SCOMPARISON OF THE PERCEIVED'NTENSITY OF SINUSOIDAL AND MULTI-FREQUENCY WHOLE-BODY VIBRATION

5D C

.~~I. SEPI I~P~ 19 11V7ft t

W)SPAC UIDCl DIVIVONAIRdW 8MEMS1 COMMANDJ

- ht US 'oenwfii~s sii-act oi0r data am usedror an piuipose athcz- tan a defi , it1y mcl'.d:~~~peuep bthe Govcrniat thereby in<ur ao laspinsilynr.r, bgnowhtcrmeta'x. ~mltd furnishrd. ot in ah) way supplied the said dmruw

,e! A otiitobe regi dcdb) imtplication or otherwise, as..itay rnanacs .. esiog thc holdeior any owhrmperonovprai 4iovci~. rights Tx permission w~ man~aac.sc use oraellahjypatetedmVeA.onittiryjq ay- a b i~~d thereto.

~s o no V4 p~ fthscpi from Ac-paeMeia Research Laboratoi,. Additionope mayj be

X4atiohal~echriicaI Information Servicet~Woya Road.

-~~~ ~Sfpned;iriia 2211 DfneDcmninCne huddrc~c~epp~oclesnt and their contractor regislxred wit

sforcopievo -thri'rporvt6:

4~ ifenstiouentation CenterCanior Station

.n4

4-- 14-oh .33'

'Thi reN-r fiibea - i ninito fie() n i eesbe0teNtba-ehia rdmt

avil-l to-_T ;t. Jjit~ W eal~pblkjncl ng e11tEHNCA ReVE AND t~oA '"aios4-,- astoaorvd ubiain

AI4RL-T-76-1 '

yn c. ThsvlnayiorMeacncr-'A-.

1~4ul~on *6~33

SECURITY CLASSIFICATION OF TIPAEWbn Dote Fntered)

REPORT DOCUMENTATION PAGE BFRE COMPRLETINORM

Lj * _; 2 GOVT ACCESSION No. 3. RECIPIENT'S CATALOG NUMBER

AMRL-T R-76-____________

4 5 TYPE OF REPORT A PERIOD COVERED

VIRATON~ hD 6 PERFORMING ORG. REPORT NUMBER

A- &RACT OR GRANT NUMBER(@)

9 PERFORMiNG ORGANIZATION NA14E AND ADDRESS 10 PR GRAWm-ICEMENT PROJECT, TASK

Aerospace Medical Research Laboratory, Aerospace ARCA IWORK UNITNMER

Medical Division, Air Force Systems Conanand, 4622W-, 7231-01-01Wright-Patterson Air Force Base, Ohio 45433/

11. CONTROLLING OFFICE NAME AND ADDRESS-I

14 MONITORING AGENCY NAME 6 ADDRESS(If different fro., Controfttg Office) IS SECURITY CLASS. (of title report)

15a. DLCLASSIFICATION'DOWNGRADING~J SCHEDULE

16. DISTRIBUTION STATEMENT (of ti. Report)

'7. DISTRIBUTION STATEMECT (of the abstfect entered in Block 20. If different frot. Report)

IS. SUPPLEMENTARY NOTES

19. KEY WORDS (Cctntintjo on reveree mide If necesary and identify by block nuamber)

Vibration (Whole-Body Vibration), Perception (Vibration), Vibration Criteria,

Psychophysical Matching, Human Factors, Biodynamics

20. ABSTRACT (Continue en, roere side If necessary etnd Identify by block tttaniber)

--- Eistngstadads orhuman exposure to whole-body vibrationarbseheavily on data f apmn~i hc the vibration inputs were singlesinusoids," emany operational environments contain comple-x vibrFatlnscomposed of multiple sine waves or random disturbances. The preferred methodset forth by the standards for assessing such environments is to evaluate eachcomponent frequency or each third-octave band independently with respect tothe levels specified by a particular criterion curve. Under this procedure,

DD I OR 1473 EDITION OF 1NOV.165 IS OBSOLETE

SECURi-LIFICAION OS THIS PAGE (07v- Del. Entered)

A-

SECURITY CLASSIFICATION OF THIS PAGE(I hen D.10 EnI.fd)

(20.)

if a complex vibration is found to contain several sinusoids (or third octavebands), each with amplitudes at tile level of a particular criterion curve, itwould be rated the same as any one of the sinusoids alone. However, as moresinusoids (or third-octave bands) are added to the combination the greater willbe the total powerqr force imparted to the man and the greater the possibilityfor_ iue--ncy ---- interaction->sAn intensity matching technique was used to testthe independent component method for evaluating complex vibration environmentscomposed of multiple sine waves. Ten subjects adjusted the intensity of a25 Hz sinusoid to match the subjective intensity of 11, 17, 40, and 63 Hzsinusoids (all with intensities at the same criterion level), presented eithersingly or in all possible combinations of two, three or four frequencies. Theresults showed a monotonic relationship between perceived intensity and thenumber of sinusoids in the stimulus (i.e., the acceleration of the matchingresponse increased significantly as the number of sinusoids increased). Thesefindings indicate (at least for the frequency range sampled) that the"independent frequency" method of evaluating non-sinusoidal vibrations willunderestimate the severity of such complex vibr-ation environments, and suggestthat the weighting technique recommiended by the standards as an al nativeevaluation method may more accurately evaluate their effects.

I )1

SECURITY CLAS SIFICATION4 )F THIS PAOE(W7,n b..r. En'e'.d)

PREFACE

Th~s research was conducted by Richard W. Shoenberger, Vibration Branch, Biodynamics and BionicsDivision, Aerospace Medical Research Laboratory, Wri ght- Patterson Air Force Base, Ohio. This worksupports Project 7231, "Biomechanics of Air Force Operations: Effects of Mechanical Forces on AirForce Personnel," Tas-k 723101, "Effects of Vibration on Air Force Crews and Personnel," Work Unit72310101, "Aircrew Performance and Subjective Response During Vibration Encountered in Air ForceOperations."

1

- . - j

iINTRODUCTION

The majority of studies investigating human response to whole-body vibration have employed discretefrequency sinusoids as the vibration input stinli. As a result, existing standards for vibrationexpos.ure (1, 4, 6) are all based heavily on data collected using single sinusoids. Yet many operational

vibration environments contain complex vibrations made up of multiple sine waves or broadbandrandom disturbances. The vibration standards specify two procedures for evaluating such environ-ments: (al Evaluate each frequency, for multifrequency inputs, or each third-octave band, for randominputs. independently with respect to the level specified for that frequency (or third-octave band) bN aparticular criterion curve; (b) Use it frequency weighting network, normalized to the most sensitiveportion of the curve, to obtain a single weighted RMS acceleration value, which is then evaluated

against the same portion of the curve on which the normalizal.on was based. There are indications thatneither of these procedures is completely adequat. For example, with respect to perceived vibrationintensity, studies by Brumaghim (2) and Dupuis, Hartung, anti Louda (3) have produced sone evidenceindicatingthat an independent evaluation ofindi% idual components would yield an underestimation ofthe subjective intensity of complex inputs, anti the ISO standard (4) points out that in the case wherethe input spectrum has a shape similar to the criterion curve the weighting procedure would lead to to,conservative an evaluation, i.e., the severity of the vibration would be overestimated.

In order to develop more adequate procedures for evaluating nonsinusoidal vibrations, additioralinformation is needed on the relationships between the effects of sinusoidal and more complex vibra-tion environments. The research described in this paper was conducted to provide such mformat on,ani was specifically designed to test the independent evaluation method (procedure "a") for evaliat-ing multifrequency vibration. U'nler this procedure, if a multifrequency vibration is found to containseveral sinusoids, each with amplitudes "t the level of a particular criterion curve. t would be rated atthe same criterion level as any one of the sinusoids alone. However, as more sinusoids are added to thecombination, the greater will be the total power or force imparted to the man and the F.Teaterpossibility for frequency interactions. Therefore. a weighting method such as that specified in proce-(lure "b" might prove to he more advantageous.

In the present experiment, the subjective intensities of several sinusoidal anti multifrequency vibra-tions (-omposed of from one to four frequencies) were measured by having subjects match theirperc tons of the intensities of the various inputs by adjusting the intensity of a single sinusoidalmatching frequency.

METHOD

SUBJECTS

The subjects were 10 male Air Forcc military personnel. They were physically qualfied volunteermembers of a vib~ration panel. and received hazard incentive pay for participation i.i vibration experi-

ments.

3

-. - - DVtG PA:1 BLUK.?M? fLMED.

APPARATUSVibration was produced by an N113 Electronics electromagnet it bratr (.Mode ( -5). which Iad beenmodi fled bt tie addition of a spri rig below the o, I nelIment 80 t iat It co~luld bandlle the luIoad of i manplus the seat and restraint harness. A lightweight alurninu m seat wa, rigidly mounted to the shakert'ead and transmitted the vibration dhrectl) to the subject without cushloning or padding. The subjectwas secured to the seat by a lap belt and shoulder harness. Figure 1 shows the 'ubject seated on theshaker and holding the potentiometer with which he controlled the amplitude of the matchingvibrat ion.

Five function generators produced the five sinusoidal frequencies used in the expenment. and theiroutputs were fed to the shaker via an EAI TR-20 analog computer. This provided the necessary gainsettings for each frequency and al!owed them to be selected singly or in combinations o" from 2 to 4frequencies. Gain settings for the individual frequencies were calibrated for each subject and samplecombined frequency spectrums for each subject were analyzed on a Time Data 1923 vibration analyzerto assure that posaible frequency interactions did not significantly affect the acceleration levels of thecomponent frequencies making up the combinations. The RMS acceleration of the vibrating .seat wasalso displayed on a meter, and the acceleration of the 25 Ifz matching frequency was read from thismeter and recorded by the experimenter for cacti matching response.

VIBRATIONThe vibration stimuli were composed of four tinusoidal frequencies (11, 17, 40. and 63 Hz) presentedeither singly or in all possible combinations of two, three, or four frequencies. Twenty-five Hz (thematching frequency) was also presented as a stimulus to pro% ide a check on possible biases or errors intne matching response when the stimulus aid response frequencies were identical. The frequenciesused were approximately the preferred center frequencies of every other third-octave band from 10 to63 Hz. However, slight departures froni some of these center frequencies were made to avoid harmonicrelationships between frequencies. This resulted in constantly changing phase relationships betweenthe frequencies in all combinations. rather than the fixed p11ing that would occur for harmonicallyrelated frequencies. All frequencies were presented (whether singly or in combinations) at accelera-

tions corresponding to the ISO 25-m Fatigue-I)ecrea.sed Proficiency (FI)P) level (4) Table I lists all ,fthe single- and multi-component stimuh used and specifies their frequencies and RMS accelerations.

PROCEDUREEach subject was required to match his perceptihm of tie intensity of eacit of tie stimulus vibrationslisted in Table I by adjusting the intensity of a 25 Hz matching frequency until he felt that itssubjective intensity matched the subjective intensity of the stimulus vibration he had just experi-enced. Each match involvedl a 30-second expostire to the stimulus vibration and a subsequent exposureto the matching frequency that lasted approximatt-k% 15 to :10 seconds. depending on how quickly thesubject achieved a match.

When each subject arrived at the laborator%. the nature of the experiment and tile intensit.-matchmingprocedure were explained. The subject was then seated lt tie ibration chair and given a siortpractice session by having him mate, the following series of six 1ingle-frequ.eny stimul. 25, 17.4). I1.63. and 25 Hz. Hti then matched cacti of the 16 vibration stimuh :iown lii Table I ,the stimuh werepresented to each subject in a different random order). After :t short rest break (about 5-min). thestimulus series was presented ;alin and tie Illljv-t mlitched eac'. stimulus vibration a second time.

Nor4

wwo u~c etdO ibao

TABLE I

VIBRATION STIMULI

.VUMBIR OF STIM'' FI'RE(,I'I.VN' .A'('ELERATIM.C'OMIPONENTS C'ODE 11. tl-S .' ;

A I 0.25

B7 0.39

On! 25 25 0.57C 10 0.92I) 63 1.43

A 11 17 .41;

AC 11 40 0.95

AD I11 63 1.45TwoBC 17 40 0.09B) 17 63 1.48CD 40 #;3 1.70

ABC 11 17 40 1.03ABI) 11 17 * 63 1.50

Three ACD) 11 40 63 1.72

BCD 17 + 40 63 1.74

Four ABCD 11 + 17- 40 -63 1.76

RESULTS AND DISCUSSION

The 20 matching responses (2 matches for each ef 10 subjects) collected for each of the 16 vibrationstimuli used in the experiment were averaged to obtain the mean response accelerations shown foreach stimulus in Table 2. Also presented in Table 2 are the mean responses for all stimuli containingequal numbers of sinusoidal components (i.e., one, two, three, or four).

Since the acceleration levels of the sinusoidal frequencies were ail st" at the ISO 25-min FDP boundary,

their subjective intensities should have all been approximately equa!. Inspection of the mean responseaccelerations in Table 2 for the single-frequency stimuli indicates some variability in the level ofaverage response. A few subjects felt that the highest frequency stimulus (stimulus D, 63 Hz) wasparticularly disagreeable, and accordingly set the acceleration of their matching responses ratherhigh compared to the other frequencies. This ;s reflected in the fact that the mean response forstimulus D in Table 2 is higher than for any other single-frequency stimulus. Further inspection ofTable 2 also indicates some differences among the responses for the various stimuli made up of twofrequencies as well as among those for stimuli composed of three frequencies. In order to test the

significance of these observed differences, three analyses of variance were r'erformed; one based on thedata for single frequencies, a second based on the data for double frequencies, and a third based on thedata for triple frequencies. The form was the same for all three analyses, a simple treatments xsubjects design (5).

6

(.

TABLE 2

ACCE.ERATION OF MATCHIN( RESPONSES RMS Gz

.TA 11.1 V A % RE.hoSE VI'.14R W* PEA. RKSIOVSEC 1D ,'. ¢'otli'NENT. BVNO. OF

("oMtPo V"

" VTSI

25 B 1S One 0.64

0.76

AB o.7'9

AC" O.80

AD 0 .93

Bfe 0.84 Two 0.8BD) 1.05

4"11 0.89

ABC 1.02ABD 1.14

BCD 1.20

ABCD 1.31 Four 1.31

The analysis for single frequencies revealed that none of the observed differences among the single-frequency mean responses were significant (F,.3 = 2.02; p > 0.10), indicating that the mean subjectiveI

intensities for the single-frequency stimuli were essentially equivalent, as the ISO standard denotes.However, significant differences were indicated by the analyses for double frequencies (F. = 6.61; p< 0.001) and triple frequencies (F,, 7 = 4.08; p < 0.05). Newman-Keuls tests (7) were then performed todetermine which stimuli differed s,-nificantly within each of these two groups. Results of these testsare presented in Table 3 for the double frequencies and Table 4 for the triple frequencies. Only stimulusBD(17 + 63 Hz) produced a unique response among the double frequencies. Table 3 shows that theresponse for BD was significantly greater than for all of the other double frequency stimuli. Table 4indicates only one significant difference among the triple frequencies. The stimulus producing thehighest response (BCD: 17 + 40 + 63 Hz) was significantly different from the stimulus producing thelowest response (ABC: 11 -t 17 + 40 Hz).

Although the differences for the single-frequency stimuli were not significant, stimulus D andstimulus B had the highest responses among the single frequencies, and the combination of these twofrequencies (stimulus BD) had the greatest response for the double frequencies (and the only onesignificantly different from the others). Moreover, these two frequencies were also included in thetriple-frequency stimuli producing the two highest responses (rCD and ABD). Apparently, at least fora -5ignificant number of the subjects in the present sample, the combination of certain stimuli, whichwere judged to be only insignificantly more intense than other stimuli when experienced individually.resulted in a summatitn effect and produced significantly stronger reactions than simijar combina-tions of other qtimuli.

7

TABLF 5

RESULTS OF NEWMAN-KEULS TESTS FORNUMBER OF FRFQUENCIE3 IN COMBINATION

NUMBER OF 4FREQUENCIES

Ordered Mean 0.64 9.88 1.11 1.31Response (RMS Gz)

Differences 1 0.24"" 0.47** 0.67°*

Between Means 2 0.230" 0.43*00.20"0

:1 **p < 0.01

The relationship between the number of frequencies in the stimulus and the acceleration of thematching response is presented graphically in Figure 2. The figure shows an essentially linear increasein response acceleration as a function of the number of sinusoids in the stimulus. The data analysis hasestablished the statistical significance of this increase, and the fact that the acceleration of thematching response increases by roughly a factor of two as the stimulus increases from one to fourfrequencies indicates its practical significance as well. A factor of two increase in acceleration for agiven frequency represents a change in severity in terms of the ISO standard from the FDP boundaryto the Exposure Limit. Within the limits of the present investigation, these results demonstrate that

.1 the "independent frequency" method of evaluating multifrequency vibrations will underestimate the;everity of such complex vibration environments and that the degree of underestimation will increaseas the number of frequencies in the stimulus increases. The data also tend to support the weightingtechnique recommended by the standards as an alternative evaluation method.

Continuing investigations in this area will include a similar experiment in which intensity judgmentswill be made of vibration environments composed of from one to four third-octave bands of randomvibration; comparable experiments at lower frequencies extending down into the body resonancerange; and, when sufficient data become available, evaluation of the weighting technique recom-mended by the standards as a secondary method for evaluating complex vibrations and, if necessary,development of alternative weighting procedures.

9

L-%

1.6

- 1.2

LU

1 0.8tI

.1 0.4

I I I I -

1 2 3 4

NO. OF SINUSOIDS(25-MIN. FDP)

Figure 2. Acceleration of Matching Response as a Function of theNumber of Frequencies itn the Stimulus

*1 1j

A'

2¢

REFERENCES

1. Air Force Systenis Command. 1972.Evaluation of Human Exposure to Whole-Body Vibration. AFSCDesign Note 3El, Design Handbook 1-3, Series 1-0.

2. Brumaghim, S. H. 1967.Subjective Reaction to Dual Frequency Vibration. Wichita Kan.: The BoeingCo., Technical Report D3-7562.

3. Dupuis, H., E. Hartung, and L. Louda. 1972. "Vergleich regelloser Schwingungen eines begrenztenFrequenzbereiches mit sinusformigen Schwingungen hinsichtlicht der Einwirkung auf den

Menschen." ("Random vibrations of a limited frequency range compared with sinusoidal vibrationswith regard to its effect on man.") Ergonomics. 15:237-265. USAF School of Aerospace MedicineTechnical Translation G-1155-0373.

4. Interntional Organization for Standardization. 1974. Guidcforthe Evaluation o Human Exposureto Whole-Body Vibration. ISO 2631-1974.

5. Lindquist, E. F. 1953. Design and Analysis of Experiments in Psychology and Education. Boston:Houghton Mifflin Co. '

6. U.S. Department of Defense. 1970. Military Standard 1472A; Human engineering design criteria formilitary systems, equipment and facilities. Washington, DC: Government Printing Office.

7. Winer. B. J. 1962. Statistical Principles in Experimental Design. New York: McGraw-Hill.I 1'

.7

111'rA 41 4!<j ~' ~