2014-2020 Formula One 1.6l V6 turbo engine formula

[johnny comelately]

What about henry's post? This measures total PU power, doesn't care of it's electric or there's hamsters turning a wheel.

re Henry page 600: FIA traction rules.
and keeping it in context, launch is a launch is a launch the idea being to get away the quickest possible; so no electric shove just horsepower and optimised tyre spin and FIA weight distribution it is giving me an indicative power.
With your figures adjusted by a bit for extra 180 kg (640 - 670hp) of mass weight (by the time honoured method of that looks about right ) it is looking around 760 to 800 hp ; what do you think?

I think you can't understand very simple things that have already been explained to you several times.

You are possibly right (although I am sure you are not), but that doesn't mean you can be reproachful in such a way.
I do look on this forum as a cooperative learning resource.
That is best done by respectful, civilised and professional discussion.
I do appreciate your efforts though.

[henry]

I had a go at calculating power from onboard acceleration - more of a proof of concept rather than an attempt to work out exact power.

......

A really interesting post.

I have used your raw data, I hope you don’t mind,and a small variation on your method.

The change I made was to smooth the kph and rpm against time curves. I reasoned that they would be pretty smooth naturally. I used a cubic fit which looked “OK”. I only did one gear, the lower one.

I used the rest of your method and data for Cd etc although I skipped gear ratio because I was only interested in power rather than torque.

The result surprised me. I got a nice smooth curve of power( at the road) versus rpm, with a low of 643 kw and a high of 717. A bigger range than I expected. But the real surprise was that the peak was at 11,500 rpm, with 643 at 10,760 and 681 at 11,770.

The absolute power values look too high, Cd, frontal area and density are all pretty influential.

Thanks again for such a useful resource. I might try the other gear and different fits for the curves if I find an opportunity.

One non power unit explanation for some of the offset of power to rpm might be that the road speed and rpm numbers are out of synch. If the asynchronous is constant I might test that as well.

[63l8qrrfy6]

I had a go at calculating power from onboard acceleration - more of a proof of concept rather than an attempt to work out exact power.

......

A really interesting post.

I have used your raw data, I hope you don’t mind,and a small variation on your method.

The change I made was to smooth the kph and rpm against time curves. I reasoned that they would be pretty smooth naturally. I used a cubic fit which looked “OK”. I only did one gear, the lower one.

I used the rest of your method and data for Cd etc although I skipped gear ratio because I was only interested in power rather than torque.

The result surprised me. I got a nice smooth curve of power( at the road) versus rpm, with a low of 643 kw and a high of 717. A bigger range than I expected. But the real surprise was that the peak was at 11,500 rpm, with 643 at 10,760 and 681 at 11,770.

The absolute power values look too high, Cd, frontal area and density are all pretty influential.

Thanks again for such a useful resource. I might try the other gear and different fits for the curves if I find an opportunity.

One non power unit explanation for some of the offset of power to rpm might be that the road speed and rpm numbers are out of synch. If the asynchronous is constant I might test that as well.

I would actually encourage anyone to play with different things and see how this can be made to work better.
I have intentionally avoided any curve fits, averaging or smoothing to highlight the acceleration spikes caused by the speedo readings (which I think are the main limitation).

It would be very interesting if someone had time to do whole laps and/or other cars. Then we could understand differences in drag from cornering performance, differences in drag with and without DRS, etc. Another interesting thing would be to compare high drag performance to low drag torque limited acceleration.

At the end of the day if we manage to predict power delta between cars I would say this has served its purpose.

[godlameroso]

I had a go at calculating power from onboard acceleration - more of a proof of concept rather than an attempt to work out exact power.

......

A really interesting post.

I have used your raw data, I hope you don’t mind,and a small variation on your method.

The change I made was to smooth the kph and rpm against time curves. I reasoned that they would be pretty smooth naturally. I used a cubic fit which looked “OK”. I only did one gear, the lower one.

I used the rest of your method and data for Cd etc although I skipped gear ratio because I was only interested in power rather than torque.

The result surprised me. I got a nice smooth curve of power( at the road) versus rpm, with a low of 643 kw and a high of 717. A bigger range than I expected. But the real surprise was that the peak was at 11,500 rpm, with 643 at 10,760 and 681 at 11,770.

The absolute power values look too high, Cd, frontal area and density are all pretty influential.

Thanks again for such a useful resource. I might try the other gear and different fits for the curves if I find an opportunity.

One non power unit explanation for some of the offset of power to rpm might be that the road speed and rpm numbers are out of synch. If the asynchronous is constant I might test that as well.

I would actually encourage anyone to play with different things and see how this can be made to work better.
I have intentionally avoided any curve fits, averaging or smoothing to highlight the acceleration spikes caused by the speedo readings (which I think are the main limitation).

It would be very interesting if someone had time to do whole laps and/or other cars. Then we could understand differences in drag from cornering performance, differences in drag with and without DRS, etc. Another interesting thing would be to compare high drag performance to low drag torque limited acceleration.

At the end of the day if we manage to predict power delta between cars I would say this has served its purpose.

A poster on here used to do the qualifying laps and he'd fix the telemetry on the videos so it synced properly. Shame they shut down his channel, F1 should give him a job.

[henry]

I think it will be very difficult to establish any absolute power numbers.

I think we have a better chance of understanding the relative power unit performances and particularly the shape of the power curves. It would be interesting to see the shape of the Mercedes power curve, the one I have extracted from your data for Ferrari is surprisingly peaky.

[ENGINE TUNER]

re Power:
How much is harvested on the warm up lap to assist the ICE at launch?

It depends what you mean by launch.

Off the line, from 0 kph, they can’t use the MGU-K, its use is not allowed until they reach 100 kph. After that they are traction limited to around 160 kph.

They might use some energy from the ES to drive the compressor right at the start, but probably not.

So from that perspective they harvest zero energy specifically to assist launch.

However I would expect them to come to the line with a fully charged ES since maximising first lap power is vital.

Are they forbidden to have the MGU-k active in any fashion or just as a motor?

They are prohibited from harvesting with the mguk when the driver is "demanding torque", aka applying any throttle.

[gruntguru]

Up-votes to Mudflap and Henry for quality effort.

It is theoretically possible to estimate drag losses from the acceleration data, since aero drag power loss is proportional to (approx) velocity cubed, while acceleration is proportional to whatever-power-remains, divided by velocity. So if there were no losses, the acceleration at 300 km/hr would be half the acceleration at 150 km/hr and the same engine power. If we introduce aerodynamic drag the acceleration at 300 will be less than half the acceleration 150 - how much less will depend on Cd.A.

Other complicating factors include inertia of rotating parts. Final drive and wheels simply increase the apparent mass of the car. Moving parts (including pistons, con-rods, valves etc) upstream of the gearbox will increase the apparent mass of the car by an amount which is larger (proportional to overall gear ratio squared) and dependent on the gear selected. So the car has more inertia in the lower gears.

johnny - when the car is traction limited, the acceleration is no different whether there is 500 hp available to spin the tyres or 5000 so this part of the acceleration profile cannot be used to calculate effective power.

[trinidefender]

I had a go at calculating power from onboard acceleration - more of a proof of concept rather than an attempt to work out exact power.

I started off with Seb's Singapore pole lap from last year (https://www.youtube.com/watch?v=xPXfr3f0veI) and used http://www.watchframebyframe.com to extract engine and car speed versus time. Interestingly this website would not work for official F1 videos (apparently FIA is not happy with people analyzing on board videos frame by frame). Only the first straight was analyzed as I can't be arsed to go through the whole lap frame by frame.
Unfortunately the video is only 25 FPS so the resolution (as in sampling frequency) is very poor (and so is that digital speedo), as a result I had to skip ahead frames until engine speed changed, which means that the time interval is not constant and samples are shite. Normally you'd probably need good data every hundreth of a second to be able to come up with reasonable numbers.

From speed vs time I calculated acceleration and then inertia load assuming a car mass of 628 kg.
Gear ratio was calculated from engine speed vs car speed (wheel radius =335 mm). I got a ratio of 5.84 for the 6th gear, 5.16 for the 7th and 4.55 for the 8th gear. These are pretty close to ratios calculated here forum/viewtopic.php?f=4&t=26230.

Drag was calculated using the average car speed over the time interval. (Cd=1 and frontal area of 1.442 m^2, someone posted these a while ago).

Wheel torque was then calculated by multiplying the sum of drag and inertia loads by the wheel radius. Dividing wheel torque by gear ratio gives engine torque. Multiplying by engine speed gave engine power. No transmission losses nor rolling resistance considered.

https://i.imgur.com/eYoniGM.png
https://i.imgur.com/mEGtFTd.png
https://i.imgur.com/5pP7wJ1.png
https://i.imgur.com/BOJNcH1.png

Greyed out rows mean funny data due to gearshifts. Last data point is also dodgy - I think it includes braking at end of straight.

You used 628 kg for the calculations.

According to this:
https://www.formula1.com/en/latest/tech ... 2018-.html
The minimum car mass is:
734 kg
66-81 kg for average driver (my personal guess)
100-105 kg for fuel
For a total mass of 900-920 kg

Do you mind running the numbers again with this mass?

[J.A.W.]

What about the post-race minimum weight measure?

& the best possible performance - near that min' weight?

Such as the Belgian G.P. @ Spa - last year, when the safety car left, & a full-out race-resumption, was on..

[henry]

I had a go at calculating power from onboard acceleration - more of a proof of concept rather than an attempt to work out exact power.

I started off with Seb's Singapore pole lap from last year (https://www.youtube.com/watch?v=xPXfr3f0veI) and used http://www.watchframebyframe.com to extract engine and car speed versus time. Interestingly this website would not work for official F1 videos (apparently FIA is not happy with people analyzing on board videos frame by frame). Only the first straight was analyzed as I can't be arsed to go through the whole lap frame by frame.
Unfortunately the video is only 25 FPS so the resolution (as in sampling frequency) is very poor (and so is that digital speedo), as a result I had to skip ahead frames until engine speed changed, which means that the time interval is not constant and samples are shite. Normally you'd probably need good data every hundreth of a second to be able to come up with reasonable numbers.

From speed vs time I calculated acceleration and then inertia load assuming a car mass of 628 kg.
Gear ratio was calculated from engine speed vs car speed (wheel radius =335 mm). I got a ratio of 5.84 for the 6th gear, 5.16 for the 7th and 4.55 for the 8th gear. These are pretty close to ratios calculated here forum/viewtopic.php?f=4&t=26230.

Drag was calculated using the average car speed over the time interval. (Cd=1 and frontal area of 1.442 m^2, someone posted these a while ago).

Wheel torque was then calculated by multiplying the sum of drag and inertia loads by the wheel radius. Dividing wheel torque by gear ratio gives engine torque. Multiplying by engine speed gave engine power. No transmission losses nor rolling resistance considered.

https://i.imgur.com/eYoniGM.png
https://i.imgur.com/mEGtFTd.png
https://i.imgur.com/5pP7wJ1.png
https://i.imgur.com/BOJNcH1.png

Greyed out rows mean funny data due to gearshifts. Last data point is also dodgy - I think it includes braking at end of straight.

You used 628 kg for the calculations.

According to this:
https://www.formula1.com/en/latest/tech ... 2018-.html
The minimum car mass is:
734 kg
66-81 kg for average driver (my personal guess)
100-105 kg for fuel
For a total mass of 900-920 kg

Do you mind running the numbers again with this mass?

The 734 kg includes the driver. In 2017 it was 728.

The lap @mudflap used was in qualification so only a small amount of fuel. In my calculations I used an all up weight of 740 kg.

As @gruntguru pointed out I should have included some rotational inertia. Given I used to make a living from that understanding it’s a bit of a schoolboy error to leave it out.

[Tommy Cookers]

[quote="ENGINE TUNER"]
They are prohibited from harvesting with the mguk when the driver is "demanding torque", aka applying any throttle[/quote]

in what rule did it say this ?
(the mapping rules are for accelerator/PU torque relationship not accelerator/ICE torque relationship)
haven't we seen they do this (what you say is prohibited) ? - eg from telemetry issued to the media 1 or 2 years ago
which gave some of us a cosy feeling from apparent vindication


regarding henry's apparent discovery of (Ferrari's) peak power being at 11500 rpm ie definitely not at 10500 ......
suggests that peak fuelling rate is not at 10500
again I feel cosy

as the cars have no CVT time spent between 10500 and 12000 ish rpm dominates race performance
fuelling eg at 95 kg/hr at 10500 and 100 kg/hr at 11500 is more efficient as AFR & boost are near optimal throughout
fuelling at 100 kg/hr at 10500 and 11500 does not allow optimal AFR & boost throughout

this might not apply at lower speeds ie ripping through the gears, when brief and small fuel accumulations are possible

[henry]

I took another look at @mudflap’s data.

The engine speed is the more erratic of the measurements so I smoothed the road speed and took the engine speed as a fixed ratio (rpm/kph). I calculated power for both gears using mass 740kg, drag coefficient 1.0 and frontal area 1.44 m^2. Somewhere I have a number for the mass equivalence of the rotating components but given the malign influence of aerodynamics it probably makes little difference.

The two power against rpm curves were hopelessly misaligned.

I looked for a parameter that would more reasonably align them and, not surprisingly, it was CdA.
If I reduced CdA with speed there was a more consistent view of the two. I used a linear change from 0.97 to 0.85 from 260 kph to 312. Entirely arbitrary but the results for the two gears matched up better.

Edit: I looked up prof Limebeer’s optimum control paper from 2013 To check the drag coefficients. He shows around 0.92 and 0.83 at those two speeds and a relatively linear drop. So probably not a long way off. /edit

The absolute power values are still somewhat nonsensical with a peak power of 715kw at 11300 rpm

The curve is pretty symmetrical about this value and the end points 50 kw or so lower than the peak. It is not as flat as I imagined it would be.

The power curve in the higher gear starts falling at 10800 rpm which suggests that even on a qualifying lap they switch off electrical assistance at the end of the straight. The reduction is large enough to suggest that the K is switched off and the H is used to charge the ES at this point.

The extreme sensitivity to drag emphasises the efforts that the teams put into managing it at high speed. Combined with the shape of the power curve it confirms that trying to estimate power from speed traps is chancy.

I was going to mention that @tommy cookers would not find the positioning of peak power on the rpm curve odd. He beat me to it.

[godlameroso]

They are prohibited from harvesting with the mguk when the driver is "demanding torque", aka applying any throttle

in what rule did it say this ?
(the mapping rules are for accelerator/PU torque relationship not accelerator/ICE torque relationship)
haven't we seen they do this (what you say is prohibited) ? - eg from telemetry issued to the media 1 or 2 years ago
which gave some of us a cosy feeling from apparent vindication


regarding henry's apparent discovery of (Ferrari's) peak power being at 11500 rpm ie definitely not at 10500 ......
suggests that peak fuelling rate is not at 10500
again I feel cosy

as the cars have no CVT time spent between 10500 and 12000 ish rpm dominates race performance
fuelling eg at 95 kg/hr at 10500 and 100 kg/hr at 11500 is more efficient as AFR & boost are near optimal throughout
fuelling at 100 kg/hr at 10500 and 11500 does not allow optimal AFR & boost throughout

this might not apply at lower speeds ie ripping through the gears, when brief and small fuel accumulations are possible

If the EQ ratio is as low as we've postulated then it stands to reason that the ideal mixture won't necessarily correspond with max fuel flow allowed for a given RPM.

[dren]

It's to allow a wider RPM range for the 'right' operating conditions since no CVT is used as TC said.

[3jawchuck]

this is presumably an exothermic reaction
so the air's nitrogen is acting as a fuel ? but (like 'oil') is somehow not subject to the 100 kg limit ?
....

I have no clue as to how this would work. I hope you don't mind me asking, but do you have any suggestions on how I can better my understanding in regards to the bolded statement?