Addendum Balancing and Center Distances of Multi Stage Epicyclic Gear Train 4

[RANorrisPE]

Hi all!

I've been doing some digging, and haven't been able to locate a resource to help with my question. I'm still new to epicyclic gear design, so thank you for your patience!

My company is working on a three-stage epicyclic gearbox with all three stages sharing the same ring gear. We are using GearTrax to help with the gear design, but I am running into a couple areas that I am uncertain about and no one within my company is either. I am profile shifting the addendum to thicken the teeth on the sun gears, and negative shifting all of the planets by an equal amount to maintain the center distance and to prevent undercutting. My problem comes at the ring gear - to maintain equal center distance as the sun and planet, I would have to add to the ring gear. I believe this may cause two problems, first that positive addendum of an internal gear may lead to more undercutting, and second, that each stage would require a different amount of profile shift.

As it currently stands, the center distance difference of the ring gear vs the planet/sun is about .30mm and 0.20mm on the first and second stage respectively (in the clearance/positive direction as in there is more "clearance"). If I force geartrax to use the planet/sun center distance, it calculates a contact ratio still in the 1.8 range for each. My colleagues are ok with this, but I have not convinced myself that this is an acceptable practice, but I cannot find an alternative.

Is there a resource or guideline or any guidance that anyone may have for this type of situation with planetary/epicyclic gearing in terms of how to best approach profile shifting, and matching of center distances, and the acceptability of using a different center distance than is recommended/standard?

Thank you so much for your help!

Cheers

Ryan

 

[gearcutter]

There are no rules, as profile shifting requires a life-time of experience to grasp.
Generally however, a more balanced specific-sliding ratio is what gear designers will look for with some designers looking for slightly more recess action than approach action.
Try to learn what these terms mean before attempting a design.

An annulus, more often than not, will have a minus correction. This is required to help reduce the risk of interference (non involute contact).
As far as I know; GearTrax will not alert you to this, so user beware!

The project sounds like an important one and it seems like you're trying to feel your way through it. I would highly recommend that you bring in a gear design consultant.
Or, get hold of as many gear design reference books and teach yourself the basics.

 

[israelkk]

As gearcutter mentioned internal gear interference with an external gear has other kinds of interference to care for. There are three three types of interference to be aware off:

1. involute interference
2. trochoid interference
3. trimming interference

You can read about it in details in SDP-SI METRIC catalog technical section.

http://sdp-si.com/D805/D805_PDFS/sections/technical.pdf

However, I believe this subject is too advanced for the frequent gear designer. Epicyclic Gearbox design is far more complicated than standard spur gearbox. There are issues of gearbox efficiency which are unique to epicyclic gear design. You can set the correct center distances and avoid all three interferences and get a very low efficiency gearbox.

 

[mfgenggear]

if master gears are designed to roll with each component it will test for interference. total composite & tooth to tooth composite check will provide inspection to verify the gears roll correctly. there is a calculation out there that can also provide the major, and minor diameter of the gears that there is no interference. and as said above the gears can not interface on non involute profiles. base on tight mess center distance and at the required center distance. as I have also learned here on this board there many other design criteria , beam strength, efficiency , contact ratio, vibration caused by interference is the enemy.

 

[tbuelna]

Ryan,

Take a look at AGMA A901-A92 which provides a good approach for optimizing gear drives. AGMA 6123-B06 is also worth reading since it provides lots of information regarding design of epicyclic gear drives.

 

[Windward]

You could avoid these problems with another kind of gearbox, the multi-branch parallel shaft type [MBPSG]. The input and output shafts are in line, the power density is comparable to the planetary type and the cost would be lower. I suspect that it would also be more efficient and reliable than the planetary type, but only testing would tell. At least it would probably be no worse on the last two points than the planetary type is.

No exotic gear designs are required. The MBPSG uses only standard gears and bearings and other conventional parts. One major advantage is that it does not use a ring gear, or any internal gears. Otherwise, the difference is solely in the gear arrangement.

Gearboxes of this kind are not available commercially, but I can show you how to build one. If you doubt that this technology is legitimate, see the MegaWind website:

http://www.cres.gr/megawind/Wind_Turbine_Gearbox.htm

The MegaWind investigators scarcely began to develop this technology. They show only two branch shafts when six is easily attained. That is why the power density is comparable to the planetary type. I could go into much more detail.