The following notice applies to any unclassified (including ... Analysis of the M60...of the M60 Main Battle Tank and the M151 %-Ton Utility Truck from Operating and Maintenance Data

The following notice applies to any unclassified (including originally classified and now declassified) technical reports released to "qualified U.S. contractors" under the provisions of DoD Directive 5230.25, Withholding of Unclassified Technical Data From Public Disclosure.

NOTICE TO ACCOMPANY THE DISSEMINATION OF EXPORT-CONTROLLED TECHNICAL DATA 1. Export of information contained herein, which includes, in some circumstances, release to foreign nationals within the United States, without first obtaining approval or license from the Department of State for items controlled by the International Traffic in Arms Regulations (ITAR), or the Department of Commerce for items controlled by the Export Administration Regulations (EAR), may constitute a violation of law. 2. Under 22 U.S.C. 2778 the penalty for unlawful export of items or information controlled under the ITAR is up to ten years imprisonment, or a fine of $1,000,000, or both. Under 50 U.S.C., Appendix 2410, the penalty for unlawful export of items or information controlled under the EAR is a fine of up to $1,000,000, or five times the value of the exports, whichever is greater; or for an individual, imprisonment of up to 10 years, or a fine of up to $250,000, or both. 3. In accordance with your certification that establishes you as a "qualified U.S. Contractor", unauthorized dissemination of this information is prohibited and may result in disqualification as a qualified U.S. contractor, and may be considered in determining your eligibility for future contracts with the Department of Defense. 4. The U.S. Government assumes no liability for direct patent infringement, or contributory patent infringement or misuse of technical data. 5. The U.S. Government does not warrant the adequacy, accuracy, currency, or completeness of the technical data. 6. The U.S. Government assumes no liability for loss, damage, or injury resulting from manufacture or use for any purpose of any product, article, system, or material involving reliance upon any or all technical data furnished in response to the request for technical data. 7. If the technical data furnished by the Government will be used for commercial manufacturing or other profit potential, a license for such use may be necessary. Any payments made in support of the request for data do not include or involve any license rights. 8. A copy of this notice shall be provided with any partial or complete reproduction of these data that are provided to qualified U.S. contractors.

DESTRUCTION NOTICE For classified documents, follow the procedure in DoD 5220.22-M, National Industrial Security Program, Operating Manual, Chapter 5, Section 7, or DoD 5200.1-R, Information Security Program Regulation, Chapter 6, Section 7. For unclassified, limited documents, destroy by any method that will prevent disclosure of contents or reconstruction of the document.

UNCLASSIFIED

AD NUMBER

LIMITATION CHANGESTO:

FROM:

AUTHORITY

THIS PAGE IS UNCLASSIFIED

AD482595

Distribution authorized to U.S. Gov't. agenciesonly; Specific Authority; 31 MAY 1973. Otherrequests shall be referred to Commander, ArmyTraining and Doctrine Command, Attn: ATOI-IC-SMI. Fort Monroe, VA. 23651. This documentcontains export-controlled technical data.

Distribution authorized to U.S. Gov't. agenciesand their contractors; Specific Authority; MAR1966. Other requests shall be referred toCommander, Army Training and Doctrine Command,Attn: ATOI-IC-SMI. Fort Monroe, VA. 23651. Thisdocument contains export-controlled technicaldata.

USACDC ltr, 31 May 1973

) p\

RESEARCH ANALYSIS li CORPORATION

Reliability Analysis of the M60 Main Battle Tank

and the M151 %-Ton Utility Truck from Operating and Maintenance Data

FOR OFFICIAL USE ONLY

The contents of RAC publications, including the conclusions,

represent the views of RAC and should not be considered as

having official Department of the Army approval, either ex-

pressed or implied, until reviewed and evaluated by that agen-

cy and subsequently endorsed.



SCIENCE AND ENGINEERING DEPARTMENT TECHNICAL PAPER RAC-TP-212 Published March 1966

Reliability Analysis

of the M60 Main Battle Tank

and the M151 %-Ton Utility Truck

from Operating and Maintenance Data

by

William H. Jacobson Jr.

DISTRIBUTION STATEMENT In addition to security requirement« which mult

be met, thi* document ii subject to special export controls and each transmittal to foreign govern- ments or foreign nationals may be made only with prior approval of Office of the Chief of Research and Development, Headquarters, Department rf the

Army, Washington, D.C. 20310

. RESEARCH ANALYSIS CORPORATION

MCLEAN, VIRGINIA



FOREWORD

This technical paper presents the results of reliability research conducted under RAC Research Project 124 (Work Year 1964) and Research Pro- ject 171.2 (Work Year 1965). The initial RAC research effort waj directed toward the investigation and development of methodologies for implementing the requirement of Army reliability and maintainability regulations AR 705-25 and AR 705-26.

On completion of the preliminary general methodology,RAC-TP-116, the feasibility of the methodology was tested using operating and maintenance data (TAERS-type data) to review existing values of the methodology parameters, reported in RAC-TP-149. The results derived during test of the developed methodology indicated that generally accepted approaches to quantifying reliability were not adequate, and additional study and research into the basic concepts) of reliability were required. The use of operating and maintenance data to predict either the tactical or the technical capability of materiel and equipment was the subject of serious questioning.

This paper therefore addresses itself to the reliability analysis of operating and maintenance data, the only existing failure data for Army materiel and equipment. Its hypothesis is that mission reliability can be derived from operating and maintenance data. The objective of the analysis is to provide a quantitative basis for the development of reliability-data requirements.

George A. Martinez Head, Unconventional Warfare Department

in



ACKNOWLEDGMENTS

The author expresses sincere appreciation to each RAC staff member and military officer who contributed to the development of the research results presented in this report.

Acknowledgment is extended to the RAC staff members who participated in this study: Mr. Albert Van Horn, Miss Judith Bancroft, Mr. Joseph Dixon, Mr. Max Schreiber, and Mr. Leonard Sterns.

Acknowledgment also is extended to RAC supporiir;^ staff members who participated in the study in support of the author: Mrs. Lorraine Rinehart, Miss Ann Gleason, Mrs. Avonale Stephenson, and Miss Elizabeth Farioletti.

Special acknowledgment is extended toLtCol R.T.Tierno, Military Advisor; Dr. George Pettee,RAC Review Board Chair- man; and Mr. Edwin Bjoro of the RAC staff for constructive criticism and comment.

IV



CONTENTS

Foreword ill

Acknowledgments ?v

Summary 1

Problem— Facts—Discussion—Conclusions

Abbreviations and Notations 4

Glossary 4

Introduction 5

Problem—Background—Approach

Methodology 7

Reliability—Failure Criteria—Exponential Failure Law- Product Rule—Confidence Level

Main Battle Tank Data Analysis 10

Requirements—Data Description—Failure Record- Estimated Reliability—Measurement of Dispersion- Endurance—Reliability Degradation

Data Analysis of the ]A— Ton Utility Truck 20

Requirements—Data Description—Failure Record- Estimated Reliability—Measurement of Dispersion- Endurance—Reliability Degradation

Discussion

Reliability Estimate Accuracy—Reliability Estimate Confidence— Endurance

30



Conclusions 35

Appendix

A. Tabulated Data Summary 38

References 49

Figures

3. 4. 5. 6.

7. 8. 9.

10.

11. 12. 13. 14. 15.

16.

Cumulative Average Tank Mileage as a Function of Time, x and y Battalions Distribution of Accumulated Miles, x Battalion Distribution of Accumulated Miles, y Battalion Vehicular Reliability of the M60 Tank Main-Armament-System Reliability of the M60 Tank All-Corrective-Maintenance-Failure Confidence Levels for the M60 Tank Major-Parts-Failure Confidence Levels for the M60 Tank Endurance of the M60 Tank Reliability Degradation of the M60 Tank Cumulative Average Truck Mileage as a Function of Time, M151 Utility Truck Distribution of Accumulated Miles, M151 Utility Truck Reliability of the M151 Utility Truck Confidence Levels for the M151 Truck Endurance of the M151 Truck Major-Parts-Mission-Reliability Degradation as a Function of Elapsed Miles All-Corrective-Maintenance-Mission-Reliability Degradation as a Function of Vehicle Age

12 13 13 16 16

17 17 18

22 23 24 25 26

28

28

Tabl«

2. 3.

4. 5. 6. 7. 8. 9.

1Ü.

Failure-Record Summary of M60 Tank 12 Reliability Estimate, M60 Tank 14 Reliability Specifications for the Proposed QMR for '/4-ton Truck 22 M151-Truck Failure Record 23 M151-Truck Probability of Not Failing in 12,500 Elapsed Miles 27 Reliability of MBT and''/4-ton Utility Truck 30 Required and Estimated MMBF for MBT and V4-ton Utility Truck 31 Reliabilities of M60 Tank and Ml51 Truck " 33 MMBF and Confidence-Level MBF for M60 Tank and Mi51 Truck 33 Endurance of ths MBT and '/4-ton Truck 34

VI



SUMMARY

Problem

To perform a basic reliability analysis of available failure data to determine the quantitative reliability of present Army equipment.

Focts

This study was organized at the request of the Army to develop criteria for limits of reliability and maintainability. A primary task of the study was to obtain and analyze available failure data. The only available data suitable for this reliability analysis were collected by RAC and were similar to The Army Equipment Records System (TAERS) data. This data contained the operating and maintenance history of 140 main battle tanks (MBT), M60; 70 armored personnel carriers (APCs), M113; and 100 V4-ton utility trucks, M151.

1

Discussion

This study originated with the assumption that mission reliability could be derived from available failure data (i.e., operating and maintenance data). This hypothesis has been tested by performing a detailed analysis of the operating history of the M60 tank and the M151 truck. The ieliability criteria and concepts required to perform the data analysis have been defined in the methodology. The application of these criteria and concepts has illustrated the difference between the required and estimated reliability and the asso- ciated mean miles between failures (MMBF) as shown in the following tabulation.

System Required Estimated

Reliability MMEF Reliability MMBF

MBT V^-ton utility truck

0.95 7800 0.95 8750

0.0005 50 0.68 1140

It was determined that the difference between the required and estimated reliability was not because of inaccurate data, incorrect criteria, or errors induced by using approximating analysis techniques.



SUMMAR

The results obtained in the analysis were in effect a measure of the materiel-readiness environment and not the mission reliability of the equipment as specified in the qualitative materiel requirements (QMRs). The measure of materiel readiness is a measure of the consumption and depletion of equipment(s) and not a measure of materiel reliability. Therefore the difference between the required and the estimated reliability was because the first was primarily a measure of the equipment, whereas the second was primarily a measure of the environment and operating conditions.

Conclusions

1. The results of the analysis illustrated tint operating and maintenance data cannot be used to measure mission reliability. Therefore the hypothesis that mission reliability can be derived from available failure data is not correct.

2. An interpretation of the results of the analysis has indicated that verifiable measures of mission reliability must be assessed in test and evaluation programs that are controlled to predetermined standards.



Reliability Analysis

of the M60 Main Battle Tank

and the M151 %-Ton Utility Truck

from Operating and Maintenance Data



ABBREVIATIONS AND NOTATIONS

APC In MBT MBF MMBF MTBF QMR RMO TAERS

armored personnel carrier natural logarithm main battle tank miles between failures mean miles between failures mean time between failures qualitative materiel requirement reliability for a certain mission duration Th'T Army Equipment Records System base of the natural logarithm

GLOSSARY

probability. A mathematical basis for prediction, given an exhaustive set of outcomes, i^ the ratio of the outcomes that would produce a given event to the total number of pos- sible outcomes. reliability. The probability that materiel will perform its intended function for a specified period under stated conditions. mission reliability. The probability that the materiel will give specified performance for the duration of a mission when used in the manner and for the purpose intended, given that the materiel is functioning properly at the start of the mission. failure. A condition occurring in the equipment requiring unscheduled corrective maintenance classified as major-part or minor-part failure.



INTRODUCTION

Problem

The purpose of this study was to perform a basic reliability analysis of available failure data to determine the quantitative reliability of present Army equipment. The objective was to provide a cefinitive data base that would provide measures against which developmental hardware could be evaluated. Then the Army could appraise it carefully and specify reliability requirements that would result in the acquisition of equipment that could accomplish mission objectives.

Background

A research program, which began as RAC-RP-124 and was continued as RAC-RP-171.2, "Reliability Research," was established at the request of the Army to develop criteria for limits of reliability and maintainability. A primary task of the study was to obtain and analyze available failure data, to determine the existing reliability values for this data, and to establish a quantitative basis for numerical-reliability-criteria development.

Acceptable data for reliability evaluation require that failure data indicate parts, components, and systems that do not fail, as well as those that do fail. An extensive data search, conducted as part of the work under RAC-RP- 124, indicated that RAC had the only available data that met this basic requirement.1 The RAC data originated from the same basic record documents that now are the data source for TAERS.

The Department of the Army's Integrated Equipment Record and Main- tenance System, of which TAERS is a part, was initiated 1 May 62. The system was devised to implement the policy announced in AR 750-1,2 "Maintenance Concepts," and AR 750-5,3 "Organization, Policies, and Responsibilities for Maintenance Operation." The information collected and compiled through TAERS is historical operational and maintenance data recorded for most Army equipment.4 These records are to be used for the control of operation and maintenance of all Army equipment included under the provisions as defined in Army manuals.5-7

The available data consisted of operating and maintenance records for 640 M60 MBTs, 722 M113 APCs, and 898 M151 V4-ton utility trucks. When the maintenance data of the M60 tanks, M113 personnel carriers, and M151 trucks were reviewed for consistency and completeness, it was discovered that a maintenance history had been recorded for only part of the sample. The data for vehicles having incomplete records were deleted from the analysis.



It was determined by the screening process that an operating- and maintenance-data history meeting the basic failure-d."


METHODOLOGY

This methodology defries the reliability criteria and concepts used in this report to analyze grating and maintenance data. The selection of the failure criteria was determined by the characteristics of the data. The reliability concepts of the exponents 1 failure law and the product rule are in general acceptable in reliability practice and were assumed valid for purposes of this analysis.

Reliability

Reliability was defined as the probability that equipment would not require corrective maintenance for a period equivalent to the stated mission duration under the conditions of the operational Army. Corrective maintenance was defined as the adjustment, repair, or replacement of part(s) of the equipment analyzed. Corrective maintenance was further subdivided into three categories (aj major part (i.e., replacement of a major vehicular part), (b) minor part (i.e., adjustment, repair, or replacement of all parts other than major vehicular parts), and (c) main-armainent-system part adjustment, repair, or replacement.

Failure Criteria

The failure criteria established for this study were: (1) Corrective maintenance recorded in the operating and maintenance

data that was not performed at a scheduled downtime was required, and therefore classified as a failure.

(2) The corrective maintenance performed corrected the failure, and subsequent corrective maintenance was in response to additional failures.

(3) The replacement of a major vehicular part such as an engine or transmission was classified as a major-part failure.

(4) The adjustment and repair of all vehicular parts and the replacement of all vehicular parts other than major ones were classified as ;,-, minor-part failure.

(5/ The occurrence of a major-part or a minor-part failure was an all- corrective-maintenance failure.

(6) When more than one corrective-maintenance action was performed per downtime, the occurrence was classified as a part failure of the most significant part involved (as a major-part failure or minor-part failure if no major part was replaced).

(7) The adjustment, repair, and replacement of all the parts, components, and systems within the turret and the weapons system were classified main- armament failures.



Exponential Failure Law

The exponential failure law can be used to predict the probability of survival of a part as a function of time when the failure rate is assumed constant. It was assumed in this analysis that the exponential failure law was valid for estimating vehicle reliability from operating and maintenance data. The law was derived from the Poisson distribution.8

The exponential failure-law equation can be expressed as

f-MI) MTHK (1)

where R^n = probability of success or reliability for a certain mission duration t: - base of natural logarithm

MD = mission duration (time) MTBF = mean time between failures. (During the operating period when

'he failure rate is constant, the mean time between failures is the reciprocal of the constant failure rate, or the ratio of the total operating time to the total number of failures. It is also defined as the average time of satisfactory operation of a pop- ulation of equipment.)

The exponent of e in the exponential failure-law equation is 1/failures; therefore, for purposes of this analysis, where the main operating criterion is dis- tance, the reliability figure of merit utilized is MMBF. Then

RMD^

FCS OFFICIAL USE ONLY

Confidence Level

Confidence level was defined as the probability that the reliability experienced on a given mission would be equal to or greater than a stated value. The confidence level for all values of reliability was determined from the distribution of the data used in each data analysis and plotted as a continuous function of reliability.

The confidence-level analysis was used to illustrate the significance of dispersion within the data distribution. As the dispersion of data increased, the significance that could be attributed to the average value (mean) decreased.



MAIN BATTLE TANK DATA ANALYSIS

The M60 tank data were analyzed based on the criteria presented in the methodology. The results of this analysis have been compared to the established reliability and endurance requirements specified in the MBT QMRs.

Requirements

The proposed QMRs for two new MBTs specified that:

(b) Sustained operation. (Essential) The tank must be capable of Sustained battlefield operation for 48 hours (400 miles) (640 km) of vehicle operation without requiring maintenance or servicing. Battlefield operation is defined as follows:

1_. Thirty-two hours of intensified tactical activity of which forty percent will be idling at rated engine idle speed; forty percent will be cross-country movement at speeds varying from 2x/2 to 25 mph (3 to 40 km/h) and twenty percent will be movement on gravel or dirt roads at speeds varying from 15 to 35 mph (24 to 56 km A).

2. Sixteen hours of minimized tactical activity of which twenty-five percent will be during daytirre and seventy-five percent during nighttime. The engine will be run sufficiently to maintain the battery in a charged state. The master relay switch and the radio/receiver(s) will be "on." All other components requiring electrical energy in order to function will be "off* unless otherwise required (examples: heater for crew comfort in extreme cold climate, engine ■varm-up or passive night viewing equipment). . . .8'9

The Joint Military Characteristics for the US/FRG MBT contains the reliability specification that states:

6. Reliability (Essential) a. Ignoring iiuman error, the expectancy of correct functioning of the whole system must be at least 5*. Reliability pertaining to crew and accompanying infantry safsty and to cotitinueri .^'Jitability of the tank shall not be less than that of current conventional b. , inks. . . .

,0

The significance of the reliability specification was analyzed using the basic reliability criteria previously presented in this paper. The reliability values specified were substituted into Eq. 2.

The mission duration was specified as 48 hr or 400 miles. The reliability specified was 95 percent. Using these values in Eq. 2,

0.9J5-f-400/MMBF (4)

and solving the equation for MMBF,

\1\1IIF - 400 In 0.95 ._

- 7800 miles

10



The tank has been treated as two major systems, the vehicle and the main armament system.

Assuming equal system weight, the reliability of the vehicle and the main armament system was determined by taking the square root of the tank reliability.

R400U,US ^°" (6) 0.975

The resulting MMBF for each system, using Eq. 2, is

MMHK -U)0 In 0.97,") ,„,

15.800 inilcM

The new MBT QMRs also specified that:

d. Endurance

(2) Life with normal servicing. (a) (Essential) The tank, excluding weapons shall operate for 300 hours or 3,000 miles (4800 km) with only scheduled organizational maintenance and without failure or excessive wear of any part, assembly or component. (b) (Essential) The tank shall be capable of being operated at least 500 hours or 5000 miles (8000 km) without replacement of engine or power train. . . .8'9

(3) Main Armament System. (a) (Essential) If a missile-conventional gun weapon combination is provided, the missile system must be capable of firing 1,0(0 missiles and of enduring 3,000 miles (4800 km) of vehicle operation withurt requiring maintenance other than organizational. . . .8'9

Data Description

The average number of elapsed miles per calendar year of operation accumulated by the M60 tanks of two armored battalions is shown in Fig. 1. (The data were collected from units operating with the Seventh Army in Europe and were considered representative of an army maintaining personnel and materiel in a state of combat readiness.) The average elapsed miles per interval of time was fairly constant and linear for both battalions. The tanks from x battalion averaged 1950 miles/year and the tanks of the y battalion averaged 1400 miles of operation per year.

The distribution of accumulated miles for the two tank battalions during the data-collection period is shown in Figs. 2 and 3. All vehicles of battalion x vsre operated a minimum of 3600 miles with one tank accumulating approximately 5500 miles of operation. An average mileage of 4150 was attained by the tanks of battalion x. All vehicles of battalion y were operated an approximate minimum of 2200 miles. The maximum miles of operation accruing to any tank of this latter group was 4100 miles. The average tank mileage attained by the tanks of battalion y was 2700 miles.

Failure Record

Table 1 summarizes the failure record of the major parts of the M60 tanks assigned to the x and y battalions. The apparent rate of failure of individual-like major parts was random until 25 to 30 percent of the parts failed. The only M60 tank parts to exhibit a 25 percent or greater failure

11

FOR OFFICIAL USE 0MLY


5000

4000 _

*> 3000 _J

5 Q

< 2000

1000

I l I l l—I" I 1 I 1 I I 1—I I I I I I I I I i i i i i i i

x battalion 1950 miles

battalion 1400 miles/year

' ' MAMJ JASON DJFMA MJ JAS OND.JFMAMJ J AS

1961 | 1962 I 1963 CALENDAR YEAR

Fig, 1—Cumulative Average Tank Mileage as a Function of Time, x and y Battalions MBT, M60; sample size—x, 72 tanks; y, 68 tanks.

TABLE 1

Failure-Record Nummary of M60 Tank"

x battalion y battalion

Tank Percent Predicted Percent Predicted

part of average Standard of aver .ige Standard sample life,

mean miles deviation1" sample life,

mean miles deviation

failed failed

Kngine 35 S.HSO 3.070 27, 6100 5120 1 ransniission 10 Ino, ..'' Imlet. Negl-° Indet. Indet.

r* mal drives Negl. Indel. Indel. Negl. Indet. Indet. Sprockets 19 3,243 587 14 3896 1075 Tracks :>o 2.423 300 2K 3109 600 Road wheels

and idlers 60 4.790 2.079 Negl. Indet. Indet. Starters 38 10,355 12.255 52 3056 2835

Use characteristics:

Hme frame: Apr f"l through Sep 6,1.

Sample size; x battalion, 72 tanks; y battalion, 68 tanks. Type o unit: armored.

Values derived from fteibull probability function, which is a mathematical tool that was used to

project the component-part failures to approximate the average life. c Magnitude indicates variance, i.e., when the standard deviation (a) equals the average value, a con-

stant rate of failure is indicated, (b) is greater than the average value, a decreasing rate of failure is

indicated, and is (c) less than the average value, an increasing rate of failure is indicated. The Weibull standard deviation indicates the variance in mileage of component part failure.

( Indeterminate.

Negligible.

12


I


14

12

5)0 or UJ

5 8

i

OLJ- 3400

Average accumulated miles during data collection

4150 miles vehicle

III EE 3800 4200 4600 5000

ELAPSED MILES

5400

Fig. 2—Distribution of Accumulated Miles, x Battalion MBT M60, sample size—72 tanks.

10

14 __, Average accumulated miles

during data collection _

^ 12 > or UJ 1= 10 z

_

2700 miles vehicle

_

uj 8 a.

-J ° ij I UJ 4

-

2

n i n ~n n p 2000 2400 2800 3200 3600

ELAPSED MILES

4000

Fig. 3—Distribution of Accumulated Miles, y Battalion MBT W^O, sample size—68 tanks.

13



level in approximately the first 2500 miles of operation were the engines, tracks, and starters. The only additional parts to exhibit a 25 percent or greater failure level in approximately the first 4000 miles were the road wheels and idlers.

Estimated Reliability

The estimated reliability of the M60 tank for a 48-hr (400 mile) battlefield day is shown in Table 2.

TABLE 2

Reliability Estimate, M60 Tank

Miles between failures (MBF) Bat lefield-day reliabil

48 hr or 400 miles >y.

System Mean

average 0.90 confidence

level Mean

average 0.90 confidence

level

Individ- Aggre- Individ- Aggre- Individ- Aggre- Individ- Aggre- ual gate ual gate ual gate ual gate

\ ehicluar major purls replacement

r.ngincs 6000 1500 0.935 0.773 Power train Indet." ludet. 0.999 . 0.999 4 Suspension 2100 1700 0.827 0.791 Fuel 3 UK) 100 0.890 0.018 Klectrical two 60 0.913 0.001 Other 1:200 66 0.717 0.002

Subtotal 500 ,15 0.45 0.000b

Vehicular minor parts adjustment, repair and or replacement

I nschcduled downtimes I00 :12 0.45 0.000''

Subtotal 500 32 0.45 0.000b

Overall "0 1 0.0005 0.0()0b

''Indeterminate.

"Approaching the limit 0.000.

The average 48-hr battlefield-day reliability derived from the analysis of operating and maintenance data was 0.0005 with an MMBF of 50 miles. The 0.90 confidence level reliability approached the limit 0.000 with an MMBF equal to or less than 1 mile.

14



The M60 tank was considered as consisting of three systems: vehicular major parts, vehicular minor parts, and main armament parts. The aggregate battlefield-day reliability of the vehicular major parts was 0.45 with an MMBF of 500 miles. The 0.90 confidence level reliability approached the limit 0.000 with an MMBF of 35 miles. The major parts were composed of all tank parts that would immobilize the vehicle if they failed. They were divided into six systems: engines, power tram, suspension, fuel, electrical, and "other systems." The parts included in the other-systems category were all the major parts not included in the first five systems.

The aggregate battlefield-day reliability of the vehicular minor parts was 0.0025 with an MMBF of 67 miles. The 0.90 confidence level reliability approached the limit 0.000 with an MMBF of 3 miles, The data analysis indicated that an average of approximately two corrective maintenance actions were performed for each unscheduled downtime. The battlefield-day reliability attributed to unscheduled downtimes due to vehicular minor parts was 0.133 with an MMBF of 198 miles. The 0.90 confidence-level battlefield-day reliability was approaching the limit 0.000 with an MMBF of 8 miles.

The vehicular reliability of the M60 tank as a function of the battlefield day in hours is shown in Fi£. 4. The battlefield-day reliability requirement of 0.975 was plotted as a reference for comparison of major-parts battlefield- day reliability and the all-corrective-maintenance battlefield-day reliability. The all-corrective-maintenance battlefield-day reliability was the aggregate reliabilities of the vehicular major-parts ar.d the unscheduled vehicular minor- parts downtimes.

The battlefield-day mission duration corresponding to a 0.975 reliability requirement is shown in Fig. 4 as 1.5 hr or 13.3 miles for major parts and 0.43 hr or 3.8 miles for all corrective maintenance. The 0.90 confidence- level reliability of the M60 tank also was plotted in the figure, and it approached 0.000 in a battlefield day of less than 20 hr or 178 miles.

The aggregate battlefield-day reliability of the main armament system was 0.45 with an M^BF of 500 miles. The 0.90 confidence-level battlefield- day reliability was approaching the limit 0.000 with an MMBF of 32 miles. The data did not list the number of rounds fired by the gun, and the battlefield-day reliability as a function of rounds iired could not be determined.

The reliability of the main armament system as a function of the battlefield day in hours is shown in Fig. 5. The battlefield-day-reliability requirement of 0.975 was plotted as a reference for comparison of the main-armament- system reliability and the 0.90 confidence-level reliability. The battlefield - day mission duration corresponding to a 0.975 reliability requirement is shown in Fig. 5 as 1.5 hr or 13.3 miles. The 0.90 confidence-level reliability of the M60 tank main armament system approached 0.000 in a battlefield day of less than 12.5 hr or 111 miles.

Measurement of Dispersion

The foregoing section on the estimated reliability of the M60 tank presented just the mean (average) and the 0.90 confidence level of reliability as an indication of dispersion. The confidence level as a continuous function of reliability for a stated failure criterion is shown in Figs. 6and7 for the M60 tank.

15



- -,— T " T 1 1 ! 1 i

T ""„ ~ _. _ __ — _. _ _. . _ j — _ — — _

1 «-v *» c _

i ° E

i is 0) u —. - CO c

1 J • D o Hi —

f \ a; o or 4J Ö

1 «u. \I i 1 1 O CO E u~T o

QJ > 0> oj o m D C

4) ! / (J 4> U / 0J C o E

c 4»

0 5 "D _ 5 V *C o

u < > u o Q) o

Ö 0 u .c

i \ 1 I

<

1 1

j>

o ■o

Q

E o E <

I

AJ.niavnaa

< Q

< CO

o»

01 a:

>

Aimavnaa

16



I.U i 1 1 1 I ! 1 1 T

0.8 -

0.6

>i Moin armoment system,

-

^^iverage reliability 0.45

0.4 -

v Vehicular system, Nw -

0.2

n 1

^V average reliability 0.06 ^V

i i i i* >■ i i >

-

0.2 0.4 0.6 0.8

RELIABILITY

1.0

Fig. 6—All-Corrective-Maintenance-Failure Confidence Levels for the M60 Tank

1.0

0.8

w 0.6

u z

z °'4 d u

0.2 -

1 1 i i 1 1 1 1 !

-

-\ -

- >w -

>. Vehicular system,

, ^average reliability - 0.45

- -

1 1 1 1 1 1 1 1*^.

-

0.2 0.4 0.6

RELIABILITY

0.8 1.0

Fig. 7—Major-Parts-Failure Confidence Levels for the M60 Tank

17



The failure criterion used for Pig. 6 was all-corrective-main.'enance failure. The reliability computed from the MMBF for the vehicular system and the main armament system was 0.06 and 0.45, respectively. The confidence levels for these reliability values were 0.31 and 0.43, respectively, because of the nonlinear dispersion of failures (i.e., tie lower reliability values occur more frequently than the higher reliability values).

The failure criterion used for Fig. 7 was major-parts failure of the vehicular system. The reliability computed from the MMBF was 0.45, and the confidence level for this reliability value was 0.32 because of the nonlinear dispersion of failure. The measurement of dispersion shown in Figs. 6 and 7 illustrates the broad range of the dispersion that will affect the significance of the estimated battlefield-day reliability.

Endurance

The endurance of the M60 tank is shown in Fig. 8 as the probability of the M60 tank operating for a stated operating life (miles) with normal servicing. The endurance specified in the QMR was plotted on the figure. These values served as a reference for comparison of the endurance experienced by the M60 tank.

The QMR specified that the tank, excluding weapons, would operate 300 hr or 3000 miles without scheduled organizational maintenance and without failure or excessive wear to any part, assembly, or component. The M60 tank had a probability of less than 1 percent of not having all-corrective-maintenance failure before 1000 miles of operation and a probability of only 1 percent of not having a major vehicular-part failure within 3000 miles. One hundred percent of the M60 tank suspension systems experienced a failure between 1200 and 3000 miles of operation.

The QMR further specified that the tank would be capable of operating 500 hr or 5000 miles without replacement of engine or power train. The M60 tank had only a 50 percent probability of not having an engine replacement within 5000 miles of operation.

Reliability Degradation

The 48-hr battlefield-day reliability of the M60 tank changed as a function of elapsed miles. The change in the battlefield-day reliability of the M60 tank due to major-parts failure is shown in Fig. 9.

The average major-parts reliability was 0.45. The degradation function was cyclic and indicated that more parts failed between 1500 and 2900 miles and after 3500 miles of operation than at other times. The major-parts reliability varied from a high of apt-roximately 0.90 at issue *o a low of approximately 0.14 at 2600 elapsed mil-."1 An analysis indicated that the tracks, sprockets, road wheels, and idlers contributed to the cyclic function. However, the cause of the basic cycling effect was not isolated.

The track and sprocket replacement, the roadwheel and idler failures, and the failure of seal and excluder were separated from other major-part failures to demonstrate their effects on the changing failure rate. The major- part reliability of the M60 tank was at a low of 0.14 in the 2500- to 2750-mile range. Therefore the unreliability of the M60 tank was 0.86. The track and sprocket, road wheel and idler, and seal and excluder failures contributed an aggregate of 0.33 to the unreliability; and 0.53 of the unreliability was due to other major-part failure.

18



Engine replacement

u

1000 2000 3000 4000

OPERATING LIFE WITH NORMAL SERVICING, MILES

Fig. 8—Endurance of the M60 Tank

8 8. O M

»a UJ to

5000

1.0

>- 0.8 -\

Less tracks and sprockets, seals and excluders, road wheels and idlers

1000 2000

ELAPSED MILES

3000 4000

Fig. 9—Reliability Degradation of the M60 Tank

19



DATA ANALYSIS OF THE V4-TON UTILITY TRUCK

The Ml51 data were analyzed based on the criteria presented in the methodology. The results of this analysis have been compared to the established reliability and endurance requirements specified in the proposed '^-ton utility truck QMR.

Requirements

The proposed QMR for the new vehicle specified that:

... it must be capable of being operated in all seasonal conditions with a mission reliability of 95$ with none other than driver maintenance and in subartic and tropical climates with a mission reliability of 95$ with the use of modification kits. This mission is defined as operations in a combat zone for a period of 6 days with an average utilization of 75 miles per day, 40$ of which is idling, 40$ cross-country, and 20$ over secondary roads. This vehicle will be used in the temperate, subartic and tropical zones and in jungle, desert, mountainous and savanna-type terrain. It must have the capability to negotiate commonly encountered terrain types and obstacles, including shallow ditci. -, bushes, grass, shallow streams and beaches, and to pass moderately forested and rocky areas over routes receiving some preparation. . . .l:

The significance of the QMR specification was analyzed using the basic reliability criteria previously presented in this paper. The mission duration was specified as 6 days at 75 miles/day or 450 miles.

Vehicle reliability was specified as 0.95. Using these values in Eq. 2,

0.95 f-4-,0 MMBK (8)

and solving the equation for MMBF,

MMBK - tfo In our,

-8750 miles (9)

The seven systems of the utility truck are functionally in series. As- suming equal weight, the reliability of the individual system was

R jsoy v'0.95 . ^Q

0.993

The resulting MMBF for each system using Eq. 2 was

MMBF : - 450 In 0.993

- 62.000 miles

20

(11)



The power-plant system contained two components, the engine and the transmission. The MMBF for each component, assuming equal weight, also was computed by taking the square root of the system reliability resulting in a component reliability of

R,M2 VTT593 (12) 0.9963 miles

The resulting MMBF for each component using Eq. 2 was

MMBF -450 In 0.963

123.000 miles (13)

The MMBF for both the engine and the transmission must be 123,000 miles. The proposed QMR specified that it was essential for the vehicle to

possess "... a 90$ probability of being operated for 15,000 miles with only scheduled organizational maintenance, without failure of either accessories or integral major components, and for 25,000 miles without the need for major overhaul or the replacement of a major component. . . ."10>u

This specification overlapped the 0.95 reliability specification and resulted in two different values of the required mean life of the components.

The significance of this specification was illustrated by substitution in Eq. 2. For a 0.90 probability of satisfactorily operating 15,000 miles the MMBF was

MMBF -15.000 In 0.90

1-12.000 miles (14)

For a 0.90 probability of satisfactorily operating 25,000 miles the MMBF was

MMBF = -25,000, In 0.90

- 237,000 mile«. (15)

These values illustrate the order of magnitude of the incompatibility that separates the mission reliability and tha 0.90 probability specification.

The significance of the proposed QMR is summarized in Table 3.

Data Description

The average number of elapsed miles per calender year of operation accumulated by 100 selected M151 trucks of two armored battalions is shown in Fig. 10. (The data were collected from units operating with the Seventh Army in Europe and were considered representative of an Army maintaining personnel and materiel in a state of combat readiness.) The average elapsed miles per interval of time was constant anc linear during the time interval analyzed. The M151 trucks averaged 9500 miles of operation per year.

The distribution of accumulated miles for the M151 trucks during the data-collection period is shown in Fig. 11. All vehicles were operated a minimum of 12,500 miles with two M151 trucks accumulating approximately 27,500 miles of operation. The average accumulated miles of operation was 17,300 miles.

21



TABLE 3

Reliability Specifications for the Proposed QMR for '4-ton Truck

Specified reliability,

probability of

survival

Specified time intervo1

or mission duration,

miles

Required failure-free life, MMBF

Mode A° Mode Bb Mode Cc

().


u I

16

14

12

10

8

6 -

4

2

0 10,500

:■«■&*:;

:■-:':■:

Average occumuloted miles 1 during data collection

17,300 miles vehicle

I

15,500 20,500 25,500 30,500

ELAPSED MILES

11—Distribution of Accumulated Miles,

M1S1 Utility Truck

TABLE 4

M151 Truck Failure Record

(Sample six«, 100 vehicles, average age, 17,000 milei;

failure criteria, major parts replacement)

Component Quantity

Subtotal Total

Power plant 12 Engine 12 Transmission 30

Power train 80 Clutch* 43 Differentia! 5 Wheel hearing 24 Propeller shaft 2

Universal joint 6 ("doling system 61

Radiator'1 29 Hose 12 Pump 3 Kan belt i:

Fuel system 51 Fuel pump 17 Carburetor* 34

Ignition system 40 Distributor" 35 Coil 5

Electrical system 83 Battery IS Generator 16 Regulator" 52

Brakes 3 Master cylinder 3

Total 300 360

igii failure part.

23



Failure Record

The replacement record of 100 M151 trucks is summarized in Table 4. Replacement of the components listed was assumed to be an indication of failure or pending failure. Six of those components accounted for 223 of the 360 replacements.

Estimated Reliability

The mission reliability of the M151 truck as a function of the mission duration in miles is shown in Fig. 12. The 6-day mission-reliability requirement of 0.95 was plotted as a reference for comparison of major-part mission reliability and all-corrective-maintenance mission reliability. The mission reliability was estimated as 0.86 with an MMBF of 3030 miles, and the 0.90

<

0 100 200 300 400 MISSION DURATION, MILES

Fig. 12—Reliability of the M151 Utility Truck

500

confidence-level mission reliability was 0.02 with an MMBF of 115 miles because of major-part failure. The mission reliability was 0.86 with an MMBF of 1140 miles, and the 0.90 confidence-level mission reliability approached 0.0Q because of all-corrective-maintenance failure.

24



The mission duration corresponding to a 0.95 reliability requirement is shown in Fig. 12 as 120 miles (less 'han 2 days) because of major-part failure and 60 miles (less than 1 day) because of all-corrective-maintenance failure. The 0.90 confidence-level mission durations were not significant numbers.

Measurement of Disperston

The previous section on the estimated reliability of the M151 truck presented just the mean (average) and the 0.90 confidence level of reliability as an indication of dispersion. The confidence level as a continuous function of reliability for a stated failure criterion is shown in Fig. 13.

7.

Z o u

0.4

0.2

jor-ports failure, overage reliability 0.86

All-corrective-maintenance""* failure, average reliability 069

0.2 0.4 0.6

RELIABILITY

Fig. 13—Confidence Levels for the M151 Truck

The mission reliability computed from the MMBF for the all-corrective- maintenance failurea and the major-parts failures was 0.69 and 0.86, respectively. The confidence level for both of these mission-reliability values was 0.33 because of the nonlinear dispersion of failures (i.e., the lower reliability values occur mure frequently than the higher reliability values).

Endurance

The endurance of the M151 truck is shown in Fig. 14 as the probability of the Ml51 truck operating for a stated operating life (miles) with normal servicing. The endurance specified in the QMRs was plotted on the figure. These values served as a reference for comparison of the endurance experienced by the M151 truck.

25



The QMR specified that the truck would have a 0.90 probability of operating 15,000 miles with only scheduled operational maintenance without failure of either accessories or integral major components, and for 25,000 miles without the need for major overhaul or the replacement of a major component.

enJuronce 0.90 _i

a: a D u u o

to < SO o a. a.

4,000 8,000 12,000 16,000 20,000 24,000 28,000

OPERATING LIFE WITH NORMAL SERVICING, MILES

Fig. 14—Endurance of the M151 Truck

The M151 truck had a probability of 0.38 of not having all-corrective- maintenance failure before 1000 miles of operating life. At an approximate 12,000 miles of operating life there was a 1.00 probability that all M151 trucks would have experienced failure. The M151 tru^k had a probability of 0.64 of not having a major-part failure (i.e., replacement of such parts as an engine, transmission, or clutch) before attaining 1000 miles of operating life. At approximately 20,000 miles of M151-truck operating life there was a 1.00 probability that all M151 trucks would have experienced a major-part failure requiring replacement of a part in that category.

26

FOR OFFICIAL USE MLY


The M1S1 truck hud a probability of 0.90 of not requiring an engine replacement in 15,000 miles of operating life, and a probability of 0.76 of not requiring a clutch repair or replacement within the same period. For a life expectancy of 25,000 miles the M151 had a probability of 0.87 of not having an engine replacement, and a probability of 0.66 of not requiring clutch repair or replacement.

TAE-_. 5

MISl-Truck Probability of Not Failing in 12,500 Elapsed Miles (Sample *ue, 100 vehicles, failure criteria,

rooair or replacement)

Probability o f survival System and component Failures

Component System

Power plant 0.7') Kngino (i 0.91 [ ransmissioji If, 0.81

\ «wer train O-.H Cluteh 57 0 t.'i Differential 5 0.975 Propeller shaft -. -- 1 niversal joint I 0.999 Wheel bearing 21 0.970

(.noting system 0.68 Kadiulor 24 0.76 Hose 2 0.9«) Pump :i 0.97 Kan belt 13 0.935

I'uel system 0.68 1' uel pump (l 0.9 V (iarburctnr 2«) 0,80 Air-hose ie'.ake 10 0.90

Ignition system 0.78 Dislributoi • >2 0.78 Coil :\ 0.97

hie« irical system o.r>: 1 lattery 7 0.965 Generator 14 0.86 Regulator 36 0.64

Hrake system 0.99 Master cylinder i 0.99

Overall 0.06

The probability of the M151 truck not failing in the first 12,500 miles of operation is shown in Table 5 as 0.06. All the M151 trucks of the 100 truck sample had only operated 12,500 miles. Therefore an elapsed mileage of 12,500 was used to evaluate the 0.90 probability of the truck to operate for 15,000 miles without failure of accessories or integral major components (major parts).

27



80

1

^s>>>>>^ Statisnco

1 1

average 0.86

Mission reliability' —

_

60

40

— —. _ _Ä __^

6?

> _l

to < -i UJ

/ _ Asymptotic limit of the mission

reliability 0.7]

-

20

1

0 1 1 1

1 1 1 1

0 4,000 8,000 12,000 16,000 20,000 24,000 28,000

ELAPSED MILES

Fig. 15—Major-Parts-Mission-Reliability Degradation as a Function of Elapsed Miles Mission duration - 6 days (450 miles^'4-ton utility truck, Ml51.

Sample Size, 100 vehicles; average age, 17,300 miles, failure criterion, major parts replacement.

Statistical computation •"^~ Projection

80

i i i i

>v Statistical average 0.6

1 1

3 -

60 -

Mission reliability ^^ „_ -

40 / ^

Asymptotic limit of the mission reliability 0.435

20 -

n 1 1 1 1 1 0 4,000 8,000 12,000 16,000 20,000 24,000 28,000

ELAPSED MILES

Fig. 16.—All-Corrective-Maintenance-Mission-Reliabilify Degradation as a Function of Vehicle

Mission duration - 6 days (450 miles),'/i-ton utility truck, M151.

Sample size, 100 vehicles; average age, 17,300 miles; failure criterion, all corrective maintenance.

— Statistical computation — — Projection

28



Reliability Degradation

The 6-day (450-mile) mission reliability of the M151 truck changed as a function of elapsed miles OP. the truck. The change in mission reliability of the M151 truck because of major-parts failure and all-corrective-maintenance failure is shown in Figs. 15 and 16, respectively.

The mission reliability of the M151 truck because of major-part failures was predicted to be greater than 0.71 during the useful life of the truck. The mission reliability of the M151 truck because of all-corrective-maintenance lailures was predicted to be greater than 0.435 during the useful life of the truck.

The degradation function was cyclic as a function of M151-truck elapsed miles. However, the shape of the cyclic function could not be satisfactorily isolated. Several methods were used. The lower limits of the cyclic functions (of the several methods) were used to estimate the mission-reliability degradation rate, and a smooth line was drawn through the lower-limit points. The resulting smooth line approximated the exponential. The statistically derived mission reliability over the average truck elapsed miles and the asymptotic limit of the projected exponential degradation function are plotted in Figs. 15 and 16.

The major-parts and all-corrective-maintenance mission reliabilities (statistical average) were 0.86 and 0.69, respectively, over the average truck elapsed miles. The mission reliabilities during the first 450 elapsed miles after issue were approximately 0.96 and 0.90, respectively, and degraded grad- ually until during 450 miles of operation after an elapsed 17,000 miles the mission reliabilities were approximately 0.74 and 0.47, respectively. When the mission-reliability degradation trends were projected, an asymptotic limit of 0.71 (for major-parts failure) and 0.435 (for all-corrective-maintenance failure) was predicted.

29



DISCUSSION

A basic reliability analysis has been performed on the failure data of the M60 tank and M151 truck to determine the quantitative reliability of present Army equipment.

Reliability Estimate Accuracy

The results of the data analysis have been compared to the requirements specified in the QMRs for new equipment to illustrate the difference between the required and the estimated (calculated) reliability.

The required and estimated reliability of the MBT and the V^-ton utility truck are summarized in Table 6.

TABLE 6

Reliability of MBT end '4-ton Utility Truck

Required Estimated System reliability reliability

MBT 0.95 C.0005 Vehicular system 0.975 0.0011

Major parts 0.987 0.45 Minor parts 0.987 0.0025

Main armament system 0.975 0.45 !i-ton utility trui'k 0.95 0.68

Major parts 0.975 0.86 Minor parts 0.975 0.79

The orders of magnitude of difference between the required and estimated reliability values are more ciearly illustrated in terms of the MMBF in Table 7. I

The difference that has been illustrated to exist between the required and j the estimated reliability was hypothesized to be the result of one, or a combi- | nation, of the following assumptions: (a) the failure data were inaccurate, t! (b) the failure criteria were incorrect, (c) the estimated reliability value reflects the true mission reliability of the equipment, and/or (d) the reliability requirement specified was unrealistically high.

30



Inaccurate Failure Data. The validity of the first assumption was evaluated by determining the most probable effect of data inaccuracy. The failure data, derived from specific organizational and supply records were verified by auditing procedures. The replacement of major parts war> recorded with considerable accuracy and was considered to be a firm indication of failure. The data on the adjustment and repair of the equipment and the replacement of minor parts were not recorded with the same level of confidence. Some unscheduled corrective maintenance was believed to have been omitted from the data. Therefore the most probable data inaccuracy resulted in a high estimate of the reliability, and the difference between the required and the estimated reliability was probably less than the real difference.

TABLE 7

Required and Estimated MMBF for MBT and '4-ton Utility Truck

System Riq uired MMBF Estimated MMBF

MBT 7.800 50 Vehicular system 15.800 59

Major parts 31.500 500 Minor parts 31.500 67

Main armament system 15,800 500 '4-ton utility truck 8,7.50 1140

Major parts 17,700 3030 Minor parts 17,700 1907

Therefore the assumption that the difference between the required and the estimated reliability was because of inaccurate failure data is not valid.

Incorrect Failure Criteria. The validity of the second assumption was evaluated by determining the effect of reducing the severity of the failure criteria. Failure has been defined as unscheduled corrective maintenance. The failures were divided into major-part and minor-part groupings, and the reliability of each of these groupings was estimated.

When the failure criteria were changed to include only major-part replacement as indicative of failure the difference between the required and the estimated reliability was still significant. Even this drastic reduction in the failure-criteria severity would not eliminate the difference between the required and the estimated reliability.

Therefore the assumption that the difference between the required and the estimated reliability was because of incorrect failure criteria is not valid.

True Mission-Reliability Estimate

The validities of the third and fou-th assumptions were evaluated by ex- amining the reliability demonstrated by M60 tanks during Operation BIG LIFT. (Approximately half the tanks used in the data analysis also participated in this operation.1'12) The tanks traveled an average of 369 miles during the maneuver and experienced a mean miles to "initial major-assembly replacement" or MMBF of approximately 2500 miles (5 times greater than the MMBF estimated from operating and maintenance data for major-part replacement). The major- assembly 48-hr-battlefield-day reliability demonstrated by these tanks is 0.85.

31



An analysis of the difference in the operation and maintenance of M60 tanks during readiness and maneuver operations provides insight in explaining why the maneuver-demonstrated reliability was greater than the reliability estimated from operating and maintenance data. When the M60 tanks were not on maneuvers they operated on restricted utilization policies. The 400-irile equivalent of the 48-hr battlefield day was spread over a period of 3 to 4 month.« This time period is approximately 50 times greater than the operating time of an actual 48-hr battlefield day.

The inspection and maintenance of each M60 tank required an estimated average work effort of 500 man-hours or 80 maintenance hours during the 3- to 4-month time period. This workload is greater than 1 man-hour/tank/ operating mile. Approximately twice as many hours were spent in tank maintenance, personnel maintenance training, and tank inspection during overall operating usage periods than were spent in tank operation (e.g., each tank experienced at least one quarterly maintenance period and several organizational and command inspections and underwent daily operational maintenance during the period considered). The recorded maintenance workload is approximately ^n times greater than the anticipated average maintenance workload (due to failure) during the actual 48-hr battlefield day (i.e., with a reliability of 0.95).

It is axiomatic that, when the time available for maintenance increases, the time involved in maintenance will increase, resulting in a corresponding increase in maintenance activity. Since this reliability analysis is based on operating and maintenance data, there will be a corresponding decrease in the estimated reliability as the time available for maintenance increases (i.e., the estimated reliability derived from the data decreases as the result of corrective maintenance, whereas the true mission reliability increases).

The axiom was demonstrated in Operation BIG LIFT. The maneuver tanks were prepared for the operation, and they proved they were in a state of materiel readiness by demonstrating a level cf reliability much greater than the estimated reliability. However, the maintenance effort required to prepare these tanks consumed parts and labor. The parts consumed were recorded as parts failure in the operating and maintenance records. These data show a decrease in reliability whereas the true reliability of the equipment was proved during the maneuver to have increased. Therefore the assumption that the required and the estimated reliabilities differ because the estimated reliability reflects the true mission reliability of the equipment is not valid. The estimated reliability is not a valid indication of the 48-hr battlefield reliability of the M60 tank or mission reliability of the Ml51 truck.

The 48-hr battlefield day is designated as a reliability requirement of the new MBT. The M60 tank was designed to have a battlefield-day capability of 21.3 hr.13 The mileage traveled during the 21.3-hr battlefield day is assumed to be 178 miles (based on the 48-hr battlefield day of 400 miles). The major- assembly leplacement battlefield-day reliability of the M60 tank is computed using Eq 1.

RMI)=e~l7"/25l,0=0.93

where MD is 178 miles and MMBF is 2500 miles. The fact that the maneuver tanks performed with a major-assembly

battlefield-day reliability approaching 0.95 indicates that if the M60 tanks are

32



properly maintained they are capable of accomplishing a level of reliability approaching that designated for them. Because these tanks had traveled approximately 1500 miles over their "retention life" and were probably experiencing a higher than normal wearout rate, the performance capability of newer tanks should be much higher than the demonstrated reliability on Operation BIG LIFT. (At least one company of the M60 tanks on Operation BIG LIFT had accumulated an average of 5573 elapsed miles. The retention life of the M60 is estimated to be 4000 miles.1)

The demonstrated reliability of the M60 tank on Operation BIG LIFT indicates that the true mission reliability of the tank will be much higher than the estimated reliability. Therefore the assumption that the required and the estimated reliabilities differ because the required reliability of the equipment is unrealistically high is not valid.

Reliability Estimate Confidence

The results of the data analysis have been presented in terms of the mean (average) reliability and the 0.90 confidence-level reliability. The difference between the mean reliability and the 0.90 confidence-level reliability indicates the precision of the measurement of reliability.

The mean reliability and the 0.90 confidence-level reliability of the M60 tank and the M151 truck are summarized in Table 8.

The dispersion within the distributed data was so large that the 0.90 confidence-level reliability approached 0.000 irrespective of failure-criteria variation (i.e., even when the failure criteria included only the replacement of major parts).

TABLE 8

Reliabilities of M60 Tank and M151 Truck

Mean 0.90 confidence- System reliability ieve 1 reliability

M60 lank O.OOOS 0.000" Vehicular system 0 0011 0.000"

Major parts 0.45 0.000" Minor parts 0.0025 0.000"

Main armament system 0.45 0.000" Ml 51 truck 0.68 0.000"

Major parts 0.86 0.020 Minor parts 0.79 0.000"

Approaching the limit 0.000.

The order of magnitude of the difference between the mean reliability and the 0.90 confidence-level reliability values is more clearly illustrated in terms of the MMBF in Table 9.

33



The large difference between the mean reliability and the 0.90 confidence- level reliability shown in the preceding tables illustrates that the precision of the reliability measurements is extremely low.

TABLE 9

MMBF and Confidence-Level MBF (or M60 Tank and M151 Truck

0.90 confidence- System MMBF level MBF

M60 tank 50 1 Vehicular system 59 2.8

Major parts 500 35 Minor parts 67 3

Main armament sy stem 500 32

M151 truck 1140 60 Major parts 3030 115 Minor parts 1907 125

Endurance

The results of the data analysis have brven compared to the requirements specified in the QMRs for new equipment to illustrate the difference between the required and the estimated endurance.

The required and estimated endurance of the MBT and the l/4-ton utility truck are summarized in Table 10.

TABLE 10

Endurance of the MBT and '4-ton Truck

Required endurance, Estimated endurance,

System miles miles

MBT Vehicular system 3000 Approaching 0.0

Engine 5000 100

Main armament system 3000 Approaching 0.0

/4-top utilitv truck 0.90 probabil ty 0.90 probability of 15,000 of 60

Major p^rts 0.90 probabil >y 0.90 probability of 25,000 of 115

The illustrated difference that existed between the required and the estimated endurance for the M60 tank and the M151 truck was as large if not larger than the difference between the required and the estimated reliability. The same arguments used to refute the hypotheses of the previous section discuss- ing the reliability differences also applied to the results of the endurance analysis.

34



CONCLUSIONS

The operating und maintenance data and the failure criteria used in the reliability analysis of the M60 tank and the M151 truck were reasonably ac- curate and acceptable. The estimated reliability and the endurance of these equipments were significantly lower than the requirements specified in the QMRs.

Given the previously mentioned definition of reliability, the following conclusions were derived:

1. The results obtained in the analysis were in effect a measure of the materiel-readiness environment and not the mission reliability of the equipment as specified in the QMRs. The measure of materiel readiness is a measure of the consumption and depleticr: of equipment(s) and not a measure of materiel reliability. The results of the analysis illustrated that operating and maintenance data cannot be used to measure mission reliability. Therefore the hypothesis that mission reliability can be derived from available failure data is not correct.

2. Although a quantitative reliability of two Army equipments was determined, the objective of providing specifiable measures against which developmental hardware could be evaluated was not achieved. An interpretation of the results of the analysis has indicated that verifiable measures of mission reliability must be assessed in test and evaluation programs that are controlled to predetermined standards.

35



Appendix A

TABULATED DATA SUMMARY

Figures

Al. Example of Weibull Probability Paper and Corip .tation Nomograms 39

A2. Miles between Failures 46 A3. Time To Correct Failure 47

Tables Al. Summary of Miles between Failures 40 A2. Summary of Hours To Correct Fail* re 40 A3. Miles between Failures, MBT M60 41 A4. Miles between Failures, APC M113 42 A5. Miles between Failures, Vj-ton Utility "i uck, M151 43 A6. Time To Correct Failure, MBT M60 44 A7. Time To Conect Failure, APC M113 45 A8. Hours To Correct Failure, '/^-ton utility Truck, M151 45

37



The operating and maintenance data compiled for RAC study RP-124 were processed aecordingto the failure criteria presented in the text of this report. The miles between failures and the time tocorrect failures were computed for major parts, all corrective maintenance, and the individual parts that fail for the M60 tank, the Ml 13 APC, and the M151 truck.

The raw data subsequent to processing were arranged in the increasing order of a cumulative distribution that was nonlinear. When data of nonlinear distribution were plotted on the nonlinear scales of statistical graph paper, the oata plot could sometimes be approximated by a straight line. A straight line has been plotted on Fig. Al to illustrate the procedure for computing distribution parameters that describe the characteristics of the data (e.g., the mean and the standard deviation). The procedure is as follows:

(1) Plot the cumulatively distributed data on the graph paper. (2) Construct a straight-line approximation of the plotted data. (3) Draw a straight line through the origin of the ß estimator parallel

to the data plot and read the ß value where the line intersects the small ß estimator scale.

(4) Draw a vertical line through the point where the data plot intersects the horizontal 0.0 axis. Read the TJ value where the line intersects the miles- bet ween-failures scale.

(5) Draw a vertical line through the ß value on the computation scale A at the top of the page. (Computation scale B is used for ß greater than 1.5.) Read the a/ r] and/u/r? ratios on their respective scales.

(6) Multiply the o/r) ratio times r) to compute a, and multiply the ß/r\ ratio times r\ to compute ß-

(7) The four Weibull parameters {q, ß, ß, and a) that completely describe the characteristics of the data are:

ß — Slope of data plot 7} — Characteristic value (for reference) ß — Mean value (central tendency) a — Standard deviation

The statistical parameters that describe the characteristics of the operating and maintenance data are presented in Tables Al to A8. Each line item in these tables is a straight-line approximation that represents a specific plot of individual data (i.e., the major-part items represent the result of plotting approximately 500 data points, the all-corrective-maintenance items represent the result of plotting as many as 2000 data points, and several individual part items represent the result of plotting only 2 or 3 data points). Each line item can be used to reconstruct the straight-line approximation of the plotted data of 117 graphs.

The tabular columns of each table labeled "0.90 confidence level* and "0.80 confidence level" indicate the parameter value that 0.90 and 0.80 of the

38


-8


H8 "ffi ,_•

H2 -™

t k|fr» Sd 9 II 1 3 § 3 S a § _ i 'itii i]11 i 11111111111111,111 i i i i i i i—1111 i—i—i—i i

I II llun T-ii ii ii "i i iff iu ili-mlii i if M ii hnn m i1 n nil 111111 i i i i i I IT

g 9 , , a ? .?■■..?.■■■?■■■ .a ~III i 1 i i i i r i i .. I i—... i .... i .... i .... a ca ©

g a|D3j uoijDindwcQ

-§

-8!

-SI

r-a

HÜ 1-

_


TABLE Al

Summary of Miles between Failures (Foilur* criteria—major port, all corrective moinltnanc«,

iampU. units—M60 tank, Ml 13 p«r»ortn«l carrier, M151 truck )

0.90 0.80

Characteristic confidence confidence Standard Vehicle Slope0 life" MMBF0 level level deviation0

Major-parts replacement

M60x .82 450 502 29 70 617 \160 y .87 530 570 39 94 652 Ml 13 z .89 500 530 .10 87 595 M151 .83 2730 3030 115 330 3686

All corrective maintenance

M60x .82 130 144 8 20 177 M60y .78 130 150 f 18 195 MU3z .86 210 259 10.7 41 305 Ml 51 .79 1050 1197 60 155 1544

Weibull distribution psramcters.

TABLE A2

Summary of Hours To Correct Failure (Failure criteria—major part, all corrective maintenance;

»ample units—M60 tank, M113 personnel carrier, M151 true« )

0.90 o.&o Characteristic Mean confidence confidence Standard

Vehicle Slope0 life0 time0 level level deviation0

Major-parts replacement

M60x 1.34 4.63 4.32 0.76 1.4 3.5 V160 v 1.10 3.50 3.67 0.49 0.96 3.33 M113 z 0.78 1.65 1,91 0.094 0.24 2.48 Ml 51 1.24 2.5 2.10 0.37 0.68 1.70

All corrective maintenance

M60x 0.265 3.0 3.27 0.2 C.5 3.81 M60y 0.83 2.73 3.03 0.18 0.44 3.66 Ml 13 z 0.68 3.6 4.68 0.123 0.38 7.20 Ml 51 1.36 1.6 1.42 0.305 0.52 1.05

Weibull distribution parameters.

40



TABLE A3

Miles between Failures, MBT M60

0.90 0.80 Percent Characteristic confidence confidence Standard of sample

Port Slope0 life0 MMEJF0 level level deviation0 foilH

hngine \ 2.0 6.r>0() 5,850 2000 3100 3,070 35 V 1.21 6,500 6.100 1000 1870 5,120 27

1 Yansmission X ludet.b ludet. lndet. lndet. lndti. Indel. c

y Met. lndet. lndet. lndet. lndet. lndet. c

Kngine and

transmission

X 2.6 ",.000 1,440 2050 27,50 1,850 45.5

y 1.1 6,500 6,273 850 1625 5,688 29

Final drive

X lndet. lndet. lndet. ludet. lndet. lndet. c

y Met. ludet. lndet. Mel. lndet. lndet. c

Sprocket X 8.0 3,450 3,243 2750 3000 587 19

y 4.0 1,300 3.8% 2500 3000 1.075 14

Track, shoe

assembly X 10.0 2,550 2,423 1900 2100 3C0 50

Y 6.0 3,350 3,109 2200 2.500 600 28 Road v:\ir els

and idlers

X 2,14 5,400 4.790 2100 2900 2,079 60

y lndet. lndet. lndet. lndet. Met. lndet. c

Excluder, retainer

X 11.52 2,1.50 ludet. 2025 2175 lndet. 98.4

y ?t.2 3,400 3,046 1700 2150 1,037 43

Seal and

excluder

X 10.31 2,350 lndet. 1875 2000 lndet. 99.9

y 4.12 3.000 2,724 1750 2100 744 51

Seals

x 5 5 2,960 2.680 2000 2225 570 98.5

y 3.4 3,950 3,547 2225 2700 1,157 17.1

Fuel tank, left X Irdet. lndet. lndet. 4300 7200 lndet. 9.75

Starter, elec-

trical

X o.a 9,200 10,442 665 1600 13,248 38 Hatt'jry

X lndet. lndet. lndet. 4300 7200 lndet. 11.1

Regulator.

generator

X 1.92 11,000 958 4300 5000 513 18.2

Link assembly

X 0.905 2,550 2,678 215 480 2,932 17.6

aWeibull distribution parameters. L r

Indeterminate. (Insufficient.

41

FOR OmCIAL USE ONLY


TABLE A4

Miles between Failures, APC M113


Part Slope" life" MMBFa level level deviation0 failed

First engine replacement 1.9 7,600 6,740 2.2 3.3 3,648 28

First transfer replacement 3.1 6,750 6,035 3.41 4.3 2,126 12.5

First transmission re placement 2.2 7,400 6,553 2.7 4.3 3,185 20

First differential replacement Indet.b Indet. Indet. Indet. Indet. Indet. 4

First final replacement 0.81 37,000 41,440 2.4 6.0 51,430 16.5

First sprocket replacement 2.35 6,000 5,315 2.35 3.11 2,400 33

First track teplacement 3.45 2,500 3,147 1.75 2.15 1,0; 2 70

Second track replacement 4.4 2,200 2,006 1.35 1.6 594 31

First pad replacement (with shoe replacement) 2.53 6,500 5,758 2.7 3.6 2,607 16.5

First pad replacement (with or without shoe replacement) 3.18 3,500 3,133 1.66 2.14 1,120 34

Second pad replacement (with or without shoe replacement) 6.5 2,000 1,863 1.4 1.57 337 32

First road-whee replacement 2.15 9,100 8,058 3.2 4.5 4,004 14

First idler-whee 1 replacement 2.0 10,800 9,574 6.0 8.5 9,990 98.5

First starter replacement 1.7 50,250 44,898 10.4 20.2 27,637 51

First generator replacement 1.4 17,500 16,188 3.6 6.1 11,375 11.5

First radiator replacement or repair 1.4 11,700 10,823 2.5 4.2 7,605 20

ftWeibull distribution parameters. indeterminate.

42



TABLE A5

Miles between Failures, '/4-ton Utility Truck, M151


Port Slope0 life" MMBFa level level deviation" failed

Kngine 1.60 54,000 48,438 12,700 20,500 30,780 20 Transnission

and transfer 1.37 58,000 53,012 10,500 18,500 39,150 19 Clutch 0.85 68,000 74,120 5,100 12,000 87,720 28 Carburetor 0.96 140,000 142,660 12,000 26,500 148,400 17.5 Fuel pump. engine 1.68 66,000 59,004 17,000 27,000 36,300 11.8

Generator, electrical 1.30 56,000 51,688 7,800 15,500 40,040 17.5

Regulator, generator 1.05 99,000 97,020 11,500 23,000 91,575 14.5

Belt, V-generator 1.45 82,000 74,374 18,000 30,000 51,660 9

Distributor 0.78 140,000 161,700 8,400 21,500 210,000 19.3 Spark plugs 0.80 72,000 81,720 5,400 13,000 103,680 25 Tires 1.60 83,000 74,451 19,000 31,000 48,140 11.8

aWeibull distribution parameters.

43



TABLE A6

Hours To Correct Failure, MBT M60

0.90 0.80 Characteristic Mean confidence confidence Standard

Part Slope" life0 time" level level deviation0

Engine 2.18 4.05 3.59 1.42 2.01 1.74 Transmission 2.65 2.1 1.87 0.9 1.2 0.76 Final drive 1.45 2.2 2.13 0.47 0.78 1.39 Sprocket 2.0 8.0 7.27 2.69 3.9 3.77 Track, shoe assembly 1.18 3.85 3.68 0.56 1.09 3.12 Road wheels and idlers 1.225 2.07 !.92 0.33 0.6 1.57 Road-wheel arm assembly, left 2.18 3.2 2.83 1.15 1.6 1.38

Road-wheel arm assembly, right 1.65 6.19 5.55 1.54 2.45 3.35

Seals 1.08 2.35 2.28 0.29 0.58 2.10 Excluder, retainer 0.95 2.4 2.46 0.22 0.5 2.,50 Link assembly 0.87 1.8 1.13 0.138 0.315 2.23 Starter, electrical 1.45 3.23 2.93 0.68 1.15 2.03 Modification kit, starter relay 2.51 '.32 1.17 0.53 0.72 0.50

Relay solenoid, starter-switch assembly 0.89 1.46 1.54 0.115 0.274 1.73

Generator, electrical 1.35 4.05 3.76 0.78 1.35 2.81 Regulator, generator 1.19 0.93 0.89 0.14 0.265 0.75 Fuel pump, engine 1.38 1.75 1.6 0.34 0.59 1.17 Fuel tank, left 3.83 10.02 9.23 5.6 6.8 2.66 Fuel-injector metering pump 1.93 9.4 8.34 2.9 4.3 4.51

Battery 2.53 0.52 0.46 0.21 0.27 0.20 Radiator, engine oil cooler 2.8 4.4 3.74 1.9 2.59 1.43

Radiator, transmission oil cooler 0.935 2.1 2.16 0.19 0.42 2.3-

Motor, dc, engine- exhaust assembly 0.92 0.63 0.64 0.06 0.121 0.70

aWeibull distribution parameters.

44



TABLE A7

Hours To Correct Failure, APC Ml 13

0.90 0.80 Characteristic Mean confidence confidence Standard

Part Slope0 life0 time0 level level deviation"

Kngines 1.7 5,5 4.92 1.4' 2.25 2.93

Transfer 1.05 5.4 5.24 0.8 1.5 3.17

Tram-mission 2.09 8.1 7.18 2.71 4.0 3.69

Differential 0.89 2.9 2.57 0.255 0.59 1.36

r inal drive 1.05 1.75 1.72 0.20 0.41 1.66

Sprocket 1.3 1.65 1.52 0.309 0.54 1.18

Track and shoe 0.885 7.75 8.22 0.615 1.41 9.3

Pads 0.94 2.0 2.06 0.19 0.419 2.18

Idlers and wheels 1.81 1.35 1.20 0.4

_b 0.6 0.68

Starters 0.61 1.18 1.73 0.11 2.90

Generators 0.96 0.98 1.04 0.104 0.22 1.00

Radiator 1.08 2.5 2.43 0.321 0.635 2.24

''Wcibull distri bution par ameters.

Insignificant.

TABLE A3

Hours To Correct Failure, %-ton Utility Truck, M151

0.90 0.80

Characteristic Mean confidence confidence Standard

Port Slope0 life0 time0 level level deviation0

Battery 1.81 0.84 0.75 0.25 0.37 0.42

Belt, V-gener-

ator 2.08 0.54 0.48 0.185 0.269 0.24

Carburetor 3.15 0.92 0.82 0.46 0.58 0.29

Clutch 1.31 2.6 2.40 0.478 0.82 1.85

Distributor 1.98 1.0 0.89 0.321 0.57 0.47

Control difference 1.35 4.4 1.03 0.82 1.43 3.01

Kngines 1.51 4.6 4.11 !>.08 2.7 2.99

Generator, electrical 1.69 1.5 1.33 0.4 0.621 0.53

Fuel pump,

engine 1.9 0.74 0.66 0.23 0.34 0.36

Radiator 1.6 1.85 1.66 0.459 0.738 1.05

Regulator,

generator 1.98 0.74 n.66 0.24 0.351 0.36

Shock absorber 1.7 0.92 0.82 0.255 0.39 0.49

Spark | lugs 1.81 0.71 0.63 0.22 0.32 0.36

Transmission 1.98 3.35 2.97 1.05 1.55 1.57

"Weibull distribution parameters.

45



$1

B *

8 t- CO 3

« X 1 " 1 i

1

8 1

HI

8 • 3

3 — 1U — a

3 si < u- H

9 z S o UJ « c LU * ■*? UJ m S'l o * s HI i •- 1 r.

O* S 3 s SI u

U- =

a st, i a s

% 'Noungiaisia

TT'TH T r -j-T-T-r- r r T_rT-r-p rfTTrn"

46



$1

3*

CO

s

o 2 h-

— _t>

'5 a. ii. E u „_

.< o c L. U ° HO*

u I I

o u

'Noiinaiaisio

i r i ■ i I ' l I ' 1

47



occurrences are expected to exceed. The difference of these parameter values indicates the broad variation in occurrences in these specific instances.

The distribution characteristics for the M60 MBT, i.e., miles between failures ai:d time to correct failure, are illustrated in Figs. A2 and A3. A review of the figures shows the accuracy of the straight line to approximate the data and the large characteristic variation for each parameter.

48



REFERENCES

1. Research Analysis Corporation, "Operation, Maintenance, a?.d Lifetimes of M60 Tanks, M113 Armored Personnel Carriers, and M88 Tank Recovery Vehicles (U)," RAC-T-460, Feb 65. CONFIDENTIAL

2. Dept of Army, "Maintenance of Supplies and Equipment, Maintenance Concepts," AR 750-1, 30 Oct 63. UNCLASSIFIED

3. , "Maintenance of Supplies and Equipment, Organization, Policies, and Re- sponsibilities for Maintenance Operation," AR 750-5, 3 Aug 64. UNCLASSIFIED

4. , "Army Equipment Record Procedures," DA Pamphlet 750-38, 25 Aug 64. UNCLASSIFIED

5. , "Army Equipment Record Procedures," TM 38-750, Jan 64. 6. , "Maintenance Management Field Command Procedures," TM 38-750-1,

Jul 64. UNCLASSIFIED 7. , "Maintenance Management National Agency Procedures," TM 38-750-2,

Jul 64. UNCLASSIFIED 8. , Combat Development Command, "Proposed Qualitative Materiel Require-

ments for the US/FRG Main Battle Tank," Dec 63. SECRET 9. , , "Proposed Qualitative Materiel Requirements for a New Main Battle

Tank," Sep 63. SECRET 10. Joint US/FRG, "Tank Development Program, Joint Military Characteristics—Main

Battle Tank 1970," Dec 63. SECRET 11. Dept of Army, Combat Development Command, "Proposed Qualitative Materiel Re-

quirement for Truck, Utility, 74-Ton," Combat Development Objectives Guide, Par 1636c(9), an unpublished working paper issued subsequent to Aug 63. UNCLASSIFIED

12. Research Analysis Corporation, "Study ot the Prepositioning Concept: Operation BIG LIFT Final Report (U)," RAC-T-462 (FOE), Nov 65. CONFIDENTIAL

13 Dept of Army, Ordnance Tank-Automotive Command, "Main Battle Tank Components and Concepts Review (U)," Mar 62. SECRET

14. John H. K. Kao, *A Summary of Some New Techniques on Failure Analysis," IRE Proceedings Sixth National Symposium on Reliability and Quality Control in Elec- tronics, 11-13 Jan 60, pp 190-202. UNCLASSIFIED

49


Documents

The following notice applies to any unclassified (including ... Analysis of the M60...of the M60 Main Battle Tank and the M151 %-Ton Utility Truck from Operating and Maintenance Data