Chainless Mecanum Drive

[Ether]

Originally Posted by Ether

Frankly, I don't see where nonadrive is simpler

Just to be clear: For a fair comparison of the code required, you have to compare apples to apples. Mecanum requires no special operator intervention or programming to actuate raising or lowering of wheels. To be functionally similar to this for comparison, nonadrive requires extra operator input(s) and extra code to respond to those extra operator input(s) to raise and lower the traction wheels and control whether or not to power the center wheel. To be functionally equivalent (same driver interface and same robot behavior), nonadrive requires extra logic to react to the normal joystick axis driver commands and decide automatically when to raise or lower the traction wheels (for example, when strafing or tight turning is commanded).

This is not an argument that mecanum is "better". Just a comparison of the code involved. The code for either is fairly straightforward.

It would be interesting to hear from teams that have fielded successful nonadrive robots. Did you add extra input(s) and associated code for the driver to manually raise and lower the traction wheels, or did you add logic to process the "normal" driver commands and let the robot automatically handle this decision in order to keep the driver interface simpler, or did you design a "hybrid" compromise with manual operator input(s) plus some automatic behavior?

[Tom Line]

Originally Posted by Andrew Schreiber

Actually, I would make the opposite argument. Mechanically a Nonadrive (one that does not need to cross bumps) is only slightly more complicated than a standard 6wd robot. Remember, the perpendicular wheel does not need to raise or lower if it is a flat field. From a programming standpoint it is much much simpler. The hardest part about mechanum drive machines is controlling them. It involves either lookup tables and interpolating results or using lots of sins and cosines or matrices (Thank you Ether). A nonadrive is simple in that for translation all you need to do put the Y component of your translation stick to the left and right drive motors and the X component to the cross drive motor(s) Rotation is a little trickier but you could just scale the left and right based on your rotation input. Either way, much simpler than mechanum.

It should also come out near the weight of a mechanum drive train. Remember mechanum needs 4 gearboxes, Nonadrive needs 3. Yes Nonadrive requires 4 cylinders but if you were planning on using air anyway that is not a ton of weight.

I will admit that the Nonadrive has, at least after a cursory glance, more failure points.

TLDR, there are benefits and drawbacks to both systems (like any systems) but neither one is inherently more complicated.

EDIT: Appended note about Ether's solution.

One item to consider with the nonadrive is that even on flat ground you will VERY likely want your center wheel to still be actuated by a pneumatic cylinder.

There is a reason stools usually have 3 legs. Having a 4 legged stool is more difficult. From experience (we attempted a solid 6 wheel rectangular drivetrain in '08 with all omnis), having your corner wheels and the middle wheel all touch if they are fixed is nearly impossible. This is why you must have pneumatics - if you rock on the center sideways drive wheel you will end up going in circles. It needs to push against the floor, but not with enough force to actually support the robot. That means some sort of suspension is required.

We used a laser to get our frame and 6 wheels straight, then realized that the floor many of the arenas was imperfect as well. We didn't much want to go back and engineer suspension, so we dropped the system and went standard 6 wheel.

It looked like this:

.......Wheel.........
........................
W....................W
h.....................h
e.....................e
e.....................e
l......................l
........................
W....................W
h.....................h
e.....................e
e.....................e
l......................l
........................
.......Wheel.........

We saw several teams using this same drive train this year, and they all seemed to experience the same problems we did during our prototyping before we threw it out - with no suspension the robot reacts differently depending on which wheels are touching the ground (Obvious when looking at it, perhaps not so obvious when designing it).

[Siri]

Originally Posted by Tom Line

One item to consider with the nonadrive is that even on flat ground you will VERY likely want your center wheel to still be actuated by a pneumatic cylinder.

There is a reason stools usually have 3 legs. Having a 4 legged stool is more difficult. From experience (we attempted a solid 6 wheel rectangular drivetrain in '08 with all omnis), having your corner wheels and the middle wheel all touch if they are fixed is nearly impossible. This is why you must have pneumatics - if you rock on the center sideways drive wheel you will end up going in circles. It needs to push against the floor, but not with enough force to actually support the robot. That means some sort of suspension is required...

This gave us major steering problems back in Overdrive as well. (In Lunacy we went with it.) That whole "3 [noncollinear] points define a plane" thing means extra contact points need to be somewhat forgiving. As pointed out, life isn't perfect even if your chassis somehow is. How you address that is up to you, whether suspension or actuation (or even somehow in driving), etc, but I would recommend addressing it. Not that this would necessarily add an unacceptable (or even particularly significant) level of complexity.

[spiffyspleen]

Originally Posted by Ether

12.75:1 seems to be a reasonable compromise between speed and torque (more or less depending on the game, of course)

What ratio do the Andy Mark gear boxes come in? We have several assembled from last year so would we actually have to change them?

Also, I am new to mechanical and am not too familiar with gearboxes. What does 12.75:1 actually mean?

[Ether]

Originally Posted by spiffyspleen

What ratio do the Andy Mark gear boxes come in? We have several assembled from last year so would we actually have to change them?

They are 12.75:1

Also, I am new to mechanical and am not too familiar with gearboxes. What does 12.75:1 actually mean?

It means that for every 12.75 turns of the motor shaft, the gearbox output shaft turns once. You are trading speed for torque.

[apalrd]

12.75:1 is the default gear ratio for the AM toughboxes and toughbox-derived gearboxes (Nano, etc.)



12.75:1 is the ratio between the input (in this case the output of a CIM motor) and the output (The output shaft sprocket or direct-driven wheel). With it you can find the output speed and torque of a gearbox if you know the speed and torque of the motors going into it.

If you need the long output shafts (you might not) you would have to get new long output shafts from AndyMark and put them in the existing toughboxes.

Edit: Beaten to it

[spiffyspleen]

Thank you guys! It looks like doing the mecanum drive is going to be a lot easier than I originally thought

[JesseK]

Originally Posted by Chris is me

@JesseK: I was suggesting 1 CIM + 1 FP in each of the forward wheels and then just a CIM on the side wheel since strafing is secondary to forward motion for a slide drive (if it isn't secondary, you should probably be using a holonomic chassis). With a crab module in the center I would probably suggest 2 CIMs on that and 1 CIM + FP on the outsides (off the top of my head here, not based on math or anything)

Many teams get away with 1 CIM per side for a 10fps (or less) drive train and do just fine. Calculating the motor load based upon robot weight and gearing inefficiency puts the motor efficiency within a nominal amount of its maximum efficiency while just driving. If sound judgment is used on the drive train with respect to turning (wheel base for traction, or use of omnis, 6WD drop-center, etc), almost the same results would apply. Thus, as long as the team doesn't try to push another traction robot they'll be ok.

Your initial instinct is correct for what many teams want -- 11+ fps.

Originally Posted by AdamHeard

You'd also have to design carefully to ensure that when the module rotates, the module rotates relative to the floor, not the module stays in place and the robot rotates around it.

Good point, I didn't even think about that... It'd almost have to be an omni wheel at that point in order to prevent this completely.

Originally Posted by spiffyspleen

Thank you guys! It looks like doing the mecanum drive is going to be a lot easier than I originally thought

The first iteration of Mecanum is always easy! If you build your drive train in the off season, play around with some extra weight. Specifically, see the effects of pushing a robot when the contact point removes traction from the front (lifts your robot slightly), and see the effects of just driving when your c.g. is too close to one of the wheels. Some interesting behaviors come out.

[Chris is me]

Originally Posted by JesseK

Many teams get away with 1 CIM per side for a 10fps (or less) drive train and do just fine. Calculating the motor load based upon robot weight and gearing inefficiency puts the motor efficiency within a nominal amount of its maximum efficiency while just driving. If sound judgment is used on the drive train with respect to turning (wheel base for traction, or use of omnis, 6WD drop-center, etc), almost the same results would apply. Thus, as long as the team doesn't try to push another traction robot they'll be ok.

I disagree, really. I can't think of any teams that enjoyed competitive success this year with a 2 CIM drivetrain, so I would hardly say "many teams" did. Who are some examples?

Even without pushing, running just 2 CIMs in a drivetrain even geared at 9 fps has many negative drawbacks. For one thing, at 9 FPS your CIMs will stall against a wall. Unlike a 4 CIM drivetrain, which just dips deep past the circuit breaker limits to get the wheels to slip, a 2 CIM drivetrain can't ever slip the wheels, so pushing other robots becomes basically not even an option. Turning is also pretty slow since skid steering relies on wheel slipping. Acceleration is also hampered a bit: In terms of actual distance traveled from a dead stop with 0.5 second of acceleration, a 2 CIM drive goes only 2.6 feet, while a 4 CIM drive goes 3.1 feet. (This effect is a lot more noticeable / dramatic in reality than "on paper").

If pushing is of zero concern for whatever reason (though, in my opinion there has not been a single FIRST game other than 2001 where at least being able to resist pushing was vital), it would make more sense to use 2 CIMs with ball casters supporting some of the robot's weight. This lets you draw less current while driving the same speeds, since your wheel force is lowered.

[Ether]

Originally Posted by Chris is me

it would make more sense to use 2 CIMs with ball casters supporting some of the robot's weight. This lets you draw less current while driving the same speeds, since your wheel force is lowered.

Could you articulate this argument in a bit more detail? It's not entirely clear what you are saying.

[Chris is me]

Originally Posted by Ether

Could you articulate this argument in a bit more detail? It's not entirely clear what you are saying.

By using a drivetrain where roughly half of the robots weight is supported by non driven wheels, the normal force on the driven wheels decreases. This puts the motors under less load at the expense of traction and pushing power. Generally a bad thing, but if pushing isn't called for at all it can be a good idea, since it lets you gear your drivetrain for a higher speed than otherwise possible.

[Ether]

Originally Posted by Chris is me

By using a drivetrain where roughly half of the robots weight is supported by non driven wheels, the normal force on the driven wheels decreases. This puts the motors under less load...

Robot A has 1 CIM+gearbox on each side, which drives the front wheel and the rear wheel via a belt or chain.

Robot B is identical to Robot A in all respects, except that the belt (or chain) drives only the front wheel. (The rear wheels are not driven).

How would your analysis apply to this scenario?

[Chris is me]

Originally Posted by Ether

Robot A has 1 CIM+gearbox on each side, which drives the front wheel and the rear wheel via a belt or chain.

Robot B is identical to Robot A in all respects, except that the belt (or chain) drives only the front wheel. (The rear wheels are not driven).

How would your analysis apply to this scenario?

Assuming a perfectly balanced CG:

With only two wheels connected, only 50% of the robot's weight is "over" the driven wheels (the other 50% of the weight is born by the non driven wheels which don't contribute traction since they are essentially just big bearings). For this case, the robot's max pushing force is effectively half (the mass portion of the formula to overcome static friction is halved). Since it requires less force / torque to slip the wheels, the motor connected to the wheel does not need to draw as much current, since the motor "reaches" the necessary torque at a lower amperage. However, the robot can only push with half as much force.

With only 2 CIMs in the drivetrain, a 148 pound robot geared to 9 FPS with roughtop tread (1.3 CoF) has a max pushing force that is torque limited. The motor's stall torque is greater than the torque necessary to slip the wheels. In addition, stalling a CIM draws 133 amps of current which quickly trips the robot circuit breakers. By halving the pushing force, you can make the drivetrain traction limited at some amperage. One can do this by lowering the traction of their wheels or by lowering the robot's traction overall by not driving a set of wheels / casters that support the robot weight.

[IndySam]

Originally Posted by Chris is me

I disagree, really. I can't think of any teams that enjoyed competitive success this year with a 2 CIM drivetrain, so I would hardly say "many teams" did. Who are some examples?

I'm with Chris, we will never do another 2 CIM drivetrain on a carpet field again.

It's really (strictly from my experience) a mater of drive-ability. Turning acceleration and other things that make for a good, easy to drive robot all suffer without the torque you can get from 4 motors.

[Madison]

We have competed with only 2 CIMs for drive with only 2 CIMs on our drive.

In that time, we've been a regional finalist, regional champion twice over and division finalist at the Championship. We're not a powerhouse by any means, but with the proper application of resources, it's certainly possible to find success with a less powerful drive.