4 wheel ackermann steering?

[javitopia]

I've started a new project for a local line follower competition, and i was thinking if there was a good idea to implement an ackerman steering in both axles, since using only one is not enough for the turning radius i have in mind.

I'm using ackerman since the robots are fairly large (30x20, almost a 1/10 rc car, pretty bigger than the jap ones) with the track having an 8 shape, and my investigations suggest that the ackerman has the advantage here, but i don't know how a 2 axle ackerman steering would behave (and i couldn't find any meaningful info in the matter apart from a paper about 6 axles military trucks)

¿Any suggestions?

[Soeren]

Hi,

I'm using ackerman since the robots are fairly large (30x20, almost a 1/10 rc car, pretty bigger than the jap ones) with the track having an 8 shape, and my investigations suggest that the ackerman has the advantage here, but i don't know how a 2 axle ackerman steering would behave (and i couldn't find any meaningful info in the matter apart from a paper about 6 axles military trucks)

Not sure why you think that the Ackermann steering geometry should be favorable for a figure-8 track, but it will be expensive and a lot of work transferring the drive power to it via a differential, as the wheels will go at different speeds when turning, not to mention the work involved in setting up a proper Ackermann in the first place.

A much easier and just as useable (for your purpose) way, would be having the 2 (horizontal) axles on a sort of turntable (with the power going to one of them) and steer their turning in synch. That way, a common axle will be all required for each set, as all wheels will go at the same speed all the time, whether going straight or turning and the steering can be made by a single timing belt pulley directly on one (vertical) axle and via a separate (vertical) axle connected to the other through a 1:1 gear. An encoder disc on one sub-assembly can tell you how much you turn if needed.
There will be a limit to how hard you can steer of course, as when it approaches 90° (each sub-assembly 90° relative to the length of the vehicle), all wheels will be in line, with nothing to balance it - around - 45° turns should be OK, if your CoG is chosen with care and you don't go too fast.

With a 45° turn of each "disc", the turns radius (to the center of the vehicle) will be the axle distance divided by sqr2.
 With an axle center distance of eg. 20cm, the turns radius will then be a little over 14cm.

[javitopia]

Not sure why you think that the Ackermann steering geometry should be favorable for a figure-8 track, but it will be expensive and a lot of work transferring the drive power to it via a differential, as the wheels will go at different speeds when turning, not to mention the work involved in setting up a proper Ackermann in the first place.

that was wath people saying, altought there wasn't any solid evidence, so this could be wrong. We've opted for the ackerman because we want to try something new basically.

A much easier and just as useable (for your purpose) way, would be having the 2 (horizontal) axles on a sort of turntable (with the power going to one of them) and steer their turning in synch. That way, a common axle will be all required for each set, as all wheels will go at the same speed all the time, whether going straight or turning and the steering can be made by a single timing belt pulley directly on one (vertical) axle and via a separate (vertical) axle connected to the other through a 1:1 gear. An encoder disc on one sub-assembly can tell you how much you turn if needed.
There will be a limit to how hard you can steer of course, as when it approaches 90° (each sub-assembly 90° relative to the length of the vehicle), all wheels will be in line, with nothing to balance it - around - 45° turns should be OK, if your CoG is chosen with care and you don't go too fast.

With a 45° turn of each "disc", the turns radius (to the center of the vehicle) will be the axle distance divided by sqr2.
 With an axle center distance of eg. 20cm, the turns radius will then be a little over 14cm.

Nice idea, we might give it a try.

Thanks for the help

[Soeren]

Hi,

[...] That way, a common axle will be all required for each set, as all wheels will go at the same speed all the time, whether going straight or turning and the steering can be made by a single timing belt pulley directly on one (vertical) axle and via a separate (vertical) axle connected to the other through a 1:1 gear. An encoder disc on one sub-assembly can tell you how much you turn if needed.

Just a correction...
I may have made it sound like the wheels would go the same speed turning and that's impossible of course. The angle is what is the same for both wheels in a set, so either use a single wheel drive or a differential for the powered wheels.
Sorry if I confused things!