Live data from Hacker News

How a differential gear works (1937) [video]

youtube.com

51–59 of 59 posts

Re: How a differential gear works (1937) [video]

#51
post #50
post #46

Earlier quoted context omitted.

So, one wheel starts spinning, and then, at a certain speed, it gets immediately locked into the other wheel which isn't moving at all? How does that not blow up the entire differential?

With a hard-locking differential ("lockers", generally used off-road), engagement while one wheel is spinning is ill-advised. They're generally toggled on at low/zero speed by the operator in anticipation of a traction-loss situation. The variety of differential which does not immediately lock one wheel into another is http://en.wikipedia.org/wiki/Limited-slip_differential .

Yup. And they typically have to be unlocked when not needed as it is almost impossible to turn with the rears locked. However, with pneumatic or electrical operation from the driver's seat this is pretty convenient.

The type of locker I'm more familiar with takes a different approach: the rear end is locked by default but unlocks when it needs to. e.g., going around a turn causes the outside tire to turn faster than the inside one, and the teeth in the locker are angled so the speed difference causes them to cam out and decouples one axle shaft from the diff, letting it spin independently while the inner axle shaft is under power.

I had that system in my Toyota pickup and even in 2WD, it was absolutely unstoppable offroad.

Re: How a differential gear works (1937) [video]

#52
post #49

This is the best technical education video I've ever seen. If all information were available in such an easy to understand form, the world would be full of geniuses. The level of effort is astounding. The "support" part holding the spokes looks cast. They couldn't have possibly cast that part specifically for this video... could they?

casting one-offs with a disposable form (wax, styrofoam) isn't particularly difficult or expensive. They use big basins full of super-fine sand: a great insulator and the grains are too tightly packed for the liquid metal to flow into the tiny spaces between them. We used to do it in high school shop class.

Re: How a differential gear works (1937) [video]

#54

Wow, that was a fantastic video. I could understand it easily only because they iteratively added complexity to the models. It's amazing how well they have made the tutorial for the layman. Thanks a lot for sharing.

If you liked that, you might like this similar video about how a mechanical watch works:

http://www.youtube.com/watch?feature=player_detailpage&v=508...

Re: How a differential gear works (1937) [video]

#55

Is it possible to create a "negative differential gear" that puts more torque into the wheel that is stuck? (in the hopes of forcing it unstuck)

Yes. You are referring to a locking differential, of which there are several types. Some use clutch plates, and will work even if one wheel has no traction at all, others use more complex gearing and can transfer torque from one to another at some multiple (called a Torsen Differential) and cannot drive when one wheel is completely unloaded.

Awesome, thank you!

Re: How a differential gear works (1937) [video]

#57

I was generally not bad at physics but I remember not getting my head around this one... What a great lesson for decoupling systems. I wonder what's the software equivalent in design pattern land. The facade maybe ?

I never really got my head around electricity and magnetism. That stuff is very abstract.

Re: How a differential gear works (1937) [video]

#58

Wow, that was a fantastic video. I could understand it easily only because they iteratively added complexity to the models. It's amazing how well they have made the tutorial for the layman. Thanks a lot for sharing.

If you liked that, you might like this similar video about how a mechanical watch works: http://www.youtube.com/watch?feature=player_detailpage&v=508...

Enjoyed watching it. Thanks. :-)
Post reply on HN