Drivetrain Stiffness

[Phil1988]

Hey guys,

got a question about the drivetrain.

How stiff are your axles? Whats the maximum torsion angle you calculated? Why do so much teams use axle made of full material? We use a big tube with thin wall thickness! Is it of the rotating mass?

regards Phil

[mech5496]

I think it's a matter of rotational inertia. Plus on a tubular halfshaft, you need to build separate end sections which you need to weld on...think about welds, post heat treatment and fatigue. Even CF tubular ones (with bonded end sections) suffer from increased inertia due to fairly large diameter.

[Racer-X]

I'm not drivetrain, so I don't know how stiff the half shafts are. What I can tell you though is we go with the solid units because they are simple and don't break. The drivetrain guy can make them in a day in the shop.

I'm not sure how much rotating inertia you save though, as the the the wall thickness gets thinner the diameter gets larger. I think most the gains are in overall weight of the car. Half shafts are low on our list of things to take weight out of.

So the positives are:
-Strength
-Stiffness
-Less design time (one round of FEA)
-Less manufacture time

Negatives:
-Weight
-Less cool factor

[mech5496]

I'm not sure how much rotating inertia you save though, as the the the wall thickness gets thinner the diameter gets larger.

There is a point where inertia actually INCREASES despite the decrease in WT. What I do not get...why stiffness should be a priority in halfshaft design? As long as the material does not yeld, they can twist as far as they want to. Just my 2 cents...

[Paul Achard]

After building an equivalent masses table (i.e. converting rotational inertia to an equivalent linear inertia considering speed reductions), I found that rotational inertias of most driveline components downstream of the final drive are negligible compared to their mass. The only components you should be worried about in that respect are the tires, wheels, your diff if it's not radially compact and components upstream of your transmission. The angular inertia of the driveshaft is negligible in any case.

I haven't studied drivetrain compliance in depth, but think about two things:
-What happens when left and right sides have different stiffness?
-Flex would affect the response time of torque transfer to your driven wheel and can potentially damp this transfer. Is this a bad thing?

Discuss.

[mech5496]

I haven't studied drivetrain compliance in depth, but think about two things:
-What happens when left and right sides have different stiffness?
-Flex would affect the response time of torque transfer to your driven wheel and can potentially damp this transfer. Is this a bad thing?

Discuss.

Not a drivetrain guy here, but I know from experience that the above translates in "torque steer". Same happens if you have two equal diameter halfshafts but with different lengths. Actually one of our halshafts (the longest one) is slightly stiffer on purpose, so they both have the same angular distortion during loading. Any thoughts on that?

[dmacke]

Our team uses halfshafts with "matched" spring rates. This reduces the effects of torque steer. There is a thread on here that discusses the topic of matching spring rates of halfshafts. There is debate on whether the driver will notice a difference but on the other hand it's easy to do and once less thing you have to defend in the design review.

[Owen Thomas]

I'm thinking the effect of relative half shaft stiffness ("torque steer") would be almost completely soaked up by a torque biasing differential. Granted, it would lower the overall torque bias ratio, but I would have to suspect this is less than a 2nd order effect.

Regardless, I guess if you're running a spool and are designing/manufacturing your own half shafts, it's worth the small effort (as dmacke mentions).

[Flight909]

I have question about this, specific for the electric teams. How do you determine the required stiffness for your drivetrain? What parameters are you considering.

I thinked about this from a traction control perspective.

[Francis Gagné]

I did some TC simulation for an electric system. My advice for you is to estimate if your drivetrain stiffness (Motor to Tire) is going to be significantly higher than your tire longitudinal stiffness or not. If the stiffnesses are in the same order of magnitude you should probably consider it in your model and see the effect it has on the slip response.

In both cases you should also verify that your TC system doesn't excite any mode of the drivetrain.

[Dunk Mckay]

If you've got a huge difference in half shaft length you might think about look into stiffness and rotational inertia. Our shafts for the past 2 years have been 1mm different in length so not worth it. This year could be the same, or they could be as much as 10mm difference, on a 3500-400mm shaft it's just not worth it.

As far as solid vs tubular we go tubular, but only in a simply way. We restrict the out diameter to the inner diameter of the splines (as defined by the CV's we use) and then take out material until they hit our target strength/stiffness. We then make them out of stupid strong steel. The only difference in manufacture over solid steel is that we gun drill down the middle.

For same strength shafts, solid ones would have an OD of 19.1mm. But increase that to the maximum 19.5mm imposed by the splines and you can bore out a 9.7mm hole in the middle! I know it sound insane, I had to do the math 3 times before I believed I'd done it right, and tbh I'm gonna check it again after posting this, now that I am older and "wiser". But that extra 0.4mm on the outside does a lot, and the middle section does next to nothing. Weight savings please.

[Phil1988]

Hey Guys,

thanks for your replys.

We used a tube with an outer diameter of 42mm an 1mm wall thickness. Calculated on 450Nm torque.

#I remember an value of 7deg of torsion from munic two years ago. They made a video with an high speed cam at accelerating.

But is it better to get a very stiff drivetrain?
I heard about things like to bias the drivetrain at acceleration and so on...

[murpia]

Originally posted by dmacke:
Our team uses halfshafts with "matched" spring rates. This reduces the effects of torque steer. There is a thread on here that discusses the topic of matching spring rates of halfshafts. There is debate on whether the driver will notice a difference but on the other hand it's easy to do and once less thing you have to defend in the design review.

Here are the relevant previous threads:

http://fsae.com/eve/forums/a/t...5607348/m/1466038355 (http://fsae.com/eve/forums/a/tpc/f/125607348/m/1466038355)

http://fsae.com/eve/forums/a/t...607348/m/26910051431 (http://fsae.com/eve/forums/a/tpc/f/125607348/m/26910051431)


Originally posted by Owen Thomas:
I'm thinking the effect of relative half shaft stiffness ("torque steer") would be almost completely soaked up by a torque biasing differential. Granted, it would lower the overall torque bias ratio, but I would have to suspect this is less than a 2nd order effect.

Regardless, I guess if you're running a spool and are designing/manufacturing your own half shafts, it's worth the small effort (as dmacke mentions).


I haven't changed my personal opinion in the last 8 years, which is that with a diff, unequal length / stiffness driveshafts are not a problem. With a spool I can see some issues.

'Torque steer' is not the correct terminology for the rear of the car. With sufficient toe stiffness it will not be a factor. Torque imbalance at the rear contact patches will create a yaw moment, but any imbalance due to driveshaft length / stiffness will be minor, and transient.

Driveline oscillations always have the potential to be a problem, and driveshaft length / stiffness will contribute to the fundamental frequencies at work here. So a possible 'fix' for an oscillating driveline with unequal driveshaft length / stiffness could be to equalise them, but that's not the root cause, just co-incidence.

Even with a spool (which has 100% torque bias potential) one concern would be oscillating yaw moments generated by tyre stick / slip & driveline resonance, shuffling from contact patch to contact patch and generally being nasty. It may be that unequal length / stiffness can help you here, by spreading the resonance peak out a bit.

Regards, Ian