Those of you who remember Aris' tutorial video about getting the most out of the engine might recall that one of the cars he talked about was the M235i.
At about 4m30s into the video:
he talks about not really gaining anything by accelerating beyond about 6000 rpm. Now, this didn't seem quite right to me at the time, because the power curve he's showing actually peaks at 6000 rpm, so clearly you need to take the revs well beyond that point to get optimum shifts. (When he then drives the car on the track, he chooses to shift at a little more than 6k - roughly 6300 rpm.)
Based on how slowly the power curve dropped after the peak, my feeling was that you would need to go pretty close to the red line to get the best shifts and that's how I was driving it myself. I later started using Sidekick (great app), which computes the optimum shift points in real time, and it also took me very close to the red line for most shifts.
When we recently decided to use the M235i as the new car for the Tuesday night social events, we considered using a restrictor to make it easier to handle (we decided against that in the end though!
) and because of this I fired up my half-baked homegrown telemetry app and calculated the power curves with and without 'restrictor=100'.
To my surprise, I found that the stock power curve which you see in the launcher is incorrect. In essence, the true power peaks a little earlier (~ 5700 rpm) and falls much faster after the peak - see below for graphs.
The curve you see in the launcher is stored in the ui_car.json file, in the
C:\SteamLibrary\steamapps\common\assettocorsa\content\cars\ks_bmw_m235i_racing\ui
folder, and that happens to be the same file that Sidekick uses to work out the optimal shift points...
With some input from @RasmusP I have concocted a new ui_car.json file (attached) which can be dropped over the old one if you use Sidekick and want the optimal shifts to work properly (or if you just want to see the true power curve in the launcher).
tl;dr: optimal shift points for the real power curve are (gear, shift-rpm):
1 6850
2 6826
3 6285
4 6377
5 6343
6 6089
i.e. in 4th gear, you want to be shifting to 5th at about 6377 rpm. (You won't ever want to shift from 7th to 8th unless you're short of fuel (!) because the revs in 7th never even reach the power peak.)
And yeah, the irony isn't lost on me - although the logic Aris used was broken, the actual shifting RPM he used on track (~6300 rpm) was more or less bang on, once you're out of second gear If you didn't know any better you'd think he somehow knew the optimum shift points!
Btw, I haven't (yet!) worked out the actual impact of shifting too late. It's likely to be only a few tenths of a second around a typical lap though...
For the original (aka wrong) power curve, the optimal shift points would have been:
1 limiter
2 limiter
3 6868
4 6951
5 6919
6 6630
Some comparison curves below...
NB: I scaled up the measured power by around 17% to make the old and new curves have roughly the same peak power. (This means the peak torque is significantly different though.)
In the curve below, you can see (for example) that the acceleration in 4th gear drops below the acceleration in 5th long before you reach the limiter in 4th. You'd want to be shifting to 5th at around 153 kph.
At about 4m30s into the video:
Based on how slowly the power curve dropped after the peak, my feeling was that you would need to go pretty close to the red line to get the best shifts and that's how I was driving it myself. I later started using Sidekick (great app), which computes the optimum shift points in real time, and it also took me very close to the red line for most shifts.
When we recently decided to use the M235i as the new car for the Tuesday night social events, we considered using a restrictor to make it easier to handle (we decided against that in the end though!
) and because of this I fired up my half-baked homegrown telemetry app and calculated the power curves with and without 'restrictor=100'.To my surprise, I found that the stock power curve which you see in the launcher is incorrect. In essence, the true power peaks a little earlier (~ 5700 rpm) and falls much faster after the peak - see below for graphs.
The curve you see in the launcher is stored in the ui_car.json file, in the
C:\SteamLibrary\steamapps\common\assettocorsa\content\cars\ks_bmw_m235i_racing\ui
folder, and that happens to be the same file that Sidekick uses to work out the optimal shift points...
With some input from @RasmusP I have concocted a new ui_car.json file (attached) which can be dropped over the old one if you use Sidekick and want the optimal shifts to work properly (or if you just want to see the true power curve in the launcher).
tl;dr: optimal shift points for the real power curve are (gear, shift-rpm):
1 6850
2 6826
3 6285
4 6377
5 6343
6 6089
i.e. in 4th gear, you want to be shifting to 5th at about 6377 rpm. (You won't ever want to shift from 7th to 8th unless you're short of fuel (!) because the revs in 7th never even reach the power peak.)
And yeah, the irony isn't lost on me - although the logic Aris used was broken, the actual shifting RPM he used on track (~6300 rpm) was more or less bang on, once you're out of second gear
If you didn't know any better you'd think he somehow knew the optimum shift points! Btw, I haven't (yet!) worked out the actual impact of shifting too late. It's likely to be only a few tenths of a second around a typical lap though...
For the original (aka wrong) power curve, the optimal shift points would have been:
1 limiter
2 limiter
3 6868
4 6951
5 6919
6 6630
Some comparison curves below...
NB: I scaled up the measured power by around 17% to make the old and new curves have roughly the same peak power. (This means the peak torque is significantly different though.)
In the curve below, you can see (for example) that the acceleration in 4th gear drops below the acceleration in 5th long before you reach the limiter in 4th. You'd want to be shifting to 5th at around 153 kph.
Last edited:
