Hey all,
Our mentors and I had a question. Altough it is hard to fit the Mecanum on the KOP chasis, we want to check this posability. I know we can fit it in with 4 CIMple boxes connected to the chasis, one for each wheel. We only saw one problem with that, Can the mecanum wheels pass the scoring zone? We are afraid that because the Mecanum has no friction with the floor, it will slip. It is the first time our team thought about the possability to build a Mecanum drive, and we have no information about it.
Thanks
Good luck! and see you in the competition!::ouch::
Mecanum wheels certainly do have friction. VEXpro states the CoF of their mec wheels are 1.0 (on carpet). You might be overestimating how slippery the HDPE ramp is. You also might be thinking the rollers will roll sideways when driving forward and back. They don’t. Take a look at this whitepaper and many of Ether’s recent posts to learn more about mecanum drive and its characteristics.
In 2012, we were able to balance on the bridges made from the same material. They don’t behave perfectly on the surface, but I think you’re overestimating just how much of an effect it has.
Although the posts here are very helpful so far, and I agree with them, I seriously do not recommend building a drive train you have never even prototyped before. If you believe your team has the resources, time, and know-how to successfully build a new drive-train, by all means, take a stab at it. But if you don’t have very strong programming, I recommend against it.
However, a mecanum drive with directly-driven wheels is incredibly simple to build, and the programming for teleop omnidirectional drive is pretty much built right into the libraries.
A team we share space with tried driving over HDPE with a mecanum drive. It could drive straight forward with no trouble, and strafing over it wasn’t bad either. The main issue had nothing to do with friction: if the drive train went over diagonally, the bump lifted two of the wheels off of the ground, and they could only get off again by flailing around with the wheels.
I will second pfreivald here. It is easy to build a mecanum drive and easy to program with the already available resources. As with any drive train, uneven surfaces can be a problem, but that is one that you can solve by practicing. Having the bump lift wheels off the ground can largely be dealt with by wheel size and/or where the wheel is mounted. My biggest concern with mecanum this year is actually that it tends to produce a bumpier ride than standard wheels, which might present a stability problem, depending on your mechanism for picking up and carrying totes/containers. That said it is one of the drive train choices we are prototyping now.
We ran mecanums for rebound rumble & ariel ascent. It handled the bridge & running over piles of Frisbees. As any drive train, it requires driver practice & a thorough understanding of its strengths & weaknesses. In you plan to use them in this game, you need a realistic scoring ramp to practice with.
This is easily accounted for by modifying the strategy and match flow. I don’t envision too many times where a Mecanum’s diagonal strafe is terribly useful around the bumps in this game. Personally, I would stick with 90 degree angles as much as possible and completely forget field-centric drive.
The cool thing about Mecanums this year - in order to properly orient yourself to the totes that start on the field, just smash into them from an angle where you’re already mostly aligned to them. They’ll square you up pretty nicely :D. Then strafing to pickup totes works like a charm.
I would say to go for it. Nice thing about mechanum is that if you run into problems you can always just switch them out for standard wheels of the same diameter and run that way. No need to make major mechanical changes.
Biggest thing I would say is to get your drivetrain built early and into the hands of the students. Driving practice is going to be more important than anything else if you want them to use it as anything more than a tank drive robot. We have ran mechanum for a few years now and the new kids tend to forget about the strafe and rotation abilities until after they have had a lot of practice with it.
My number one advise for any team considering mecanum for the first time is get a drive train done early, like by the end of week two. Get your drivers practicing by the end of week three.
If you can’t get that done then switch to 6 wheel tank. You will be much happier.
We have done mecanum before and there is literally no substitute for practice. We already have an old bot in the programmers hands and should be driving next week.
We decided (As you said here) that we need more resources and time in order to try Mecanum for the first time. So we will probably go with omni-drive ( KOP with omni and two more wheels).
I will say as being last year our teams rookie year, that mecanum is very doable. We had only met for 9 hours a week and execute a successful mecanum drive and hope to work out the bugs this year.
I have attached a file on programming mecanum, but now a days labview included a base code to help get started.
We almost always design our drivebases that are trying new setups in season to accept two kinds of drive. One of our standards, and then the experimental drive. That way we can easily swap between them if needed.
We successfully drove over a platform with 8" AndyMark Mecanum wheels on a Nano Tube Chassis. The frame tubing is 3" tall, giving us 2.5" of clearance.
I wasn’t the one putting the last year’s Mecanum drive upgrade on this year’s KOP, but they were able to put them outboard using existing holes and they never mentioned it being hard in any way. However using existing holes makes the base too low and it will just barely drag on the scoring bumps so we’ll be adding some holes to the KOP chassis to raise it up, but we should be running it as-is tonight as a base and electronic board only (No lifting mechanism installed).
They are going to use a Z axis joystick (that took me 10 minutes to find in a back room behind stuff) but I’m thinking that an XBox controller might work better. I guess we’ll find out.
We did mecanum wheels last year as our rookie year. We seeded 15th at Kansas City. They were surprisingly easy to do. If you hit the scoring platform at a diagonal things will be odd. If you try to strafe with your front on the platform and your rear on the carpet it will probably get weird because of the different friction levels. The weight shift of the totes may make it not track in the exact direction commanded. But all that is not too hard to deal with.
We plan on doing them again this year with field centric again as well. I find that fresh drivers do better with field centric. Experienced ones are better with robot centric. I also believe that aligning with mecanums is easier than skid steer for all but the most experienced drivers.
