Planetary Gear Help! 2

[md0501]

Hello,

I am having some issues with a planetary gear system I am working on. All the calculations that I can find seem to be working without any interference. But as soon as I model the system it shows some interference. I used online guides to model the actual involute curve of the gear teeth. Is this a modelling problem or a gear problem?

My biggest concern is the lack of a profile shift. I could not find a mathematical reason to use one or how to calculate one, so that could be the issue but I am not sure. Can someone help me with this? I have the minimum of 17 teeth for no undercutting of the gears with a 20deg PA.

Module: 0.4
Pressure Angle = 20deg
Sun Pitch Diameter = 7.2mm (18 teeth)
Planet Pitch Diameter = 10.4mm (26 teeth)
Ring Pitch Diameter = 28mm (70 teeth)
Center Distance = 8.8mm

 

[mfgenggear]

md0501

quick observation of the picture,
how much backlash ? it appears none, and at tight mesh which is no go.
contact ratio,
the sun would require a + profile shift, and the planets a - profile shift.
did you run the numbers manually, and which specification, AGMA, Iso ?
purchase a gear program. and run the numbers again.

 

[md0501]

Hi mfgenggear,

Thank you for the response. I have not looked too much into the backlash, so I will have to run those numbers.
The contact ratio calculates to 1.58, which I figured was alright since it is over 1.2. When trying to calculate the profile shift, I do not get any hard values as to what it should be. Should I just start guessing with a small shift and go from there?
I have been using the SDP/SI "Elements of Metric Gear Technology" handbook, which I would assume is based off of some standard, but I am not sure which one.
Do you have any suggestions for gear programs?

Thank you

 

[mfgenggear]

MD0551

Here are two ANSI AGMA recommendations 6123, & 917-1
I believe Kissoft is one of the best, but very expensive, and a long learning curve.
take a look at gear trax as well

the gear programs will have recommendations on profile shift, and will give the safety factor results, & if there is interference.
looks at contact ratio, scuffing, contact stress, tooth bending. the root radius will be important,

what is nice about these programs is one can make changes and visualize instant changes,

 

[spigor]

md0501
Will you only have one planet gear?
I have found no reason for gear interference, so it looks to be a modeling problem.
Your planetary gear should work as it is, however, you might want to optimize it in some direction (capacity, noise, wear). Information on how to apply profile shifts are e.g. in:
DIN 3992 Addendum modification of external spur and helical gears
DIN 3993 Geometrical design of cylindrical internal involute gear pairs

 

[md0501]

Thank you for all the help pointing me towards the different standards!
I will be looking into Kissoft. We use Creo3 which Geartrax does not support.

spigor- I am planning on having 4 planet gears. I checked for interference between the planets and that they can be evenly spaced, and there are no issues there.

 

[3DDave]

From

https://camnetics.com/faq.htm

- GearTrax Standalone is available - it will generate STEP and IGES versions which you may be able to import to Creo.

 

[Windward]

"...I am planning on having 4 planet gears..."

With more than three planet gears, the arrangement is statically indeterminate and load predictions are unreliable.

It is about load sharing between the planets. It is never perfect in a planetary gearbox for at least three reasons: manufacturing tolerances in the parts and assembly, the effect of centrifugal forces on the revolving planets, and the impossibility of rigidly supporting the planet gears.

These problems exist even with only three planets. Even if the load balance were perfect in a new unit, it would deteriorate in service because of uneven wear.

There are methods for reducing the load imbalance, but they cannot eliminate it. One method is to increase the quality of the gears and other parts and the precision of the assembly, but this is expensive. Another is to use a floating sun gear, but this introduces problems of its own.

The best you can do is to live with the load imbalance after doing the most you can afford to reduce it. Obviously this problem has not prevented the wide use of the planetary design, but you must deal with it.

 

[gearguru]

If the planets in the planetary with 18T sun, 26T planet, 70T annulus have to be equally spaced, then there can be just 2 or 4 of them.

 

[tbuelna]

Great points about assembly requirements and load sharing for a simple epicyclic. ANSI/AGMA 6123-B06 section 6 provides some excellent guidance.
- The 18T/26T/70T configuration is non-factorizing for 3 evenly spaced planets and factorizing for 2 or 4 evenly spaced planets.
- The AGMA recommended planet mesh load factor for grade 4 or better gears without any flexible member is 1.00 for 2 or 3 planets, and 1.15 for 4 planets. With a flexible member it's 1.00 for 2 or 3 planets, and 1.08 for 4 planets.

 

[gearguru]

tbuelna: As I put it before: the 18T/26T/70T configuration will not assemble with 3 equally spaced planets.

 

[mfgenggear]

md0501
curiosity what is the torque and rpm requirements, to my opinion as well a 3 planet configuration is what I have always worked with.

 

[Windward]

Another consideration for the planetary design is that external loads can increase the load imbalance of the planets and reduce the life of the gearbox.

It would be difficult and expensive, maybe impossible, to straddle mount the input and output shafts of a planetary gearbox. Straddle mounting would also increase the length of the gearbox. This would reduce the power density, which is the only virtue of the planetary design.

Therefore, the planetary design requires overhung loads on short shafts. External loads on these shafts will cause greater misalignment compared to that of straddle mounted shafts. The greater misalignment will increase the lead error on the gears and reduce their life.

The greater misalignment will have no good effect on the load sharing of the planet gears either. It would result in constantly changing differences in the center distances of the planets as the they revolve. In other words, the planets would constantly assume and shed load, never taking an equal share. The designer must allow for the greatest load on the planets although they will periodically assume a lower load. The greater the load the planet gear must handle, the larger, or more expensive, or both, it will be.

Misalignment of the driving and driven shafts with the input and output shafts of the gearbox will impose these harmful external loads. Such misalignment is common, for instance in wind turbines. That is one reason wind turbine gearboxes have a reputation for early failure. One means for reducing the problem is to mount the generator directly to the gearbox, as in the hybrid drive.

A multi-branch parallel shaft gearbox of six pinion gears on the low speed end will have a power density equal to that of a three-planet planetary gearbox. All of its shafts will be straddle mounted. It will require a load-balancing mechanism, but that mechanism will be simple and will not require much room. It will also ensure perfect load sharing under all conditions. It is a mystery to me that such gearboxes are not the standard for high power density.