differential design

[Luiz Henrique]

I am designing a differential for my university's FSAE vehicle and have some doubts. First thing is that I got the CAD modeling of the torsen model that is sold to FSAE teams and I have verified that the large gears (attached to the semi-axles) are helical gears about 48 mm diameter with 15 teeth. I'm using MITcalc for the calculations and I can´t get even close to those dimensions (it indicates pitching failure). I am using an input torque of 900 Nm, which means a torque of 450 Nm at those gears. I am not understanding how can it be. I was thinking about maybe using a design similar to the wavetrac (http://www.wavetrac.net/technical.htm), that I know that have a large load capacity because is used in trucks and stuff. In fact, I found this post in the ford raptor forum (http://www.fordraptorforum.com/f5/my-take-wavetrac-differential-12856/) and you can see a picture of a guy holding one of the gears and, man, they're small!
So I'd like to ask what do you think about that and if anyne has any advice. Sorry about grammar mistakes, but I'm brasilian.

[Paul Achard]

May I ask why you're designing your own differential gearset? Are you looking to obtain a specific bias ratio? What do you mean by "gear pitching failure"?

[Luiz Henrique]

In fact, the idea is to design an active limited slip differential, so the advantage would be the controlled slip. It was our professor who came up with the idea. I also think that buying the university special is a much more reliable option. Nevertheless, I learned a lot so far and am really enjoying the designing process. By pitching failure I meant contact fatigue.

[Paul Achard]

That is a cool project to pursue. It seems to me a salisbury-type diff would a more convenient baseline design for your application... Straight-cut gears are much easier to source and the basic design already incorporates variable stiffness due to clutch pressure.

As for the gears, I'm no expert, but I know you usually deal with the high contact stresses by case-hardening the parts. Are you accounting for that?

[Luiz Henrique]

To be honest, I started designing straight tooth gears, but at the end I saw that the whole assembly (excluding the housing) would be about 140 mm diameter. That is too much compared with the torsen FSAE model and would not be competitive. The fact is that the higher load capacities you can achieve with helical gears are determinant in this case where the rotational inertia is so important... I don't know, man. There's something about this torsen design that increase the load capacity and I am not seeing. Thaks for the help anyways.

[Paul Achard]

I think it might help you to put things in perspective:
If you're worried about the diameter to reduce the rotational inertia for acceleration purposes, I would set up an equivalent mass spreadsheet where you "transform" the rotational inertia to an equivalent linear inertia, taking into account reduction ratios between your component and your wheels. I think you will find that the diff contributes the equivalent of much less than one kg to the mass of the car, only due to inertia.

Secondly, question yourself as to the real goal of your project. If it is to gain points through vehicle performance, I could think of lower fruit to pick than the problem your are currently attacking.

If it is to gain points through the design event, I think you have more to gain through a simple yet effective and validated design than the "optimal" design which will be a headache to manufacture and might not work at all.

If your goal is just to make something awesome à la WWU, then by all means challenge yourself with a complex design. I wouldnt expect to win a competition that way however.

If your goal is to please your advisor... remember that this a student engineering competition, not an advisor ego-trip.

[Luiz Henrique]

Those are really good advice and I will certainly take in consideration. Thanks a lot, man!

[Z]

Originally posted by Luiz Henrique:
I don't know, man. There's something about this torsen design that increase the load capacity and I am not seeing.
Luiz,

Are you sharing the torque amongst the MULTIPLE pairs of teeth in the gearset?

That is, planetary gears (as in diffs) have 2, 3, or more "planet gear teeth" that mesh with the output gear teeth. This is why these arrangements can carry very high loads for their overall size. By comparison, conventional "parallel shaft gearsets" only have one input tooth pushing against one output tooth...

Z

[M. Nader]

the only advice i can give on the subject is to take good care of the concept of "recirculating power in planetary systems" as this is a problem some overlook.

[Luiz Henrique]

Originally posted by Z:
Luiz,

Are you sharing the torque amongst the MULTIPLE pairs of teeth in the gearset?

That is, planetary gears (as in diffs) have 2, 3, or more "planet gear teeth" that mesh with the output gear teeth. This is why these arrangements can carry very high loads for their overall size. By comparison, conventional "parallel shaft gearsets" only have one input tooth pushing against one output tooth...

Z

Yes, Z. I took that into consideration. Nevertheless, I still cannot get a set of gears as small as the set in the torsen university special. Besides that, I found out that that model have a torque capacity of 5200 Nm, which is much more than the value I used in my calculations. I think the load capacity calculations of the geras must not be made the conventional way. Don't know yet...

[JT A.]

Originally posted by Luiz Henrique:
To be honest, I started designing straight tooth gears, but at the end I saw that the whole assembly (excluding the housing) would be about 140 mm diameter. That is too much compared with the torsen FSAE model and would not be competitive. The fact is that the higher load capacities you can achieve with helical gears are determinant in this case where the rotational inertia is so important... I don't know, man. There's something about this torsen design that increase the load capacity and I am not seeing. Thaks for the help anyways.

The helical gears serve another important function besides higher load capacity. As I understand it, the helix angle of the gears in a Torsen type differential is what creates the outward thrust when a wheel starts spinning. The output shafts are thrust against a friction plate which causes the diff to lock up & reduce the wheelspin. If you used straight cut gears I don't think the diff would function as a limited slip.

Someone correct me if I'm wrong about that, that is just how I think it works from taking one apart & playing with it.

[scotty young Taylor Race]

JTA....

you are correct...

http://www.taylor-race.com/pdf...ng_differentials.pdf (http://www.taylor-race.com/pdf/understanding_differentials.pdf)

[Luniz]

Hi Luiz,

some assumptions first: "Pitching failure" in MITCalc relates to some kind of fatigue (maybe you actually meant to say pitting?), and you are using the maximum input torque for your calculations, is that correct?

In that case I wouldn't bother... Just design your gearset to withstand these loads statically (strength at tooth root and sufficient compression strength through material properties) and it should be fine.

In the automotive industry, gearsets are calculated using a load spectrum, so the gearset is not constantly loaded full bore... This is why your reverse gear on your passenger car might only do a few miles before it is broken! Usually it doesn't get used very much, and even if so, not in high speed.

[Markus]

Originally posted by JT A.:The helical gears serve another important function besides higher load capacity. As I understand it, the helix angle of the gears in a Torsen type differential is what creates the outward thrust when a wheel starts spinning. The output shafts are thrust against a friction plate which causes the diff to lock up & reduce the wheelspin. If you used straight cut gears I don't think the diff would function as a limited slip.

Someone correct me if I'm wrong about that, that is just how I think it works from taking one apart & playing with it.
That might be a type of Automatic Torque Biasing differential but Torsen differential works purely mechanically on the principle of worm-gears being one direction drive only.

[duanpeng]

Originally posted by Luiz Henrique:
I am designing a differential for my university's FSAE vehicle and have some doubts. First thing is that I got the CAD modeling of the torsen model that is sold to FSAE teams and I have verified that the large gears (attached to the semi-axles) are helical gears about 48 mm diameter with 15 teeth. I'm using MITcalc for the calculations and I can´t get even close to those dimensions (it indicates pitching failure). I am using an input torque of 900 Nm, which means a torque of 450 Nm at those gears. I am not understanding how can it be. I was thinking about maybe using a design similar to the wavetrac (http://www.wavetrac.net/technical.htm), that I know that have a large load capacity because is used in trucks and stuff. In fact, I found this post in the ford raptor forum (http://www.fordraptorforum.com/f5/my-take-wavetrac-differential-12856/) and you can see a picture of a guy holding one of the gears and, man, they're small!
So I'd like to ask what do you think about that and if anyne has any advice. Sorry about grammar mistakes, but I'm brasilian.

[duanpeng]

Originally posted by Paul Achard:
May I ask why you're designing your own differential gearset? Are you looking to obtain a specific bias ratio? What do you mean by "gear pitching failure"?

[allster]

Originally posted by Luiz Henrique:
I am designing a differential for my university's FSAE vehicle and have some doubts. First thing is that I got the CAD modeling of the torsen model that is sold to FSAE teams and I have verified that the large gears (attached to the semi-axles) are helical gears about 48 mm diameter with 15 teeth. I'm using MITcalc for the calculations and I can´t get even close to those dimensions (it indicates pitching failure). I am using an input torque of 900 Nm, which means a torque of 450 Nm at those gears. I am not understanding how can it be. I was thinking about maybe using a design similar to the wavetrac (http://www.wavetrac.net/technical.htm), that I know that have a large load capacity because is used in trucks and stuff. In fact, I found this post in the ford raptor forum (http://www.fordraptorforum.com/f5/my-take-wavetrac-differential-12856/) and you can see a picture of a guy holding one of the gears and, man, they're small!
So I'd like to ask what do you think about that and if anyne has any advice. Sorry about grammar mistakes, but I'm brasilian.

[William Porcher]

Cara, tu já tentou trocar o material da tua engrenagem? que material tu tá usando?

[PatClarke]

William,

English is the language used on this Forum

Pat

[William Porcher]

Originally posted by PatClarke:
William,

English is the language used on this Forum

Pat

I know man, is just i'm brazilian as is Luiz, i'm just making it easy for both of us...