,~13W3~G ; &$~ ; ~ i::::: ::::::: :::: ::;: ;::

. . . . . . . . ...+..... . .. . . . . .. .. . . . . .. . .. . . . .. . . . . . . . . . . . . . . . . . . . . . . . . .. . . .>..<...:..+<.:.. . . . . . . . . ...}..... . .:.....:..-.:...:.. +..;..+:,:::: !.

413,j3E+G }+o~:;:::::: ::;: :::: :::

:;: ;:::!: :::; ::;:,: ;:::: : ::: .:::

::. : :::: :::: :::613&3B;G ; ++D[ ~ f

::;: :::: :::: :::

~ m-------”------”:-”-j-”!--!---.:

““””--7Kjtititi”:G---jti+::$~*$~*:::: :$,. . . .

‘>;::;\ ; : ::

::::,, .. . . . . . . . . . . . . . . .. .. .. ..y.

:::.:::;::::

. . . . . . .. . . . . . ... . . . . . . .,::::::

.........

... ..

.... .....

lE13 ,a+

b) Frequency

FREQUENCY [HZ)

Response Function Phase

38

cRHEi?ENcE

1

0.5+2

EI.96

13.94

0.g~

0.s

0.88

0.86

0.84

U.8%

0.810 103

Freauencv

,5+

FREG!UENCY (HZ)

c) Frequency Response Function Coherence

la

Response Functions at Ambient, 70”F as Measured witha Titanium Hopkinson Bar

(DC-1OkHz Non-Dispersive Bandwidth).

T

39

rJER

ENT

DI

FERENcE

PHRsE

DEGREEs

40

1

0.99

0.98

E1.97

0.96

0.95

0.94

a.93

E1.92

D.s!l :;:::;:::; :;;: ;:: ;

::: .: .,: . . :,:0.3 I I I I I I I 1 ! 1 I 1 I I

!!32,g3

c) FrequencyFREQUENCY (HZ)

Response Function Coherence

Frequency Response Functions at -65°F as Measured with aTitanium Hopkinson Bar (DC-IO kHz Non-Dispersive Bandwidth).

41

;;;::; ;;:; ::, ;::, ,., . ::. ,., ,. j,:::;;:-2Ei I t ! 1 1 I I ! I I 1 I 1 I I 1 I ! 1 Ii I I 1 I ! 1

102 la3 lb+ IE15

FREQUENCY (HZ)

b) Frequencv Response Function Phase, * J A

42

~

uH

NcE

FREQUENCY (HZ)

c) Frequency Response Function Coherence

Frequency Response Functions at +185°F as Measured with aTitanium EIopkinson Bar (DC-1OlcHz Non-Dispersive Bandwidth).

43

; *3 4

FREQU&CY (HZ)

,05

Response Function Phase

44

1

0.9

0.8

0.7

0.6

0..5

0.4

0.3

0.2

0.1

0.10’ 4 ,05

FREQUpNCY (HZ)

c) Frequency Response Function Coherence

Frequency Response Functions for the ENDEVCO 7270AM6Mounted on a 2.0 in. Diameter Beryllium Hopkinson Bar

(Ambient Temperature, DC-30 kHz Non-DispersiveBandwidth).

45

Appendix B

ENDEVCO 7270AM6 in a Ballistic ShockEnvironment

International Explosive Tests in Germany

46

.

DEPARTMENT OF THE ARMYU.S. ARMY ABERDEEN TEST CEWER

400 CDLLERAN ROAD

ABERDEEN PROVING GROUND. MARYLAND 21005.5059

February 27, 1997

BallisticsDivision

Dr.Vests BatemanSandiaNationalLaboratoriesP.O. BOX 5800, Mail Stop0555Albuquerque,NM 87185-5800

DearDr. Bateman:

The AberdeenTest Centerwould like to thankyou for fabricatinga mechanicalisoIatorfor the7270accelerometer.TM devicewaswed to measureexplosivelygenerated“ballistic shocY at levels whichareknownto destroyhardmounted7270accelerometersandotherstandard,commerciallyavailableaccelerometers.

The mechniea]ly isolatedaccelerometerthatyouprovided survived5 testshotsinLechtenau,Germanyand7 testshotsatAberdeenProvingGroundwithoutftilure.Mr. ScottWalton(410-278-3396)andiMr.MichaelClark(41O-278-5120)canprovidefurthertectilcal details.

p+ercly,

/’1 1

Y 1&.L. ‘j&k;J Id L. Nook

hector.EngineeringDirectorate

47

oEXPLOSIVE TEST RESULTS FROM *?.andia National Labs

TEST DESCRIPTKNIS

1. The data on tie SRS plot labeled “SO gram German~ were generated in Germanyon 22 May 96. The actual test vws 50gm of M with a 90mm standoff from the centerof a small, fiat steel pl,ate. Dimensions of the plate were 4SOmm x 460mm x 4Dmm. Allgages were mounted ~dially equidistant (127mm) from the center of the plate. TheLOFFI record was genemed by a mechani~lly fiiterad (80W-lz rail offj Endevc.o226zA-2000. The BOBKAT record was generated from a mechanically filtered(8-kHzrolloff)Endavco 7270A-6K. The AT(2 VeIocity gage is Scmti Walton’s veloci&’ gagedesign.

2. The data an the SRS plot Iabe!ed “1OOOgram GermanF were generated inGermany on 22 May 96. TfYetest ansisted of detonating 1000gm of C4 placeddirectly against ths German BaIlistic Shock SknuIatrx (SBS). The expbs-kfe charge wasplaced against the front plate (which was perpendicular to the grnund) of the simulator-The gages were mounted on a plate (parallel to the ground) that butted into the platethe explosive was placed against The mass and geometry of the SBS is roughlyequivalent to a turret of a main battle tanlc

INSTRUMENTATION FOR EXPLOSIVE mTESTS INLICl+TEhlAll,GERMANY Smd,a National labs

SOgm Against Flat PJa@ Wimm Stend-oi7

Gacie TYoe Gaqe Factor Sample Rate Anti-Alias

ATC VeI~ Gage Y956$ 2ao,oao SO-kHz

BOEKAT (7270A) 36,032g/ 200,000 80-kHz

LOFFI (226X) 3,084g/ 200,000 SO-kHz

Sandia (7270A) 670,691g~ 200,000 None

fOOOgm Against German Ballisdc Shock Simulator (S0S), NO Stand-oti

Gaqe Twe Gage Factor SamrNe Rate Anti-Alias

ATC Velo~ Gage 12.70; 200,000 80-kHz

BOSKAT (7270A] 36,032 g/V 200,000 %C-kHz

LOFFI (22624) 3,084g/ 200,000 80-kHz

San&a (727oA) 599,520~& 200,000 None

48

ACCELERATION COMPARISON m50 GRAM TESTY 80 KHZ FILTER s.mdja=b

L’” ~........ .Smdia 7279 C6iSE&)

tt)!lrl~! !l’!!:ll l),:,0.0s 2Q.EUS S88.EIIS ?s8

Ims

r t 1 J 1

s

Note: High Frequency on 7270A-AM6 is at 48 K1-lz.

ACCELERATION COMPARISONo50 GRAM TEST, 30 KHZ FILTER san.,a.g.,s

1.o~+4I I I I i I I 1 I I I

~I [ [ I

l“i I

: — DerivATC Velocity 6age;; . . Smdia 7270::

Hi5000.0

-5000.0t -i

tfJllll 111 J!lllll, ill, ,l0.0s 250. Dus 500.Dus 750.OW I .Oms

TIME

For a bandwidth of 30 KHz, both signals are reasonable.

49

WIDEBAND VELOCITYmCOMPARISON =50 GRAM TEST.,..,.....=.’..

2.0

1.5

1.0

0.5

0.0

-0.50.0s i.Oms 2. Oms 3, Oms

TIME

oACCELERATION COMPARISON m50 GRAM TEST9 30 KHZ FILTER ...m.....,~.~.

-wal

11’’’’ 1’’’’ 1’’’’ 1’’’’ 1’””’&

2.0E+4

F /1 —French 7255----French 7255-+andia 7270

----DerivativeofATC Vel

1- fl

i.0E+4

I HA..,.,.,......L

0.0..

:..“

I I I I 1 I I I I I I I 1! 1,,1,,,,-50.OUS 0.0s 50. &Js Ioo.ous 150.Ous 2oo.ous

TIM5

Comparison shows 7255 #l is clear outiier.

50

WIDEBAND SHOCK SPECTRAmCOMPARISON “ 50 GRAM TEST.n.ra.a*om].~s

CO.g-l

(uL “orl)-

0)

:CL “o

o-1French Sandia 7270

----------------- FITCVel aci ty-------- BOBKFIT———— LOFFI

c +a. “o ;/$“ 1 i I Wiil

2 3 u 56769’1 I I II(1

; 3 U 567B9i[ I I I 1111 [ I I 1 1 I2 3 4 567!391 2 3 uS67#

.+ 10’7 102 ,031o~ ~~s

Resonant Frequency (Hz)

WIDEBAND VELOCITYmCOMPARISON =1000 GRAM TEST,,..,.,.,..,,.,.

I 1 I I I I 1 I I [ I 1 i I I I I 1:3.0 — Integrated %ndia 7270

,... ATC Velocity Gage

2.01’-

0.0+--+

...

I

-2.0 I I 1 I 1 1 I I d,, , F I I I i0.0s 2.5ms 5. Oms 7.5ms

TIME

51

WIDEBAND SHOCK SPECTRAmComparison u1000 GRAM TEST...,~atitnanLa,s,s

%c “7o

Fren>h Sandia 7270---------------- FiTC Veloci ty

s++ - -------- BOBKRT

(u——— — L(3FFI

C “a_(u -z

Q)

8a ‘o_

o..4

*

cd

’70 _w) “z

-P’

c

Q)

-o

(d” I I 1 1(11 I 1 I I [Ill I I I I illl 1 I 1 1[1

3 2 3 ~ ;6?891 2 3 Ii 56789’ 2 3 q 567891 2 3 i5678J.+ [01 [02 103T I 04 I 05

6-W Resonant Frequency [Hz]

52

Appendix C

Bobkat Isolator and the ENDEVCO 7255

Frequency Response Functions

53

Percent Difference

I WA

,. -- . . .!------- -------

.---- ..;..---..: ------

. . . . . ..!-..--- . . . . . . . .

-.-. - .;. ----- +..-....I 1

1

-. ..--, ~----..-+.,-.--

. --- . -,------- +------, ,,

. “---------- .--- .”.- ------------

iiii

. . . . . . . . . . . . ... . .....-+-... . ...1 ------

---- .L . . ..--.l ------- L.. - . - .-J..-...

----- . . . . . .. .. . .. . . . . . . -------- . . . . . .

--- --:------- ;.”.----;-..-.-.;-------...:..

I. . . . . ..+. -.--.. J - “-- . --,-------- :. . . ----

, ,

1 ,$

, I------ ~-------r-.- . . . ..*-_----,------’

4 ,8

,,,

, t..---..-, . - . . . . --------- -------- ,------, 8

, , ,t-----: -_. -..-l - ~ . . . -------------, . .. -_-z

, t

t # ,, , ,

,!,

,t ,8 ,, 8

, ‘—

,

I 8, , I ,

,,

,------- ; --. -----)--- . . . . . . . . . . . . . . . . . . ..;. .. ....-- +......--, I , ,

--------- ------------ ----- ---------- * . . . . .. .. . . . -k -----

, t , ,, , , ,.1

. - . .;....--+ ----- +----- ------- ----- ;----- ; ----- ;----- ;-----8 1I t t , I ,

-.. -A--- . . .. L-------- -----: ----- ------ $ . ----: -----, L . . - ..-;-----

, , 4 ,. - . - . - . -- ..,----- , -----, .- ..-. . . . . - . ----- 4----- : . ---- ;-----

$ , ! 1 I 1

. ---------- +---- +-- ...-.: . ----- --- . --: . . ----: ”..---: ------; -----1 I , , i , I

, , , ,t ,

t

,,

, I, I

I II , , 1

1 ,

,, , ‘

I , : ,

,-_--l----;-L------;------L ----- ---..+-.-__._-.- ... .. . . .. . .. . . . .

I , , ,1

,... . .. r------ ,----- :.. -..--, --- -:--- _ .. : .. -- .. -:------ -------------

# I I I

--- ..+------ * - .. _ .._:.-... , ----: ----- ;----- : ----- + ----- i---- -# , , ,

, , i, . -..-,---- ~ - .. ---; . . . . -. . ... ----

, ,,

. ----: ------: ------:---- -+- . - .:- . . - . .:. . - . . -:------: ------ :------

. ----} -- . . _} --- ._;_-_--’

, ,-.. -.+... . . . ..+.....- ,. __-.+ . - . ..-+----- ,---- _+----- ,- . . ..- ,------

-- * ,,4

,, ,

1 .,4 , 1 , 1, , , 4

,,

, I8 I

I$ , 8 ,

I

,, I

, , ,I , , , a ,

(I t

, $

, t I ,

----- ; -----: . --- .; .---- ;-...-; -. . . -; --. --; ---- .;.-..-; ----, , , , 1 I I

c) Frequency Response Function Coherence

55

Bobkat Isolator Frequency Response Functions at 5000 g Amplitudeand Ambient (70°F) (DC-1OkHz Non-Dispersive Bandwidth).

Percent Difference

VIm

, , 8 0 ,. .-...+.....-..:-- ----- L . - . - . ., . . -.-.-;..- ___

, I t ,

, 1 t ,----- q------- ,-- --- <’------- -, ------ ~------- ,------

,

#

. . --------------------t

t,

7-------; --- . . -------- .:------,, , ,

, I~.. -----:. ------ + ----. -.,....--

,.----- .-------’ ~. -----

.-- . . . +. ..--...-+------- }. . ----- .. . . . . . . +... ..- .-,------ ..---- .:... -.-.: ----- ....... . . . . . . . ------:. ----- .+----- -——, t.—

, 8

-.---

-.---

-----

. ..-.

. . . . .. . . . .. . . . .

.~

, I ,------. ------- . -- . - . - 7-------- .---- .-’ , --...--.’ ,------ ‘. . .. .. .. . .

, ---- t ,

.---- ------ . . .. ---- ------ ----- . . . .... -------- . . . . . .,,, , 1,!,,), , ,!!!,, ,, ,,!,

. . . . . ---- .+---- . ; -------------- +. ---- ------------ . - . . . . . . . . . .,,, , , f,!,,,, , , ,,8,

.--- AL . . . . . . . . . . . L ---- ..!..--., . . . . . 1.. ---- L.----L ----.L----., ,!1,

,,, , , ,,, ,------------- --- . . . . --------- 4.. . . . . . . . . . . . . . . . . . . . . . . . . . . .

,,! 4,, ,,,.. --, -- .-..; . . . . . . .-- .---, ------ .- - . - q . . . . . . .-. --., -. . . . . .-----

!,(, Ill,——

,,, , ,!, ,,(1! ,,, ,,,, , ,,, ,,,, , ,10,0,s, $ ,,8,,,, , , ,,, ,!,, , , !,4,,elt 4 8,1,!,, ! , ,,, ,,,, , , ,,, ,,,, , , ,,!,,,!, , ,,!,

,,, , , ,,, ,

. ---,----- . . . . . . . . . . . . . . --.-,---.- ,----- , -----, -.---, -----

0,4 ,,, ! ,

. . . . . . . . . . , -----+--- - ; -----: ----- ; --- . - . . ...--..+ . --- .+-----,,, ,It

1,8 ,,,4 4

-r -----} -----r -----,1! ,,, , 8

. ...+.... .. :----- +_---- :----- f------ :--- ..+-- .- +----- .. . . . .,,, ,

-----,,!, 1,

,,!,. . . . . . . . . ., . . --------- , --...:...

,

__J.-.+---- .: ,,*, t

,,, , ,,, ! ,,,, , ,,, ! ,,,, , ,, ,,,,!} !!,‘,, , :’ <:,1 ,t,, , ,!, ! ,,,,8 ,,81 ,,,, , ,,, , ,,!!1 ,,, ! ,,,,1 ,$48 ,,,, , ,,, ! ,,,!! 1,, ! ,,,!! !,!1 !,,

. - . -, -.- . -,. --- .:.--..’ , Ir-----r-.---r.......................,,1, 1,, ? ,

1.2 ,II,,: ,;,

e

(J, ,, , ,

I,\$, ,; :1, :,,, ,, ,, ;;:,,;, :, ,,

!,;:; !,, ?,: ,,, ,, ,,,, ,( ,,,: $:, ,, ;:; ,‘ ,

!,, ,,!;,, ,,, ,: ,

1.0 ;,,, ,; 1 ;:,; ‘,,, ,, c !,, ,$

,:, , ,,, ,: ,,,, ,

c , ,,, ,,::, ,, , $,,,;>,~,f,

t ,, ,:,, ,, ;;:,

:~,;;,

,,,:, ,, :;;,

,,:t

, !,, , (:0.8 7;:,‘------ :,.-.--L- ----------------------- :-----:--_-:---:---:--:-- ,-

8 ::::: 1!,, ,, --------------:),, ,: ,!,

6i

:,, : , ,,~:1 ,1 ,K

.,, t ,, ,,,~,:: ,, :::,,

y if, 1,,;, !:

,,, ;;:,Io ;::;:

,,

0$,;;;,

0.6 +--:--4--:--:- - ------------- ;--------:-----i----;----;---:_-:-- :- --------------, $ $,

*,, ,; $,;;:*!,~,,, ,,,1,1, ),; ;;$,,, ,, , (,,:, ,, ,, ::: f, ,;, !,~t ,1, :! ,,*, , !,8,t !,8, ,,,, , ,,, ,, I

0.4 :---+-+ ;-- -------------- ;--------:-----;----k---:---L--;--+-- -------------- ‘

,f, ,; !:::, , ,,!, :,

, ! ,,, ,: ,,,, ,,

,,t~ ,,,, ,

,, ::;,,,, ,,

:,, ,: ,,, , ,, ,

,,, ,, ,,, ,: ! ,!,,:, ,,, ,, ,,}, : , :,, ,:,

0.2 ‘ !I I !,! I 1, I 1

5*102 Id 104

FREQUENCYHZ

c) Frequency Response Function Coherence

Bobkat IsoIator Frequency Response Functions at 10,000 gAmplitude and Ambient (70”F) (DC-1OkHz Non-Dispersive

Bandwidth).

57

70$.0

75.0

50.0

m 25.0Inal%

% 0.0~

2a

-25.0

-50.0

–75.0

-100.0

1.2

1.0

0.8

0.6

0.4

0.2

5*102 10’ 10”FREQUENCYHZ

a) Frequency Response Function Magnitude

;:;,, ,: ;;<,,,1,

,,, ,;,

,,!,,,, ,,!,, ,, ,

,,!:t,,

t,, ,: ,

,,!,1,, ,

,,, ,‘, .,: :,, ;,,

.,, ,,,, ,

?, r,: !

,,

!,, ,, ,

. .. ----- .-. L- ------ .-----.-, -------- ,----- :- .-:---J ---t-_,-- ,_ -------------- ~.,,,,,’, 2;;:;’,,*, ,, ,!,, , ,, ::, ,,,, , , !, f

!,,’ ::;:: !

‘,, , t ,,, ,!,!, , ,, ,;,,!, !,;;;,,,, , ‘,, ,, ,

$,-+ --;--: --,L- . ----- +---- ~ ---;---:--:-- ,-

!,1, ------------- ;-------- , l!!,!,

,,, ,, 1, ‘-- ‘------ ----,,,, , ,1, : :

,,, , :,, ,,,!,,!,:, ,, ;;;,,,, ,, ,,,t, , ,,,<!, ,, ;:) ,r,, ,

,,, ,,

,,!: ,, ::;, h--.--;-..-AL- . ----- . --------: -.------, -- .--+---- :--+.-:-+. .-,,!, ,, 1--------------:.!<,!!,!,

,, ;:;!!,, ,

,,, ,, ,,,

,,, , !,;:: ,!,, , ,,,,, , ,, ;;;: Ii,,!: ,,,,, ! ,, :;: , ~~,,, ,

,,,,, ,, ,

--,--<--.--L_._-------------;--------:-----.----’r--+--: -.:-- +-.!,!,

--- ---------- ,.I ,,, ,,

,,!, ,,, ,: ,,,t~ !,, ,,!,!) ,,, ,; ,!,1, !,, ,

,,!, !, ,: Iv!,’, ,, ::: ,!,!, ,,,,!, !,::, ,!,!,I ,, t,,,;,

) , 1 I I , 1 I I

. . .5*10” lo la

FREQUENCYHZ

b) Frequency Response Function Phase

58

.

25.0 ,, 1,,, II,,; !:: ,, ;:, :,,,, , ,, ,,!l l,, ,,:: j;,

2G.O :---{ --;--:--:-- -!,1 ,,

. . . . . . . . . . . ---: -------- . -----, ---- ~-.-:.--~--:.- ,- ----------.- ----!,

,(!: !,, , :;:., ; :, !,, : I,, l,, !,

!:, !,,, l,,15.0 ~---: --<--1--:-- - . -------------- -------- . - . --- . ----c ---,--- ,--;--, -. -----.-.------ L-

,,, : :!: 1:, ,,! ,,:,, >,, !,,, :4,, !,: (;

;: ;) (!:10.C $---: --<--:--;-- ---------------: -------- : -----;----:---:---:--:-- :-- --------------:-

,:!:: ,,! l!,,,,!,

OJ,,, : !

,, :;o

:, l:, ,

--------- .-----:--------:-----:----;----;---:--:--:-- --------------;-Sb

!,:: l:, .,, ,! :!; “

~Q ---- -------------- L-

!;, ;f? ,, !,:

,, :,,:

-5.0 ;---:--;--w- - ------ -------------- L-&

CL !,!:, ,,:: ::,,: 1,:

,<,l:!

i,, ,

-10.0 *--. --:--4--:--,,!,::+:;,,!, ,,,!,l, ,

_~~,() *--,-.J+--.-+ -!;;::,,

!,!,,:~

:;, ,-’20.0 . . . . . . . . . . ------,,

,; :,:,,!,!;:

I 1t

-25.0 ~ ‘ :

5*102 Id 104FREQUENCYHZ

c) Frequency Response Function Coherence

Bobkat Isolator Frequency Response Functions at 15,000 gAmplitude and Ambient (70°F) (DC-1OW Non-Dispersive

Bandwidth).

59

25.C

20.0

15.0

10.0

5.0

0.0

–5.0

–10.0

–15.0

–20.0

–25.0

100.0

75.0

50.0

25.0w

20.0

%

–25.C

–50.0

–75.(?

–100.0

...’,..”,’”,../

,,.““/ “’\...”

r140016000 g. . .. .500 g--------- ..-.5000 g—.— .9000 g—- —

...”

-------------......... .....

5*70* 104FREQUENCY W

a) Frequency Response Function Magnitude

I-...,.,.,..........

16000 g..... .500 g------------ .5000 g—.— .—9000 g—- —

I I ! I I [ I I I 1. 1 I

5*102

b)

10’ 104FREQUENCY HZ

Frequency Response Function Phase

xl

60

1.2

1.0

0.8

0.6

0.4

0.2

0.0

—. — -—

1400 q

16000 g.. .. ........500 g. . . . . . . . . . . . .5000 g—.— .—9000 g—- —

—-— .— .— -— -— -

5*102 103FF/ZQU~NCY~Z

c) Frequency Response Function

104

Coherence

ENDEVCO 7255 Frequency Response Functions at Ambient, 70”F(DC-1OkHz Non-Dispersive Bandwidth).

61

Appendix D

TNO Building and Construction ResearchThe Netherlands

TNO Report 94CMC-R1366, Mechanical Isolators

62

TNO reporl

94-CI?C-K1366

TNO Building andConstruction Research

Mechanical fi.leers

Lange KIeiweg 5. Rijswii< Oa!e

P.o.aox 4s2600iW CMThe Nettmiants

Au:hor(S)

Pr!0ne+31 15(2)&42CWFax +31 15 (2)843993Telex 3L?270

Ail figh!s resewedNO pan c1[his publication maybereprcd~ce dandforpublii%by PfintPPOIOprinLCIiC1OfihOrany Othefmeanstimx!{ Ihe ?retiOuS writtenconsent o: TNO.

Ir. casethisqmtwas dd. mimtructiom.the rights andobligations@ contracting paitks are subject 10either the Sandard COnd%nS fOfResesrch fmm&ma given to TNOor the rtievant agreement concludedbehveen the contractingpates.%bmtiii me repal forin$w2ionto*e&s whohavea directinlemstispermitted.

Sponsor: i-iinissryof DefesceDirectorateof Material

i?oyd Hetherkwis lizvyP.O.BOX 207022500ES TiiEHAGUETHE NETHERLANDS

Monitoring Agency : TEO DefenceResearch

Classifiedby : ir. P.J.Keunir.g,RNINClassificaczon6ace:30 Novenber1994

Project number : Q77L$613 - A92/KM/12~”Author : ing. B. Bosrnzn ~yApproved :lr. A.IJ.VredeveldtVisa : ir. G.T.M..3anssenPages 23 (excl.RDP & Distr. +

List)Classification : ONGERUBRICE,ERE /

The classificationdesignation‘O$KXWJBRICEERD’is equivalentto‘UNCMSSIFIED’.

0 TNO

.... ..... --:...,-. 4.,

63

TNO-report94-CMC-R1366

Date Page’30 November 1994’ 2

ManaKeme%tuiCtreksel

Titel . :Auteur :Datum :Opdr.nr. :lWP nr.Rapportnr. :

t4echanLcalfiltersIng. B. 130sman30-11-’94r4277461.3 - A92fKM/1227929.4-CMC-R1366

Ken viertal mechanische filters, te gebruiken voor &e Endevco 7270Aversnellingsopnemers, zijn beproefd.De frequentiekarakterisciek (zowel in meetrichcing als dwars da.arop),ce lineairiteit ev.de temperatuursinvloeden zijn coc een frequenciesvan10 ldizge=valueerd.He: is mogelijk OE, net gebruikmaking van speci.alezechaaische filiers,zwzre schoksignalen in een frequer.tiegebiedtot 10 kHz juist teregiscreren.De ge~vzlueerde mechanische filters vertonen klei~e nieclineairiteiten.Van de geSvaLueerde mechanische filters heeft het Disk-isolacion-cechnique filter de besce karakteristieken.Ch het bruikbare frequentiegebied van een mechanisch f~lter te bepalendienen de werkelijk opgetreden omstandigheden (zoals amplitude,cemperacuur, etc.) in beschuuwing te worden genomen.

64

TNO-report94-CMC-R1366

Contents

1.

2.

3.

4.

L.1

4.2

4.3

5.

6.

6.1

6.2

7.

8.

Dare Page.30 Novenber 199L’ 3

SUMARY

INTRODUCTION

EVALUATED MECH~’ICAL ‘FILTERS

EXPERIMENWXL SETUP AND WXSWREMENT SYSTE??.“

Transducer calibration

ExecuCe6 measurements

Usetiequipment

DATA /WALYSIS

~SULTS

Tests on the MIL 901D light weight shock mchine

TesCs on che hopkinson bar

CONCLUSIONS

REF-ERENCES

FIGURES

&?&

&

5

6

7

7

7

7

8

9

9

9

11

12

13

65

TNO-report94-CNC-R1366

Date Page’30 Novem”Der 1994” &

1. SMMARY

Four diEferent mechanical .filcers,to be usetiwith Endevco 7270Aaccelerometers, have been tested.The frequency characteristic (measurement direction and transverse),linezrity and influence of temperature have been evaluated till 10 kHz.It is possible co measure high level shock signals in a frequency rangeup CO 10 !&z properly, using spec<al mechanical filters.The evaluated mechanical filters show a.small nonl<neari~y.Of che evaluated mechanical filters the Disk-isolation-techriiquemechanical filter “showsthe best c’naracteriscics.One has to consider the actual circumstances (such as amplitude,temperature, etc.) to define the usable frequency range.

TNO-report94-CMG-R1366

2. INTRODUCTION

Tine calculation tnethods

Date Page

- 30 November 1994 5

for the effects of underwatershock on a

construction are improving strongly.Of course there is a need for verification by ex~eriments.Because of the strong improvecientsin the calculation methods thefrequency contencs of the experimental data has co be much higher tha~in the past.Several accelerometers with a very high natural frequency arecommercially available. TNO experience is that in praczice thoseaccelerouie.tersc~ be exciCed in resonance and fail, even though this

resonance is 180 kliz.The only way Co zvoid failure, due to excitation in resonznce, is the

use of mechanical fil~ers or internaly damped transducers.Tineaim of this investigation is to design a mechanical filter for cheErxievco 7270A transducer with a usable frequency band=idth of 10 Id-l-z.FOX che usable frequency bandwidth a maximum amplitude deviarion of 5%,a v.aximumphase deviation of 5 degrees and a maximum transversesznsiti.vityof 5X is acceptable.

66

TNO-report Dzce Page94-CMC-R1366 ’30 l?ovember 1994” 6

3. EVALUATED MECHANICAL FILTERS

Four different mechanical filters, to be used with Encievco 7270Aaccelerometers, have been tested.The first one is the filter as used by TNO in the past at full scaleexperiments and small scale experiments. For small scale experiments a

large bandvidth is necessary. This fil~ex will be called TNO-old

filter. This filter is show in figure la.The second one is the nodified filter of che previous one (thinnerrubber). This filter is called TNO-new filter and is shown in figure

lb.The third is a further modified filter called TKO-new-round filter.This filter is shown in figure lc.The fourth is a filter developed by Sandia National Laboratories andmade available by Endevco. This filter is called Disk-isolation-Cechnique filter. [1],[2]Endevco scaced that this filter was not the best they had of this type.Ike rubber of Che first two filters (TNO-old and TNO-new) were glued onthe steel surface. The rubber of che TNO-new-round filter was glued onChe aluminium surface. The accelerometer was clamped between VO rubberlayers ac the Disk-isolation-technique fil~er.The TNO-new-round fil~er and the Disk-isolation-techniaue filter bothhas a stud for attachment to the test structure.The TNO-old fil~er and TNO-new filter were glued to the test structure.

67

TNO-report94-CKC-R1366

&. E.X2!XIHF.NTAL

Date’30 Xovember

SETLIPAND MEAWRE.MEWT SYSTEM

Page1994” 7

4.1 Transducer calibration

The accelerometers used for this study were calibrated by TNO using twomethods; a shaker calibration according to 1S0 5347 part 3 “Secondaryvibration calibra~~on” and a dropball calibration according to ISO 5347

pare 4 “Secondary shock calibration”. [3],[4],[5]Figurs 2 shows a dropball calibration &=ta record.

4.2 Executed nteasurenencs

Some neasuresents took place on the klIL901D light weight shock machineat a cemperacure of 20 degrees Celcius. Ilifferenc shock levels vereused (200 g, 300 g, 500 g and 3000 g).All other measurements were carried out using z hopkinson bar.On the hopkinson bar some different amplitudes, pulse durations andcenperztures were used.

The hopkinson bar (made of.AISI -304) is shown in figure 3.The support allows free movement of the bar in axial direccion.The hopkinson bar easily lends itself to temperature conditioningbecause the end of the bar can simply be inserted into a temperature

chamber. In literature the theory of stress wave propagation is welldocumented. [6],[7],[8]A controlled impact of a projectile a.gainscche bar generates arepeacdole one-dimensional ekstic stress w-avein the bar which resultsin a motion at ‘the free end of the bar of : V-2CC

vith v the velocity at the end of the bar.c the wave propagation speed in the bar.e the strain measured in the bar and not affected by reflections.

The shape (approximately a half sine) and durztion of che pulses arecontrolled by impact speed and interface.Acceleration amplitudes of 2500 g, 6000 g and 15000 g and pulsedurations of 45 p, 55 ps anti70 KS were used zt temperatures of 20 and50 degrees Celcius. The pulse durations were measured at the 10Z levelof the first peak in the acceleration signal.The four mechanical filters as well as a hard mounted accelerometerwere measured.On the hopkinson bar also transverse measurements were performed.For transverse measurements an accelerometer mounted at the.end of thebar vas used as the reference acceleration.

4.3 Used Equipment

All transducers used for this evaluation were Endevco 7270A-20K (range20,000 g) accelerometers.For the measurements on the MIL 901D light weight shock machine and onthe hopkinson bar a Scadas-11 frontend with PDFA-ETD modules and 16bits ADC (make Difa) was used.A sampling frequency of 125 ldizand a analogue low pass filter of 61.7

68

.

.

,

TNO-report9L-CMC-R1366

s. DATA ANALYSIS

Date Page’30 November ‘L994” 8

For che MIL 901D light weight shock m=chine the frequency responsefunctions of the mechanical filters were calc=lzced, using the data oftransducers mounted on the mechanical filters and -a harL mountedCransducez. The frequency resolution for this calculation Was 6.25 Hz.For the calculation of the frequency response funczions 18 zverzgesWere used.

For the ~opkinson’bar considerable Preparations of the datz ~zs

required before frequency response functiOnS could be cdculaced.

Firsz che data had tO be corrected ‘or ‘he time delay of theacce~eracion record compared to the strain record (this citne~elay is

equal to the time for the G“ave CO propagate fIOm the .sCrain gage to the

end of che bar).Then the acceleration was int:grace~ to ~btain velocity azc che srrain

record was conver~ed CO ‘eloc~tY using tne factor 2c.From these data the freq”encY resPonse functions were calculated using

a impzce window co Prevent leakage errors. The frequency resolution for~hi~ calculation is 488 Hz.~oc the calculation of the frequency response functions 5 averages wereused.

69

TNO-report94-CtfC-R1366

Date Fage30 November 1994 9

6. RESULTS

6.1 ~~ Tests on the MIL 901D light weight

Figure k shows a time hisio~ measured on theshock machine. The anplituciespectrum of this

shock machine

MIL 90LD light weightdaCa is shown in figure

5. The frequency response funccion of the TNO-new filter, usir!g3000 gshock level ad 18 averages, is shoc+min figure 6.AC frequencies higher than 3000 Hz the coherence is 10L’.Thetransducers are mounted directly next co each ocher St che centerline.Figure 5 shows Chat che high frequencies are present in the aeasured.accelerztionsignal. Apparently the data is very pl~ce dependent forche high frequencies.‘i’hesignals ge~erated by the MIL 901D light weight shock mzchine cznnot be used so judge che functioning of the r.echanical filters at h%ghfrequencies.

6.2 Tests on rhe hopkinson “oar

A time history of a 2500 g shock pulse (pulse duration of 70 As) on thehopkinson b=r is shown in figure 7.Figure 8 shows the integrated and time shifsed acceleration signalcombined with che velocicy signal of the strain gages.The amplitude spectrum of the 2500 g, 70 ps acceleration pulse is shownin figure 9 and the amplitude speccrum of the 15000 g, 45 +acceleration pulse is show-nin figure 10. It is clear that thesesignals can be used for evaluating the filter response in che frequencyrange from 1 to 10 kHz very well.The coherence functions are calculated for the hopkinson bar data.The calculated coherence functions are unity in the presented frequencyrange (1 to 10 kHz).The frequency response func’cionsof the dLfferent mechanical filters,for a shock level of 2500 g (70 W} at a cemperacure of 20 as well as50 degrees Celcius, are shown in figures 11 CO 18.All mechanical filters show more or less acceptable characteristics a~20 degrees Celcius.A temperature of 50 degrees Celcius proves to high for the TNO-new-

round mechanical filter.The transverse sensitivity of the TNO-new-round mechanical filter isshown in figure 19. The transverse sensitiviq of each mechanicalfilter is given in table 1.

mechanical filter transverse sensitivity [%]

TNO-old (worst direction) 6

TNO-new (worst direction) 7

TNO-new-round 5

Disk-isolation-Cechnique 3.5b

Table 1. Transverse sensitivity

.

70

.

TNO-report9L-cW-R1366

Date Page

’30 Novernbex 1994” 10

The frequency response function of the TNO-new-round uiechmical filterac a shock level of 15000 g (45 ps) and a temperature of 20 degreesCelcius isshown in figure 20.The frequency response function of the Disk-isolation-techniquemechanical filter ac a shock level of 15000 g (45 PS) and a cempexzcureof 50 degrees Celcfus is show-nin figure 21.Both filters show a small nonlinearity.

TNO-report94-cMC-R1366

Date Page. 30 November 199L 11

7. CONCLUSIONS

It is possible to measure high level shock signals ix?z frequency rzngeup co 10 ldizproperly, using special mechan~cal filters.

The signals generated by the MIL 901D light weight shock machine can= the mechanical filters at highnot be used to judge the functioning o.

frequencies.

A temperature of 5,0&egrees Celcius is to high for the TlfO-new-rountimechanical. “filter.

The $xzrusverse sensitivity of the TNO-old and TNO-new =echaniczlfilters are too h<gh.

The evz~uazed mechanical filters show a small nonlinearity.

Of rhe evzluated mechanical filters the Disk-isolation-techniquemechanical filter shows the best characteristics.

One has CO consider the actual circtunstances (such as axplitude,

temperature, etc-) co define the usable frequency range.

71

TNO-report

94-CMC-R1366Date Page

-30 November 1994” 12

8. REFERENCES

[1] Bateman V.I., Bell R.G., Davie N.T.; EvaluaCio~- of shock isolation

techniques for a piezoresistive accelerometer.

Proceedings of the 60th Shock and Vibration Symposium, Da.viti

Taylor Research Center, Portsmouth, VA, November 1989.

[2] Bateman V.I., Bell R.G., Brown F.A., Davie N.T., Nusser M.A.;

Evaluation of uniaxial and triaxial shock isolation techniques for

a piezoresistive accelerometer.The Shock and Vibration Bulletin, No-61 VO1.4, Ott.1990.

[3] ISO 5347 part 3. Methods for the calibration of vibration and

shock pick-ups - .Secondazy vibracioc calibration. X994

[L] 1S0 53&7 part 4. liechods for the calibration of vibration znd

shock pick-ups - Secondary shock calibration. 199L

[5] Kandboek Kwaliteitszorg TNO-BOUW.

Centrum VOOK Mechanische Constructi.es.

Faciliteiten- Kalibratie Shaker. 1991

[6] Lekkerkerker J.G.; Dynamica van machines.

Vakgroep Technische Mechanica Afdeling der Werktuigbouwkunde

Technische Universiteit llelft. College bw73 1981.

[7] Goldsmith W.; IMPACT The theory and physical be’nav-iourofcolliding solids. Edward Arnold (publishers) LTD. 1960.

[8] Kobky H.; Stress waves in solids.

Oxford at the Clarendon press 1953

72

TNO-report94-CXC-R1366

Date Page.30November 1994” 13

a) l?m-ola b) ‘IWO-WW c) mo-N.%-Roum

FIGURE 1. MECHANICAL FILTERS 1’------ ---

.SE43C8SI

?JHM8J

s

II 3s.000 41.55:

x

!!

w—

1

&, I4 ~ 1* .E-n .11 ,Jl, Max U*” .

~~ *“+@TiizizG-. . ... ------ .. . . .

73

TliO-report9&-CKC-R1366

DaCe Page-30 Xovembez 1994’ 14

/

IAONCEXXG CIIJh~E2 STRAIN GAG-&

/

?Rc.JZCHLS/ /

//

,//,

‘@ PKINSON BAR

/

AA r

1000:1

-------- ---- .- -.-r. -.-. -.--- .---. -.4--. -.-. J.-. ---. -}---------------------------------

/i

h

i

\

Como?-?-JL 3s00 c-

FIGURE 3. $:OPKIXSON ERR

i “-.... . . . -.”-—---- --2 —-----2

md

4M0.1 I

Into-

1 -

.lcao-

FI(3JllI! .4. l!im history masured on the MU901DLI?S[WSZJ

74

.

~0-repori94-cKC-R1366

Date Page“ 3Q Eovember 19$’i 15

(kntre forliechmical Engineering~’~

I

.[ 1 1 IL

L[ 1 f I 1 IIfl! I I I I 11111 t I i IL(1

u-l 1%-1 M m+kHz

A?.?IMUDE

COEEMNCE

75

‘rNO-reportg.&cKc-R1366

Date Page30 November 199& 16

iCentre for MechanicalEngineering

L1‘-1.$-O.\-L$-c.[-

R.2

$.&

IWXIIW7. ‘13mhistoryof hopkinsonhr 2500g stiockpulse (70nicrosec)

1 / &ntre forliecha.nkal Engineering1.1 ‘

; . .>:.-->,... ....-4. ,.., .-:.. .. .. ..+-....,: .,.:....

-1.1 -M 4.3 4.

+illstrain signal and integrated and time shifted acceleration [ids]

.

76

TNO-report9h-CMC-R1366

Date Page“ 30 November 1994 17

(kntre for MechanicalEngineeringWE

‘ }ACCELg~~~ION

1; E/s2

m:

ii. .

iii

‘~VELOCITY

/1 ~]s‘~

U.1: !

lx-l I 1 \ I 1’ I I Iw

--@z“~FUURE9.Amplitudespectrumof 2500 g (70nicrosec) acceleration pulse

(Mkre for MechanicalEngineeringlu

ACCELER..TION

d S2

u ii

R‘! VEUX!ITY.

Dlis1U-1z

is-l& t ! I ! I ( I IMO

+~fl~WURE 10.M@itde spectrun of 15000g (45microsec) acceleration ~ulse

77

TKO-report94-CMC-R1366

Date Pageso Novenfoez 1994 18

Cwtre for MechanicalBn@eeriDgL20!

1.s

Liy

LC5.—--—-—-----—-:

. .,.

0.9s

o.$~k

6

-5- --—. -. ——.,.. .__...

-ta22$

FWURE11, FRYof TNO-old(2500Q, 20 degrees Celcius)

I Centre forlieclxmical Enginee.riDg1.28“

1.2s MHJIUDE

lJ$ ._..—- -...-... -- ....—..—

LIS

1.M-–

0.S- ;—

$.9! ‘

Iu

tPEASE 0

-5 -

-

-uw

+~z’:~

FIGURE 12. FRFC)fTNO-dd(25.00g, 50 degrees celci.w)

78

TNO-report Date Page94-CMC-R1366 “ 30 November 1994 19

Centre for MwhanicalEngineeringLlo

l)J-

1.14

1.8,-————-.1

i.w

qns

U.PJ$-5———-1?13$ x~

FIGURE13. I’RF of TNO-new[SW) g, 20 degrees C!elciw}

79

TNO-report94-C?lC-R1366

Dare Page30 November 199L 20

FIGURE16, FRFOf TW-w-round (2500g, 50 degrees Celcius)

80

‘mo-zeport94-cliC-R1366

Date Page- 30 November 199’L 21

[ +~z ~1FK!URE17. FRFof Disk-isolation-technique (2500g, 20 degrees (!elcius)

+~~ ~1FKUIU318. FRFof Disk-isolation-technique (2500g, 50 degrees Celcius)

81

‘SNO-report94-CHC-R1366

Date Page- 30 Rovenber 1994 22

Centre forlkchanical Enffineerinu I

I +~z ~JFIGURE19.FRFof TNO-new-roundin transverse direction (2500 g, 20 deg.C)

II Centre for MechanicalEngineering1.1?‘

L13

1.IG “--.-—---–-— -—-. . . .. . ,—------- .+-----

1.C5

0.5$

6 — ,

~- 1~1

Pm% o-s

$ ~1.1$m

+~Hz”~FIGURE20+ FRFof TNO-new-round(15000g, 20 degrees C.elcius}

82

.

.

,

TNO-report94-CKC-R1366

Date Page

“ 30 November 199f+ 23

Centre forlleckmical Engimertig1.16r i

m

1.14

S.c$ ———— ___. -_-;. —._——.— —

i

1.W —

$.s$

$.q 1/1

-s ---

-M ,m l!l~l

1 -JHz I IilWX13.E21. FRFof Disk-i. solation-teclu@e (15000g, 50 degrees Celcius)

.

83

Appendix E

ENDEVCO 7270AM6

Performance Specifications

84

L

PiezoresistiveModel 7270AM6

Accelerometer

Mechanical Filter6000 to 60000 g Full ScafeDC ResponseRugged

Units——- .. -6KM6 -20KM6 -6QKM6

: RANGE [7]..— — —.-—— —--—. .

.._-_..-.3_,L ..... . . . .-.-.— .:5Ow m lxx -50 C:’cl. .. .. .._;SENSITIVIW

—.——— —-—-.iv J .,,._,”~_ 3C - 2CJ,’-:5 j:+

——..— .—. — .? -2?. 1 5\ FREQUENCY RESPONSE [2]

.-.. ..”-. —.....-. ..-.——.- . .. . .. ..- ...”

I izi(% max. ret. ‘.00 Hz] ~~~ — . . C[om ~t5 :~,c:c,~c . .. ——— ——— . .— —....-~hiOUNTE D RES-~fi”~~&~~KCY kt+? T;D_ --_”,-”. ...,, ,,._,,,-- :E,.~,-,,,,,-,,,,-,,,,-,,,””.~~-?e,,______Wjc ---- .-,,,-~rmizKHwT”Y-— ~: wax -.2 up m zccelera:m :ofresmnrlm~ la IP,e :ecomme KJ9~ :am;e.,.-— .. ._ -—..: (.%CT:ea,m~.:0 fbk ran$el

.—..——.McasurefnfjP! CC,ceEaIntIeS pfe,.,en: sta:mc IP,l$as a s~ec,- cawm I ni;

wove 10 WC a... . —... ____________TRANSVERSE SENSf71’/l TY’ ‘.. ?ys 5 5—.... -.. . . . . .-. —.— >

ZE?.O ME..ISURSNCI i3U-PIJT—. . . . . ..-. .

zl\J Mai .:1 ,>!: .:o~ :.10:. .. —- .-— — .- —.. —— .- —. . . . . . . .

,

85

Piezcwesistive Accelerometer

SPECIFICATIONS-continued

PERFORMANCE CHARACTERISTICS-sonIifsued

! Units -6KM6 -20KM6 -60KM6 I\ THERMAL ZERO SHIFT! From -30-F ?0 +150=F (-34:C to +66-C)

I THERMAL SENSITIVITY SHIFT

m\/ ),i~x :50 =~~ +50—.I

/ From -3wFand+150>F (-34eCand*66’C) %.,Max :1 c, xl o t? c\ OVERRANGE LIMIT g Pk Zla 000 50 coo :lCO 0C3/ WARM-UP TIME

.—..—Mirmes Max ~ 2 ~

ISeconds [Typ) [15) (Is) (15)

ELECTRICALEXCITATION 10.0 Wk. 12 Vdc maximumINPUT RESISTANCE 550400 ohmsOUTPUT RESISTANCE S50 z200 ohmsINSULATION RESISTANCE 100 megohms minimu,m w 100 VrX: beween sensor. cable shield m! c?.se

PHYSICALCASE. MATERIAl S:ain!eSs Sieel{17.4 PH CRES)ELECTRICAL, CONNECTIONS Integral cabie, 4 conductor iio. 36 AWG Te:lcm’9 me.ula:ed leads. fxalced

—— ;1

shield, tlourwar bon jacke:IDENTIFICATION ManufaclWir’sJogo, model rlumbcrand seria:cumber

-/

MOUNTINGmORfX.JE [? ] 75 Ibf-in (S,4 Nm)WEIGHT 8.4 qrams excluding cable (cable weighs 3.6 Qrams{meter)

..- . . .....-— —.. .—-

ENVIRONMENTAL

ACCELERATION LIMITS zlOOOOOghalfsine pulse or~ree:imes the:ecommcnded ra:>ge, ic anydireclion, whichever is smaller. Pulse duration should ae the greater of 20miCrOS@cofidsor five periods of the re$onance f[equency

TEMPERATURE

~ww!... ..:.30’F 10 Y150”F (-3C’C tO +66’Q-65’F 10.r250-F (-54’C to +121 ‘C)Unaif ec.ted. Uni[ is epoxy sealed

. . .. .

—!

CALIBRATfONDATA StIPPt-lED

t SENSlTW~f51 ;lv?Qat ?XWOQ I

ACCESSORY3123C Wrench

0PTK3NALACCESSCMES250% 4Ccndwtor Shielded Cable32153 Tri-axial Mwn!ing wack

NOTES:. Removable Loc-Tke sh3vlCbeappi!edio 10-32 sludhelore inSaflzicn

cm?

Cors1inue6prc!uucli?nprwemen: ncmess!Latesmat Endewm reacwc the right lC modify these spec.ihcatlons Wnos: notice. EnCevco mamlasns a program oic@w-anlsuweiUnce o\`wall prOducts loensure atigalevel olreliabili~. This pro~ram,ncl~desalteni`tin !or&iab,l iYfaclor$ nur,ngprOductdcsi~c, thcsuJps<01stnn~er,t%ak~ Comroirewwements,andcompulsmycorrec:!veact!onprocedures.Thesemeasures..wether WA consem’atiwspect!tczl!$ns have OIate...- .. . . .

86

Distribution - External:

1

1

1

1

1

1

1

1

Bob ClarkENDEVCO30700 Rancho Viejo RoadSan Juan Capistrano, CA 92675

Randall D. CopeNAWCWT’NSCode 478600DChina Lake, CA 93555

Mark Gelak (Code 9520, Bldg. 559)IHD NSWC101 Strauss AvenueIndian Head, MD 20640

M. Shannon HaatajaAMSMI-RD-ST-WFRedstone Arsenal, AL 35898

Brad HannaNSWCDD/G2217320 Dahlgren RoadDahlgren, VA 22448

Danny HaylesAFRL/MNMF306 W. EgIin Blvd., Bldg. 432EgIin AFB, FL 32542

George N. HenningsNAWCWPNSCode 478COODChina Lake, CA 93555

David HollingsworthNAWCWPNSCode 473320DChina Lake, CA 93555

1

1

1

1

1

1

1

1

Ron LeeL-281Lawrence Livermore National LabLivermore, CA 94550

Richard Mabry101 West EgIin Blvd., Suite 219WL/MNMFEglin AFB, FL 32542

Brian MaryUS Army ARDECAMSTA-AR-CCF-A2800 Powder Mill RoadAdelphi, MD 20783

David MontoyaLANLP.O. BOX 1663, MS C931Los Alamos, NM 87545

Art SavareseUS Army ARDEC, Bldg. 6Picatinny Arsenal, NJ 07806

Steve SmithAFRL/MNMF306 West Eglin Blvd., Suite 432Eglin AFB, FL 32542

Peter K. Stein5602 E. MonterosaPhoenix, AZ 85018

Keith ThomasAFRL/MNMF101 West Eglin Blvd., Suite 246Eglin AFB, FL 32542

87

Distribution - External (cont.):

1 Chris Janow 1US Army Fuze Management OfficeB1Picatinny Arsenal, NJ 07806

1 Jim Kennedy 1LANL, MSP950Los Akmos, NM 87545

1 William L. Konick 1US Army TACOM-ARDEC, Fuze DivisionAttru AMSTA-AR-CCF-A2800 Powder Mill RoadAdelphi, MD 20783-1197

CommanderUS Army Aberdeen Test CenterATTN: STEAC-SL-B (Mr. Scott Walton)Building 359Aberdeen Proving Ground, MD21005-5059 .

Dr. Patrick L. WalterTexas Christian UniveristyDepartment of EngineeringBOX298640Fort Worth, TX 76129

Bob WhittierENDEVCO30700 Rancho Viejo RoadSan Juan Capistrano, CA 92675

88

Distribution - Internal:

11

1111116

11111111

21

MS0303MS0314MS0328MS0328MS0427MS0453MS0453MS0847MS0553MS0553MS0555MS0665MS0665MS0828MS0986MS0986MS9018

MS0899MS0612

R. G. Lundgren (15414)M. R. Scheffer (15425)J. A. Wilder, Jr. (2612)M. J. Retter (2612)W. R. Reynolds (2103)D. L. McCoy (2104)N. R. Hansen (15414)R. A. May (9126)V.I. Bateman (9126)F. A. Brown (9126)M. A. Nusser (9122)L. J. Azevedo (2541)R. R. Romero (2541)T. C. Bickel (9100)D. E. Ryerson (2660)T. A. Rohwer (2665)Central Technical FilesDept. 8940-2Technical Library (9616)Review & Approval DeskDept. 9612For DOE/OSTI

89