"Float" test -- Thoughts please!

[Vitamin K]

So, after a car's been built, and lubed and aligned, there's a test I like to do with it that I call the float test...where what I'll do is set the car on a smooth, flat surface (like a countertop), and just tap the front and rear of the car in alternating fashion to let it roll back and forth between my hands gently. What I'm looking for is how smoothly the car rolls, and how far it floats with just a minimal tap. My experience has been that the better the car's aligned and rolling, the more easily it "floats".

So I'm wondering: Can I translate this into some kind of metric that's repeatable? Consider if I took my tuning board with the yardstick down the middle and placed the car over the yardstick with the DFW on the rail and rears in straight position. Then, I gave the car a repeatable amount of impulse (say, from a pendulum that I let swing down and tap the car), could I measure the distance that the car travels and use that as a performance metric?

Now, clearly, the material used for the rolling surface would affect the travel...but if I test all of my cars on the same surface...have I got something useful there?

[Stan Pope]

Interesting idea. The following variant has fewer uncontrolled variables:

Set up your tuning board with a variable slope, allowing calibrated slope adjustment from 0 to 8 degrees. Starting at 8 degree slope, place the car at the top of the ramp, give a gentle nudge UP-SLOPE and watch for one of the following:
1) car stops doesn't roll back down the slope
2) car reverses course and rolls back down the slope

When you get a "1" response, increase the slope by 1/2 the prior change amount and repeat the test or stop if the change amount is below your threshhold.
When you get a "2" response, decrease the slope by 1/2 the prior change amount and repeat the test.

Example sequence:
Degrees Result
8 - - - - - - - 2
4 - - - - - - - 2
2 - - - - - - - 1
3 - - - - - - - 2
2.5 - - - - - - 2
2.25 - - - - - 1 Stop ... close enough

The 2.5 degree measure is a transferrable measure that is independent on the shape and weight details of your car and independent of the details of impulse mechanism.

You are correct in the observation that both methods will display good wheels, axles and alignment by SMOOTH transit.

[Vitamin K]

The 2.5 degree measure is a transferrable measure that is independent on the shape and weight details of your car and independent of the details of impulse mechanism.

You are correct in the observation that both methods will display good wheels, axles and alignment by SMOOTH transit.

Hmmm, I'm not quite so concerned about removing the weight variable, since the lion's share of PWD cars weigh in at 5oz, but I guess it would be useful for Outlaw cars, or maybe the Awana car rules I had to contend with...

So if I understand correctly, the gravity-rollback action means it should perform the same regardless of how hard the initial push is (within a reasonable range)? I guess I'd need to test that.

I'm assuming that the lower the angle at which the car will reverse-roll, the better. So you could say "The car's doing a 3.5 degree on the float test" or something.

Once you had a few cars doing numbers on that, with equal CoM, it would be interesting to see how the float numbers correspond to track times.

[Stan Pope]

...

So if I understand correctly, the gravity-rollback action means it should perform the same regardless of how hard the initial push is (within a reasonable range)? I guess I'd need to test that.

I'm assuming that the lower the angle at which the car will reverse-roll, the better. So you could say "The car's doing a 3.5 degree on the float test" or something.

Once you had a few cars doing numbers on that, with equal CoM, it would be interesting to see how the float numbers correspond to track times.

I believe that the energy of the initial push is immaterial, provide that the ramp slope and surface texture are constant.
In my alignment work, I used the edge of a block as a "starting pin" and pulled the block forward away from the car, hopefully imparting no energy to the car. The reverse roll should eliminate any concern for "adding energy" or "pulling the car along."

Yes, the lower the slope angle the better!

My suggestions arise from early my alignment work ... observations that as the alignment came into true, the car's acceleration of my modestly sloped table top became smooth and consistent. Slopes of less than 1" in 42" were my usual tuning slopes ... if I did the trig right that would be less than 1.4 degrees. I did not alter the wedges under the table legs, so I did not get measures of when it cut off.

I would not slight your original suggestion ... the exercise should give you a "warm fuzzy feeling" about the state of the car, in the same sense that my "smooth and consistent" observations gave me affirmation that the alignment was complete!

One aspect of which I am unsure is the effect of DFW Toe angle ... my work mentioned above was with straight alignment i.e. 0 degrees toe in! It may be that the toe-in angle interferes. Some experimentation with that is called for.

[Vitamin K]

One aspect of which I am unsure is the effect of DFW Toe angle ... my work mentioned above was with straight alignment i.e. 0 degrees toe in! It may be that the toe-in angle interferes. Some experimentation with that is called for.

Yeah, I had thought about that. That's why in my original proposal, I suggested arranging the car over a simulated rail, so that the floating test would happen in conjunction with being rail-guided...thus, if your oversteer is eating too much energy on the rail, you won't "float" as far. Not sure how this would work in reverse, but I sort of consider Derby cars to be forward-only vehicles.

[dpatrick911]

I like Stan's idea, and I will propose a method very similar to his. If your tuning board is level. By increments of 1 degree maybe a half... you raise one side of your tuning board. This could be done with "leg levelers" like this http://www.billbentgen.com/railroad/ima ... eveler.jpg and a digital angle gauge. With levelers like that you could dial in precisely when the car starts to roll. Set your car down on that side and record the roll if any. A reasonable matrix would be at what angle does the car finally roll the entire length of the board? I understand pushing the car back and forth. Heck, I do it too. But if you knew at what angle you car has a propensity to roll X length you could base line your car's "race worthiness" by documenting that known angle that it takes to achieve the desired roll. Small adjustments to your alignment could be performed to achieve the roll if the car falls short of your "baseline". We do not have a starting block that bumps our cars to start the race. Personally, I would think that jolting the car by pushing it could be enough to break the friction possibly caused by wheels rubbing against the car body. I am not saying that it is a huge problem but something to consider. I have found while checking alignment that it is sometimes even hard to get the wheels to stay against the body. So this concern maybe negligible. But I do think a "when does this car just want to take off?" test might actually be good for all of us to perform. Our cars this year would almost roll off the table if you breathed on them. One of the design judges mentioned that to me part way thru our races. He said, "I leaned on the table with my forearm to write a score down and your cars just about rolled off the table." I need to build a pit lane/staging area for all of the cars next year. He said, "I knew right then that those were going to be the cars to watch this year!" Just keeping the discussion rolling (pun intended)!!! After rereading the previous replies. I realized that I basically just repeated what both of you said. My apologizes.

[ngyoung]

Variable slope table is a good way to test coefficient of friction. Whichever car starts to roll at the lowest angle likely has the lower CoF.

[bracketracer]

If you have a smooth tuning board with a foam bumper at the end, you can let the car roll down the board from a fixed starting point and record the distance it rebounds. The farther it bounces back, the faster it was going. Don't even need a rail.

[Speedster]

What is the goal? A straight runner will travel a shorter distance than a rail rider without a guide strip. With the guide strip I suspect the opposite is true, along with lots of other things in 40'.

[Vitamin K]

What is the goal? A straight runner will travel a shorter distance than a rail rider without a guide strip. With the guide strip I suspect the opposite is true, along with lots of other things in 40'.

This is actually part of why I incorporated the guidestrip into my original idea. The idea was to incorporate a "maximum friction" situation, that is, the car is aligned and rolling with the DFW along the rail, and the steer-in set...and under those conditions, how well does it glide?