Selecting the Number of Teeth on a Planet for a simple planetary set 4

[Spurs]

How is it possible in a simple planetary gear set that the tooth count on the planet gear can be adjusted plus or minus one tooth without affecting timing while trying to maintain equally spaced planets.

As an example, if you had a sun gear with 18 teeth and a ring gear of 60 teeth, and you wanted 3 equally spaced planets, why is possible (according to AGMA 6123) to choose the number of teeth on the planet of either 20, 21 or 22 teeth and not affect the timing or efficiency of the mesh. I would think in such a case 21 teeth would be ideal, and selecting 20 or 22 teeth would result in an unequal load sharing (and hence efficiency loss).

Cany anyone shed some light as to why this is the case

 

[tbuelna]

Spurs,

With a 3 planet simple epicyclic, one of the members should float so that the planets will load share. The +1/-1 tooth counts are adjusted for using profile shift. If the profile shifts are not too extreme, timing and efficiency should not be impacted much. In fact, even if it's not required for assembly, some profile shift in epicyclic gears can be helpful for equalizing bending stresses or producing more efficient recess-type contact action.

Hope that helps.
Terry

 

[Spurs]

Terry
Thank you for your response. Are you saying that by changing tooth counts on the planets + or - 1 tooth, that timing on the entire planetary is not affected at all? Or are you saying that the effect is minimal?

Can there for example be a slight decrease in sound quality using a -1 tooth approach. My suspicion is that timing may be slightly affected.

 

[mraabe]

If all numbers of teeth are a multiple of the number of planets there are the same contact conditions on each planet. Therefore the variation of stiffness is larger as for shifted positions.
For the purpose of reducing noise and vibration it could be better to avoid this. The variation of stiffness would be smaller.

 

[Spurs]

Mraabe

In this simple planetary set, the sum of the sun and ring gear teeth is a multiple of the number of planets. That is not the issue. When discussing + or - 1 tooth on the planet, what I mean is without changing the number of teeth on the sun or ring gear.

So do you mean that in such a case, by shifting the planet + or - 1 tooth, we get a variation in stiffness that is larger than by using +0 and hence it will be noisier?

If it increases noise, why do people do it?

 

[mraabe]

If number of teeth for sun and ring are fixed, I see the main reason in changing the number of teeth for the planet to influence the sum of profile shifts for both pairs.
Changing the center distance you get the same sum of profile shifts for both pairs (with opposite sign).

For example if you have a very small number of teeth for sun and planet you need positive profile shift for both of them. This can be achived by an increase of center distance and will result in a large negative profile shift for the ring. Decreasing the number of planets by one can resolve this.

What I meant before is not the sum of sun and ring gear teeth, but that both of them are multiples of three. The sum has to be, so you can mount under 120°.

 

[Spurs]

Yes I understood what you meant about the sum are multiples of three - and that is not a problem.

So from a profile shift point of view, dropping a tooth is done to give you a +ve profile shift for both the sun to planet and for the planet to ring gear meshes. I can see some advantage to that aspect, but how does dropping a tooth not result in a shift away from perfect timing among all 3 planets simultaneously with the sun and the ring gear?

Is this done for strength at the expense of noise?

 

[tbuelna]

Spurs,

The condition you are describing is referred to as "non-factorized" and is described in detail in AGMA 6123-B06 sec. 6. I don't know what your definition of timing is, but adjusting tooth counts and using profile shift won't affect timing in the conventional sense. It will only change the overall gear ratio and have an effect on torque ripple.

A non-factorized gearset will tend to produce less noise due to the more random meshing between planets. It has the same effect as increasing contact ratio in a single mesh.

Hope that helps.
Terry

 

[gearguru]

If you only change the no. of teeth of the planets (sometimes called pinions) in a simple planetary the overall ratio does not change.

 

[mfgenggear]

Gearguru
are planet gears considered like idlers that is why the # of teeth do not change ratio?


Mfgenggear

 

[Spurs]

tbuelna

If you look at the meshs in a planetary in CAD, I notice that when the number of teeth on the planets is +0, the involutes line up properly -backlash is all to one side.

When you go +1 tooth on the planet, I cannot get perfect meshing action in the same way. Lining up planet A with the ring gear and sun gear so the flanks kiss results in a condition where planets B & C when kissing the sun gear, have very slight interference with the ring gear flank.

Is this why you get more noise in a +1 system?

 

[gearguru]

Mfgenggear,
You can think of the planets like if they were idlers.
There are methods for calculating the ratio of the planetary drive in any configuration (depending on which part is locked and what is input/output).
In all of these calculations the no. of teeth of planets is mathematically eliminated. In better gear books you can find not just the final formulas for determining the ratios but also how the formumulas were derived.
For example in the older (1988) ANSI/AGMA 6023-A88 standard "Design Manual for Enclosed Epicyclic Gear Drives" there are the calculations on pg. 8, 9.

 

[tbuelna]

Spurs,

As I noted in a previous post, with a simple 3 planet epicyclic it is common practice to "float" one of the members (usually the sun gear or carrier). With 3 contact points, the floating member will naturally assume a stable radial position and indexing, giving equal loads at each contact point.

You can simulate this in your CAD layout by moving the carrier axis off center from the ring/sun gear and rotating each planet on its axis until there is contact at each mesh's drive side teeth. Or the same effect can be achieved by shifting and rotating the sun gear.

Hope that helps.
Terry

 

[Spurs]

Terry
Your comment about 'floating' a member makes sense. Many years ago I have found that by putting more clearance between the planet bores and the planet pins, the system becomes quieter and more efficient. Logic therefore told me that by doing that, a force balance can be achieved.

Is that the logic to the + 1 or -1 tooth on the planet? My CAD shows in pure nominal geometry that a +1 or -1 tooth planet (in a fixed system) is not totally conjugate simultaneously on all 3 planets. By floating, I can see that the planets 'self align'. I assume that this self aligning cause some extra sliding action whereas in a +0 conditon it is pure rolling.

If however you had no floating (or if the float is not adequate) is it correct to say that using a number other than +0 on the number of teeth in the planet is not conjugate?

I do n0t agree with gearguru's point that you can consider the planet to be an idler. An idler in a 'conventional' sense does not have timiing implications the way that a planet has in a multi-planet arrangement in a planetary system. I believe that it is an over simplificaiton. The CAD certainly also proves my point.

 

[gearguru]

Spurs,
For the calculation of the ratio in a simple planetary we can consider the planet as an idler. It does not change the ratio, does it? I do not talk about anything else, only about the ratio in the simple planetary.
For more knowledge about the planetary design I recommend the standard mentioned in my previous post or some specialized book on this topic.

 

[tbuelna]

Spurs,

If you define conjugate action as no relative contact sliding, then most any profile geometry mods would not produce a true conjugate action. But there are many types of gear meshes that don't have true conjugate action, and they still work very well. The most obvious example are spiral bevels. Involute geometry gear teeth are also relatively tolerant of changes in center distance, that's what makes them so nice.

Regards.
Terry

 

[Spurs]

So to sum it up; adding or dropping a tooth on the planet can be done (witout affecting ratio), and it may function within the tolerances of the system, but it may be at the expense of some efficiency and possibly noise.

 

[mfgenggear]

Spurs

It was a very interesting subject.

 

[carburized]

I think you know from reading above posts that sumof teeth is divisibale by 3 in case of three planets. What I think you are reffering to profile shift. Usually in planetary planet has a negative shift, because it is in double bending.Look at the link example.

http://www.gearsandbearings.com

 

[Spurs]

Carburized
Yes of course we are talking about a system where the sum of the tooth counts on the sun and ring gear are equaly divisible by the number of plantets.

The discussion is really whether it is possible to have proper timing simultaneously on all planet to sun and planet to ring gear interactions when the number of teeth on the planet has dropped or added one tooth and been adusted for by profile shift.

My conclusion based on CAD data and practical experience is that the action is not as smooth when a tooth is dropped.