How do differentials work-old school video

[SIMid]

We all know about diffs, but those of you who are not James May, will find this educational.

http://paultan.org/2009/08/12/how-do-differentials-work-old-school-video/#more-16110

[DirtyHarry]

nice...the way i learnt how a diff works was with "lego" and later in life pulling apart a diff at pick and pay much more blacktown.

[chrissn89]

thats pretty cool, simple to follow and educational yes.

thanks :)

[mania]

Great find, I have always had a vague idea on how they work, but that's the best explanation I've ever seen. So easy to follow. Where's the one for the atb lsds? I struggle visualising how they work from the diagrams..

[daaaaavek]

haha I love those old Jam Handy videos.

[poita]

lol love it!!

weld that shit up!

[Brazza]

great video as everyone has said very easy to follow...

Looking at the diff moving at the start of the video I was thinking there was no way I'd understand after watching it once but as he says the concept is really simple once explained.

[mmahle]

awesome video. very educational, makes sense. :P
Second questions, what do LSD's have to do with it? Or are differentials = LSD's? :P I googled it, but I didn't make much sense of it. hehe. Can anyone explain?

[ChrisMaz]

I love how they make the woman go on the easy wheel at the end :lol:

[mania]

awesome video. very educational, makes sense. :P
Second questions, what do LSD's have to do with it? Or are differentials = LSD's? :P I googled it, but I didn't make much sense of it. hehe. Can anyone explain?

I'll try and explain it ;).

A normal, "open", differential does two things:

It splits the torque 50/50 to each wheel.
It spins at least on wheel at the speed of the driveshaft going in to it. (Lets ignore for this that most difs have an effect on the final drive ratio..)

What this means is that under normal driving the car can go around corners with the two wheels going different speeds - which is desired. The problem arises though when one wheel can spin too easily.

ie, if you suspend one wheel in the air with the other on the road, the wheel in the air won't require much torque at all to spin it. Now because the torque is split exactly 50/50, the other wheel also gets not much torque - not enough to move the car. The same thing happens when the unladen wheel breaks free around a corner. It'll limit the amount of torque that can be sent to the wheel with more grip, as it takes comparatively little to keep an unladen wheel spinning.

For this reason an awd made up of three open differentials will have less grip then a fwd or rwd, because the traction is limited by which of the four wheels has the least grip.

An LSD attempts to solve this problem. Three main types that I know of (and my vague understanding of how they work):

Viscous LSD: Think of each wheel having a second connection to the driveshaft. Not directly connected though - it's broken in the middle, with a liquid filling the break. Under normal circumstances this broken connection does nothing. However when the speed of the wheel differs by a lot compared to the driveshaft, the fluid heats up. It's a viscous fluid - so that as it gets hot it'll become gooier, becoming more solid to bridge the connection. This limits the speed one wheel can rotate in relation to the other. Not very good for fwd's, as they tend to suddenly bite. Also tend to wear out over time. This is your standard rwd lsd.

ATB: Quaife's :). Rather then send 50/50 torque to each wheel, up to (in the case of Quaife) 4 times the torque will be sent to the wheel that isn't limiting the torque. Great for fwds. On their own terrible for off-roading, as if one wheel comes off the ground, 4 times not much is still not much. Hummer's solve this by applying the brakes on the wheel that's in the air - the differential will then send 4 times the braking force as a drive force to the other wheel. I have no idea how ATBs work.

Electronic LSD: Think Evos, and the future of difs. I believe similar to ATBs, but rather then the torque bias being managed by clever gear arrangements, a computer somehow manages it. Only problem is the mechatronics involved may take longer to respond then a purely mechanical solution, however with proper setup they will make better "choices" with what to do with the torque then any purely mechanical system.

Think that's about it. Longer post then I intended, apologies :o

EDIT: Does this make me James May?

[ChrisMaz]

From what I've just read mate it all makes sense, To my extremely limited mechanical knowledge, therefore, Richard Hammond :p