Optimum shift points for performance

[seangti]

On reading an article from Autospeed website (Autospeed - Finding the best gear shift points) I decided to try this on my car to find optimal shift points for maximum performance. That is, at what revs should I be changing at to get maximum acceleration? This is relevant for track days and drag racing, or the odd traffic light GP. Whilst I did this using a Gtech GPS lap timer, it’s absolutely possible to do with any smartphone with an accelerometer. You could try use your phone GPS, but I don’t think it’ll be sufficiently accurate to due to the refresh rate of a phone, and the IPhone accelerometers have been reported as being sufficiently accurate.

So far I’ve only done the second to third gear shift, I can/will easily overlay 4th when I have a sufficiently safe environment to stretch out 4th gear.

Anyway, enough waffle on how, the results. The short of it, for optimum shifts I should be shifting at 5800 rpm for 2nd to 3rd. Anything beyond that is wasted energy. I have the raw data, though yet to validate all this. It’s also not relevant for a 0-100km/h run. Not sure if the polo’s gearing would allow 2nd to stretch to 100km/h, though you’d be nudging or over 7000rpm I reckon (I have no interest in trying this).

The horizontal axis is in seconds as it’s the common point of reference. Km/h could be used if doing it from a phone as per Autospeed. As the gtech record at 10hz (10x second), it’s the easiest for my data collation. 5800rpm in second is ~90km/h, 90km/h in 3rd is about 3900rpm (as per 3rd picture below).


From this graph, you can also see my acceleration times, e.g 3rd gear 80-120km/h in 4.5seconds.


If you use your phone accelerometer, ensure its reading zero g’s before you start, more details in the Autospeed link. It will look something like this.




This also backs up my thoughts/feel for the car. On track days I very rarely rev beyond 6000rpm. Given the torque of these cars, short shifting at revs beyond 4500rpm (4.2seconds on x-axis) in 2nd also lands in the power band for 3rd gear.

For me, knowing this means I can shift at the most appropriate point, rather than revving it harder, which wastes time, generates much greater heat and added stress on engine and associated bits (thinking drive train and everything that runs off the timing belt, so alternator and air con bearings at a guess).

[rgvlee]

I usually just use dyno results to tell me when to change gears, not much point going past peak power

[seangti]

I usually just use dyno results to tell me when to change gears, not much point going past peak power

A dyno doesn't give you shift points, just which point max power is generated, which changes from gear to gear. It's communicated and displayed clearer in the Autospeed article towards the bottom. Whilst the article is based on a diesel, I am assuming these gear/power delivery characteristics are transferrable.

It's also to demonstrate that shifting at redline isn't necessarily the way to drive fast.

[s4mmie]

Brilliant data.
I've always felt this in my car too. Only from the feel, no hard data, but there's definitely a point in 2 gear where it's quicker to change into 3rd and let it do the work.
I confirmed this at Happy Laps too. From the logs I had it wound out to 6800rpms and TBH there was just no point...

[VWindahouse]

Nice work Sean!

To answer your question yes the Polo will just hit 110kph in 2nd then you will hit the fuel cutout. Also max speed in 3rd before fuel cutout is about 160kph.

So for 0-100kph and 1/4 mile times the polo's major failing is the short gear ratios for one reason and that's for both scenarios - you totally loose power and some momentum at around 90kph in 2nd doing the 0-100 run as you dramatically fall off the power band tempting you to change up to third where you loose massive time also purely because of the shift time. (My best time on the Gtech is 6.1 and that's been in 2nd gear ringing the balls off it)

And the same thing at about 160kph in 3rd doing the 1/4 mile. Just before the finish line you have the choice of changing up to 4th as you hit redline hard right about there. But funnily enough my best run of 13.9 was actually in 3rd under no power for the last 20 meters or so because I successfully found the fuel cutout My trap speed on that run was 160.98kph.

So my theory is if for example you got the 5 speed synchro gear set from SQS you can hit the following speeds in gear succession and therefore overcome the above mentioned issues for way better times;

1st - 81.2kph
2nd - 122.4kph
3rd - 179.4kph
4th - 245.5kph
5th - 294.3kph

all measured at 7000rpm.

What do you think

[rgvlee]

A dyno doesn't give you shift points, just which point max power is generated, which changes from gear to gear. It's communicated and displayed clearer in the Autospeed article towards the bottom. Whilst the article is based on a diesel, I am assuming these gear/power delivery characteristics are transferrable.

It's also to demonstrate that shifting at redline isn't necessarily the way to drive fast.

I don't consider myself an expert but I don't think engine max power or it's characteristics will change gear to gear. Your rate of acceleration will though.

[kaanage]

It does and it doesn't

Power can accelerate and it can force top speed. Purely for acceleration, the inertia of the car is acted upon by the engine power to give the accelerative force at any instant BUT the drive train inertia varies depending on the gear that you are in with lower gears having higher drivetrain inertia relative to the flywheel. Purely for speed, it is the air resistance of your car that is being overcome by the engine power and the drivetrain inertia is irrelevant.

In practice, you are always accelerating to some degree and the car is always meeting air resistance so the power usage is a combination of the 2 factors, varying with both speed and gearing (assuming drivetrain frictional losses are relatively constant in each gear which is true enough in modern indirect drive gearboxes).

But the best point to shift will definitely vary per gear as you are trying to keep the average power output at its maximum across the 2 shift points and the varying spacing of the gears mean that the rev gap narrows as your go up the gears. In practice, it is almost always somewhere between the power peak and the redline.

And there are operating practicalities like the time lost in a gear shift vs slight overrevving when the gap between 2 corners or to the end of the drag run only allows for the optimal higher gear to be run for a very short period. Now if we had cassette gearboxes...

[seangti]

Thanks for explaining kaanage. I feel I understand the application but my ability to communicate it effectively is yet to catch up. Would it be fair to suggest that as the speed increases, optimal shift point would closer reflect somewhere between Max torque and Max power, given that torque will over come resistance?

Anthony, from my understanding, the outcome to your approach would really only be applicable to reach that arbitrary figure or distance. So for quarter mile, shifting at optimal points for 1st and 2nd gear would accelerate faster than holding onto them till cutout, but holding 3rd to cross the line may be faster. Reason being, that once the gear is beyond it's efficient power delivery, the next gear WILL accelerate faster. This is probably best explained in the g-force graph, post 5800rpm for me, 3rd gear pulls more g's, which is accelerative force.

Also perhaps worth bearing in mind that each manual shift takes an average of 0.8 seconds which is also backed up from some runs I've done and Ben (aka Machine in his Golf R). If you run similar tests to me in gears 1-4 stretching them out to cut out using gtech (I'd need the raw data so must be the RR version), I could reasonably easily model the differing quarter mile times pending shift points. Though I'm not keen revving my car into/beyond red line, though your mods and tune may/could likely yield differing results to my entry stage 1 level of mods.

I had some discussion on the mkVl forum comparing DSG to manual (specific to R's), DSG shifting in about 0.1 secs. That doesn't mean you lose 0.7 off your time per shift, just 0.7 of full throttle per shift, I.e. robbing 1.5+sec of foot to floor board over a ~14second run.

Food for thought hey. (PS, if Im interpreting any of the data incorrectly, please correct me)

[s4mmie]

I'm not keen revving my car into/beyond red line...

You've got to burn the cobwebs out somehow!

[seangti]

You've got to burn the cobwebs out somehow!

Haha, gotta get the carbon build up out somehow hey

[VWindahouse]

Anthony, from my understanding, the outcome to your approach would really only be applicable to reach that arbitrary figure or distance. So for quarter mile, shifting at optimal points for 1st and 2nd gear would accelerate faster than holding onto them till cutout, but holding 3rd to cross the line may be faster. Reason being, that once the gear is beyond it's efficient power delivery, the next gear WILL accelerate faster. This is probably best explained in the g-force graph, post 5800rpm for me, 3rd gear pulls more g's, which is accelerative force.

Also perhaps worth bearing in mind that each manual shift takes an average of 0.8 seconds which is also backed up from some runs I've done and Ben (aka Machine in his Golf R). If you run similar tests to me in gears 1-4 stretching them out to cut out using gtech (I'd need the raw data so must be the RR version), I could reasonably easily model the differing quarter mile times pending shift points. Though I'm not keen revving my car into/beyond red line, though your mods and tune may/could likely yield differing results to my entry stage 1 level of mods.

I had some discussion on the mkVl forum comparing DSG to manual (specific to R's), DSG shifting in about 0.1 secs. That doesn't mean you lose 0.7 off your time per shift, just 0.7 of full throttle per shift, I.e. robbing 1.5+sec of foot to floor board over a ~14second run.

Food for thought hey. (PS, if Im interpreting any of the data incorrectly, please correct me)

No you're correct. I shift first at about 5500-6000 in first on the 1/4 and 6-7000 for the 2nd - 3rd. But as you stated a gear shift can rob you of .7 seconds so my logic for the quarter with the SQS gear set is save the time loss from the 3rd to 4th gear change but continue velocity over the line in 3rd by revving out to 7000rpm and therefore hitting a trap speed at that point of around the 179kph mark.

As opposed to 160kph and coasting over the line There has to be a few 10ths in that I would think.

[jasn78]

Hey guys, I will put my non-scientific but personal experience into this. I am honestly finding for qtr miles runs even with mods there is no point revving beyond 6grand. Obviously I am not quick enough to do this in first, but every other gear change I try my damndest to keep it under that mark and seems to work ok.

[wai]

Power and torque curves have to be read very carefully.

Generally, the curves are obtained at wide open throttle. The curves will change depending on the percentage of throttle used, and so you have to look at them to work things out properly. The part throttle curves can be very different to the wide open throttle curves.

[kaanage]

yes. understood. torque peak shifts lower down the rev range with less throttle due to pumping losses being constant vs power produced per stroke decreasing.

This thread is about maximising acceleration so unless the car has very bad tuning of some sort so it's stumbling at some point of the rev range, then WOT is a given.

Back on topic, for peak performance, you generally want to shift past the power peak so that the engine has delivered the most that it can and you would like to shift into as high in the power curve as possible so the engine keeps driving as hard as possible. The torque peak (where the most energy is delivered per stroke) is basically irrelevant as it is the RATE of energy delivery is what matters (the actual definition of power) and this is the product of the energy is delivered per stroke x number of strokes (torque x rpm).

Operational impediments are the only reason the don't do this when trying to maximise acceleration.

Going beyond the redline is a bad idea from the engine longevity point of view and also many engines will drop power dramatically beyond the power peak any way (be it naturally or with rev limiters).

jasn78, at what rpm does your engine deliver peak power?