J 00
00 o o PQ
TECHNICAL REPORT NO. 12068 /
O C_5
DRIVELINE ANALYZER M151 1/4-TON UTILITY TRUCK
FINAL REPORT
10 JULY 1975
TACOM MOBILITY S\
C2~ Northrop Corporation
,m At/ Anaheim, California
D D C
DEC 19 1975
D
For Diagnostic Equipment Function Research, Development and Engineering Directorate
MOBILITY SYSTEMS LABORATORY
ARMY TANK AUTOMOTIVE COMMAND Warten. Michu
V
4
M m - —f "*y#. uiui. • "' i "h ■■■ ■•■■ r ■■ ■-i.n»i»-.t.rf.1l;ifl-,T„'.lf„-B,,r.„ann...i.,ri.„■.«■ ,.w ■-..,.. •■■■^'iiiir'.Jt._Jj,.J,...
-W-1II1" ' "■y.j'.B'.j.ijmu'iiji '
W« miw SMIIH
«TtflUtMH _....
sr n
« ! MTiiiniM/miLAimiTr MOB
~"jk """ mil, M m VECIU m li
1 I \
l l !
I I I I I I I
i '"'■>:'»"f i I• • 8B
Technical Report No. 12068
ÜRIVELINE ANALYZER
M151 1/4-TON UTILITY TRUCK 0 /
/ fl ^FINAL REPirr, - I
iy
A. D./ßeBolt^
P. J./Leibert /
L. V./Rennlck f
Northrop Corporation
Electro-Mechanical Division
Anaheim, California
For
Propulsion Systems Laboratory
U.S. Army Tank. Automotive Command
Warren, Michigan
£ Contract/DAAE#7-73-C-0116 7
Distribution Halted to U.S. Oov't. agencies onlyf Test and Evaluation; JJJJ£C 1975. Other requests for this "document must bo referred to US/TlT^COM
j^ ^y
/j-
^'Mehr-i 5Y>»81
.£-■—
D D C
DEC 19 1975
TSEUTTE D
U?ßJ
ft 4
.«■
... -^"":'/'"
■'.-■•
■v:
,„,-.- ft» as, ,_ s 3 j Hsj at [ MHOHUSi««aMM«K0WWMSeiiWtMriKlti0^3ää ■:,.,....' ..... <:■,.;.-.:■..-■:. .,.,.-..-
"trr
The findings in this report are not to be construed as
an official Department of the Army position, unless so
designated by other authorized documents.
2 station of c«.rcU1 produc[s ln
not constitute an official -„,< -=h product8.
U1 end°"e"e"t » «"«v.! „f ]
I
I i ]
]
J
ii
"w»*!»«**^**™^ *N*«Ktt8gs^Ä^ ,.^i -j ''■--■^v'*.^;*'.--,'.-.;.
1 i J
•*v
ABSTRACT
The objectives of this development program were to determine the feasibility
of using data obtained from measuring discrete selected parts of vibration
signatures for Army vehicle driveline diagnosis and to provide the U.S. Army
Tank Automotive Command (TACOM) with an advanced development model driveline
analyzer designed to identify faulty driveline assemblies of an Army M151A2
1/4-ton utility truck. The development program performed by Northrop included
the correlation of simplified analytical models to empirical data, and the
design, fabrication, and functional test of one portable driveline analyzer
test set complete with sensors and a ramote display console.
Analytical models to determine how driveline components and assemblies gen-
erate excessive vibration were developed and used to identify vibration fea-
tures that were potentially good indications of faulty components. Following
this initial effort, the driveline assemblies of an M151 vehicle were instru-
mented for vibration and shaft rotation data. During road test, the data was
recorded on magnetic tape and was analyzed on Northrop computers to develop
power spectral density plots. These plots were used to provide empirically
derived excessive vibration features of good and bad assemblies.
The analytically derived vibration features then were correlated to the
empirically derived features to identify errors in either of the processes.
Good correlation of analytical and empirical data was used as criteria for \
selecting features that would provide the highest probability of detecting
faulty driveline assemblies and the lowest probability of identifying a good
assembly as being bad. From this analysis, vibration features were selected
that were compatible with a simplified road test and that were practical for
this application.
iii
-"■--V.>.;--._ ■, ;.; _;,., „^'.'...s^v.-,.:^ i^^,-.:lÄ^^'n."^'-:::J. U$$fä$3iJ&T/äfääfe \" ■■&'.■";--•"-'.'•-: ^A^stf-AiVW.,-^..,.^.;:- '■'-- ■'(.-:,■■ \ ji'.
i.v
A portable test set was designed and constructed, consisting of a driveline
analyzer chassis, a remote display console, and a set of sensor assemblies. ]
The driveline analyzer chassis contains amplifiers, programmable filters, i
counters, phase-locked loop circuits, and a digital computer with both random
access and read-only memories. The remote display console contains vehicle
operator instruction displays.
Digital computer programs were prepared for use witli the driveline analyzer \
test set to provide the vehicle operator instructions, gather the vibration
data during the vehicle road test, analyze the data, and display test results
in terms of faulty driveline assemblies. |
The driveline analyzer set was tested utilizing simulated test data and data
recorded on magnetic tape. Sensor installation compatibility, test sequence,
and vehicle control data acquisition processing and display have been verified I
on the M151 vehicle.
The test set was demonstrated to TACOM personnel using simulated and recorded
data in the labcratorv. Installation of the test set on an M151 vehicle was f
also demonstrated. After this demonstration to TACOM, Northrop conducted a j
number of road tests which demonstrated the ability of the driveline analyzer
to detect drive shaft, U-joint, engine vibration, and differential types of
faults.
I
This contract was conceived to assist TACOM in their goal of advancing the
technology in the field of Army vehicle diagnostics. The M151 driveline t
analyzer feasibility model has been delivered to TACOM for test and evalua-
tion. The driveline analyzer has verified many of the diagnostic concepts;
however, further testing and engineering improvements are necessary to deter- \
mine the adequacy of the diagnostic concepts.
1
FOREWORD
t
1
!
!
This report is submitted in accordance with Contract Data Requirements List
sequence number A003 of the U.S. Army Tank Automotive Command (TACOM) Contract.
DAAE07-73-C-0116. The contract was awarded in February 1973 and called for
Northrop Corporation to determine the feasibility of using vibration-diagnostic
techniques for Army vehicle driveline maintenance, and to provide TACOM, for
their test and evaluation, with an advanced development model Driveline
Analyzer designed to identify faulty driveline assemblies of an Army M151A2
1/4-ton utility truck. Work on the contract was performed at Northrop Corpo-
ration, Electro-Mechanical Division, Anaheim, California, under the technical
guidance of the TACOM Diagnostic Equipment Function Project Engineer, Peter Carland.
v/vi
"*-"" ■"-"'-—' -' »ok.
I I St rt ion
CONTENTS
P;ij...
I 1
INTRODUCTION 1-1
1.1 Background 1-1 1 .2 Ob H'ct i ves I -J 1.3 Scope ol Work I-.'
ESTABLISHMENT OF HARDWARE DESIGN REQUIREMENTS 2-1
2.1 Empirical Approaches 2-1 2.2 Analytical Approaches 2-2 2.3 Technical Approach 2-3 2.4 Selection of Fault Types 2-8 2.5 Analytical Analysis 2-10 2.5.1 Discrete Frequency Models 2-11 2.5.2 Vibration Due to Impulsive Forces 2-21 2.5.3 Cross-Coupled Vibration 2-22 2.6 Empirical Techniques 2-26 2.6.1 Vibration Recording and Analysis 2-27 2.6.2 Impulse Response Measurements 2-29 2.6.3 Analysis of Recorded Vibration 2-31 2.. Comparison of Analytical and Empirical
Results 2-38 2.8 Selection of Test Techniques 2-43 2.8.1 Wheels, U-.Joints and Drive Shafts 2-49 2.8.2 Differential, Transmission and Engine 2-5 2 2.8.3 Clutch 2-55 2.8.4 Selected Test Techniques 2-55
IMPLEMENTATION OF THE DRIVELINE ANALYZER DESIGN .... 3-1
3.1 Overall Concept 3-1 3.2 Central Processor Requirements 3-3 3.2.1 Central Processor Implementation 3-4 3.3 Vehicle Operational Display Requirements . . . 3-8 3.3.1 Operator Display Implementation 3-8 3.3.2 Description of the Displays 3-9 3.4 Test Result Display Requirements 3-10 3.4.1 Test Result Display Implementation 3-12 3.4.2 Description of the Test Result Messages .... 3-12 3.5 Sensor and Signal Conditioning Requirements . . 3-13 3.6 Driveline Analyzer Specifications 3-14 3.7 Hardware Pictorials 3-17
I vii
■..V:'~^i*l.-^J<SrtsJ^Ml»>«*i^/4aV^iv'i^';U^Ä^''.Ä.;'. ^vfe;; ;*.;v:;.ivi,t;"'
CONTENTS (Cont iniied)
Section Pajje
4 TEST AND EVALUATION 4-1
4.1 Bench Tests of the Drivelino Analyzer 4-2 4.2 Vehicle Compatibility Tests 4-6 4.) On-the-Rond Tests 4-6 4.1.1 Drive Shaft U-.loint Tests 4-7 4.3.2 No Weight on Drive Shaft Test 4-H 4.3.1 6-Ounce Weight on Rear Drive Shaft lest .... 4-10 4.1.4 6-Ounce Weight on Front Drive Shaft Test . . . 4-11 4.1.5 On-tbe-Road Differential Test 4-14 4.1.6 On-the-Road Test at Wheels 4-16 4.4 Summary of Test Results 4-lb
5 SUMMARY 5-1
6 CONCLUSIONS 6-1
7 RECOMMENDATIONS 7-1
7.1 Fault Simulation 7-1
8 DISTRIBUTION LIST 8-1
I)D FORM 14 73
ILLUSTRA'i IONS
Figure Page
2-1 Driveline Analyzer Development Procedure 2-4 2-2 Ball Bearing Relative Frequencies 2-12 2-3 M151 Basic Drive Train Data and Sensor Locations .... 2-17 2-4 Third-Gear Driveline Shaft and Gear Mesh Races 2-18 2-5 Effects of Worn Teeth on Vibration Spectrum 2-21 2-6 Typical Coupled Vibrations 2-23 2-7 Generalized Technique Used to Identify Excessive
Vibration with Elimination of the Effects from Other Assemblies 2-24
2-8 Data Process 2-28 2-9 I' 'ilse Response Measurement 2-30 2-lü M 1A2 Driveline Impulse Responses 2-30 2-11 Typical Analysis of Vibration 2-31 2-12 Synchronous Analysis of Vibration 2-32 2-13 Typical Computer Printout 2-33 2-14 Interpretation of PSD Amplitude and Frequencies .... 2-36 2-15 Relationship of Plot to Shaft Rotation 2-38 2-16 PSD of Rear Differential 2-39
viii
i 1 i
I 7.2 Test Program 7-1 i 7.1 Rotation Sensors 7-2
(
i I I 1 I I I
i
I Figure »»LUSTRATIONS ((. "»tinned)
2-17 2-18
3-1 3-2 3-3 3-4 3-5 3-6 3-7 i-8 3-9 4-1 4-2
4-3
5-1
High-Frequency PSD Plot Driveline Analyzer Feature Extraction and Analysis Processes
Drive line Analyzer Block Diagram Power/Performance Comparison of Various Minicomputers, Nortlirop CMOS Processor Remote Display Console Panel Drivellne Analyzer Chassis In Place Remote Display Console in Place Front Differential Sensor Assembly Instated Rear Differential Sensor Assembly Installed Transmission Sensor Assembly Installed Synchronous Signal Simulation riench Test - Input Synchronous Signal Amplitude Versus Driveline Analyzer Reading Operated in Drive Shaft Test Mode
Laboratory Tests Using Magnetic Tape Recordings from On-the-Road Test
Driveline Analyzer Block Diagram . -
Pagj?
2-41
2-46 3-2 3-6 3-7 3-9
3-18 3-18 3-19 3-19 3-20 4-4
4-4
4-6 5-3
I
.Table
2-1
2-2 2-3 2-4 3-1 1-2 4-1
4-2
4-J
4-4 4-5
4-6
5-1 5-2
TABLES
Shaft Speed Tran
s as a Functi iransini8,i0n Gear Ü °f Mii*s Per n„,
.Gear and Shaft pi! Ran*e ■ . . . . "°U Shaft Fre
r and
^r^..^rs- <*«<* ^ "n-the-ltoaj 0r,,
•ÄaS-S ?-; *."-•« ood
»"^CSr*1--- , JJ MPH On-the-Road Drive Shaft Te Drive Shaft, 35
Summary of Imbal
Oun
MPH st. 6-0unc
ce weight on'R;a;
e Weight oVpront °-the-Road™a""d Drive Shaft'Test » .
35 MPH Differential Test r Res"lts
"*«-«oad Wheel Test ™«V * ' • »»Each Test . . SSt« Third Gear, 35 MPH, F0
Compliance to Hardware e Summary nf w ,a,uwar<? Specif loan y ot Hardware/Snff.. IIlcation . .
'«-/Software Verify verification.
erential,
ur Legs'
2-17 2-19 2-20 2-41 3-2 3-5
4-9
4-11
4-12 4-13
4-15
4-17 5-2 5-5
ix
****"■ •■■'-.,
I
I I I 1 I I 1 I I I
SECTION 1
INTRODUCTION
This document is the final technical report for a Northrop Corporation advanced
development contract for one experimental model driveline analyzer. The per-
formance goal of the M151 vehicle driveline analyzer is to identify the faulty
driveline subassembly that is the source of abnormal vibrations.
The report is in accordance with Contract Data Requirements List sequence
number A003 of the U.S. Army Tank Automotive Command (TACOM) contract, DAAE-07-
73-C-0116. The contract was awarded in February 1973 and called for Northrop
Corporation to design and fabricate for TACOM test and evaluation, a driveline
analyzer designed to detect faulty driveline subassemblies for the Army's
M151 1/4-ton utility truck. Work on the contract was performed at the
Electro-Mechanical Division, Anaheim, California, under the technical guidance
of the TACOM Dia^ostic Equipment Function Project Engineer.
1.1 BACKGROUND
In the U.S. Army vehicle maintenance system there exists a difficulty for even
the most experienced mechanic to correctly diagnose a driveline subassembly
rnaifunction. Good assemblies are invariably sent to the depot level for repair
when such repair is not appropriate. After one of the driveline subassemhlies
is replaced, it also becomes necessary to validate the repair afte is made, :er replacement
This driveline analyzer has been developed with a long term developmental
objective to become an aid to the Army mechanic in solving the diagnostic prob-
lem of accurately identifying a faulty and/or misaligned driveline assembly
within an Army vehicle and also to serve as an aid in verifying that the sub-
assembly replaced is installed correctly and did, in fact, correct the fault.
1-1
-'iSt *v- ^ytfe*Ü.MH ...V^'fa.,;.;,
■ii-■-•^■r-
1.2 OBJECTIVES
The objectives of this development program were to determine if it is feasible
to use vibration techniques for Army vehicle driveline diagnosis and to pro-
vide TACOM with a.i advanced development model driveline analyzer designed to
identify faulty driveline assemblies of the M151 1/4-ton utility truck. The
development program performed by Northrop included the correlation of simpli-
fied analytical models CO empirical data, and the design, fabrication, and
functional test of one portable driveline analyzer set complete with sensors and remote display console.
1.3 SCOPE OF WORK
The contract specification required Northrop to furnish the supplies and ser-
vices necessary to design and fabricate, for TACOM test and evaluation, a
driveline analyzer designed to detect faulty driveline assemblies in the M151 series 1/4-ton trucks.
The general performance requirements for the driveline analyzer were estab- lished as:
a. Design hardware that will detect t*e following:
1. Slipping clutch
2. Drive sh^ft imbalance
3. Drive shaft misalignment
4. Excessive differential vibration
5. Wheel assembly imbalance or bent axle
6. Excessive engine vibration
7. Excessive vibration due to wheel U-joints
8. Excessive transmission vibration
9. Excescive vibration due to drive shaft U-joints
b. Have packaging dimensions approximately 12 x 18 x 18 inches
c. Be insensitive to vibration during transportation and suitable for
use on a moving test vehicle
I 1-2
d. He powered by vehicle's electrical system
1 1 !
I I 1 I i I I I I I I I I 1
e. Be simple enough to be installed and operated by one mechanic at
Direct Support level of maintenance
f. Have self-testing capabilities to determine if it is operating
properly and within tolerances
g. Have automatic test capabilities in which each driveline subassembly
is tested and results are displayed to indicate good or bad condition
h. Be capable of being installed in less than 10 minutes and having a
te.'t time of less than 8 minutes.
The following project tasks outline the approach used in reaching the objec-
tives and meeting the hardware performance requirements above.
a. Perform ai. analytical projection of the specific sources of vibra-
tion expected to be found in the M151 driveline subassemblies.
b. Obtain empirical data and correlate to the analytically determined
features. From this, make the final selection of test technique and
finalize the driveline analyzer configuration.
o. Design, construct, and test the driveline analyzer set.
1-3
.;ta#^«"'-«*i*i*««*^'''Äi» i s& ,»- ■*-.» r *r««*«v, '"v - t,.-»m.» » ti-to.*.'*
FRECKDHO PAM UUNK-NOT t waj
SECTION 2
ESTABLISHMENT OF HARDWARE DESICN REQUIREMENTS
Wlien Northrop established the technical approach to defining the design require-
ments for the driveline analyzer hardware, consideration was taken of the suc-
cesses and failures of similar programs that had attempted to use vibration as
a diagnostic tool.
2.1 EMPIRICAL APPROACHES
1 1 I I i I I r i i
Empirical approaches involved gathering large volumes of machinery vibration
data for various types of machines. Vibration data was analyzed using var-
ious statistical techniques to develop discernable features. Maintenance and
failure reports were gathered for each of these machines. Attempts were then
made to correlate specific vibration features to types of machinery fault.
Theoretically, this approach is sound and should provide good diagnostic and
prognostic features.
One of the empiri-'U approaches has met with some success when applied to
simple rotating machinery; however, this approach has been frustrated with
failure when Northrop attempted to develop reliable diagnostic and prognostic
tools for complex gear train vibrations. This failure could be related to the
highly complex nature of the vibration spectrum and amplitude. Each component
radiated at a variety of frequencies and amplitudes which were dependent on
machine speed and load and the good or bad condition of the part. A faulty
component also coupled mechanical stresses to good components and caused them
to modify their vibration characteristics.
Our attempt to cross-correlate highly complex vibration data to sets of com-
plex, unreliable failure reports resulted In frustrating lack, of success.
2-1
:.-: ■«:- -';.* ■■;V^wj«,.-^ij> '^ ,'-,.^.äii^;5©iÄ; *■; :■
I.J j.n, || nwftl' 1
B
1). It was impractical to expect to maintain or even locate a vehicle
with only one controlled fault. Normally, there would be cases where
unexpected faults would exist.
c. The data required to define the vibration related to a particular
raulty assembly must not contain a large number of complex terms.
Complicated terns would require extensive development and would indi-
cate test equipment too ; implex to be practical for this application.
d. it would be impractical to expect the vehicle to be operated under
precisely controlled conditions during the test.
e. Output data must not require interpretation by experienced personnel.
It must identify the most probable part causing excessive vibration.
f. It would be impractical to solve both the diagnostic and prognostic
problems in one step. The initial objective should be limited to the
identification of driveline assemblies which were causing excessive
vibration.
g. It is impractical to expect universal agreement as to the goodness or
badness of the vehicle's components.
h. The approach must recognize the limitations of empirical and analy-
tical approaches.
2.3 TECHNICAL APPROACH
Procedures used to develop the driveline analyzer technology are shown in Fig-
ure 2-1. The approach uses both analytical and empirical techniques. The two
techniques complement each other in a way to overcome some of their individual
disadvantages. The limitations of the analytical and empirical approaches, the
time available to conduct tests and develop test techniques, and the limited
availability of vehicles with various faults were constraints on the combined
approach.
2-3
■%-m^-i ■■■r-;,ÄiilV^.'-' ;.■>>■ ■■■■■,.',',:---.-;o--;'-W-"' .-'■r.^y«:?..---.;x'!'i:.
MT Aftl i',M »I »I ' .HM*** t K| niXfll Mt M«. AM» Arn |[ A ! if M|
• IfHNTlf V «Ml um iH*i'Hi.' f «K'l V'**AT
• TR*NWII*StnN •(iftivii INI •OUtf RINHAl
• Ulf Mt f •* SU**ORI IfVft WfCMAMir
• »RACTfC.AI (0»*i(*(i»i||vii UM
• »ORTAli f *|0« tOST •|A|V TO LOO» U*
• «Ml lit AND *«H I
VARIOUS TflT TECHNIQUES 04 V| l C*l 0 RV NO*TMROF
Stiff i PtVth
• ROTFNTlAi FEST TICMN'OII
• 1**1 ol o*r* AiouiRto ti IMFMFMFNT TM|M
UNU*|A«l| T|CMN(OH|t»IC*LW O*
• coMfitmtv or EOO"*MFNT
MfajmfWUTMo« IARQF voiiMrinr TftT DAT* FROM GOOD A*«0 IAO AUfWtlKf
• in»FA<>«A«it,Tv O* ftUCCIU
MICMANICAI OATA »OK Mttl DtPVIlINTI ANA! Vtl(.A( ANA [ Y%\% OF
ORlVlltNI AUfMfll ■( , TO IDENTIFY MO»FRTtfi ■» OIAC.NOTTIC FFATUREI
T
'Mt I MflHi. v#w*ctf« "h
Z3 RECORD Vi«RAT(ON DATA FROM Mifti D«tvf|tA» ASS4MRI (IS
Mit' VfMICi.ll »■TMi GOOD
SOME «AU
MAGNETIC TAMS
ANAL VTlCAll V DfRIVEO FEATURES
JZ roNoncT sncruAi AND
STATISTICAL ANALYSIS Of VHAAT'ON DATA
JZ FM*iRlCAt DATA
AIOU"Rf MINTS FOR IMPROVEMENT IN ANA! VT If Al. ANA! VMS
- • APPARENT VIBRATION FEATURES NOT MIVICKALV ICH NT ml 0 -{
_£_ _J
COMPARISON ANA! Y$H
Vi«RATiON FlAl»*f S TM*T ARE COMPATifHF WITH TEST TECHNKKJlS AND APPt irATION RFOUIRt MINTS
J
- 1
DI SIGN SMCtFiC HCMNIOUfS ID« tntNTII VINO DRIVE LINE ASilMRi HS WITH EXCESSIVE **~ Vti*ATlON
REOUtMEMENTS FOR IMPROVED Of SIGN
'SIGNS THAT HAvl LOWER •• "JBABiiiTV 0» SUCCfSS ANO REQUI^F FUttHJffl TO EVALUATE DESIGN
\ / DESIGNS THAt HAvF HIGH fACSABILTV OF *>. CESS
SRI ADROARp« LECTIO TECHNIQUES
TEST BREADBOARD Df SIGNS
MAGNETIC TAU HFCOflDINGl Mill ROAQ TlfT
J_ OF SIGN DATA
i
I i
I
OF SIGN AND MANUFACTURE: IXPtHIMINTAi MODCL
«ADJUSTMENT Of DECISION PARAMETERS
TEST EXPERIMENTAL MOt. '. AT NC.MTHROP ANO NEARBY MILITARY BASIS USING Mtftl VEHICLE*
MAGNETIC TAPES M>«1 ROAD TEST
ANALVTICAl ANALYW TU OF TERMINI IMMIOVED SETTINGS OF DECISION PARAMETERS
r~^ EVALUATION OF TEST OA\A
HI COMMENDATIONS FOR FURTHER DESIGNS ANO IMPROVEMENT« IN ANALYSIS TECHNIQUES
DRlVFUNE ANALYZER DEIIVtRED 10 TACOM
Figure 2-1. DRIVELINE ANALYZER DEVELOPMENT PROCEDURE
2-4
1 i 1 I
As depicted in the driveline analyzer development procedure of Hgure 2-1,
t lie following steps were taken:
a. LsLablisiiment oi 1'erj ormance Requirements and Constraints
1. Development application w is limited to one chicle to reduce the
complexity of having to test several vciiicles. This allowed more
time to concentrate on one vehicle, thii." enhancing probability of
success and proof of feasibility,
2. Failure levels were define ' s excessive vibration; thai is,
vibration levels tnat would cause a vehicle operator to request
maintenance ai.tiin. This allowed the basic principles of selected
techniques to be tes'ed and also eliminated the time-consuming
tasks of determining whether a marginal assembly was good or bad.
This constraint eliminated tests that related more to the engi-
neering design evaluation or to the production testing requirements
for the venicle.
3. The test set was to be designed to detect which of the driveline
assemblies was causing the most serious vibration. There would
be no attempt to detect which component within the assembly
had failed.
4. The test .^et was to be practical for use on a simple road test.
This meant that precise speeds and loads could not be a require-
ment and that road vibration must be expected as an input to the
system during the diagnostic test.
5. All techniques considered must be compatibl with the concept of
a test set that is portable, low cost, and practical for attachment
and use by a Direct Support level mechanic.
k* Selection of Candidate Test Techniques - Candidate techniques were
selected from various machinery vibration test techniques that had
been developed by Ncrthrop and others. This selection process took
2-5
_-k- i. ■*'■•■ :j-*.v^"'-. ... ■,;'■;.. ,:.•?...-., ■■ '.^V: VKSVOVM... ^■■■^^^.^.■i^xi&i-M'zw ^.^fe;fev^.:i^■y:v..:;^W ^■<--l^^^'.'.^\,r:^ ■£<&■ ,
into consideration the constraints and test requirements developed
by the previous step. Techniques that required large expensive equip-
ment or which did not provide identification of faulty assemblies
without manual interpretation were discarded. Techniques that required
large volumes of empirical data or involved long expensive analysis
were also discarded. The probability of success based on experience
with the technique and engineering judgement was also used as a
selection criteria.
Analytical Analysis of Drivellne Assemblies - The Ml51 vehicle
driveline assemblies were analyzed to determine characteristic vibra-
tion features that were applicable to the selected test techniques.
Quantitative data was provided for frequency and qualitative data was
provided for amplitudes or left for empirical evaluation. Also, math-
ematical models were developed for cross-coupling factors between
assemblies.
In all cases, attempts were made to limit the analytical work to prac-
tical levels that could be accomplished within the time and cost con-
straints of the program.
Empirical Analysis of Driveline Assemblies - This analysis consisted
of recording and analyzing vibration data to develop power spectral
density plots. These data were recorded while operating the vehicle
on the road. Analysis was oriented to supporting selected test tech-
niques and to complement the analytical analysis. Other tests were
conducted to determine the relative coupling of vibration between
assemblies.
Comparison of Analytical and Empirical Data - Analytical data was
used to identify sources of frequency lines in the vibration data and
to provide explanations for vibration characteristics. The empirically
derived data was used to provide characteristic amplitudes and coupl-
ing factors that are much more difficult to derive analytically. The
comparison process provides an accuracy check on both analytical and
empirical work. Areas which did not correlate and could not be ex-
plained pointed to the need for additional analytical and empirical
2-6
W «Mant«i>w --
work. In ill cases, these investigations wen- limited to areas t li.it
would support tin' objectives of the driveline analyzer and would
support the selected test techniques. The end result was a set of
vibration features that .:ould he used for driveline diagnostic
analysis.
f. Test Technique Design - A detailed design ol the selected test
techniques was conducted. This design was based on test requirements,
selected test techniques, and on the analytical and empirical data.
This process considered the impact on hardware and the possibility of
using common hardware to conduct more than one test.
8- Design and Manufacture - This procedure involved the selection of a
■pecific test equipment design and the fabrication of an experimental
model of the driveline analyzer. Trade-off studies were conducted
to determine whether analog or digital processes should be used, and
to determine whether or not the driveline analyzer test set should use
a hard-wired or stored program technique. The use of commercially
available or specially designed equipment was also considered. The
design was completed and an experimental model of the driveline
analyzer was manufactured using good commercial practices.
h Test of the Experimental Model - Bencli tests were conducted to
demonstrate that the driveline analyzer performed to the design re-
quirements. These bench tests were conducted using laboratory equip-
ment to provide simulated vibration data. In some cases, actual
vibration data recorded on magnetic tape was used.
The driveline analyzer test set was attached to the M151 vehicle to
demonstrate compatibility with the vehicle and performance of the
test. Limited evaluation of the driveline analyzer test set on the
vehicle, using good and bad assemblies, was conducted.
2-7
'A- «ito
t ... ■■..■-. .-.."■ ... - ..-"- .: -/i/---^--—■" — ■^•■--■~^:y-~~^--^-~r-K^ ä-ii£3.'-£': ■-■*&^r-.-^->>^'*4*^- ; .rf^*-;v';^::./;i;,^V.Ä*:,^^fi-vi^'^i^»K:
^m'
2.4 SELECTION OF FAULT TYPES
The requirement for fault detection was to identify driveline assemblies that
were causing excessive vibration. Excessive vibration was defined as an on-
the-road vibration level that would cause the vehicle operator to request
maintenance action. In addition, the driveline analyzer was to provide an
indication of the relative degree of badness. The clutch assembly was to be
tested for slippage.
Driveline assemblies were fault identified for excessive vibration due to:
a. Engine
b. Transmission
c. Rear drive shaft imbalance
d. Rear drive shaft alignment or U-joint bad
e. Front drive shaft imbalance
f. Front drive shaft alignment or U-joint bad
g. Rear differential
h. Front differential
i. Rear right wheel imbalance, bent wheel, or bent axle
j. Hear right wheel U-joints
k. Rear left wheel imbalance, bent wheel, or bent axle
1. Rear left wheel U-joints
m. Front right wheel imbalance, bent wheel, or bent axle
n. Front right wheel U-joints
o. Front left wheal imbalance, bent wheel, or bent axle
p. Front left wheel U-joints
q. Excessive clutch slippage
During the course of the contract, it became desirable to define the above
qualitative faults ir: terms of quantitative mechanical deviations and to pro-
vide examples of these faulty assemblies. It was also desirable that these
faults be typical of thosa in field occurrences and that the probability of
occurren.-'u be known.
2-8
This requirement to provide definitive data became on<> of the more difficult
tasks in the program and was only partially fulfilled. The Army could not
provide statistically significant failure reports on the driveline assemblies.
The only significant failure identified was a manufacturing deviation in the
differential which was causing failure in the field. Also identified was that
axle U-joints and drive shaft U-joints were also major sources of failure.
itne example of a differential assembly that had failed in the field and one
example of a differential housing that had a manufacturing flaw was provided
by the customer.
ihe alternative to the lack of other actual failed assemblies was to modify
good assemblies in an attempt to simulate faults that might occur. Some of
the techniques used to simulate failures were:
a. Removing of a needle bearing from a U-joinf
j b. Adding weight to drive shaft to cause imbalance
c. Adding weight to wheels to cause imbalance
d. Improperly adjusting the differential I I e. Hlsadjusting the clutch pedal to cause clutch slippage
f. Adding wrong size shims and spacers.
in many cases, the degree of part modification did not cause sufficient vibra-
tion to be considered excessive. Only one M151 vehicle with one set of spare
parts was used or. the program. As ,',. result of these deficiencies, it became
necessary to estimate the vibration that would be caused by bad parts using
extrapolation from good part data. One exception was the wheel and drive shaft
imbalance which was simulated by adding weights to the drive shaft and wheels.
The combination of empirical and analytical analysis was used to identify
vibration frequencies that were due to various gears, bearings, and shafts
within the driveline assembly. By assuming that these various parts had
failed, a partial estimate of the effects of failure were made. Road tests
later showed that these estimates were within reasonable ranges.
1 I I I I I 2-9
J.5 ANALYTICAL ANALYSIS
Theories liave been developed that relate the amplitude and frequency charac-
teristics of machinery vibration to moving parts within the machine. The
theories were used in conjunction with M151 driveline dimensional and rate of
motion data to predict the nature of its vibration and to predict how fre-
quencies and amplitude would change when various components within the drive-
line failed. Both qualitative and quantitative types of information were
derived. Quantitative data was primarily limited to frequencies generated by
moving parts. Amplitude data, changes as a function of degree of failure, and
the effects of cross-coupling between assemblies were limited to qualitative
derivation. These theories could also be used to predict how vibr.ition char-
acteristics would change as a function of vehicle speed, transmission ,;ear in
use, vehicle load, and path of operation.
The vibration characteristics derived in this way could be grouped according
to type of source and information provided. They are:
a. Vibration frequencies that can be directly related to the movement
of a specific part. The sources are mass imbalance, misalignment of
shafts, meshing of gear teeth, and bearing contact.
b. Mechanical ringing that is similar to the ringing of a bell when
impacted by the clapper. Frame assemb ies, shafts, ind gear disks
vibrate in this manner. The frequency of vibration is characteristic
of the part dimensions, material, and how it is mounted. The vibra-
tion mode, amplitude, and modulation is a function of the mechanical
forcing function.
c. Cross-coupling of forces and vibration from one assembly to another.
There are two modes of coupling: those related to variations 1n tor-
sional force transmitted along the coupling shaft, and vibration
coupled through frame and shaft members.
The analytically derived data was compared to data from actual vehicle test.
Differences indicated the need to revise analytical approaches or to improve
the data acquisition and analyses processes.
2-10
1 I I J I I
—!r~4 -,.—— •
As iho theoretical models wore confirmed by experimentell data, the model could
he treated with more confidence and extrapolated to cases that were not practical
to he confirmed by experiment. This theoretical understanding oi how vibration
is generated ami the nature of iis change as a (unction of part failure was used
to develop and select Improved test techniques.
2.5.1 IH_screte •" rJ'_9ü^l_ll<' Models
Moving parts such as shafts, bearings, and gears within a machine generate dis-
crete frequencies that can be directly correlated to the motion of the part
and to the repetitive contact of one part with another. The amplitude jf the
frequency is proportional to mass rate of displacement, contact force, and
their time rate of change.
For example, sinusoidal force or vibration generated by a mass imbalance in a
rotating member is known to have a frequency equal to the rotation rate, and
its amplitude is known to bo proportional to the mass rate ot displacement. In
their pure state, these vibrations are single frequencies that are coherent with
shaft rotation. Bent or misaligned shafts generate a similar set of frequencies.
A ball, bearing generates a family of frequencies that are functions of shaft
rate, inner and outer race dimensions, number of balls, etc. Figure 2-2 shows
a derivation of these frequency elements.
The inner race is allowed to rotate through an angle ..>. such that the original
contact points (A) of ball and inner race are now at B. and B , respectively,
and the ball contact point with the outer race is at point B . The contact o
surface covered on the inner race C. = contact surface covered on the ball (• l s
= contact surface covered on the outer race C , because there was no slippage.
= Angle through which the ball has rotated about the bearing axis.
(ball train angular motion)
= Angle through whicli the ball has rotated about its own axis.
Angle through which the inner race rotated.
2-11
'*■2\1 Bt>Jfing Relat Ive Frequencies
Consider ball bearing where r = inner rate radius,
and r, ■ ball radius. b rQ = outer rate radius,
No slippage between races and balls has been assumed in the followinj derivations.
I I I !
3300
J
3
I i
Figure 2-2. BALL BEARING RELATIVE FREQUENCIES
2-12
I I I I I !
1 1 i
1 I I
|! •" Diameter nf outer race o
D ■ Diameter of inner race
Ball Turn Kate
c = c = t; 1 ll s
b =
2C o
I) 0
(' V'o 0 2
'h =
2C. l
I).
(1)
(2)
(3)
u. D Combining 1 and 2r C, = -H-2
i 2
and combining this with 3: (. - . ) U = . (U ) i b i bo
solving for ., in terms of . . b i
i). i
'b I), + I> 'i (4) 1 o
This is the rate of precision of the ball train about the bearing axis; irreg-
Ball Spin :. ,
t ularity of ball's size,
1 I
C C 2 so
~s 1) " u s s
i 2C = J, Ü
o bo
combining 5 and 6 we get
I I) o
"'s "b I)
(5)
(6)
(7)
2-13
,*'-Vv'&*»-v;;.< S*Ütt}^ -ij&.-,T-'&v!v'--.>. )>■/-£ ■-.■'Air ^ä^SJfSäS&äEi * :• . ■ ■-■ - . '&& - ■ • -'»
Replacing .. with . from 4 wc got:
1) 1) . . . L--S. (8) s i I) + I) L> v y
i o s
is the spin rate of the ball; a flat spot on the ball will cause a dis-
turbance at twice tliis rate due to its passing over the inner and outer race.
Inner Race Rate
(;i " (i " V T (9)
Frora4 b " (iTT-5")i i o
Ci ° (wi" innr -i> Di 1 o
ci ■ "i (1 -D-TV) Di (10)
i o
If there are M Balls, their angular spicing is < = — ^nd the distance between
balls along the inner race is C... = OÜ./2, in l
2« Di CiM = IT (11)
Ci
A point on the inner race will contact a new ball at the rate of -—. IM
From 10 and 11
C| ^ (1 " ITTD-> UI
i _ 1 o
CiM ä °i
Ci '"i Di ~- = — (1 " n * n ) M (12) CiM n Di + Do
This is the rate of a disturbance generated by a fault on the inner race.
2-14
I I I I I I
Outer Kace Kate
C = ... D (13) o bo
TT The distance spacinc of the balls on the outer race is C .. = — 1)
on M o
C ,, D u).M q_ b o _ b
L'oM -" D M o
and from 4
^- - - °i M C M i n X n ? (14) oM D, + D TT
i o
This is the rate of disturbance generated by a fault on the outer race.
NOTE: The above equations are in terms of radians; when used in
terms of frequency, they are 2nf.
For practical calculations of bearing frequency, the following were used for
normalized multipliers of the bearing shaft rate.
Nf ■ 1 ■ shaft frequency
= ball train precision frequency DJ + F) i o
D N =» 2N, (-—) = bail spin frequency s b b
s
N. = (1 - N )M = inner race frequency
2-15
N = (N, M) = outer race frequency o b
The above numbers apply to Timkin roller bearings as well as ball bearings
that do not have angular contact. For angular contact ball bearing, the con-
tact angle must be known.
mm*m*i* ...m-nun'Warn »ft, . ,r , ,- ,u,„; -*~U..~;«Ä^ÄSSÄ *.a>JW«- '
(»ear Frequencies N, UETH
INPUT«
N,
OUTPUT n7
NjTEETH 3300
R2 " N^l
f = t< = eccentric gear or shaft on input
Nl f„ " R_ » — f * eccentric gear or sliaft on output »2 £ W» 81
f " N R = N f = gear mesh frequency gm 1 1 1 Sj
N - N2 f , ■ —- f„ = gear beat frequency gb N2 gi
Normalizing to input sliaft, we have
N,
g2
I ft'
gm
gb
Nl-N2
Determination of various shaft speeds as a function of vehicle speed and
transmission gear is necessary for this type of analytical analysis. Figure
2-3 is a model of the M151 drive train showing the various speed ratios and
proposed sensor locations. From these data, the wheel, drive shaft, and engine
speeds were derived as a function of transmission gear ratio and speed in miles
per hour as shown in Table 2-1.
"I
1 I I I I
2-16
*$£%W$i^sJ^^
I 7 R. in, si
6 ;i7
IGNITION SENSOR
(SI SLIP
ENGINE
ROTATION SENSOR
DRIVE SHAM
Figure 2-3. M151 BASIC DRIVE TRAIN DATA AND SENSOR LOCATIONS
30 IN
Table 2-1. SHAFT SPEEDS AS A FUNCTION OF MILES PER HOUR AND TRANSMISSION GEAR RANGE
1
Speed (miles/hr)
Wheel Speed
(re"/sec)
Drive Shalt
(rev/sec)
Transmission Input Speed or Kngine Speed (rev/sec)
1st (5.712)
2nd (3.179)
3rd (1.674)
4th
(1)
5 0.93 4.5 25.6
10 1.8b 9.0 40.5 28.6
20 3.72 18. 1 57.2 30.1
40 7.44 36.3 ol.O 36.3
65 UA 59.0 59.0
t I I I I I
Wheel speed and drive shaft speed are related to vehicle speed as follows:
Wheel Speed, Rev/Sec = 0.186 ~'-—- x (miles/hr) Mil-Sec
Drive shaft Speed, Rev/Sec = 0.904 -^"^ x (miles/ ir)
The easiest to Implement and the most promising vehicle operating conditions
for measurement of clutch slippage appears to be a road test in which the
2-17
.<-.-■ •.- .-- »' ■< I ,-M, ■
vehicle accelerates in fourth gear to approximately 35 miles per hour. Con-
ducting the test in fourth gear eliminates the requirement to insert elec-
tronic dividers to compensate for transmission fjear ratio. It also provides
a maximum load on the clutch at a reasonably high revolution rate.
The above- formulas were used to derive discrete frequencies generated by com-
ponents in the K151 driveline. Figure 2-4 is an example of gear mesh and
shaft rates for third gear. These rates are given in terms of drive shaft
rate. A complete list of sources and frequencies as a function of vehicle
speed in mph and transmission gears used is shovn in Table 2-2. Frequencies
generated by various bearings in the transmission, differential, and wheels
are shown in Table 2-}. In this later case, the bearings are identified by
tiieir part number. For each bearing, the frequency for inner race and outer
race, ball spin, and precision rate are given. Frequencies for speed values
from 5 to 45 miles per hour are given.
TO CLUTCH
i TRANSMISSION
1= 7
35* 20*
■11 J2i_r o>d
SHAFT SPEED OAR MESH RATIO
1 • wt 1 • 20 wd
2 • -066wd J) • 20 Ud
3 • wd 3 " 20 Wd
4 • wd 4 ■ 35u>d
5 • -0.897(jd 5 • 34.98u>d
6 • -0.0W7wd 6 ■ 42.20 wd
< • l.bHuifj ■ <"d / ■ 30.13 «d
DIFFERENTIAL
SHAFT SPED GEAR MESH RATIO
1 • u>c 1 • lijtf
A • 4.86wd A • 7wd
B • 4.86(i>d B ■ 7ud
1903
Figure 2-4. THIRD-GEAR DRIVELINE SHAFT AND GEAR MESH RATES
2-18
3
I I I I
i 1» I 1
I i
h i | i
i
UJ
_)
H 5 B2 H
M"
u w
H
3 a Q UJ H < E? u
u
CN I
CN
Cfl H
o -'1 ■n in O in V* in in m in O —4 r^i .-4
.n ^> rp —4 j> o ON o> ■> o o o 0> O-i ,—1 00 ^ *J «n ■ • • * • ■ | <t -J- o —i r^ o r-— O o O CT\ J. HI I i i 1 o r-4 D c r»
<f 00 —i t - 1 <t -t ~t r~4 1—4 r-4
00 lO 00 <f 00 J0
rN r-4
_ O m
in ,n O .n r"J in in in m O m c- -.t x: 00 r— ,—i o O ■"0 00 'XI o o o 00 r- in o in *J in • • • • • <t m r-l m ci —i ~i r-H ~-4 —1 t-- r- r-- I i i 1 r-4 c-n r-l r-l Q
m D .-<•> nj ro ^n m m m m m c ni rj
,—I
o r-4
o in
<~"> *jO <r rvl C-J X) ^n r-4 —1 r-4 C\| D O K CN -T -T in r-- "C O O in r^ O r^-. ,n O <f <f ^t CN h* 00 iTl r^~ <r O m 0 u in • • • • ■ ■ ■ • • • . ■
(•"> CN x> -O vt CN in r-l o >; ^r <r -j- m <r 00 -J* rj in co> O- o- m r-~ 30 CN r-4 r 4 nj ^n in in in J^
<f -1- <f r-x m
5- 4D r-J -1
i—i
t
Cr> JO -M O r—1 O c-l T> -T r^ o O -. X < T3 <f 3-1 10 D o rsj r*l <t o o o o O o m r-4 -t —' -1 K
l" o • • • • • • • ■ ■ • ■
M ~J O o 00 » ^4 00 i0O <t <f sf ^ r-~ 00 oo 00 40 N rx m m c
in r-4 ~4 r-4 m -o -JO c m
rn m m ^o m
rx m —4 m CN
O i/i o O O r-^ O r~ m CN o O O O O
-a CN m in -D O 43 JJ N o o o o O O o-. JO C4 CN c 00 c iTl • . > • • • • • • • • • • •
CM r-4 m X> 00 m CJN mom rj IN N K(J x> >—t l~- <*1 rx m in CN <r 00 H <-H CN <f
O m -4-
r-. r^. r^. r^. CN CN CN <f
t- CT-
1-^ r^ CJ O
~J 00 r\| O m O o <t CN o O 00 <f r-~ "0 0> 00 Cjn t—i O H CO 0^ O O O m m r^ in r-4 CN i.n r— 00 c O • ■ • • • • . • • • • .
CN r^ 00 r^ O o^ D O m oo CM N M 0O oo ty* O C3N oo m m r-4
CN un CN —< nj 00 00 00 l-l —4 <t CN ^-1 D m
—
r-4 rH r-4 m m -C in
. m
00 vD <r in O in .o oo m Ln m o m o O^ r- m u a> 01 X) in o 1/1 OA Ol O O O f 1 CN ■o .40 m —4 r-- oo a-- U) l/"> • • • • . • • • • • • • * > . • • • • •
r-4 m r-4 r-~ <!■
m <t Qy in r-4 i—4 r-l O-i 0^ oo m <r 0"! r-4 —1 o CN m X> c-J o> cji o> in
r-4 m —i
CN m
u r-N * ffi OJ 01
u M 3 0
j 00 r^ X
IH
-C
a; e
4-1 14-4
CO <T 43
X
01 E x: 4-1
01 0)
x: aj e u CO
w OJ 4J CO
x. •—i CN 1 1 I-- m in
oi r-i (N m <j- 4-1
m .o r^ c
en 4-1
0) 00
ci r-l ctj 0> CO c 0) .. 0 CO In oo n U r-4 -r4 M) r-4 r-l
•p4 4-1 ro CO CO 0 c X CO 4-1 c M 4-' ai 0) -0 "-) •r-J CO 4-1
CO oi 01 iw t>0 00 6 1 r>4 c u 1 15 CO •r4 3 i-J 01
XI CO C x: c d r-l c e •H T3 ■rn C/J 0 0 4.4 4-4 CO •r-i a > C r-l e —4 •H 14-4 U-4 ■r-l 0 -a
3 >> CO u) UJ CO CO w "~1 c MH u Ü I«
■-4 1/} rC C
04
1 3 14-4
a; <y 01 a B r4
C c a w w 0) OJ 0) r-4 r-4
•H ■-4 ■rJ e C > > IM OJ 0) 00 00 00 « CO ■-4 ■r-4 U-4 OJ 0> (3 c c U 1-4 M t-l ■r-l X! _d
Cd Eil w H H Q Q a "3! •"5
2-19
I Ä*'ÄÄ'iA*w»tes WJ. . . .'*>^ 4rJW> tk
1 —■ J T L I
r, — |
N H OS UJ a
h
c r t ^i
i i -
~* ! -.? Cr <f f -t ....
f 0 T
-j *n iA
■ t ~4
T. —> —4 XI
r '--I r O 1 --* y { m t i ~* T> O T^
i
i i *
'-I
i t r, ; a -* j
W
UJ
O*
-,? o -t o
fM d ,r' 0
O IN ' - m
D 00 1 C
— "
-j JO f** r-
'.T* x? O O
0 C ..T O
I I t I
<~i 'T —« X JO rsj j
CP> or- | —« ,n O i r-( fM ^ u*i N U 00 N
(N —^ nj
tu
Q W H
—ICO
JO ri o -'-IM "J -- x
ri C ' ^ r^ f^ "i
-n r-. r> I
u
I
10 H
—
r- ■-< .r: mo.«
r-s a- O O fi f~- f"i r-^
• m O 00
m rj ' -- c"*
O en o <f
r> —* r~ JJ
—« xj GO D
o ^r m —* o
O mo '-
CP n- in <.r\
l*> <M rt IN ; O m
O O O m I —< O O rj »-< m rs|
»_' o m -J ' —> r- —■* rjo Lf> •J -t X i —i \r\ O —i -<
•J o O t-» r-, '-j j no -orj -1- in : x N -ß I
<-J ' - O X . <V -T O m r-4 in m , OJ ' ^
J .-- - uo m m■ m ;
in cr ' t I
i - m r-j <" J ro <7'. ao rj
o —J yo m ■-I XJ O m 1 rM
r- O r-v -* 00 -t
CX) v.t r-l
r: d r: c
2-20
I
2.5.2 Vibration^JHic to !inpu 1 s i ve Forces
I.
Wien moving parts that come in contact with each other fail, they have n ten-
dency to develop flat spots or spall pieces from the contacting surface. When
in good condition, t lie parts transmit force from one part to the other in a
smooth transition. A galled part transfer« force with sharp steps or discon-
tinuities. These forces are shown in terms o! gear teeth in Figure 2-3. The
smooth force transition of good gears generates a relative low amplitude
frequency at the fundamental toolh rate. Harmonics of this tooth rate are
usually quite low. In the case of bad gears, the sharp transition in force
generates a high amplitude signal at the fundamental and has many harmonics of
high amplitude. In a classical sense, the spectrum lias a sine x over x
envelope.
[ 1 i i i I I
J^Xn
GOOD GEAR
BAD GEAR
TOOTH RATF
TIMt.
-—S
TIME
FREU'JFNCY
Figure 2-5. EFFECTS OF GOOD TEETH AND WORN TEETH ON VIBRATION SPECTRUM
i-^UiC'i^ i.vi'V-.i^".:J -;..-.-;-.i=-->-;
2-21
v,^.^;;^«:.*;-^^ ^■44««^*^^«^Sfi5^^K^^^^^^^8^1ÄiS
W
The important diagnostic feature is the ratio of i In- iniplitude of I he higher
harmonics Lo the fundamental. For the same sei .>1 ;•,. irs, the amplitude of
the fundamental will be sensitive to the speed and load. The raLio of the
amplitude of the harmonics to the amplitude of the fundamental can be expected
to be sensitive to the amount of wear.
From an analytical point of view, specific amplitudes ot the fundamental and
its harmonics could be calculated if the forcing wave shape is known. This,
however, is lmpractic.il for complex machinery, sino tue amount of wear and
discontinuity is generally unknown. As a qua]itat ivi concept, this under-
standing of hew vibration is generated is useful. i! Identifies potential
approaches to the design of diagnostic tools and can he used to explore diagnos-
tic features using empirical techniques.
A second characteristic of the impulsive forcing fund ions are that they
excite higher frequency modes of ringing in structural elements.
in summary, failures that cause repetitive impulsive forces excite the higher
frequencies and the rates of high frequency energy tu low frequency increases
as the part fails.
2.5.3 Cross-Coupled Vibration
Vibration measured at the surface of an assembly consists of components that
are due to mechanical disturbances internal to the assembly and due to dis-
turbances that are transmitted to the subject assembly from other assemblies.
These transmitted or coupled disturbances occur in the following ways:
a. Torque variations developed by other assemblies are transmitted
through coupling shafts to the subject assembly. These torque var-
iations set up mechanical disturbances which result in additional
vibration that take on the repetitive nature of the variations in
torque. These torques can also cause gears, bearings, and shafts of
the subject assembly to generate greater amplitudes of vibration at
their characteristic frequencies.
2-22
i
b. Another assembly t_ii.it is vibrating will transmit a pan of this
vibration to the subject assembly through shafts and vehicle frame.
These vibrations are typically characteristic of the vibrating assem-
bly modified by the vibration transfer characteristics of the coupling
structures and the response characteristics of the subject assembly.
Sharp force transitions coupled through frame elements can also cause
the subject assembly to vibrate or ring in its characteristic modes.
When the subject assembly induces vibration in another assembly, this
induced vibration can be reflected back or coupled back to the sub-
ject assembly. This reflected vibration will be modified by the
coupling transfer functions and response functions of both assemblies
and their connecting members.
Ihe direction of flow of coupled vibration is shown in Figure 2-6. It should
be noted that reciprocity of coupled vibration does not always apply. There-
tore, the coupling factors in one direction is different than the coupling
in the opposite direction. A generalized technique for separating or isolat-
ing the cross-coupled effects t
Figure 2-7. -om the aetual assembly vibration i
s shown in
REAR VIBRATION INDUCED IN THE TRANSMISSION BY WHEEL FACTORS
VIBRATION INDUCED IN THE WHEEL BY TRANSMISSION FACTORS
3300
Figure 2-6. TYPICAL CUUPLEI) VIP.RAT TIONS
2-2'J
■•w
<9hK-
11 i E
£8» 'Ah t
as;«
g him
V ii i! ill l!
II"
is ftj"
15
8sp
3
!fii S K28
< Q ^
1* 95
el
NOIIVMM cm*na>
if'
(VIA1»
e » »IHBY mom NOu»ugi/\ ajianoo
* »»I »
» I
, 1 - «a - "
UUil
► 8
II?
c v—,
H --- aj =G —* £> -r.
UJ .^ *—* -» »J — S3 v: V tr. L.; L_," tr. 'O c/i
V". ^ ^) hLi
w u; Ui u^ ^■s X t—
1—<
!<5 ~* u. -)
UJ H ^
1—1
w fc-l
N <" >-4 >—*
:i l™
W ^ «- -. U W o S
H ~-i
r^
CN
V i.1
3 M
.'
i
J
2-24 I I
If t lit- vibration generated by assembly B is V(B) and the coupling factor
from assembly B to assembly A is KJ , then the vibration induced in assembly
A by assembly B is Kl V(B). Where VB is made up of a number of complex
vibration and mechanical stresses, Kl is the respective coupling or transfer
function factor from assembly B to assembly A and includes the response
functions of assembly A.
11 assembly A vibrates with a vibration V(A) that is due to internal sources,
then the total vibration of assembly A is V(A) + K VU. Similarly, assembly B
will have a V(B) + K VA vibration where K is the coupling factor from
assembly A to assembly B.
These expressions can be taken as the vibration detected by sensors mounted
on the assemblies and to include the transfer functions of the sensor. As
expressed in this manner, thr equation V(A) + K VB represents a voltage
output of an accelerometer mounted on assembly A.
If the accelerometer output is applied to a feature extractor that has a
transfer function F, then the output of the feature extractor is a set of
values representing long time averages of extracted features. These fea-
tures may be numbers or analog voltage representing the magnitude of a
specific frequency, average peak values over a specific time period, the
power within a specific band of frequencies, etc. In addition, an essential
quality of the feature extractor is that its averaging time be much longer
than the transport delay between assemblies and must be much longer than
the period of the lowest vibration frequency of concern. The output of the
feature extractor is therefore a set of stationary values represented by
the equation
FA x V(A) + FA x K. V(fl) A A i
2-25
,-*-.:J,.vH---:^;^--^-v.--. ■ .z*i ^i->i,.W.''>&- &£&i&9&ffii&&ii': ■-'.- ■
where F is the transfer function of the A assembly feature extractor,
Assembly B has a feature extractor output:
lß VfB) + FB x K„ V(A)
The B assembly feature may he multiplied DV a constant ''., and subtracted from
assembly A features as shown in Figure 2-7. Tills results in an output
F, x V(A) x (1-K.K,) which ccntaiis onlv the assembly A 1 eat lire and is Inde- A 12
pendent of vibrations generated by assembly H. Threshold or comparison teeh-
niques may then be applied to these features to determine whether or not the
assembly is faulty.
As derived here, the approach is qualitative in natuti, .A.i.i l< t ical derivation
of numbers for the coupling factors and constants i> impractical for complex
mechanical systems but can be evaluated experiment.! 11>. The derivation does
help in the understanding of the interactive eiiects o! one assembly on another
and to indicate processing techniques that can be used Lo separate the effects
of cross-coupling. The understanding of this process is useful in the
interpretation of power spectrum plots of vibration dat-a from the various--,
assemblies. Analysis of vibration data from tin assemblies operated under a
variety of conditions can be used to empirical!', derive values for the
coupling factor and constants.
2.6 EMPIRICAL TECHNIQUES
Empirical techniques involve the gathering, analysis, interpretation, and
evaluation of vibration data from the various assemblies while the vehicle is
I ',v.
ft r
2-26
operated. The vibrations of the vehicle assembly may be classified as being
made up of the following characteristics and related sources:
a. Those vibrations which are coherent with specific time rates of change
of components within the assemblies. These functions may be corre-
lated to functions of output shafts and their frequency value will
varv with the machine speed.
h. Ringing frequencies of frames and components within the assembly.
These frequencies are relatively constant with respect to shaft
speeds.
c. Rattles that are usually due to loose or cracked parts. These usually
have low frequency repetitive characteristics that occur at random
intervals.
d. Random vibrations that are due to road roughness, etc.
A second approach uses vibrations induced by an external source, such as a
vibration machine or impulsive driver, attached to an assembly. Northrop used
an electrically operated hammer to develop an impulsive excitation applied to
various assemblies.
2.6.1 Vibration Recording and Analysis
The equipment used to record and process vibration data is illustrated in
Figure 2-8. The M151 vehicle was equipped with accelerometers mounted on the
engine-transmission housing, rear differential, and front differential. Rota-
tion sensors were mounted to detect rear drive shaft rotation, engine rotation
and to detect each of the four wheel rotations. Signal conditioning ampli-
fiers and filters were ustd to condition the accelerometer signals for mag-
netic tape recording. The output of t ,e rotation sensjrs was processed to
sharpen the pulse shape and to condition for magnetic tape recording.
2-27
S
o^^:rU
"1 I r
VIBRATION ROTATION
COHTRINTAWRAC.I -
PtOTS
POWER SPECTRAl _ DENSITY PIOTS
NOP ?7'j-lbS
COM PURR
i
DIGITI/fD DATA j, ,__jC)
I ON MAC 1APE
SIGNAL CONDITIONER
VIBRATION ROTATION OUT OUT
! SYNCHRONOUS DIGITIZING
RATE I . J
ROTATION
• o
NCI STA'.D-
AlONE DIGITIZER
Q (k j) — /
RECORDER
<~>
Vi if
RECORDER
O VIBRATION
/ RECORDED DATA
Figure 2-8. DATA PROCESS
A portable magnet ic tape instrument recorder was used in record the accelerorr.-
eter and rotation signals. Data was recorded while operating the vehicle under
the following conditions.
a. At various speeds and gears while jacked up and suspended op spring
blocks. Each axle was supported on a spring that would permit the
wheel to vibrate.
b. While operating the vehicle on a dynamomet. .
c. While operating the vehicle on the ro.id at various speeds .nd in
various gears.
These data were analyzed ii. the following ways.
a. In real time observing the general nature and amplitude of the vibra-
tion signal.
b. Preliminary spectral analysis using a Tektronix spectrum analyzer.
2-28
i
I 1 I I I I 1 I I I I
c. Computer processing to develop coherent average and synchronous power
spectral density plots.
The process used for the computer analysis is illustrated in Figure 2-8. The
recorded data was played back on the magnetic tape recorders. Analog data
was filtered to prevent aliasing and amplified for input to Northrop Rssearch
& Technology ("enter digitizing systems. Digitization rate was selected and
controlled by a synchronous digitizing unit. The digitizing s.-.iit used recorded
shaft rotatior pulses to generate N digitizing pulses per sha '' revolution.
These digitizing pulses were phase locked to the shaft rate. Tu« digitized
analog values were therefore coherent with various mechanical functions in the
assemblies. This digitized data was recorded on computer compatible magnetic
tape.
These magnetic tapes were then procef -ed on Northrop 370-165 IBM computer to
develop coherent average plots and power spectral density plots.
2.6.2 Impulse Response Measurements
The impulse response technique involves inducing an impulsive stress in an
assembly, measuring the resulting vibration of that assembly and measuring the
vibration induced at other driveline assemblies through connecting shafts and
structure. This technique is illustrated in Figure 2-9.
An electrically operated hammer that delivers a controlled repetitive impulse
was used to excite the transmission assembly. Accelerorneters placed on the
transmission and on the differential were used to pick up the resulting vibra-
tions. Thi_ ringing frequency f1 and amplitude A. at the transmission were
observed. Similarly, the coupled frequency f„ and amplitude A„ induced in
the differential were observed. In this case, a frequency of 6 kHz and ampli-
tude of .10g peak was induced in the transmission and resulted in a frequency
of 1.5 kHz and amplitude of ,05g in the differential. The ratio of amplitudes
is —— = 7.00 or 46 dB. Other frequencies and amplitudes can be derived by
analysis of the two waveforms. If the case is reversed, a 2.5 kHz signal is
induced in the differential and a 5 khz signal is coupled to the transmissions
that is attenuated by 25 dB.
2-29
-■■■■■.;, ".-■ .i:**;^'--'
IMPULS! Drivw
tj — 6 kHz
M LOG
COUPLING fACTOR
M 10
OR IVI SHAH
JU--
DIWSENTIAl
ACCfLLROMfTTRi
FORCING fUNCTION
A2 0 05 200-46dB
I 1
Figure 2-9. IMPULSE RESPONSE MEASUREMENT
As seen in Figure 2-10, the coupling trom transmission Lo rear differential was
depicted by an error from transmission to differential indicating a 6 kHz sig-
nal in the transmission and a 1.5 kHz signal in the differential with 46 dB
attenuator.
Nf ■ FREQUENCY NOT DLFINED
REAR FRONT
Figure 2-10. M151A2 DRIVELINE IMPULSE RESPONSES
2-30
1903
I i
I I I
I I I I I !
I 1 1
The fundamental coupling factors between major assemblies Is illustrated in
Figure 2-10.
2.6. ) Analysis of Recorded Vibrat ion
Vibration signals from accelerometers were recorded on magnetic tape and then
played back for analysis. One technique used by Northrop was spectrum anal-
ysis. The effects of using different spectral line bandwidth in the analysis
is illustrated in Figure 2-11. Note that octave band and 1/3 octave band
analyses shows only the general energy content over the spectrum. The narrow
band 4 percent filter analysis shows more detail at the lower frequencies, but
is similar to Liie wider band responses.
1 500 1000
RKQUCNCYOt)
I 1 I I I
Figure 2-11. TYPICAL ANALYSIS OF VIBRATION
The 2 Hz bandwidth filter shows that the vibration is made up of a large
number of discrete frequencies. These discrete frequency lines were identified
as belonging to specific mechanical sources as illustrated in Figure 2-12. A
smaller bandwidth analysis with higher signal-to-noise ratio would reveal
additional detail.
2-31
1 ■■■*
4
.»
0 1 10
DRIVE SHAFT ROTATION
40
Figure 2-12. SYNCHRONOUS ANALYSIS 01' VIBRATION
High resolution frequency analysis required sampling periods of one or two
seconds to provide frequency resolutions uf 1. Lo 1/2 Hz. To provide high
signal-to-noise ratios that reject road and other random vibtations, a number
of spectrums were averaged to obtain signatures of the mechanical parts. An
impractical requirement of holding the machine speed constant within a few
percent is imposed if the high resolution frequence analysis is done from a
constant time base. One alternative is » ■> analyze the vibration data refer-
enced and synchronized to a shaft rotation. J The spectral lines are now expressed i.i terms of shaft rotation. To accom-
plish thi?, the driveline vibratior data was digitized in synchronism at
various multiples of the drive r'naft or axle speed. This digitized data
was then subjected to a power spectral density (PSD) analysis and plotted as
illustrated in Figure 2-13.
The abscissa of these plots are those functions of N times the shaft rate
where N may vary from a number less than 1 to a number greater than 1.
Typically, a plot contains 256 spectral lines, each one of which is called a bin.
2-32
7?
i
I I I I I
J
!
> i i r r > r i * i i * 7 1 ft J T » r s i » i x 7 I ff S J 7 * J r T f » » * T T r i T T j J * 7 7 I 7 7 9 7 T 7
i 7 -9 i a 7 a 7 » » » > 1 ? ft J 7 ft 3 3» j :* a * > ft J » » > J * > J ft f * ft :• ft ft ft ? * ft 3 > 3 > ft » a a > * a ft ft ft ft ft 1 1 » » 1 > » X ft ft y r * i ft. > ) * > ft « > i ft a ft > x > » - r i s Til 7 » i ft ft * I » ? 7 1 ft ft ? ft
*• f 1 J ft a
> r > ft » ft » > » > ft a i > > > ft > ft > » ft » i J » > ft ft > * > » » > » a > ft ft
> ft > J > ft l > ft i > ft » a i a > ft ft ■> > a i > a ft a > z ft a a a ft ft ft a > » ft i J a » i ft a ft ft » > j > ? a > » > a » ft > a 7 a a ? a. a ft a j a » » ft > * s : a. > a. a 2 a » a ft a ft ? * j r T 7 ft ? r
r : > j »
i .■« * :* ft
•»\j^«^r--*r^OJ,(y'^'NO'**^CrO',j'\I,j'*»*-^'Sjw^'*-»-jfk-< 'N N 4 ,/ « CO' j j^ *^ r* a
' (V (" ■# If1 4 >• (T ^ O ■ ry rf\ .f ir O I ■ »*» Id <** *•■> (•*■ m r ■*# .f .ft ^ .ft ,» ^4'^^f^/^^^'J^|J^^^l^'l^^^^rl^•O■O•^«0'^•C'i,•
r
A*
s»»»>iiif»i> » i i J i -I»>>J>JII>J— -.»»— ** *« r 1 s. x * .>>>**>>>> **i#a»>». <»>1»»>J>» >* j > J * » > ► *»t*a»»»»l**»*»)*j»>»»»**»»T » * » j J * * J a i»»»2»*i*ikxii»?>»>:i»>>)»»>}-r i * » i z » i z x t
i»ixij»ai^»»»»i>i»»>*»»>*»>»»»»»*,»»l,i****
» * » a » > »t ii * * * * * y » » » ? I »I>J»>»*:»:»»>:»I»**>»*
t»>>!i9>»*>>>ri>»a?>>>>>i>f>>*>*>'*a,*v:l'
»ii**»>iii*j»i>»»*»;.»»»»*»»».>»J:*»a,a-,-lllx
»»!»>! 1 1 > 1 1 I » I 1 1 I ) I I )»>)>>'»»'» ''', '^ * "
*5»»f**i*»j»»i#iti**f?f»»-> *■''»>*»*?****''''* * 5 » • »s*«***i»i •»>>**>*»>* » j »» } J» !> 1 M M > » I ) ) t 1 I I I I ! I t I S t • I ! -JJTJi-r-JJJJJJ?: > X » * » 3 » » i J i » ' i > i*» f I t I • f r > i 7 ) i ? J n M M > ) M I i r i n ,,v,.,;';;,;;,,;;«;:i?«>«?*>*39*xxz»?izxz i*^^*i*i»»«l*»»-»l3*sSJ*i»x:,»XXJ»:.'Xl.f'»¥l»•,» i»ii**'.J'iS»;»*-J**-«-***>*y>*',?*»Sl'*X-r**
« ) 1 > i > i 5 > « ! i ' I j i ' i > ' ) ■ > » J ' 5 J ' > ) M r J > J 1 J > f
i —• 'J " *J"^Na fl <*v -f ' * #■ » ff o -* * »or *~* c* — -* -* -
■ ** • ^ <r c
Figure 2-13. TYPICAL COMPUTER PRINTOUT
2-33/2-34
.i«*<SS*; ;..-,,_-■ ;.,.'..:
,V^ iffiÄ «si :■,;;■,'-;;; l-®&&i
The exact shaft rate frequency may be (for example) In bin 16 (N*16). Bin
4 would therefore contain the magnitude of vibration that occurs at one fourth
of the shaft rate and the 80th bin would contain the magnitude of vibration
that is 5 times the shaft rate. In the computer printouts the abscissa con-
tains two sets of numbers as shown in Figure 2-1). The lower set (first
column at the bottom of the page), ranging from 1 to 256, is the bin number
and corresponds to N+l. Therefore, bin 17, in this case, corresponds to the
shaft rate and bin T2 corresponds to two times the shaft rate. The upper set
of numbers (second column at the bottom of the page) aJong the abscissa repre-
sents the relative vibration amplitude in dB.
An explanation of this type of power spectrum density plot is as follows:
T ■ Period of the sample
N * Number of clocks per sample period; 128, 256, 512 or 1024 C = Clock Rate = NT
f = Maximum frequency resolved max
= £ _ I _ CT-2 2 T " 2T (f-')l
min 1 f
K = N The bin number with values from 1 to —
Bin separation frequency = ~ T
•k - Frequency of the Kth bin = (K-l) I T
Note that bin 1 K=l TU„ r "in x K=I. The frequency is al ways equal to zero.
2-35
.'.'.;;;-;:.'..,.,.■:.,'-.-■,''■ , . ..- » i -. - ' * * « • *• s " *< -»- "«-'■■*-'■'
EXAMPLE
N is picked to be S12 clocks per period
T = 0.1 second (600 rpm)
f I . 10 Hz (appears in Bin 2) nun
max . (All , \ J_ = 25S0 Hz
2 f 0.1
= i = 10 Hz T
K = 256
.'he plot would appear as shown in rig Kifiure 2-14.
i x i 2 S 8
a -m-
XXX
8 % I
J U 1
fi s s
Figure 2-14. INTERPRETATION OF PSD AMPLITUDE AND FREQUENCIES
In Terms of Shaft Rotation
1 T
1 DT
where T is the shaft period and —— s T s
— = shaft frequency, f
2-36
max
min
= Number of docks per sample period (T)
I-JL -I, T DT I) o
s
1 . I f DT D o
s
!'he bin numbers can now be expressed In terms of f
fk - (K-l) i - (K-l) I fo
Note that this gives the ability to plot frequencies below the shaft frequency
and gives a higher resolution plot.
An example is shown in Figure 2-15 and is interpreted as follows:
Assume f = 600 rpm or 10 Hz o r
D - 8
N = 512
max
8 (f0> 31.87 3 f
min
= 318.7 Hz
fo m 10 8 ~ 8
1.25 Hz
~ f - 1.25 Hz D o
2-37
*wi'3*;s.M-.-:!'-«.*-.i'.. : -i .:«i->.-w;-fa.i*s .*i^iAAh«Jii^wu*tffe«/!vVJ:-'*"S-&-:>
a
K < 3s
H n n ir-
u. < 5
-4f- 2 £
K t < s £
J L
Figure 2-15. RELATIONSHIP OF PLOT TO SiiAF! ROTATION
2.7 COMPARISON OF ANALYTICAL AND EMPIRICAL RESULTS
Analytical analysis indicated that there would be a large number of discrete
frequencies that could be attributed to specific components within an assembly.
The majority of these frequencies were below TOO hertz. A typical PSD com-
puter plot from a differential is shown in Figure 2-16. In this case, the
M151A2 vehicle was operated in third gear at approximately 23 miles per hour.
By using analytical techniques, the various lines can be identified as belong-
ing to specific mechanical components.
Sources of lines are identified on the plot in Figure 2-16. Many of the lines
show spectral lines that are due to other assemblies such as wheels, wheel-U-
joints, drive shaft, etc. The effects of the drive -haft U-joints on the
differential gear mesh frequency is identifiable as the gear mesh frequency
minus and plus the drive shaft and U-joint frequency.
The plot illustrates the highly complex nature of the vibration and how i£ is
mac'e un of frequency components attributable to components within the subject
assembly, frequencies from other assemblies, and induced frequencies that are
frame-coupled due to torque variations.
A high-frequency plot from a differential is illustrated in Figure 2-17. Be-
cause of the broad band of frequencies covered, the frequency resolution is
2-38
\
5 o
3? u_ u. 5 $ «/> 1/1
3 t 3
r
IT
I
3
S3
8
5
Q
e
5
s
a t
2 v/l in
2
or
5 o z
Of
o
400 HZ
,#J--O*-. f '« 3 • — , -l.«.#-%-j-i »*-•«. / N »••- .( -ü — «• 0 ^ • ) ^ N i» # ( O^^AAjl^A/f^' . -* D w * '5 -• -
#^>««»»»0""',fc*
-TMT
J 2
100 HZ
FRAM£ VIBRATION 0OIS NOT CHANGE WITH VtHICU SPUO
taoHZ
I 9 l % 1
> * i >
• f • I » t
T iJ - - U ? • • ou«: i w -• D^^ -• -i -v ■« # .# * » c *•-*»- r> a o a .
*i » 1 — '„ — 4 iT e ? * > y ;
.i.„. „ _.J_Jt-.
3
DirrtRCNTIAL RLAR 3RD GEAR 25MPH
ACTUAlFREQUENCY IS APPROXIMATELY 4 (BIN NUMBER - 11
4POIE IOVY POS FILTER SET TO 800 HZ
KEY TO VlBRATIONLINE IDENTIFICATION
EXAMPIE . . . . 2!ÜIFF.GEAR.MESHt.-<CYU,\DER.RATE>
MEANS TWO TIMES THE DIHfRENTIAl GEAR MESH RATE MINUS THE ENGINE CYLINDER FIRING RATE. SEE BIN NO. 61
\
S t -r- ~", r> *• £> j-. t~ a * >*•■ j*. i |««i< • 4 * * *
«N^MNNNNf
, ■•» Jt ^ C — BOO — «v ••* * W «•-«-<* O — 'S. ■« .* r«»«.-p»rt-t-n^*^rfi~ «»^-a-w-M«*»«
Figure 2-16. PSD OF REAR DIFFERENTIAL
2-39/2-40
■i,'.'.-..».-. ~~~" -*T
I I I I
1. 1 1
or
Z g 3
5 i
13
r
or
5 o
«CO HZ 2
3
.. _ _
I.
■"■■r «-*»■»»»■
s FRAM£ VIBRATU'N
1000 HZ 6» HZ
:*.
y \ x v i » » v : » t r •
> ■ » a i. - x. > * ' X >
-—>*:* — O-i^f ■ü c £ x. T
■»^■»iii» ■»■■■, M,..*^.
""" -~^—~> — sir - ~—
r ip.
1600 HZ
DIFFERENTIAL REAR 3RD GEAR 20MPH ACTUAL FREQUENCY IS APPROXIMATELY 8 (BIN NUMBER-II
4 POLE LOW POS FILTER i a TO 2000 HZ
t • t I 1 1 1
*»>*»*>» * a i->-i-'--»»ria*
M ) i • I i > M a, »a>k**t *^.a»!ta»:»»>»aa
• *'^*»i:«»t-»Ta>i»«XX J a * - : * a » a >i-:->a>?»*r?a r -•• i * y i i » -. > i » T i j i j > 7 r i » j * i ) U ' r l > * i i t >M t : > - ^ . i M ) i L i t — . --I:..,. .- , . , . j . , : .> a » $* x f
: *»-s»a**-«»a :-»„»»>■. • a « T a » r r * r • * • - : ■> * : ■ * s' j • a i » j a !> X i »
-»>.•■.■.-,. a «» «r;rrr'«Trr/ictffi
• * - - i ". ' ► T at I■ K - -»; IT»:*.* xiKvrxxa ! ■ J i ■ * ' ■ ■• i : > v » - * ' i. » *. v * « a >- 1 * r »- r r * ■ » •-■>**■». -• i 4 >--*■•>;>■ i *'*..-. t r. y x a *■ r j- i r r !»■--<:. *. *a-*y»i »*»»*- *tr»T»*'">»'»>rarrT_-,.J.-.H«__*^
* s • ••» - * a . ' r > J • n n »i » i < n H » U :-'^-*'* J», i > v i r : ü r i it ^ I r > n n t | ( * t 1 1
- * - *»>'•»/"*■«. *»*;>*» >>..x» a. ai«».ir:.aa.J»«'rrx II ;*>»*-**i.»*> *»■ •►»■-^■'Ji»»a»yii»r»»ar»<:a,i»'cx
»■> »•--»•jar»* -»• ': f r»i < TT* wrMrici »i »u. t f r i r t , > « . > r •*>>**. * ay i*-.-..*a-*a>a»a:">aa*»aa»'Txxa-xr»xx
?-.. *'» »>■*»*#.-. ' t k > i > > > >. M i , > i M ' r ; r ic > r i i t t i i . . >« »-. . «>..••')»» j >>,»>)>»> i :»>» i »»»» i >»•>> i > i i i j «'*»•»> a ai~**"*:xx*a>»»"\-*J«ir»'*ra«»1 if ■/»«»?»*"» a- *jrj JI»J. >. 1 ' i J •>»>»»'«»» > a J a a > >. >>>ajt>>}, aA.a.aa.a>. a » * »»»»■'•«. »»»■*' ' > : * ■ * . a a > * . .4,>>i" k > > 1 ) k It I M t 4| ^ | • » > * ->•*>. •/>■#_>- i < ■ i i i ,* ; t ) ^ i i » i i r ;: > i t >■ t ^ A i t. i
> a j - - a I ■ ••: < ^ • ' ;. ' . f > . (. > i • i r * « i > £ ' < ). k t i < t > ^ ) .■?,■»■•*•:'.: *.».*i,i+.*>*ii4.:if>*»i.*i«.:?j.x*;«rji.x a >*•■-_,* ,■ *. * > - •«*•»- a i -* a -•>>•»>, a*j»--rr*ij>r».t» , » j » . . a » .» ■ : »*. . : M > U > > > » > > i > 1 » > i 1 i, M i m >. »i-JT».
. • .?•-'.. - i - : J . : :■ *. r a « > > > n < ) ;-■ >>a, -.«-. aa»j-ji.T» • a . r » • t »*»t> * : /j»y>>»,i-*i*7>aaiTiar»ji-:>-fa)'aaJa:a.a • i * ,« ->»: - • * 7 * . - ». > >/**--'.-. • . ; J i * i ■ J < T ' J ' 1 T » f * » T • »*?■- ».' ^ * r » ) u ; :• J » u > i .* i j j M > i ; ,* i m i i ;, t s. n »*»-i>. :;' '»v»: >>/■ ,>>rr>.-:*'--i>;li-^JXj: t x. a » » a
- : i . * »- i .. 4 t ■ ' *' •■» * i - ■ . : a = > ? > * : > a. r > ■<. J. * *. a a .. a i. a: ^ a ., » , a * a .■•> ■- «*. a*.*i-* * ? i » * * a* L. a a • a •-* a. i I :,. a a x *. a a. a a
X ' » T I
a a * x a. a * a X > » X -*■ a » * * r > a. i ;-. > a. X l. - A J- *■ .■ ■ >. » X * r * » * a a a » fill a a r a X a a. x X '-'■ X X 1
a a
a a
i a >. a
> > ' J a a
a. r ? a *. ■> x » > » a » v a a r > -. x T a > »f a a
v. a a a ; T a
I x x x X X
* > it a x x
J r o o - - 'ir- JJ--< OO' J » » ^ o o — a
Figure 2-17. HIGH-FREQUENCY PSD PLOT
2-41/2-<'2
-A-^..
low. All of the data from the previous Figure 2-16 is grouped in the first
15 or 20 bins of this plot. This plot illustrates the effects of frame ringing
and high frequency elements. Those areas that are due to frame ringing do not
change frequency as the speed of the vehicle varies. The amplitude varies con-
siderably as a function of speed, load, and the goodness or badness of the
part. These bands of frequencies followed Hie characteristics qualitatively predicted using analytical techniques.
By using botli analytical and empirical analyses, it was possible to identify
specific sources of vibration and to measure their amplitudes. Frequency com-
ponents that were not predicted analytically were observed empirically and
were then traced to their source analytically o.- shown to be an error in instru-
mentation. A summary of these features is shown in Table 2-4.
As a part of this comparison analysis, Northrop concluded that discrete fre-
quency techniques were useful to identify specific parts. Not only could
wheels, drive shafts, U-joints, etc., be identified, but gears, bearing races and balls could be identified.
For this program, the discrete frequency technique was useful for identifying
the wheel, wheel U-joint, drive shaft and drive shaft U-jolnt problems, but
was not practical for identifying transmission and differential assemblies.
Each of these assemblies generated between 50 and several hundred frequencies
which would require individual measurement and comparison to specific limits.
The memory and time required to do these measurements and comparisons was not compatible with a low-cost portable system.
The ratio of frequencies in selected low and high bands did, however, prove
to be good indicators that something was wrong in an assembly as predicted analytically and identified empirically.
2'8 SELECTI™ 0F TEST TECHNIQ„ES
Any drlveli„e a„al„er t av
««»2-M. rhe syste„ mus;y; ~ ~ * tmam ulustMted ln
or texture analysis n-,- 7 1S' Data acquisition
2-43
CO US
D H
o
i CM
CO H
* 4J
u ~-4 _l ■*w
3 4» d TO XT 1> .-* x: —t x: 0 m c 3 •—1
SI 1 <b
(J C > ftl ■"*
H irt
B O
Li
TJ
C *J u 0 x; w ** i0 *-* u M V) TO 0< ■*>
tJ —4 :* x: 3 "O Qfl
0 V 0' -* u 01 0*
i/j
4j
d to 3
u a« d x: 01 > (j
3 w u TO TO TO G <■» *J
U, a.
.-
V tx.
Oi *j <
—- -. „.4.—
i —
^ 0 x: OJ 1 x: TO u
d i ./! u d w u a —( 1 —■*
TO c OJ ■.1 01 JJ TO z n u 1* a TD U 0 1 —< c -J 1M -- '
L< O 01 bfl * TO tJ Q. •' U9 ;*. OJ u i„ '.J M --. V 'J c X d . tJ u <*-* TO a.i*- TO n r. 0( tn d a-
U c *** U 01 ■^ O 3 ,* u 01 TO i-
01 0) —. <c *-J d -0 -J) lA U d a; 3 cr
2
<ii G O 11 <+-i 3 -a TO TO -O U-d '»-. \*-< cr c u u d c d 0 '*- u ■It *-
UJ a* c 0 0 0 TO ■ -< TO *_i -~*
L. 0 ■~4 r~4 •> ■ ^ X) in TJ 0) TO T? no U-4 iti V >, !/) l/I '1* U 'J, L,
c w Ifl V O T3 Ifl "0 ■^-i C d ■H .-< T3 ■mi JC d C r- ^ C U-i 0 c *j ~j ~* OJ C 0 J 0- TO e L/l i ^ '<-. a v TO in 3 -C ul —( Jl M TO w 3 x: x> d + i tr OJ d Li 0* ■0 (fl JC tJ
0 3 4) TO OJ T3 TO TO Oi r w « d
u TJ M ■XI U -H b ül x: X. TO '-J V« TO 1 ■*^ —i 4J 01 ■*J to *j 0/ 3 X3 E 01 ^ n m 0 ifl ■-4 -r-4 U in >_< TO 0 m •5 u -< 01 t) U XI 01 XI C U TJ 0 >> 01 d e 0) t-4 C d Ü .-( TO d u —* u TO OJ ai cu TO 1» -0 —* ■A u Of .-* V)
u ~* 3 n -d •-* 3 o> 3 »J 3 p-j
TO -a -^ v 3 " "( cr u u u cr TO OJ 3 0 TO 0» -" TO GJ TO u V ai 3 U - m u M w 3 U ^^ 3 c IM w 3
ZD a u >*j < a *J i*«i cx < T3 TO '*-. LJ T3
'■
0 ifl 11 TO c —4 » .^ Oi u TO V -~j —i K l/l X! Xl u
t/1 V 01 0 *j iV i-j 3 - •»J Lw
'_J X 0) c TO to TO >l jt H tfl '■"t 3 JC u TO x: ifl ^-J 01 X! u 3 OJ U i-j OJ r-< f-d B m ifl u "0 ^-i R T TO ■V WJ
•>-i Ifl *J 3 •J QJ c u -0 u 0 u TO t-> OJ 0 0 *J ■-* '.1 k-1 •H TO a; >1
■^ X! OJ ■■-* a 0 X c *J u i*. ^^ 3 ^ i-j r^J >i ifl 01 § u c U B- i-j u H 3 O TO a> d 0 w *-» X TO ^M d in L. d U) 3 Oi TO •*** TO OJ U) 0 r-4 .^ TO 4) TO cr 3 1) a y • in x; 3
x: 01 N cr 3 ^ >\ d TO --j >1 M tJ cr N
u w X Oi 0 *H O M d d u X Cfi r-4 u 0' ^3 4-1 u. r-J c ai
§ c 01 71 ■V ki
c o ■*-< TO OJ 01 O U TO .c ■*- in
o » r~* >• 3 • at u 3 OJ 3 '*- -£ <f ■r4 —^ r—« Ld in tr N u TO 0- TJ ■FH
w TO 0 TO OJ TO (L ac 3 J3 01 9 ^ M JZ TO 0 (0 -a «J T3 u* M •0 u U. t~< TJ *■' M -t •<-< c 4H O 0 c •*- O, V ■•*
X) o -> 0 u <—t u ■H r-l u 0 0^ --< (fl ^j . -u u P. (j I~J 3 UJ > 3 01 3 n y •^ 0 n •^ XI r-H d XI 3 Oj
C M C -^ d in i.j d u VI ..J 3 0 C w -J (5 ■*H TO 0 TO TO ■^J TO TO TO ■ ^ L« — TO :/l u 'T\ u O DO rM 'J i' 0Ü u *> Q. (/) L.
u
i*-f 0 0 0
i_* 01 <V fJJ (V -J u u 'J
k. 3 c c Ml OJ c 3 TO d> TO d >M TO *-t U< 3 A3 TO *J TO ii u ^-j TO W
55 X) c
äs XI D
d
01
TO ■*4
TO X ^
M w «■} |" N -^
*J
V >, '!-* c -< tn TO
■-< X) J= ^. B c t/1 oj a* .-( > <n t—« 0 0)
—» w «i 0» V •*~i 5» u < 0) -—t 1 a x; X X ru Li
:* < < a
% li
m ■-* « w
C "J « '*-•
3 X a- 01
0 d L. v ex u
H ^ 11) y 11 •H -•* OJ X 3 3 TJ V! >XI rT
*-• O 0 TO 0
I -, J L. ^■^ TO
■J) '> J TO d
3 OJ — 1 e m H U 3 k. 0
■) TO U <H OJ
OJ TJ - XI m ■>
' 0 aD M ■M
-f-
c -a
^C OJ 4M w Oi •*A
d M 0* 0 '41 -O L.
-w 3 ao . -1 r: ■^4
L> r-* TO in TJ ■*4 4_(
■-j TO e 3 -Hi in V) J u rr U 30
M 1 Oi X tr
tq cr u 0) 'H
0) d ii '4-4 Ul
u 0 L. >. 3
3 ^* (1
73 « >i-* d c 0 ji 3 TO <nH
Li r* 0 Cl 2 n a. 0 .-*. yi
■-•
TO d
41 01 H TO ■V
0 ■1 *.J ai
c rj OJ
cr 30 "0 w « c o u
g
a. *« o «1 8
m Fi
u 'V cfl 0 n 0 u H <u w .« N V m n tfl X •o .J X r: T) -^ V ti 31) r. M 0 n -1 0 O ■-( -> C 1 ■^
u. hJ 00 £. u. 01 -d w
2-44
I 1
I I I
c
H
H
OS
$
'X.
-1 i
.a 10
-J ■4) 1 .a r. LJ
H -■• -1
V) us a.
0 c
tj —i (ft 1 ,-w T3 » !/) 3 <fl x: u
u. 7) Ul (■3
■1)
2: B 12 V!
to <fl c tu ift c 0) V !1J U 11 N C B ■-' 90 u 1 -1 i.J ■** « V u C ft 1« rr U X -' ■~-4 X to —* M V 0) 3
10 c
C 3T 5 « K C U •*-
M O -< -J 0 0 -n
.? ^ ■*■» <y
ao
J3
a ^ C.l. n *-» h —4
td tf) -j 1/. m ;« OJ V) ■■^1 V u 0
■a 'j u u !fl R u u V) li fl c c > HJ ■J ■J >-» 1' 1) —4
U 1 U -1 -1 —4
u rr IF c c,r rr (fl u V b a» 41 u 3
B e c u-
V <V *-t »H <ft
^, '*' -< -C c u u -* n « 0 c
3 -o
O <D C «
<u a> C
n N u. 3 --<
c o
o «-* a- « ■■*- B v
v (N 0 ^ x:
ti c ~* c: tg
cum ?U .J L-
o
«1 ;u C t) 4i *-' 0 r -eg TJ
m C u nj c:
w *. B - E
2-45
"UK»
DATA ACQUISITION AND FEATURE EXTR ACTION
MATURE ANAtYSIS
VIBRATION SENSOR
ROTATION
SENSOR
TEA TÜRE EXTRACTION
>!♦
VEHICLE OPERATIONAL CONDITION
ANALYZE* AND DISPLAY
MATURE ANALYSIS
EXTRACT ION Of CONSTANTS COUPLING
»ACTORS AND NORMALIZING
IIMITS ANALYSIS AND ORDERING
Of FAULTS
FAULT
DISPLAY
Figure 2-18. DRIVEUNE ANALYZER FEATURE EXTRACTION AND ANALYSIS PROCESSES
involves the instrumentation needed to convert mechanical responses such as
vibration and shaft rotation into electrical signals. Feature extraction in-
volves the processing of these signals in any one of a variety of ways to
identify and measure specific characteristics. Typical feature extraction
techniques include spectrum analysis, peak detection bandpass analysis, cor-
relation techniques, etc. In any case, the output of the feature extractor is
a series of voltages or numbers representing the magnitude of specific features.
Typically, this may be the amplitude of a specific frequency or the energy
within a specific band of frequencies. Others may be the rotation speed of a
shaft or the phase relationship of two shafts or frequencies.
The outputs of the feature extractor are used for two purposes:
a. To provide inputs to the vehicle operation display. These, are used
to indicate to the vehicle operator whether or not he is operating
the vehicle within the require 1 limits.
b. To provide inputs to the feature analysis portion of the system.
2-46
..-.,.-,*.,.„ ,,■.-,„■.,,
Feature analysis consists of performing a set of specific operations on one or
a collection of basic features to convert them into one voltage or number that
is an indication of degree of fault. Typical operations include offset and
scaling, normalizing, applying cross-coupling factors, combining several
measurements, comparing to test limits and fault ordering. Also included as
a part of this function are the test results displays.
Some of the initial considerations in selecting test techniques were:
a. The driveline faults could be grouped into two categories.
1. Faults related to a specific part or type of part rather than an
assembly that contained a large number of parts. Specific part
faults included the following:
Wheel imbalance or bent
Wheel axle U-joints
Bent axles
Drive shaft imbalance
Drive shaft U-joints
Drive shaft misalignment
Clutch slippage
Faults related to complex assemblies that were made up of a num-
ber of moving parts, any one of which could be faulty. The
requirement in this category was to identify the assembly with
no requirement to identify the specific part causing the excessive
vibration. The assemblies in this category were:
• Differentials
• Transmissions
• Engine
2-47
k^jJAUiiUui--. :- '■,*■>■■ ■ "-..-■-- - __^_ ^ , ■--.*.-"' 'v^J^ÄVJe^-
Analytical analysis and empirical data indicated the following:
a. Discrete parts faults such as wheel imbalance and U-joints generate
frequencies that aro functions of the shaft rotation and whose ampli-
tude is a function of speed, load, and degree of fault. These parts
when faulty also generate higher frequencies and excite the trans-
mission and differential into generating additional frequencies.
In general, faults due to imbalance and bent shafts generate
sinusoidal frequencies that are at the shaft rate. U-joints
generate frequencies that are at twice the shaft rate and have a
tendency to be richer In harmonic content.
One major factor to consider is that all four wheels generate very
close to the same frequency since the diameter of the wheels are
nearly the same. A similar condition exists with the front and rear
drive shafts.
A rough running engine such as skip, missfire, etc., has a tendency
to generate frequencies at the engine rate and twice the engine rate.
Knocks generate frequencies at the engine rate, but are characterized
by generating a sin x over x envelope for higher harmonics. They
also set up ringing conditions in various frame members as well as
causing the transmission and differential to increase the amplitude
of the gear mesh frequencies, etc.
b. The transmission and differential will potentially generate up to
several hundred discrete frequencies related to gear meshes, bearings,
and shafts within the assembly. In general, these frequencies or
their near harmonics are higher than those due to wheels and drive
shafts. Faulty parts within these assemblies also have a tendency to
generate a large number of high amplitude harmonics. Faults in the
transmission and differential may be categorized as an increase of
mid-frequencies with a larger amplitude increase of the higher
frequencies. Gear and bearing faults also have a tendency to excite
the higher modes of frame ringing.
I 1 I I I
2-48
2.8.1 Wheels, 11-Joints and J)rive_ Shafts
The wheels, U-jolnts, and drive shafts all generate single frequencies that are
in synchronization with the shaft rates and whose amplitudes are proportional
to the degree of fault. This indicates a test technique that is very narrow
band to eliminate potentially interfering frequencies, and is phase locked to
the shaft rates. Since cross-coupling from other components are negligible at
the phase-locked frequencies, normalizing limit setting and fault ordering is
all that is required. For these applications, Northrop considered synchronous
integration or discrete frequencv feature extraction technique as being the
most practical approach to these problems. The technique uses a rotation
sensor to generate pulses in synchronization with the rotating member and then
phase locks the analyzer to either the fundamental or second harmonic of the
rotating shaft frequency. The feature to be extracted is the amplitude of the
frequency that is phase locked to the shaft of interest.
Since some of the rotating parts of the vehicle may rotate at very nearly the
same frequencies, some means must be used to introduce a change in rotation
rate or change in phase. To apply synchronous integration techniques to the
detection and measurement of wheel, axle, and some differential faults, it is
necessary to develop different angular rates for left and right components.
This can be thought of as developing different frequencies or ensuring that
there is 2 or more phase angle differences between wheels.
The knowledge of wheel rate differences as a function of vehicle path is
necessary to the design of diagnostic equipment and for the analysis of vibra-
tion data. Consider the following ; "oblem.
2-49
p
'p 'A'p
s
S ♦ AS
RADIUS OF INSIDE WHEEL PATH
RADIUS OF OUTSIDE WHEEL PATH
OISTANCE "RAVELED BY INSIDE WHEEL
DISTANCE TRAVELED BY OUTSIDE WHEEL
VEHICLE PATH
Op • ANGLE THROUGH WHICH VEHICLE TURNS
'„ - RADIUS OF WHEEL
u>„ - ANGLE THROUGH WHICH WHFEL TURNS
I?or the inside wheel t! ie angular motion is w = 1
wi r
F°r the °UtSlde wh«l the angular motion i
The difference in angular
w
WO r w
»tlon. of inside dnd outslde whteis
w^ WO wi = s +AS • s 2-^5 _ AS
«- , +AS . (% +Arp)6n
AS
s (■
P P
5 = e p
Ar P
_ G Ar —P-
1 P r
2-50
I 1
1 I I 1 1 I
"w •
Note that the phase differences between inside ind outside wheels is a function
of vehicle path angle and is independent of radius.
^w ') 1.467
P
J^r P
T 52 1 r = 15 1
For the Ml5.1
J.8.1.1 Typical > l.tues - If tiie vehicle is driven on a circular path of
300-foot radius at a speed of 10 miles per hour, the angular rate about the
path is.
10 mi/hr (5280 A/ml) . n._ „ ,, 6 ■ -,,vnn /,- v ,:mf = 0.049 Rad/sec p (3600 sec/hr) 300A
The difference in wheel frequencies is:
= 0 3.467 = (0.049X3.467) w p
w = 0.18 Rad/sec
f - v - 0.0272 Hz w 2
The average wheel rate = f = 1.77 Hz at 10 mph.
An imbalance vibration frequency would be:
Inside wheel « 1.77 =—^- = 1.756 Hz
Outside wheel = 1.77 +°'0.2.7 = 1.734 Hz.
The imbalance frequencies are sensed at the differential and are made up of
both inside wheel and outside wheel imbalances. The sensing device must mea-
sure the amplitude of frequencies at 1.756 Hz and 1.784 Hz to discriminate be-
tween right and left wheels.
2-51
_"2£^M
.".8.1.2 Kstlmates of Analysts Requirements - Some preliminary estimates of
analysis requirements can he developed. Consider the NCL PSD analysis technique:
to differentiate between right anil left wheels, the two frequencies should be
separated h\ .;t least one bin; that is, the two frequencies cover three bins. 0 0 '8K
The bin width is then ■-—~— or «0.01 Hz the sample period (T) should then be 1
~—. = 100 seconds.
If we use r)12 samples, there will be 25b bins giving the highest frequency in
the plot at 2.55 hertz. The inside wheel component would be in bin 177 and the
outside wheel component in bin 179. To analyze the r-joint frequencies at
3.512 and 3.568 hertz, a separate run would have to be made. Synchronization
must be held to about 1.5 percent over the digitizing period. The distance
around the circle the vehicle must be driven for one sample is 100 seconds x
.18 rad/sec, or 18 rad = 5 tines around the circular track.
This approach will effectively resolve right and left wheel faults, etc. It
is impractical to expect circular or oval tracks at most Army installations,
and the approach would require some modification of path requirements. An
approximate right angle turn will introduce a wheel phase difference of approx-
imately 300°. If the vehicle was driven along several legs of a block square
(or rectangular area) and a phase-lock analysis is conducted for each wheel
along each leg and an average for sp"°ral legs is taken, the interference from
other wheels would average out, leaving cnly the component related to the
wheel of interest.
The future extraction approach selected for wheels is a phase-locked synchro-
nous-integration analysis that is phase averaged for four different right wheel
to left wheel phase shifts equal to or greater than 90°.
2.8.2 Differential, Transmission and Engine
The analysis of the vibrations generated by faulty components in the trans-
mission showed the following:
a. Each gear generated a fundamental frequency at its tooth rate and
each bearing generated frequencies corresponding to inner-race,
r. 2-52
5 •wr
outer-race, hall spin rates, and hall pre« ision rate. bach shaft
generated frequencies .it thus.- rates.
h. All of the above gear and bearing frequencies have a tendency to be
rich in harmonic content.
c. Cross-coupling within an assembly and Iron assembly to assembly modu-
lates the fundamental gear and bearing frequencies generating many
sidebands about each.
d. The amplitude of the fundamentals and their harmonics is proporticnil
to load. Faults, however, increase the amplitude of the harmonics
much more than they do the fundamental.
e. The higher frequency modes of the housings and frames are excited
to higher amplitudes by faulty gears and bearings.
f. Bent shafts have a tendency to increase the amplitude ol the gear
mesh frequencies.
In summary, vibration energy in the band from 20 to 400 hertz is proportional
to the amount of power being transferred. in the higher frequency bands,
energy is proportional to power transfer but increases much more rapidly with
part failure.
Selection of test techniques that utilized individual frequency 01 each compon-
ent would result in a large list of synchronous frequencies and amplitudes for
eacli part within the transmission and differential. In addition, equations
for the cross-coupling of each of the frequencies to other parts would be
involved. This approach would work and appears desirable for production test-
ing and engineering evolution of new designs. The approach is impractical for
field maintenance applications for the following reasons.
a. The identification of Individual gears and bearings would be confus-
ing to the field maintenance mechanic. His only interest is the
complete assembly.
2-5 J
■ <4p~
b. Memory required for storing constants and cross-coupling factors for
components and the related calculations is impractical for met field
r.ppl icat ions.
c. The number of measurements required and the related caJculation.s
Rre.it lv increases the test time.
d. Experimental dita requirements for various conponent failures is much
more extensivt than would be permitted by this program.
When there is only the requirement to identify the assembly as being bad with-
out the need to identify components within the assembly, the bandpass ratio
technique becomes more practical. This approach takes advantage of the fact
that the high frequency vibrations increase more rapidly than the lower fre-
quencies as components fail within the assembly. This approach was selected
for the engine, transmission, and differential driveline assembly. It has
the following advantages.
a. For each assembly, only two measurements are required: one energy
in a low-pass band and two energies in a high-pass band.
b. The calculation involves taking the ratio of the high-band energy
to the low-band energy which eliminates the major effects of power
transfer.
c. Selection of appropriate bandpasses potentially eliminates or greatly
reduces the effects of cross-coupling between engine and transmission
and between engine-transmission assembly and the differential
assemblies.
d. Since accelerometers will be mounted on both differentials, it would
be practical to separate front and rear differential faults by magni-
tude comparison. Experimental data indicated that the coupling be-
tween front and rear differentials was comparatively low.
2-5/4
:.:"!Wipr ■... ,"V3»,-W«*i-*---
I I I I 1 I
1 [ i I I I I I
2.8.3 Clutch
In a vehicle, a defective clutch slips under conditions of high vehicle accel-
eration or heavy load, and the amount of slip was found to be several revolu-
tions. For this application, the vehicle will be assumed to be unloaded and
that clutch slippage is induced by high acceleration in high gear. The tech-
nique selected counts the revolutions or the engine, effectively divides it by
the transmission gear reduction, and compares this to the drive shaft revolu-
tions. Differences in the two counts indicate the amount of clutch slippage
when measured over a specified time and under specified operating conditions.
2.8.4 Selected Test Techniques
In summary the selected test techniques were:
a. Discrete synchronous frequency analysis for wheels, axles, U-joints,
and drive shafts.
b. Bandpass ratio techniques for differentia!, transmission, and engine.
c. Engine shaft rate versus drive shaft under acceletation conditions
for slipping clutch.
2-55
sa«w-ife^w^ÄS^KA^^^
PRECBDUO PAOB DUNK-NOT tl "f
SKCTION i
IMPLEMENTATION OF THE DRIVELINE ANALYZER I)I-SK;N
:f'.
This suction describes the implementar ion ol the performance requirement:- into r.!u
design of the driveline analyser hardware rind the rational,1 employed in resolving critical development decisions.
J.1 OVERALL CONCEPT
The performance requirements for the driveline analyzer have established that test set must:
Be a stand-al
to he tested one portable unit that is easily attachable to the v,
Us e power from the vehicle battery
• Be automatic in its operation
• Be capable of displaying the results to indica
of the vehicle. te good or bad condition.1-
I I 1 I I
Figure 3-1 depicts the resultant block diagram for the driveline analyzer which
I is an integrated series of tests which will provide a tool for automatically
I I I
required to perform the data gatherin g portions of the tests.
diagnosing the vibrati on source in the vehicle as it is driven en the road.
-fe<w;.-N*>^O^.^
3-1
',i- ^ir^afcnÄ- ^■^■&^&i^i-s:v&v'--M^K!-l.\ vwy^iji&^Vik
^^^^m^%%m^^^^^^^sm^^u^-,, %&$
TfcST RESULTS OISPIAY
AND CONTROL
VEHICLE OPERATIONAL
OISPLAV
ENGINEERING OlSltAV
Figur*. 3-1. Mil VELINE ANALYZER BLOCK DIAGRAM
This configuration will detect abnormal vibrations that indicate faults in the
driveline subassemblies listed in Table 3-1.
Table 3-1. DRIVELINE SUBASSEMBLY FAULTS
3-2
.. ,
Subas. jmbly
Clutch
Ft ilt Mode.
Slippage
X
Imbalance Misalignment Excess
Vibration
Drive Shaft Drive Shaft U-Joinjs
X X Wheel Assembly
X •fc
Wheel U-Joints Axle X Engine X
Transmission X Differential X
X
i I I I I I 1 1
,2 CENTRAL PROCESSOR REQUIREMENTS
'Hie large number of data points •, ,
—.. -^ JUZ'Z :::::::: - " —• - the ot
Preliminary analysis of the data showed that between 8- ;ind '6-bit accuracy was
required for most data. The analysis of wheel and axle data would result in summa-
tions that exceeded 25 bits. This large number of hies was required for the long
time period? of synchronous integration in the presence of large noise-to-signal
ratios. To conduct the analysis in real time, thus eliminating the require.', nt for
storing large quantities of vibration data, it would be necessary for the processor to add in approximately 7 microseconds.
Analysis of control and processing requirements indicated the following:
a. Between 40 and 50 measurement;: rau.«f bt ::.«<!"'.
b. A number of measi
LJ minutes. dements must be made and stored for periods up t P to 10 or
c Accuracies of requirements for the measurements
better than 1 percent. were typically equal to or
d. To offset, scale-, normalize, and compensate for rross-coupling and fault
order, there was a retirement for approximately 400 additions and sub-
tractions, multiplications, and divisions. One hundred constants were used in calculations.
3-3
-vr
e. There were approximately 100 control commands to signal conditioning and
display functions.
f. There were 40 numerical numbers involving spc-ed, gear, and other opera-
tional conditions that must be displayed to i:h*» operator.
g. There are approximately 80 pieces of variable data that must be stored
and used during one complete set of tests.
It. Approximately 1200 words of program and constants, and 100 words of
read-write memory for data and test variables would be required.
To satisfy these requirements, Northrop considered a number of commercially available
computers. Analysis, as indicated in Table 3-2, showed that most of these computers
were not suitable for portable application and tl at their power was much higher than
could be supported by vehicle power.
A CMOS computer being designed by Northrop was also considered. A comparison of
power consumption versus performance is shown in Figure 3-2. In this comparison,
Relative Performance is taken as one over required instructions times execution
time, \ftien required, Instructions is the number of instructions required to
perform a stanaard sequence of operations, and the Time is the time required to
perform a standard sequence.
3.2.1 Central Processor Implementation
Selection of a specially designed CMOS computer was made and was based on the
following:
a. The low power requirements of the CMOS system. Less than 1 ampere of
vehicle power required. This la up to 100 times less than the power
required by commercially available models.
b. High noise immunity of CMOS components to the vehicle ignition environment.
The CMOS approach has five times the noise immunity as compared to other
approaches.
1 I
3-4
■**-
-~ ii .,.,. " ~ } i I I I I ~ --~
--~ . i I I 1,~ ' ~;
w
I Ln
~~ f·
(;_
-~
~
~ >
w
r m
0 0 "t2 -<
----
--·-
·---
.,.-
-r-
---·
-·-
isin
g:"
::>r
I Do
u~le
!
:...·a
:-d
Co
r::.
?u
ter
Instr
Tab
l,.;
3
-2.
CO~Pl3"TER CO
~~~\
i~IS
\)~;
-~-
Fix
ed
o
r j
Var
ic1!
:>le
i
:r!s
tr
Fie
lds
.!;..c
ct:..
..'":!
?.e
gis
::ers
I:
--.d
ex
? ~~ ,-
.: <:
:-r_.
:-
,.
··~!------r-~
.~.d
d
B::
te
.=-·.,
.:ay
l:!.
-::2
;~. !"
".:. :
:-::
ex
. .!.~
:: c
-:·
:=~~
=-~
:-. .: :
-~· :-
_,_.:
:' ·..
;--3
.: :5
)
Lo
ckh
eed
, S
ue
S,D
F
7 G
.P.
I (s
an
e)
__ _
~les
Yes
2
.8
:iex
.
TI
960A
.-\d
vdnc
e 6
13
0
SEL
.S
lO.\
(S
ys
En
g
Lab
s)
D
CO
HP-
16
(Un
ico
mp
I S l
Inc.)
!
CD
20
0
(Co
mp
ute
r ~.5,5~
(Dev
.)
{8
b
its)
Ph
ilip
s,
P/?5
0
Dat=
te,
16
Rec
ico
r,
RC
7
0
! In
terd
ata
, 80
Mic
ro
81
0
S,D
s S,D
S,D
I ~.5
,5!~
1<8
bit
s}
l '
Nov
a I
S ! N
ov
a 1
20
0
, l
~~----
I i
Alp
ha
16
I
S I
I HP
2114
A_ --
1 s
Un
dete
rcln
ed
=
?
F .- v v F F v v v
-------------------~-:-
·---
----
----
----
----
--8
G.P
. 8
(sa:
':le)
~~
0
---------
l +
1
---
2 1
1 +
1
(in
co
re):
14
G.P
.
1 +
1
1
16
G.P
.
1 +
1
3 ~0
l ~;
0
7 ~'
o (i
n co
re)
1 +
l
Yes
(i
n
co
re)
14
(s
an
e)
1 1
15
(sa:
::e}
1
~ I
Yes
(A
ll
Col
:lb)
No
Yes
Yes
Yes
~;0
~;0
~;o
~0
Yes
~\o
~0
No
No
3.6
] • 3
3.5
., -
-·:>
9.6
I H
eY.
I
Octa
l --~----
I
0-:
tal
Octa
l
He
x \A
lph
aS
=
Dis
pla
y)'
3 .,
:.,
. h
• ..
I ~'e
J.t~
er
2 O
cta
l
1.9
H
ex
Hex
5 E.
ex
j i
! i
I .
r--
----,
! ,
i 4
2 (s
=.e
) ! 1
No
l S
o 5
.9
! Oct
.. 1
I !
I l.
35 I
~;o
~0
~·0
~0
~~0
~0
~0
So
i_ __
___ _
75
};.)
0
90
:..:.cj
:
( :A
J5
::e
::::)
l.2
J'J
t.JQ
3:30
2so-
.:..oo
175-~75
I 1
1 I
" I
,. 3
? .
..
. . r~ 1
.~o
• • _
:
ttex
'
. ~-
_ ~'
j N
o 1
NJ
.. j Oc
ta~-
~---
N,:o
.::>
,,,_
35
33
<)
~0
2 ~cne
-------------------
;
~~'"';~~~~4.~..!~::'6:·:: .. :,~
-:~~
)t~~
~~M;
'~v'
.,.,
;~-.
=· ...
-.i-, ---
~w.~·,'*-----·
-~
_.j.
~?"'~
~-:·
t~,
...
'1:.~"_:~;
r t; f I f ~' ~- {~ ~· t- ~:· ,._ .. I < ... l f I ! L •.r:•
g.oow
10W
m
MP7114A SUE
NOVA TIMO ALPHA It
MICRO 810 NOVA
1200
CDC 4«
I "8' I
' NORTHROP' I CMOS I ^PROCESSOR]
MCS4
RELATIVE PERFORMANCE (REQUIRED INSTRUCTIONS) (TIME)
Figure 3-2. POWER/PERFORMANCE COMPARISON OF VARIOUS MINICOMPUTERS
c. Compatible with the low cost read mostly memory devices.
d. CMOS devices are not damaged by power shorts, etc., do not require logic
ground planes and low switching transient generation that would interfere
with sensitive analog circuits.
e. Optimum instruction site. The ability to use micro-instructions to develop
instructions specially oriented to driveline testing greatly reduced the
number of instructions required.
f. A large number of device interrupts and special device commands could be
provided for vehicle operational monitoring and control of: sensor devices.
A functional block diagram of the Northrop designed CMOS processor is shown in
Figure 3-3.
3-6
I I I !
I !
1 i 1 1 I I l l
Figure 3-3. NORTHROP CMOS PROCESS
ANALOG INPUTS
OR
EXTERNAL EVENTS
3300
3-7
:--v^h^.j.^:a;,---:.,;,-.L,\: t_,-: *~-&SU,l-
i. 1 VIIHICLK OPERATIONAL DISPLAY REQUIREMENTS
There are a largo number of vehicle operational modes during a complete series of
tests on the Mlr>l vehicle. These operational modes involve operating the vehicle
in a specified gear at approximately a specified speed along a prescribed path.
Because tbese tests are conducted on the road, traffic and other considerations will
cause interruptions in the test sequence. In some cases, a test mode will have to
be repeated because it was impossible to hold the vehicle operational cond'tions for
the specified time. Provis ons must, be made to provide the vehicle operator with
indicators that instruct him concerning vehicle gear, engine rpm, vehicle speed, and
the identification of the test being conducted. The vehicle operator should also be
provided with indications of whether or not he is meeting the required operational
conditions and if it is necessary to repeat a test sequence. The types of informa-
tion that must be presented to the vehicle operator are:
• Engine rpm for engine and clutch tests
• Transmission gear for specific test modes
• Speed in mph for clutch, transmission, differential, and wheel tests,
The monitored conditions display requirements are:
Engine rpm
Speed in mph
Waiting for conditions to be met
Invalid test
Test complete.
3.3.1 Operator Display Implementation
The remote display panel on the operator display assembly provides the vehicle
operator with the indications as illustrated in Figure 3-4. This assembly is
packaged in a box 6 by 7-1/2 by 2 inches and attaches to the metal panel just below
the windshield.
I I f
3-8
1 I I I I I 1
REMOTE DISPLAY CONSOLE
I
RPM H O REQUIRED .
SPEED REQUIRED
* »
- (Z
RPM Ä i» GEAR $> VCTUAL j REQUIRED
, 0WAITING # CLUTCH
• 4fc TEST IN A n.r_ Ä W PROGRESS h rr,
mm • INVALID ÄTRANA TEST X
SPEED f> ACTUAL ■F
Ä
# ENGINE •WHEELS
Figure i-4. KEMOTK DISl'I.Ai n'N'Snu PANFL
i.J.2 Description of th_e Misjilay.s
'Hie operator display assembly has the following operate! c-r iented display functions.
WAITING Initially illuminates when analyser has completed a power-up sequence,
successfully completed a processor seM'-test, has initiated the program,
and is waiting for operator action.
I I I 1
During a test mode the lamp will illuminate whenever an operator action
is required to change gears in tin transmission, rpm of the engine,
speed of the vehicle, or the direction of the vehicle to cause wheel
phase shift.
3-9
TKST IN PROGRESS
Illuminates whenever the vehicle is heins operated in the appropriate
test mode and the processor is conducting data acquisition or data
analvs is.
INVALID Illuminates to inform the operator that the required test conditions
were not met during the previous test program and that the test should
he repeated.
ENGINE, CLUTCH, TRANS, DI FF, WHEELS
These five individual lamps will illuminate as the diagnostic test
progresses through the individual test modes to enable the operator
to determine that test conditions are being successfully met.
RPM REQUIRED
The numbers (times 100) designate the revolutions per minute that
the vehicle operator should operate the vehicle engine.
RPM ACTUAL
The numbers (times 100) display the actual rnm at which the engine is
being operated.
SPEED The numbers designate thp speed in mph that the vehicle operator should REQUIRED . , fcl . . ,
drive the vehicle.
SPEED ACTUAL
The numbers designate the speed in mph at which the vehicle is actually
moving.
GEAR REQUIRED
The number designates the gear position in which the transmission
should be placed:
1 ■ 1st gear 3 ■ 3rd gear
2 = 2nd gear 4 = 4th gear
5 = reverse gear
0 = neutral
3.4 TEST RESULT DISPLAY REQUIREMENTS
There are a number of assemblies that must be identified with an indication of the
degree of fault indicated. A method to indicate the degree of fault is to use
numerically normalized values from 0 to 9. During engineering evaluation it would
3-10
■w
I i be highly desirable to display a higher resolution number, for example 00 to (*9
The faults which must be identified include the following:
i I I I
a. I'ngine
b. Clutch
c. Transmission
d. Rear Drive Shaft Imbalance
e. Hear Drive Shaft U-joints
f. Rear Differential
g. Right Rear Axle U-joints
h. Right Rear Axle Bent
i. Right Rear l/heel Imbalance
j. Left Rear Axle ü-joints
k. Left Rear Axle Bent
1. Left Rear Wheel Imbalance
m. Front Drive Shaft Imbalance
n. Front Drive Shaft l-joints
o. Front Right Axle U-joints
p. Front Right Axle Bent
q. Vront Right Wheel Imbalance
r. Front Left Axle ll-jcints
s. Front Left Axle Bent
t. Front Left Wheel Bent.
Typical display arrangements for these faults might be as follows
T~~ ■
A | U F 7
♦ Degree of fault
-Left or Right
] Measured value or
normalized value for
engineering use only
—Front or Rear
■Assembly Axle U-joint, transmission, etc.
In this case, a technique must be provided for presenting the faults in descending
order of the des-ree of fault. Provisions for manually sequencing from one fault to
the next must be. provided.
3-11
^;.■-:,.>■,-:_ v-y.'«f*:-H-Wt.,*JV,>^W-'-..^.:.- JXiQ.*
w
'.-».I Test_ JiesulJ: JUs£lay Implementation
The display selected for the hardware implementation is a l^-character alphanumeric
am! is contained in the logic and control assembly. This display is used tor iden-
tification of the faulty driveline components. 1'pon completion of a valid test
program, the front panel FAULT DISPLAY will illuminate <.• ith a TEST COMPLETE message.
To observe the test results, the operator actuates the FAULT DISPLAY INDEX pushbutton
also located on the logic and control assembly.
The faultv assemblies will be identified in the hierarchy oj faults. The FAIL!
DISPLAY, will first display a message identifying the most significant fault, together
with a normalized degree of fault and the order of faults, as multiple faults can
occur through the cross-coupling of vibration. The most likely i. 'use of vibration
fault will be the first fault displayed.
3. A. 2 Description jof_ the Test Kesult Messages
The format utilized for these messages is as follows:
7 8 9 10 11 12 13 14 15 16
Fault Identifier
Degree of Fau 11
Rank Order of Fault
The abbreviations utilized for the fault identifier messages are defined below.
Test Name
REAR RIGHT WHEEL
».EAR LEFT WHEEL
FRONT RIGHT WHEEL
FRONT LEFT WHEEL
REAR RIGHT AXLE
REAR Li..'7 \XLE
FRONT RIGHT AXLE
FRONT LEFT AXLE
REAR SHAH' FUND
FRONT SHAFT FUND
PJ-?play Code
WHB-RR
WHB-RL
WHB-FR
WHB-FL
WHA-RR
WHA-RL
WHA-FR
WHA-FI.
R-DS-F
F-OS-F
3-12
I I
I I I I I 1 I I I I I I I
I I I 1 1 *
1
1 1
1 1 t 1 I I I
Test Name Display Code
REAR SHAH' 2ND R-DS-2
FRONT SHAH 2ND F-DS-2
REAR HIFF HI RIUF-H
REAR IUFF RATIO RD1F-R
FRONT IUFF 111 FD1F-H
FRONT 1)1 I F RATIO FDIF-R
IRANS GEAR 1 Hl TR-l-H
TRANS ('.EAR 1 RATIO TR-l-R
TRANS GKAR 2 HI TR-2-H
TRANS GEAR 2 RATIO TR-2-R
TRANS GEAR 3 HI TR-3-H
TRANS GEAR 3 RATIO TR-3-R
TRANS GEAR 4 Hl TR-4-H
TRANS GEAR 4 RATIO TR-4-R
ENIJIIV. 600 Jii EGOO-T.
ENGINE 600 RATH) E600-R
ENGINE 1000 HI E 1K-H
ENGINE 1000 RATIO E 1K.-R
ENGINE 2000 HI E 2K-H
ENGINE 2000 RATIO K 2K-R
CLUTCH SLIP CLUTCH
Following are definitions of trims used in Test Name:
FUND - Tlii' fundamental frequency oj rotation tor the component ol interest
2ND - The second harmonic ol the rotation frequency for the component oi
interest
HI - The measurement value of the high frequency bandpass
RATIO - The ratio of the high frequency bandpass to the low frequency bandpass
values.
J.5 SENSOR AND SIGNAL CONDITIONING REQUIREMENTS
The requirement of applying the driveline analyzer to the vehicle in less than the
10 minutes dictates the use of a small number of easily connected sensor assemblies.
This requirement excludes the ability to instrument areas of the driveline subassem-
blies, which would require excessive time to install or modification of the vehicle.
3-13
From the test technique selection, the rotation sensor requirements become six:
• Four for the wheel rotation sensing,
• One for the engine rpm sensing
• One for drive shaft rotation sensing
The rotation sensors will require phase-locked-loop logic. Multiplexers are adaed
to allow for time sharing of the hardware.
Three accelerometers will be required. This was determined empirically and fron
the requirements imposed by the fault separation test technique selected for the
front and rear differentials.
Fach of the ac'» aromerers used to sense vibration must have the following signal
conditioning provisions:
a. Programmable preamplifier with 0 to 42 dB dynamic programmable range and a
bandpass of 0.1 Hz to 10 kHz.
b. Programmable filter wich four-pole high pass and four-pole low pass.
c. Progiammable post amplifier with 0 to 42 dB programmable gain and a band-
pass of 0.1 to 10 kHz.
d. An analog-to-digital converter programmable to select any of three
accelerometer channels and with a conversion start control.
3.6 DRIVEL1NE ANALYZER SPECIFICATIONS
CJMMPONENTS One Logic and Control Assembly (Part No. 307225)
One Operator Display Assembly (Part No. 307226)
Three Sensor Assemblies (Part Nos. 307227, 307228, 307229)
One Power Cable (Part No. 307231)
MEASUREMENT Four counter-t imer cnanneib v. 'J-Lll UlUXLJ-^J-*'"- — r
CAPABILITY
1 Hz 10 kHz PL1SF'.5) CTK3r,C
10 Hz 100 kHz PL2SB(6) CTR4TC
100 Hz
1 kHz
1 MHz
PLiSA(l)
CTRJ rc
CTRT' r
PLLl
PLL2
I 1 1 !
3-14
-sr
Two amplifier-filter channels with multiplexed controls of:
Input Select: Ground, Accel No. 1, Accel No. 2
Input Gain: Unity, 6 dB, 12 dB, 18 dB, 24 dB, 30 dB, 36 dB, 42 HR
High-Pass Filter: Bypass, 10 Hz, 30 Hz, 90 Hz, 700 Hz, 2200 Hz,
5 kHz
Output Gain: Unity, 6 dB, 12 dB, 18 dB, 24 dB, 30 dB, 36 dB, 42 dB
One A/D converter with multiplexed controls of:
Input Select: Chan 1, Chan 2, Chan 1 Avg, Chan 2 Avg, DC Test,
Ground
PROCESSOR CAPABILITY
One 16-bit data bus (processor and I/O functions)
One 16-bit address bus (address memory and I/O devices)
Eight 16-bit general purpose registers (data opera-.ds, accumula- tors, or address index registers)
One 16-bit arithmetic logic unit (ALU)
Four 16-bit memory address registers (memory address pointers)
256 words alterable read-only-memory (microinstructions)
1536 words alterable read-only-memory (macroinstructions)
1028 words random access memory (read-write)
READOUT Operator Display (Remote Unit)
I
WAITING
TEST IN PROGRESS
INVALID TEST
ENGINE
TRANS
DIFF
WHEELS
Waiting for operator action
Performing data acquisition or analysis
Instruction to operator
Programmed for test related to engine
Progi-anuned for test related to transmission
Programmed for test related to differential
Programmed for test related to wheels, wheel axles, propeller shaft, and U-joints
3-15
4V ;ti*.ja'"-- <.».*.«»».»■*•*>*"•■■ ■■•■ *w»«tj«'*t«»*'i*W*
CONTROLS
POWER
REQUIREMENTS
RPM REQUIRED Designates speed of engine required
RPM ACTUAL Dicplays actual speed of engine
SPEED Designates speed of vehicle required REQUIRED
SPEED ACTUAL Displays actual speed of vehicle
GEAR Designates transmission position required REQUIRED
Front Panel (Logic and Control Assembly)
Alphanumeric 16-character diagnosis display
Engineering Control Panel (beneath front panel)
Sequence states
Address bus
Data bus •*
Front Panel (Logic and Control Assembly)
POWER On/off control for 24-volt source power
RESET Program reset to start of selected data acquisition
FAULT DISPLAY Allows for stepping through fault display messages INDEX
RUN Initiates the test j
Engineering Control Panel (beneath front panel) I
Processor Control Switches (5) 1
Register Selection Switches (7)
Address/Data Input Switches (16) "f
Voltage: 25.2 f3 volts DC ^
Polarity: Negative terminal grounded to frame (Reverse polarity |
is protected)
Current: 7-1/2 amperes maximum, 3 amperes nominal 1
I I
MECHANICAL
I I I I I I 1 I I I 1 I I I K I I
DIAGNOSTIC PROGRAMS
Size: Logic and Control Assembly - 12 x 14 x 17 inches
Operator display Assembly - 6 x 7-1/2 x 2 inches
Weight: Logic and Control Assembly - 40 pounds
Operator Display Assembly - 2 pounds
Sensor Assemblies (Total) - 11 pounds
Test Select and Control
Wheel Data Acquisition and Analysis
Propeller Shaft Data Acquisition and Analysis
Differential Data Acquisition and Analysis
Transmission Data Acquisition and Analysis
Clutch Control Data Acquisition and Analysis
Engine Data Acquisition and Analysis
Test Results Analysis and Display
Engineering Data Display Control
Analyzer Self-Check
3.7 HARDWARE PICTORIALS
The driveline analyzer chassis installed in the M151 vehicle is shown in Figure 3-5,
and the vehicle operator's remote display is shown in Figure 3-6. The rear differ-
ential sensor assembly, transmission sensor assembly, and the front differential
sensor assembly are shown in Figures 3-7, 3-8, and 3-9, respectively.
A detailed description of the installation and use and interpretation of the drive-
line analyzer is given in the Instruction Manual M151 Driveline Analyzer Feasibility
Model.
3-17
Figure 3-5. DRIVELINE ANALYZER CHASSIS IN PLACE
I
Figure 3-6. REMOTE DISPLAY CONSOLE IN PLACE
i
\
I I
3-18 I
Figure 3-7. FRONT DIFFERENTIAL SENSOR ASSEMBLY INSTALLED
I Figure 3-8. REAR DIFFERENTIAL SENSOR ASSEMBLY INSTALLED
I 3-19
Figure 3-9. TRANSMISSION SENSOR ASSEMBLY INSTALLED
I
3-20
i
I I I I I
I I E I !
I t 1
T
I 1 I I I I
SECTION 4
TEST AND EVALUATION
(Summary of Tests)
The test and evaluation portion of the drive]ine analyser program involved the
following phases.
• Bench Test which included the following:
a. Computer program sequence checkout.
b. Use of simulated transducer signals as inputs to the analyzer to
determine proper measurement of specified signal features.
c. Use of magnetic tape recordings of transducer signals from on-the-road
test to show compatibility and reproducibility of readings using actual
vehicle data.
• Vehicle Compatibility Test which included:
a. Attachment of sensors to the vehicle and installation of the driveline
analyzer and display in the M151A2 1/4-ton utility truck.
b. Operation of the driveline analyzer from vehicle electrical power
while operating the vehicle on the road.
• Preliminary Road Test to:
a. Demonstrate the ability of the test set to measure vehicle operational
parameters and display them to the vehicle operator with desired modes
of operation during the various test sequences.
b. Obtain preliminary magnitudes of extracted features for a good vehicle.
4-1
.-äf*'-34ft'?attr4tftf' «W/ÖftT «SOW* -: s«»*f» *S «B»$ "'■
c. Obtain preliminary magnitudes of extracted features for certain
inserted faults.
After completion of these tests the driveline analyzer was shipped to TACOM.
4.1 BENCH TESTS OF THE DRIVELINE ANALYZER
The bench test portion of the program involved operating the driveline analyzer in
the laboratory to determine if the driveline analyzer performed to the design goals.
The bench test included the following phase=>
a. Computer Program Operation - All feature extraction, data processing,
test faults display and the vehicle operational display and control program-
were checked, using simulated digital data. The test, programs were modified
to use special digital words as substitutes for data inputs from the trans-
ducers and signal conditioning equipment. Each test was initiated to
determine if the program was extracting the desired feature, analyzing
data, and displaying the results properly. The advantage of this approach
was that the simulated input data could be controlled precisely so that the
accuracy of the feature's extraction and analyses could be determined.
These programs included the following:
• Driveline analyser computer program process flow and control routines
• Tnput device control programs for signal conditioning, analog switching,
counter control, analcg-to-digital converter, binary-to-BCD conversion
and phase lock loop controls
• Vehicle control monitor programs; these programs included the function
of display of desired vehicle operational conditions, monitorirg of
vehicle speed, gear and engine rpm
• Driveline analyzer bandpass ratio tests; this included input device
setup, speed and gear control setup, high-pass and low-pass band
measurements and ratio calculations
• Wheel/axle data acquisition program
";: 4-2 I
w
I I I I I 1 1
1 I I 1 1 I I \
Wheel imbalance data analysis
Data analysis for propeller (drive shaft) imbalance
Data analysis for wheel U-joints
Transmission test
Differential test
Data normalizing and ordering program
Fault display program
Each of the above programs were checked individually and then linked together to
check data acquisition, feature extraction analysis, and display functions as a
complete sequence.
b. Checkout Using Simulated Analog Signals - Laboratory oscillators, pulse
generators, attenuators, etc., were used to simulate signals from the
transducers. These signals simulat-.ed the output of accelerometers and
rotation sensors. The amplitude and frequencies were controlled so that
precise results could be predicted. The technique made it possible to
determine the proper operation of signal conditioning equipment, feature
extraction programs under dynamic data conditions, and analysis and display
routines.
A block diagram of the setup used to check synchronous analysis routines is
shown in Figure 4-1. The arrangement makes it possible to simulate accel-
erometer outputs that have frequency components that are in synchronism
with rotation pulses. The ratio of the oscillator frequency to rotation
pulses can be selected by the divider circuit. Interfering signals can
also be simulated by mixing in signals from oscillator no. 2. The second
pulse generator is used to simulate variations in vehicle speed, engine
speed, etc. The results of a typical test are shown in Figure 4-2. In
this case, the linearity and dynamic response of the driveline analyzer
4-3
"W"
OSCILLATOR NO 2 !
SIMULATED
DRIVELINE ANALYZER
AMPLrrUOE MONITOR
MIXER CIRCUIT
ACCELEROMETER INPUT
OSOLLATOR NO. 1 ATTENUATOR
' k
SIMULATED I i i i
< ' SENSOR INPUTS
SQUARING CIRCUIT DIVIDER -*
PULSE GENERATOR
NO. 1 ' ' RATIO
MONITOR
PULSE GENERATOR
NO. 2 t
Figure 4-1. SYNCHRONOUS SIGNAL SIMULATI ON
3 4 B« 7 » 9 xo 11 13 ij
INPUT VOLTAOE, PEAK-TO-PEAK 3X0
Figure 4-2. BENCH TEST - INPUT SYNCHRONOUS SIGNAL AMPLITUDE VERSUS DRIVELINE ANALYZER READING OPERATED IN
DRI"E SHAPf TEST MODE
4-4
I "W
synchronous measurement capabilities were checked in the presence : an
interfering signal.
Simulated bench tests were conducted for the following:
Differential bandpass ratio
Transmission bandpass ratio
Wheel U-joints test
Differential imbalance
Differential U-joints test
Transmission gear ratio check
Vehicle speed monitoring
Magnetic Tape Recordings of On-The-Road Tests - During the empirical
data acquisition phase magnetic tape recordings were made of vehicle
vibration and shaft rotation. Recordings were made for both good and bad
assemblies. These recordings were played back into the driveline analyzer
for fault analysis. A block diagram of the technique is shown in figure
4-3. These tests demonstrated the ability of the driveline analyzer to
operate on actual vehicle data. The advantages of this technique were:
1. A repeatable set of vibration data could be provided, thus eliminating
the variations of vehicle, operator, and road conditions.
2. The tests could be conducted in the laborai y where a full set of
laboratory instruments are available for verification of the driveline
analyzer signal conditioning and data processing.
3. The modifying of analysis procedures could be accomplished without
setting up a new road test for each change.
A technique like this was used to demonstrate to TACOM personnel the
ability of the analyzer to detect a bad differential.
4-5
~<7t
ON THE ROAD RECC1DINC OF VEHICLE VIBRATIONS ANO SHAFT ROTATIONS FOR VARIOUS OPERATING SPEEOS AND WITH GOOO AMD BAJJ ASSEMBLIES
DI»LAV
LABORATORY TEST OF THE DRIVELINE ANALY2ER US'NG MAGNETIC TAPE RECORDINGS FROM ON THE ROAD TEST
Figure 4-3. LABORATORY TESTS USING MAGNETIC TAPE RECORDINGS FROM ON-THE-ROAD TEST
4.2 VEHICLE COMPATIBILITY TESTS
The objective of this phase of the program was to demonstrate the compatibility of
the driveline analyzer with the M151A2 vehicle.
It was demonstrated that one man could install the test set in less than five
minutes.
A second portion of this _est demonstrated that the driveline analyzer would
operate from vehicle power under typical road operational conditions and would
not be affected by ignition and other electrical noise or road vibration. Tests
included:
a. Determination of vehicle speed
b. Determination of vehicle gear
c. Display of required vehicle operational conditions
d. Sequencing through each of the program tests
e. Display of test results.
4.3 ON-THE-ROAD TESTS
A limited number of the tests were conducted while operating the M151 vehicle on
the road. The purpose of these tests was to establish some baseline capability of
the driveline analyzer. These tests were limited due to lack of conTact funds and
4-6
I I !
I
tine. Plans call for extensive evaluation tests to he conducted by TACOM to
determine feasibility of the concept and to establish limits for many of the test
parameters.
4.3.1 Drive Shaft U-Joint Tests
A series of tests was conducted to demonstrate the capability of the driveline
anal/zer to:
a. Detect imbalance in the rear drive shaft
b. Detect imbalance in the front drive shaft
c. Differentiate between front and rear drive shaft faults.
The tests also demonstrated the driveline analyzer's ability to measure vibration
due to U-joints and due to engine firing. Three conditions were simulated:
a. With no weight added to either drive shaft (no weight condition)
b. With 6 ounces of weight clamped to the rear drive shaft
c. With 6 ounces of weight clamped to the front drive shaft.
The driveline analy °.r was set up to displayed measured values for the following:
Rear drive shaft first
fTn-phase component of vibration.
Rear drive shan't second
(Second harmonic component of vibration)
Front drive shaft first
(In-phase component of vibration)
Front drive shaft second
(Second harmonic component of vibration)
4-7
,.■ tfj$ . v..._-^\: ),Ü?zä&'iä-ü'vS&iy&r ■ '"■■ *-*"*_■ Jt%-:;':J''-. ■^■I,M'.-.#^, :-^.^:;w,:>Vi,v^--;^^iir.■.ife^/-'-:,!,-::-'.uL^iA-^^h'. iJ-.^^i^Di^: *j-5-;.;_^-ii
"W"
The vehicle was operated on the road at a speed of 35 mph for the following
conditions:
Test No. 1 - Third gear
Test No. 2 - Fourth gear
Test No. 3 - Third gear, four-wheel drive.
4.3.2 No Weight on Drive Shaft Test
The test results for no weight added to either drive shaft, presumably good vehicle,
is shown in Table 4-1. The values provided here are considered baseline values.
Sorae significant changes in measured values as of function of third and fourth gear
operation and four-wheel drive operation are as follows:
a. Rear Drive Shaft First (In-Phase Vibration)
1. Changing from third gear to fourth gear increases the rear first from
an average of 32 to 111, a factor of 3 increase. This increase is due
to cross-coupling between engine imbalance and rear drive shaft
imbalance. Going to fourth gear causes the engine to rotate at the
same rate as the drive shaft. Drive shaft imbalance should not be
measured in fourth gear.
2. Operating the vehicle in thK'd gear and four-wheel drive caused the
output to increase from 32 to 46 average, which is insignificant. Some
small amount can be attributed to additional drive shaft loading when
in four-wheel drive. This is the same as increased loading. A part
of this increase may also be attributed to additional cross-coupling
between front and rear drive shafts.
b. Rear Drive Shaft Second
1. Changing from third {sea** to fourth gear increases the second harmonic
from 86 to 193, or a factor of a little over 2. This increase can be
attributed to cylinder firing rate and is cross-coupled to the drive
shaft imbalance wh»n the transmission is in fourth gear.
4-8
«^«^f-tmmsmmmmm^fm, smui^m^-«^
■ 1>" ■■ "'JJ'1 f WJl
«u ii HI 'T-Ä***.'!-!1.-..11' -
Tablj 4-1. ON-THE-ROAD DRIVE SHAFT TEST, NO WEIGHTS (GOOD »RIVE SHAFT),
35 MPH, NO FAULT
I I I I I L I
I I I I I 1
1 ,
Condition Run No.
Rear Drive Shaft Front Drive Shaft
1st Harmonic
2nd Harmonic
1st Ha rmon i c
2nd Harmonic
Test No. 1
Third Gear
1
2
3
4
28
13
46
43
SI
70
98
95
210
204
227
203
113
201
23
13
23
13
04
03
0
3
Test No. 2
Fourth Gear
I
1
2
3
4
5
6
60
91
104
149
121
143
51
39
6 )
40
57
89
550
500
527
519
458
474
Test No. 3
Third Gear
Four-Wheel Drive
1
2
3
4
5
63
20
57
53
37
97
151
158
226
119
20
48
28
45
15
44
46
56
54
53
Units are Kg, where K Is a programmed constant and g is an acceleration at the discrete frequency.
2. Increases in second harmonic of drive shaft vibration when vehicle
operated in third gear and four-wheel drive third gear, 86 versus
150, respectively. Four-wheel drive increases the loading on the
U-joints and this increases the second harmonic output.
Front Drive Shaft
A similar set of conditions exists for the front drive shaft. Of
significance is the much closer coupling of engine firing rate to the
front drive shaft, increasing from 3 to 504.
4-9
aHaiagaittewiaa^Bgwwij^^ £^-Vsrt
From this series of tests, the features were traceahle to the following driveline
assemblies:
Baseline vihration for the following:
Rear drive shaft imhalance
Rear drive shaft U-Joint
Front drive shaft imbalance
Front drive shaft U-joint
Cross-coupled effects included:
Engine imhalance
Cylinder firing rate
U-Joint and drive shaft loading due to four-wheel drive.
4.3.3 6-Qunce Weight on Rear Drive Shaft Test
The test results for 6 ounces clamped to the center of the rear drive shaft are
shown in Table 4-2. While driving the vehicle on the road, there was no significant
increase in vibration that could be detected by the driver.
Changes in measured value as a function of gear and front wheel drive were similar
to the baseline test. There was, however, a significant increase in rear drive
shaft synchronous vibration when operated in third gear. The value was 32 for no
weight added to the rear drive shaft versus 365 for 6 ounces added to the rear
drive shaft. Also, there was cross-coupling to the front drive shaft, increasing
the value to 181.
Rotation of the weight 180 degrees on the» drive shaft changed the value of the
in-phase component from 365 to 449, indicating that the new position was more in-phase
with the unfaulted imbalance of the drive shaft.
4-10
^■wyi^apip...,,,!!!! *'
I I I I I 1 I I 1 !
I I I I I I I I
Table 4-2. ON-THE-ROAD DRIVE SHAFT TEST, 6-OUNCE WEIGHT ON REAR DRIVE SHAFT, 35 HPH
i
Run No.
Rear Drive Shaft Front Drive Shaft
Condition 1st
Ha rmon i c 2nd
Harmonic 1st
Harmonic 2nd
Harmonic
1
| Test No. 1
Third ('.ear
j
1
2
3
388
385
324
48
87
40
188
186
169
11
11
11
Test No. 2
Fourth Gear
1
2
3
467
334
336
227
182
194
309
183
168
141
499
499
j
i Test No. 3
Tliird Gear
| 6-oz Weight
Rotated 180° i
1
2
3
463
474
412
60
5
44
183
173
188
17
12
11
Test No. 4
Fourth Gear
6-oz Weight
Rotated 180°
1
2
457
380
266
185
179
240
499
450
Units are Kg, where K is a programmed constant and JJ is an acceleration at the discrete frequency
4.3.4 6-Qunce Weight on Front Drive Shaft Test
A 6-ounce weight was clamped to the center of the front drive shaft and the vehicle
was operated at 35 mph in third and fourth gear. The results of these measurements
are shown in Table 4-3. Significant changes in measured value occur in the first
harmonic of the front drive shaft measurement when operated in third gear and four-
wheel drive. The value was 31 for no weight versus 280 for a 6-ounce weight on the
front drive shaft. These measurements were taken with the rotation sensor on the
rear drive shaft.
4-11
- - .»j.«»*a»iif<ainji—«umwuifmiTiiw rinn i*in;iiiiii.iri|imii).*>sfcrej.^iM-g,>m.::. ™*«wmmm^mm^&,4i%i&i*iii&i&!t&ä. >*.
«A«. IM» "m-m9ymmnmtm >> •' — 1 m mm i-- ^*W
5*
s
Tablc 4-5. ON-THE-ROAJ) DRIVE SHAFT TEST, 6-OUNCK WEICHT ON FRONT DRIVE SHAFT, 35 MPH
When four-wheel drive is disengaged, there is a lack of good synchronism between
front and rear drive shafts. The low reading for the front drive shaft
the four-wheel drive is not in use.
Oondition
Test No. 1
Third i.e.ir
ve Shaft
2nd Harmoni c
82
64
5 7
72
26 L
310
276
25 5
Run No.
1
■■>
3
4
Ke.-ir I)ri Kronl Or [ve Shaft 1st
Harmonic
32
41
60
24
1st Harmonic_
16
16
04 1 20
2nd Harmonic
07
12
0 3
I 11
Test No. 2
Fourtli Gear
1
■>
3
4
81
88
101
95
34
6 3
40
34
1
565
j 614
1 600
553
Test No. 3
Third Gear
Four-Whee1 Or i ve
— ~ — ■■-- —
1
3
34
34
57
186
147
16 3
83
anil g is an
279
260
301
1
61
60
62
Test No. 4
Third Hear
Check on Test No. 1
, ... ,,„ , .
1 30 ,
23 0 3
Units are Kg, where K discrete frequency
is a programmed constant ecceleration at the
was when
A summary of imbalance drive shaft test results is shown in Table 4-4. An analysis
of the data shows the following:
a. A drive shaft imbalance of 6 ounces is easily detected and as little as
1 or 2 ounces can probably be measured.
4--12
'^MinimMtm^mimmiVammiiitmmtMui»^
Table 4-4. SUMMARY OF IMBALANCED DRIVE SHAFT TEST RESULTS
f jMeasured Value
Condition 3rd ('.ear Ird (.ear t-wneei Drive 4tn t.ear
Weight Shaft 1st 2nd 1 St 2nd 1st 2nd
GD GD CD (TvT) CD CD Basel'ne-- Baseline-- Ratet ine-* In« rease Im rease Increase
Rear l.ood Cood (iood over base- Ilne due to U-)olnt stress in 4-wheel
over haae- 1 lne doe to engine Hull« lam e
over base- 1 ine due to cylInder firing
No weight on drive shaft CD CD
drive
CD GD CD Values low. Rotation Front and rear drive Increase Large In- sensor on rear drive shaft locked together due to crease due shaft. Ur1ve shaIts rotate avmhronlned to engine to cylInder
Front are not synchronized. This It due to small differences in front and rear wheel diameters
front drive shaft Imbalance firing rate
(365) GD CD CD Large In- In base- (This series of tests Show* Shows
Rear crease due to weight on rear drive shaft,
line range was not run) weight on rear drive shaft plua engine
effect of cylInder firing
Six ounces on Bad rear imbalance
rear drive shaft drive shaft ——j
CD CD CD CD I tu rease In base- (This series of test Shows Shows due to 1 lne range was not run) cross- effect of
Front iross- coupl ing from weight on rear drive shaft
coup I ing from rear drive shaft plus engine Imbalance
cyl Inder firing
CD GD GD CD GD CD Rear BaselIne-- Basel lne-- BaselIne-- Shows U- Value high Value high
C.ood rear Cood Good rear Jolnt due to shows drive shaft drive shaft stress due engine cylinder
Six ounces on front drive shaft
to 4-wheel drive
Imbalance firing
CD CD GD GD GD (583j)
Values low. Rotation Value high Basellne-- Value low Value high sensor not synchronized due to Cood. due to lack shows
Front to front drive shaft weight on Slightly of rotation cylinder front drive hinti due to sensor firing shaft. Bad 11- joint synchron* rate drive shaft stress izalIon
4-13
; ._,-,„■:.: ^Mi^'&jiie&ax-.t.'- < r 'ji^mmmmmmmmmamm^m^Ai,-.,:'
in SJSJ "" I' '
h. Hrive shall imbalance should he made with the vehicle operating In third
gear. I'DUILII gear operation (1 to 1 transmission ratio) permits excessive
cross-coupling of engine imbalance that interferes with drive shaft
imbalance and cylinder firinj
measurements. ig cross-couples into drive shaft U-joint
c, Front drive shaft measurements must use the front drive shaft rotation
sensor. (iood results were obtained with the vehicle operating in four-wheel
drive when the two drive shafts were mechanically locked together.
d. Measurement of U-joint source frequencies showed increases in measured
values directly correlating to U-joint loading. This indicated U-joint
faults, if inserted, would be easily detectable.
e. Contributions of engine imbalance were easily measured as indicated by
the fourth gear operation.
Contributions of cylinder firing rat
the fourth gear operation. e were easily measured as indicated by
*• 3.5 iinjjLhe-_Koad Differential Test
The fU51 vehicle was operated at 35 mph on the road under various gear and front
wheel driv<:> conditions while conducting front and rear differential measurements.
The results of these tests are shown in Table 4--5. The driveline analyzer program
was changed to provide readouts for the magnitude of the high bandpass vibration
(High-IJ) and the ratio of the high bandpass to the low bandpass (Ratio). The
results of these tests indicated the following:
a. Operation in third gear shows that the rear differential has a higher high
bandpass output than the front differential. The reason for this is that
the rear differential is operating under load.
Operation in third gear and four-wheel drive greatly increased the rear
differential vibration level. These increases in rear differential vibra-
tion can be attributed to the differences in front/rear wheel circum-
ferences which results in the production of a higher level of internal stress.
4-14
-: ■■ ilMM0KWlHtltlMMi(t
*•«•»«•■«••»«
1 tt*Sp. 3'»».V.' L*^&mmX*.*Ki.^s. XK. *M<- ■
Table 4-5. ON-THE-ROAD DIFFERENTIAL TEST,
GOOD DIFFERENTIAL, 35 ?0'H
I I I I I I I I I r
i i i
Condition Run No.
Rear Differential Front Differential
High-B Ratio
1
1
3
2
2
2
3
High-B Ratio
Test No. 1
Third Gear
i
2
3
4
5
151
294
191
9 3
301
86
84
!i 4
89
87
91
87
81
82
83
>-
Test No. 2
Fourth Gear
1
2
3
4
5
464
394
381
359
358
Test No. 3
Third Gear
Four-Uheel Drive
1
2
3
4
5
551
733
669
593
684
3
3
3
2
3
200
192
190
180
208
Test No. 4
Fourth Gear
Four-J/heel Drive
1
2
3
4
1011
911
666
789
4
4
3
3
188
189
170
168
1
1
1
1
High B: The magnitude of accleration of the programmed high band frequencies. Ratio: The ratio of the high frequency bandpass measurement to the low fre- quency bandpass measurement. Units are Kg, where K is a programmed constant and g is an acceleration at the discrete band of frequencies.
c. Operation in fourth gear causes an increase in rear differential vibration
over the vibration experienced in third gear. Also the bandpass ratio has
increased from 1 to 2 or 3. While operating in fourth gear, the engine is
directly coupled to the drive shaft such that cylinder firing and engine
torque variances are more directly coupled to the rear differential. This
causes an increase in ring gear tooth chatter.
4-15
^i^sffii^^amtMr»mwaiKW6ii*^9imi»Msimi^,A? MB—WWW» mmmmmmmmmmmmmmmmmmmiitsmumms^W^^k^mi ;$$&.
Htm ■: '.Uli... '!■■! 1
(i. "peration in fourth (tear and four-wheel drive lncrea«ed both front and
roar differential outputs as would he expected. Also, the bandpass ratio
for the rear (inferential increased fron 1 to 1 or •'« indicating much more
tooth or bearing chatter.
It should he noted that the above test data was talen on a vehicle that was believed
to have good differentials in both front and rear. The above test data, however,
would indicate that the rear differential was marginally bad. This rear differential
had been removed and adjusted several times during the program and, therefore, could
have been misadjusted during these tests.
These preliminary tests indicate the following:
a. The effect of loading variation indicates that the bandpass ratio technique
used in the driveline analyzer can be used as a diagnostic tool for
differentials.
b. The differential vibration can he separated from the effects of other
assemblies and from the effects of road noise.
c. Differential tests should be run in third gear and four-wheel drive, as the
vibration levels of faulty components are maximized in this mode,
4.3.6 On-the-Koad Test of Wheels
The vehicle was operated on the road in third gear at 35 mph. Weights of 3 ounces
and 20 ounces were added to the wheels as shown in Table 4~o. Analysis of these data
indicated no significant definitive features for weights being added to the wheels.
Some trends <:re indicated, however.
a. Weights added to the right rear wheel did increase the second harmonic
component. It did not significantly increase the fundamental.
b. Weight added the right front wheel decreased the second harmonic
component associated with that wheel. This may be an indication that the
wheel was previously seriously out of balance and that the addition of
weight tended to balance the wheel.
4-16 * I I
I I 1
Table 4-6. ON-THE-ROAl) WHEEL TEST, THIRD CHAR, 35 UPH, FOUR LEGS ON EACH TEST
I I I
Condition
Rear Wheels Front Wheels
Right Wheel Left Wheel Right
First
Wheel
Second
Left Wheel
First Second First Second First Second
8-oz Ueight 12 248 34 84 56 122 19 205
Right Rear 73 139 31 21 70 239 35 158
63 162 28 57 87 59 21 31
j 20-oz Ueight 21 211 25 60 96 20 35 71
Right Rear 31 200 10 53 45 323 19 203
41 270 46 33 24 189 12 295
18 239 18 41 42 228 22 80
Ho Weight 12 133 40 30 bl 100 31 59
35 42 22 37 56 150 35 98
42 21 43 47 48 96 43 70
20-oz Weight 30 89 23 121 58 81 49 76
Right Front 28 62 29 144 18 83 23 157
58 89 34 155 24 74 17 134
First - A measurement value of the vibration in s^ /nchroniz ism with the funda- mental frequency of wheel rotation. Second - A measurement value of the vibration twi< :e the wh eel rotation frequency. Units are Kg, where K is a programmed constant an< i g is an accelers tlin at the discrete band of frequencies.
c. First harmonic measurements did not vary significantly with addition of
weight, and the measured values were all very low, indicating a need to
increase the sensitivity of the fundamental feature extractor for wheels.
Additional test of wheel imbalance showed careful balancing of wheels should be done
before addition of test weights, and the program should be modified to increase the
gain of the fundamental extraction feature.
4-17
.tmimmmtämmmmmimeitxsxmMwmim
4.4 SUMMARY OF TKST RKSW.TS
The tests previouoly described showed the following:
a. The driveline analyzer was capable of conducting all tests set forth in
the design objectives as demonstrated by the bench test.
b. The driveline analyzer was compatible for use on the M151 vehicle and
can be installed by one man in less than 5 minutes.
c. <>n-the-road tests conclusively demonstrated the ability of the driveline
analyzer to detect the following assemblies:
Front drive shaft imbalance
Rear drive shaft imbalance
Front drive shaft U-joint
Rear drive shaft U-joint
Fngine imbalance
Engine vibration
Cylinder firing
Front differential vibration
Rear differential vibration.
Preliminary tests indicated the probable success of wheel imbalance and wheel U-joint
vibration; however, additional control of faults and adjustment is required.
The bench test indicated that the «slipping clutch test will work on the vehicle,
but limitations on available time prevented conducting this test.
The TACOM test and evaluation should include a program to define how the driveline
assemblies fail and should include provisions for controlling the inserted driveline
faults.
4-18
^*&*&flH&^ii^ S»*WK«WM»WiftW«^Mfcfe>frT^iK .^
h SECTION r>
SUMMARY
The scope of work specified that Northrop furnish the supplies and services necessary
to design and fabricate, for TACOM test and evaluation, a driveline analyzer designed
to detect faulty driveline assemblies in the Mlr)l stries 1/4-ton trucks. This
effort has been completed and the driveline analyzer has been delivered to TACOM.
Compliance to the hardware requirements is shown in Table 5-1. The procedures used
to define and improve the test techniques utilized for the analyzer design involved
both analytical and empirical approaches. Comparison of the theoretical vibration
properties of the driveline assemblies to the recorded features from actual baseline
and a simulated faulty M151 vehicle led to the validation of the selection of the
following test techniques:
I I I
a. Discrete synchronous frequency analysis for wheels, axles, U-joi ts, and
driveshafts
b. Bandpass ratio techniques for differential, transmission, and engine
c. Engine shaft rate versus slipping clutch.
The test techniques selected established the need for the quantity and types of
sensors, the specific signal conditioning and processing, and the various portions
of the vehicle's drivelines to be operated under controlled conditions.
Figure 5-1 depicts the resultant block diagram for the driveline analyzer which is an
integrated series of tests that provide a tool for automatically diagnosing the
vibration source in the vehicle as it is driven on the road.
The large number of data points projected for the test techniques, the accuracy
required, the number of constants, and the number of calculations required clearly
indicated a requirement for some form of digital processing and control systems.
Since the program is a feasibility investigation, test techniques are likely to
5-1
i-.xmm' *^^^*iJW3&fe»w™^***^^ - ?^l;-.■^,J.v.l^^^■-:;^-5■v^^^-=^:.:l■.
TabJe 5-1.
Requi rement
Dimensions:
]2 x 18 x 18 in.
Insensitive to ve- hicle vibration
Powered by vehicle electrical system
Simple installa- tion by one mechanic
Operate by one mechanic
Determine If oper- ating properly and
Autotest with re- sults displayed to indie :e Hood/Bad
COMPLIANCE TO HARDWARE SPECIFICATION
jCompl ic -ompliance
Yes
Yes
res
Yes
Yes
Yes
Yes
Logic/Control, 12 x 14 x 17 in. Operator Display, 6 x 7-1/7 x 2 in.
Logic/Control, 40 lb;
Senso.- Assemblies, 11 lb (total); Operator Display, 2 lb
Designed to operate on +25.2 + 3V DC (battery cell voltage range); current required, 3 amps
Logic/Control, display, and 3 sen- son assemblies with 5 cables
j One button to in it tat • selected tesü
Exercises internal logic and verifies circuits prior to each test
Test is automatial.lv sequeneed; display of faults is programmed to be shown in order of significance
change. This established a need for a stored program rather than a hard-wired
approach. Requirements for displaying some 30 test results were identified and
indicated the requirement for a small alphanumeric display. The alpha portion of
the display was required to provide name identification for the large number of
tests; numerics were required to indicate degree of badness and order of fault.
The operational modes involved in performing the Lests require operating the vehicle
in a specified gear, at a specified speed, along a prescribed path. Because these
tests are conducted on the road, traffic and other considerations will cause inter-
ruptions in the test sequence. Provisions were made to provide the vehicle operator
with indi alors that instruct him concerning vehicle gear, engine rpm, vehicle
speed, and the identification of the test being conducted, together with indicationr.
of whether or not the operator is meeting the required operational conditions. The
types of Information that are presented to the vehicle operator are:
• Engine rpm for engine and clutch tests
• Transmission gear for specific test modes
5-2
-V K * -
f*W
ROTATION SENSORS
«COUNT!« TIMER
CHANNELS
PHASE LOCKED LOOPS
CENTRAL PROCESSOR
MULTIFKXEM
AC« L EROMETERS
SWITCHES SIGNAL
CONDITIONERS
TEST RESULTS DISPLAY
AND CONTROL
VEHICLE OPERATIONAL
OISPLAV
ENGINEERING OISPLAV
Figure 5-1. DRIVELINE ANALYZER BLOCK DIAGRAM
Speed, in mph, for clutch, transmission, differential, and wheel tests
Engine rpm
Speed, in mph
Waiting for conditions to be met
Invali". test
Test complete.
The test and evaluation portion of the driveline analyzer program involved three
phases:
a. Bench test of the computer program sequences and measurements of simulated
transducer signals
b. Bench test of the diagnostic program sequences, using magnetic tape
recordings of transducer signals from on-the-road tests
5-3
t^nmmmnmmmmimmmm NMMS (ßÖBäfe-
l"*t>,r'f*i'»»«'r« '"'* I'm'»« ■ M i»111.
c. Road tests to verify the ability of the test set to measure vehicle
operational parameters and to display them to the vehicle operator.
Table 5-2 displays, in summary form, the results of this hardware and software
verification effort.
The resultant hardware developed during this program allows for flexibility In vary-
ing the parameters of the individual diagnosti techniques. Instructions to the
operator for conduct of the tests, the gains and bandwidths of the sensor signals,
the cross correlation of selected sensor signals, and the resultant decision and
display of fault indications are programmable to allow for selection variation of
each parameter. The effort resulted in the development of an effective engineering
tool nat can be utilized to closely examine the feasibility of using vibration
signatures as a method for Army vehicle driveline diagnosis.
1 3 1
t
5-4 I 4
^a-jw*re!w»u*ytfiWiiw-'^^^ *«&*'* ■*&*&&**&*■ - -,*:Ä««Jl(«IS-<*
I
X t9 H
C 01
i
(A 1 OJ T3 0 1 1 *-i C ifl > 19 x; >" X 01 00 T3 ~H c 01 Ui 0 0 1 •-H O 0 O If; C <H 1 C
01 0 X 19 X U 01 0 01 u 0 —< 19 TJ 19 CJ 'H • a 19 ta u 4J 01 Q. 1- ~( — C X 4J >.u m C 4J Ü 19 <H u lj u Ul (S h n it » 0 JZ c T3 C TJ 0 C X X TJ
f—1 01 41 01 *J 0 01 (9 01 >. k4 01 01 01 3 > h U u T3 i-( u 3 -• 4J 4-i 4-1 li 9 41 4-t • E 01 X c oi :* • 0 19 19 —i C 01 c 19
TJ •*-« V) <-) 01 19 •H oo X X Ü 3 0 14-1 19 t-l u 95 MH X •H 3 X ^M <^t -—( • c x e ft u U u-i Oi f«
M. 1-4 —4 19 •P4 «M •O i-> M-4 T3 ~< f< x x TJ u <9 X x -o X ki 01 ig 4J u r, <4- TJ T3 60 4J C (9 > u UIM Ü E Hi S o in 19 01 '■4 U4 C C l- ••H E c (9 O TJ *M > l-i 01 01 u •-< 19 X -H 3 o X Hi 0 -H T3 4J 60 3 4J Q ig u u 4J X
OC 1* JO U X > 4-1 u> oi c (A Hi 60 4J 9) X 1! a-- o iw 4-1 ^ 0 T3 X 19 01 c 01 0» 01 oi oi 0) 6C a. nj k. 19 u u 01 4-1 01 4-t • «H B i« U 4J u a c (9 r a. X <u 0 4J U u e X 4-i 19 ft u a •H M. cr. <« > « in oi 3 >. X (9 X C 3 • >, 3 B IT. U ••* U Uj 4j 01 T3 0 iH 14 1IH -4 a" 4-t U X
3 •H tf) (9 0) Oi 01 (9 «J c 14-1 <H 19 U 3 01 19 01 C 19 X 4-1 0) > 3 T3 1-* -H CUE 3 00 u 01 C 01 u m lA 0 60 •H 00 CT •a x; c > •■-I 3 •* er E •^ ^H (9 u 0 4J Wi C -H T3 01 •^ Oi B x x <n •H rH X U E X 0 1—1 C TJ D •H C 01 *J 4-1 01 C ■M 4J C 19 X C iH 19 k- 3 RO 01 U C ■a £ ts •H ■-» 19 3 -H H Q X H X C *J ft <-< X a x m > -H o Oi It U k E 3 4-i cr 00 Oi —i 4J 3 U 60 "0 <4-i 01 0 a. a» •H -< £ t(
Q 4J H 4-1 01 0 01 0 •H 19 it «< C W X ft 19 Oi •-< ft 01 IM WI < u Vi *J H U J «4-( O C 0) Ü4 flj J u-i H W5 3 U 0- D.
■a 4-i
<9 0
r-t > > 06 19
c 0
(V.
01 C
13 •H r-4 4) > > > > > >
X VI <u 19 H CO
u 19
3 > > > > •*«. / > >
ij eg 19
TJ
U.
t) -O 01 c 0) 13 ■H
(C 0 l-l
01 "> > > > > > > Hi u m H 3J m
as to
«o 4-1
4J H 3 > > > > > > > > >
4j e n)
U -a it
U.
01 Tf *J c o « 'f-1
X 3 0) > > > > > > > > > v e X H ft «
w CO
u »f-t
u E tfl 19 C W > > > > "> > > > > OC 0 m w>
•H a. ä 1 X <L) / u U X / 4-i C QJ 4J / 3 i-l fH (9 rH c 3 / ■H u w C (9 .H X •H O i-i / CJ IM u-i a»
19 B •H 19 < 0 •H 14-1 / 19 <u *J C XI c n c X C 19 C / 0© X 0 r c R 0 B «J o 1 0 oi o X O / n W C (/) 00 0» -H M C •H 3 ft •H -H W 14 'rl / •r-l tB •H l-l 4J 0) 01 4-1 44 E u C 4J / a. <D i-l 01 ^H 01 19 IH CQ C 19 rH 19 W fl) OJ -H 19 / a > rt > 19 U-4 IH V •H 1-c 01 b C M > O t-i / •H £•§ ■H (fi <4-l XI 0» 13 oox 01 X 19 X •H i-J X
/ 1 H U -H •H -H 5 (9
C -H gs U -H t-l 1 -r-t ( \ W Q M Q X Q > Ul > H > a a >
5-5
,-v.:^i't l:«;^^«M^^^^^--ite^'3^^«^
PRBpMDtm
I I I I I I I I I
r
i^fcÜLm, FQMm
SECTION 6
CONCLUSIONS
1. The definitizing of qualitative faults in terms of quantitative mechanical
deviations proved difficult and requires more study by the Army (para-
graphs 2.4 and 4.3.3)
2. Simulation of fualts on the vehicle proved difficult to define and control
(paragraphs 2.4 and 4.3.3).
3. The effects of a fault in a defective assembly can cross-couple to a good
assembly, causing the good assembly to appear fault (paragraphs 2.6.2, 2.7, and 2.8).
4. Spectral lines synchronized to a shaft rotation exhibits very good signatures
of mechanical parts (paragraphs 2.6.3 and 3.8.1).
5, Discrete frequency techniques are useful to identify specific parts, including
wheels, drive shafts, U-joints, gears, bearing races, and ball bearings (para-
graphs 2.7, 2.8, and 2.8.1).
6. Discrete f
en screte frequency techniques are not practical for the transmission or differ-
tial assemblies due to their large quantities of frequencies requiring neas-
urement and comparison (paragraphs 2.7 and 2 8)
7. Ratio of high- to low-frequency bands did exhibit an increase when misadjustment
or fault was involved in the differential (paragraphs 2.7 and 2.8).
8. A specially designed CMOS computer was the optimum unit to process and control
the driveline analyzer (paragraph 3.2.1).
9. The driveline analyzer performs in the manner for which it has be
and programmed (paragraph 4.1). en designed
6-1
'-'-'»WW'IWtflWLJttfiB.-
~W.»-
. f&ti&ys* , $J% sv-y^^-y-, ■■ »gsKro-sjä*^!^.,
■ ■ ög^(»3*^i*JM(» «*"!*; •■'■ »jWp.' iMti^V;i-'
10. The driveline analyzer is compatible wiili the physical and operability require-
ments of the M151A2 vehicle (paragraph 4.2).
11. The driveline analyzer is capable of performing its design and programmed
functions on the M151A2 vehicle (paragraphs 4.3, 4.3.1, 4.3.2, 4.3.3, 4.3.4,
4.3.5, and 4.3.6).
12. The diagnostic programs of the analyzer require more test and refinement, with
the following controlled variations (Section 4):
• Variations in road conditions
• Variations in driving patterns, including speed variations and driver
characteristics variations
• Variations in degrees and types of faults
• Variations in vehicle conditi on
• Variations in degrees of multiple faults.
6-2
-^ ---- ~i\\yvr •in-'--^-"]-*-'—*-■"—rt-—--—-■---■— -—-■- -- rirniwMii rijiiMiinnpnimi (MÄ£Bä$95W!-j*fc»i; ,,■„.: ...... -, .
SFXT10N 7
RECOMMENDATIONS
Three actions are recommended based on the results of this program:
a. Refine the methods for simulating faults in a measurable controlled
manner (Section 6, paragraphs 1, 2, and 3).
b. Conduct a test program to gain data related to the fault detectability of
the analyzer's test techniques (Section 6, paragraphs 5, 7, 9, 10, 11,
and 12).
c. Develop an improved method for sensing the rotation of the drivelines and
engine (Section 6, paragraph 12).
7.1 FAULT SIMULATION
The simulation of field-faulted assemblies requires extreme care to verify that the
faulted condition results in vibration that is truly representative of field failures,
The Army should identify a set of real-world driveline failures that occur in the
field and develop controlled w?.ys of inserting these faults into the test vehicles.
Levels of unacceptable vibration should be separately determinable as a function of
the degree of fault.
7.2 TEST PROGRAM
The driveline analyzer should be evaluated using these controlled faults singularly
and in combination. Analysis should be made of these results to evaluate how future
extraction, cross-coupling, weighting, and limit parameters should be modified to
improve the test capability and repeatability of the driveline analyzer's measure-
ments. This must be done not only for the M151A2 vehicle but for any other vehicle
to which the Army intends to apply the driveline analyzer.
7-1
mmtmmimim
»L— 1 '.." .,UH,g.M!>Wll^UftW
.1 Once effectiveness has been established for the M151A2 vehicle, the analyzer should
then be reprogrammed and tested with other vehicles. Having done this, the Army can
determine whether or not such a test set would be truly effective in the depot
environment.
The empirical results of this program and the preliminary evaluation of the drive-
line analyzer have shown the ability to measure vibrations well below levels that
would result in maintenance action. These vibrations are traceable to specific
assemblies and components. With additional testing, the Army may find the driveline
analyzer effective as a prognostic tool.
7.3 ROTATION SENSORS
The rotation sensors that were utilized in this application proved very effective
in the laboratory and during most of the empirical testing. There were road test
conditions that did result in misleading test results. These have been traced to
be the effects of early morning or late evening sunlight shining directly on the
sensor. The sensors are very workable, as this condition is realized by the user,
but the sensing approach on the sensor design should be improved before a prototype
phase is undertaken.
7-2
^BiMCIWIIWIWMTi ""~™ ■^m^ii!^m^m^>vm^sm^^^'^'^^^-^tV «-* #*r ^, ,,., .-.
■ ipi'M^», ■■■■'> niijuj , | i HI U.I».«I w '■■". mmf*mmm\itptiw "'<"•<""'* '" "'"■
SECTION 8
DISTRIBUTION LIST
ADDRESSEE
NO. OF COPIES
Commander U.S. Army Tank-Automotive Command Warren, MI 48090 ATTN:
Research, Development and Engineering Directorate, AMSTA-R Chief Scientist, AMSTA-CL Propulsion Systems Division, AMSTA-RG Engineering Science Division, AMSTA-RH Armor, Materials and Components Division, AMSTA-RK Systems Development Division, AMSTA-RE Maintenance Directorate, AMSTA-M Product Assurance Directorate, AMSTA-Q MICV Project Manager, AMCPM-MCV XM1 Project Manager, AMCPM-GCM Technology Library Branch, AMSTA-RWL
1 1
22 1 1 1 1 1 1 1 I
Commander U.S. Army Material Command ATTN: AMCRD-GV
AMCRD-RD Washington, D.C. 20315
Office, Chief of Research and Development Department of the Army ATTN: CRDCM Washington, D.C. 20310
Commander Defense Documentation Center ATTN: DDC-IR Cameron Station Alexandria, Virginia 22314 12
I 5-1
■EAD IHtTWICnOlt*^
-Mtotliifi WKSo NIW
i«. TlTtt fart MWK)
DRIVELINE ANALYZER M151 1/4 TON UTILITY TRUCK
AUTHOUT*)
A.D. DeROLT P.J. LEIBERT
Northrop Corp., Electro-Mechanical Division 500 East Orangethorpe Avenue ,/
I Anaheim California 92801 lt. COMTMH.LIMO orriCt KAMI MO AOOMM
U.S. Army Tank Automotive Command Diagnostic Equip., AMSTA-RGD Warren, Michigan 48090
11. mi OP RtNRT * PIWOO covtnto
FINAL Nt»ORT NUMM
AMT NUMOtnräT
DAAE07-73-C-0116 fAt'
It. *MPO«*m"<j 75Y081
"SfHo%?5s?2SS2 •'""
II. IW^O«TO»Tt
10 h.lv 1975
11NCLASSIFIF" , ■ ^4WAMI#ICATIOH/OOW«0«AD1II<»
CMlOUU«
nSiSliSS (TATKMtMT f3 M« SSSOTQ*
19 DEC " b»iT«»U»»CH« |IATtM«H^ C»' —• —^- '
distribution limited to U.S.£T^*g^Z''SS& Other requests must be referred to.
hf-bHTWOUTION IT»T«W <»' 55 SS555 www S m—k n,» 5555« i
lit. »UP»V.CMENTA*V HOTtt
I» "«V «OHO» fC-KI— -
Analyzer Diagnostics Monitoring Testing mhration —
It». AMTHACI (C««!*-««"
• «I* if MMMa» 25 MSNTS »r SS3)
Driveline Synchronous frequency analysis Bandpass ratio analysis
The objectives of this development program were to determine the feasibility of using data obtained from measuring discrete selected parts of vibration signatures for Army vehicle driveline diagnosis and to provide the U.S. Army Tank Automotive Command (TAC0M) with an advanced development model driveline analyzer designed to identify faulty driveline assemblies of an Army K151A2 1/4-ton utility truck. The development program performed by Northrop included
DO t'tTn 1«73 tomo-oriNOvMiiowoutTE
,.1>.iiE^j*u*"r.'»ft&:-.^
^-JAWSWÄ» iffis«H»*,:*1-s*M'"*"";
, , „, , ,*- '■'•-■•
«ECU Kl TV CL»SHriC«TIQN Of THI» PAQUWhmt Out» ggtggg
20. ABSTRACT
the correlation of simplified analytical models to empirical data, and the design, fabrication, and functional test of one portable driveline analyzer test set complete with sensors and a remote display console.
Analytical models to determine how driveline components and assemblies Ren- erate excessive vibration were developed and used to identify vibration features that were potentially good indications of faulty components. Follow- ing this initial effort, the driveline assemblies of an Ml 51 vehicle were instrumented for vibration and shaft rotation data. During road test, the data was recorded on magnetic tape and was analyzed on Northrop computers to develop power spectral density plots. These plots were used to provide empirically derived excessive vibration features of good and bad assemblies.
The analytically derived vibration features then were correlated to the empirically derived features to identify errors in cither of the processes. Good correlation of analytical and empirical data was used as criteria for selecting features that would provide the highest probability of detecting faulty driveline assemblies and the lowest probability of identifying a good assembly as being bad. From this analysis, vibration features were selected that were compatible with a simplified road test and that were practical for this application.
A portable test set was designed and constructed, consisting of a driveline analyzer chassis, a remote display console, and a set of sensor assemblies. 'Hie driveline analyzer chassis contains amplifiers, programmable filters, counters, phase-locked loop circuits, and a digital computer with both random access and read-only memories. The remote display console contains vehicle operator instruction displays.
Digital computer programs were prepared for use with the driveline analyzer test st!t to provide the vehicle operator instructions, gather the vibration data during the vehicle road test, analyze the data, and display test results in terms of faulty driveline assemblies.
The driveline analyzer set was tested utilizing simulated t»st data and data recorded on magnetic tape. Sensor installation compatibility, test sequence, and vehicle control data acquisition, processing, and display have been verified on the M151 vehicle.
The test set was demonstrated to TACOM personnel using simulated and recorded data in the laboratory. Installation of the test set on an M151 vehicle was also demonstrated. After this demonstration to TACOM, Northrop conducted a number of road tests which demonstrated the ability of the driveline analyzer to detect drive shaft, U-joint, engine vibration, and differential types of faults.
This contract was conceived to assist TACOM in their goal of advancing the technology in the field of Army vehicle diagnostics. The M151 driveline analyzer feasibility model has been delivered to TACOM for test and evaluation. The driveline analyzer has verified many of the diagnostic concepts; however, further testing and engineering improvements are necessary to determine the adequacy of the diagnostic concepts.
KCUNITY CLAMiriCATION O* THIS PAO€(Whit Data Bnfrtd)
HfHSffHi'yf/t^mm w» .^'WWWWM