precision Polymer gears 2

[subsearobot]

Hello folks,
I am involved an instrumentation application where molded gears are needed.
Originally, the gears were customized by a consultant using Kissoft. When we went to tooling, the molding team redesigned the geartrain using AGMA profiles (to ease the inspection).
This molding house makes gears for some big name electronics companies. So, despite everything I'd read about the benefits of modified involutes, the gears were molded to AGMA. (this was 5 years ago. yes, I work for the slowest product development company in the universe.)
Fast forward, and management wants changes which require a redesign of the gears.

this is a precision handheld instrumentation application.

Question to you polymer gear folks:
Is it worth a change in gear molders?
My long ago understanding was that modified profiles could reduce friction and backlash in a gear train. We do not have the equipment inhouse to measure the backlash in the system. However, my intuition tells me that there is excessive play in the display gearing- as compared to competitors in our market.

please, comment on the benefits that you have seen using customized gear profiles, and whether they are applicable to a hand powered instrument. (low power, very low duty cycles, 2-3 million input revolutions over the 10 year life of the product). We have the opportunity to re-do this correctly, but it will be a push on my part to convince the team to change molders.

thank you

 

[mfgenggear]

subsearobot

I not a plastic gear specialist. how ever that said gears are gears. however changing suppliers is always risky.
questions :
Has the supplier forwarded a sampling inspection plan with certifications, gear involute charts and total variation, and tooth to tooth variation charts.
or have a master gear to verify total composite errors.
The only reason to change design if it over design with to stringent of AGMA or DIN gear quality lever EG AGMAQ 12 vs 10.
when designing gears a very complex analysis is required. Torque, RPM , efficiency, noise levels, # predicted Cycles before failure or wear.
I do known plastic gear molding is an art. there has to be allowance for shrinkage, and to maintain the correct tooth thickness. which to
a set center distance maintains the correct backlash.


take a large sample of gears to a reputable gear shop, have them tool up if necessary and inspect the gears to the drawing. inspect and record the attributes
record gear charts and analyze if the supplier is meeting or exceeding the requirements. MOW, Tooth Span or Master Center distance verification is crutial.
no quick answer here.

 

[mfgenggear]

also there are many gears out there designed to AGMA standards, and to Din, the issue is not the standard, the issue is if the gears are being being manufactured correctly or are not tight enough quality requirement. so instead of AGMA Q10 the gears may need to be AGMA Q12, as an example, if you can give details what the issues are with the gears and more specific detail of what expected from management, then probably better advice. also the design of the profiles and if the teeth have profile shift to strengthen the pinions,
and the gears were designed not to have interference. so many attributes to look at. Kissoft is one of best work horses out there, and is one of the best. I am not affiliated with them, but I had the opportunity to work with it in a limited capacity. but there are a many great programs out there that are very good as well.

 

[3DDave]

The following applies for the sort of gearing you seem to have - not high loading, high power transmission gearing which takes other factors such as tooth bending and problems from harmonics into consideration.

Ordinary involute gearing is essentially perfect gearing - that is the ratio of angular motion being transmitted is invariant with alterations in center distance. The payback for that ability it that there is sliding contact as the teeth come in and out of mesh which can produce friction and wear. Perfect involute teeth have no necessary backlash. Backlash, the making of gear teeth to a smaller profile, is geometrically to accommodate center distance variations or tooth form manufacturing variation. Unless some magic is happening, whatever the reasons for backlash it will affect involute and modified involute the same so whatever competitors are doing has no relation I would expect from custom tooth profiles.

Lower wear/friction teeth are cycloidal** - these are used in wrist watches and clocks where the minimum friction and wear is a premium, but center distance must be much better controlled or the motion will not be transmitted uniformly - less a problem in a ticking watch or clock than a 1000 Hp transmission.

Then there's most every variation. For example, to prevent tip contact in high-force gears the form is modified to ease the tooth into mesh, at the loss of some contact ratio and increased load at the pitch line, but much better than high sliding loads or teeth jamming.

I don't see that the company using involute form would ease inspection, but it would certainly easy fabrication. Typical tooth generating techniques require relatively simple and somewhat universal tools. A custom shape needs custom tools to generate. The inspection should include lead error; if you specified a different form for the tooth that would show up as lead/involute error and just be compared with whatever delta was desired. I do acknowledge they might software that doesn't allow that comparison, but getting the cutting tools right is more difficult.

**

https://en.wikipedia.org/wiki/Cycloid_gear

and

http://www.csparks.com/watchmaking/CycloidalGears/index.jxl

Not to be confused with cycloid drives.

 

[subsearobot]

3D Dave, thanks for the info- good stuff. I need to read up on cycloidal profiles.
our molds are made with wire edm, so the shape is totally independent of the tooling needed to create it (except wire diam, of course). we are not bound by hobbers.

my assumption regarding backlash, based on my *limited* understanding:
I understood that one of the benefits to modified involutes was to allow for a higher variation to center distance. there for, the AGMA gears woudl have a higher ratio between CD error and backlash, all other things considered equal. Since molded parts have a higher dimensional variability, we would tend to see less backlash with custom gears than AGMA due to center distance differences batch to batch. What is the flaw in my logic? (not disputing you, I am definitely the less informed one here). I think i need to learn more about AGMA too- I suspect I've made assumptions there also.

thank you for helping spread the knowledge!

 

[mfgenggear]

all gears require back lash, if it is at tight mesh they will have tooth interference, and will wear prematurely.
it's actually better to have to much back lash than not enough.

 

[drawoh]

subsearobot,

I have successfully rapid prototyped plastic gears, using the fake "involute" profile described in my 1941 edition of French's Engineering Drawing. They worked fine. I would expect plastic gears to be low friction, and I would expect backlash. If the gears had been larger, I would have read up on how to model involutes in SolidWorks.

If backlash is an issue, you need the precision of metal gears. You can spring load the gears together. I have had a gear vendor offer to fabricate a gear oversize, so that when pre[‑]loaded, it would run on the correct line of action. Is your plastic process accurate enough to take advantage of this?

--
JHG

 

[mfgenggear]

important data to look at is contact ratio, tooth bending, root radius, and no interference. old timers would change the pitch diameter on one component
to change sliding action of the gears. but maintaining the base pitch. so there is proper tooth conjugant contact.
I imaging plastic gears the torque and RPM would be low. contact stress would be a big factor. if plastic gear was still around this was his expertise.
as dave said Lower wear/friction teeth are cycloidal" gears are more efficient but are more expensive to tool. since it is molded, it may be an option.
the budget will dictate the design method. but a new redesign will be expensive. look at the existing design, and run a new analysis. on high velocity
gears (high RPM) the profile is very important, due to the dimensional change from moments of tooth as high speed and torque. causes tooth interference, thus the
modified involute to compensate. with a minus tip on the profile. here are the advantages on a standard teeth with a modified profile.
The amount of Total indicating reading Total composite error, Tooth to Tooth error, will be higher on a standard profile.
the objective is to obtain precise TCE, very little error. even if it is much better than the requirement.
the advantages , less wear, quieter noise level, smoother action. on small fine pitch gears there is no involute tolerance, however the profile of the involute
will immensely effect the TCE error. of course the index variation has to be precise. problem I envision as a molded gear having to much index error.
and is not as precise as a finished ground steel gear. in aerospace it is not uncommon to maintain .0001-.0003 total tooth variation.
molded gears may obtain .0005-.001 total variation or more. that is why it is important on my previous post to obtain analytical inspection data.
and verify if it is a design issue or a manufacturing issue. my personnel thoughts. involute gears will do and have done the job. in many designs.

 

[mfgenggear]

also as important look at the housing, verify that the shafts have no misalignment. improper misalignment of the gears will cause
error in the tooth contact will cause premature tooth failure. take an assembly and apply redline on
a single gear and look at the tooth contact pattern, if it has proper tooth pattern then it is not the housing.