Recommended Posts

Hello all,

I have a question regarding torque distribution in all wheel drive system. Maybe someone with better knowledge could help me. I cannot build right now, so I am not able to see if it works.

Let's say I would like to implement torque vectoring to the drivetrain - more torque to the rear wheels and less torque to the front ones. Am I correct that I should gear down the axle coming out from the center differential to the rear axle and gear up the axle to the front axle? Won't the wheels slip because of different rotational speed?

Thank you!

Edited by tomek9210

Share this post


Link to post
Share on other sites

I think the only way this would/could work is running an xl at the rear axles and a geared down L motor for the front, the rear would produce more torque but both would be running at the same speed. I know Brunojj1 used it in his Mercedes DTM car and it works really well, however he hard coupled the L-XL so I'm not too sure if it would work that well being entirely separate. I hope this helps and or makes some sort of sense. Unfortunately what you described  I can't wee working as wheels turning at uneven speeds would just cause binding, unless the differential would soak up the difference but I don't see that happening? Now I'm intrueged and I'll see if I can't try this later! 

Share this post


Link to post
Share on other sites

Well, when I did my AWD I made a mistake by gearing my front and rear axles differently, and I ended up destroying my center differential... so I guess that answered your question :laugh:.

Share this post


Link to post
Share on other sites
3 hours ago, tomek9210 said:

Hello all,

I have a question regarding torque distribution in all wheel drive system. Maybe someone with better knowledge could help me. I cannot build right now, so I am not able to see if it works.

Let's say I would like to implement torque vectoring to the drivetrain - more torque to the rear wheels and less torque to the front ones. Am I correct that I should gear down the axle coming out from the center differential to the rear axle and gear up the axle to the front axle? Won't the wheels slip because of different rotational speed?

Thank you!

yes:

you need this setup: a differential chase moved by your main motor, and the two differential output A and B (not connected between them so they can rotate independently) one for front one for rear, what you'll get will be that you need same torque one the same output "of the differential" this means that if you gear down A by 1:3 ratio, the other will have 3 time less torque than the first, but mind that the sum of both toque will be equal to the main motor torque minus the amount taken by friction, and since it is a differential  both output will tend to rotate at the same speed (when they have same friction, so like 4 wheel on the ground need to move 1/4 of car weight each) in a low friction condition the geared down part will slip "3 times easier than the other" because it actually have more torque.

 

But remember this is not really torque vectoring but only torque distribution (if i'm correct)

Share this post


Link to post
Share on other sites
5 minutes ago, PorkyMonster said:

 I ended up destroying my center differential... so I guess that answered your question :laugh:.

Thanks, that saves me from destroying one myself :wink:

Share this post


Link to post
Share on other sites

@PKW I think you're right, torque vectoring allows torque distribution to send more to the front or more to the rear and from side to side and vise versa but what you've described is only giving you higher torque in either front or rear and cannot change once selected. I guess what I described would be doing the same thing as it's not able to move around from wheel to wheel or front to rear.

 

 

Share this post


Link to post
Share on other sites

There seem to be a couple of different ways of achieving torque vectoring, but for your "mostly rear wheel drive with a bit of front" doesn't seem to be done anywhere without electronic control. The only mechanical torque vectoring i've seen is a Torsen differential, but it's aiming to act more like a limited slip differential than perpetual difference.

The idea of gearing up and down from one motor doesn't work because you have to end up with all axles spinning at the same rate - at least on a hard surface.  Friction losses aside, this means a purely mechanical system will end up with the same torque everywhere.

Modern TV units use mechanical clutches to allow torque distribution and/or they use braking systems to control not the engine torque delivered to the wheel, but the nett torque the tyre sees.

The last way is independent drive, which is only done (as far as i'm aware) by electric motors. There aren't too many production cars with more than one mechanically driving liquid fuel engine.

Your only way to achieve this in lego is for two motors that can be geared to the same output speed but have a different torque at that speed. An L and XL with half torque/double speed and one is geared by a factor of 2:1 will not achieve the output you're searching for. If you can get some useful data from Philo's site, you may be able to achieve something if the available motors are sufficiently different, but if the true electric motor inside all of lego's PF series is the same (and only the gearing internally is different) then you'll be out of luck.

Share this post


Link to post
Share on other sites

The easiest way: use 1 motor for the front wheels and 2 hard coupled motors for the back wheels.

10 hours ago, tomek9210 said:

Won't the wheels slip because of different rotational speed?

Yes if they don't have much grip - but if they don't slip, something is sure to break!

Edited by mocbuild101

Share this post


Link to post
Share on other sites
11 hours ago, tomek9210 said:

Am I correct that I should gear down the axle coming out from the center differential to the rear axle and gear up the axle to the front axle? Won't the wheels slip because of different rotational speed?

No, the central diff will compensate for the gearing up / down and effectively nothing will change. Only the diff will have a lot of unnecessary work to do.

8 hours ago, bonox said:

The idea of gearing up and down from one motor doesn't work because you have to end up with all axles spinning at the same rate - at least on a hard surface.  Friction losses aside, this means a purely mechanical system will end up with the same torque everywhere.

As far as I know, the only purely mechanical way to control torque split is inside the differential. Either with clutch-packs or with worm gears you can make an open diff close gradually once torque starts leaking to one output axle. These can be tweaked to have one output axle close (force it to rotate along with the diff-house) sooner (with less leakage) than the other. That way you can obtain an uneven torque split, like for instance 60%/40%.

Share this post


Link to post
Share on other sites
9 hours ago, bonox said:

Your only way to achieve this in lego is for two motors that can be geared to the same output speed but have a different torque at that speed. An L and XL with half torque/double speed and one is geared by a factor of 2:1 will not achieve the output you're searching for. If you can get some useful data from Philo's site, you may be able to achieve something if the available motors are sufficiently different, but if the true electric motor inside all of lego's PF series is the same (and only the gearing internally is different) then you'll be out of luck.

A M motor geared down 1:3 goes at pretty much exactly the same speed as an XL motor geared down 3:5, and the XL motor will have significantly more torque at those speeds.

Share this post


Link to post
Share on other sites

Real AWD cars with unsymetrical torque distribution usually have planetary gears as central diff (e.g. early BMW xDrive) because those can be combined with viscous couplings or clutch packs as central diff lock. Your lego version would work as long as you dont add a diif lock directly to the central diff housing, but you could use a clutch to lock the front and rear driveshafts after they were geared up/down.

Share this post


Link to post
Share on other sites
On 24/5/2017 at 0:12 AM, DugaldIC said:

@PKW I think you're right, torque vectoring allows torque distribution to send more to the front or more to the rear and from side to side and vise versa but what you've described is only giving you higher torque in either front or rear and cannot change once selected. I guess what I described would be doing the same thing as it's not able to move around from wheel to wheel or front to rear.

 

 


Sorry for the late reply, there are two options to compare our solutions:

if with your solution both motors will be connected together, the we will have a sort of electronic differential, so current may flow more in one motor than in the other (this is not correct but it is a easy-way explanation) so the speed of the two motor will "more or less" balance to the same, and will be helped by the gear ratio.

however if the two motors are, for example, powered by different receivers on different batteries then each motor will try to force the other to go at the same speed, with a significant loss of power

in my suggestion  (that was also the first idea of tomek) the central diff will always balance speed by acting on modifying the amount of torque on the rear axle, in the same way that does when you have a central diff and the front axle is on a slippy surface while the rear one no: it will give more speed to the front one and more torque to the rear one (but since the toque on the rear one is forced to be  equal to the resistance on the front axle you car won't move).

http://www.brickshelf.com/gallery/PKW/MISC/schermata_2017-05-27_alle_13.27.01.png

so let's examine the mechanism up here:

the differential would give all the speed to the 3:1 geared axle, but doing that the axle would be braked by the other one, so it moves from having 1/3 of the fricion to have the same friction but requires 1/3 of the torque to balance it, at this point is t clear than the 2 axles will move on the same speed but with different relative torque, until the external friction will become too low to keep the geared down axle to slip, at this point we return to the start position where all the speed will be given to the geared down axle, more over it is important to say that with friction x to the ground you need to have less than 3 times the friction on the rear axle to make it slip, but only less than the rear axle friction to the front one to make it slips
 

Share this post


Link to post
Share on other sites
10 minutes ago, PKW said:


Sorry for the late reply, there are two options to compare our solutions:

if with your solution both motors will be connected together, the we will have a sort of electronic differential, so current may flow more in one motor than in the other (this is not correct but it is a easy-way explanation) so the speed of the two motor will "more or less" balance to the same, and will be helped by the gear ratio.

however if the two motors are, for example, powered by different receivers on different batteries then each motor will try to force the other to go at the same speed, with a significant loss of power

in my suggestion  (that was also the first idea of tomek) the central diff will always balance speed by acting on modifying the amount of torque on the rear axle, in the same way that does when you have a central diff and the front axle is on a slippy surface while the rear one no: it will give more speed to the front one and more torque to the rear one (but since the toque on the rear one is forced to be  equal to the resistance on the front axle you car won't move).

http://www.brickshelf.com/gallery/PKW/MISC/schermata_2017-05-27_alle_13.27.01.png

so let's examine the mechanism up here:

the differential would give all the speed to the 3:1 geared axle, but doing that the axle would be braked by the other one, so it moves from having 1/3 of the fricion to have the same friction but requires 1/3 of the torque to balance it, at this point is t clear than the 2 axles will move on the same speed but with different relative torque, until the external friction will become too low to keep the geared down axle to slip, at this point we return to the start position where all the speed will be given to the geared down axle, more over it is important to say that with friction x to the ground you need to have less than 3 times the friction on the rear axle to make it slip, but only less than the rear axle friction to the front one to make it slips
 

640x490.jpg

I'm sorry, but I'm quite convinced that this does not give an uneven torque distribution. The central diff will completely undo the different gear ratios for the front output and the rear output, causing the whole setup to behave as a completely open diff. Torque vectoring withd differentials can only obtained with some kind of limited slip mechanism, which controls the amount of torque flowing from one output to the other output.

.

Share this post


Link to post
Share on other sites

Gentlemen,

thank you for your answers and explaining this topic to me.

Look here at Madoca's Pickup. There is a reduction in rear hubs, meaning that rear wheels get more torque than front wheels and their rotational speed is lower than front ones'. Is it compensated by the central diff? As we can see in the video, the model drives well, no slip occurs.

Share this post


Link to post
Share on other sites
1 hour ago, Didumos69 said:

 

I'm sorry, but I'm quite convinced that this does not give an uneven torque distribution. The central diff will completely undo the different gear ratios for the front output and the rear output, causing the whole setup to behave as a completely open diff. Torque vectoring withd differentials can only obtained with some kind of limited slip mechanism, which controls the amount of torque flowing from one output to the other output.

.

the point is that the differential is there to avoid different speed, also, we won't have more toque than the input at front but just more torque than the rear axle, in this cases the "limited slip" is the friction of the non geared axle to the ground (trough wheels), because as long as it doesn't slips the mechanism tend to give torque to the other wheel, this would give more speed to it but the differential and the fact that both wheels are mounted to the same chassis (so inserting a difference in speed will increase toque required for the faster wheel) prevent this.

i hope i can make a super small video to convince you
 

 

here is it, you can notice that only the geared down wheel slip (more torque) but the car move in a straight line that is the average of the speed of the two wheels (so actually a 1.5:1 ratio is applied to the output of the diff)

Edited by PKW

Share this post


Link to post
Share on other sites

The interesting thing about torque vectoring is, that its a misleading term as long as we use it to describe the rear axle mechanical torque vectoring diffs which can be found in Evos or Audis Sport Diff. Basically, they do not have the ability to 'send' more torque to a specific wheel than a classic Lsd or torsen could do, instead, their real advantage is to force both wheels to turn at different rotational speeds, just like a tracked vehicle. As the gearset of a TV diff causes the outer wheel to rotate faster, the maximum torque is even smaller as it would be without this device (more speed means less torque).

In terms of centre diff, torque distribution is a more precise term regarding the way it works and does not necessarily need any limited slip device to work. Adding such a devise would upgrade it to a variable torque distribution.

Share this post


Link to post
Share on other sites

Seriously, differentials are meant to kick in only when neccessary - e.g. when turning, when encountering different road traction, etc. and not ALL-THE-TIME...

So when the front and rear axles are driven at different speed (as opposed to same speed but different torque), they are forced to rotate at the same speed since they belong to the same vehicle, and the little gears within the center differential will end up working ALL-THE-TIME, which is very bad especially for Lego, because heat will build up, and with heat, plastic will soften, and torque + soft plastic is BAD... 

Share this post


Link to post
Share on other sites
16 hours ago, tomek9210 said:

Is it compensated by the central diff? As we can see in the video, the model drives well, no slip occurs.

So yes, the central diff compensates for the different gear ratios.

15 hours ago, PKW said:

the point is that the differential is there to avoid different speed

No, an open differential is unbiased towards the output ratio.

15 hours ago, PKW said:

i hope i can make a super small video to convince you

Sorry, still not convinced. I think the effect in your video has to do with friction.

5 hours ago, Lucullus said:

In terms of centre diff, torque distribution is a more precise term regarding the way it works and does not necessarily need any limited slip device to work.

Indeed, but this requires a planetary differential with different ratios for the two output axles. A 50/50 torque distribution implies that when one axle halts and the other slips, the rotation speed of the slipping axle is mutiplied by 100/50 (=2). This is the case with a normal diff, but also the case in @PKW's setup. A 60/40 torque distribution would imply that when one axle halts and the other slips, the rotation speed of the slipping axle is either multiplied by 100/60 or by 100/40.

5 hours ago, PorkyMonster said:

So when the front and rear axles are driven at different speed (as opposed to same speed but different torque), they are forced to rotate at the same speed since they belong to the same vehicle, and the little gears within the center differential will end up working ALL-THE-TIME, which is very bad especially for Lego, because heat will build up, and with heat, plastic will soften, and torque + soft plastic is BAD... 

This is also what I meant to say in an earlier reply. @PKW's setup will only will only lead to a lot of work for the diff.

Share this post


Link to post
Share on other sites

I found a solution to avoid permanent internal diff movement while keeping the uneven torque distribution. As a classic diff is nothing else than a planetary gear made with bevel gears, it is possible to swich positions of output and input. Something like this:

800x600.jpg

In this case engine input comes from the u joint, while rear axles output is on the right. As long as both axles rotate at the same speed, the bevel gears wont move, but there is still more torque sent to the rear output, so the rear wheels will start to slip earlier if both axles have the same amount of traction. I tested it and it works fine.

Share this post


Link to post
Share on other sites
Just now, Lucullus said:

I found a solution to avoid permanent internal diff movement while keeping the uneven torque distribution. As a classic diff is nothing else than a planetary gear made with bevel gears, it is possible to swich positions of output and input. Something like this:

800x600.jpg

In this case engine input comes from the u joint, while rear axles output is on the right. As long as both axles rotate at the same speed, the bevel gears wont move, but there is still more torque sent to the rear output, so the rear wheels will start to slip earlier if both axles have the same amount of traction. I tested it and it works fine.

You just locked the diff.

Share this post


Link to post
Share on other sites

Ok, my bad, i did not note the changing in rotational directions, its not locked but its a senseless construction (more like a complicated rear wheel drive). Please ignore it.

 

Edited by Lucullus

Share this post


Link to post
Share on other sites
5 hours ago, Didumos69 said:

So yes, the central diff compensates for the different gear ratios.

No, an open differential is unbiased towards the output ratio.

Sorry, still not convinced. I think the effect in your video has to do with friction.

Indeed, but this requires a planetary differential with different ratios for the two output axles. A 50/50 torque distribution implies that when one axle halts and the other slips, the rotation speed of the slipping axle is mutiplied by 100/50 (=2). This is the case with a normal diff, but also the case in @PKW's setup. A 60/40 torque distribution would imply that when one axle halts and the other slips, the rotation speed of the slipping axle is either multiplied by 100/60 or by 100/40.

This is also what I meant to say in an earlier reply. @PKW's setup will only will only lead to a lot of work for the diff.


i agre with what is there, and yes my mechanism lead to a permanent differential "work" but the results is that torque is not distributed equally to both axles and so you have "kind of more torque" to the front one, this is why it slips and the other stall because the dill gives more torque to that one.

and off course the effect has to do with friction, how can we easily measure torque if not putting it agains an equal friction and see what "win it" more easily?

also i don't know where that mechanism would be used, but i have to agree that is is not suitable for high speeds

but for a supercar model or some small rc veichle it should be ok, even for a torque distribution on a drifting car

Share this post


Link to post
Share on other sites
18 hours ago, PKW said:


i agre with what is there, and yes my mechanism lead to a permanent differential "work" but the results is that torque is not distributed equally to both axles and so you have "kind of more torque" to the front one, this is why it slips and the other stall because the dill gives more torque to that one.

and off course the effect has to do with friction, how can we easily measure torque if not putting it agains an equal friction and see what "win it" more easily?

also i don't know where that mechanism would be used, but i have to agree that is is not suitable for high speeds

but for a supercar model or some small rc veichle it should be ok, even for a torque distribution on a drifting car

Out of curiosity, I've built a simple model just to compare the difference in performance between same gearing ratio between front and rear axles, and different gearing ratio. What I've observed is that not only does the center differential work harder (that's a given, since all of us here agree with that), the resulting performance of the vehicle is worse-off than if both axles are geared the same way - which means, before 35 seconds, while the model can overcome the hump, it does so more strenuously compared to after 35 seconds.

The construct is rather brute-force :tongue: - just to test the idea. Here's the underbelly picture:

800x1067.jpg

 

Here's the video that compares the difference (before 35 seconds - different gearing ratio, and after 35 seconds - same ratio): 

So I think it is not worth the trouble to use different gearing ratio for front and rear axles... both speed and torque suffers for the whole vehicle.

Edited by PorkyMonster

Share this post


Link to post
Share on other sites
19 hours ago, PKW said:

and off course the effect has to do with friction, how can we easily measure torque if not putting it agains an equal friction and see what "win it" more easily?

I was referring to friction in the drive train.

17 minutes ago, PorkyMonster said:

So I think it is not worth the trouble to use different gearing ratio for front and rear axles... both speed and torque suffers for the whole vehicle.

Thanks for your experiment! Your observations may have to do with the following effect: When a diff needs to 'work' (has one output axle rotating faster than the other) under torque, the bevel gears inside the diff will push each-other against the diff housing, which will in turn cause friction. This would mean the diff isn't really open.

Share this post


Link to post
Share on other sites

According to manufacturers of cars with unsymetrical torque distribution, its main purpose is increasing the vehicles balance while cornering/corner exit. Therefore I think its hardly possible to observe any difference in performance of a lego car equipped with it (even if someone would be able to create a flawless lego representaion of the real system). PF and even buggy motors seem to be not powerfull enough to use such a systems potential. But if someones goal is to create a detailed representation of a vehicles drivetrain it might be worth to think about it.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Recently Browsing   0 members

    No registered users viewing this page.