Help on helical gear design

[circuit]

I am working on a hub motor design with internal RC outrunner. It is going well, however I don't have much practice with gears. I would like to use helical gear to minimize noise. Problem is that helical gears have relatively large axial forces and simple bearing on planet gears seems to be not enough.
If someone here has some practice designing gears or has read some theory, could you point me to right direction?

 

[circuit]

This is how Peugeot did their helical gears:

This is a two-stage gearbox. Bearings only on input and output shafts... No bearing on middle gear. Shaft goes directly in a the hole in aluminum block. There are two cutouts for oil to get in and lubricate the surfaces. It also had a couple of crush washers on each end.
So is this all that is needed? On other hand, these 50cc engines require a full maintenance after few 1000's of kms, so probably gearbox is also not built to last?

 

[Punx0r]

I would look at photos of car gearboxes for inspiration (especially automatic boxes, which use planetary gears like yours). It's been a while since I worked on a car gearbox, but I recall a mixture of ball-race and taper roller bearings. Taper rollers will take all the axial thrust you want, but have their disadvantages. There are a number of different designs of ball race bearings that can take differing axial loads.

I imagine you will be able to calculate the thrust produced by a helical gear based on the input torque, gear radius and tooth angle?

 

[circuit]

I would look at photos of car gearboxes for inspiration (especially automatic boxes, which use planetary gears like yours). It's been a while since I worked on a car gearbox, but I recall a mixture of ball-race and taper roller bearings. Taper rollers will take all the axial thrust you want, but have their disadvantages. There are a number of different designs of ball race bearings that can take differing axial loads.

I've seen several photos and drawings. It is more or less clear on sun and ring, but not so much for planets. I've seen many pictures and drawings of planet carrier and it's crown, but nothing about what kind of bearings are they using. It is needle bearing, but is that enough? To my understanding, needles would see a very uneven load on their edges.
For example this is very simple, but only on straight spur gears:

I imagine you will be able to calculate the thrust produced by a helical gear based on the input torque, gear radius and tooth angle?

I have a theory about calculating that, but a confirmation would be great Don't really know how gear radius comes in.

 

[Punx0r]

Yes, needle rollers do seem very common for the planet gears and it is a good question about how the axial force is resolved. I imagine that helix direction of the sun and ring gears would be opposite, so one side of the planet gear is pushed and the opposite side pulled, cancelling out the axial thrust, but making the planet gear want to lean.

The gear radius should matter as the force on the gear teeth for a given shaft torque input should be dependent on radius. So a small diameter gear will have greater tooth force and so higher axial thrust.

I did a quick google search and this seems to be the case: http://books.google.co.uk/books?id=kotyj80OMCoC&pg=SA3-PA19&lpg=SA3-PA19&dq=calculating+thrust+helical+gear&source=bl&ots=uqMcFMW2dM&sig=pZncf9YA8LLyDgqfn6Kub7jIWg4&hl=en&sa=X&ei=MkJCVLv5EabT7Qbo8oGoBg&ved=0CCYQ6AEwAQ#v=onepage&q=calculating%20thrust%20helical%20gear&f=false

(Handbook of Gear Design by Gittin M. Maitra, Section 3.12: Force analysis for helical gears)

This paper confirms that there is a tipping force on the planet gears: http://neugartusa.com/Service/faq/helical-gears-in-planetary.pdf

I think you will find it interesting as it lists the advantages and disadvantages of spur Vs. helical planetary gear sets. It seems the effect of the tipping force in the helical set reduces the planet bearing capacity, making the gearset larger overall to take a given torque.

So if you want helical gears I think the answer is to use needle roller bearings, but derate them to take account of the tipping force (uneven bearing loading). You may be able to find a calculator on a gear supplier's website? Or perhaps calculate the tipping force using the thrust force on the teeth of the planet gear, multiplied by the radius to give force at the bearing centre line, divided by half the width of the bearing (wider bearing being better) and assume the whole thrust force (after account for leverage) is taken as a radial load by the very edge of the bearing (perhaps 1mm?). This would give you a radial load by unit width on the bearing and you must then find a bearing that can take that force over its whole width.

I.e. say force on planet bearing edge (1mm) is 100N. You must find a 10mm wide bearing with capacity of 1000N.

This is just a guess and would do some reading on the subject and try to find out the proper way to calculate it!

 

[Miles]

You seem to have plenty of room. I don't see why you need to use needle rollers? You'd need a separate thrust bearing if you did...

What helical angle are you using? Ordinary deep groove radial bearings will handle an axial load of half the static load rating. Wouldn't this be sufficient?

 

[spinningmagnets]

The greater the angle of the helical gears, the quieter they will run. However, as the angle increases, the forces that try to push the gears away from each other axially increases. On the BBS02, the primary reduction gear-set is clearly very helical. When I first wrote about that drive, I was certain that the secondary reduction were "straight-cut" gears, and since the secondary spins much slower than the primary, the potential noise wouldn't be very bad at all.

Later, I happened to see a picture from a more optimum angle, and there actually was a slight angle to the gear-teeth on the secondary. Much less than the primary, but definitely enough to classify as helical.

In a gear-set with a high angle to its helical gears, I have ocassionally seen an axial "thrust" bearing. It can be anything between a simple precision bronze washer, all the way up to a needle bearing. I'm sure there are many factors to take in consideration. Is the lubrication grease? oil immersion? oil "splash" lubrication? Of course the optimum for extended severe duty would be a pressurised oil system, probably overkill for anything you'd need here.

If immersed in oil, a bronze washer can take up the wear (the sides of the steel gears wouldn't wear), but if you go with an annual cleaning combined with a smear of high-temp synthertic grease...perhaps go right to the needle thrust bearing. They are actually quite affordable.

 

[peters]

I don't know much about this, but what about double helical gears?
http://en.wikipedia.org/wiki/Gear

 

[Miles]

I don't know much about this, but what about double helical gears?
http://en.wikipedia.org/wiki/Gear

There's the small matter of assembly, with a planetary set...

 

[Miles]

Another possibility? http://simplybearings.co.uk/shop/Bearings-Double+Row+Deep+Groove+Ball+Bearings/c3_7/p150718/3800B2RSRTVH+Rubber+Sealed+Double+Row+Angular+Contact+Ball+Bearing+10x19x7mm/product_info.html

 

[peters]

I don't know much about this, but what about double helical gears?
http://en.wikipedia.org/wiki/Gear

There's the small matter of assembly, with a planetary set...

ok, I see

 

[Punx0r]

Another possibility? http://simplybearings.co.uk/shop/Bearings-Double+Row+Deep+Groove+Ball+Bearings/c3_7/p150718/3800B2RSRTVH+Rubber+Sealed+Double+Row+Angular+Contact+Ball+Bearing+10x19x7mm/product_info.html

That seems good as it also claims to be good for tilting loads.

I figure only the sun gear will be subject to an axial thrust that the bearing must accommodate.

I don't know the capacity of needle Vs. roller, although I suspect the needle roller bearing can take more load. Durability is also another factor that has probably been discovered by experimentation and experience.

There must be good reasons that planet gears in commercial gearboxes normally use roller bearings. I think it would be good to determine what those reasons are to see if those reasons are applicable to this applications. For example, if the advantages for a commercial gearbox of using roller bearings are low cost, speed of assembly etc, we can ignore them. If the advantages are a smaller bearing, or two times the bearing life then this is worth paying attention to IMO.

 

[circuit]

Yes, tilting of planet gears is what concerns me. That would lead to uneven tooth engagement, uneven wear, etc.

If immersed in oil, a bronze washer can take up the wear (the sides of the steel gears wouldn't wear), but if you go with an annual cleaning combined with a smear of high-temp synthertic grease...perhaps go right to the needle thrust bearing. They are actually quite affordable.

Probably I will have an oil bath here. Bronze is one way to do this, but that would mean occasional maintenance. It's not like I will assemble and forget it, it's a DIY thing, means I will be playing around with internal anyway. But still would like as little problems as possible.

Another possibility? http://simplybearings.co.uk/shop/Bearings-Double+Row+Deep+Groove+Ball+Bearings/c3_7/p150718/3800B2RSRTVH+Rubber+Sealed+Double+Row+Angular+Contact+Ball+Bearing+10x19x7mm/product_info.html

I was actually thinking of something similar. Probably the best solution so far...

There must be good reasons that planet gears in commercial gearboxes normally use roller bearings. I think it would be good to determine what those reasons are to see if those reasons are applicable to this applications. For example, if the advantages for a commercial gearbox of using roller bearings are low cost, speed of assembly etc, we can ignore them. If the advantages are a smaller bearing, or two times the bearing life then this is worth paying attention to IMO.

Yup, that's the question. How do they do it in automatic gearboxes? I'm almost sure they use needle bearings. But never seen a disassembled carrier unit... These are often presented as "single piece" units.

 

[circuit]

Here's an informative thread:
http://www.dieseltruckresource.com/dev/beefing-up-my-a518-transmission-updated-occurs-t273018p2.html

So it looks like a steel/brass plate bearing is all that is enough for thrust bearing... But it needs a cut to get new oil in. Interesting. Wonder if these washers are possible to buy.

 

[Punx0r]

Interesting. It seems crude having a steel plate as a thrust washer, but thinking about it, the gear will be very hard and the steel relatively soft, and I imagine there is a constant supporting film of oil, as in a plain journal bearing. Speaking of which, it's interesting to see the sun gear has a plain bearing!