NAVY HIGH-IMPACT SHOCK MACHINES FOR ... HIGH-IMPACT SHOCK MACHINES FOR LIGHTWEIGHT AND MEDIUMWEIGHT...

0'

C> NRL Report 5618

NAVY HIGH-IMPACT SHOCK MACHINESFOR LIGHTWEIGHT AND MEDIUMWEIGHT EQUIPMENT

I. Vigness

I, Shock and Vibration BranchI Mechanics Division

-j

L:..IJ 7

June 1. 1961

•242

SNA V h'

U. S. NAVAL RESEARCH LABORATORYWashington. D.C.

Best Available Copy

C'ONTE NTS

All.;trartP 'obt, in•1Stati

iiAut horniz7at ioi

INTRODUCTION

Objhct1Earlv ]lbslozrv I

EQUIVALEN(': OF SHIPBOARD SH(X'K ANDITS I,ABOIiATORV SIMULATION 2

ANALYSES OF SHOCK MO. IONS 3

Shon' l.t t.-iVvhwcily- Shock 6Sinpllh Shock Puises

-7Misoel lalrous 7

"Sl"UI'FYIN(; A •!C"K TYST 7

NECE'SS"ITY FOR PREC:(ISION AN'))A(-( t('

FOR TESTS ON SIIOiK MACHINES 8TIIE NAVY II1SIiO('K MA(CINE FultLIGHtTWE;IGIT EQUVIPMNT

9

Loading Ai ralalrimirts le p'rle1 ('laiigtrstw. 9bI;titruin itai i m;i d Mi *j t.uiniiliI

TIlE NAVY !II SII(X'K MACHINE FOfRMEDIIUMWEIGIIT E,;UIPMENT 26

Dh•sr 2i6ill 26

mi'a i l - l s 28CONC LUDING DIS(33SSION

ACKNOWLEDI)GMENTS 35

RIE FERENCES 36

ABSTRACT

Descriptions are given of the Navy HI shock machines for light-weight and mediumweight equipment. Shock motions are given forstandard loading conditions. These are illustrated by acceleration-,velocity-. and displacement-time relations. Maximum values ofvelocities and displacements. and of accelerations passed by variouslow-pass filters, are presented. Shock spectra are presented forselected conditions. Equivalent displacement- and velocity-shock,together with maximum values of acceleration, can be established fortheir respective effective frequency ranges from observations of theshock spect ra.

Concepts relative to the specification of shock tests are con-sidered. These include brief considerations of analyses of shockmotions. trethods of specifying a shock test, and what is meant bysimulation of field conditions. It is indicated that shock tests shouldnot be specified in terms of shock motions, or spectra, unless thevalues specified be considered only as nominal values.

PROBLEM STATUS

This is an interim report on one phase of the problem; work iscontinuing.

AUTHORIZATION

NRL Problems F03-02 and F03-08Projects RR 002-03-42-57"54 (RN), SF 013- 10-01,

Tasks 1790, 1793 (BuShips), andSF 013- 10-03, Task 1804 (BuShips)

M;Lnuscript submitted March 9, 1961.

Best Available Copyii

NAVY HIGH-IMPACT SHOCK MACHINES

FOR LIGHTWEIGHT AND MEDIUMWEIGHT EQUIPMENT

INTRODUCTION

Object

This report will consoliziate information contained in previous publications, many ofwhich are out-of-print, relating to the characteristics and use of Navy HI (Hfigh-Impact)class shock machines. In addition the report will present reCent views relating to shocktests and test procedures.

There are presently two principal classes of Navy HI shock machines: the HIShock Machine for Lightweight Equipment (1-5) and the HI Shock Machine for Medium-weight Equipment (5-7). Equipment for use on naval ships is classified as lightweightif its weight does not exceed 250 Ib, and as mediumweight if the weight is between 250 andabout 5000 lb. These two classes of machines were designed primarily to siniulate shol;ks

probable on shipboard. Their characteristics and performances will be considered indetail. Other shock machines, such as the Shock Machine for Electronic Devices (8.9),the JAN-S-44 shock machine (10), various air guns (11-13). and drop-tables (11, 14-16)are also used by the Navy in common with other services, but will not be specificallyconsidered here.',

Early History

Prior to the early stages of the Scond World War the major causes of shock to equip-ment aboard ships were direct hits by enemy shells, torpedoes. and the firing of the ships'own guns. It was then generally conceded that the only practical protection of equipmentagainst enemy action was to mount the equipment as far as possible from the hull platingand to use as much armor as possible. A '3 ft-lb" (17), a "250 ft-lb" (18), and a combina-tion roll. shock, and vibration machine (19) were developedi during thi: period to simulateship environments caused by the action of its own machinery and ,rvuua. e and to generallyimprove equipment reliability.

During the Second World War the use of underwater mines that exploded some distancefrom a ship often resulted in little structural damage to the ship but caused considerableshock damage to equipment in the ship. This was in consequence to the large area of thehull that was exposed to the undervatrr pressure pulse. In addition, the greater quantityof, and the greater rvliance on, electronic and other complex equipment for the operationand eontfrol of the functions o, a ship required greater reliability for this equipment.

In 1939 the British developed a shock testing machine for lightwevight equipment thatprcxiuced damage to items under test that was similar to that caused by shipboard shock.The U.S. Navy shock machine for lightweight equipment was then designed similar to thatof the British. and the first such machine was built in 1940. In order to perform shock

"*hMort.C complete ger. ral des criptions of slaock machinles arc giit'n in Chapter lo of

R, cf. 11 and in R tf. IZ.

Best Available Copy

2 NAVAL RESEARCH LABORATORY

tests oit heavier equipment the shock testing.1machine for mediumnweight equipment wasdesigned and built in 1942. The shock outputs of the two classes of mnachi~nes were suchas to have about tile same maxi mumi values of accelerntion, velocity, and displacement foran equivalent shock condition.

EQUIVALENCE OF SHIPBOARD SHOCK ANDITS LABORATORY SIMULATION

One of the mo:st characteristic features of shock motions is their infinite variety. Itis neither desirable nor practical to construct a shock machine which produces a shockmotion equivalent to that of a given field condition. Rather thie Shck moltionI genleratedshould have a damiag~e potential at least as great as any probable field shock for whichprotection is required.- The shock machine, thereorint egndt sml te ,iefield condition; and the question frequently posed.. as to how accurate this simulation is,cannot be given a sensible answvr.

Nevertheless it is presently the objective of shock tests to provide -an accurate simiu-lation of field conditions -. ý not a giveii field condition,' but a shock motion that p)ossessesthe important characteristics of all probable field shock motions. This objective requiresthat sufficient field data be obtained so that they canl be treated in a statistical manner.The field data must be analyzed so that their damage potentials can bl. assessed. Anienvelope of all of the values of shock intensity. or their damiage potentials, obtained,( fromfield measurements is drawn. A shock machine* must then he devised which will provide

Ishock miotion repjresented hyMo envelope. This simla)llted( shock mnotiton thenl has, forthe methods of analyses used, a damage potential equal, at least, to the maximumi valuesencuuntered under any probable field condition for which protection is required.

This statistical approach, at first Lglance. appears straightforward and valid. And soit is for equipment that is relatively light. However, it is knuwn that equipment reacts tofoundation motions inl a manner similar to dynamic vibration absorbers (20), and that thleequipment will reduce thle amplitude of frequency components of foundation motions thiatare equal to "fixed-base" 7 natural frequencies of the equipment (21-23). It is also knownthat thle frequenicy components of shock motions that are the most claniaging' to an equip-mernt are those that correspond to the equipment fixed-base natural frequencies. Thus inla statistical study of damage spectra (shock spec ti-a), where an envelope of the rmximuzn,11shock spectra vailw's ol all appropriate field shock motions are normally' used to indicatethe sp('ctria of a su itab~le simulated shock Inot ionl thle sinmu it di motion would be overseverefor relatively heavy I equipment. For inl the, Ftatistical approach the envelope is determinedhy the maximumi values. But equipmient ron,:,tions causev values to be systematically at111(1r miniimu in t only the frequencies that are important for damage considerations. Thestatistical ap~pro~ach o ftenl results inl impossibly higgh design requirements which may be Intorder of magnitude above the true requi rement s. Such fa 'ctors as these indicate that the

S Oenefi shock testing is still inl an youthful stag' and that considecrable judgment maybe lif-cv ssa rv inl formulating tests, and that consider-able changes may be expected in shockmlachlici s and proc -du res in the fu turc

11W ll;itu ratI f ' i t'., (i f l t-e veqc~ijj:llcri 'o..1:1 tI1W f-Aind.ttiOl I0 whiCh it iS ;lttadlicd isI :Fiiiit.'I ' rivli;n 1lce;lvV i.' 0w fouIncidationi is fixeud, or1 docs 11I1, Iliove.

!Ai .rIl-11111*.Zlt I.,n- ;dv .I .eivy v-Iou its offect ivi- weight lý sufficienrt it) cono-iderablyIii 1h-i v h. I shock toutlotiols. 't'hi. is clcp-i~ldert 11polr filh- rvlative uIass, of the. equipijineuut

1!.' IX.o;?ltiatiocil ;1.s ),.e.,I 's 11pIORI heI. iltvt rn'ilu (l.1iflhiil. andl thte zIlodtes of vibrattion of the.it uuiutv I). 111 '-t hat i.1 t, 111 ili % lIl S Ics H a t1 0 1) mgt so et1inic S t

Best Available Cop

NAVAL RESEARCH LABORATORY3

ANALYSES OF SHOCK MOTIONS

Some examples of commonly idealized shock motions, and a complex Shock actuallyexperienced, are shown in Fig. 1. It is difficult to use such information as is given "orcomplex shock, it is even difficult to tell when one type of shock is more severe thananother. In order that the damage potential of one shock motion can be compared withthat of another some quantitative means of measuring the damage potential must bedevised. Various means have been used but, perhaps, the following predominate: shockspectra, velocity- shock, and simple shock pulses.

(a) (b) C

0 00

0 0 ^AV Time

0i; I Difeen tye ofsokmtos

(a) eloity-shok, ste-veocit chnge(b afpro ieaceeainCls;

0c atohaclrto lle dcrpe

shc0 ae h Isokrahnspo

system of d ifferent taualfepeces.o hc mto

The mxmmrsossothal-erio elmnsie cclrto pus.to(ca sawtoothsaccelerftihn palse;plottcomplex

-hcspcrmothtshock. waefh Isokmcintpo

Shode comlexrtrks

FiA Z-series of masnglc-dgrsyse-o-feso (snl-ereo-reoi rsml ytm)ish)nI i.2 ,ihsystems ofs different natural frequencies. th t- to t •hy.oeThfeqmaeimu ranespofnteest of c thstee elements rctinl atofcitcldal

hock Thectarem alf atahed thok acmo uwihsmdetprakofashoc moio

HIMiwi isi tog 2. allyel as yto it hs da ifferpentl nAtcural frepreenvsothnt tileter they cover

ec pi nse i f (-: eli cif tl-i systenins to the shock mot ion` expressed a4aflunctionl of thlell:ltto rýc frocilenc es iiý til systemls, ;(; thll damlage potential of that shock to thel series

!) *11pi !-,VýJt' ii" This 'ccc (-1: caIc I "s oc "I W I'l t 1 Th us o Shock simeetr~i,

Best Avai~lable Copyý


besides having other valuable applications, has uecome an acceptable means of expressingin a quantitative and standard manner the intensity and nature of the damage potential* ofshock motions (24). Shock spectra can be expressed as the maximum relative displacementof the masses (Fig. 2) with respect to their base, as the maximum absolute values of theaccelerations of the masses, or as pseudo maximum relative velocities. If the maxiurumrelative displacement response is given as X, then the displacement, velocity, and accel-eration spectra (or responses) have the relation X:2nfX:(2irf) 2 X. This relation is onlyapproximate and applies accurately only if the damping is negligibly small. In specifyinga shock spectrum the amount of damping should be given; if it is not, then the amount isassumed to be zero. The spectra arc usually plotted as the maximum values irrespectiveof the sign of the response. However, somewhat more information is available if both thepositive and negative maximum values of responses are plotted to provide positive andnegative shock spectra.

Examples of shock spectra are shown in Figs. 3 through 6. In Fig. 3 the accelerationof the shock motion are shown by the inset curve. The ordinate represents accelerationin units of gravity. The maximum accelerations experienced by the masses (Fig. 2) forseveral given amounts of damping, when the base is subjected to the accelerations shownin the inset, are shown by the three principal curves. It is thus noted that the shock spectrado not represent the shock motion, rather they show what the shock motion does to astandard set of simple systems. The shock spectra of the simple pulses of Fig. 1 areshown in Figs. 4, 5. ond 6.

The four sets of coordinates of Fig. 4 are of particular interest. The three ordinatevalues, velocity, acceleration, and displacement, represent amplitudes of sinusoids andso are all ritpentevnt upon 9ach other. If the velocity amplitude, V, is assigned, then thedisplacement amplitude is X - V/2nf, and the avccelrration amplitude is 2,fV. These willappear as straight ines with 45-degree positive and negative slopes respectively for thempthod of plotting employed. This t -o'rdinate system is not only a covenient way tosimultaneously represent the three types of sh'cck spe,'tra, but also represents directlythe shock spcctra of velocity-shock. Velocity-shock is discussed in the following section.The numerical value of it velocity-shock spectra is equal to the magnitude of a step-velocitychange (for conditions of negligible damping) and is independent of frequency. The corre-sponding acceleration and displacement shock spectra are given by the sloping coordinates.

Figure 5 illustrates shtock spectra for a half-period Since pulse. A curve representingthe oaxmlluilul positive valuu:c would be the positive shock spectrum, its previously defined,and a curve representing the maximum negative values would be the negative shock spec-trum. These are shown respectively by the curve lnbeled "primary" and by the curvebelow the zero axis. However, as illustrated, the responses during the time of the pulseare Feparated from those following this time and are respectively identified as "primary"and 'residual." In Fig. 6 the same separation is provided for the shock spectra of asawtooth pulse. A primary shock-spectrum is defined as the maximum response of thesimple systems (Fig. 2) during the time ot the shock. The residual shock spectrum isthe maximum response after the completion of the shock. Positive and negative shockspectra may exist for both primary and residual shock spectra.

Figure 6 illustrates why a pulse of a sawtooth shape has become popular for shockspecifications. Its positive and negative shock spectra are equal and rise smoothly to amaximum value, after which they remain close to a constant value. All symmetricalpulses, on the other hand, have different pisitive and negative spectra with the residualspectra periodically becoming very small.

'The shock spectra rcprcscnt damnag, potciiti.d for til',dle systcms. The relative valuesmay be diff,.rent fczr nonlinear atnd mnore complicated ,,ystems.


60'

PIAUI C5P'IKr0go L_ I '.* - _

io 60 90 '20 '30 'Na 2,0 ?4t Ai t 0 ISO MD t 9 AliuihUmil MOIuitci 1gFtI

Fig. 3 - Aert-le ration shock spet rmin for shoe k

ruotitln shown by in.set figurv. Both uIacip'd awxilundam ipedc elemein tots are c onside red. (AftecrEllett (25).)

'J%

4-I

'4-k

'2

C; ~~100 0(

RE Out NCl IP

Fig. 4 -Shock spectra fori velocity-shock. Thevelocity shock-spectra are numerically eq ualIto) the Values of the velocity- shock (the magni-

tude of step-change of velocity). The displace-

ment shock-spectra are given by the lines of45-degree-positive slope und the acceleration

shock-spectra ail! ~i ven by the lines of 45-

degree-negative slope.


~;' ~ 1 -~ - F~ Fii!. 5 - Shock sIp.ct ra for a half-

pe io-sine-wave nrccleration-pulse..'I'le pulse (inset) has an amplitude of

r~/-- and a dluration of T/2. 'he max-PC imor response is

0 RM* rjhe primaryI /-2--~ shock s 1) c c t rit tii is !he miaximuni

I j cc tsp on si.do ring [ihe pulse i nt erval andI .1. s ,otiv c (t.'it: sanie di rectionj as the

iri-~). heresidual spI~tra havecquj a II' poiv and( negative LIOlbS

Fit4.I 1, Shu k ). 1r o 6 5 t!.I t o

(Aft"' .11.,Iil.- t

IiiQL-t N, -

Velocity-Shock

The most important characteristic of many types of shock motions can be expressedats U step-change in velocity (27). This concept is sufficiently accurate for a large classof shock motions and, outside ot purely static considerations, is the most simple in itspractical application. The velocity-shock is expressed quantitatively as equal to themagnitude of a step-change of velocity of the excitation as shown by Fig. l(a). Frequentlyan equivalenlt velocity-shock is taken ats the maximum c'hange of velocity of the center-of-mass of anl equipment subjected to a shock. It is to be noted that the velocity-shlock spec-trum of a velocity-shock (no damping), ats shown on Fig. 4, is numerically everywhereequal to the value of the velocity-shock.

The use of a sin-gle number to describe the intensity of shock over given ranlges oftrequeracies can be useful and sufficiently accurate for many applications. The termvelocity-shock :.s one such number. Obviously under real conditions, where displacementsand acc'--rations remain finite, the spectra for velocity-shock involving a finite step-timemust d(-,(ease below and above certain frequency limits. Shock spectra, shown later inFigs. 28-29 and 42-43, give a middle range of frequencies where the velocity-shock spectrais about constant. *The single numbers corresponding to these valuc~s can be givent as theequivalent veloc tty- shocKs for these frequency ranges.

NAVAL RESEARCH LABORATORY 7

For some lower frequency range one observes a constant-displacement shock spectra.The value of this displacement is equal to the value of "displacement shock" over thatfrequency range, where displacement-shock is defincd as a step-change of position and itsmagnitude is equal to the magnitude of thc step-change. The shock spectrum in this regionis cquivalent to the disolacement-shock, and is equal to the maximum displacement involvedin the shock motion.

Above a certain frequency the acceleration shock spectra becomes constant. For thisfrequency range tile peak, or maximum, value of acceleration present in the shock e-citationis equal to the acceleration shock spectra. This condition exists when the highest ,6 nifi-cant fecquencies associated with the shock motion are low compared with the shock-spectrafrequencies considered. Thus the accelerations can be considered as equivalent to staticvalues.*

Simple Shock Pulses

When the time for the velocity change associated with a shock motion cannot be con-sidered short (compared with periods of significant modes of vibration of an item subjectedto the shock), then it is sometimes sufficient to construct a shock motion of mathematicallysimple shape which will be equivalent in important aspects to some shock motion of interest.Some of these pulses are shown in Fig. 1. Responses of simple systems, to a variety of typeof pulses can be found in Ref. 28. An application of this type is to obtain a simple shockpulse which has a shock spectrum approximately equal to an envelope of maximum valuesof a statistically significant quantity of shock spectra obtained from field measurements.A shock machine, buil' to provide this pulse, would then provide a simulation of the damagepotential of the field environment. In addition the responses of an item to the pulse excita-tion can be theoretically determined more easily. As has been mentioned, there are validobjections to the statistical procedure involved ill establishing the shock spectrum, becauseof the effects of equipment reactions in modifying the shock motions.

Miscellaneous

It is acceptable under certain conditions (29) to give maximum values of accelerationsor velocities and associated frequencies which one observes in a shock motion. Fourierintegral and series techniques have also had limited use. The Fourier integral methodhas considerable potential value.

SPECIFYING A SHOCK TEST

Shock tests cn. be specitied by three methods (12): First, a shock motion can bespecified. A shock test then consists of causing t,.Ž points of attachment of the itemunder tust to par'take of this motion. Second, a shock spectrum can be specified. Ashock test hl.en consists of causing the points of attachment of an item under test topartake of a motion that has this spectrum. Third. a shock machine can be specifiedtogether with a procedure for its operation. A shock test then consists of mounting anite,01 under test to the .iachine in a prescribed manner and of operating the machineaccording to the given proce('dIu 'e.

W:I[t ,wnoul tt.- l l t.r, 'l, t ilicl<t:le "OCcccJerat ion-shtc)c1" for thie upper range of frequtencies,wle it, ;i ('elt'rcitora i-shck would be dcfine'd is t0c iagnitude of a st, p-(han.,L of acccicra-ti,,ti; h(,wV ril, l(- .hIck sij•ctrulnm of all ;lcccleratiot n-shock would he equal to twice thev dlue of (,ccelk. ratiom-shock , wihich is ,not analotgotis to the similar situation for displace-11, rtt ;nd velOt it ,.

Best Available Copy


The first and second methods of giving specifications are somewhat similar, althoughthe second places more burden on the test engineer, as he may be required to devise ashock motion corresponding to a given shock spectrum. However both methods areimpractical of achievement unless the items under test are perfectly rigid or relativelylight. The cause of the difficulty is the reaction of the load on the test machine. Thisreaction causes the applied shock motions to become dependent upon the nature of theequipment under test, so that, unless large tolerances are permitted, the test cannotpractically bz. made as prescribed. The only practical solution to this problem is toconsider that the numerical values of shock motions, or spectra, given in the SPec.fic:-ttio.1be considered as nominal values. They should be complied with for rigid loads rigidlyattached to the machine, and calibrations of the machine should be made to assure thatthis is so. Tests of real equipments should then be made according to prescribed pro-cedures with no further concern as to the specified motions or spectra. However. careshould be taken that sustained natural frequencies, not typical of field conditions, are notintroduced by the shock machine and the mounting arrangements. This procedure ispreferred, even though it would be possible to generate tile specified motions, as thisprocedure may prevent the overtesting which would result if prescribed shock motionswere maintained in spite of equipment reactions. If this procedure is followed, the shockmachine should have a mechanical im-pedance, as seen bv the equipment, at least as greatas that of the structure to 'Which the equipment will eventually be attached.

The third Int,tlod of specifying shock tests requires that thc agency rispons ible fortile test provide a ia;-chine design which, when the machine is built and 11sed in a prescribedmanner. will provide a suitable shock motion. This reduces the trials and tribulations or

the test engineer to relatively small proportions. The Navy HI machines are in this cate-gory. The design of the HI Shock Machine for Lightweight Equipment has been standard-ized (4) by the American Standards Association. Complete working drawings of themachine together with operating instructions are available from this source. Becauseof the small number of mediunmweight machines in existence, it has been subjected only

to Navy standards.

When a machine is specified for a shock test (the third method) it is the "esponsibilityof those who spe-cIfy the test to provide information as to its shock motions and theirspectra. These are normally given only for rigid loads rigidly attached to the shockmachine.

NECESSITY FOR PRECISION AND ACCURACYFOR TESTS ON SHOCK MACHINES

Great accuracy can seldom be justified for shock tests on the basis of field informa-tion. However three factors require that shock machines be constructed so that they canrep)roduce shock conditions with considerable precision: (a) Shock tests may be legalrequirements for the acceptance of an equipment. Whether a test has been performedarcording to specifications within acceptable limits is of great concern because it involveswhether or rn It the (qavpmnent is of acceptable quality for centract fulfillment and pa yment.(0) In developmental work it is not possible to tell whether or not a significant improve-ment has been made unless the magnitude and nature of the shock can be accuratelyrepeated. (c) A shock machine at one location should be able to provide, within reasonable

hi)ler-ances, tne san test to a given equipment as would be provided by the same type ofshock machine at another location.

The mootion nomi,' of shock machines (that do not suffer plastic deformation ofse 01 ipernianent j;;m s) should not differ by more than 5 percent for frequency componentsbelow about 200 cjIjs. Greater variation can be expected in thil shock-spectra peak values,as these values are determined by damping losses in the ma'chine. As the damping dependslargely on bolt tightness .ad friction between surfaces this factor may show considerable

Best Available Copy


variations. Considerable variations mnay exist at higher frequencies. If machine parts(such as the anvil of the HI Shock Machine for Lightweight Equipment) graduilly deformwith use, then greater variatigo..s in performance can be expected. The HI Shock Ma'chinefor Med juminwight Equipmtnit suffers no ap~preciab!e perinnnent deformation (if its part.

THE NAVY HI SHOCK MACHINE FOR

LIGHTWEIGHT EQUIPMENT

Description

Tile III Shock Machine for Lightweight Equipment (Fig. 7) c(onsists of a welded framleof staniard steel sections, two hamimers, one, of which drops vertically and the otherswings in a vertical arc, and an anvil plate which may be placed in either of two positions.The combination of two hammners and two anvil-plate positions permits blows to bedelivered in each of thiree mutually perpendicular directions without renmount ing the test

equipment. Each hammeir weighs 400 lb) and may be raised to a maximum height of 5 ftabove its impact position. ito de-liver am maximiUm of 2000 ft-11i of energy at impact.

The anvil plate consists of a1 steel, plate mel(asUring~ 34 *, 48 ",5 '8Bin., reinforced arrossits back surface by f-beam stiffvnvrs. Steel shioek-pads are we ldt-d to the lop and sideed(ge~s amnd at tilt' Center. of tht, back facc ovicr thc sli fteiters ait fih p~oiints of hamimer impavt.For back and top blows the anvil plate is positiomwcl across thev main framne and rests an .1pair of enclosed helival sprintvs 11 is teid: in'. al vIerticail position bV .1 sI t of sjpriiig..,and throungh bollts h a riang againist ilhe oaim i o1priight i Wash. r s aindI sparers pr rvent Iiind ismiof llthe.u1ii% m platsdun top tilow~s. For ctg.' blows. tlhe anivil plate is rotate-d 90 (legrees,aroutnd a vi'r-tc.1I ,I~is andi is S11111)[1-11,1 Ity rdIlers bcarinig .mgaimnsl steel traicks. 11 ispoisit io~ned by a set if sp ir ins ni' iunt ed fill thlt forwa rd :;iPpl iii -hbrace edge. lcct(gc' blowsare elvee I~y t e1wii o i. r 1ninwr aalis th .1 i 11v a iI - Ia te vildg Inl vach of tlathree dIirect ions of hammeir impact the, anvil plate const raiining spring% are( adljusi'te topermit 1.5 in. (if forw~ar nclmoion against the( spirinigs before bottoming occurs againstli mit stopls. Rebound( springis foar kairk and edge blows are also provided with limit stops,although these springs reach thuir solid heighrt before the hirSits are reached. There areno reboundl spriiigs for top blows. thin' naxinmorn spring vxt.'nsion being govrrtiid by acaptive bolt.

Several standardized mounting plates have been devised which simulate the mountingconditions aboard ship. These are inte-rposedl Latwevin the anvil p)late and the( test vquip-mvnt and pirovide a de~greev of flexibiility and isolat ion to the shock motions in i1 mannersimilar to normal shipboard bulkhe'ads and decrks. Two mounting p~lates are used pre-dominantly for specification shock tests (4.5): the, 4A p~late for bulkhead-miounted equip-mernt and the. shelf mounting plate for platform-ni-ountenc equipmenti. The formvr deri.'edits name from its figure number in- shock-test specifications (5at) and is a flat steel plate27 x 34 y' 1/2 in., while the- latter is a similar plate to which a reinforced shelf has beenwplded. Reinforced 4-in., 13.8-1lb)c'ar-btiilding channels along the vertical edges snacethe mounting plates away from the anvil plate, Holes are drilled in tl,.e mounting plate'sas required to miount the test equipmenvt cenitrally. The plaite is discarded when the holesfronn previous tests becomne too numerous.

Loading Arrangements for Reported Character ist ic's

Reference 26 gives the results of an 4nvestigation of mechanical shock oil the( machine.Detailed drawings of the load apparatus used are given in this reference. Thle total wrightcapacity of the machine was covered by two separate load assemblies, the lighter covering

10 NAVAL RESEA'CH LABORATORY

411

; g

I'I

I.:

F ig -::. H I-, . = S h o c k .... Ef::m

'V. ), is'' i* .. . . .. ,

Fi •, n frL ,itw ib £ .q!i pnhcl "

"l':invil 1 ilate is orienited f'or itl edge- blow.


the range up to 200 lb. and the heavier rorn 200 to 400 lb.* A rugged welded framre, com-prised the base assembly in the light range to which additional steel plates were bolted toincreas'- the load weight in small incrt-ments of approximately 25 to 50 lb. In the hecavierrange the load consisted of two sections oif solid steel plating, one weighing approximatuly200 lb and the other 125 16i. These sections could be used singly or inl pairs to alter theload weight. Each load was drilled to accommnodate the &lame rectanigular mounting boltpattern so that the load distr~ibution onl the mounting plate remained the same for all loads.Two-inch cylindrical pedestals spaced thlt load weight away fromn the mounting plate toprevent binding. Examples of the loading arrangemients are shown in Figs. 8 and 9. Theloads and their mounting supports ca;n be considered ais relatively rigid compared withthe mounting plates (4A or shelf p~late) of the shock miachint.

Instrumentation and Measure mnents

Measurements %%e rc made- of veloc ityv and acc eleration as a function of time, and ofshuck spectra; ;IS indicaIted by ;I mlult ifrequeny v reed gage. Instruments for these incas-uremients w-rie bol1ted directly to the load assemnbly and oriented to mecasure, the shocknIUtMIuS inl lith' dIrect iOnf tHeP h1.1in1me11 impact. A second ac t.h lronivter was maintainedinl a fixed Position o1) echitfl c)U11iitin platu, anid pr-ovided an indication of the magnitude ofShock mnot ions at these specific hwa ;tions.

"ricl1-1f~r LZI vdgv4 bilow.

':Norrnal speCCIfication8 11111t the lUad~ iuJ 250) Ill.


"'" • SHEFLr PLATE -

Flu. Sh il jdtt' ' -Il, "" t',ii I: " ad. 'Dw i, rX i ci l't,

With "hc possiblh, vxc'ptlom of the reed gagevs, the istruonwilts were standard typeswill.'y, illrit(t.,l'istll"& , and limil.dti,,s ar, wvill kn'iwn and which havw pro'ven slatisfactoryfor shock neasure mnnts. InstlrUllmet locat ions lld d(,,tai.Is of their mounting adaptorsmay be seen ir Fivs. 8 and 9. Standard 300-, 1000-, and 5000-cps low-pass filters.inncorporatt'd in theill r''h iet ,r cpreamnl ifiers, liminited thie upper frequency responseof tile acceleration signals by re-moving acce(lerometer resonances and the higher fre-quency acceleration 'niptmnents which are of little importance, i.e., have little damagingvalue, vet predominate in the unfilterud records. The filtered output signals were displayedsimultaneously on a Inuitir'hannlI Cathode- ray osc'illograph and recorded photographicallyby a inuving-filin ca •iera as shown in Fig. 10. Sufficient recording channels were availableto permit recording each accelerometer Eignal on two channels at the same time, usingdifferent sets of filters. Thus, the 1000-cps filtered record was recorded for every blow,while the paralleled channel alternated between a 300- and a 5000-cps filter. The velocitypickup signal was unfiltered and included an ac-ceptable frequency range from about 6 to2000 cps.

Aucelerations, Veloci and Displacements - The shock-motion waveforms, producedby the Shock Machine for Lightweight Equipment exhibit the same general characteristicsfor different heights of hammer blow delivered to a particular load arrangement in a givendirection but greatlV modify their c haracteristics with changes in direction of blow, loadweight, load orientation, and mounting plates. Previous history of the anvil and mountingplates also affect the shock waveforms, but to a much lesser exL.nt, by its influence on theplate stiffnesses and shape. These change because of work hardening .znd plasticdeformation.

The acceleration traces for both the Iliad and mounting plates are shown on the typic9ltest record of F-g. 10. In general, the maximum value of acceleration occurs shortly afterimpact; it is folowed by irreguiar vibratory modes. Characteristics of the motion remainsimilar, except for amplitude, if only the height of hammer drop is varied. The magnitudeand frequencies of the accelerations which follow the maximum value are greatly influenced


*4p 3C(! CPS)

I i t f 1 i

X.~ 'C .. I,. I Altl oI-;, ., -t t:i 1- 1A Ait 9000 1o

, ~ ~ ~ ~ ~ ~ ~ I L.a1 ;.' P.M -1 il UIVI-. ..t'I -'-j ~ t -c

by a1 large niumbiler of factors suchi as bol Itightness, mass dist ribution. and energry dlissi -pation. Slight variations in any of these factors produce large changes in the waveformiafter the first pulse. Records taken using the 5000-ejps filter are worst in this respect,since the high freqjuenc y co11IMIPantS are most easily changed.

With in~formai-tion givenI aS tol thU geinera i waveform of the shock motions, it is sensibleto plot maximum values of -( celeration and velocity in order to provide comparativevalues for the shock motion. Figures 11 through 13 give maximum values of accelerationof the load and 4A plate and Figs. 14 and 15 give acceleration values when the shelf-platemounting arrangement is used. As the load was essentially rigid the location of thltaccelerometer onl the load was not of great importance. However, the value of the acce l-ecation of the 4A plate was strongly dependent uponl the accelerometer position. Theapparent erratic trends of 4A plate acceleration curves are caused by the changes o)fmode shapes of thc plate for different loads. For a complete description of results andthe factors involved see Ref. 2.

Maximum values of load velocity are plotted in Figs. 16 through 21. For any givenload these values increase approximately linearly with the hammer impact velocity, or.is the square rout of the hcight of drop. This approximation is also roughly true for themaximum values of acceleration.

The displacement- time motions of the load can be calculated from the velocity records.These motions have been plotted and are given in Fig. 22 for two different loads mountedon the 4A plate.


0 L• lC• .PS FIL T!•H,

•~~~ ~ 1 lX•t - 4

r t

H : .ýMl. OF ýAIAMtR DROP (FT 7 1i2

A UA ccPifRA T iF e tfi i A T*

,09 .02 5

400-A

S ... :.. . . . . . . . ..:.A r, a .

J' - * I-

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

S i),. .. ..... ... 1

HA ME -•M AC L T If ! . . .EC.) t

! . :2 . . . ... 'p

""ig I Iueo

an o' 0.......n 'b

400:..... i..........

.4 I

H EIGHT OF HAMMER• ROP VI ) • .

4 6 8 '0 '2 14 !6 '8HAIMMER - IMPQAC! VI" LOC',tY (F? ,SrI•

Fig•. 1i - Maxim~um accelerations of loadaned of 4A plate for back blows


400 -i-T

LOAD ACCELERATION'000 CPS FILTER

300

1 "I___ ' 1 .

I" 0 ; I. E!GHT OF HAMMER DROP (FT1 " '--- "

, . :, I , 3 I ,4

500 . • - "CL I I ! I n

4-PLATE ACCELERATION

.400

"LOAD (LB) RuN T0 • 21 2 ' ' 'X : -l .,• ; '

300 . ,45 4 . r

a 389 01 .

*20f

100 1o f t t I

SHEIGmT OF HAMMER DROP (FT)

0 . !S .! , 1,.1 l • •

4 a 8 0 I? !4 16 18HAMMFfl-IMPACT VFLOCITY (FT/SE• I

]"i:,. I,'! - laxil:1111 ;i-, of•rd~ m• , load,Iml ý)l" IA.\pli* fm-r b''Ilo)w,

200 LOAD ACCELERATION '

i(00 L-TP AFILTERATO

IOOO-CPS FILTER

S LOAD (LA) RUN -i!.. : ' • --

4II

- 00 ' 1i14 - E-I 1 F HA MRD O I F

Si '- iI-.

0

100- 0 CPSFITE

4 121 i O 1 24 IF- I-- i

TAMUFA- IMPACT V!LOCIrY (FTISEC)

Fig. 13 - Maxilmun acceicerations of loadand of 4A plate or top blows


IOAD ACCILCEATIOW I1000 O-CPS FILT!R -* -

0

-0 SHELFr ACC IA E LRO t "AIO

400 OO-CPS FILTER

3 00

192 1*

4 0 26 is. 6 '

ISI

a LOAD ACCELERATIOIII IZ. 1O000CPS FILTER,

100 vi

: -} 1 ~ H IGHY Of HAMMER DROP ITI )

IAMMR- IPAC VELCIT IFYSECFi0 1 aiu ceeain fla

andk of sdf lEatefr de lwsload on00 the sheLf plat


LOOD (LEI) RUN14 0 57 0I

_ •' •'• 4 I A •..

V~ -4

1 1 19 2 6

t, . I t 1 - ; .. - - !.

a tI

HI !HEGHT OF HAMMER DROP (FT)-- 11

26 8 0 I 14 16 18HAMMFR IMPACT vELOCITY •FT/SEC)

F ig 1" 1 Maxi-nilln Ioad vW',,citi•,s for bau-k bluwsr,,r 1),,,1• ,., t , % " plat,-

A 12 "1 2 ; I 2. 1v 4 4

1 192 611 389 '0

10

4

HMEIGHT OF HAMMER r)ROP IFT;

° , + T0 i y•i ~

2 4 6 : 10/ 4 14 16 to . . -

HA4MMER IM PACT VELOCITY IFTS:C)

Fig. 17 Maximum load velocities for edge biow,for loads on the 4A plate


I LOAD (L.83 RUN j I14 0 57

V'45 4 I i 1 ' 4

O•21 6 9; 6 i 4.. 4--'

9 8A I

'06, 2 ! 0 T 0 rI [ -• - i ! [-I --

OF ME DROP IFT)HEIGHT OF ME 5I

2 4 6 a 10 I2 4 I6 IsHAMMER -IMPACT VELOCITY (Ft/SEC)

Fit, 1 8 M;L'x1ilimnJ1II loa veoii3c'tCs for top blowsi

I 1. - ,4 on I-i--4 I u

LOA D (L B)RUN!I

I I Ig I

14 0 57 i

& :2.1 : 2 ' :l

A12 1 a I9

to I

i I I +

t t HEIGHT OF HAMMER DROP (FT) - - I

o0' I :, 3 4 5 ,

2 4 6 8 10 12 14 16 IsHAMMER - IMPACT VELOCITY (FT!SEC)

.i I ' -IQ klaxiiiwu lo ad vcl o .tit , for bark bloxwsfor 1tot 1 l):: Ihc ll lif t i o-. I ting pl ate


I-LOAD (LB) RUN f-I

1" ] * ,92 :4 -S, IIs

SI 17 .o0- tt!/ :-I- i l * l ! 1-

o .1 .

0

- , •i 0.+ +,A m P r o r )4 I

I I*HEIG-T Of t4AMMFP DROP (FT)

420 1 I4 16 l8HAMMER -IMPACT VELOCITY (FTISEC)

Fi '. M') - Maximum loaed vhocjtijes for v,'di blowsfor 1(1t.S o th0-le sltlf uimoutnting plattc

LOAD U181 RUN I14 0 o 7 I.

A 121 '2 I45 S 14 .

0 12 4 9 ;41 26' 5S

w v 38916 1 '

-I * I , ,

Ut

06

4 4 t

IEi'GHT OF HAMMER DROP (F T) 4I 3 4 5

2 4 6 a 10 12 14 16 IsHAMME A-IMPACT VE LOCI TY I FT/ SEC.)

Fig. 21 Maximum. load velocities for top blowsFrloads on the shelf mounting plate


• 0.I• ÷" • .... RACK

c,571 LLBI LOAD

0 10 20 30 40 0

MILL ISECOF•OS

1 .5 -l E D EO

389 L8LOAO

00 10 20 50 40 so

L11LISIGONDiS

i":g. 2,J- - ]i. 1 1 LcCli]vlt|-t " )111( motions (fromn intgra, tt'dVclot ity rucordb,) Of load on the 4 A pl ate rvr 5-ft back,I op, ,nad, ,..g- bluw:i

Shock Spectra - Results relating to shock spectra are principally contained in reportsby Dick (3) and Conrad (30). Shock spectra can be obtained directly fron reed gages.When this method is employed, only a small number of frequency values can be obtainedbecause of the difficulty of using maiqy iveds simultaneously. A more complicated but

preferable procedure is to record the appropriate acceleration or velocity signal on tapeand to analyze the recording in terms of shock spectra.

Shock spectra, for motions of the load in combination with the 4A plate and the shelfplate are given in Figs. 23 through 25. Figure 26 indicates how the response varies withhammer drop-height, and Fig. 27 illustrates an equivalent velocity-shock or step velocity-change. This would be the average slope (see Fig. 23) of an acceleration shock-spectrumcurve. Velocity-shock is useful for design calculations and provides a way of expressingshock intensity in terms of a single number.

Shock spectra for several different shock machines for several load conditions andhammer drop-heights have been plotted in Figs. 28 and 29 using the four-coordinatesystem. Ct.isiderable information is directly available from such a graph. For verylow frequencies the displacement shock spectra is asymptotic to the maximum displace-ment involved in the actual shock motion. For vei y high frequencies the accelerationshock spectra is asymptotic to the maximum acceleration recorded for the shock. Forintermediate frequenuies, if peaks caused by resonances are neglected, the velocity-shock

spectra can be taken as the best value of velocity-shock. Peaks in the shock spectrarepresent sustained vibrations in the shock motion.


2000[

BACK BLOWTOP BLOW

-.. .. END BLOW

CL0~

4 OCO

_j

a-.

4 o~LL

0 200 400

NAIURAL fREQUENCY (CIPS)

F~i•j, I) - ho,ki• bputctra ,,i - i'no•ti•,l,,ý of it rigidi 57-lb lo)adM)i th,. 'IA I)I lttt 101- ')-ft bi' 0tt)j), ;+tcll ed.Ige bloWJ

S00

__ BACK SLOW

TOP BLOW/S~END BLOW

I-- iO0/

Lu

I---

-J

200 400

NATURAL FREQUENCY (CPS)

Fig. 24 - Shock spectra for motions of a rigid 5Z7-lb loado; the 4A plate for 5-ft back. top, and edge blows


BACK BLOW

400-- TOP SLOW-----... BOW-END BLOW "

200-

"- - -I•1I POUND LOAD

I--

tr

U,

0 400

1- 200---

261 POUND LOAD

-LJ_1

u- 400

I -- - -389 POUNO OAG

0 200 400NATURA; FREOUFNCY (CPS)

li.g 2. - Shock sp'ctra fur mnotions of rigid 121-, 261-,and 38

9-lb1 lojitd on1 the s;ht'lf plate for 5-ft Lk, 1,2p,

and b'dl* blow.'.

1500

l,,-

5OOC FOOT BLOW

2

z\

200 0<

I-1

0 200 400NATURAL FREQUENCY (Cps,

Fig. 26 - Shock spectra for rigid "!I -lb loads on the

4A plate for varic,us heights of hammer drop


?C

H' tA(K ti OW, 4A PL AIEt I O BLOW, 4£A PLATE

t] ND fiLOW. 4A PLATE

B HACK BLOW, SHEIF PLATE

A TOI RHOW, SHELF Pt ATN

D U t ND BLOW, SHELF PLATE

I.-,(A

T I S 6\

II

5) * *p 6 8

Fip. 2 7 V,- Viu c o f v fI uitiy - sIh ic k f L)r 5 -ft bl u v sas At function of load weight


000

M''4

/

II: ' - - - -.l 1

Fig. *8 Shc sc f'a -

* - IT •IROP

'C..................................... : ......100I 0 '0C•CopS

8A;:K iJ;uWS- 57

'b LOAD

Fig. 28 - Shock spectra for m~otionls of a rigid

mass of 57 lb attached to the 4A plate of theHI Shock Machine for Lig:t•wcight Equipmenit.An eqmuvalent velocity-shock (V.S.) for fre-quencies between 10) and 40 cps is indicated.


W, t- J [JT1

'M I-

.. / I..* I

I ' , 'I :

14-A' K ITT OW',,A!

, - Shock spectlra for ni.t ons of a rigid

111sF of 2".7 lb Tn: the, 4A plahtl, (if the III SiockMachine for I.i _.htweight EqUipivi-tt

As a shock spectrum is by definition the maximum relative displacements, or themaximum accelerations, experienced by the masses of single-degree-of-freedom systemssubjected to the shock motions, these values ('an in many cases be taken directly torepresent relative displacements across flexible mounts and accelerations of itemssupportedon flexible mounts.

The points plotted in Figs. 28 and 29 represent different machines. They showconsiderable spread in the high frequency end of the spectrum. These differences arein part caused by changes resulting from cold-working and deformation of the anvil. Theshocks tend to become measurably more severe as the hammer-anvil contact areaincreases and the anvil work-hardens. This coniinues until cracks form in weld areasor the deformation becomes emcessive. When the deformation exceeds a prescribedvalue, defined in Ref. 4, the anvil is repaired or replaced.


THE NAVY HI SHOCK MACHINE FORMEDIUMWEIGHT EQUIPMENT

Debuription

The Hi Shock Machine for Mediumweight Equipment, Fig. 30, consists of a 3000-lbhammer which swings through an angle greater than 180 degrees and strikes an anvil.The anvil thereby suddenly acquires a velocity ini the upward direction. The anvil, whichwveighs about 4000 lb, can be placed in either of two vertical positions. A mnaximumivertical travel of 3 in. is permitted by hold-down bolts from the lower position, and 1 .5 in.from the upper position. Special arrangements may be used to permit other travel dis-tances. The hold-down bolts cause a sudden reversal of the upward motion of the anvil.The machine is mounted on it heavy concrete block which is isolate~d fromn surroundingareas by coiled-spring supports. This provents thev shock from being transmitted to thesurrounding area.

Equipment under test is not attached directly to the anvil table, hut is attached to aset of channels that are separated fromt thu tahle by spavers. This is illustrated in Fig. 31.The number of channels usfcI is a function of the weight ~), the equipment and is given inspecifications (5) for use of the i-machine. Trhe an1vil, cnelsystem, and equipment, soassembled can aipproximately bo represented by Fig. 32, where, M, is the anvil, the springand d~ashpot are the( c banoel1vs, ;Id M, is thev equipme(nt. The ha 01 mer i mpact is applied atthle bottom cenlter of M-

T.*. P,',

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

Fiv 4W) 1.'* nsiokfrMdii- egt q~ne Tepsto

of ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~I Ih a-nu iteis io mpc ssonb h otdlns


4116: 7z~I..Q

FI - ýI I [I ýi ... MakCI11IIC. Itir M1vditkin ijgh1

Fig. 3Z Scherit ic rupi-csetitt;lt o11 f the III -

Shock Machii r for Mediamtwei j.ht Ecjuiptnenlt.M rc;- rec cat s the loadl, M.~ the invil tablL, I V,

and F the~~'US


- As the shock motion is only in a- vertical direction, an alternative mount-

in%ý arrangement is sometimes used. Twobulkheads and a deck section are assembledtogether, as shown in Fig. 33, to form acorner with three mutually perpcndicularP surfaces. The junction of the two bulk-heads is made to incline 30 degrees fromthe vertical when the fixture is mountedoin the shock machine. The fixture thusprovides shock motions along all principalaxes of an item of equipment under testand is convenient for items that requirebulkhead supports.

Mleasurements

Reed gages and velocity and accel-eration pickups were mounted on the loadsMnd the anivil table to provide informationfor the determination of shock spectra andshock motions. Time accelferometers site-oats were- filtered by 300-, 1000-, or

- i00O- cps lox.- pass fi ltii-tZ III addition a1W-M W i silt (if bonded-wvire rvsisfanc(' straijn

gages were conmente~d to one, of the load-suwppoirtifig feet. This prov ided infornia -tion fill determin cing the force exerted on

Ii'. - hidy-dc ill Iw~ld hillk- thce load by th( channels.hrad fixturei. All itiem of cquiii!ill'.ili I

11101ii.tt-L! ill thl' fiXIttire .- iid is re-aiy A typical set of records for 4420-lbfor test. load and a 2.75- ft ha minir drop is shown

in Fig. 34. A 1000-cps low-pass filterwas used for the acceleration records. Itcan be observed that the channels provide

considerable isolation for th'e high-frequency comnponents, but introduce a strong domrinantfrequency of about 70 cps. This correspomds to the natural frequency of the single-degree-of-freedomn system shown in Fig. 32. The isolated acceleration pulse on the top curve iscaused by the sudden arresting of the upward motion of the table, by hold-down bolts, aftera table travel of 1.5 in. has been completed.

00 200g -Xtable ('000) 20-

SR4 :0 1b

X load (1000) :: duX load si SC

* 6 /see,Xtoble G/,gc~

2.75 F T HAMMER DROP -i-5 IN. TABLE TRAVELFig.. 341 - Pecordings 'f picaJ 'hock~ niotcoz'.. Bitntin -~i, o' iuc r. are ýpacea afI millisecond inturva.1s. 'l'op i rac.e a rcluratjoc of tdii,je. luo-;cps low-pa~ssfilter: second i ris( : for, c exe rti'- ! oi od: thiri r,i' aCCo'ier_,i.-,e -~f 1,-1~-I 000-cpti low-pabs filter; fourth and fifth t rac-s: velocity of load and of table.


As a result of the reversal of velocity of thle anvil table, when it recahes the lim it ofits upward motion, the load experiences an additional sudden shock. The damage potentialof this vvphwity- reversal shock depends upon the relative positions and velocity of the loadwith respect to the anvil at the time the reversal occurs. If the anvil (see Fig. 32) issuddenly stopped when the spring is at its greatest extension, then the damage potentialis the largest. That this is a matter of significancef is realized when it is observed that,depending upon the load vibration phase angle at the time of impact, the velocity changecaused by the stopping impact may vary from about 0.5 to 2.0 times the veloc~ity changecaused by the original impact. Various table travel distances and travel times involvedin a test make it probable that under some condition an item of equipment will be exposedto a phase angle of most severe danmage potential.

The relationship between ham miel imipact velocity (and height of hamimer drop) andthle ma1tXimum.11 veloCIty a~nd acceleration of thc anvil are shown in Figs. 35 and 36. Theseare linear relations. riiis is to be expcitetid when it is observed that the hanmmer-anvilimpact timec is it constaw¶, about 0.001 sec, and is independent of the drop height and ofthe load. Thel initial maximum velocity attained by the anivil is little affected by the load,az t at "tained sullclenly compared with the prriod of the load an its channels. However,;1% shown by F ig- s', the maximumi velocity experivnced by the load is affected, as onewould vxp('ct. by thev imgnitude of the load.

The d ifferente i in nia~giiitude of the, acce lerationi values passed by the 300- and 1000- (pslow-pass filtors, as shown in Fig. 36, indicated a1 considerable amount of relatively high-irl'qiii'ioc'v mtif'm ,n tt'slt' ill the. ;mi~il table. This is also shown. in Figs. 38 throu01gh 41;however, the latter, group also shows that tho maximium accelerations, of the load areab.Ixt t het 5si lflt %' hit' wlt-'lit i' or ntot they pass through the 1000- or thle 300- eps low-passfilter. Actually. as cani he seti 1bv tw he 'f'o1d oif the actual motion (Fig. 34), the maxi mu invallues of load awcvci'rationl and velocity are associate-d predominately with the 70- epsfundamiental lre('qoeicY. vLzen( ia Ily the (atliaIii's p'ov i(it' isolation from: the high frequencieswhen the load is r('lat'i~vly rigid.

I2~ ' ' t 'PER

Z3 I

W-

o t I

DRO -F

"'"Tof thAMe hnemotd loas.~~


"u,600 r7 I v E-,oot . i . ,oo tt,•,

,,4300 ltti

SI i

S I I "- , : *

z40 30 CP .LL

,. j'.i...S.... ..... .....l l ,I-7--

I , ' Ir•IPfTVI•T [' T/•},

40 0 m9

a *( CitLIIIC.IlIa* l' I I.'00

6 2 '4 .4 I 70IAUMFR IMPACT VF I OC.ITY (F T1F r.)

F'ii. 36, h NI imnuia '1z1vil -1,ilJle ILuvit-ratimiis for 1000U- .1jiud3U00-.eps Iow..pass, filt rationi. 1 'I'llevalue are i13(Iit-p13(Itlit of

S it,,•lE.-..f.t31 1 1 ( Ilads.

20

\S~NFT MAUVES DROP

-:'-. .' Fig. 37 - Maximum load velocity as a function

, _ .. ' of load weight fur different hammer drops

2

00 ,000 2000 3000 $O00 SC00

LOA0 WEIGHT (LOS.


- LOAD WEIGHT - I' LB I I801 S'ACN ;4 94.

CH -NO~ ANNELS - 3 - -- 1 -i -

400 -- --.4L I-c;,•r4•. 1 4

300 IT [ - PVT.. .

•- *1-t T - 1T - - - -

21, . '

Fig. 38 -Nlo,A~i11Um MCCelera- U) 100--4eA0 0 ~5,~tion vii "rigid" load -n of the-- ,S

anvil table - Ace clerationv sig -I I I;lnals wvre filtered by 300- or 01.! -c-ps l-v,-pzss filters as I 3 TABLE TRAvIL _

di 't5' TABLE A L

S' '5- TABLE TRAVEL

100

0i I.,•o V 130,0 .000 cps

"IFIGHT OF mBAR~d [I•ROP T , i

04 6 8 '$ 12 14 16 I]

HAMMER IMPACT VELOCITY (FT/SEC)

600 -' " --1010 wEG~ '01T TOI B I I i * IBoLT 1 PACRr- 4 IN ,

400............. • ' A A/ I

- ,000 CPS)

300 I ' 'I" I I

200

n. 1n0 ANVIL J

tS "150O0 CVS1

w1 1 ' ' I '. ' . . I , i Fig. 39- Maximum filtered

AB 0 "" TR'VELacceleration values on a rigida" 15 TABLE TRAVEL load and on the ativ.l table

150 'A 075' TABLE TRAVEL

100

' 00I a 000 cpsi,

5 0 . , hl l H IIDO-l O O (I l ,

"HtpIalT OF HANlWIER DROP-FT

4 6 a 10 12 14 16 IS

HAMMER- IMPACT VELOCITY (FT/SEC)


600 r- I I'11'i

-LOALD WEIGT-13841.SA D -t-.-4 J------

500 -- Ido CHANNELS-*

:--- 4. f-I , , - ,o • - '400

300

~200 10c

to-- L ~ •-:• - I, :-'' ' ,I I1 I

I , |*.• ] I | 'I

1j00 If: t ANV1 I ' I L '

a 305CPS)

o0 - Fig. 4 OMai in filtered, 'J :i A.V acceleration valucs on a rigid

IT0 0 ; A8Lr 'RAVEL ', load and on the anvil table

100

* I 1

50 I :I30U &LA D01 M

5O0 iIO cpsl;

iA

HEIGHTI IIOFHAMMER [ROP- It

U61 1? 1?5 20 2;5 3', _

4 6 8 TO 12 14 I Is

HAMMER-IWMACT VELOCITY (FT/SEC)

LOAD wl :[11T 442311 t • -- tII € J • •

01HOLT SACIN6 4 i

500 1.0 CANNI IS 10 *

I I ' ,I I A VL

400 ,O* I j 4c 00(15

300 ' *A

S. .. .. .. T• r•o ,

Z20 0 **I0

oo -

LA •,.• : u ;:I ;.'.L

Fi . 41 -Ia X'm u:n fi," 'rid I 0

... 4, 1 , . , + + •- • - -

~acceleratioli values on * rigid *IABL TRAVEL

load and on tii, aivil t at, uI S IAO E TRAVEL I". i ' I . 0'S AOL ?RAVEL

IS- ,,I .... _.,. -

_,, o , I .• -.- r , I , , I - -!00 i I ! .Ii •S

50,o+ O. 1 "00 a 000 -

[1!? '; ' ', ! : If A DR P - T7T V

"4 a 10 12 4 I4 Is

HIAMMER-IMPACT VELOCITY IFT/SEG)


Typical values ;:f shock spectra are shown in Figs. 42 and 43. These have beenextrapolated by dotted lines to show probable values beyond the range of measurements.The values at low frequencies hfcome asymptotic to the permissible table travel, whichis usually 3 or 1.5 in., although 0.75 in. may additionally be used. The spectra of themotions of the load* provide an equivalent velocity-shock, which is applicable from afew cycles per second to about 40 cps. The spectra then rises to a maximum value dueti the ioad-e:halmvil r1ebu'iilivV at about 65 cps. Above 100 cps the load acceleration -pect'aremains about coistant. If the load were a more flexible system higher values of accel.-eration would be experienced at the hig'h frequencies.

The relatively simple spe.ctrurn, Fig. 43, is illustrative.of that of the anvil table. Themotion of the anvil table can be cersidered a velocity-shock for frequencies between 10and 1000 cps. Below these frequencies the displacement spectra becomes asymptotic tuthe table-travel setting, and above these frequencies acceleration spectra becomes asymp-totic to the true maximum value of acceleration, which is in the region between 5000 and10,000 g.

CONCI,UDING DISCUSSION

The Navy III shock machines attempt to provide a simulation of types of shocksprobabhl on board ships. The intensity of shock is of such a value as would occur whenthe ships structure is damaged but the ship is still seaworthy. Shock tests are specifiedin terms of the shock machine and procedures for its use, rather than in terms of shockmotions or spectra.

Shock motions and sp)ectra typical for various loading cnnditions of the shock machinesare included in this report. A determination of these values under standard conditionsdescribes their performance, and is sotl*'times called the calihration of the ms(hine.

The shock motions expressed as a time function are not of themselves useful withoutfurther analyses. Perhaps the most useful, and meaningful, analysis method is in termsof shock spectra. Simpler, but less inforrmative, statements of shock intensity are givenin terms of velocity-shock, and in terms of maximum values of accelerations transmittedthrough filters of given bandwidth. All methods of anal)ri es assume something of thenature of the Item being subjected to the shock inasmuch as damage potentials of a shockmotion are as much it function of the characteristi( s of the item being shocked as theyare of the nature of the shock motions.

The use of four-coordinate log paper for the presentation of shock-spectral curvespermits the natural exteosion of the concept of velocity-shock to displacement-shock andpermits a determination of the maximum values of acceleration. The shock-spectralcurves illustrate that below a given frequency the displacements approach a constantvalue. This value, together with the frequency range for which it is sufficiently accurate,is defined as the displacement-shock value and range. A middle frequency-range usuallyexists for which the velocity-shock spectrum is relatively constant. This velocity isdefined as the equivalent velocity-shock for this range. And similarly an upper range offrequencies exist for which the acceleration-shock spectrum is relatively constant. Thisacceleration value is the maximum value present in the excitation. The maximum valueof acceleration can be considered as a static value in this upper frequency range. Maxi-mum values of changes of displacement a-:d v,..c'itv and maximum values of acceleraf-onused in this manner provide simple, significant, and meaningful descriptions of a shockmotion but of course do not incl ude all the information contained in the motion-time orshock-spectral curves.

":Sinc. t' - load is rigid, its -pectra and its motions arc assunico to bt, the same ,a5 that ofits -nountinp .::(iitlb 0n1 the channels of the hiock machine.

34 NAVAL RESEARCH LA,,ORATORY

(COOt j % (ri! °"•.v.... .. . .i' .~I: ....

10,iIPN2K 0:tK '143

'1 . • . . .. . . . . . .

lot Aq t A

r I'~4~ _W

II A\ )J .

* v '~ f '• ¢J'•• !iw' "ji ... .

-, . - .5

Vig. 4Z Shock sjnct ra for moins - of rn'td loadmiiilouted act ordlitip to jpresc r"Ibed 'specificationas (5).A. 5,5-ft 'a...iti d(IropJ, .3-in. .able travel, 1423-lbload, S lb. o n table, class A shock. B3. 2 .fi.am-mt r drop, 3-ii. table travel, 1115-lb Ioad, 1..3 lbon table, clas A shock.

- r,' . . ..1 .......

',, • " ;,';, - , :t " "• - : . . .. .' •t ,;. .. .

o F11 .... s4 , .....n 4 t j . . .•-,oei iht Equipment,

•' ,.. .. ... . • ..;,. 'i - .. L.... ..

F"ig. ,43- Shock spectrum for n 'lani| ofr ari•i tableof H Sick Mach•',ine: fo r, Mcdirnbed cighctEqipmnt;5)

3-ft harnirtme 1-r drop, 3-in, table t rravel, 1 b5-lb load,1858 lb on table, drop-height 150 percent greaterthan specified (5) for class A shock


The HI shock machines provide good duplication of shock motions for successiveimpacts of -,i given test. However if thE' equipment is dismounted and then remountedinl presumably an identical manner, differences in the nature of the high-frequency com-ponents are- observable. These are caour'd by differences in damping and elasticity under

the two conditions which is in turn caused hv differ'tit seating of parts held together bybolts. and by differences in bolt tightnoss. In general, frequency components below severalhundred cycles per second are not affected. The changcs of physical propertics of the

anvil plate of the machim't ior lightweight eqluipment. as it deforms with use, are such as

to gradually increase the severity of the shock motions.

It will occasionally happen that an item of equlipmen('t Will consistentI.y pass specifica-

tion onl one, HI nmuc hnt but fail an vqui~alent test on another similar machine. -Studies of

these. instancl's have shown that there was little miarizin of safety in the first eaSe, andthalt t hi' cr itical 1ValIe, for r i hi i' wap I-rh ' , ioei'nded in thi, second. Obviously an itemmnust pass the test onl the ..i :naew used during aecreptanice tests.

Thu maxima in the shoe k-spectral curves which are caused by resonance vibrations

(f the mountingw plates, or channelIs, of the sheu)(k mac hines art, of some concern. It mayhe that the, presencev of narrow frequency regions, inl which the damiage potential is large,

is nolil in ac ord Withi the idea,; of prensent ing a shock inot ion vquqi-ia lent to a generalizedfwie condition. It is probable that at sbo( k motion. -, i4,ut dominant frequencies would be

prefte cable.

.FKNOW1 FDI)G\IFN-TS

Ilthe atelltr al presented in this report has been 1)1incWipallIy obandby R . W. Col~a-id,1% F 1)jcp. H F lllalw. aol: F W. (leIcItis.

REFERENCES

1. Oliver, R.H., "The History and Development of the High-Impact Shock-TestingMachine for Lightweight Equipment," Shock and Vibration Bulletin No. 3, NRLReport S3106, pp. 3-8, May 1947

2. Conrad, R.W., "Characteristics of the Lightweight High-Impact Shock Machine,"NRL Report 3922, Jan. 1952

3. Dick, A.F., "Reed-Gage Shock-Spectrum Characteristics of Navy Lightweight High-Impact Shock Machine," NRL Report 4749, June 1956

4. "American Standard Specification for the Design, Construction, and Operation ofClass HI (High-Impact) Shock-Testing Machines for Lightweight Equipment," AmericanStandards Association Publication Z24.17- 1955

5. Military Specifications specifying and describing the HI Shock Machines for Light-and Mediumweight cquil)pnlnt:

(a) MTL-S-901B(Navy) April 1954MIL-S-901(Ships) Nnvewmblr 1949

(b) MIL-T-171113(ships) July 1952(c) MIL-E-005272B(USAF) .June 1957 Procedure III

6. Conrad, R.W.. "Characterirst icis of the' Navy Mediurowvight high-Impact ShockMachine," NRL Report 3852, Sept. 1951

7. Dick. A.F., and Blake. R.E., 'Reed-Gag, Shock-Spectrum Characteristics of NavyMediumweight High-Impact Shock Machine." NRL Report 4750. July 1956

8. "High-Impact Shock Machine for Electronic Devices (Flyweight)," Proposed A mericanStandard Specification S2.3/44 (to be published by ASA when approved); also MIL-E-ICof Oct. 1956 (Drawing 180-JAN)

9. Vigness, I., Kammer, E.W., and Holt, S., "Mechanical Shock Characteristics of theHigh-Impact Machine for Electronic Devices," NRL Report 0-2497, Mar. 1945

10. "Shock-Testing Mechanism for Electrical Indicating Instruments," American StandardsAssociation Publication C39.3- 1948; also Method 202A of MIL-STD 202A of Oct. 1956

11. Harris, C.M., and Crede, C.E., "Handbook of Shock and Vibration Control," New York:McGraw-Hill, 1961

12. "Shock and Vibration Instrumentation," ASME publication, pp. 12?-145, June 19%8

13. "Shock Testing Facilities," U.S. Naval Ordnance Laboratory Report 1056, Mar. 1956

14. "American Standrrd Specification for Design, Construction, and Operation of VariableDuration, Medium-Impact, Shock-Testing Machine f.r Lightweight Equipment,,Pu.:blication S2.1/1, American Standards Association (in preparation)

36


15. Lowe, R., "Barry Shock and Vibration Control Notes," No. 7, Barry Controls,Watertown, Mass., Aug. 1957 (see also bulletins by Lycouuing Division of AVCO Mfg.Corp., Stratford, Conn., relative to a New Precision Shock Test Machine)

16. Morrow, C.T., and Sargeant, H.I., "Sawtooth Shock as ;. Component Test," J. Acoust.Som. Am. 28:959 (1956)

17. Conrad, R.W., "Characteristics of the 3 ft-lb Vibration Machine," NRL Report S-3186,Oct. 1947

18. Conrad, R.W., "Characteristics of the 250 ft-lb Shock Machine," NRL Report F-3328,July 1948

19. Norg-rden, 0., and Shau.ahan, F.J., "A Devicc for Mechinical Test of ElectronicEquipment," Shock and Vibration Bt.letin No. 3, NRL Repo)rt S-3106, p. 43, May 1947

20. Don Hartog, J.P., "Mechanical Vibrations," New York: McGraw- Hill, 4th Edition,po. 87-102, 1956

21. Bllshii m, R.O., and Blake, R.E., -Effe(t of Equipment l)ynam ic Reaction on ShockMotion of Founudations," NRL Report 5009 (Confidential Report, Unclassified Title),Oct. 1957

22. O'Ha ra, G.J., "Eff,'vt Upon Shock Spectra of the l)ynamic Reactions of Structures,"

NIRL Rk-1),,rt 5236, 1)t-. 1958

23. O'Hara. G., . "Shock Spectra and D)esign Shock Spectra." NRL Report 5386. Nov. 1959

24. Walbh, J.P., and Bllake, R.E., "The Equivalent Static Accelerations of Shock Motions,"Proc. Soc. for Exper. Stress Analysis 6(No. 2):150 (1948) published 1949

25. Klein, E., cd., "Fundamentals of Guided Missile Packaging," Report RD 219/3, Officeof Assistant Secretary of Defense, Research and Development. Chapter 5, p. 24, 1955

26. Durr, G.W., "Investigation ul Characteristics of Mechanical Shock on H.I. ShockMachine," NRL letter report 3853-323A/50 GWD:mb to BuShips, Sept. 6, 1950

27. Crede, C.E., "Vibration and Shock Isolation" (see especially Chapter 3) New York:Wiley, 1951

28. Jacobsen, L.S., and Ayre, R.S., "Engineering Vibrations" (see especially Chapter 4,"Transient-Response Spectra"), New York:McGraw-Hill, 1958

29. Goodier, J.N., and Hoff, N.J., eds., "Structural Mechanics" (see especially pp. 512-517),Ncw York:Pergamon, 1960

30. Conrad. R.W., NRL letter reports to BuShips: NRL Code 6250. Shock and VibrationFolder 767, Nov. 1956 and Aug. 1957, Files 6251-386A/56 mb and 6251-236A/RWC:mb

c~c.

C~22

tc -

0, oO0 m

ýn L4

a= -

.I C. t t

- ri

C, U.

1r t" - CL

F-. -ý -Z 0 ,

-0 C Z cZ; 1 C? :z >

tU -- -C

-~~~C. L.U ~ C *

U ~ ~ ' Z n ~ C ~ ~ d

zza) u ) L

r- CLL

tz- C, C

F- U~~~) ~ F ,-0 C.Ci

to c- V)14

rd E 0 L 0

o9 .- r .2- 0 2,

tj .-

u0 0 0Z

.0 .

- U n

z Ln z> zý

tr~- cd tr C-*0 V.. r-

TI Tl

00

00

I c co)

CEn D

U, - .,'--.U

5~-a. 'A Lný

L..

=L cctUun~~~ A U,',4

L.n

Date post:	19-Jun-2018
Category:	Documents
Upload:	votuyen
View:	217 times
Download:	0 times

NAVY HIGH-IMPACT SHOCK MACHINES FOR ... HIGH-IMPACT SHOCK MACHINES FOR LIGHTWEIGHT AND MEDIUMWEIGHT...

Documents