Scrub friction

[Mschmeh144]

Hello all
Just a question

We just ran the first test on our new robot, tote bot,which is set up for tank drive, and it seems to have a lot of scrub friction. I don't understand it too well, so I could use some pointers on how to reduce this.

[Mark Sheridan]

This is a great guide on this subject:
http://www.chiefdelphi.com/media/papers/1443

[asid61]

If you are using 4 traction wheels in a aquare or long configuration, consider changing the back wheels to omnis.

[GeeTwo]

Originally Posted by asid61

If you are using 4 traction wheels in a aquare or long configuration, consider changing the back wheels to omnis.

I'm not sure if a tote bot has a front/back orientation.

If the weight is not centered equally over the axles, change the wheels on the axle away from the center of gravity to omnis.

Better yet is to go to a 6 (or 8) wheel drop center configuration. This essentially turns a single long configuration into a number of wide ones.

[Ether]

Originally Posted by Mschmeh144

I don't understand it too well

Here's an updated explanation of the physics based on non-cleated wheels with front and rear chained together.

[Mr V]

Originally Posted by GeeTwo

I'm not sure if a tote bot has a front/back orientation.

If the weight is not centered equally over the axles, change the wheels on the axle away from the center of gravity to omnis.

Better yet is to go to a 6 (or 8) wheel drop center configuration. This essentially turns a single long configuration into a number of wide ones.

While that is conventional wisdom the actual physics are that it turns it into a situation were two wheels have a high normal force while the other two have a very low normal force. Note the assumption behind that is that the center of gravity is near the center wheels and not close to one of the end sets of wheels.

[GeeTwo]

Originally Posted by Mr V

While that is conventional wisdom the actual physics are that it turns it into a situation were two wheels have a high normal force while the other two have a very low normal force. Note the assumption behind that is that the center of gravity is near the center wheels and not close to one of the end sets of wheels.

This is true at constant speed. When accelerating, the normal forces increase on the wheels away from the direction of acceleration. That is, in acceleration, the weight goes onto the rear wheels, and in deceleration the weight goes onto the front wheels.

[Mschmeh144]

Here is the bot

[Oblarg]

Originally Posted by Mschmeh144

Here is the bot

You're going to need omni wheels on either the front or back to turn that thing reliably.

[GeeTwo]

For a 12x18" robot going to six wheels would be awful tight unless you went smaller than 4" wheels. I would recommend changing the two at the radio end to 4" omnis to minimize the scrub force resisting rotation, and to maximize the scrub force in resisting strafing motion, especially if it is applied near the CoG.

You may have some swaying motion when you go to two omni wheels, whichever end you put them on. You will probably have more swaying one direction than the other - experiment with this to determine which end should be the preferred "front". As I recall from our Ultimate Ascent 'bot, we did better to push from the solid wheels towards the omnis, but we were not driving the omnis, so YMMV.

I'm curious - you referred to this as a tote 'bot. How are you going to secure the tote to the chassis? One big hole or four little ones?

[Mschmeh144]

Originally Posted by GeeTwo

For a 12x18" robot going to six wheels would be awful tight unless you went smaller than 4" wheels. I would recommend changing the two at the radio end to 4" omnis to minimize the scrub force resisting rotation, and to maximize the scrub force in resisting strafing motion, especially if it is applied near the CoG.

You may have some swaying motion when you go to two omni wheels, whichever end you put them on. You will probably have more swaying one direction than the other - experiment with this to determine which end should be the preferred "front". As I recall from our Ultimate Ascent 'bot, we did better to push from the solid wheels towards the omnis, but we were not driving the omnis, so YMMV.

I'm curious - you referred to this as a tote 'bot. How are you going to secure the tote to the chassis? One big hole or four little ones?

This robot will have the tote sit on top of it, and connect to the frame of the tote using some sort of easy or quick connections,as I don't want to have to take parts off to remove the tote.there is no bottom to the tote we selected for the robot
We also added a set of omni wheels to the back of the robot. Which did give it some success. It was cruising around the lab this morning, until I floored it and reversed it too hard.
I underestimated the tourqe of the CIM motors we used. Resulting in a bent front axle. That part will be reinforced heavily to prevent damage like that ever again. Here is a photo of the front axle in its bent state