60 ft acceleration

[Bus_Lengths]

So, our teams data acquisition system sucks major donkey balls. We're limited to about 12 hz. Our accelerometer output pretty much looks like a seismograph in an earthquake (I had no idea we hit both 2g's and -1g on a WOT pull). Also, we can only get wheelspeed from engine RPM, so thats pretty much worthless during the launch. So, I'm asking out of curiosity what other teams seem to be pulling for 60 ft times. 2.0 seconds? 1.9 seconds? Maybe even some 1.8's?

The OBDII data aq on my street car is far better than what we have on formula and thus its the only testbed I have for tractive effort modeling. I've found that engine power alone is enough to predict 60 ft times, but only up until 2.0 seconds. Beyond that, the math breaks down.

On a 1.7 second 60' with a 6000 RPM launch in my street car, I've noticed that for the first 30 ft or so, I'm seeing accelerations WAY higher than I should be able to achieve on engine power alone (upwards of 2 g's). However, the next 30 ft fit perfectly with calculated 1st gear performance (obviously taking into account the huge inertial losses in 1st). Obviously the clutch dump is probably responsible for this - the kinetic energy from the clutch helps bring the car off the line.

I'm more or less curious as to what people have seen on FSAE cars with regards to the launch. Obviously, with the tires/surface we race on, a huge clutch dump isn't going to be as beneficial because its not just going to hook. However, have others still noticed some 'extra energy' during the launch? Maybe g forces that you know your car should not be able to hit based on torque/weight/gearing?

From a pure engine power perspective, I calculate our car should run a 'true' 2.05 60' time and 'measured' 60' of 1.8 seconds (measured takes into account that you always have about 10-11" of rollout until you trigger the beams).

[Bus_Lengths]

So, our teams data acquisition system sucks major donkey balls. We're limited to about 12 hz. Our accelerometer output pretty much looks like a seismograph in an earthquake (I had no idea we hit both 2g's and -1g on a WOT pull). Also, we can only get wheelspeed from engine RPM, so thats pretty much worthless during the launch. So, I'm asking out of curiosity what other teams seem to be pulling for 60 ft times. 2.0 seconds? 1.9 seconds? Maybe even some 1.8's?

The OBDII data aq on my street car is far better than what we have on formula and thus its the only testbed I have for tractive effort modeling. I've found that engine power alone is enough to predict 60 ft times, but only up until 2.0 seconds. Beyond that, the math breaks down.

On a 1.7 second 60' with a 6000 RPM launch in my street car, I've noticed that for the first 30 ft or so, I'm seeing accelerations WAY higher than I should be able to achieve on engine power alone (upwards of 2 g's). However, the next 30 ft fit perfectly with calculated 1st gear performance (obviously taking into account the huge inertial losses in 1st). Obviously the clutch dump is probably responsible for this - the kinetic energy from the clutch helps bring the car off the line.

I'm more or less curious as to what people have seen on FSAE cars with regards to the launch. Obviously, with the tires/surface we race on, a huge clutch dump isn't going to be as beneficial because its not just going to hook. However, have others still noticed some 'extra energy' during the launch? Maybe g forces that you know your car should not be able to hit based on torque/weight/gearing?

From a pure engine power perspective, I calculate our car should run a 'true' 2.05 60' time and 'measured' 60' of 1.8 seconds (measured takes into account that you always have about 10-11" of rollout until you trigger the beams).

[VFR750R]

Take your car to a test and tune night at a local dragstrip....might even attract some sponsors, and it will only cost you $10 a person (usually with or without a car, so everyone should bring their cars and play too). You can get lots of practice and see how different launching procedures effect your 60ft time.

The only thing will be the track will be prepped allowing faster 60ft times then you will be able to do on an non prepped surface like MIS.

[Bus_Lengths]

I would love to get our car down to the drag strip sometime. However, its always usually well prepped and with slicks on the car, I'm pretty sure we could raise the rev-limiter to 15,000 RPM, clutch drop, dead hook, and break an axle. This is not really indicative of your average FSAE launching surface. At the earliest, we probably wouldn't be able to go until way later this summer as the drag strip would literally destroy everything in the rear end of the car.

I figured there were some other powertrain guys out there with significant interest in straight line performance. I really wanted to just have a technical discussion on what takes place during the launch and see how people are finding this correlates to 60 ft times and tractive effort models.

In my tractive effort model, I've given up on even calculating the launch and instead just made the 60 ft time an input and assumed 2 stages of constant acceleration to get there. This is the only thing that seems to work for any vehicle on sticky tires. What do other teams do to simulate the launch in your tractive effort analysis? Is anybody actually using a detailed clutch model and calculating the kinetic energy transfered to the tires on a clutch drop?

I suppose another way of asking this is to ask what g's people are pulling off the line? I would imagine 0.8 and 0.9 are the norm. However, for the high 3 second 0-75 m runs, I would imagine that teams have to be seeing upwards of 1.3g's off the line.

[exFSAE]

How good is the tire data you're using for this predictive analysis?

[Bus_Lengths]

The tire data is not the limitation. I would use 'Tractive Force vs. Slip Ratio' scaled by the peak coefficient of friction measured on that surface. This would be good enough to get you within say 10% or so which would be great - if you had a launch model.

What is not great, is being 30-50% off on launch g's because of the damn impact load from the rotating assembly.

[exFSAE]

How do you know the tire data is any good to begin with? How well does quasi-steady state F&M data help you for such a dynamic event (launching)?

Honestly I think this is one of those things better suited for testing. Get the car set up with the CG as high and far rearward as possible.. sweep air inflation and launch RPM.

[Bus_Lengths]

I have respect for the 'It's too complicated, just test it method'. However, I'm an engineer, not a technician. Thus, not being able to figure out something that pretty much involves the fundamentals of mechanical engineering greatly frustrates me.

Here's where I'm going with this. Imagine, that as an engineer, you are actually tasked with engineering something before building some prototype. I'm going to use as an example a common cam/bolt-on LS1 powered Camaro.

3500 lbs w/driver
400 lb-ft at the rear wheels (4th gear)
330 lb-ft at the rear wheels (1st gear)
2.97:1 1st gear reduction (manual transmission)
3.55 Final Drive
12" radius tire

Now, lets say your boss for some reason wanted you to find out how fast this car could cross 60 feet with infinite traction. You conservatively assume that the car outputs peak torque the entire way to the 60' mark.

330 * 2.97 * 3.55 = 3479 lb-ft
3479 lb-ft/1 ft = 3479 lb of thrust

This is roughly the weight of the vehicle (3500 lb), thus you can assume the car will see a maximum acceleration of 1 g (32.2 ft/s^2). Assuming the car could accelerate like this all the way to the 60' mark, how long would it take?

d=(0.5)*a*t^2
t=(2*d/a)^0.5

If 'a' is 32.2, and 'd' is 60 ft, then time = 1.93 seconds.

Congratulations, you have just been fired. You assumed that this engine could magically output peak torque for 2 whole seconds (which it can't) and hoped this gave you the minimum time possible with that torque. Instead however, people routinely pull 1.4 second 60' times with the top drivers pulling 1.39x 60' times - with the same amount of power listed above.

Now wait, we can plug this time in and back calculate the torque.
A 1.39 second 60' time equates to averaging 62.1 ft/s^2 across that distance or 1.93 g's. This is a far cry from the 1 g calculated earlier. It comes out to be 6750 lbs of thrust or 640 lb-ft at the rear wheels.

How is this possible? How can a vehicle accelerate almost twice as hard as engine power alone should dictate? The only thing I can think of is the clutch drop. This obviously has a huge impact on street/strip cars. I was curious what effect this has on FSAE cars and if people had cared to model this, or if most people just 'fudge' there launch models to account for this.

[exFSAE]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Bus_Lengths:
I have respect for the 'It's too complicated, just test it method'. However, I'm an engineer, not a technician. Thus, not being able to figure out something that pretty much involves the fundamentals of mechanical engineering greatly frustrates me. </div></BLOCKQUOTE>

Believe me, I understand. However, and perhaps this was different at your university, transient tire dynamics are not included in fundamentals of MechEng.

IMO, the wise engineer understands when something is too involved to be modeled with any shred of accuracy.. and is better off spending the time testing it.