3 vs 4 satellite gear planetaries

[rmw]

Being an old boiler engineer, I once sat in a lecture by a renouned expert who said to never use a 4 roll tube roller on a boiler tube, because, as I remember it, if the tube had any ovality at all, and most have some, the two rolls opposite each other in contact with the short side if the oval would load up, while the other two rolls would carry no load at all, and for all practical purposes it would be like trying to roll a tube which was not restrained in a tubesheet at all. A 3 roll tube roller will center itself and all rolls will be in contact with the tube surface, overcoming the ovality. I hope I stated what I remembered correctly. (I don't want this to be a thread about tube rolling. I will post a thread in the boiler forum for that topic. This is a gear question.)

My question to the gear forum gurus is; does this same concept have any bearing on a planetary gear set with 4 planet or satellite gears versus one with three.

Here is the application. It is an automotive application with a planetary gear set in each wheel hub driving a device in the center of an axle. The sun gear is mounted stationary, and the shaft is increased in speed 6:1 by the wheel motion through the planetaries. The driven shaft is not constrained in any way radially at the planetary end, except by the planet gears themselves. In other words, the shaft at the planetary end is not bearing supported. The other end is splined into the driven device, which is rigidly held in conical roller bearings. The shaft is restrained axially, of course.

The shaft RPM can approach 3000-3600, and the torque output of the planetary gear set is, round figures, 1,200 ft-lb. OD of the sun gear is 8", planetary gears are about 2" OD, each, and the OD of the gear teeth on the shaft is roughly 1-1/2". These are approximate, and I do not know the actual pitch diameters. Sorry. If more information is needed, I can dig for it.

The axle carries beam load, so there is normal deflection during operation. (Remember the old civil engineering adage; 'if it doesn't deflect, it breaks.') So, there is a miniscule amount of angularity at times with respect to the axis of the shaft to the axis of the planet gears. (Wheel bearing clearances can also cause some angularity as well.)

I cannot get the old boiler experts advice out of my mind as a concern that a 4 planet gear planetary set up might get the shaft located between two gears, and possibly during a deflection "pop" the planetary. I can see in my mind the 3 planet gear set up always centering the shaft, and equalizing the load to each planet gear, but I cannot mentally visualize a 4 gear set up doing it.

I have to make a recommendation to a client as to whether or not to go with a 3 or a 4 planetary set up. The 4 planetary set up is touted by the manufacturer to be more rugged than the 3, but I have my reservation based on what I have described. Is there any basis to my concern????

rmw

 

[diamondjim]

What you state makes sense but
because the system should have
backlash built into it to help
overcome your concern. It will
wear in very quickly and the 4
will be much stronger as the gear
manufacturer suggested and also
reduce the tooth loads which may
be the greatest advantage.

 

[rmw]

Diamondjim,

Thanks for your response.

In the backlash situation, I can visualize the shaft gear equalizing itself among the three planet gears, and finding a natural center.

I can't visualize this with the 4 gear set up, and am concerned that it might try to center itself (instantaneously at the point of a sudden load surge) between two of the 4 gears, and overload those teeth.

The claim to robustness is made with regards to previous problems with the sun gear, some of which in the older 3 gear planetary, I have seen rip out individual teeth. (the sun gear has about 65 teeth.)

rmw

 

[EnglishMuffin]

One way this is addressed sometimes is to use flexibly mounted planets. For a discussion of load sharing in planetaries, you might find the following of interest: thread406-59082

 

[rmw]

EM,

That is an interesting read, and yes, I must confess that I did not search the site before posting.

I can see from reading the thread that I have mis-used some terminology. This axle is not made in an english speaking country, and when I think in a technical manner about it and all its parts, I just revert to the other language, and do all my technical thinking in that language, (and metric as well) so I never have to use english terminology for the parts. The names of the parts in the English parts manual are translations and not always the best. (the stop that locates the shaft axially on one end, if translated directly from the mother language would come out in english as 'butt washer'. Fortunately, that one was caught and was changed to 'stop washer.' The word satellite in the title of the thread might be a bit of a give away.

The sun gear in my application is the floating shaft gear end, that is the output of the planetary, and what I referred to as the sun gear is really the ring gear (that is fixed). I would expect that a real good gear guy would have seen right through that, however.

So, it is ring gear fixed, rotating planet spider, and sun gear speed at 6:1 faster than the rotating spider which is mechanically fixed to the wheel hub in this situation. Plain thrust washers for what that is worth, and dual needle bearings that the planets rotate on.

And, yes, back to the other thread, I have seen the pitting on the ring gear faces, and have even had gears analyzed, and it was said to be pitting due to lubrication failure. This little planetary application can trash some well known brands of gear oil fast. This is lubed per factory recommendations with SAE 90 weight gear lube. I had to learn all about parafinnic and asphaltic bases in gear lube manufacturing that I did not really want to know.

After reading what I read in the referenced thread, I am still concerned with the 4 planet set up. What I have a mental picture of is like when a round grinding stone is put into a hole of a larger diameter than the stone itself, and as the stone begins to rotate with the drill motor, as it strikes the wall, it begins to ricochet off and sets up a very violent motion that is detrimental to the grinding process.

With a 3 planet set up, I stll see the sun gear shaft just centering itself in a 3 point stance, while with the 4 gear set up, I can see the shaft ricocheting around looking for a happy place to be, just to be dislodged by gravity on each wheel revolution.

Maybe I am making a mountain out of a mole hill.

rmw

 

[EnglishMuffin]

Well, as I said, that is why people sometimes use flexibly mounted planet bearings when there are more than three planets, especially on very high powered drives. Otherwise, it's mainly a question of how accurately you think you can position the planet shafts in the spider. There was another thread about planetaries where various guys had some useful things to say about this subject. Try a search.

 

[rmw]

EM,

I read a lot of the threads that the search produced for planetary gears. Good reading.

You are going to have to elaborate on "flexibly mounted" planet gears. The axle I am dealing with does not have such a thing, and I am having a hard time envisioning it.

rmw

 

[EnglishMuffin]

To get some idea what I am talking about, see the following for example:

US Patent 3964334 (actually, this looks a bit too flexible to me)

Timken have just come out with a new "Flexpin" bearing which has an integrated flexible shaft - currently being used in wind turbines. Similar idea. Maag (and others) have been doing this in their planetaries for a long time, but I think they have just adopted the Timken Flexpin in their latest designs.

http://www.solardaily.com/news/timken.html

http://www.timken.com/media/kits/flexpin1.asp

http://www.timken.com/media/kits/flexpin3.asp

 

[EnglishMuffin]

By the way, you will note that the bearings appear to be cageless. Roller riding cages don't work well in orbiting applications because of centrifugal force, as I think I remarked in one of those old threads.

 

[rmw]

EM,

Following up on your links and then googling the words, Timken and flexpin, I found a link that shows the system real well.

http://www.timken.com/hannover/flexpin/Flex_Pin_files/frame.htm

While this is informative, and interesting information, the only consolation I have at this time, regarding the system that I have to work with, is that there may be just enough clearance between the planet gear, the cage type needle bearings, and the pin to allow a little "flex", although it would not be much.

You have my curiosity up, now, and I will have to get the old dial indicator out, and see if I can measure the radial clearance of the planetary sets I have to work with.

The fits aren't real loose, but also aren't tight enough so that there is any type of a "press" or "force" fit in order to assemble/disassemble the gears and pins. The pins, however, are press fit into the spider on one end, naturally, and quite tight on the slip fit end. Obviously, I don't have the manufacturers tolerances, nor dimensions, and would have to reverse engineer the individual pieces to find out what they are.

I can make this observation; Wheel bearing clearance on this axle is critical, since the planetary set is cantilevered outboard of the wheel hub hanging on the wheel bearings, and I have seen worn planetary sets where the WB clearance was excessive, and there is always a distinct (and ugly) irregular wear pattern in the contact side face of the ring gear teeth, so that indicates to me that there is not enough "flex" to accomodate very much excessive WB clearance.

In all of this new information, I still haven't seen anything that would relieve my concerns about the self centering capabilities of a four gear planetary set, and the tendency I see for the unsupported shaft (sun gear) to 'hunt' within it's backlash for a true center.

I appreciate the help you have given so far. I have a lot better understanding of some of what is going on in the heart of the system. I am still open to input.

rmw

 

[EnglishMuffin]

Well, all I can tell you is that 4,5,6 or 7 planet designs are not unusual, but the 3 planet design is the most forgiving, especially if one employs only floating sun, ring or carrier (spider) designs. If you don't have perfect load sharing, whatever the number of planets, you could well get hunting effects. And excessive flexibility or clearance in the wrong place can screw up any design. Good luck with your investigations.

 

[plasgears]

The question of three vs. four planets may be resolved as follows:
Visualize that the chosen sun gear leads to the number of planets. A sun gear divisible by three suggests three planets. Further, the resulting ring gear needs to coordinate with the planets in their orientation in space. Designing for load sharing in planetary and alternate gear path systems needs to be designed carefully.

 

[EnglishMuffin]

plasgears: Actually, for a simple planetary with equi-spaced planets, it is only necessary that the total number of teeth in the (ring gear plus sun gear) divided by the number of planets comes out to be an integer. There are more complex requirements for other more involved arrangements, but as far as I am aware, none of them require that the sun gear have a number of teeth which is an integer multiple of the number of planets. But are you seriously suggesting that when designing a planetary one should decide on the numbers of teeth before one decides on the number of planets??? I cannot think of any justification for such a procedure other than a situation where one was forced to use existing components for some reason.