Gear Train Design Question? 2

[fsincox]

In my machine tool days, I was taught that you designed a gear train with the pitch circles touching. The place I am at now wants to leave a .002 gap between pitch circles. I believe the proper offset is built into the actual gear tooth design. Grade 8 gears used in aerospace butterfly valve rotary actuators. Heat may be an issue?
Thanks,
Frank

 

[3DDave]

It depends. Produce clearance by thinning the teeth which increases stress levels and decreases contact ratio or by increased spacing which lowers contact ratio and increases stress.

One can do both at the same time. My experience for such has been that it's cold that is the problem. Housings were aluminum and gears were steel, so at lower temperatures the housings contracted more than the gears, decreasing the center distances. I suppose if there is enough power in the train the gears could also warm and expand, making low temp operations worse.

 

[gearcutter]

Backlash is generally introduced by thinning of the teeth allowing mounting/centre distances to remain nominal.
The amount of thinning is negligible and has very little effect on the tooth's strength and mesh geometry.
Other than bevel gears; I've never seen a design which specifies changing the standard mounting distance so as to increase backlash.


Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[tbuelna]

The beauty of involute geometry gears is that they are relatively insensitive to small changes in PCD. Since gears and housings are manufactured and inspected at room temp, but usually operate at a temperature range above and below room temp, it is quite common to adjust the PCD of the housing to compensate for any CTE mismatch between the housing and gear materials. The CTE mismatch problem can be quite significant with large diameter steel gears in an aluminum or magnesium housing. The relative change of a steel gear mesh with a 12" PCD in an aluminum housing would be about .006" smaller from RT to -20degF. Without compensating for this effect, the gears would jam/bind if operated at -20degF conditions.

 

[israelkk]

From AGMA 917-B97, Design Manual for Parallel Shaft Fine-Pitch Gearing

"Tooth thinning coefficient is a normalized measure of the amount by which the tooth thickness of a pinion or gear is reduced from its nominal value at the standard pitch diameter. The tooth thicknesses are reduced by feeding the cutter deeper into the gear blank. The tooth thinning coefficients are considered independent of the profile shift in order to have the outside diameter independent of tooth thinning for backlash."
.
.
.
"The maintaining of a desired minimum backlash at the tightest point of mesh is the first objective of the design procedure. The minimum backlash is desirable for one or more of the following reasons:

-- To provide room for the fluid lubricant film that separates the loaded surfaces of the gear teeth;

-- To provide a flow channel for the fluid lubricant trapped in the root areas, especially in high-speed, wide-face spur gears, thereby relieving the pressure that adds load to the shaft and bearings and reduces efficiency;

-- To permit the passage of small dirt particles carried by the lubricant;

-- To provide additional clearance between the gears as an allowance for possible gear and mounting dimensional variations and thermal effects not considered elsewhere in the design procedure.

An interference condition, or negative calculated backlash, should be avoided. Even when it does not prevent assembly of the gears, the interference will cause deflections of the gear teeth and of the shafts, bearings and other supporting members. Such deflections often result in unpredictability high forces, greater than the forces from the design load, and lead to reduced efficiency and premature failure. The value for minimum backlash is used as input data in the design calculations. It should be selected after review of the above considerations and their importance to the application."

 

[gearcutter]

Thanks for that looking up that info israelkk - a star for you.

tbuelna - the situation that you describe is usually compensated for by the addition of backlash derived from tooth thinning.
Both AGAM & ISO talk about the variations in thermal conditions as you describe but I can't say that I've ever seen where the standards talk about varying the mounting distance to compensate. My impression is that it's always done with tooth thinning.
Of the hundreds of gearboxes that I've seen; mounting distances have always been nominal with the only variation being the manufacturing tolerances.

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[tbuelna]

The OP asked about a .002" gap between the pitch circles of their nominal gear geometry layout. Unfortunately, there was no other detailed information provided regarding the gears, but we all seem to have assumed they are an external spur gear mesh.

Maybe fsincox can provide more details about the gear set.

 

[fsincox]

I am sorry, thank you all for feedback. Yes, they are grade 8 external spur gears generally 64P, The heat issue was something I had not considered. Machine Tool gear trains have oil cooling and all steel parts, here we are talking aluminum housings and steel gears spinning on hardened steel shafts.
Frank

 

[tbuelna]

Sorry for the late reply. But 64DP spur gears having class 8 tolerances is a bit unusual. It can be difficult to get precise contact geometries with such gears.

Can you provide more details?

 

[fsincox]

tbuelna,
I would be happy to if you let me know what you need. The gears turn on fixed shafts. Very small stuff. Can you explain class 8 id too high or low?
Thanks all,
Frank

 

[tbuelna]

Can you provide details of the gear mesh such as numbers of teeth, face width, etc.

As I noted in my previous post, with very small size teeth like your 64DP gears have, tolerance accumulations can be problematic if the gear mesh is transmitting any meaningful torques, operating at PLVs of any significance, or must provide any amount of positional accuracy. Since this is an aerospace application for a valve rotary actuator, it seems likely you might want to specify a higher quality class than AGMA Q8.

 

[fsincox]

Most of the gears have a working face width of .062. They are stacked on offset fixed pins and the bores through the gears act as their own plain bearing for rotation. The current train runs from as small as 9 on the drive motor to up to 46 teeth.

 

[tbuelna]

A 9T 64DP pinion with a .062" face is a tough gear to get right. Depending on what the mating gear looks like and how you intend to manufacture the 9T pinion, it might require quite a bit of tweaking to prevent undercutting and get it to mesh acceptably with the mating gear. One thing I would recommend is doing a quick design study of how much modification (like profile shift) your 9T pinion might require based on the AGMA recommendations. Then if you're feeling ambitious, you might try doing a tolerance sensitivity study using the extreme combined conditions of the profile & runnout tolerance limits from the AGMA quality class you intend to use, the gear shaft positional tolerances, and the bearing journal tolerances.

 

[israelkk]

Class 8 is not a clear definition. You have to specify the governing spec the gears should meet, such as according to AGMA 2000-A88 (standard for gear accuracy) Q8C etc. Class 8 by itself says nothing!!

Added to the definition such as AGMA 2000-A88 Q8C you need to specify maximum and minimum testing radius when test with double flank master gear testing machine (or at least and minimum dimension over wires which is less desirable). There is a newer AGMA spec (I think it is AGMA 2015 which is quite similar to the ISO 1328 standard but it is more complicated tan AGMA 2000-A88).

 

[tbuelna]

israelkk-

You are correct in that it is better to use the newer AGMA radial and tangential profile standards. And the new AGMA standard requires each gear documentation to explicitly define things like profile and face tolerance limits (slope, fullness, hollowness, etc). This approach means that a CMM can be used to inspect a gear. But when you have a 9T 64DP gear, the tolerance limits one must specify to ensure proper meshing with a high performance drive can be too tight to manufacture reliably. The OP asked about a variation of .002" in PCD, which is quite substantial in relation to a pinion having a .14" PD and a .039 WD. It would also be difficult to get an accurate measurement of the tangential profile of a gear having just a .063" face width.