Ozite skid pad for test & tune?
03-12-2009, 03:36 PM
#1
Ozite skid pad for test & tune?
Like many others on this forum my frame of reference is 1/1 scale racing.
Has anyone ever used a garage size (8' by 8' square) carpet skid pad for tuning their chassis? I ask because 1 race a month is not the fast track for a newbie trying to develop and set-up a 1/10 VTA car, TIA.
Has anyone ever used a garage size (8' by 8' square) carpet skid pad for tuning their chassis? I ask because 1 race a month is not the fast track for a newbie trying to develop and set-up a 1/10 VTA car, TIA.
03-12-2009, 04:12 PM
#2
Tech Adept
I never did any skidpad testing personally, but I've done some cornering power system models on my personal computer.
Google: wiki skidpad.
It says a typical skidpad radius is 300 feet. That scales to 30 foot (1/10).
The wiki links to a great online article at Sport Compact Car, g Masters Lateral Acceleration Testing. I used this data in my system models.
According to this article, and my cornering power research, in a full scale car the skidpad predicts performance mostly in low speed corners. Downforce changes handling in high speed corners, so one must consider both factors on course with tight corners and high speed corners.
When you scale down a vehicle at constant density, the mass goes down by the scale fraction cubed (1/1000), but the aerodynamic body area goes down by the scale fraction squared (1/100). The parts at scale are less dense so this effect is amplified in RC racing.
Thus a small car is more responsive to aerodynamics than a full scale vehicle, even though top cornering speeds are relatively small, they are probably significant in fast sweeping turns or fast banked turns. My little brother races midgets, and the wing has a big impact on cornering performance, even at moderate cornering speeds. Downforce should be considered at scale.
In theory you could place the 1/10 car on a board covered with carpet, so the tires run across at 90 degrees to the tilt angle, and tilt the board until the car tires lose grip. Be sure to secure the car so it doesn't roll right over, since I suspect a high angle.
Traction coefficient u = tan(theta), where theta is the tilt angle at which slip occurs. Changes in tires, weight distribution, CG, etc., in theory should show up in this simple test as an increase or decrease in the tilt angle. I would be interested in comparing this simple test to elapsed time on a skidpad, and see if the theoretical correlation shows up on the stop watch.
Google: wiki skidpad.
It says a typical skidpad radius is 300 feet. That scales to 30 foot (1/10).
The wiki links to a great online article at Sport Compact Car, g Masters Lateral Acceleration Testing. I used this data in my system models.
According to this article, and my cornering power research, in a full scale car the skidpad predicts performance mostly in low speed corners. Downforce changes handling in high speed corners, so one must consider both factors on course with tight corners and high speed corners.
When you scale down a vehicle at constant density, the mass goes down by the scale fraction cubed (1/1000), but the aerodynamic body area goes down by the scale fraction squared (1/100). The parts at scale are less dense so this effect is amplified in RC racing.
Thus a small car is more responsive to aerodynamics than a full scale vehicle, even though top cornering speeds are relatively small, they are probably significant in fast sweeping turns or fast banked turns. My little brother races midgets, and the wing has a big impact on cornering performance, even at moderate cornering speeds. Downforce should be considered at scale.
In theory you could place the 1/10 car on a board covered with carpet, so the tires run across at 90 degrees to the tilt angle, and tilt the board until the car tires lose grip. Be sure to secure the car so it doesn't roll right over, since I suspect a high angle.
Traction coefficient u = tan(theta), where theta is the tilt angle at which slip occurs. Changes in tires, weight distribution, CG, etc., in theory should show up in this simple test as an increase or decrease in the tilt angle. I would be interested in comparing this simple test to elapsed time on a skidpad, and see if the theoretical correlation shows up on the stop watch.
Last edited by SystemTheory; 03-12-2009 at 04:16 PM. Reason: spelling
03-12-2009, 04:32 PM
#3
Like many others on this forum my frame of reference is 1/1 scale racing.
Has anyone ever used a garage size (8' by 8' square) carpet skid pad for tuning their chassis? I ask because 1 race a month is not the fast track for a newbie trying to develop and set-up a 1/10 VTA car, TIA.
Has anyone ever used a garage size (8' by 8' square) carpet skid pad for tuning their chassis? I ask because 1 race a month is not the fast track for a newbie trying to develop and set-up a 1/10 VTA car, TIA.
03-12-2009, 05:34 PM
#4
Tech Adept
A simulator is good for improving driving skill. It must have a good physics engine, or it does not faithfully reproduce the real driving experience. You would be better off making a practice course if the physics engine is bad.
Two cars go through a turn with the same driver. Which one has the lower elasped time? The car that can pull more gees goes through the same turn with lower time and greater exit speed. So to win races, you must have the skill to drive a car that can pull more gees than the next guy. This is why race teams invest so much effort in tire testing, wind tunnel testing, and skid pad testing matters in Sports Car performance.
Two cars go through a turn with the same driver. Which one has the lower elasped time? The car that can pull more gees goes through the same turn with lower time and greater exit speed. So to win races, you must have the skill to drive a car that can pull more gees than the next guy. This is why race teams invest so much effort in tire testing, wind tunnel testing, and skid pad testing matters in Sports Car performance.
03-12-2009, 08:58 PM
#6
http://www.redrc.net/2009/02/vrc-col...ng/#more-15997
03-13-2009, 02:38 PM
#7
Tech Adept
I attach three diagrams of the significance of cornering power. The basic idea is that centripital acceleration aC <= uT, tire traction coefficient. The front unit and rear unit may have a different uT under dynamic loads (weight transfer and suspension geometry), hence oversteer and understeer occur. I don't have time to develop theory now but a picture is worth a bunch of words.
03-13-2009, 03:07 PM
#8
Skidpads & VRC
ST thanks for the engineering lesson. I have seen people use smaller radius circles (75 ft) to sort a full size car. I.E. loose or tight, bump stops, springs, tire pressures, shocks, etc. That said I think a 10 foot radius would work for my RC car. I am not looking for ultimate grip and max. cornering speed just a way to get the car neutral before a race. Also this would allow me to make small adjustments on the car and evaluate and log the effect. I would rather fine tune a good basic set-up during the limited practice I get on race day. Right now me and the other new guys end up using heat races to dial in the car if we can.
I know this is hard to believe but in the twin towns of 2.5 million people we do not have a permanent carpet indoor RC race track. Yet every small town east of Madison Wisconsin has one?
Rob and Gordie I do practice on VRC but a 1/8 scale Nitro Can Am car on a flowing asphalt track with sweeping turns is very different from the tight and twisty carpet tracks preferred by some of the local guys. They are at such an advanced RC skill level I think they like the technical challenge of constantly changing direction while maintaining speed. I do think VRC has helped me develop the muscle memory required for improved car control.
Thanks again for the input.
I know this is hard to believe but in the twin towns of 2.5 million people we do not have a permanent carpet indoor RC race track. Yet every small town east of Madison Wisconsin has one?
Rob and Gordie I do practice on VRC but a 1/8 scale Nitro Can Am car on a flowing asphalt track with sweeping turns is very different from the tight and twisty carpet tracks preferred by some of the local guys. They are at such an advanced RC skill level I think they like the technical challenge of constantly changing direction while maintaining speed. I do think VRC has helped me develop the muscle memory required for improved car control.
Thanks again for the input.
03-13-2009, 04:27 PM
#9
Tech Adept
Parnelli,
As you probably know the traditional skid pad test is based on the radius and elasped time for a circuit completed on the edge of traction. This lets you compute the average velocity v and acceleration aC, and that's the effective traction coefficient uT.
You need to compute distance s as the circumference of the circle, then divide to find v = s/time. Cornering power is then aC = v^2/(R*g) where g is standard gravity in the same displacement units as velocity.
I estimate about a 0.2 second error for a human starting and stopping a stop watch, based on informal experiments. For a tight circle the ET might be very small, and the start/stop error more significant.
If you're into electronic sensing, a two axis g sensor could be installed with a data logger. I think a problem with these sensors may be vibration or a noisy signal. I've never used one but it seems like a neat way to experiment with scale racing data on a skid pad or race track.
I don't know if one of the new commercial loggers has a g sensor, but Pololu supports a sensor for robotics etc.
Please share your results if you develop a basic testing system?
As you probably know the traditional skid pad test is based on the radius and elasped time for a circuit completed on the edge of traction. This lets you compute the average velocity v and acceleration aC, and that's the effective traction coefficient uT.
You need to compute distance s as the circumference of the circle, then divide to find v = s/time. Cornering power is then aC = v^2/(R*g) where g is standard gravity in the same displacement units as velocity.
I estimate about a 0.2 second error for a human starting and stopping a stop watch, based on informal experiments. For a tight circle the ET might be very small, and the start/stop error more significant.
If you're into electronic sensing, a two axis g sensor could be installed with a data logger. I think a problem with these sensors may be vibration or a noisy signal. I've never used one but it seems like a neat way to experiment with scale racing data on a skid pad or race track.
I don't know if one of the new commercial loggers has a g sensor, but Pololu supports a sensor for robotics etc.
Please share your results if you develop a basic testing system?
03-13-2009, 06:15 PM
#10
My only concern would be that we don't have a lot of part throttle, long sweeping turns. The few sweepers on most rc tracks tend to be ones that even a poorly set-up car can take wide open. Therefore, our cars handling during transition is far more important. How does it handle entering a hairpin while standing on the brakes? What does it do when grabbing a fist full of throttle exiting that really slow corner, etc?
Then again, stock classes tend to be a little more wide open so maybe you could get some valuable info for that.
Then again, stock classes tend to be a little more wide open so maybe you could get some valuable info for that.
03-13-2009, 06:16 PM
#11
03-13-2009, 10:17 PM
#12
Tech Adept
I asked John Stranahan about his measurements on some other threads. Here is how he found a 2.8g traction coefficient:
John leaves out the division by standard gravity g in the formula, but you get the picture.
Wingracer I respect your driving experience but it still makes sense to tune for more g's in every corner. Being aware of the estimated entrance/exit velocity in a corner might help you set and focus on making braking points with smoother transitions, so the concept of focusing on the corners is consistent with learning to make better transitions. In theoretical terms, you want to brake hard while you're still going straight, and just enough to get to your entrance velocity for the corner, and that goes up if your car can pull more g's, so it all goes together.
We measured the radius of a medium turn using a rope to swing the arc of the race line; we put some cans out to mark 1/4 of a full turn and used a stop watch to get the 2.8g. Ac=v^2/r. The turn was traction limited. A 3.5 g tipover result (traction roll) was determined with the weight transfer formula. I calculated at what g was 1/2 the weight transfered. The center of gravity height was determined by balancing the entire car on a tape reinforced section of the body. It was about .75 inch up on a touring car at 6mm ride height.
The few sweepers on most rc tracks tend to be ones that even a poorly set-up car can take wide open. Therefore, our cars handling during transition is far more important. How does it handle entering a hairpin while standing on the brakes? What does it do when grabbing a fist full of throttle exiting that really slow corner, etc?
Last edited by SystemTheory; 03-13-2009 at 10:29 PM. Reason: typo
03-14-2009, 07:41 AM
#13
I'm not saying there would be no value to it. I just don't think it will tell the whole story.
As for straight line braking, Nigel Mansel and Michael Schumacher kind of disprove that one. They are both legendary trail brakers.
But then, I drive everything like it's a 1/8th onroad car. Brakes are needed to get those things to turn sometimes.
As for straight line braking, Nigel Mansel and Michael Schumacher kind of disprove that one. They are both legendary trail brakers.
But then, I drive everything like it's a 1/8th onroad car. Brakes are needed to get those things to turn sometimes.
03-14-2009, 08:22 AM
#15
Tech Adept
Trail braking a dump truck will not win a formula race. There is nothing I have offered here that would not help you understand what wins races. If you can trail brake a car that pulls more g's, you'll go faster than the guy in a dump truck! But if racing without thinking about setup/theory is what you enjoy, go for it!
Wingracer, maybe this comment is too harsh. What you are describing in terms I understand is that the good driver has an optimized control over his braking and handling, is smother and saves energy in every corner, which is critical to racing in addition to having a good driving machine. So I agree with you there.
A few years back I drove the NASCAR simulator at the local mall. Despite motocross experience I drove like a total rookie. The print sheets listed the Kinetic Energy all around the track. I asked myself, why? A smoother (faster) driver is actually conserving kinetic energy in the corners, thus maximizing the power available from the engine, wasting less energy by over-using the brakes (which convert kinetic energy to heat). Thus, the printing of KE on the simulator makes good sense if one can digest the data and improve one's driving. Me? I stopped driving the simulator since I get more fun from "bench racing."
Wingracer, maybe this comment is too harsh. What you are describing in terms I understand is that the good driver has an optimized control over his braking and handling, is smother and saves energy in every corner, which is critical to racing in addition to having a good driving machine. So I agree with you there.
A few years back I drove the NASCAR simulator at the local mall. Despite motocross experience I drove like a total rookie. The print sheets listed the Kinetic Energy all around the track. I asked myself, why? A smoother (faster) driver is actually conserving kinetic energy in the corners, thus maximizing the power available from the engine, wasting less energy by over-using the brakes (which convert kinetic energy to heat). Thus, the printing of KE on the simulator makes good sense if one can digest the data and improve one's driving. Me? I stopped driving the simulator since I get more fun from "bench racing."
Last edited by SystemTheory; 03-14-2009 at 08:31 AM. Reason: 2nd paragraph & 3rd