# COF help?

Has anyone measured below .1 for COF? We have tried measuring with force gauges and doing the incline test and both were giving .13 I’m not quite understanding how they are quoting COF = .05

Am I also correct in assuming that if they were going to treat the regolith surface with something that this would definitely be in the competition manual somewhere?

It is in the manual. gel-coated, fiberglass-reinforced, polymer material. Now wether they are referring to the frp being naturally gel coated or it being an additive is another story.

I have tested the wheels on linoleum. I get .12-.18. (inline)

It’s naturally gel-coated; or rather, that’s how it’s sold. The only treatment is to put it on carpet and tape it down with fiberglass tape.

This has been dicussed on another thread, but man that floor is gonna be gone in a day. We drive on it and it just powder. Happy Snow Day teams.

We get repeatable results with both the incline and pull force methods.

The conditions are new undamaged wheels and new undamaged Glassliner FRP from home depot.

We see static mu of 0.12 ± 0.02 and transverse of 0.13 ± 0.02. All measurements taken with 4 wheels.

At this point, I for one do not believe static mu of 0.05 without some form of lubrication on the floor. But those are the FRC numbers and they are sure sticking to them.

As a further aside, I cannot think of any way to traverse 48 ft of floor in 6 seconds in a standard kit bot frame without mu > 0.083, but then again I could be missing something.

In any case, we are preparing for static mu to be in the range from 0.06 to 0.15 and for the dynamic/static ratio to range from 1.0 - 2.5. As long as you can work under the full range of conditions you should be OK.

Conjecture:
The powder is the difference between the .06 quoted and .12 measured. After all, regolith means a layer of fine powder/dust on the moon. Maybe they are assuming the surface will be all jacked up and we will be driving on the powder which will then make the coefficient of friction more like .06

???

We got the same results this weekend. COF around 0.12 to 0.14.

Good news for most (more traction), bad news for anyone who was planning on using smaller motors on the drive train (Banebots or Globes), like us.

Oh well, bring in the CIMs.

Equations:
F = ma
F = u
N
ma = uN

Plugging in:
(mtrailer + mrobot) * a = u * mrobot * g
a = (u * mrobot * g)/(mtrailer + mrobot)

Let mrobot = 120lbs + 18lbs (bumpers) + 7lbs (battery) + 5 lbs (trailer tongue weight) = 150 lbs = 68.0 kg
Let mtrailer = 35lbs (not supported by the robot) = 15.9 kg

a = u68.09.8/(68.0+15.9)
a = u*7.94 m/s^2

We know that:
x = .5at^2

If x = 48 ft = 14.6 m and t = 6s, then
14.6 = .5a36
a = 0.811 m/s^2

Plugging back into our first equation…
0.811 m/s^2 = u*7.94 m/s^2
u = .102 (at least)

Possible Conclusion 1: If we are to believe Dave’s claims about traversing the field forward and back in about 12 seconds, then the published FIRST figures must be wrong.

Possible Conclusion 2: FIRST measured their CoF on a worn floor, while Dave drove on a new floor.

FWIW: Our new wheels on new regolith came out to about u=0.10

This is the type of thing I would expect to have happened

If the CoF really is about 0.1, then that is good news for linoleum, because we measured about 0.13-0.16 on linoleum, so it turns out to be a better analog than we initially thought.

Lovely math.
Your model is driving full speed end to end, without even a turnaround and decel. Top speed before hitting the wall is 0.811m/s^2*6s = 4.866m/s.
Does the stock transmission setup support that top speed?

To get to half-field in 3 seconds with this same accel, your initial velocity has to be 1.21m/s, or about 1.5s worth of accel. To summarize: u=0.102 also supports a model of accellerating to midfield, instant reverse accel, and bounce off walls perfectly elasticly at a bounce speed of 1.2m/s. This model drops the top speed to 3.64m/s by raising the average speed.
I wonder how it works out if you assume a lesser elasticity, like 80%, and a late non-midpoint reversal. ie faster crash, slower bounceback

A video of someone doing laps with trailer on a real field would be awesome.
Mass of MoonRock payload is assumed to be neglegible.

Re COF measurements: Is there a difference between static friction with a rolling wheel, vs static friction on a fixed wheel? It sounds like most people are measuring with a drag sled fixed wheel.

I have observed, but not measured, significantly lower COF when static friction is lost.

Rolling friction (where the surface of the wheel at its bottom point doesn’t slip relative to the floor) should be equivalent to static friction for non-deformative surfaces.

Yes, so I have been taught as well. I’ll rephrase my comment/question as
"When reporting local COF measurements, please include details of your methodology, such as using a fixed wheel drag sled, vs using a weighted robot with KOP wheels pulling a scale pulling a dragsled.

So I assume you have an algorithm in your drive code that determines when your wheel is slipping and backs off the power? Certainly possible - compare omega R of the wheel to integrated speed at the wheel from the yaw rate and accelerometer. Or just calculate it from static friction and never apply more than that amount of power, but if you do start to slip it’s not self correcting since the power requred decreases. Otherwise if you’re applying full power to the wheels you can expect them to slip.