7/18/2019 Ruido y error.pdf
http://slidepdf.com/reader/full/ruido-y-errorpdf 1/10
BACK TO B,ASICS...
R~E ~~ C~
R. E. Sm Ith • Co., Inc.
Roc:hester , New York
The
Gleason
Work s
Roche s t e r ,
New
York
F lg .. 1 1 1
-. ...
'1, 2
pU lftI_
Z a
r
,
- , - I
Ill-
40 m m - ~
on:
r n n t A w L . ' . . • ,001 or 1
0 1 1 > _
i : ~ . . : . : . . . .
J
40 m m 4 AlAe MINUTES
Rg.lb
Gear
Tect l i lO logy
~I
AbItnct
Gea r noise is not just a 8W p rob l em. It is
a system
prob lem and th e
gears are the ex-
citers o f th e system. T ra ns m ia sio n e no rs in
th e gears due to geometric: v ar ia tio n s a re the
sourt:e
o f
the
eKCilat ion .
This
aJtide
drM
with
t he ins trumen ted test ing of nolle in th e aystem
and how to
s u c c : e s s f u I l y
rela te th e r e su l ts b ad e .
to the mea su r ed tr an sm is sio n e rr or s in
th e
sean. Th e
bigest dHficuJty in
es tab l i shing
th e
re latio ns hip is h ow to c ompa r e visua l ly dis-
p layed aound ana lysis
data
to anj udsmen t s
of
the
hu m an r ate r, and t h e n rela te th e r e su l t s
back to m eas u re d tran sm ill iio n e r ro rs o f th e
gears. This is r e f e r r ed
to as the
missing
l inkH
in th e
article.
Introduction
Vehicle gear noise
test ing
is a complex
a n d often misunde rs t ood subject. Gea r
noile is reaDy a system prob l em.a> Mo s t
gearing
us ed
for power transmission is
enc lo s ed in a h o us in g and, t h e r e f o r e , lit-
tle or no audible sound is actuaUy heard
from th e gear pair.
(2)
The vibrations
creat ed
by
th e
gears
a re a m pl if ie d
by
r e s onance s of s t ruc tural elements. This
amplification occu rs when the speed of
th e
gear se t
is
such
th at th e
meshing:
f r e -
quency or a multiple o f it is e q u a l to a
natural f r equ ency o f th e system
in
which
th e g e ars are
mounted.
In order to make quiet g e a r s , there
must be
feedback
from vehicle testing to
gear
manufacturing. Unfo r t una t e l y ,
this
is whe r e a missing l ink usua l ly occurs.
Much o f the vehic le testing is done in a
subjective
mann e r ,
which
is
of
little
help
to th e g e ar m ake r as far a s k no w in g what
e l em en t of gear quality to improve to
reduce noise.
If
instrumentation
is
u s e d ,
such testing is often done by noise
a n d
7/18/2019 Ruido y error.pdf
http://slidepdf.com/reader/full/ruido-y-errorpdf 2/10
Noise To easured ear
vibration people who know little about
gear geometry.
Thei r
data may show lots
of
peak
vibrations that bear no relation-
ship to what the human ear is hearing
relative to gear quality. Many times the
excitation may even be coming from
another s o u r c e . such as a rolling element
in a bearing.
F r e q u e n c y
is the key to understanding
gear noise. The problem is how to con-
vert aural images to visual ones; in other
words. how to find a way to display on
a CRT or on paper what the human ear
is hearing.
Gears As
Exciters
of Structures
Most of what we hear as gear noise
does not come directly from
th e
gear pair
at all. The minute vibrations created by
the gears as they move
th rough
mesh are
amplified by r e s onanc e s in the structural
elements of
th e
housing. such as panels.
ribs, beams, etc. Gear noise in a vehicle
generally occurs at tooth mesh frequency
or harmonics of it and occasionaUy at
sidebands of mesh frequency. The gears
are
th e
exciter. The structure resonates,
amplifies the force and converts it to
airborne noise. Transmission error or
non-uniform motion of th e gears is the
mechanism of excitation. Transmission
error(3) is the result of small variations
from
th e
correct or ideal tooth geometry,
as well as runout. and can be measured
in the production process by the use of
s ing le f l ank composite testing techniques.
In the case of vehicle noise, these varia-
tions can be as small as 25 to 150 micro-
inches. depending on the frequency.
Transmission Error Testing
Transmission error testing is done by
rolling gears together with backlash and
at
their
proper operating center distance.
T he input and output angular motion
characteristics are measured by an en-
coder system and electronics as shown
in Fig. la. The data can be presented
S 'ID , E B A N D S
Pinion Run o u t
Bol t
Hole
I
1 st H a rm o nic ,o f M e sI1
//200 (Ghost.)
--...... .. 2 .6 (Gho s t)
1 1
5
4
HARMONICS OF MESH
Fig. 2
in graphic analog form, as shown in
Fig. lb. or can be further processed by
Fast Fourier Transform (FFT) techniques
to aid in pinpointing th e s ou rc e o f excita-
tion. The Fourier technique
take s
an
analog signal and breaks it down into a
spectrum of
f r e quenc i e s
that relate to
the
various elements of gear tooth geometry
present in the transmission error.
(See
Fig. 2.) Fig. 3 shows how the geometry
of a gear tooth (3a), including flats or
waviness. rolls with a master and gen-
erates the transmission error curve (3b).
This tooth to tooth transmission error
curve would be further ana lyzed by FFT
techniques to yield a spectrum as shown
in Fig. 3c. The general tooth form or
deviations from conjugate tooth shape
give rise to the integer harmonics in the
spectrum; waviness or flats relate to the
ghost harmonics; and the long tenn
errors (runout) cause the s idebands of
tooth mesh frequency. S e e References 1
AlJIHOR:
MR. ROBERT E. SMITH lu is 0 f1flr thirty
y ea rs e xp e ri en c e
i
th e
g l f lF ' industry.
He
is
c urrently presid mt of
R . E .
Smith
a n d
Com-
pany.
a con sultin g firm in RoclwstB .
New
Y ork . H e received h is fnlining
f r o m
Roclwstn-
I ns titu te o f T ec hn olo gy . Whik employed by
The G leason W orks. h is eng inRmng auign-
m e nts inc lu de d g lfl F' methods. r tVllIu f r rc tu r ing.
research a n d gear quality. T I w s e QSSignmen t s
i nvo lv ed t he u se
an d appliaztion
o f inst 'runvn-
tatio for the study of noiM, uibnI t io
an d
str uc tu ra l d yr um tic s. F rom t~ taignments ,
he expanded his Ukas r elm in g to gNr
m m o l -
ogy.
C urre ntly , M r. S mith
is
chaimum of the
Meas urin g M e th od s
a n d Pract ices
an d
Master
Gear SubcommiHee, AGMA. a n d is also a
member
o f the
Ro ch e st er l ru Iu st rW I E n g m « r -
in g S oc ie ty
nd th Society
o f E x pe rimen ta l
S t r E s s Analysis.
January February
1988
7/18/2019 Ruido y error.pdf
http://slidepdf.com/reader/full/ruido-y-errorpdf 3/10
T O O T H F O R M
F1a.3a
and 3 Ior further discussion .o f this
subject.
The Missmg Link
After the gears have be en measured by
single flank composite testing and all of
the possible excitation sources have been
identified,
it
is necessary to,put the gears
in
the final application and
mea su r e
the
results. This is where the missing link
usualJIy is obvious. The missing link
is
frequency discrimination .
(S ee F ig .
4.)
Many facilities at this point only use sub-
iective rating techniques. A hu man. rater
udges the level of gear noise ona scale
f one to 'ten. Ten is inaudible; five is
orderline acceptable, and four is,reject,
nfortunately, the rater may be rating on
irst harmonic of mesh in one test and
econd harmonic or a ghost harmonic in
nother test ..He or she may even
by
lis-
ening to noise as excited
by
bearings
ather than gears, even though the noise
ounds like the gears ..
The problem is that the vehicle is gen-
rating noise at many frequencies due to
oad and tire conditions, wind, engine,
ransmission components, final drive
ears, etc. This masking noise or ambient
oise level (Fig. 5) is several d B , higher
han the gear noise the human ear is
orting out and finding objectionable ..Zl
F FT S P E CT RU M FI a I s
/ (Ghost
Harmonics )
1
2 3 4
H A RM O N I C S O F M E S H
Fis ,3c
Even if instrumentation is used to
evaluate the vehicle test, the displays or
hard copy of data may show many peaks
or frequencies unrelated to what the sub-
jective rater is hearing. The challenge is
T R A N S MI S SI O N E RR O R C U R V E
FIg.3b
I PITCH
T OOTH f t f S t U
I PITCH
T OOT H E 8H)
0 Gear Technology
F ig . 4
to be able to sort out the
important f r e -
quencies and relate visual data to aural
observations.
Types of Vehide Sound Testing
Two basic types of vehicle noise testing
will be
dis cu ss ed h er e. T ype tfrequency
disc.nmination, involves identifying the
problem - the missing l ink re fe r ru l . to
above. Type II involves comparing the
relative noise levels of various gear boxes
as traditionally done 'Withtracking filters ..
Type 1 Noise Testing.
Thjs is 'the most important first step'
to vehicle noise testing. As mentioned
above, fllequency is the key to under-
standing gear noise. The purpose o E Type
I
noise testing is to properly use fre-
quency discri_mination to' relate what the
human ear is hearing to what the eye sees
in graphic data related to gear transmis-
7/18/2019 Ruido y error.pdf
http://slidepdf.com/reader/full/ruido-y-errorpdf 4/10
OMPOSITE OE R NOISE
a · S
sion 'errors. The necessary test equipment
is pictured in. Fig. 6.
Test sqyipmenb
1. Real
t ime
spectrum analyzer (RTA)
with bubble memory.
2. Sound pressure level meter.
3. 12VD C to
no
VAC Inverter compat-
ible
with microprocessor equipment.
4. Tachometer pulse generator and
readout atta.ched todriV'e shaft.
(optional)
5. Tape recorder. (eptional)
6. Digital plotter for hard copy of RTA
data. (optional)
T es t P ro e ed ur er
1 Calculate
an d
plot
ahead
of 'ti,methe
drive s hM t R PM vs .. g e ar m esh fe e -
,quendes at
al l
vehide'
s pe ed s. T his
in-
fannalion can be used , along with the
tachometer readout in the vehide to
identify mesh frequencies at anytime
during th test. (See Fig.7a.)
2.
Insta l l
the ,equipment listed above, 'ex-
eept for the digital plotter,
in .
the vehi-
cle under test
B e
sure to choose a gear
box
that
is
a d e fin ite
reject or noisy
example. It may
be
neeessarytorun
several tests to find the ,optimum Ioca-
tion for the miaophone.
3. Make a Itest drive with the vehicle
piloted
by
a qualified subjective rater
and an instrument It,echni.cianto
operate the RTAandadjust
it
to op-
timize the display to relate to the sub-
jective rater.
4. Se t the analyzer to operate ~n Mode
Io detect the peak noise frequencies
at different vehicle speeds
er
Mode
II
to detect
th e
resonant frequencies of
the structure.
S. Mode I setup-Peak noise
mquencies
Sound Pressure
Level Meter
RTA Frequency
Range
Voltage Scale
Y
Axis Oog )
Input
R M S exponential
A Weighted
o
to .2000 Hz (1)
o to 1 0 volts R M S
A Weighted (2)
averaging (3 or 4 averages)
CQr50r identifying peak ampli-
tude
NOTES:
(a) .Frequency range will vary accord-
i n s :
to application.
(b)
This results
in
an unconventional
double A weighting. This is in
order to enhance data
in
relation
to what the ear is hearing and
could
be
done
by
the use of other
high and low pass .filters.
6. Operate the
vehicle
,at a speed and
condition that allows the
driver
to
hear gear noise. Then change vehicle
speed and load conditions up and
down through this noise period. The
noise should change in amplitude or
come and ge, While Ithe driver is
doing this, the instrumentation
operater can. observe whi.ch peak in
'the frequency spectrum is varying in
relation to what the driver is
he.ar ing,
Th e C 1 L r s o r o n t h e S C l 1 e e n wil1 indicate
the fr1 quen.cy of in'terest. The 1 -
quency and RPM should
be
wnt'ten
In a. test log. Tape recordings can be
made, and the data 'on the RTA
screen can best.ored iin bubble
memory for plotting, at a later time.
The
vehicle
can theabe driven. at any
other sp e ed of interest ,and the test
procedure repeated.
7.
I f
desired, run the test ina. slJghtly diJ-
fe:rent
manne r .
The
vehicle can be
accelerated or decelerated throughout
a t
broad sp e ed range. The RTA screen
is observed throu,ghout the test. Th
cursor w iD automatically shift loth
highest noise level at any given
peed .
The frequency of the curser location
would
be
noted, on the f1y and may
peak at one, two or several frequen-
cies throughout the speed range.
S. Mode II setup - Structural
resonances
The se tu p fo r M ode
n ,operation is
the same as for Mode I, ,except for
the
cursor
operation.
The
'cursor
should now be set fo r peak hold
mode. The test would
be
run as
described
in Step
7. At the end of
an aceeleratlon all'deceleration run,
the display is held (pause) and
stored in bubble
memozy
or ana-
lyzed 1 0 · 1 1 '
peak
frequencies. The
As. 6
.Januarv/~ru.ary
1988 4
7/18/2019 Ruido y error.pdf
http://slidepdf.com/reader/full/ruido-y-errorpdf 5/10
I
I
_ i
• • • • M
t fa •
C IR CLE A -1 4 O N
IR 1 A J D lR IR E f lY
C A R D
1 0
,
Vrms
10
dB
I~
r A V
J ~
I
I
I
1
I
mVrms
III
S t a v t :
a
H z
x: 386Hz
t tt
10
Ytma
1 0
e m
fDy
1
~~~ ~
x:eHz
fIa. 7 c
peaks in. this display represent the
structural. resonant frequencies.
These frequencies could
be
excited
at any vehicle speed by the ap-
p ro pr ia te m e sh
frequency
o r h ar -
monics that happen to coincide.
9.
Compare
th e f re qu ' l \c ie s o f g e ar m e sh
and the harmonics and sidebands
determined from this test with the
gear mesh and harmonic data. gen-
erated in the singleflank transmission
erJIortests
o f
the same gear sets.
In
7/18/2019 Ruido y error.pdf
http://slidepdf.com/reader/full/ruido-y-errorpdf 6/10
IS E P EA K
385Hz
43 M PH
1st HARMONIC
ME SH
I D R I V E
X C I II I SA
53 IMPHI
1 8 1 H A R M O N i C M E S H I
this way, eharaeteristics of the gear
toofh geometry can
be
related to the
causes of gear noise in the vehicle ..
, ( S e e
Figs.
7h.
7c, 7d.)
Type II Noise Testing.
Comparing the relative level of noise
from one gear box Ito another has been
do ne fo r m any ye ars by a m e tho d as
described
in
R e fe re nc e 2 . This involves
the use of tr.acking, filters. These are nar-
row band pass
f i l te r s
'of 5 or 10
Hz
band-
l
S O O H z
S O M _ P H I
2nd H A R M O N I C
M E S H I
10
VIms
M E D IU M D R IV E
IN OIS E I PE AK
:SPEED
I E X C I T E R
10
ce
tr:IV
1
mV lm s
SIIrt
0
Hz
900Hz
FIa·7d
width that are tuned by pulses from a
tach generator on the drive shaft, such
that the center frequency is always
at
me sh
frequency or some hannonic of it.
A s 'th e ve hic le var ie s
speed,
th e f il te r is
always tuned to a particularorder of
r o ta tio no f the dr ive s haf t. A plo t o f th e
output of the filter
will
show any peak
noises associated with the gear
mesh
at
that particular
order of
rotation.
The
vehicle can
be
driven at various rates of
acceleration and deceleration and a com-
ple t e
hjstory recorded Ior analysis. The
r e s u lt s a re quite accurate an d
r epea tab le .
so it
is u se fu l fo r the
comparison
, o f , tw o
orseveralgear boxes. fig. 8 shows the
Veh ic l e
Recording
Un i t
January/February
1'988 4.
7/18/2019 Ruido y error.pdf
http://slidepdf.com/reader/full/ruido-y-errorpdf 7/10
H i g h In t e n s i t y
M u lt : i .Fr e q u e n c y S c a n n e r s
R E D U C i E
H E A T TRE A Ti lNG
C O S T S up to 4 0
A re ce nt T OC C O d ev elo pm e nt, H igh In te ns ity
l Iin d u c tio n H a r d e n in g (H I IH ) n A ,r ed u c e s h e a t
tre atin g co s ts s u bs ta n tla jly in a v arie ty o f a ppli-
ca t ions.
HIIH
h as b ee n u s ed s u cce ss fu U y w ith l
a p pllic atio n s o n a u to m o tiv e a n d o ff-h ig h w a y
e ngin e a nd d riv e tra in co m po ne nts , a s w ell a s
sp e cia l iz e d 'ge a r a p p :lic a tio n s . T h e
HIIH
scan -
n in g p ro ce s s, w i,th p re s en t te ch n olo gy, ca n
be a da pte d to pa rts u p to
15
d ia m e te r a n d
I ~e n g th sto 1 2 0 .
P a rts th at p re vio u sly w e re fl'a m e h ard en ed in a
ba to n p ro ce ss , ca n n o w be in du ctio n .f1 a rd en ed ,
in - line , w i , th grea ter flex ,ib il ity mo re prec is ion ,
in d iv lid u a l q u a lity m o n ito rjn q , a n d a t lo w e r co s t.
H ig h In te n s ity In d u ctio n H a rd en in g is in d ee d,
a pro ces s o f T he F utu re .
C ro s s s ec tio n o f c am s ha ft lo be d em o n stra te s u n ifo rm ity o f h ard en in g.
M u lti-fre qu en cy, h igh in te ns ity h ard en in g o f ge ars pro du ce s a h igh -
ha rd ness
cheractertsuc
for Im p ro v ed s u rfa ce c on ta ct fa tig ue H fe .
H ig h re sid ua l c om p re ss iv e s tre ss in th e
root
a re a Is a ls o a ch ie ve d
lo r im p ro v ed io o th S lr en g th .
., S ys te m
,- T-OS .. .
d cu m e n t In d u ctl~n . n e e d
N E W . . . tics afe offered
1 0
°onditions.
T h ls ,a dv a 6de m
T O C C O S y $ \e ~ r~ ~ ~ gd n y ~ S a m \ c;ault o~: l~~~~~ ~ecording,P~~~~ ~v~ua.-, .
I)ertorma~cedig\la\ andanalog $l g f ons to r systemperlo
r
t
.r fu r ther details.
T O S mO':utor~t a t o d es igna ted , o c - t d c c o sales . e n g m e e r 0
translernng, - a hoo t i ng . As K y o u r
tion
o r
trouble-s
F or th e la te s t s ta te -o f- th e -a rt h e at tr ea tin g t ec hn o lo gy ,
contact y ou r T OC C O represen.tative,
o r
call: rocco ,
me;
3 01 00 S te p he n s on H ig hw a y,.
Madison Hts.,
M I,
4 80 7 1. P h on e : 3 13 ·3 99 - 86 0 1
o r
1·800-468·4932
o u t s id e M i ch ig a n ),
O IR C L E . A -l 2 O N .R E .A D E R ,R E P L Y C A R D
A u nit of.P srk O hio In du s trie s , in c.
7/18/2019 Ruido y error.pdf
http://slidepdf.com/reader/full/ruido-y-errorpdf 8/10
C OM P O S IT E G E A R N O I S E
· ·
and dynom om ete r te s ts vs . su bje c tive
ra ting s , the c o r re la tio n w as po o r . The
s am e w as tru e w he n com par in g o nly th e
s e co nd hanno nic data . The as s u mptio n
w as tn af a ll gear n ois e w as coming f r om
th e
same Source of excitation.
A new te s t w as 'the n s ta r te d by us ing
th e
t echniques .
de sc r ibe d abo ve fo r Type
E Noi s e T e s t i n g . Itw as no 't f e as ib le to ru n
th e
RTA
in
th e ve hic le at tha t
tim ,
s o a
d ec is io n w as m ad e to tap e re oo rd
I t h e
r oad
te s t data and an ll ly ze it ~a.te r in th
laboratory,
This
was do ne fo r se ve ral
r oad
tests with
d if fe re nt g e ar s e ts .
Subse -
qu e n tanaly s is s how e d that the no is e
h e ard at 80 K PH was ,e xc it,e d by s e co nd
harm cnic o f m e sh , at 1 60 KPH by f ir s t
harmonic of mesh, and at
40 K PH
at
fourth harmonh; of mesh.
In
a U cases the
fn quency
'o f n o is e w as ,approximately 93 0
H z. Now whe n the s ing le
f lank
data
ef
f i rs t
hanno nic transm is s io n e r ro r w as
compared to
s u bje c tiv e r atin g s
at 160
K PH ,and the s ing le
f lank s e co nd har -
m o nic w as c om pare d to s ub je ctive rating s
at , 80 < PH mdp e nd e nt ly a g o od c or re la -
tio n w as sho wn .
te st s etu p and re su lts
from
s uch a sy ste m.
The re is a tr e nd to day to try to do this
kind
of analysiswith
d ig ita l F as t f ou rie r
Transfonn (FFT)
t e c lm ic r ue s ,
rather than
with
ana l og
tr ac kin g f il te rs . H ow e ve r, i. t
is qu e stio na ble w he th er the digital tech-
niqu.es are fast , e n o u g h to d ete ct th e peak
noise periods with s u ff ic ie n t a cc u ra cy
wh e n the vehic le
is
Icon~tan t ly changing
sp e ed , The F FT techn iques ar e acceptable
in
Type I
te s t ing , because
only
th e fre-
quency, no t the relati.ve amp l i t ude s , is of
interest. In
Type II
testing,.
one
is
trying
to c atc h th e abs o ~u te
maximum
ampl i -
tude for
e ach set
ofgears bein~ tested.
There are times when th e noise is heard
on ly as the ve h ic le is changing s pe e d o r
l oa d c o nd itio n s.
C ase H is to ry
N OTE : B e cau se pe rm is s io n w as g ive n
by a Eu r op e a h auto m ake r to u s e ac tu a l
te s t r e su l ts for this artic le , it w as d e cid e d
to pre se n t Lhe data in m etr ic u nits ,
k ile me te rs pe r
h ou r (KPH), rather
than
conve r t
to
MPH.(4)
During
Ithe
p ro ce ss o f in sta ll in g a s in gle
flank transmission
error tester
at an
automotive p~ant..a s tu dy w as
run .
to
correlate tr an sm is sio n e rr or with vehic le
g e ar n o is e . S ing le f lank data was ana-
lyze d byF FT te chniqu e s to qu an tify e x-
c ita tio n re la te d to first, s e co nd and o the r
harm on ic s o f m esh fo r se ve ra l gear s e t s .
The s am e gea r s e ts w ere bu il t in a te s t
ve hic le and ra te d s u bj,e ctive ly fo r noise
le ve l du rin g ; ro ad 'te sts . Th e ve hic le w as
a ls o in s tr um e n te d
and noise
te ste d o n
a.
dyn amome t e r
test
s tan d u sin g
FfT
ana-
l vt ic al t e ch n ia u e s.
The
data
from th e s in g le f lank te s t,
s u b je c tive ro ad te s t and i n s t r umen t ed
dynam om ete r te s t w as the n compared
fo r c or re la tio n. Th e re su lts did no t lo ok
g o o d for s eve ral re as on s.
Ina
s ub je ctive te st. the driver ra te s
n ois e h ear d
at any
speed ,o n a scale f r om
o ne to te n . He' o r she m ay rate o ne gear
set fo r n o is e h e ard at 80 K PH cau s e d by
s e c o nd harm onic o f m esh and ano th e r se t
fo r no is e
he ard at 1 60 K PH cau s e d by
first ha rmon ic of mesh.
T he n ois e s o un ds
th e s am e ,
and the ra te r doe s no t re -
alizethe e xc ita tio n is co m -
ing from d if fe r en t s o u rc e s.
I n . fac t .
at
so me o th e r s pe ed
a similar
n o,is e w as being
c au se d by e xc ita tio n from
be arin g s u s e d to m o u n t
th e pini.on.
In th e instrumented
dynam om ete r te s t, F F T
analyses were run at
g iven
fixed
s pe e ds . T he s e
r e s u l t e d
In
spectral data
showing
peaks.
at many
f re qu e nc ie s . T he p ro b -
lem is to ,de te rm ine
th e o n e to wh ich th e
d riv er is [1 l s po nd in g.
\l \lhe n lo ok ing ; a t f ir s t h ar -
m on i c data. fro m s in gle R a n k
(can tin ued on page 47)
Gea-, Sound
A na ly sis C o ns ole
7/18/2019 Ruido y error.pdf
http://slidepdf.com/reader/full/ruido-y-errorpdf 9/10
CLASSIFIED
R a te s : C la s s ifie d Display-per l in c h ( m l n i-
m um 3 ) 1 X-$1 20 , .3X -$1 10 , 6X ·$1 00 , T ype
will be s e t 1 0 a dv ertls er's la yo u t o r G e a r
T e c h n o l o g y w il l s et type a t n o e xtra ch arge ,
W ord C ou nt: 35 ch a ra cte rs pe r l in e, 7 ,l in es
p er in c b.
I P a y m e n t : Full payment must a c c o m p a n y
cla ss ifie d a ds , M a ill co py to G e a r T e ch n ol og y ,
iP .O. Bo x 1
'
4 26 , E lk G r ov e V illa ge , III..6 00 07 ,
.A ll en cy C .o m m l ss lo lil : N o a g en c yc o m m is io n
o n c la s s ifi ed s .
M a te ria ls D ea dlin e: A ds m u st be re ce ive d
by Ih e .2 Sth of the m on th , tw o m on lhs prio r
1 0
p ub lic atio n , A cc ep ta n ce : P u blis he r
re se rv es th e rig ht
10
accep t
or
re ject
c la s sifie d a d ve rt·is e m e nls a t h is d is cre tio n .
COMPUTER ArCS
G E A R S - S P L I N E S
D ES IG N A ND T OOL IN 'G
• C us to m ge ar d 'e sign lin clu din g n on .
s ta n da rd pre ss u rs a n gl'e s fo r m o re
strength.
• P ro gra m s
t o' e n te r
too l ing
da ta in -
to co mpu te r file s
a n d search
fo r
e x-
isting cuners
to cu t a n ew
gear
o r
sp l ine.
• G ea rin gl computer software for
sa le .
.. Consullin9 services for gear and
s p ll n e p ro b le m s .
V A N G E R PE 'N ·F lE EC E E :N G IN EE RIN G I
1 502 G ra nd B lv d,.
C ed ar F a lls ,
' Iowa
50 6 1 3
(31'9)
266 4614
C IR CLE A~1 5, 'ON R EA DER R E P L Y CA:RD
G E A R E S T :I M A T IN : G
T he C OS T IM AT OR ® co m pu te r
aidedl cos t
,est imat ing
system
inseres
speed and
cons is tency
in th e
difficult
ta sk ,0 1 '
est imat ingl
the co s ts o f
a n
type s o f ge ars ..
Used by sma l l s ho ps a nd :F o rtu n e
. 500
c o ma a r ues
th ro ug:ho u t the
I loountry.
II
:Fo r
co m ple te m to rm a tlo n
, contact us
today.
Manufacturers Technologies, Ine,
59G Iinterstate Dr,
W es t S prin gfie l'd , M A .01 089 '
(413) 733-1972
C ilR C LE A -1 1 6 O N ; R E A D E R I R E P L Y CARD
HELP WANTED
W A N T E O
Gearse l u p man fo r c:I 'ose
to lerance wmk on
I Il a te mode l
g ea r m a ch in e ry: R e qu ire m e nts :
s h a pin g , h o b bin g a n d i~ io n
experien ce. M id wes tern lo ca -
tion; g c x x : t
pay
a n d frin ge s . S e n d
re so rn e to ,:
Box .A T
cto
G e a r
Techno logy
P.O. Box 1426,
E lk 'G ro v e V J n age ,.It. 6 0007
G EAR P ER SON NEL
ENG'INEERS
G EAR G RIN DER S
MACHINISTS
P R O :D U C TI IO N O O NT IF IO L
I
Q U A liIT V C O N TR O L. II NS P E CT IO N
SALES
SH OP SUPERV IS ION, IETC.
If
yo u
are in
the gea r bus iness
in
a n y
capaci ty,
are
a mbitio us a nd th in kin g
of
I
rn ak in g:a m o .
v e , sen d
u s
y ou r. ~ es u m e.
The a bo ve and
o th er n ew po sn o n s
are
be in g cre ate d by
mterna t
growth al
both
o u r S a n D iego , California & Roseville,
I Michigan plants,
A C R IN D U S T R IES , IN C .
29200 Calahan
Rosev i l le , Michigan 48066
Brin g in n ew cu s to m ers
fo r
yo u r bu s in e s s by a dvertis in g in
G E A R T E CH N OL O G Y , T h e
.J o urn a l o f G e a r M a n u fa ctu rin g.
Call f31'21437-6604
46 GaorTechno loQ,y
G , EA R C A LC U IL AT IO N
P R OG R AM S F OR TIH 'E
G E AR M IA NU FA CT UIR IE R
IN:EXP,ENSIVE,
IE AS Y T 'O ' iR U N
'O'N I IBMI 'COMP ATIBLES
I IEXTEI ' 1 NAt
I
I N T I E R N A t
S PUR
H E L l C A t
I N C H
C D P )
M ET RIC ,(M Od ,)
SHIAIGHi BEVEL (1
i
04),
S P I R A L .
BEVEL
1 1 6
& . 27 GRD
B LA NK A ND , S UMMA 'F UE S
IF or B ro ch ure a nd M ore ,
Informatl'on
C AL L O R: W A IT E
IUIIN IV EIR SAL G EAR CO. I:NC .
P. O. Box A .G .,. A nz a,. ()A 9 23 06 ,
(714)
76346,16
CIRCLEA. ·17 ON
R EA DE R R EP LY C AR D
BUSINESS OPPORTUNITIES
W A N T E D
H igh q ua lity 'ge ar m a nu fa c-
tu re r o pera tio n . S m a ll to
m ed iu m (25 ) ge a rs . W ith o r
w itho u t pro du ct lin e . A ll in -
q u irie s h el'd j n s tricte st co n -
fid en ce . P rin ciple s o n ly.
Bo x G M
clo G e ar T ech no ~I'o gy
P .O . B o x 1 4 2 6
E : l k
G ro ve V lilla ge , IL 60 00 7
There 's s til l t lime , • .. .
order
clo sin g d ate
fo r
a
c:lassif ied
a d in the M arJ A pr.
is su e is
Jan.
10 th .
7/18/2019 Ruido y error.pdf
http://slidepdf.com/reader/full/ruido-y-errorpdf 10/10
REBUILD. G
H OB BER R EBU ILD :IN G
S PEC IIALlS,TS
H a v in g tro u b e m e etin g
leday's de-
mand ,quality oontrol tolerances?
~e t o u r fa cto ry tra in ed a n d e .x
perienced staff return y,our
mach ine
to optimum operating
condi t ion.
We specializ,e in r,epai'ring,
l iebui lding
and l modemizingaJl
m a k e s 1 0 1 hobbers.
.' C le ve la n d
IRigidhobbers
.' G o l il 'd & Ebelil 'lardl
.' B arbe r Colman
IP ;R iE ,SSt t14TION I'NC.
5 22 C o tta ge G ro ve IR oa d
IB loomf ie l 'd . Conn . 106002
(2 0 0 ) 2 4 2 :.8 5 2 5
C IR ClE A I',a O N R E A D E R R e Pl V C AR D
R E LA TIO NSH 1 P 'O 'F
M EASU R ED GEAR NOISE ... '
c,ontiP lued
from
pAge 45
With this knowledge, another pro-
gram was started with a greater number
of axles and
yielded excellent results,
even when comparing single flank trans-
mission 'errors ,against subjective
Call'
ratings. (See fig. '9.) From this data
it
was
possible to establish an acceptable m'ange
of single f lank transmission errors for
SUBCONTRACT
0
I
I :
o 0.'
o .a o .a
OA
G. I o .e
G.7
o .e
U ,.
G E A R T OO TH G FlIN D ,IN G
H O N I N G ONLY
Product ' ionQuant i t ies ,
3/4 P.O.,
to
27.5 P.D ' . ;
3 ,.5 D .P . a nd 1 1 IF ace
Gear Tooth Finishing
Is eur
'O n ly B u sin e s s
W e h a v e n o . tu r n i n g , h o b b in g Of
s h a p in g c a p ab il it y
A L L E G H E N Y G E A R C O R iP .
23 Dick R.oad
~t..'U~ 0..
Depew, NY 14043 ~
71 6 -68 4 -38 1 1 ~
O IR C t J E . A -2 1 O N R E A D E R R E P l Y C.MD
iirstand second harmonics of mesh.
Conclusio.n
Many di f f e r en t types · o f in~t rumen ta -
tion are available for the sound testing
of gear noise. However, the procedure
used with the instruments is the
key
to
successful
application.
Relating
th e
fre-
quency of the noise aUl'ally and visua]y
in the data is 'the most import.ant step
in
the analysis.
fQl
~
BUSINESS ACQUISITIONS
t o I : :
I ••
I
,
. ::
7
~ l
i . -
j : · · · · - -
2 I
I
I
:
• 7 • • 10
W A N T E D T O B U Y
fll.9
L arg:e ' pre cis io n g,e ar' co m -
pany. M u st have m od'em
equipment, lnterested ln
o o rrpa n le s th at a re ,cu rre ntl:y
,o pe ra tin g ,a t lo w Ip m du ctio n .
IRepli:es confident ial .
Bo x BT
c lo Gear l e c hno l o gy
P.O.
B < r x
1,426,
Elk
'Grovel
VIII 'age
IL
16(X)()7
h
is possible to
apply ehe
above
eeeh-
niques and sort out 'the facts from mys-
tery.
Onceth:js
is done,
3
correlati:on
between noise and transmission error can
be established and lim its can set on th e
process for quality ,eontJIot
R:efe : rences :
1. SMITH, R.E. Identification ,o f Gear
Noise With Singl Flank Composite
Me4SUreD' Jen l , AGMA.
FaIl l
Technical
Meeting, San .Francisco, CA, October
14·16, 1985 .
.2.. PITTS, loS. Bevel and Hypoid Cear
Noise Reduction, SAE Paper No.
720~4.
3. SMITH, R . E . . 'What Single Flank
Me as ur em e n.t C anD o E io rY o u. AGMA
Fa l l
J:echnJeill
M e e tin g. W a sh in g to n,
D.C.,
October
]5..1.7, 1984.
4. V OND ER SC HM lOT. D ata f ro m te s tin g
do ne at
A dam 0pe II
AG. ,
Rl l e s s e l she im ,
We s t
'Genna.n.y.
Acknowledge.ment: The author wishe to thin .:
A dam O pel A G .• RU eJ54 llsheim .
Wes t Germany:
RItd Dr. Vcmde r5 Chmi dt fo r tM u se 0/ thei r restl'iatA.
Rqrinted wilh pmnwion of thl ArrIt'riam G_
Mmlufactums Association. The opinio1JS.
st l~
mmts a n d conclusions presented in this p.aperan
those of tne Aulnc,.,
,llI'Id in riO' TDtl j rrpraml
Ih
p ositl O o r o pin io n o f
the A MERICA N GEA R
l1. tANU fACTU .RERE ASSOCIAT ION .
January /February 1988 47'
Top Related