This morning I did an informal experiment in power distribution with my car. Like many people in the Cold North right now, my driveway is a complete sheet of ice. The road, however, is actually down to pavement, or at least very sandy pavement. The car is a 2003 bmw 325xi wagon, 5 speed manual. Here’s the experiment: put some wheels on pavement, others on ice, let out clutch, add throttle, see what happens.

Please note there is no traffic on my road and anyway I’m at the end, so this was very safe. When I put my front wheels on the pavement / rear wheels on ice, I can get the rear wheels spinning pretty fast, the car moving only slowly forward, until all 4 wheels hit the pavement. But when I turn around and put the front wheels on ice / rear wheels on pavement, I can barely get any wheelspin on the fronts before the whole car is up on the pavement - the rear must grab and pull the car before the front has much chance to get power.

I have read that the power is distributed 60/40 or even 70/30 to the rear. Can anyone explain how this is achieved through the transfer case? Shouldn’t I be able to spin the front wheels on the ice nearly as much as the rears anyway? One difference is that I’m in reverse when trying to spin the fronts. Does that matter? I wouldn’t think so. A final note, I tried this with traction control on and off, with similar results. The only difference being the TC (while on) came in to apply brakes to the spinning rears in that scenario. Maybe I need to try this with the fuse pulled?

This does explain the general tendency of the car to behave like RWD in the snow, as far as ease of throttle oversteer. I should also try putting just one rear wheel on the ice. Hmm...