Greetings and salutations one and all, I want to start out this thread with a hypothetical question. In a friction-less world would mecanum drive run at the same speed in both directions as an X-drive? I have worked out the math for an X-drive, but have no clue how to start in on how to solve the hypothetical torque and speed of a mecanum drive. any help is appreciated, and I am aware that in reality mecanum wheels strafe slower than the x-drive.
In a frictionless world wheels wouldn’t work. Traction (the maximum amount of force a wheel can apply in a certain direction before wheel slip occurs) is force normal (usually just the amount of weight pushing down on the wheel) times coefficient of friction. In a frictionless world the coefficient of friction would be zero so the wheel would start slipping immediately meaning no traction and therefore no force to get the robot to start moving.
Note: I could be wrong about the terms I’m using here. If I am, please feel free to correct me.
That being said, the reason X drives move faster is because of the way the forces are angled. I’m not particularly good at explaining why, just read this.
Because the direction of force in mecanum wheels are not at 45 degree angles (I believe it’s more like 30 degrees or so) you don’t get anywhere near as much of the movement directed to the outside which means VEX mecanum wheels will never be able to strafe as quickly as an X drive with the same gear ratios.
EDIT: Never mind, totally wrong about this, sorry.
I’m pretty sure the wheels on a mecanum drive are angled at 45 degrees, and both forward/backward motion and strafing “should” have the same speed. (Of course, unlike a x drive, there is no added gear ratio on a mecanum drive). From what I understand, mecanums work just like x-drives except that the rollers are arranged so that they fit on a forward-facing wheel.
I’m guessing that vex is making their mecanums just like how every other normal mecanums are made. If the rollers were angled at something like 30 degrees, it would be very obvious in sight, but they appear to me to be angled at 45 degrees, not to mention that lighter robots that don’t have to deal with much friction have strafing that is almost as fast as forward motion.
If the rollers on both types of wheels were frictionless, then an X drive and a Mechanum drive with the same ratio (which is hard, since the x drive wheels would have to go 1/root(2) as fast as the mechanum drive) would both strafe and drive forward equally fast. Why mechanum drives strafe slower (and stall more easily when strafing), is because a strafing mechanum drive spins the rollers root(2) times the speed the robot is traveling, whereas the x drive spins the rollers 1/root(2) times as fast as the robot is traveling, meaning the Mechanum drive has twice as much resistance. That is why the Mechanum drive strafes slower.
EDIT: The Mechanum drive does not have exactly twice the resistance of the X drive when strafing, but it has twice as much resistance from the rollers. The rollers are most of the resistance, though.
Thanks that last answer is exactly what I am looking for.
Hey 26, this problem can actually be solved using a little physics and trig, seeing as the components of the mecanum wheels are positioned at a 45 degree angle they actually don't want to move in a straight line, when you run both motors in the same direction they end up moving in their straight line pattern because the opposite angle of the front and back wheels cancels out any sideways motion that they would produce, resulting in a velocity vector which moves forward.
-
For the sake of this explanation let’s imagine our robot is sitting on a unit circle, the front of the robot is pointing forward, in the 0 degrees direction, the left and right sides of the robot are pointing in the 90 and 270 degree respectively directions and the rear is pointing in the 180 degree direction, 90 degrees represents the positive y direction, 270 as the negative and 0, positive x, 180 negative x…*
The story changes however when the wheels move in the opposite direction. In this situation the the two forward/backward velocity vectors move in an equal and opposite magnitude, this cancels them out, meaning the bot will no longer go in the x direction, however the y or side to side velocity vectors (from the wheels mecanum components being in opposite direction) no longer cancel out! However seeing as the motor moves the wheel only in the x direction, we have to use the unit circle to find out how much of our origional speed is actually going to go into this new angle, it comes out to around .7 or 70% (using the unit circle we know that at a 45 degree angle the net x/y change will be root 2/2 or around .7 (70%). However this isn’t the whole story, seeing as the x components of these new vectors still move in opposite directions (one is at an angle of 45 degrees while the other is at 135 degrees, so part of one is moving positive x and positive y while the other is moving in positive y and negative x, the different x values cancel out) this means that the vertical, or y component can be found with root2/2*.7, this nets .5 or 50% of the original velocity of the motors/of the velocity you get then moving forward/backwards, when moving in a sideways direction with mecanum.
So to answer your question the mecanum wheels should move faster than an x drive while going forwards/backwards (plus or minus x direction) because in the x drive all the wheels are moving in the 45 degree angle, thus their velocity vectors in any of the main directions will always be .7 of the velocity of the motors (70% speed in all four of the main directions (forward backwards right left…) however mecanum will move around 30% slower when moving to the sides, it’s not friction it’s just physics
I’m afraid your math is wrong - the vectors on an X-drive actually are additive and as a result, an X-drive actually travels at a speed that is 1.4 times that of a tank/mecanum drive driving forwards, and the same applies to strafing.
Also, I remember seeing a post or thread somewhere where the mecanums were found to have an angle that wasn’t 45 degrees, I’m not sure where though. When I get back to school on Tuesday I may be able to measure it myself.
Since we all have heard of AURA, here is their archive on it. (Were I took the picture from so I should give them credit. It’s a great article too.)
http://www.aura.org.nz/archives/1137
Basically an X-Drive creates a 45°-45°-90° right triangle. The equation for the legs to the hypotenuse, with the leg length being X, is «X multiplied by the square root of 2».(see picture)
http://www.aura.org.nz/wp-content/uploads/2012/05/xdrive2.png
Your final ratio is about 1.41421356237:1 (“1” being the speed of a tank drive with the same wheel size and gearing. 1.41421356237 is the equivalent of the square root of 2.)
As a side note your pushing force is only the equivalent of a 0.70710678118:1. (Less pushing force than a tank drive)
That’s as simple as I can put it. Remember this is not factoring in the friction differential.
EDIT: I’ve never used mecanums, so I don’t feel it’s my my place to comment on them, and because I’ve never used geometry to figure it out. I also don’t know the angle of the rollers on the mechanums.
Here’s the numbers you probably want:
I apologize, i now realize i was writing about speed when i really was talking about force :(… sorry about all of that, so i guess to correct my self, yes you go faster with an x drive however the forces are much lower
This is a really good explanation.
Thanks! That was my first time trying to do something like it.
Solution:
Build your own mecanums.
The Green Eggs did it.[LIST]
*]Less Resistance
*]Non-weird vector angles
*]Non-cantilevered rollers
*]You look way cooler if you build your own
[/LIST]
Less resistance??
Rollers on vex mechanum wheels are shaped such that the outer circumference of the wheel is as close to a circle as possible. Using rollers from 4" omni wheels does not do that at all, making the ride bumpier, thus less efficient. Using my hand, I could get the rollers on a mechanum wheel to spin for 2.5 seconds. Rollers from omni wheels don’t spin for any longer.
It could be possible to get more efficient mechanum wheels by building them out of vex parts, but I don’t see how you would do it.
Click on the link in the post you quoted. Green eggs built their own mecanums.
Yes, I clicked on the link and know Green Egg robotics made their own 2.75" mecanum wheels. However, they were not as smooth as vex mechanums, and they were built before vex had there own mechanum wheels. They were not built to be better than vex wheels, they were built so that the team could have a mecanum drive at all. I am saying that creating a mecanum wheel with less resistance than a vex mechanum wheel would be difficult.
You should also note the date when Green Eggs built their own, and compare it with the release of the official VEX EDR ones.
I’m proud of how our 4" mecanum wheel design turned out. 
I’m happy to see teams loving them.
After messing with our mecanum wheels today, I found that the cantilevered structure on the wheel rollers create more resistance when the wheel is touching the ground near the edges of the rollers.