Quote:
Originally Posted by swamp2
OK since you asked, I will humor you. There is no reason not to share the capabilities or validation of a software tool I like and call accurate.
Using the exact figures in your post and making only the following minor changes to the default CarTest parameters: weight - to match the measured curb weight (software includes options and defaults for driver weight and gas weight and they did not match the number you posted for this particular car), shift times - default is .5 s a more realistic figure for a good driver in a MT is .3 s. Conclusions:
-0-60: right on the money or within 3/10th however you like to call it
-0-100: within 7/10th or 1/10th
-1/4mi: within 1/10th to 4/10ths on ET
-1/4mi: trap within 1.5 - 3.5 mph
I think this is very typical of what one car get with CarTest. Do you notice how much closer this is comparing simulation to one actual test than the case we argued about for pages. I suspect that some tests do show better numbers and others worse for the E46 M3. My conclusion from this exercise is that both the reported figures for the car, the inputs to CarTest, the actual test itself and the simulation outputs are "consistent". There is no glaring/obvious problem with any of the pieces of the puzzle.
My C63 AMG results today were just as good!
Enjoy.
P.S. One big reason the E46 M3 is fairly fast given its peak hp/weight ratio (not all that hot) is it's high redline and short gear ratios i.e. torque multiplication, just like the new car (I mean wrt gearing not a moderate hp/weight ratio).
|
OK, now that the smoke has cleared and everyone has cooled off a bit, I thought I'd address these simulations.
Swamp, the thing is, useful tool or not, the simulation results seem to be fairly poor - notwithstanding your feeling that the results are just fine.
Using the above E46 M3 results as an example, the simulator seems to show a power shortfall of around 10% compared with C & D's "new car" results - and C & D's results are not atypical of what one might expect of such a vehicle, corrected to approximate SAE "gross" Standard Day meteorological conditions. Our car went a best of 13.12 at 107 and change, if memory serves.
A simulation that represents an approximate 10% power loss is seriously deficient.
Doing some minor analysis, the ET-to-speed relationship (ET times speed) is about right, with the simulation coming in at a 1396 compared to C & D's 1402. (What one should look for is a number of about 1400 or better, with very powerful street cars struggling to reach this number. Anything much over 1400 means you've got a significant traction or gearing problem - or a pilot unfamilair with basic starting line procedures. Much under 1400 shows traction and torque.)
One thing that does in fact stand out a bit is that the simulator shows a pickup of less than 20 miles per hour (19.81, to be exact) between the 660 and 1320 foot markers. Not to put too fine a point on it, but this car-made-of-electrons is a flat pig on the top end, which is just where you'd expect a power-laden, torque-challenged car to really be coming on strong, particularly with the relatively close gear-spacing BMW uses after second gear. Our E46 picked up an average of nearly 24 mph as a comparison, and other cars we've documented also have done comparitively well on the top end, as well.
Examples:
'85 Vette, 230 HP, 3230 pounds. Average 13.77 @ 101.36 MPH, picked up an average of 21.24 MPH in the last 660 feet over 20 runs.
'91 Saleen, unknown power and weight. Average 13.67 @ 102.11 MPH, picked up 20.71 MPH over a dozen runs.
'93 Vette, 300 HP, 3340 pounds. Average 13.13 @ 107.04 MPH, picked up 22.59 MPH over 31 runs.
'95 M3, 240 HP, 3220 pounds. Average 14.11 @ 98.42 MPH, picked up 20.15 MPH over 30 runs.
'04 M3, 333 HP, 3410 pounds. Only made three runs and didn't save the slips, but I checked while at the scene, and the car averaged around 13.2 @ about 107, with a 660 foot speed in the 83 range.
Now (and finally getting on point), your simulation in note #89 shows that the Lexus is also a pig on the top end, doing only around a 20 MPH pickup in the last 1/8th mile. A car with this power to weight will just flat do better than that, especially compared with the examples already given in this note.
I mean no disrespect, but to me, this is symptomatic of a general problem with these particular simulation results. I don't know the solution to the problem, but I do know there is a definite problem which makes these results pretty much without significant value, and by inference, other simulations you generate with this tool without a serious re-evaluation.
Bruce
PS - A high redline and short gear ratios have very little to do with actual quarter-mile results. Gearing essentially only matters a damn in the first 60 feet or so, and on a traction-challenged car, it won't make a significant difference even in that range. It's power and weight that mostly matter with cars in this performance range.