Professional Documents
Culture Documents
Gear Inspection Method PDF
Gear Inspection Method PDF
Rg. 1- Involute charts were made on a variable involutemeasuring tnstrumem and charted
Much information has been written on gear inspection, A typical functional rolling chart is shown in Fig. 4. Runour
analytical. functional. semiautomatic and automatic. In most and wobble can be readily found from this type of check,
cases, the charts, (if you are lucky enough to have recording but involute, spacing and lead 'errors are difficult to find. The
equipment) have been explained. If there is an error in the tooth-to-tooth composite variation is not the part's spacing
gear, however, there is little information to tell you what the errors, but is a composite of lead, involute and spacing errors,
causes of the errors might be, or to oHer some idea on how as weU as nicks, etc. Wobble or face runout will show up
to correct them. This article attempts to offer some rudiments as a single or double runout pattern. However,. sometimes
of "what to do next". the pattern of the error in the functionalcheck can lead you
Figs. 1~3show typical involute and lead charts.as well as to the cause of the error. Fig. 5 could be of some help and
the methods used to check runout and spacing on all the is self-explanatory.
Analytical Gear Checks shown in this presentation. These
charts can be used as a guide to test results of a normal, error- Runout
free gear. Fig. 6 shows all the analytical checks (lead, involute, spac-
ing and runout) on a given gear, when the gear is mounted
and inspected on the same center as it was machined. This
particular gear is of very good quality, AGMA Class 12 or
13. Note that the gear was checked in three different pesi-
AUTHOR: tions, and all the checks are basicaUy alike. In order to
demonstrate the chart readings indicating runout, 'the gear
ROBERT H. MODEROW is Training Man.ager for the was then mounted with runout and checked again. The results
lllitrol1 Division of Illinois Tool Works, Chicago. He has over are shown in Fig. 7. All three checking positions gave basic-
twenty-five years experience in gear tooling, gear desl~gtIand ally the same checks, but the runout shows up, as it should,
inspection. Mr. Moderow has conducted gear seminars
and this runout affects the tooth spacing and the involute.
throughout the Un,jted States, Mexico and Carll.de:! and
The lead is not affected, and would not be, unless the gear
presently conducts four day gear seminars at llIitron dealing
with the basics of gear .theory, production and inspection. He had lead and involute modifications, or was helical (Note
is a member of the Sodety of Manufacturing Engineers - runout does affect the gear's involute, but this error shows
(S.M.E), American Society for Quality Control (A.S.Q.c.), up more en small numbers of teeth. Also, the involute varies
and' .the A.G.M.A and has spoken at many of their meetings. from plus to minus, depending on the tooth checked.)
30 'Gear Technolog,y
,ACl'UL[AII'
AL jr
'01' HELII - ,',',"".,,'.
•
-,DE,SIRIO, ILIAD
01' 1lEl.:I!I
!~I I
J
' I
I 'I
I"
'I:
,
,I
,
I
I'
I
'.
.1
=I:c ,
II I I
!
lJ ell II I
I
II
I'
II, .llli:1
.. I-I I
,
liJI •
, , I
o.lMI
I II I IW I
I.:i-I ,.
,
,
II! :1 !JII r I, ,~ I
1 ,
, I.
~la. Z-Lead chKb wel1t made an iI he1ic~ lead measuring inslrument and charted.
SPACING
RUNOUT
i;1g,. 4A - (BelowJ Typicil1 chart obtained
when using a "Gear Roller",
\ I -
"
. . ONE
~~TURNI
4J!~A
-B
-C
CHECKING ~
POSITIONS
34 Gear Technol:ogy
o WOBBLE
Fig. 7 - Analytical inspecnon charts showing Ihl- ef· GEAR DATA
fects of runout, either at sh lime of machining or 6.3 D.P.
al the time of inspection. 01'1 the various gear 30° P'.A.
elements, Note that the charts are ba ically alike,
no matter where the gear i checked. I I 9 TOOTH
SPUR
.0034 RUNOUT
04. I I
- '-I
__
I
I
,I,_,_I
CHE~KING
POSITIONS
)
I'
__ C
I I
~~~-=-:-
R.o~
May/.:kJne 1985 35
.005 WOBBLE
GEAR DATA
Fig. 8 - Analytical inspection charts, showing the
effects of wobble, either at the time of machining
or the time of inspection on the various gear
elements. Note that the charts are different, depend-
A 6.3 D.P.
30° P.A.
9 TOOTH
ing on the position of the check, relative to the pivot
point of the wobble.
SPUR
-- - ..... -- -A
'J
, I
-,~ .... - - ..."
-c
36 Gear Technology
AN INVOLUTE ClilECKII'IG F HIlER
WITIIINSUFFICIENT HOOIIWLL NOT A "PROPERL'f HOOo<EO
SHOW TH£ COIIAECT iUIOUHT C1# INIIOUIfE C>IECIIING FINQ£R WILL
UNDERCUT NOR THE" COIIAECT SHOW THE cOlVlECT .....otJNT OF
SWEEP-OIJT DIAMETER UNDERCUT AND THE CORRECT
UNIl£RCUl' SWFEP-o\Jr DIAMETER
Fig.9A Fig.9B
CORRECTWON
Rg. 9C
May/June 1985 37
D ·3
r-
!-IUSriler;;
OF 'Hoe
.'.'
• AVEHAGE POS'lriON
.'.'
'.••
OF THREAD "2
IN FIRST CONVOLUTION
.'
.1 '~
I
,0
t
•I ,~
''.. •i
•~
• ;
•i
:i
.B
•
AVERAGE posmON
Of THREAD'"
IN FIRST CONVOLUTION
• II
••
•
Fig. 10
Hob Errors ing a hob incorrectly. Number 1is a very Because of this, when hobbing with
The lead chart ofa hob is an indica- good hob mounted properly and the re- heavy feed rates, the entire involute fonn
tion of the accuracy of the helical path sulting gear involute produced. Numbers will tend to drift. (Fig. 12) The resulting
of a hob's teeth, A single thread hob can 2-4 show the same hob mounted incor- involute checks may vary on the same
affect only the profile of the gear tooth. rectly, with the resulting effect on the tooth across the face of the gear, depend-
If the hob has a good lead, the gear's pro- hob's lead chart (checked as mounted) ing upon just where the involute check-
file can be good. If the lead is had, the and on the involute form produc d on ing finger is positioned. This is not to say
profile will most likely be bad. Fig. 10 the gear. there is an error here, but just 'to explain
shows a hob lead chart. As a hob cuts a gear, the top of the why two involute checks may vary when
The charts shown in Fig. 11 illustrate gear's involute form is finish cut at a dif- checking the same tooth.
the problems which result when mount- ferent point in time than the bottom.
38 Gear TechnolOOv
~" dc
.2'
I
o
1 o o
~,o ~':0007 ,
o
! e , I
o
I
o , o
o o
o o
I
o o
~ o I o
.rt< o o
.i, I.OOOl ooo2jf1 o
c
,
0007 ,
o
o :
o i o r
I,.
.1
o o ~
o !
o
;
,
I o o : •
• o
o
, o
o
~'m~"m~~WW~~~\U
,
=
1
I ,
,
,
o o
~~. o o
""
Fig, n - These charts show the type> of weave produced by various conditions of runout and the amount of error each type produces in il gear tooth.
3 CHECKS ON
SAME TOOTH
Fig. H.-Involute checks on the same tooth vary if the gear is hobbed with a heavy feed rate. the involute drifls and one check may high
point of feed. scallop at pitch line. and another check m y have high point at another place along the involute.
Muv/Ji.Jne1985 39'
INVOLUTE
rCHECKING
®
PROBABLE CAUSE-
FINGER
INVOLUTE CHECKING MACHINE
GENERATING SUDE OR
FINGER HOLDER NOT ON CENTER
t-·
i _
, -;
;
""'-- FINGER
TIP ON CENTER !,
: ..
PRODUCES CHART
(CORRECT CHART)
!
j-
,.
j -. - '
! -- ~ -
l"
:.--~t·- , 1
1
I
·1 -c; '-
1-
'
•
1
12.° .j.
I
,..-INVOLUTE
CHECKING
FINGER .. '
I -- FliNGER TIP
NOT ON CENTE,R
SHARPENED
BACK?
PRODUCES CHART
(SHORT CHART)
Ag.13
40 Gear Technology
I -.~ + ".
'
1 . ..- ___ . ,_,'
GEAR HAS LEAD ERROR
.I~"i, '
PROBA;8LE CAUSE-
MAC~I.!NE MIS-ALIGNMENT .J
SET-UP ERROR-GEARS I.
HOB ARBOR ENI) PLAY )
'"
Fig. 14
Moy /JuneI98541'
.1 >
,
I
I.,
~_t"
I~ 1__
- I
I,
i
-t--
- -
I I U jJJ-j!
1 ---1~-r
' I
.-
"~
I~.
I
I.
' •
PROBABLE CAUSE-
MACHINE WORN OR LOOSE
WORN HOB SPINDt.E SLEEVE
BEARINGS
EXCESSIVE HOB ARBOR
END PLAY
WORM WHiEEL WORN
WORM RUNNING OUT
INVOLUTE
UNIFORM EXCESSIVE :ERRO;R
(WIGGLE)
--'"..J
Fig. 15
42 Gear Teohnology
HOB MOUNTING ERRORS
MULTIPLE THREAD
THO'1 THO"
MOUNTING ~
CENTER."r .+,
" ,----......
HOB TRUE
CEiNTEB
J ,~
THO 12
2 THREAD HOB
At IS LAAGER THAN R1• AND WHEN CUTTING Ra AND LARGER
EVEN NUMBERS OF TEETH W1LLCUT WIDER CUTT1NG A NUMBER OF 1EIftIIL~."L.'
AND DEEPER TOOTH SPACES DMStBLE 8V S. WU
DEEPER TOOTH SPACES
RUiNOUT
- ·T-- I I
Fig. 16
TEETH APPEAR ,EaUAlLY
Fig. 11 - Possible cause: .
SPACED AT THESE DIAMETERS
II Hob hubs running out-out of phase (high with 10~1
2) Hob has weave error in lead, out of phase, from thread 10 thread
r 1 ·1 1.1 ' . j
GEA!A HAS INVOLUTE '.
-'VARIATION-ON i
CONSIECUTIIVE TEETH
TWO THHEAD HOB
11~~'CUnIIN_GIEVEN NUMBER
OF TEETH
'" i,I
I '. I I
t
I·;, !I. 1I • I· '] .>
~.~al'
~_-:.
~-----.-,
...-:C·-I-~·l---l- ~'
1-'
POSSIBLE
r , ,'.,.'
•. -
.--
..
:§§:
~. ---+'T
-"~--I .' -- "_'1
'I~:
"I e'
_EQUAl. SPACING
ERROR
\
PR()[)(JCED BY OOUBLE
f
THREAD HOB
\. ~INC) fN 000 NUt,!SEA ,Of' TEETH I
44 Gear Technology
ALTERNATELY WIDE ANID
NARROW FEED SCALLOPS
POSSIBUE CAUSE-
1') HOB IHUB(S) RUNNING OUT
2) HOB HAS THREAD TO THREAD
SPACING ERRORS
Fig. 19
POSSIBLE CAUS!E-
1) H:OB HUBS RUNNING OUT -IN PHASE
(HIGH WITH HI:6H)
2) HOB HAS THREAD TO THREAD
SPACING ERRORS
Ma,vl.Ame1985 45
Inspection Equipment Errors Multithread Hobbing the part. However, with today's high feed
Fig.. 13 shows twcJnvolute charts rates, much of this "canceling effect" does
which have errors, resulting from inac- not occur, and the errors are produced
Multithread hobbing has its own
curate inspection equipment, Chart #1 peculiar problems, somewhat similar to in sections along the gear's face width,
(Fig. 13) shows actual charts of a gea:r single thread hobbing, and yet quite dif- in each individual feed scallop. Many
having a 2.3492 base diameter. Chart "A" times, feed scallops can tell a story.
ferent. Since more and more companies
was made with a new involute finger are going to multithread hobbmg, it is Fig. 18 illustrates a Ieed scallop pattern
which had its tip oncenter, Chart "13. was important to cover some of these which varies i.nwidth along the involute.
made with a hooked finger which was Since each feed scallop is produced by a
problems.
"sharpened back" and the finger tip was A drawing of a. gear hob and the lead
then off center by .022. This made a charts from both a single thread hob and
"short chart", being off one degree of roll. a three-thread hob were shown in Fig..
(Note that alJ modifications will also be 10. The three-thread hob lead chart
displaced one degree of roll.) Chart #2 shows a thread-to-thread spacing error
(Fig. 13) shows 'two charts again, AWand
betweenthread #2 and thread #3. It is
H
'1r, whicha:re of the same flank of the generally shown that if a gear with a FEED SCALLOPS NOT SAME
same gear tooth, but turned end for end number of teeth evenly divisible by the ONI BOTHSI'DES ON TEETH
in the involute machine. VVhenthe charts number of hob threads (3) were cut by
do not Come out identical. as to the posi- this hob, the gear cut would have a spac-
tion of their modification relative to the ing error approximately equal to the
degrees of roll. usua11y the machine is not error in the hob. However, assuming a
centered property. Most involute check- 3~thread hob had little or no spacing
ing instrumentsare furnished with a cen- errors, because just a section of the hob
tering gauge to check out this problem. teeth (one axial pitch or so) is finish cut-
Lead Errors Hng the gear, hob mounting would be
Five lead charts with various errors very important. If not correctly
have been grouped in Fig. 14. When mounted, it would have thick and thin
dealing with lead errors, it is important teeth, spacing errors and runouts. which
to remember lead errors seldom, if 'ever, show up in patterns equal to the number
come from the cutting tool. They almost of threads in the hob. Fig.. 16 shows the
always come from machine misalign- condition and the result on the gear. In
ments, looseness, lack of rigidity, gear each case shown, the effective radius to
I 1.1!'llmrmi.'''I''~I'ttrI~~Hi
thr,ead#l (Rl) is shorter than the effec- INVOLUTE· "l·l.l1j"
blank accuracy and mounting, etc. Chart
tive radius to the other th:read(s), and Rl VARIATIION .
fl4, showing a lead error only on one side
of the gear tooth near one end, usually will cut shallower tooth spaces than the
occurs as the hob is breaking out of (or other threadts), On the examples shown,
into) the cut. This is a result of machine the dimension over wires on the gears
or arbor deflections. This can be cor- would be considerably different, from
rected by climb cutting, if the gear was one tooth space to the next.
cut by conventional cutting and vice Again, when cutting a number of teeth
versa. A marie rigid setup of machine in a gear which is divisible by the number
(such as larger diameter hob arbor) or of hob threads (non-hunting too~th), each
lighter cuts may be in order. Chart #5 tooth is cut by one given thread on the
(Fig. 14) is a difficult problem, and is us- hob ..Therefore, each thread will produce
ually caused by machine misalignment, a. given involute, and involutes may vary
although, the hob could be at fault, if from tooth to tooth. This is shown in Fig.
you arediagonall or oblique hob bing .. 17. In this case, the gear's tooth spacing
mayor may not vary, depending upon
lnvo.lufe Errors the position of the spacing check along
Fig. 15 shows involute charts with the involute.
various errors, all of which are self- The usual "rule of thumb" in
explanatory. It is important to note, multithread hobbing is to cut a number
when using a single th:read hob, if the in- of teeth not divisible by the number of
volute error is uni£·orm (essentially the threads (hunting tooth). This is definitely
same on an teeth), the error may be com- advisable, as many ·of the errors may
ing born the hob (accuracy, mounting, tend to "cancel out", as each thread of the
sharpening, etc.), but if the error is non- hob eventually passes through all of the
uniform, it is more likely to be coming gear's tooth spaces ..This would work ef-
from the machine. fectively if the hob was not fed through
46 Gear Technologv
differe.nt hob thread, the involutes can dueing that area (scallop) of the gear. In- tooth ratio rule" was followed. but the
vary by as, many 'threads as there are in volute trace 1/2 isa combination of /II hob was mounted with runout On on
the hob and more. At the junction of two and #3. end, or had severe 'thread spacing on one
adjacent feed scallops, there is an. Figs. 19 &; 20 show variations of feed flank. If the hob is running out on one
"overlapping" area. where the "canceling scallops, in patterns relative to the end, the side of the gear tooth with the
,efEect" does occur and another involute number of threads in the hob. tWlc,e normal size feed scallop win also
variation results. This is shown by the Examples of feed scaJlops not being the have severe involute error, The other sid
involute checks (Fig .. 18). Trace eland same on both sides of the gear tooth or of the gear tooth will have a much bet-
(13 result from. differ,ent hob threads pro- space are show-n in Fig. 21. The "hunting ter jnvolute.
)PPOSITIE SIDES
)F SAME TOOTH