AD-A248 670III~iill IIII lUU liUSAARL Report No. 92-17

An Experimental Basis for the Estimationof Auditory System Hazard

Following Exposure to Impulse Noise(Reprint)

DTIC BySElEC.'E

APR 21 1992 James H. Patterson, Jr.

D Sensory Research Division

and

Roger P. Hamernik

Auditory Research LaboratoryState University of New York

at Plattsburgh 92-10055

February 199292 4 20 073

Approved for public release; distribufion unlimited.

United States Army Aeromedical Research Laboratory

Fort Rucker, Alabama 36362-5292

Qualified requesters

Qualified requesters may obtain copies from the Defense TechnicalInformation Center (DTIC), Cameron Station, Alexandria, Virginia22314. Orders will be expedited if placed through the librarianor other person designated to request documents from DTIC.

Chanae of address

organizations receiving reports from the U.S. Army AeromedicalResearch Laboratory on automatic mailing lists should confirmcorrect address when corresponding about laboratory reports.

Animal use

In conducting the research described in this report, the investi-gators adhered to the Guide for Laboratory Animal Facilities andCare, as promulgated by the Committee on the Guide for LaboratoryAnimal Resources, National Academy cf Sciences-National ResearchCouncil.

Disposition

Destroy this document when it is no longer needed. Do not returnit to the originator.

Disclaimer

The views, opinions, and/or findings contained in this report arethose of the author(s) and should not be construed as an officialDepartment of the Army position, policy, or decision, unless sodesignated by other official documentation. Citation of tradenames in this report does not constitute an official Departmentof the Army endorsement or approval of the use of such commercialitems.

Reviewed:

THOMAS L. FRMALTC, MS

Director, Sens y Research

Released for publication:

ReV W. ,.DPh.D. DVDH EChairman, Scientif-ic- Colonel, MC, FSReview Committee commanding

UNCLASSIFIED

SECURITY CLASSIFICATION OF THiS PAGET

Form Approved

REPORT DOCUMENTATION PAGE OM 4e.7O0-0188

Ia. REPORT SECURITY CLASSIFICATION i'. RESTRICTIVE MARKINGS

UNCLASSIFIED

2a. SECURITY CLASSIFICATION AUTHORITY 3. DISTRIBUTION/AVAILABILITY OF REPORT

Approved for public release; distribution2b. DECLASSIFICATION /DOWNGRADING SCHEDULE unlimited

4. PERFORMING ORGANIZATION REPORT NUMBER(S) S. MONITORING ORGANIZATION REPORT NUMBER(S)

USAARL Report 92-176a. NAME OF PERFORMING ORGANIZATION 6 6b. OFFICE SYMBOL 7a. NAME OF MONITORING ORGANIZATIONSensory Research Division (If applicable) U.S. Army Medical Research and DevelopmentU.S. Army Aeromedical Each Lab Comand

6c. ADDRESS (City, State, and ZIP Code) 7b. ADDRESS (City, State. and ZIP Code)P.O. Box 577 Fort DetrickFort Rucker, AL 36362-5292 Frederick, MD 21701-5012

Ba. NAME OF FUNDING/SPONSORING 8b. OFFICE SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBERORGANIZATION (if applicable) Partial effort under contracts #DAMDl7-86-C-

.1 6172 and DAMDl7-86-C-6139

8c. ADDRESS (City, State, and ZIP Code) 10. SOURCE OF FUNDING NUMBERS

PROGRAM PROJECT TASK WORK UNITELEMENT NO. NO. NO. ACCESSION NO.

0601102A1 3M161102B5 5 282

11. TITLE (Include Securrzy Classification)An experimental basis for the estimation of auditory system hazard following exposure toimpulse noise

12. PERSONAL AUTHOR(S)

James H. Patterson, Jr. and Roger P. Hamernik13a. TYPE OF REPORT j3b. TIME COVERED 14. DATE OF REPORT (Year, Month, Day) 1S. PAGE COUNT

TROM TO 1992 February I 1316. SUPPLEMENTARY NOTATION ...

Reprint from Noise-Induced Rearing Loss, Dancer, Henderson, Salvi,

and Ramarnik. St. Louis, MO. Mosby-Year Book

17. COSATI CODES IS. SUBJECT TERMS (Continue on reverse if necessary and identify by block number)FIELD GROUP SUB-GROUP Impulse noise, hearing, chinchilla, audiometry, and

ZU UL histology

U• ABSTRACT (Continue on reverse if necessary and identfy by biock number)eAenergy spectrum of a noise is known to be an important variable in determining the

effects of a traumatic exposure. However, existing criteria for exposure to impulse noisedu not consider the frequency spectrum of an impulse as a variable in the evaluation of thehazards to the auditory system. This report presents the results of three studies that weredesigned to determine the relative potential that impulsive energy has in causing auditorysystem trauma. Four hundred and seventy five (475) chinchilla were used in theseexperiments. Pre- and post-exposure hearing thresholds were measured on each subject. Inthe first study, the noise exposure stimuli consisted of six different computer-generatednarrow band tone bursts having center frequencies located at 0.260, 0.775, 1.025, 1.350,2.450, and 3.550 kHz. Each narrow band exposure stimulus was presented at two to fourdifferent intensities. An analysis of the audiometric data allowed a frequency weightingfunction to be derived. This weighting function de-emphasizes low frequency energy morethan the conventional A-weighting function. In the second study, the exposures consisted of

20. DISTRIBUTION/AVAILABILITY OF ABSTRACT 21. ABSTRACT SECURITY CLASSIFICATION

q• UNCLASSIFIED/UNLIMITED 03 SAME AS RPT. C DTIC USERS

22a. NAME OF RESPONSIBLE INDIVIDUAL 22b. TELEPHONE (Include Area Code) 22c. OFFICE SYMBOL

Chief, Scientific Information Center ._ (205) 255-6907 1 SCD-UAX-SI

DD Form 1473, JUN 86 Previous edittons are obsolete. SECURITY CLASSIFICATION OF THIS PAGEUNCLASSIFIED

Block 19 continued

two-types of broad band computer synthesized impulses. Subjects were exposed to 100impulses at a rate of l-per-3-seconds. Each type of impulse was presented at 3 intensities.The third study used impulses generated by three different diameter shock tubes. Subjectswere exposed to 1, 10, or 100 impulses at one of three intensities. The results of thesecond and third studies ware interpreted using the weighting function derived from tlefirst study. The hearing loss from all three studies is a linear function of the weightedSEL calculated using the weighting function, derived in the first study.

CHAPTER 30

An Experimental Basis for theEstimation of Auditory SystemHazard Following Exposure toImpulse Noise

JAMES H. PATrERSON, JR.

ROGER P. HAMERNIK

There are a number of different suggested A more direct spectral approach to thestandards for exposure to impulse/impact evaluation of impulses and impacts was pro-noise (Coles et al, 1968; OSHA, Dept of Labor, posed by Kryter (1970). His suggestions. ai1974; Smoorenburg, 1982; Pfander et al, though based on sound reasoning, never1980). Although each of these criteria has its gained acceptance. The Krytei approach wasproponents, none of them is in complete attractive in its ability to predict the amountagreement with existing data (Smoorenburg, of temporary threshold shift measured 2 min-1987). What is needed is a new criterion. Un- utes after exposure (MS 2 ) to a noise tran-fortunately, there is an extremely limited em- sient. However, this approach was limited topirical database on which a new standard can situations in which the M2 was not exces-be built. The difficulties associated with gener- sively large or, alternatively, the levels of theating such a database are compounded by the transient in any given frequency band wereextremely broad range of high-intensity noise not excessive.transients that exist in various industrial and Price (1979, 1983, 1986) has built on andmilitary environments. For example, in indus- extended the Kryter approach by consideringtry, impacts with variable peak intensities and the spectral transmission characteristics of thea reverberant character often occur. At the peripheral auditory system. Price's reasonkigother extreme, the diverse military weapon led to the Woliowing conclusions: (1) There issystems produce impulses that originate as the a species-specific frequency, f., at which theresult of a process of shock-wave formation cochlea is most vulnerable and that impulsesand propagation following an explosive re- whose spectrum peaks at f, will be most dam-lease of energy. These waves, which can have aging. This would appear to be true, accordingpeak levels in excess of 180 dB, can be either to Price, regardless of the distribution of en-reverberant or nonrc.ierberant, depending on ergy above and below C3. For man, the sug-the environment in which they are encoun- gested frequency is 3.0 kHz; and (2) Relativetered. Trying to develop a single standard to to the threshold for damage at fo, the thresh- + 1cover this broad range of "acoustic" signals is old for damage should rise at 6 dB per octavea formidable task. when fp is greater than f, and at 18 dB per oc-

Existing or proposed exposure criteria tave when fp is less than f., where 4 is spectralgenerally lack specific consideration of the fre- peak of the impulse. Thus, a model for perma-quency domain representation of We impulse. nent damage was developed that is amenable ...........This point has been raised frequ.arly by Price to experimental testing. In subsequent studies,(1979) and others. However, some deference Price (1983, !b86) has tried to relate, with ------------is given to the spectrum in these criteria, in an varying degrees of success, experimental dataindirect manner, through the handling of the obtained from the cat to the predictions ofA and B duration variables, this model. More recently, Hamernik et al :odes

336 Of

AN EXPERIMENTAL BASIS FOR THE ESTIMATION OF AUDITORY SYSTEIV HAZARD 337

(1990) and Patterson et al (1991) have re- MethodsporLcd oai an extensive series of parametrics:udies in which the spectra of the impulses The noise-induced permanent thresnoldwere iaried. A. re; icw of the literature indi- shift (NIPTS) data presented in thi. reportcates that, except for the studies mentioned were acquired from 475 chinchillas exposedabove, there arc few other published results to high levels of impuls:;- noise. Audiometricobtained from experiments specifically de- data on each animal wtre obtained using ei-signed to study the effects of the spectrum of ther a shock avoidance procedure (Pattersonan impulse on hearing trauma. et al, 1986) or measures of the auditory

This chapter presents an analysis of the evoked potential (Henderson et al 1983). Per-Patterson et al ( 1991 ) data from which a spec- manent threshold shifts were computed fromtral weighting function is derived. This the mean of three preexposure audiogramsweighting function will then be applied to the and at least three audiograms taken 30 days af-blast wave data of Hamernik et al (1990) and ter exposure. The behaviorally trained animalsto the synthetic impulses from Patterson et al were tested at octave intervals from 0.125 kHz(1986) in order to develop a relation between through 8 kHz including the half-octave pointsthe permanent threshold shift (PTM) and the 1.4, 2.8, and 5.7 k-Hz. Evoked potential thresh-sound exposure level (SEL). The intention olds were measured at octave inte-rvals fromhere is not to present a set of conclusive re- 0.5 to 16 kHz and at the 11.2-kHz point. Forsuits, but rather to illustrate a new approach each animal, mcasurc., 3f compound thresholdto the analysis of this type of experimental shift, VPS, and quantitative histology (cochleo-data. It is an approach that develops a direct grams) were obtained. In the analysis that fol-relation between frequency-specific measures lows, only PTS data will be discussed.of MTS and the frequency domain representa-tion of the impulse. The results of this ap- Series I Exposures (N = I 18)proch can be related directly to the Price(1983) model and can be used to estimate thepermazent effects of a traumatic impulse noise Animals were exposed at a normal inci-exposure in a manner similar to that approach dence (i.e., the plane of the external canal wasproposed by Kryter (1970) for estimating parallel to the speaker exit plane) to 100 irm-temporary threshold shift (TS) after a ra. pulses presented at the rate of 1 every 3 sec-pulse noise exposure. onds, This series of exposures consisted of 20groups of animals, with five to seven animals

per group. The stimuli were narrow-band im-pulses produced by passing a digital impulse

TABLE 30-I Exposure Conditions through a four-pole Learner-type digital band-

for the 20 Groups of pass filter (Gold and Rader, 1969). FollowingAnimals Used for analog conversion, the signal was transducedSeries I Exposures through an Altex 515 B speaker in a model

815 enclosure. The filter bandwidth was inde-CF (Hz) PEAK SPL (dB) TOTAL SEL (dB) pendent of center frequency, with steep atten-

260 139 uation outside the passband permitting the260 146 139.8 synthesis of equal energy impulses at a variety775 134 124.8 of center frequencies while assuring minimal775 139 129.4 spread of energy to other frequencies. The775 144 134.8 center frequencies of the six sets of impulses

1025 129 119.8 varied from 260 to 3,350 Hz. The bandwidth1025 134 124.2 of the impulses was approximately 400 Hz.1025 139 129.11025 144 134.6 Impulse peaks were varied from 124 to 1461350 129 119.8 dB. For each of the exposure conditions listed1350 134 124.2 in Table 30-1 the total SEL was computed as1350 139 129.0 follows (Young, 1970):2450 129 120.62450 134 124.92450 139 129.6 P2 t t2450 144 135.0 SEL = 10 log,,) 2

1S50 124 113.0 P t3550 129 119.93550 134 12 i.23550 139 129.5 where t, 1 second, Pr = 20 LPa. Figure 30-1

illustrates an example of the pre-suce-time his-

2

"'9-

338 PARAMETERS OF EXPOSURE

20

775 1& d, 139 d8 peak SPL

40.

260

-a

" "401

10 100 1000 10000

Frequency (Hz)

20--1350 fI, 139 d6 peak SPL

0-

3 -20- 0 T~Sn) 25'

-tt

10 100 1000 10000

Frequency (Hz)Figure 30-1 Examples of the 775-Hz (A) and 1,350-Hz (B) center frequency impulses of the Series I exposuresalong with their respective spectra.

TABLE 30-2 Exposure Conditions for the Seven Groups Used forSeries II Exposures

WAVE TYPE PEAK SPL (dB) TOTAL SEL (dB) TOTAL P-SEL (d6) TOTAL P'-SEL (dB)

High Peak 147 130.8 127.6 133.4Low Peak 139 130.3 127.2 132-9High Peak 139 123.0 119,9 125.6Low Peak 131 12Z4 119.3 125.0High Peak 135 119.1 115.8 121.6Low Peak 127 118.5 115.3 121.0High Oeak 131 115.1 111.9 117.5

tories of the 775-Hz and 1,350-Hz center fre- 1 every 3 seconds. There wert seven differentquency impulses along with their respective exposure conditions (Table 30-2) to whichspectra, seven groups of animals were exposed. Each

group contained six animals. Two types (lowpeak and high peak) of relativel, broad-band

Series 11 Exposures (N = 42) impulses with identically-shaped amplitudespectra were synthesized digitally (Patterson

Animals were exposed at a normal inci- et al, 1986). The peak sound pressure leveldence to 100 impulscs presented at the rate of (SPL) of the impulses was varied from 127 to

3

AN EXPERIMENTAL BASIS FOR THE ESTIMATION OF AUDITORY SYSTEM HAZARD 339

147 dB. Hearing threshold data were obtained from each set of the two to tour groups of an-using the avoidance conditioning procedure. imals that make up an intensity series for aFigure 30-2 illustrates the pressure-time histo- specific characteristic frequency (CF) impulseries of typical high- and low-peak impulses behaves in an orderly manner, with fi.2,4 in-along with their common spectrum. creasing in an approximately linea: fashion

with increasing SELSeries III Epsre The relative susceptibility to NIPrS is(N = 315) seen to be a function of the impulse center

frequency, with th, lower-frequency impulses

producing relatively little NIPTS even at theAnimals were exposed at a normal inci- higher SELs. A relative frequency weighting

dence to either 1, 10, or 100 impulses, pre- function can be derived from the data pre-sented at the rate of or 1 every 10 seconds at scinted in Figure 30-5 by shifting each fre-intensities of 150, 155, or 160 dB peak SPL All quency-specific data set along the SEL axis theof the above combinations of number, repeti- amount that is necessary to collapse the datation rate, and peak yielded 21 different expo- into a si-gle PTS/SEL function using one of thesure groups with five animals per group. The exposures as a "zero" reference.impulses were generated by a compressed-air- Such a data-shifting process was carrieddriven shock tube. This set of 21 exposures out "by eye" to produce a best fit using thew-.s repeated using waves generated by three 1,350-Hz series of..data as the reference point.shock tubes of different diameters that pro- The amounts shifted were 260-Hz CF im-duceO blast waves whose spectrum peaked at pulses, -20 dB; 775-Hz CF impulses, -7.2 dB;three different locations of the audible spec- 1,025-Hz CF impulses, -4 dB; 1,350-Hz CFtrum. The pressure-time traces and spectral impulses, 0 dB; 2,450-Hz CF impulses, -4 dB;analysis of these waveforms are shown in Fig- and 3,550-Hz CF impulses, +4 dB. The re-ure 30-3. In addition, the A-weighted octave alignment of the data that such a shift pro-band energies are shown in Figure 30-4 so duves is Shown in Figure 30-6, and the weight-that comparisons could be made for each ing function, thus obtained, is shown plottedwave from each source. Because of the high (solid line with symbols) in Figure 30-7,levels of very-low-frequency energy in these where it is compared to the conventionalblast waves, the resolution at the high fre- A-weighting function (solid line). The newquencies is poor if unweighted energies are empirical weighting function is referred to asplotted. For further details ,,ee Hamernik and 0-weighting in the legends for these figures. AHsueh (1990). Table 30-3 summarizes the linear regression through the shifted data setconditions for the Series III exposures. Only showed a correlation coefficient of 0.89 with athe SELs for the 100-impulse conditions are slopc of 2.6 dB PTS per decibel P-weightedtabulated. Successive 10-dB adjustments need SEL (P-SEL) and a threshold for the onset ofto be made to obtain the 10-impulse and the FI.2,4 of 116 dB P-SEL The empirical function1-impulse SEL values. All animals in this series derived from the narrow-band impulse data iswere tested using the auditory evo'ed poten- seen to differ from the A-weighting functionti-. procedures. by as much as 10 dB at the low frequencies.

Also evident in this figure is the anomalous be-havior of the data point produced by the cx-

Results posures to the 2,450-Hz, CF impulses.

The results of each series of exposure,; are Series II Exposurespresented -eparately, and the methods used toaealyz the NIPTS data from each series are The detailed his:ologic and audiometricexplained, results of this series , f exposures have been

published by Patterson et al (1985, 1986).Series I Exposures The fi.2,4 data from this series of seven expo-

sures is shown plotted as a function of the SELFor each of the 20 groups of animals that and the P-SEL in Figure 30-8. The latter was

were exposed to the narrow-hand impulses. a obtained by applying the empirical weightingmean PTS evaluated at 1, 2, and 4 kHz (M.2.4) function (Fig. 30-7) to consecutive octavewas computed, and the groups were com- bands of the spectrum of the Series I expo-pared on the basis of SEL This data set is sures. Also included in this figure are theshown in Figure 30-5. The group mean PMS shifted (or P-weighted) data points from the

4

340 PARAMETERS OF EXPOSURE

High Poo 147?d8

2DOC

0 S' i's1 20 2

Low Peak 139 a

4001

0 I' 10S ;50 is

Tlme (me)

0-

100 1000 10000

Frsauencyf(H)

Figure 30-2 Examples of the Series U impulses and thewr common spectrm. A, The high-peaked 1 47-dB peAk SPLimpuLsc. A, The law-feaked 139-dB irnpuLse. C The speeruni 0( ec"- of the above, appruximarely equal-energ,impulses.

eee5

AN EXPERIMENTAL BASIS FOR THE ESTIMATION OF AUDITORY SYSTEM HAZAW 341

._ 0 -I0-160 d6 peak SPL

-20"

> 40-

_j ~ 15

"J 60- 4D '1 ý H

L_ -801- 0W a. _ _ _

L-J _t O0 Time (ms)

0.k 0 160 dBpeak SPL

. 4F0ure W3 Exsw uBof die 1605dB pe > -S F. impul se p roduc ed by the th m differ- 0)

So rc

e at s ho c k ti t he s a n d i d i c k r s pe c t i v e s p e c -6 0 . , _ ,

Cibee oTJo are ypa of dKW used >%forthc Series III aPONMas I8. 0 H

0 -IO0 Time (ms)

0 60 0BpeakSPL

-20-

0ýý

60F ret: 2.5 X 1O Jy m /Hz

C 0 Time (mns) 1

01L I 100

Frequency (kHz)

Series I exposures. It is evident that the extrapolated as shown by the dotted portionP-weighting function does not have the de- of the function in Figure 30-7, and then used

sired effect of increasing the degree of congru- to weight the Series II impulses, the agree-ence between the Series I and Il exposures. ment between the Series I and Series 11 dataBecause the Series H exposure had substan- becomes good, as wen in Figure 30-9. A linear

tial energy in the 2-kltz region of the spec- rpgression anal)yss (solid line) of the entiretrum, it was apparent that the effect of apply- data set from the Series I and Series IU expo-ing the empirical weighting function to this sores shows a correlation coefficient of 0.91, a

region of the spectrum would shift the Series slope of 25, and an X-intercept of 116 dB.II data points in the wrong direction. How- This modified weighting function is referredever, if the empirical P-weighting function is to as P-weighting.

6

342 ~PAIAMflMM OF IXFWUU

IN aPak PL N 0- 1

Dam-

0.00 UAscu

0.12-lS d@ pokSPL U .cI

0.10.*~

0.6.figur 30.4 A wvighted octnve band

spectra of each of the waves that were

030

0.1.

.126 .255 I 2 4 a

Octow kW (kHZ)

TABLE 30-3 Exposure Conadions for the Nine Groups Used for

IO.-mplubu Sories III Exposures

SOURCE PEAK SPI. (do) TOTAL S.L (di) TOTAL P'-SEL (dB)

ISO 140.3 129.2I 155 141.8 133.61 160 146.4 138I8!1 ISO 131.4 130.3II 155 136.5 135.3II 160 140.6 131.6*I ISO 129.0 130.8Ut 155 135.0 136.2IMI 160 139.1 139.9

sCavvapnki SEL and Pv-EL vakme for the m0-imulse aid l-bnuls conditins can be obwind by Wwcig e-Pia 1046 a*Aumft

80

o 21OHzd

* 775 H. do l•:Ht €t

60

1

0 13501.ad 0rnA 2460 M d

13A 3550 Ht d

2 40 4

paUhanewl tttnsboid s"if (M~) W&Acvahwed at 1, 2, and 4 kHz Of(M12.4 23 fiUntion Of the tOai I-son expo& lee for th si ( 20,

groups expo2sed to th Senics I~ A U-narrow.bknd fmpuism 0. U c

* 0- A C0

10 120 130 140

Sound Exposure Level (dB)

80.

0 260HI d

s-a 1025 I* c60 1350 lFe d

A 24,504 H ,liA 3.=~ Hz c

Fgure 306 The permanent 40dwthrsod shift at 1, 2, and 4 kHl J(FiA,) as a function of the Rtcmpwficalky-dcrived P-weightedsound exposue kv ld for all the 20.Series I eP6MUrts- The recgrssionline has a slopeo 2.6 and an CAX-intercept of 116 dB. 0

o.y - -*269152.2584.3X

.20.110 120 130 140

P-Weighted Sound Exposure Level

8

34PARAMETERS OF EXPOSURE

30-- -- P-wlhtlng

P.. .. F-weihting

- 4-"h*tg20,

V10.

.C

~0

-10 .100 1000 10000

Frequency (Hz)

Figure 30-7 The empirical P-weighting function derived from the Series I exposures along with the conventionalA-weightng function and the P'-weighting function inferred from the Series 11 and III experiments,

80

* High Pet1A LOW Peak Seris II. P-Wihted

60 + High Peak0 hLow Pe } Series II, Unweighted *

N0 Sries I, P-Weighted

- 40 * • •Figure 3048 The permanent40 threshold shift at 1, 2, and 4 kHz0• (z�2.,4) from the Series 1H

*¢exposures shown as a function oft + * unwcighted and P-weighted sound

S 20K expos'z," Ivc el compared to the

p.. 1.2.4 versus P-weighted sound+o exposure level of the Series I

JO texposures.

-201

110 120 130 140

Sound Exposure Level (dB)

Series III Exposures waves that were similar to some of the im-

pulses in the Series III exposures. AnotherOne problemf that seems to characterize problem is the excessive time necessary to

the measurement of M13 following exposure run an experimental animal through a corn-to these high peak levels of impulse noise is plete experimental paradigm of audiometricextreme iritersubject variability. A number of and histologic protocols, thereby effectivelyauthors have commented on this problem in limiting the number of animals in each exper-the past, including Kryter and Garinther imental group and hence the statistical power.(1965) and Henderson and Hamernik (1982). On the ',asis of a preliminary analysis of thePrice (1983, 1986) also reported large inter- PTS data (using analysis of variance), it wassubject variability when measuring threshold apparent that the effects on PTS of the differ-shifts in cats that had been exposed to blast ent impact presentation rates were, at best,

9

AN EXDWI" NTAL BASIS FOR THE ESTIMATION OF AUDITORY SYSTEM HAZARD 345

8o0 swi I ExGpscwe. P-Weigt

70 0 High Pet. P-WeilghA LOW P**. P-Weigh

S60 0

Figure 3009 Permanent threshold 50

obtained from tie Serie and m. 4Seres a expoures as a function ot 1 40 aP- and P'-weighted sound exposure rlevels, respectively. The linear 0 30.reession line was computed using W0ail the points shown plotted In thisfigure. X-it-fcept = 116 dB; slope r 2S2.5.

S 10 y.- 24.69. 2t53Sx

S• R^2 0.0=3 R. 0.9070.

110 120 130 140

Sound Exposure Level (dB)

marginal statistical effects. Thus, a decisionwas made to evaluate all the PTS data without 2.0 kHregard for presentation rate. Also, because re-

lations between PTS and the increasing energy 60 8of the stimulus were being sought, presenta-tion rate did not affect the independent vari- a 0able. This effectively increased the number of 40

animals at each SEL to 15 except for the l-im- a o N

pulse exposure conditions. Total sound expo- (a , 0%: i

sure br exposure level is increased by increas- 80o 0ing the peak SPL or the number of impulse f08 0presentations. B 8 000 o

For each audiorneric test frequency, the o

ndividual !?:;i,! FPTS at that frequency was .20 - Iplotted as a fu.ctton of the total unweightcd GoSEL in tb,? octv--t band centered on that test 4.0 kH2frequ.. y Twc' examples of this analysis at 2 0Sitz ind 4 klii tkr Cource II are shown in Fig- 'a

ut, 30-10. For impact Sources I. II, and 111, 3105 individual data points fbr each source at . 0 0 0 0 0each audiometric test frequency were plotted In 0 Uover a range of SELs of approximately 30 dB. V 0 *The actual number of data points in each e ; 0 0panel of Figure 30-10 is less than 1O5, because -o 8 0

a number of animals had the same data coordi- 0 o *; • x onate. Using data sets such as those shown in 0 8 §figure 30-10, the 90th percentile hearing loss 0

kP¶&g) was computed for each SEL at each oc- -. ,0tave frequency from 0.5 to 16 kHz. The PTS90 at so 100 tic ,20 130 -oany frequency was computed as follows: Sound Exposure Level (dB)

PTSg-- + st.,o Figure 30-10 Two examples that illustrate the individ-ual animal permanent threshold shift (iris) values at 2

anid 4 kHz MoDwing the Series III exposures to Sourcewhere x is the group mean PIS; t.10 is the il. The solid symboLs represent the 90th percentile val-value of t below which 90 percent of the PTh ues of the PTS at the various exposure energies.

10

346 PARAMETERS OF EXPOSURtE

70 -y -223.54 + 1.9789x

60 AA2 -0.824 R60..0

'' 501 40 0 Oo 90th percentile Permanente•" 40so hrshold shift (PrS.) measured a

1, 2, and 4 kHz for all of heQI 30. roups expsed to the SeriesII

mpulses as a functio o the20 P'-wclghred sound exposure

level. A linear regression analysis

10 0(solid line) yileds aslope of

:0 X.intcrcept of 113 dB.0.

-10 - --110 120 130 140 1SO

P'-Welghted Sound Exposure Level (dB)

80

E sII High PO*a Se iILtow I" Pk

60 AS Srod I

S40 t A Figure 30-12 The meanpmnnancnt th'reshold shift (PTS)

_ produced by exposures to the0 Series 1, 1I, and III Impulses as, a

& £ . function of the P'-welgtcd sound204 exposure level. The equation for&A ifthe linear regresion line (solid

0 line) is also give (s

y- 246.33 + 2.1520X

RA2 0.804 R 0.897

-201100 110 120 130 140 ISO

P'-Welghted Sound Exposure Level (dB)

data lies; s is the group standard deviation, percentile points, a 90th percentile il.2,4 wasThis procedure yields nine percentile points computed for each exposure group and plot-for each test frequency, shown by the filled ted as a function of the P'-weighted SELs (P'-symbols in Figure 30-10, i.e., three peak levels SELs). These results are shown in Figurefor each of three numbers of impacts. This ex- 30-11. The P' weighting has the effect of col.ercise was repeated for each of the six octave lapsing all the shock tube data into a reason-test frequencies and for each of the three ably cohesive pattern for which a linear re-sources. gression produces a relation between pi.2.4

From this set of frequency-specific 90th Prid P'-SEL whose correlation coefficient is

1,.

AN EXPERIMENTAL BASIS FOR THE ESTIMATION OF AUDITORY SYSTEM HAZARD 347

0.91. A threshold for the onset of Fi,2.4 of 113 The empirical P'-weighting function pre-dB SEL and a slope of approximately 2 dB sented in Figure 30-7 has a low-frequency seg-ns1 .2 4 for each decibel of P'-SEL describes the ment (i.e., below 1.5 kHz) with a slope of ap-equation of this regression line. proximately 10 dB per octave, which is

Figure 30-12 shows the entire data set greater than the low-frequency slope of eitherfrom the Series 1, II, and IlI exposures plotted the A-weighting function or the "relative sus-as a function of the P'-SEL. As a first approxi- ceptibility" curve presented by Price (1983),macion the P'-weighting function has the de- This indicates a much smaller hazard from thesired effect of unifying the PTSSEL relation lower-frequency components of the impulsefollowing a diverse series of impulse noise cx- noise spectrum than previously believed.posures. The correlation coefficient between Above 1.5 kHz the A-weighting function is rel-the PIS and weighted SEL variables is approx- atively flat, whereas the Price susceptibilityimately 0.9. curve rises monotonically at about 18 dB per

octave above 3 kHz. The P'-weighting curveprovides no evidence relevant to this part of

Conclusion the spectrum. The unusual feature of the em-pirical P'-weighting function is the 2,450-Hz

We have presented a preliminary analysis point. When the weighting indicated by thisof a large experimental database obtained point is applied to the 2-kHz octave band en-from 475 chinchillas that were exposed to a ergy of the impulse of the Series II or Series IIIvariey of impulse/blast wave noise transients. data, the effect is to decrease the correlationThis analysis, although encouraging in its abi- coefficient between the H3 .2,.4 and the P-SEL.ity to unify the PITS data, is considered prelim- (The actual weighting used at the 2-kHz oc-inary because only a portion of the data that tave band is the value obtained by linear inter-will eventually be available have been ana- polation between the 1,350-Hz and 2,450-Hzlyzed. In addition to the results presented, the data points.) Although the 2,450-Hz point ap-following data sets will ultimately be entered pears to be inconsistent with the rest of theinto the database for a final analysis: (1) non- P'-weighting function, it should be noted thatreverberant, high-frequency, Series III-type this point is the result of a consistent set ofimpulses (N = 105); (2) a more detailed ex- data that was obtained from four different ex-ploration of the 1- to 8-kHz region of the em- posure groups (N = 24). If, however, the P'-pirical weighting function using the Series I weighting function is used-i.e., an attenua-narrow-band impulses (N = 50); (3) highly- tion factor of -5 dB is applied to the 2-kHzreverberant Series Ill-type ihpulses (N = octave band energy of the Series II and Series300); and (4) all sensory cell loss data from IlI impulses-the correlation coefficient be-the above exposures. tween FRU., 4 and the weighted exposure level

The surprising order that is imposed on increases to more than 0.9 (see Figures 30-9the PTM data by the P'-weighting function is and 30-11). This result seems to indicate thatencouraging and tends to lend some validity the appropriate weighting function to be ap-to the methods used in the analysis, i.e., the plied to an impulse spectrum is not a simpleorganization of group mean data averaged monotonic function, as implied by A-weight-ovet several frequencies and, in the Series III ing or the Price susceptibility curve, butexposures, the use of a 90th percentile IMS. rather a more complex function (at least inThe analysis presented would indicate that de- the chinchilla) at frequencies above approxi-spite the problems and inconsistencies in mately I kHz. The data of von Bismarcksome of the data obtained from high-level im- (1967) on the external ear transfer functionpulse noise that have been described in the lit- and the multifrequency impedance data oferature, the use of large samples and the sys- Henderson (personal communication), alongtematic variation of exposure conditions can with the intracochlear pressure measurementsyield a database that reflects some underlying of Patterson et al (1988), would indicate thatorder and can be useful in developing expo- such nonmonotonic behavior is to be ex-sure criteria. These data have shown that us- pected.ing electroacoustic methods and narrow-band In conclusion, if a suitable weighting func-impulses, a weighting function appropriate for tion can be established empirically it couldhigh-level blast waves can be established. This then be applied to the spectrum of an impulseweighting function also may be appropriate to develop an energy-based approach to thefor use in the evaluation of industrial impact establishment of criteria for exposure to anoise data. wide variety of noise transients.

12

3.PARAMETERS OF EXPOSURE

Bases Exptrimentales ReferecesRelatives A Il'Estimation Coles RRA Garinther OR, Rice CO, lodge DC. Hazard-ous exposure deto impuls noise. J Acous ScAdes Risques de 1968; 43:336-343.I'Exposition aux Bruits Gold B, Rader CM. Digital processing of signals. New

York: McGraw-Hill, 1969.

Impulsionnels Hamernik Rt, Abroon WA, Hsueh KD. The energyspectrum of an impulse: Its relation to hearing loss.J Acoust Soc Am 1991; 90 (In press).

Latalyse des rtsultats de deux stries ex- Hame.nik RBP, Hsuch KD. Impulse noise: Some defini-ptrimentales portant sur 1'exposition i deux tions, physical acoustics and other considerations. Jtypes de bruits impulsionnels trks difftrents Acoust Soc Am 1991: 90 (In press).est prtsentie. Les valeurs sont bas6es sur des Henderson D, Hamernik RP. Asymptotic threshold shiftrksultats obtenus sur plus de deux cents ani- from impulse noise. In: Hamernik RP, Henderson D,

maux de laboratoire (chinchillas) chez Salvi RJ, eds. New perspectives on noise-inducedhearing loss. New York: Raven Press, 1982:265.

lesquels les pertcs auditives (PTS) et les Henderson D, Hamernik RP, Salvi RJ. Ahroon WA. Com-pertes de cellules sensorielles (SCL) ont ete parison of auditory-evoked potentials and behav-

mesurtes. Les premikres stries d'expositions ioral thresholds in the normal and noise-exposedfurent rtalistes en utilisant des impulsions rt- chinchilla. Audiology 1983; 22:172-180.alistes caractkristiques des tirs de trois armes Kryter KD, Garinther GR. Auditory effects of acoustic

impulses from firearms. Acta Otolaryngol Suppldilffirentes (type Friedlander). Ces impulsions 1965; 211.sont produites en utilisant trois sources difftr- Kryter KD. The effects of noise on man. New York: Ac-entes actionntes i P'air comprimt (tubes I ademic Press, 1970.choc). Elles comportent une distribution spec- OSHA, Dept of Labor. Occupational aoise exposure.trale d'tnergie de large bande avec des pics de Proposed requirements anI procedures. Federalbandes d'octave pondtrts A situts A 0,25; 1,0' Register 1974; 39(207):155-159.

1 Patterson JH Jr, Lomba-Gautier IM, Curd DL, Hamerniket 2,0 kHz. Les niveaux de a d 0RP. The effect of impulse intensity and the number160 dB SPL Les secondes stries d'impulsions of impulses on hearing and cochlear pathology inttaient synthftistes par ordinateur A partir de the chinchilla. USAARL Report No 85-3, 1985.bandes ýtroites (- 250 Hz) reproduites par Patterson JH Jr. Lomba-Gautier IM, Curd DL Hamemikun haut-parleur de forte puissance, Ces impul- RP. The role of peak pressure in determining the

sions, dont le niveau crtte variait de 124 A auditory hazard of impulse noise. USAARL Report

146 dB SPL avaient des frtquences centrales No 86-7, 1986.Patterson JH Jr, Hamemik RP, Hargett CE. et al. The

de six valeurs difftrentes situes entre 0,15 et hazard of exposure to impulse noise as a function of3,50 kHz. A partir de chacun des deux frequency. USAARL Report 1991 (in press).groupes de rtsultats, un niveau ltsionnel con- Pfander F, Bongartz H, Brinkmann H, Kictz H. Dangerstant, dtfini en termes de PTS et de SCL fiat of auditory impairment from impulse noise: A com-

mis en relation avec le spectre d'tnergie et les parative study of the CHABA damage-risk criteriaand those of the Federal Republic of Germany. J

niveaux d'exposition globaux de chaque expo- Acoust Soc Am 1980; 67:628-633.sition. Les difftrences et les similitudes trou- Price GR. Loss of a'iditory sensitivity following expo-vtes parmi l'ensemble des relations de ce type sure to spectrally narrow impulses. J Acoust Soc Amubtenues avec [tune et l'autre sources 1979; (66:456-465.d'impulsions ainsi que la valeur prdictive de Price GR. Relative hazard of weapons impulse3. Jces relations sont discuttes. Acoust Soc Am 1983; '"3:556-566.

Price GBL Hazard from intense low-frequency acousticimpulses. Acoust Soc Am 1986; 80:1076-1086.

ACKNOWLEDGMENTS Smoorenburg GF. Damage risk criteria for impulsenoise. In: Hamernik RIP, Henderson D, Salvi RJ, eds.

The support of the U.S. Army Medical Re- New perspectives on noise-induced hearing loss.search and Development Command through New York: Raven Press, 1982:471.contracts DAMD- I 7-86-C-6172 and DAMD- 17- Smoorenburg GF. Effects of Impulse Noise. NATO Doc-

ument AC/243 (Panel 8/RSG.6) D/9, 1987.86-C-6139 is gratefully acknowledged. We von Bismarck GV. The sound pressut - transformationwould like to thank C.E. Hargett, Jr. and Dr. function from free-field to the earC'-um of chin-WA. Ahroon for their assistance with the audi- chilla. MS thesis, Massachusetts Institute of Tech-ometric protocol; G. Turrentine for his pa- nology. Cambridge, MA, 1967.tience and skill in preparing the figures; and Young RW. On the energy transported with a soundRenee Johnston and Sandy Nease for preparing pulse. J Acoust Soc Am 1970; 47:441-442.

the manuscript.

Copyright a 1992 by Mosby-Year Book, Inc.A B.C. Decker imprint of Mosby-Year Book, Inc.

13

Initial distribution

Commander, U.S. Army Natick Research, U.S. Army Avionics ResearchDevelopment and Evaluation Center and Development Activity

ATTN: STRNC-MIL (Documents ATTN: SAVAA-P-TFLibrarian) Fort Monmouth, NJ 07703-5401Natick, MA 01760-5040

U.S. Army Communications-ElectronicsNaval Submarine Medical Command

Research Laboratory ATTN: AMSEL-RD-ESA-DMedical Library, Naval Sub Base Fort Monmouth, NJ 07703Box 900Groton, CI 06340 Library

Naval Submarine Medical Research LabCommander/Director Box 900, Naval Sub BaseU.S. Army Combat Surveillance Groton, CT 06349-5900

and Target Acquisition LabATTN: DELCS-D CommanderFort Monmouth, NJ 07703-5304 Man-Machine Integration System

Code 602Commander Naval Air Development Center10th Medical Laboratory Warminster, PA 18974ATTN: AudiologistAPO New York 09180 Commander

Naval Air Development CenterNaval Air Development Center ATTN: Code 602-B (Mr. Brindle)Technical Information Division Warminster, PA 18974Technical Support DetachmentWarminster, PA 18974 Commanding Officer

Harry 0. Armstrong AerospaceCommanding Officer, Naval Medical Medical Research Laboratory

Research and Development Command Wright-PattersonNational Naval Medical Center Air Force Base, OH 45433Bethesda, MD 20814-5044

DirectorDeputy Director, Defense Research Army Audiology and Speech Center

and Engineering Walter Reed Army Medical CenterATTN: Military Assistant Washington, DC 20307-5001

for Medical and Life SciencesWashington, DC 20301-3080 Commander, U.S. Army Institute

of Dental ResearchCommander, U.S. Army Research ATTN: Jean A. Setterstrom, Ph. D.

Institute of Environmental Medicine Walter Reed Army Medical CenterNatick, MA 01760 Washington, DC 20307-5300

14

Naval Air Systems Command U.S. Army Materiel SystemsTechnical Air Library 950D Analysis AgencyRoom 278, Jefferson Plaza II ATTN: AMXSY-PA (Reports Processing)Department of the Navy Aberdeen Proving GroundWashington, DC 20361 MD 21005-5071

Director, U.S. Army Human U.S. Army Ordnance CenterEngineering Laboratory and School Library

ATIN: Technical Library Simpson Hall, Building 3071Aberdeen Proving Ground, MD 21005 Aberdeen Proving Ground, MD 21005

Commander, U.S. Army Test U.S. Army Environmentaland Evaluation Command Hygiene Agency

ATTN: AMSTE-AD-H Building E2100Aberdeen Proving Ground, MD 21005 Aberdeen Proving Ground, MD 21010

Director Technical Library Chemical ResearchU.S. Army Ballistic and Development Center

Research Laboratory Aberdeen Proving Ground, MDATrN: DRXBR-OD-ST Tech Reports 21010--5423Aberdeen Proving Ground, MD 21005

Commander CommanderU.S. Army Medical Research U.S. Army Medical Research

Institute of Chemical Defense Institute of Infectious DiseaseATTN: -SGRD-UV-AO SGRD-UIZ-CAberdeen Proving Ground, Fort Detrick, Frederick, MD 21702MD 21010-5425

Commander, U.S. Army Medical Director, Biolopi--alResearch and Development Command Sciences DivisionATTN: SGRD-RMS (Ms. Madigan) Office of Naval ResearchFort Detrick, Frederick, MD 21702-5012 600 North Quincy Street

Arlington, VA 22217

Director CommanderWalter Reed Army Institute of Research U.S. Army Materiel CommandWashington, DC 20307-5100 ATTN: AMCDE-XS

5001 Eisenhower AvenueHQ DA (DASG-PSP-O) Alexandria, VA 223335109 Leesburg PikeFalls Church, VA 22041-3258

CommandantHarry Diamond Laboratories U.S. Army AviationATTN: Technical Information Branch Logistics School ATTN: ATSQ-TDN2800 Powder Mill Road Fort Eustis, VA 23604Adelphi, MD 20783-1197

1i

Headquarters (ATMD) CommanderU.S. Army Training U.S. Army Aviation Systems Command

and Doctrine Command ATTN: SGRD-UAX-AL (MAJ Gillette)ATTN: ATBO-M 4300 Goodfellow Blvd., Building 105Fort Monroe, VA 23651 St. Louis, MO 63120

Structures Laboratory Library U.S. Army Aviation Systems CommandUSARTL-AVSCOM Library and Information Center BranchNASA Langley Research Center ATTN: AMSAV-DILMail Stop 266 4300 Goodfellow BoulevardHampton, VA 23665 St. Louis, MO 63120

Naval Aerospace Medical Federal Aviation AdministrationInstitute Library Civil Aeromedical Institute

Building 1953, Code 03L Library AAM-400APensacola, FL 32508-5600 P.O. Box 25082

Oklahoma City, OK 73125Command SurgeonHQ USCENTCOM (CCSG) CommanderU.S. Central Command U.S. Army AcademyMacDill Air Force Base FL 33608 of Health Sciences

ATTN: LibraryAir University Library Fort Sam Houston, TX 78234(AUL/LSE)Maxwell Air Fore Base, AL 36112 Commander

U.S. Army Institute of Surgical ResearchU.S. Air Force Institute ATTN: SGRD-USM (Jan Duke)

of Technology (AFIT/LDEE) Fort Sam Houston, TX 78234-6200Building 640, Area BWright-Patterson AAMRL/HEXAir Force Base, OH 45433 Wright-Patterson

Air Force Base, OH 45433Henry L TaylorDirector, Institute of Aviation John A. Dellinger,University of Illinois-Willard Airport Southwest Research InstituteSavoy, IL 61874 P. 0. Box 28510

San Antonio, TX 78284Chief, Nation Guard BureauATTN: NGB-ARS (COL Urbauer) Product ManagerRoom 410, Park Center 4 Aviation Life Support Equipment4501 Ford Avenue ATTN: AMCPM-ALSEAlexandria, VA 22302-1451 4300 Goodfellow Boulevard

St. Louis, MO 63120-1798

1.6

Commander Aeromechanics LaboratoryU.S. Army Aviation U.S. Army Research and Technical Labs

Systems Command Ames Research Center, M/S 215-1ATTN: AMSAV-ED Moffett Field, CA 940354300 Goodfellow BoulevardSt. Louis, MO 63120

Sixth U.S. ArmyCommanding Officer ATTN: SMANaval Biodynamics Laboratory Presidio of San Francisco, CA 94129P.O. Box 24907New Orleans, LA 70189-0407 Commander

U.S. Army Aeromedical CenterAssistant Commandant Fort Rucker, AL 36362U.S. Army Field Artillery SchoolATTN: Morris Swott Technical Library U.S. Air Force SchoolFort Sill, OK 73503-0312 of Aerospace Medicine

Strughold Aeromedical Library TechnicaiCommander Reports Section (TSKD)U.S. Army Health Services Command Brooks Air Force Base, TX 78235-5301AITN: HSOP-SOFort Sam Houston, TX 78234-6000 Dr. Diane Damos

Department of Human FactorsHQ USAF/SG1PT ISSM, USCBolling Air Force Base, DC 20332-6188 Los Angeles, CA 90089-0021

U.S. Army White SandsU.S. Army Dugway Proving Ground Missile RangeTechnical Library, Building 5330 ATTN: STEWS-IM-STDugway; UT 84022 White Sands Missile Range, NM 88002

U.S. Army Yuma Proving Ground U.S. Army Aviation EngineeringTechnical Library Flight ActivityYuma, AZ 85364 ATTN: SAVTE-M (Tech Lib) Stop 217

Edwards Air Force Base, CA 93523-5000AFFTC Technical Library6510 TW/TSTL Ms. Sandra G. HartEdwards Air Force Base, Ames Research CenterCA 93523-5000 MS 262-3

Moffett Field, CA 94035CommanderCode 3431 Commander, Letterman Army InstituteNaval Weapons Center of ResearchChina Lake, CA 93555 ATI'N: Medical Research Library

Presidio of San Francisco, CA 94129

17

Commander Directorate of Training DevelopmentU.S. Army Medical Materiel Building 502

Development Activity Fort Rucker, AL 36362Fort Detrick, Frederick, MD 21702-5009

ChiefCommander USAHEL/USAAVNC Field OfficeU.S. Army Aviation Center P. 0. Box 716Directorate of Combat Developments Fort Rucker, AL 36362-5349Building 507Fort Rucker, AL 36362 Commander U.S. Army Aviation Center

and Fort RuckerU. S. Army Research Institute ATTN: A1ZQ-CGAviation R&D Activity Fort Rucker, AL 36362ATITN: PERI-IRFort Rucker, AL 36362 Commander/President

TEXCOM Aviation BoardCommander Cairns Army Air FieldU.S. Army Safety Center Fort Rucker, AL 36362Fort Rucker, AL 36362

MAT Terry NewmanU.S. Army Aircraft Development Canadian Army Liaison Office

Test Activity Building 602ATTN: STEBG-MP-P Fort Rucker, AL 36362Cairns Army Air FieldFort Rucker, AL 36362 German Army Liaison Office

Building 602Commander U.S. ,Amy Medical Research Fort Rucker, AL 36362

and Development ConmmandATTN: SGRD-PLC (COL Sedge) LTC Patrice CottebruneFort Detrick, Frederick, MD 21702 French Army Liaison Office

USAAVNC (Building 602)MAJ John Wilson Fort Rucker, AL 36362-5021TRADOC Aviation LOEmbassy of the United States Australian Army Liaison OfficeAPO New York 09777 Building 602

Fort Rucker, AL 36362Netherlands Army Liaison OfficeBuilding 602 Dr. Garrison RapmundFort Rucker, AL 36362 6 Burning Tree Court

Bethesda, MD 20817British Army Liaison OfficeBuilding 602 Commandant Royal Air ForceFort Rucker, AL 36362 Institute of Aviation Medicine

Farnborough Hampshire GU14 65Z UKItalian Army Liaison OfficeBuilding 602Fort Rucker, AL 36362

'.8

Commander Col. Otto Schramm FilhoU.S. Army Biomedical Research c/o Brazilian Army Commission

and Development Laboratory Office-CEBWATTN: SGRD-UBZ-I 4632 Wisconsin Avenue NWFort Detrick, Frederick, MD 21702 Washington, DC 20016

Defense Technical Information Center Dr. Christine SchlichtingCameron Station Behavioral Sciences DepartmentAlexandra, VA 22313 Box 900, NAVUBASE NLON

Groton, CT 06349-5900Commander, U.S. Army Foreign Science

and Technology Center COL Eugene S. Channing, O.D.AIFRTA (Davis) Brooke Army Medical Center220 7th Street, NE ATTN: HSHE-EAH-OCharlottesville, VA 22901-5396 Fort Sam Houston, TX 78234-6200

Director, LTC Gaylord Lindsey (5)Applied Technology Laboratory USAMRDC Liaison at AcademyUSARTL-AVSCOM ef Health SciencesAITN: Library, Building 401 ATTN; HSHA-ZAC-FFort Eustis, VA 23604 Fort Sam Houston, TX 78234

U.S. Air Force Armament Aviation Medicine ClinicDevelopment and Test Center TMC #22, SAAF

Eglin Air Force Base, FL 32542 Fort Bragg, NC 28305

Commander, U.S. Army Missile Dr. A. Kornfield, PresidentCommand Biosearch Company

Redstone Scientific Information Center 3016 Revere RoadATrN: AMSMI-RD-CS-R Drexel Hill, PA 29026

/ILL DocumentsRedstone Arsenal, AL 35898 NVEOD

AMSEL-RD-ASIDDr. H. Dix Christensen (Attn: Trang Bui)Bio-Medical Science Building, Room 753 Fort Belvior, VA 22060Post Office Box 26901Oklahoma City, OK 73190

U.S. Army Research and TechnologyLaboratories (AVSCOM)

Propulsion Laboratory MS 302-2NASA Lewis Research CenterCleveland, OH 44135

19