NATL INST. OF STAND & TECH R.LC.

A11103

NIST

PUBUCATIONS

^ ^

NISTIR 4511

i/

1 Measurement of Structural Responsef'

I

Characteristics of Full-Scale Buildings:

i Selection of Structures

c

t

/

NISTIR 4511

Measurement of Structural Response

Characteristics of Full-Scale Buildings:

Selection of Structures

R. D. Marshall

Long T. Phan

M.Celebi**

*Branch of Engineering Seismology and Geology

U.S. Geological Survey

Menlo Park, CA

February 1991

U.S. Department of CommerceRobert A. Mosbacher, Secretary

National Institute of Standards and TechnologyJohn W. Lyons, Director

Building and Fire Research Laboratory

Gaithersburg, MD 20899

11

ABSTRACT

This report describes the selection of existing building structuresfor subsequent field measurements of low-level ambient vibrations.By comparing measurement results with previously recorded high-level responses, it is anticipated that guidance can be developedfor improved measurement practice. The buildings selected for thiseffort represent a cross-section of contemporary structural systemsand materials, foundation types, and a range of building heightsand aspect ratios. Each building was subjected to strong shakingduring the Loma Prieta Earthquake of October 17, 1989.

Keywords: buildings, earthquake, instrumentation, dynamic response,field measurements, structural dynamics

IV

TABLE OF CONTENTS

ABSTRACT iii

TABLE OF CONTENTS v

INTRODUCTION 1

BACKGROUND 1

TEST PROGRAM 2

CANDIDATE STRUCTURES 2

GROUND RESPONSE RECORDS 8

STRUCTURAL RESPONSE DATA 13

FIELD MEASUREMENTS 13

ACKNOWLEDGMENTS 15

REFERENCES 15

V

VI

INTRODUCTION

This report presents a summary description of building structuresin the San Francisco Bay area that have been selected for use in astudy of dynamic characteristics. Specifically, the validity ofsmall-amplitude testing methods for the assessment of dynamiccharacteristics will be studied. Each of the buildings describedherein was instrumented with strong-motion accelerometers prior tothe Loma Prieta Earthquake of October 17, 1989 (M^ = 7.1) andyielded excellent sets of response records during the earthquake.The buildings have been selected to cover a range of structuralsystems, construction materials, aspect ratios, and foundationtypes

.

BACKGROUND

Reliable estimates of modal frequencies, stiffness and damping ofstructural systems are essential for predicting the dynamicresponse under loading conditions associated with serviceability orstructural safety. Although response characteristics of isolatedstructural components usually can be predicted with acceptableaccuracy using simple models, this is not the case for completestructures where primary and secondary systems can interact incomplex ways. Examples of such interaction include theparticipation of cladding in overall system stiffness and damping,the participation of infill walls and non-structural partitionwalls, and the interaction of floor systems with columns andbearing walls.

Analytical models for calculating building response usually involvesubstantial simplification of the real structure and, for variousreasons, detailed physical scale models for controlled testing inthe laboratory often prove impractical. Consequently, each ofthese approaches must depend on reliable full-scale fieldmeasurements for validation and/or improvement of the modelingtechnique. Changes in the dynamic properties of structuressubjected to strong seismic excitation are related to the intensityand extent of damage and, therefore, reliable full-scale responsemeasurements are of potential use in the assessment of structuraldamage

.

Numerous studies of structural response have been carried out ontall buildings, long-span bridges, and large dams. The most widelyused technique has been to rely on ambient vibrations forexcitation or, in certain cases, to use one or more mechanicalshakers to excite specific modes of vibration in the structure. Inrare cases response measurements have been obtained during extremeevents such as earthquakes or wind storms. Generally, thesemeasurements suggest a strong dependence of response parameters ondisplacement amplitude, thus raising questions about the validityand proper interpretation of response measurements obtained under

1

low levels of excitation. It is not unknown for seeminglyidentical structures to exhibit basic characteristics such asstiffness that differ by as much as 50 percent (Ellis and Littler,1987) . In the case of structural damping, determinations maydiffer by a factor of two or more. How much of this disparity canbe attributed to actual physical differences and how much can beattributed to measurement errors and incorrect interpretation oftest results is unknown.

TEST PROGRAM

The Loma Prieta Earthquake of October 17, 1989, provided a uniqueopportunity to carry out a program of field measurements andanalytical modeling of a select number of existing structures tobetter understand the significance of factors such as displacementamplitude in full-scale response measurements. In particular, itshould be possible to interpret characteristics measured under low-level response in the light of known (recorded) high-levelresponse. The approach will be to select a number of undamagedbuildings that were subjected to strong ground shaking during theLoam Prieta Earthquake and for which valid response records areavailable. It should be possible from these records to obtainestimates of overall damping, the first two or three modalfrequencies and the peak accelerations and displacements.

Existing analytical models can be used to estimate transverse andtorsional modes of vibration and to predict response to low-levelexcitation. These predicted responses can then be used as a basisfor selecting and installing suitable instrumentation with which toconduct low-level response measurements. To the extent possible,instrumentation systems installed in the buildings at the time ofthe Loma Prieta Earthquake will be utilized in the fieldmeasurement program.

CANDIDATE STRUCTURES

Buildings that are considered to be good candidates for additionalstudy are listed in Table 1 and are located on the local area mapof Figure 1. This map also indicates the major ground failurescaused by the Loma Prieta Earthquake. The instrumentation systemsin these candidate structures were installed and are maintained bythe United States Geological Survey (USGS) and by the CaliforniaDivision of Mines and Geology (CDMG) . The structural systemsinclude steel frames, reinforced concrete moment frames andcombinations of the two. No attempt has been made to includeshear-wall systems in the selection although some of the structuralsystems include shear wall or core walls in the lower stories ofthe structure. The data included in Table 1 were extracted frompost-earthquake reports by Maley et al., 1989, and by Shakal etal., 1989. Photographs of the buildings are included in Figures 2-6 .

2

BUILDING/

STRUCTURAL

SYSTEM

BUILDING

BASE

TYPICAL

FLOORS

FOUNDATION

YEAR

YEAR

INSTRUMENTED

HEIGHT

DIMENSION

FLOOR

PLAN

ABOVE/BELOW

TYPE

DESIGNED

CONSTR.

BY

(ft)

(ft)

(ft)

GROUND

o o CO COs s O OQ Q c/5 COO CJ D 2

I

O'

00

COp^

I

CNJ

pvO'

cn00ON

u C

C ua (Q do

00 u) cC Cu O

3 w“d 00^ C>

•d 0c oM m in e

0) D > o DP D z P

>; d HH Co

•d d u D Q uu u D

T) d 1—

(

Pc 4J c 3 CO 4J

>N d c d ca U cd D 0 CJ DZ ID E u w z E (9

u 0 d u 0O 4J 0 E 1—

(

D Q E cz d y tn 0 b 4J D uHH D 0 Pu D 4J hH D O 4J 0O D 4J f-4 —4 o D (9 to 3 <9 4J z D t4H

hJ > p D d 1+4 V4 pa •»4 D HH D •*4

hH c3 U 3 >• iJ C 00 c U Q U <2 c V4 y d H CO u C u D u y hJ U f9 MCQ z u i+H c u c Z 0 u 3 0 C HH CO CJ <

0 0 d 0 ro 00 M-f 4J hH C m 0 z zZ U y D y o c to h D y 133 Cu.

o 4J 1^ c Z D u ,—4 pH > u 0 DCO D d to to d (9 to O < c9 d < D y EH s V <; d CJ D CJ D CJ E <0 f9 ct

< CJ 0 y D 0 ry D V4 >. y HH 0 U <Z e u 4J 4J z U 0 u Oi 00 y 'M cuH c 0 D J D u u c 0 w u cCO u •d <4-1 U 4J u u to c CJ D 3 M-l y c f9 u>—

»

0 u D c y m to 0 D Qi r; U c 0 U D HHz d D c CM < D n E CO W a CO CO z Ui D in (44

hH 3 U D 0 r*. H 3 0 y D 2 4J HH

z >N CO U o to z z 2 O C Di < o C CO UQ d 0 < O c9 0 o O c9 0 {X o f9 0 << u z z in P^ CO 2 u C3N CO 2 H v£) CO 2 CL.

w 6

to q;

00 oc

D u a> 4J

c0) mh

<

^ 4H 0} 0)

u > c s ^>N •-» nj oO

cn u ly u oj

\0 0/ 0^ '-H "TD

e

00-ToCO

o

0)

o2

CDMG

=

California

Division

of

Mines

and

Geology

USGS

=

United

States

Geological

Survey

Pacific

Ocean

SAN ^

ANDREAS, FAULT

^ Selected Measurement Sites

• Ground Failure

o Fault Zone Ground CrackingX Liquefaction— Tension Crack

® EpicenterMonterey

HAYWARDFAULT

CALAVERASFAULT

SARGENTFAULT

30 KM

Figure 1: Locations of Selected Measurement Sites and Major Faults.

4

Figure 2: Administration Building, CSUH, Hayward.View to northwest.

5

Figure 3: Santa Clara County Office Building, San Jose.

View to southeast.

Figure 4: Commercial Office Building, San Bruno. View to southeast

6

Figure 5: Transamerica Building, San Francisco. View to southeast.

7

Figure 6: Pacific Park Plaza, Emeryville. View to west.

GROUND RESPONSE RECORDS

Table 2 contains a summary of peak ground accelerations at basementlevel in each of the five selected buildings and at strong-motionsites in the vicinity of the selected buildings. In most cases thesites represent free-field conditions, not being affectedsignificantly by nearby buildings or other structures. In somecases the instrumentation is either in or adjacent to a single-story building such as a fire house. For the Transamerica Buildingin San Francisco, the strong-motion sites at Diamond Heights,Pacific Heights and Telegraph Hill are located within two-storybuildings.

8

TABLE 2 SUMMARY OF LOMA PRIETA GROUND RESPONSE RECORDS AT ORNEAR BUILDINGS SELECTED FOR ADDITIONAL STUDY

STATION NAME EPICENTRAL ACCELERATIONDISTANCE COMPONENTS

(km) (g)

ADMINISTRATION BUILDING - CSUH 70N. Lat. = 37.655W. Long. = 122.056Site Geology: Franciscan metavolcanic

rock

050 = 0.09Up =0.05320 = 0.08

Fremont - Mission San Jose 55N. Lat. = 37.530W. Long. = 121.919Site Geology: Alluvium

Hayward - CSUH Stadium Grounds 71N. Lat. = 37.657W. Long. = 122.061Site Geology: Franciscan greenstone

Hayward - Muir School 71N. Lat. = 37.657W. Long. = 122.082Site Geology: Alluvium

Hayward - BART Station Parking Lot 73N. Lat. = 37.670W. Long. = 122.086Site Geology: Alluvium

090 = 0.11Up =0.09360 = 0.13

090 = 0.08Up =0.05360 = 0.08

090 = 0.14Up =0.10360 = 0.18

310 = 0.16Up =0.08220 =0.16

SANTA CLARA COUNTY OFFICE BUILDING 35N. Lat. = 37.353W. Long. = 121.903Site Geology: Alluvium

San Jose - Santa Teresa Hills 21N. Lat. = 37.210W. Long. = 121.803Site Geology: Alluvium over serpentine

Saratoga - Aloha Avenue 27N. Lat. = 37.255W. Long. = 122.031Site Geology: Alluvium

067 = 0.10Up =0.10337 = 0.11

225 = 0.28Up =0.22315 = 0.27

090 = 0.34Up =0.41360 = 0.53

9

(Table 2 continued)

STATION NAME EPICENTRALDISTANCE

(km)

San Jose Interchange - 101/280/680 34N. Lat. = 37.340W. Long. = 121.851Site Geology: Not available

Halls Valley - Grant Park 37N. Lat. = 37.338W. Long. = 121.714Site Geology: Alluvium

Agnew - Agnew State Hospital 40N. Lat. = 37.239W. Long. = 121.952Site Geology: Alluvium

Sunnyvale - Colton Avenue 43N. Lat. = 37.402W. Long. = 122.024Site Geology: Not available

OFFICE BUILDING - SAN BRUNO 81N. Lat. = 37.628W. Long. = 122.424Site Geology: Alluvium

Upper Crystal Springs Reservoir 63(Skyline Blvd.)N. Lat. = 37.465W. Long. = 122.343Site Geology: Sandstone

Upper Crystal Springs Reservoir 63(Pulgas Water Temple)N. Lat. = 37.49W. Long. = 122.31Site Geology: Sandstone

Foster City - Redwood Shores 63N. Lat. = 37.55W. Long. = 122.23Site Geology: Alluvium to 210 m,

serpentine

ACCELERATIONCOMPONENTS

(g)

232 = 0.13Up =0.08322 = 0.18

090 = 0.11Up =0.06360 = 0.13

090 = 0.16Up =0.10360 = 0.17

270 = 0.19Up =0.10360 = 0.22

245 = 0.12Up =0.12335 = 0.14

090 = 0.09Up =0.04360 = 0.10

090 = 0.09Up =0.06360 = 0.16

090 = 0.29Up =0.11360 = 0.26

10

(Table 2 continued)

STATION NAME EPICENTRALDISTANCE

(km)

ACCELERATIONCOMPONENTS

(g)

San Francisco International Airport 79 090 = 0.33N. Lat. = 37.622 Up =0.05W. Long. = 122.398 360 = 0.24Site Geology: Deep alluvium

South San Francisco - Sierra Point 84 115 = 0.06N. Lat. = 37.674 Up =0.05W. Long. = 122.388 205 = 0.06Site Geology: Not available

TRANSAMERICA BUILDING 97 083 = 0.12N. Lat. = 37.80 Up = 0.07W. Long. = 122.40Site Geology: Not available

351 0.11

San Francisco - Golden Gate Bridge 100 270 = 0.24N. Lat. = 37.806 Up = 0.06W. Long. = 122.472Site Geology: Not available

360 0.12

San Francisco - Presidio 98 090 = 0.21N. Lat. = 37.792 Up = 0.06W. Long. = 122.457Site Geology: Serpentine

360 0.10

San Francisco - Diamond Hts. 92 090 = 0.12N. Lat. = 37.74 Up = 0.05W. Long. = 122.43Site Geology: Franciscan chert

360 0.10

San Francisco - Pacific Hts. 97 270 = 0.06N. Lat. = 37.79 Up = 0.03W. Long. = 122.43Site Geology: Franciscan sandstone,

shale

360 0.05

San Francisco - Telegraph Hill 97 090 = 0.08N. Lat. = 37.80 Up = 0.03W. Long. = 122.41 360 = 0.06site Geology: Franciscan shale,

sandstone

11

(Table 2 continued)

STATION NAME EPICENTRALDISTANCE

(km)

ACCELERATIONCOMPONENTS

(g)

San Francisco - Rincon Hill 95N. Lat. = 37.79W. Long. = 122.39Site Geology: Franciscan sandstone,

shale

090 = 0.09Up =0.03360 = 0.08

Yerba Buena Island 95N. Lat. = 37.81W. Long. = 122.36Site Geology: Franciscan sandstone

Treasure Island 98N. Lat. = 37.825W. Long. = 122.373Site Geology: Fill

090 = 0.06Up =0.03360 = 0.03

090 = 0.16Up =0.02360 = 0.11

PACIFIC PARK PLAZA - EMERYVILLE 97N. Lat. = 37.844W. Long. = 122.295Site Geology: Silty fine sand fill and

Bay mud on hard silty clay

Pacific Park Plaza 97Free Field - South

080 = 0.21Up =0.06350 = 0.17

090 = 0.26Up =0.06360 = 0.22

Pacific Park PlazaFree Field - North

97 090 = 0.22Up =0.09360 = 0.20

Piedmont - Piedmont Jr. High Grounds 93N. Lat. = 37.823W. Long. = 122.233Site Geology: Weathered serpentinite

Berkeley - UCB Memorial Stadium Grnds. 98N. Lat. = 37.87W. Long. = 122.25Site Geology: Not available

Berkeley, U.C. - Strawberry Canyon 98N. Lat. = 37.87W. Long. = 122.24Site Geology: Not available

045 = 0.08Up =0.03315 = 0.07

135 = 0.13Up =0.03225 = 0.07

135 = 0.04Up =0.02045 = 0.08

12

(Table 2 continued)

STATION NAME EPICENTRALDISTANCE

(km)

ACCELERATIONCOMPONENTS

(g)

Berkeley - Lawrence Berkeley Lab 99 090 = 0.12N. Lat. = 37.876 Up = 0.04W. Long. = 122.249 360 = 0.05Site Geology: Thin alluvium ion shale.

siltstone

Richmond - 2501 Rydin Road 101 057 = 0.08N. Lat. = 37.884 Up = 0.04W. Long. = 122.302 327 = 0.11Site Geology: Not available

Richmond - City Hall Parking Lot 108 280 = 0.11N. Lat. = 37.935 Up = 0.04W. Long. = 122.342 190 = 0.13Site Geology: Alluvium

STRUCTURAL RESPONSE DATA

Representative structural response data obtained during the LomaPrieta Earthquake are listed in Table 3. A considerable amount ofresponse data from these and other structures has not yet beenprocessed or analyzed. Also included in Table 3 are selectedresults of forced vibration studies carried out prior to LomaPrieta and measurements of ambient vibrations carried out by USGSand NIST personnel on May 10-11, 1990.

FIELD MEASUREMENTS

To the extent possible, use will be made of the permanentlyinstalled instrumentation in the candidate structures. Theadvantages of this approach are that it allows a direct comparisonof low-level measurements with high-level measurements andconsiderably reduces the amount of time and effort required tocarry out the field measurements. The major disadvantage is thatthe sensitivities of the permanently installed sensors may incertain cases be too low to obtain reliable accelerationmeasurements. In most cases the permanently installed sensors areforce-balance accelerometers (FBA's) with a nominal sensitivity of2.5 volts/g and a resolution of approximately 1 micro-g. Ifnecessary, additional point measurements can be obtained usingportable accelerometers provided by USGS.

The NIST data acquisition system to be used in the fieldmeasurement program has an input range of + 10 volts and employs

13

BUILDING

DISTANCE

NUMBER

LOMA

PRIETA

EARTHQUAKE

RESPONSE

DATA

NATURAL

FREQ.

FROM

OF

DATA

PEAK

ACCELERATION

(r)

NATURAL

AMBIENT

SURVEY

EPICENTER

CHANNELS

GROUND

ROOF

FREQ.

(This

Study)*

(km)

(Hz)

(Hz)

W 3I i

Z U]

3 u\ I (U

2 W >

CM (N

o o o

w 3 00 3I I I I

Z til z Uoo 3 c/5 3

I I I I

Z ti4 2 UC/5 3 CO 3

I I I I

2 U 2 oq 2 W 2 Wo uo in T)

CO cn -4- Ocn CO CO cn Z

4J O O O O AJ o o o o -T <f -T

CO 32 til

14

CO 3 5-«

I I 0)

2 W >

rH vO.-(NOo o o

1 u 2 C«

uo 1—

(

Q C2 ,-1 U CO

»—

i

U4 -H nJ 0 0Q 0 fC => z 4-j C

00 00 C 00 t—

1

—4 00 u oO

0) > G V u QJ Q a; < 0 OJ

O C Q t-l u Q X 0 Q n5 Q tsl >44 QCQ U z 0 u 4-1 h-

1

a <0 O =) M •-d n 3 1-4 0

Z C^4 0 CO tt4 CO rQ in cO no CO u CO U4 cn 0 cn CJ cn-H cn 0 CJ < CJ

H 05 II <; D II II U in II os II

< O a M •«-4 < o;

2 X < -C < 0 x: 0 JZ a. -CH w j (V -> hH c u u c 4J ^-4 win TJ u u w u a G u (fl M u •-4 l4

M U 0 0 0 od u 0 u 0 k-i > 0Z 2 < n 2 [ij CQ 2 t/3 tK4 2 fa 2hH H ? z )—

(

u2 -Q Z G JD 2 G X < G -Q u a; jQQ O < ci5 0 0 CO 0 X CO 0 < e 0< X t/3 CO CJ CO n (r> CO fa w n

NOTE:

Forced

Vibration

a 12-bit A/D converter capable of performing up to 250,000conversions per second. A total of 16 data channels can bemultiplexed at the 250,000/sec rate. Each data channel is providedwith a differential input amplifier, a 3-pole active filter, and asample-and-hold amplifier. Thus the samples obtained in each scanof the data channels have zero time skew. The differentialamplifiers have nine jumper-selectible gains covering the rangexl to x500. If desired, DC components in the amplified signal canbe nulled with a programmable offset voltage of up to 4 0 percent ofthe dynamic range. The effective dynamic range of the dataacquisition system is extended by a programmable-gain amplifier inthe A/D section with a gain selection of 1, 2, 4, or 8 . The filtertype and cutoff frequency can be changed by changing resistor andcapacitor networks on the signal conditioning cards. Theconfiguration to be used in this measurement program is aButterworth low-pass filter with a cutoff frequency of 10 Hz. Thesample rate will be 50 samples per second per channel. The minimumrecord length will be 2,048 samples or approximately 40 seconds.It is anticipated that several minutes of recordings will becollected at each field site.

ACKNOWLEDGMENTS

The authors wish to acknowledge the assistance of Dr. M. J. Huang,Office of Strong Motion Studies, California Division of Mines andGeology, who provided valuable information on structural systemsand Loma Prieta response data for the field sites. Mr. MarionSalsman, Branch of Engineering Seismology and Geology, U.S.Geological Survey, provided valuable assistance in conducting thefield measurements.

REFERENCES

1. Celebi, M. , et al. (1985). "Integrated Instrumentation Planfor Assessing the Seismic Response of Structures - A Review ofthe Current USGS Program." U.S. Geological Survey Circular947, 1985, 38 pages.

2. Celebi, M. and Safak, E. (1990). "Recorded Seismic Responseof Transamerica Building. Part I - Data and PreliminaryAnalysis." Paper submitted to ASCE Journal of StructuralEngineering . 1990.

3. Ellis, B.R. and Littler, J.D. (1987). "Lessons From DynamicTesting of Buildings." Structural Assessment . F. Garas et al.Eds., Butterworths , London, 1987, pp 63-70.

4. Maley, R. , et al. (1989). "U.S. Geological Survey Strong-Motion Records From the Northern California (Loma Prieta)Earthquake of October 17, 1989." Geological Survey Open-FileReport 89-568, October 1989, 85 pages.

15

5. Shakal, A.F., et al. (1989). "CSMIP Strong-Motion RecordsFrom the Santa Cruz Mountains (Loma Prieta)

,

CaliforniaEarthquake of October 17, 1989." Report No. OSM 89-03,November 1989.

16

NIST-114A U.S. DEPARTMENT OF COMMERCE(REV. 3-90) NATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY

BIBLIOGRAPHIC DATA SHEET

1. PUBUCATION OR REPORT NUMBER

NISTIR 45112. PERFORMING ORGANIZATION REPORT NUMBER

3. PUBUCATION DATE

February 19914. TITLE AND SUBTITLE

Measurement of Structural Response Characteristics of Full-ScaleBuildings: Selection of Structures

5. AUTHOR(S)

Richard D. Marshall, Long T. Phan, M. Celebi6. PERFORMINQ ORGANIZATION OF JOINT OR OTHER THAN NIST. SEE INSTRUCTIONS)

U.S. DEPARTMENT OP COMMERCENATIONAL INSTITUTE OP STANDARDS AND TECHNOLOGYGAITHERSBURG, MD 20899

7. CONTRACT/GRANT NUMBER

8. TYPE OP REPORT AND PERIOD COVERED

9. SPONSORING ORGANIZATION NAME AND COMPLETE ADDRESS (STREET, CITY, STATE, ZIP)

10. SUPPLEMENTARY NOTES

11. ABSTRACT (A 200-WOR0 OR LESS FACTUAL SUMMARY OF MOST SIGNIFICANT INFORMATION. IF DOCUMENT INCLUDES A SIGNIFICANT BIBUOGRAPHY ORUTERATURE SURVEY, MENTION IT HERE.)

This report describes the selection of existing building structures forsubsequent field measurements of low-level ambient vibrations. Bycomparing measurement results with previously recorded high-levelresponses, it is anticipated that guidance can be developed for improvedmeasurement practice. The buildings selected for this effort representa cross-section of contemporary structural systems and materials,foundation types, and a range of building heights and aspect ratios. Eachbuilding was subjected to strong shaking during the Loma Prieta Earthquakeof October 17, 1989.

12. KEY WORDS (6 TO 12 ENTRIES; ALPHABETICAL ORDER; CAPITAUZE ONLY PROPER NAMES; AND SEPARATE KEY WORDS BY SEMICOLONS)

buildings, earthquake, instrumentation, dynamic response, fieldmeasurements, structural dynamics

13. AVAILABiUTY

UNUMITED

FOR OFFICIAL DISTRIBUTION. DO NOT RELEASE TO NATIONAL TECHNICAL INFORMATION SERVICE (NTIS).

ORDER FROM SUPERINTENDENT OF DOCUMENTS. U.S. GOVERNMENT PRINTING OFFICE,WASHINGTON, DC 20402.

ORDER FROM NATIONAL TECHNICAL INFORMATION SERVICE (NTIS). SPRINGFIELD, VA 22161.

14. NUMBER OF PRINTED PAGES

24

15. PRICE

AO 2

ELECmONIC FORM

.iil

. 'f

5u#"’3UT^' 1

'

.,, 's;,.^ .‘jyf-^sa .*.>' '

; V

'

tV:”

; !':

<-,