14 replies to this topic

[Greg Locock]

 

Hat's off the best demo I've seen for a long time.

[gruntguru]

Very cool. It even goes backwards. I suspect at speed it would need different control parameters when going backwards.

[desmo]

At speed on a bike, you need to countersteer and use weight shifting to steer. I'd like to see this go well beyond walking speed where this effect kicks in.

[gruntguru]

It seemed to be able to take corners OK - at low speed but still requiring noticable lean.

 

Actuall just reviewed it. Cornering at 9:35. Only a slight lean so perhaps he hasn't incorporated a velocity x steering angle term in his algorithm yet. Should be relatively easy compared to what he has achieved so far. The flywheel roll stabiliser (plus the steering effect) can still simply target an "apparent vertical" condition - even through corners.

Edited by gruntguru, 23 June 2021 - 02:43.

[desmo]

No, I meant that the steering dynamics of bikes changes significantly at speed. It's very different from the dynamic of slow walking speed we see in the video.

[Fat Boy]

At speed on a bike, you need to countersteer and use weight shifting to steer. I'd like to see this go well beyond walking speed where this effect kicks in.

This is steering the bike more like a car. A bike, be it pedal or engine driven, corners much more on the rear tire than the front when done properly. The turn is initiated with the handlebars and rider weight shifts, but, once initiated, it's much more a product of lean angle than of handlebar angle. Using the front tire to turn is just a good way to low-side for us who lack professional level skill.

[gruntguru]

No, I meant that the steering dynamics of bikes changes significantly at speed. It's very different from the dynamic of slow walking speed we see in the video.

Not sure if I agree with this. Sure the input amplitudes and response rates are very different but I think the strategy is the same.

 

The only major difference I can think of is that pedalling torque becomes a significant input at low speeds but the reasons for that only relate to maintaining enogh speed for the steering inputs to control roll.

Edited by gruntguru, 23 June 2021 - 21:05.

[desmo]

There is a ton of controversy about bicycle dynamics. It may have been fully solved, but maybe not. Tire dynamics and profiles might be confounding variables.

[gruntguru]

It is almost unbelievable that that article ignored the obvious source of stability in the riderless bike at speed. Gyroscopic torque steers the front wheel left when the bike falls (rolls) to the left. This steering to the left tends to roll the bike to the right and counters the fall.

Edited by gruntguru, 23 June 2021 - 23:01.

[Bikr7549]

It is almost unbelievable that that article ignored the obvious source of stability in the riderless bike at speed. Gyroscopic torque steers the front wheel left when the bike falls (rolls) to the left. This steering to the left tends to roll the bike to the right and counters the fall.

 

I believe that at least one experiment (or series of experiments) disproved this, but I am not sure if this test included a rider or not. In Bicycling Science (Wilson and Whitt) it was reported that a bike was built and ridden that had 'wheels' rotating in parallel but in the opposite direction to the bikes rolling wheels, which would have negated the gyroscopic torque. I am not at home now but will look it up when I get back. 

[gruntguru]

Yes I've seen a video but I think it was debunking the idea that gyroscopic torque resists the roll directly. I don't think it considered gyroscopic torque applying a steering input.

 

Ah - here you go. Discusses both plus lots of other factors. http://www.phys.lsu....59no9p51_56.pdf

Edited by gruntguru, 24 June 2021 - 03:31.

[Ross Stonefeld]

No, I meant that the steering dynamics of bikes changes significantly at speed. It's very different from the dynamic of slow walking speed we see in the video.

 

Terrible mountain biker here, can confirm. 

[Greg Locock]

I'm feeling lazy, has anybody seen plots of fork angle and frame lean vs time for different speeds and radii?

Ditto for hands free riding?

[gruntguru]

Have a look at the link above.

[Greg Locock]

Apologies, what a great article.