Help with gear material selection for gear bearing

[maxoverclock]

I would like to request some help in gear material choices. I have a lot more information but want to start with a summary of what I am putting together.

I am wrapping up a design for a gear bearing drive actuator prototype. It is a "gear bearing" based planetary gear reducer (

http://ipp.gsfc.nasa.gov/ft-tech-gear-bearings.html)

wrapped around a compact external rotor brushless DC motor (

http://www.patentstorm.us/applications/20080045374/fulltext.html).

The actuator is a proof of concept prototype and will be used to drive a rehabilitation robot.

The gears will be EDM wired to save money. It is also important that the sun and pinion gears are non-magnetic.

I designed the gears using Integrated Gear Software by UTS. Since the gear drive will be backdriven, the gears were optimized to be 50/50 approach/recess action for both the sun/pinion and pinion/ring meshes.

The stage one sun gear is input and houses the DC motor. Stage one ring gear is fixed, Stage two ring gear is output and stage one pinions and stage two pinions are concentric and locked together in rotation. Since it is a gear bearing there are no carriers and roller bearings are used instead of ball bearings. Stage one is 30dp with a 52t sun, (3) 26t pinions, and 104t ring gear. Stage two has a dp of 30.923 with 25t pinions and a 108t ring. Max input speed is about 3800 rpm with torque of about 6.2 lbf-in. The reduction ratio is 40.5:1.

The system is sealed so it can be lubricated with grease or oil. I had a set of gears wired out of aluminum and assembled them using carriers and bearing to check the gear design. see pic...

I was looking at using a a bronze alloy for the sun, steel or stainless for the pinions, and i am not sure about the rings yet. Looking for anyone that may offer some advice....Thanks

 

[tbuelna]

maxoverclock,

Just a couple quick recommendations:

You must have done some clever carrier design to get that compound planetary to assemble with those tooth counts.

I would not use wire EDM if your gears need to be accurate or have a good surface finish. The wire EDM will not give good gear tooth quality and the recast surface layer wire EDM leaves is very poor for a gear application. Hobbing or shaping would be much better and is not that expensive.

For the planetary stage in your picture, you might be better off using more planets with less face width if possible. Your planets look like they have a W/D aspect ratio of 1.0 or more. That's at the limit of what's normally acceptable with spur gears. Less face width would give a better face contact load distribution, especially with a planet carrier that is deflecting torsionally.

A full recess action mesh at the ring gear will give better efficiency, if that is important to you.

As for what materials to use, good gear design practice dictates that all of your gears should be designed to have equivalent fatigue lives in contact and bending. So select your materials and heat treatments accordingly.

Hope that helps.
Regards,
Terry

 

[maxoverclock]

Thank you....

Some more information. The picture is of a mock up assembly I made to test the gear design. For the actual actuator design the face width of the gears is .3in and the PD of the pinions are .783 and .867 for stage two and one, respectively. W/D < .4, is this within reason?

After researching my material options, I was thinking about making the whole prototype out of Nitronic 60. On paper the properties seem like a good fit.

For hobbing, would having a non-standard dp of 30.923 for the second stage be a problem. Would I need to have special tooling made?

Thanks, Brian

 

[maxoverclock]

Also, would adjusting the meshes to full recess action vs the currently balanced recess/approach action effect the backdrivability of the system?

regards, brian

 

[tbuelna]

Brian,

A W/D ratio of 0.40 is normally very good for spur gears. But what you need to watch for with planetary gearsets is torsional wind-up in the carrier structure. Your gears and torques are very small so torsional wind-up in the carrier will not likely be an issue.

Nitronic 60 is an austenitic alloy that cannot be heat treated, so you would need to specify what (cold worked) condition your raw material is when the parts are machined.

If you use grease lubrication or operate the gear meshes with oil in boundary conditions, you will need to make sure that your gear tooth contact stresses are always below the fretting limit for your gear materials and heat treatments.

Having a custom hob or shaper tool made is quite common and is normally not very expensive.

Good luck.
Terry

 

[maxoverclock]

Thanks again. I'll make some calls after the holiday and price alternate methods of making the gears. Right now cost is approx $1600 for the EDM job out of stainless or regular steel. That's for 6 pinions (2 sets of 3), 2 rings and one sun gear. I am curious to see how hobbing compares. Do you know of any skilled reasonably prices gear cutters on the east coast?

Brian

 

[israelkk]

Just curious, are you punlicly trying to infringe NASA patent of gear bearing? Hope you received their permission. If not, I suggest you contact NASA about royalties and legality before continuing. "Everybody piss in the swimming pool but not from the high jump".

 

[maxoverclock]

Please look at the patent I linked to in the first post. I am Brian, John Vranish (from NASA) is the original inventor of the gear bearing .....

I am in the process of fabricating the first fully functional metal prototype right now. The one in the patent images was fabricated out of plastic on our 3D systems Viper stereolithography machine....

brian

 

[israelkk]

Usually gear material should be tough and surface case hardened to 60-65 RC if they are for such use as aerospace.

There are special alloys such as the Nitralloy family specifically intended for nitriding or the AISI 9310, 8620 etc, intended for carburising.

The rollers in the gear bearing should also be fully hardened or case hardened to 60-65 RC because the are basically rollers of a bearing.

How are you going to produce an integrated gear with rollers if the roller diameter is the same as the pitch diameter of the gear?

It means that you can not use hobbing, grinding, etc.

In plastics it is easy because you inject them into a mold but with high end gears such as for helicopter transmission you will not be able to do that. Therefore, you will have to assemble the rollers (press fit?) on the gear but then it will be more difficult to keep concentricity between the rollers and the gear.

 

[israelkk]

Just wanted to add the Nitronic alloys are not high strength alloys. Therefore, you will not be able to exploit the compactness ability of the gear bearing arrangement. For example, a common gear made of AISI 416H or 15-5PH-H1025 may be smaller because the alloys are stronger. If galling is your concern you should use the lubrication and the alloys I mentioned. See AGMA Gear handbooks for gear materials.

 

[maxoverclock]

Gear and rollers have to be two parts for the system to be assembled. The rollers actually serve several functions. They radially locate the gears, serve as bearings, and also axially support the system. The gear bearing structure can inherently support some level of thrust loading. Hopefully, I will be able to get a better idea about how the different loading will effect its performance after the prototype is up and running. I am also designing the system to be sealed so it will definitely by lubricated by either grease or oil....

What I expect is that if the mechanism proves successful then a commercial version would need to be produced using the correct materials. I would also expect that these gear would be designed and by an expert in gears and materials. I am using a great software package that was generously donated by UTS for this project, but it doesn't replace experience so in the end I am trying to learn as much as possible and to fabricate the best prototype within my tight budgetary constraints.

I am worried that if I go down the road of hardened steel alloys then my costs are going increase by an order of magnitude due to having to add several additional processes: hardening, grinding, etc to get everything working. What I am hoping is that using NI60 will give me enough use out of the system to test its performance and try to find a company to take it to the next step.

Thanks again for the comments, Brian

 

[israelkk]

From 30 years of experience in the aerospace area it will be a wrong approach to neglect the technology aspect. For example, for the rollers function as roller bearings you need a bearing quality processes to achieve low friction, low heat generation etc. and high speed rotary motion.

See the ABEC classes to get the feeling how accurate bearings are (they are very accurate especially precision bearing with high ABEC numbers). The rollers in bearings need to the smallest as possible if high speed and low friction is needed. While in the gear bearing you are very limited in the roller diameter options. The rollers and balls of bearings are highly accurate.

For example, a grade 3 ball has a roundness of 0.000003" and diameter tolerance of ±0.000030". I have no knowledge of a gear technology that will give you composite tolerances so accurate. It is possible to reach such high accuracy in round objects but not in gears with tooth profile, lead etc.
,
This means that if you manufacture the rollers to bearing quality (and you should if you want high efficiency gear) then you will need to introduce backlash in the gears to avoid interference.

Note the gear movement is not purely rolling, they also have sliding between the teeth below and above the pitch diameter, and I am not sure how such sliding will affect the rolling motion of the rolls.

The fact that the roller doesn't rotate (as I understand they are integral part with the gear) with reference to the gear it is mounted on, may cause stickiness problems or at least limit the effectiveness of the rollers/races interaction as bearings.

To my opinion you need to combine the expertise of bearing design and development professional/company and gearing technology specialists.

 

[TVP]

Brian,

In general, stainless steels have problems with galling when used in mating components subjected to high contact stresses (gears, threaded fasteners, etc.). Lubrication considerably improves this issue, but does not always alleviate it. Depending on the desired/expected operating life for these parts, you may want to use Nitronic 60 for two of the parts, say the sun and the ring, and use a copper alloy for the pinion. Nickel aluminum bronze (UNS C63000) is a high strength wrought gear alloy that is one possibility. Continuous cast aluminum bronze (UNS C95400) is another popular gear alloy with somewhat lower mechanical properties. Both are available in bar, so you should have no problem finding a distributor. Here is a Thomasnet listing of bronze bar suppliers:

http://www.thomasnet.com/products/bronze-bars-3256005-1.html

 

[maxoverclock]

I remember reading somewhere that when the aluminum bronze alloys wear that the particles can become very abrasive. how would this compare to 932 bearing tin bronze?

thank you,

brian

 

[unclesyd]

If your sun and pinion gears have to non-magnetic here is some information brochure for Nitronic 60 with it's galling and wear properties.
Stay away from the HS Nitronic 60 if possible as the in the process of cold working some of it's anti galling properties are lost.

http://www.hpalloy.com/alloys/descriptions/NITRONIC60.html

I would also look at the SAT12 Process for surface hardening austenitic stainless steel from Swagelok.

http://www.swagelok.com/services/sat12.htm

 

[TVP]

The tin bronze alloys are the best with respect to wear against mating steel gears. They are lower in strength than the previously mentioned alloys, so you would need to take that into account.

 

[maxoverclock]

i read some more on the ampco 45/63000 bronze. it has really good wear characteristics against N60. much better than N60 to N60. Thanks for the suggestion. Brian