Alright, I wouldn’t go so far as to call this a retraction, but let’s call it a little clarification of the last article I wrote on the optimum shift RPMs for the Skip Barber car. Based on the calculations of horsepower and RPMs, I put together a prescribed set of shift RPMs for each gear change. But like all theories, it doesn’t hold much value until it is put to the test. The proof is in the puddin’.

So I set out to confirm if my shifting advice would put a driver at an advantage as they scream off the grid towards turn one. To do this I ran a number of starts at the beginning of an 800m long straight stretch and tried to hit a variety of shift RPMs and evaluate what shift technique was quickest to a given distance.

Well, it turned out that the run that was closest to matching my shift recipe was a few milliseconds slower than another run which was slightly off. So I decided to plot all the runs I did and compared them to the fastest run of the experiment. Then I looked at each shift from each run and evaluated whether it gained on the fastest run, or lost to it. I quantified the difference and came up with a much more reliable technique to see which shift RPMs offered the optimum acceleration.

The result is shown in the following graph:

It was odd how identical full-throttle starts resulted in different launches. It appeared that when the engine was increasing in RPM after the brief drop due to the rev limiter, and was close to 6300 RPM, the result was the optimum launch off the line. This was confirmed when the RPMs were held close to the rev limiter without triggering the rev-limiter. Thus, you could take your chances with a full throttle start, but you may get a slightly more reliable result if you hold it close to the rev limit during the launch into first gear.

The shift from first to second did not seem as critical, since it was changing in RPM so rapidly that the difference in shift points was not pronounced, but the trend did indicate that 6200 RPM was best. Both second to third and third to fourth shifts were optimized when just over 6000 RPM, but the most critical shift point appeared to be the final shift into top gear. Shifting too early is a big disadvantage here and you would need to get close to the 6300 RPM before you lose the advantage.

So the new shift advice for the Skip reads as follows:
Neutral to 1st: 6300 RPM (without triggering the 6400 RPM rev limiter)
1st – 2nd: 6200 RPM … not as critical as the others
2nd – 3rd: 6000 RPM … plus or minus 100 rpm is ok
3rd – 4th : 6000 RPM … plus or minus 100 rpm is ok
4th – 5th: 6200+ RPM … stay well above 6000; shifting lower than 6100 will slow you down a lot compared to people shifting at higher RPMs.

So my old shift points were not that far off the mark, and followed the same trends as was found in the actual testing. Only the second to third upshift was off by about 150 RPM.  Comparing this new routine to someone who starts on the rev limiter and shifts always at 6100 RPM would result in an advantage of about 1.16m after 800m.

But you know how the old saying goes: Even a broken clock is right twice a day.

Share Button


i wish i could see my rev meter when i use realistic FOV on my triple monitor setup 🙁

Jose Afonso
January 17th, 2010 at 12:01 am

I was thinking exactly the same thing as Jose … iSpeed and an iPod help the situation.

Christiaan LeGrand
January 17th, 2010 at 1:59 am

wow.. wish I read this before my skip barber race today! Nice to finally know the shifting recipe for this car though. Cool article.

CJ Modiano
January 17th, 2010 at 6:48 am

Another great article Ray. Have you tried plotting the applied torque in each gear vs speed? Depending on the gear ratios and the shape of the engine torque curve it may be better to stay in a lower gear past the intersection of the power curves (as given in the previous article). Unless the torque curve falls off dramatically and/or the gear ratios are quite close, the lower gear can give more applied torque near the rev limit because of the shorter ratio even though the engine power and torque have have dropped somewhat. It’d be interesting to see what shift points the torque curves would suggest.

Ken M
January 28th, 2010 at 4:07 pm

Yes, I have that data as well, Ken – as a matter of fact I was plotting some curves of torque v speed last night. And I think when it comes down to it the torque curves are critical for the performance at lower speeds, and the power curves dictate the pace at higher speeds, so really it is important to consider both.
The procedure above was a nice way to evaluate the optimum shift recipe from a more pragmatic approach.

Ray Bryden
January 29th, 2010 at 6:47 pm

Yeah, some one once told me that power gives top speed and torque gives how quickly you reach top speed. 🙂

Since you have the torque curves and drag/friction losses, shouldn’t it be possible to determine speed and distance over time (acceleration = applied torque-drag)? If you can do that then studying the effect of shift points, wing settings and gear ratios becomes possible without getting into the car. Though that wouldn’t be as much fun. 😉

Ken M
February 2nd, 2010 at 3:16 pm

The “point” of engine power output is to be a rating of engine performance that considers the effects of both torque and gearing. Were gearshifts instantaneous, you would want to be in the gear that resulted in greater engine power at all times, because this results in the greatest engine torque multiplied by gearing – that is, the highest rear-wheel torque. The only reason that you might want to shift at an engine speed other than the intersection in the power curves is to attain a higher speed before you lose acceleration for a gearshift – something that is more likely to work in a low gear where the rate of acceleration is high.

Damon Ethakada
February 23rd, 2010 at 1:24 pm

Damon has it spot on . To get maximum performance you need maximum average power delivered to the rear wheels. This will always occur at an engine RPM just slightly above the power crossover point, since the car is slowing down (if even slightly) during the shift operation.
I have always been intrigued by the belief that engine torque itself is a factor at different car speeds or engine speeds. The important thing is the value of engine RPM times engine torque (i.e. horsepower) and then matching the car transmission to deliver this power to the current car wheel speed.
If there was a highly efficient continuously variable transmission (CVT) that was suitable for racing, then the transmission control system would be programmed to retain the engine at its peak power point at all times of full throttle application and the car would achieve maximum distance, speed and acceleration.

Joe K.
January 5th, 2018 at 4:53 pm

Interested in special offers, free giveaways, and news?

Stay In Touch