Sliding spline design in linear actuator 4

[Aconcagua]

Hello - I am designing a linear actuator using a relatively large (60mm) roller screw. I am driving the screw and the output is attached to the nut. Obviously I need to oppose the torque that the screw imparts to the nut so that I get the desired linear motion/force. I don't have a lot of radial room though, so I think a linear ball bearing is out of the question. My thought was to use an involute spline that will allow the input and output to slide concentrically with each other, but will also prevent them from rotating relative to each other.

I am concerned about the spline design though. The travel is about 6" and the diameter is about 8". Can I make an internal spline that is very long (6" + length of the external spline), that will fit into my internal spline along the whole length? Will I have large binding concerns? Are there any issues with sliding during operation? There will be a maximum of 20,000 in-lbs on the spline, and it will be in an oil bath. The good thing here is that I am not concerned about a little slop in relative rotation between the two, so the tolerancing can reflect that. Also the sliding speed will be slow, around 0.5 in/sec.

For a reason of a minimum minor diameter on the internal spline, I have tentatively selected a spline of
Fillet root side fit
DP - 10
Teeth - 76
Pa - 30 degrees

Please let me know if you have any suggestions here, tolerances, coatings, etc. I don't have any experience with splines that slide against each other during operation, and I can't find anything about it in the machinist's handbook. Also if I moved this into a salt water bath, would that change this design?

Thank you,
Adam

 

[Tmoose]

So You need to slide while transmitting over 1000 lb-ft of torque?
How much axial travel?
How much thrust can you afford to lose to spline friction, and how much jumpiness in the output thrust is OK?

" Society of Automotive Engineers (SAE) Universal Joint and Driveshaft Design Manual "
$500 bux (!) on Amazon.
100 US bucks from SAE.

http://books.sae.org/ae-07/

http://www.gkn.com/landsystems/bran...s/Documents/GKN-PMD-MechanicsFlyer_GB_web.pdf

"> Nylon Coated Splines (Optional)"

 

[Aconcagua]

Thanks for the reply

Yes it needs to slide while opposing the torque. The 20,000 in-lb should be only to overcome initial stiction, after which it should drop by half at least.

The axial travel is 6 inches.

I can afford to lose a substantial amount of force due to spline friction. Thousands of pounds if necessary.

 

[MikeHalloran]

I think the 30 degree PA involute was developed and standardized so that external splines could be rolled into axle shafts, etc. at very high speed, like a few seconds per shaft.

I don't know if anyone is tooled up to roll the large, long splines you require.
It should be possible to cut them, and cutters may be available, but if you can't roll them, then I don't see any virtue to using such a large PA.



Mike Halloran
Pembroke Pines, FL, USA

 

[Aconcagua]

Well right now I need a long (~7") internal spline. There is a possibility that I could create the part in a way that allowed me to EDM those internal teeth. That would leave a short external spline. Is that your concern?

 

[mfgenggear]

&' LONG SPLINE IS NO PROBLEM FOR EXTERNAL, HOW EVER INTERNAL WOULD HAVE TO BE BROACH , 7 " IS PUSHING IT.

 

[MikeHalloran]

I don't think I'd want spline teeth cut by EDM; short/zero fatigue life, and you can only get 'good' finish by cutting even more slowly than what is normally a glacial process.

<tangent>
The Atomic Cannon's barrels were produced at Watervliet Arsenal, near where I grew up. I think its 38.5 foot long, 280mm bore barrel was rifled by a series of broaches pulled through by a cable, as were other smaller but still very large rifle barrels produced in larger numbers at the same facility.

According to unsubstantiated rumor from now-dead insiders, the manufacturing engineers could not figure out why the last broach in the series typically wore out faster than any of the others.

Everyone else on the production floor knew that, to save time or to meet unrealistic production schedules, only alternate broaches were used on each barrel; s/n 1 was broached with only odd numbered broaches, s/n 2 was broached with only even numbered broaches, etc., but every barrel saw the last broach, so it wore twice as fast as the others.

</tangent>
There's no substitute for "being there". You need to talk to broaching machine producers, and try to arrange visits to their customers' facilities to see the machines in action, and talk to the people who actually run the machines.


Mike Halloran
Pembroke Pines, FL, USA

 

[gearcutter]

6" of internal spline at that pitch & # of teeth is a relatively easy job.
Gear shapers that will cut up to 8" are easy to find, 10" somewhat more difficult and there are some shapers that will do up to 16" and higher...........not too many of those machines around though.

The most difficult thing when doing these long internal splines isn't so much finding a machine that can do the stroke; the problem is finding a suitable cutter and its mount. The problem gets worse as the number of teeth become less.
The 76 teeth for your job would be an easy job.

I think the 30 degree PA involute was developed and standardized so that external splines could be rolled into axle shafts, etc. at very high speed, like a few seconds per shaft.

30dg pressure angle may have been introduced purely for ease of manufacturing; but there is a very important reason why a higher pressure angle is usually better for a splined connection. Due to the higher-radial tooth loads the higher-pressure angle induces on the coupling, the better the coupling will self-centre under load.

 

[Aconcagua]

Thanks for all the advice! I will reach out to some gear cutting shops and see if they have that capability with a shaper.

As far as the pressure angle goes, are you suggesting I move to a 45 degree? Because of some diameter restrictions I have pretty thin walls on the internal gear. I care about concentricity, but also about limiting radial loads.

 

[gearcutter]

As far as the pressure angle goes, are you suggesting I move to a 45 degree?

No, I was not suggesting that.
You need to determine that yourself.
But check the increased hoop stresses on the female component and also be mindful of the resistance to axial sliding that higher pressure angles create.

 

[mfgenggear]

yes depending on the stroke of the machine that is the limiting factor. ad I am thinking of a US BS92.1 grade 5 Spline. even at 76 teeth, there is many factors that are necessary to include. like. what type of material. what type of heat treat, what hardness is required? shaping or broaching a soft part is no issue. heat treating and maintaining the spline geometry is an other issue. how many parts are being made, should gages be use to inspect. what is the pitch diameter run out required.
holding a .0003 lead at 7 inches is a little fun. .0013 index error. and p.d. run out of .0015 inch. good luck

 

[gearcutter]

holding a .0003 lead at 7 inches is a little fun. .0013 index error. and p.d. run out of .0015 inch. good luck

Yes all good points mfgenggear.

But the OP did say - "I am not concerned about a little slop in relative rotation between the two, so the tolerancing can reflect that. Also the sliding speed will be slow, around 0.5 in/sec."

It doesn't sound like an Aerospace application, so perhaps the tight tolerancing that you've quoted might not be so important given the OP's application.

Also, for ANSI B-92.1, 10/20DP, 76 teeth - my charts are showing me Total Index variation of 0.003" for class 5. That's not too difficult to achieve on a machine in good condition.
The OP is after an easy sliding fit under load; so class 6 would be more in-line with the requirement........that would allow up to 0.004" TIV.

 

[mfgenggear]

good points gearcutter star for you.

 

[tbuelna]

I agree with the comments about the face length limit for shaped internal spline teeth being the max stroke of the shaper. But another consideration is can the part be installed on the shaper while still allowing the full spline length to be produced.

Lead variation over the long length of the internal spline should not be a problem if you keep the length of the mating external spline to the minimum required for stress requirements.

 

[Aconcagua]

Thank you everyone for the help.

I'm thinking that the external spline will only be about 1/2"-1" in length. That is 10-20 times larger than it needs to be to take the torque, but I'm thinking it might be smart to not go too short so that the face to face compressive pressures don't get too high, increasing friction.

Is there any reason not to go with class 7 tolerance? Assuming I can handle the backlash. Is there a risk that by making the tolerancing worse, the load will be shared by fewer teeth, thus increasing friction, and decreasing concentricity?

 

[gearcutter]

Given the application; I would use a 'top fit' spline..............also known as a 'major diameter' fit spline. The tip diam of the shaft will have a neat fit with the major diam of the hub. This helps to keep the coupling halves centred and may also help to reduce the risk of binding due to any possible misalignment.
Use a spline standard & pitch (consider standards based on metric module) that is in-line with tooling your gear shop has available. The equivalent of 10DP would be 2.5 module.
Usually the shaft of a major diam fit spline will require some tooth tip chamfering to help avoid interference with the hub's spline root radius.

I wouldn't recommend a class 7 profile in this application.