Number of shifts on Detroit Track Lap

[rjwoods77]

Does anyone have any data from their car reguarding how many shifts were made on average per lap of Detroit? I am trying to determine time lost per lap from shifting assuming average shift time to be 100ms to compare it against the shift ratio of a cvt. Thank you.

[lporter]

0 http://fsae.com/groupee_common/emoticons/icon_wink.gif

Guess thats what happens when you rupture a CO2 line and get stuck in 2nd gear. Still finished 12th though.

[jsmooz]

0.....for the last three years http://fsae.com/groupee_common/emoticons/icon_smile.gif

[absolutepressure]

Originally posted by jsmooz:
0.....for the last three years http://fsae.com/groupee_common/emoticons/icon_smile.gif

Well, I think technically you had an infinite amount of shifts, but each one took (1/infinity) seconds.

[oz_olly]

Hey Rob,

I am very interested to know how you plan to compare the time. From work I have done the lost lap time is not the number of shifts multiplied by the shift time. My thesis will be finished in about 4 weeks if you would like to read it. The CVT is very close in lap time to a 93% efficient manual transmission with zero latency and optimised gear ratios, however that is assuming a CVT efficiency of 80% which is about the best you will ever get based on the sources I have cited.

Cheers

Olly

[rjwoods77]

Olly,

Yeah I would like to take a look at it when you are done. My thought on the matter, and I may be wrong, was to look at it just from an accel run standpoint just to make it easy. By comparing the rpm of the crank vs rpm of the jackshaft you will be able to log shift ratio and lay it over the power curve on a tractrive effort diagram and use it as a tuning tool to get the best area under the curve. Aaen always shows the difference between shifted and cvt cars in his book and I wanted our team to replicate that and use it as a tuning tool for all the weight,springs, helixs etc. In the long run if it possible to figure out how to do this by referencing power vs map and logging your shift ratio you can do the same kind of comparison. I know it is difficult and may be unlikely but at least for an accel run standpoint it will make for good comparisons. I wonder in your thesis if you are accounting for the other problems in throttle off downshifting and braking. Dearborn looked awfully slow in the autocross but posted a real good time. Corner entry and corner exit you never once saw the car jerk or hobble in the transfer from power on to braking where as almost every shifting car had some crazy studder from too much/too little braking along with the engine braking involved. This may just be an unmeasurable factor as driving skill can be which is something cvt's are incredibly forgiving over that shifter cars are not. I think teams lost more time in these power on/power off with braking/turning then they did actually shifting the car. I would imagine in doing a comparison of the two you would have to included data/discussion on the backshifting characteristics of the cvt and how they dont really "upset" the car whereas the shifting cars do. This may be the function of Dearborn having less power but they looked smooth as glass coming in and out of corners.

[adrial]

A well shifted, well setup car will not be unstable under braking. I am no schumacher but the car did not feel (or look) at all unstable during heavy braking & clutchless downshifts.

The CVT is definetely easier to drive, but you are also missing the ability to use engine braking to make slight changes to the brake bias for different corners.

That said -- FSAE is all about making a car that any idiot (most of us) can drive fast.

IIRC RFR shifted twice during the endurance in '06.

P.S. whats the fun in driving a car thats rock solid stable? Balls out, Ass out...thats the way we like it.

[Charlie]

If I recall correctly, we up-shifted about 5-7 times a lap in 2003 (when we finished 3rd in Endurance time) and about 6-8 times a lap in 2004 when we finished 2nd in Endurance time (7 seconds behind Cornell, I take the blame for being super slow my first two laps! Just 5 days of testing will kill you :P).

I would guess we shifted a lot more than other cars. But our shifting setup was pretty good and from my work this should have been a performance improvement (we only dyno'd about 68 HP that year and the car was just under 500 lb, yet we finished 3rd in Acceleration)

Anyway, I think the comparison you are trying to make is too simplistic, though it may show a big number in favor of your CVT.

I also agree with adrial, the unstable behavior you saw at the event that you attributed to shifting, is merely poor setup, or poor driver behavior. Not because the car is shifting.

I would guess if you watch the 'fast cars' part of the endurance race, they won't look like that.

[Biggy72]

If you want a non cv car to turn in like a cv car just hold the clutch in and see how fun it is... I personally like having the control of engine braking and being able to kick the rear out a little getting the car to rotate when entering a corner. With a little practice that should be considered fairly predictable and with a decent suspension setup it should be very driveable.

[rjwoods77]

Biggy,

If you setup a weak backshift then you will get a freewheeling effect liek you describe. You can actually setup a backshift so aggressive that it will almost lock the wheels if you so desired. It all depends on what you want for balance.

[jsmooz]

Rob - I'll credit the smooth as glass look to my awesome driving skills. http://fsae.com/groupee_common/emoticons/icon_smile.gif I also sometimes dislike the backshifting of the CVT. Even a decent driver would be able to use different amounts of powertrain drag in different corners.

Honestly, I think that it all has to do with driver skill. The benefit of the CVT is that the driver can focus on the line and not add shifting into the equation. Given that most teams never have enough drive time, the CVT puts you that much more ahead of the game.

With that said, I also agree with Charlie....the top teams are fast and smooth. However, this year we (Dearborn) were in the 'fast car' group for endurance. Granted we broke so we don't have a complete comparision but that should be some indication. Remember, while the uber gearhead, 'wanna go fast' in all of us makes the car complicated, the target audience is the weekend autocrosser. They could be fsater/more sucessful with a little less power and a CVT, that's what we've tried to prove.

[Pete M]

With driver-controlled shifts, the driver can pick the exact moment to lightly upset the rear end of the car on corner entry. Or choose not to if the they don't want to in that particular corner. CVTs reduce the number of things the driver has to keep in mind but they also reduce the amount of control the driver has over the car.

If your drivers are struggling to shift and drive at the same time, maybe you need to spend more time on driver training and less time on innovative gearboxes?

[rjwoods77]

Charlie and Jsmooz,

I totally agree with you about the top teams but it seemed like that number was about 15-20 out of a very large field. I am also with you on the driver input and target audience as well. I understand that not every shift is a perfect time "loss" since there are a bunch of things going on. I am just trying to quanitify the number shifts and do a best case worst case senario.