## Can the MGU-K harvest or deploy under its own control? And related questions.

All that has to do with the power train, gearbox, clutch, fuels and lubricants, etc. Generally the mechanical side of Formula One.
saviour stivala
-13
Joined: Wed Apr 25, 2018 11:54 am

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

“Today’s ERS takes the concept of KERS to another level, combining twice the power with a performance effect around ten times greater. ERS comprises two motor generator units (MGU-K and MGU-H), plus an energy store (ES) and a control electronics (CE). The motor generator units converts mechanical and heat energy to electrical energy and vice versa. MGU-K (where ‘K’ stands for kinetic) works like an uprated version of the previous KERS, converting kinetic energy generated under braking into electricity (rather than it escaping as heat). It also acts as a motor under acceleration, returning up to 120kw (approximately 160bhp) power to the drivetrain”.

henry
242
Joined: Mon Feb 23, 2004 7:49 pm
Location: England

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

Last try.

For anyone who’s interested a really useful resource on this topic is in the Honda thread.viewtopic.php?p=733944#p733944

It’s a magazine article of Honda’s developments.
On page 019 of that article is telemetry showing PU behaviour in qualifying at Barcelona.

The trace marked Opening Sprint is Accelerator Position
The trace marked Gap Position is Gear Selection

Looking at the area marked as Turn 3 we can see:

Up to the Apex the MGU-K generating during braking, with spikes where it stops briefly during gear shifts.

After the Apex the MGU-K continues to generate slowly transitioning to driving mirroring the driver accelerator demand. This is an example of the ECU deciding the mix of ICE and MGU-K to meet the driver demand, and harvesting to the ES, extra fuel is being burned to charge the ES.

Moving to turn 9.

The driver makes a partial lift and the MGU-K goes to full regen before slowly returning to driving again. By slowly I mean over a period of about 0.3 seconds.

It is a possibility that as well as lifting the driver also presses the brake pedal. I haven’t seen an example of a driver braking through turn 9 but maybe others have.

This set of data also shows other behaviours. Spikes of MGU-K activity through upshifts and high MGU-H activity at the beginning of straights where the wastegates are open and it is driving the compressor.

In this magazine article there is lots of other information on operating modes and the engineering associated, it shows how the complex engineering reality differs from the simple marketing message.

I continue to be very grateful to @Hino for posting it. It would be great to see the equivalent for the other PU manufacturers.
Fortune favours the prepared; she has no favourites and takes no sides.
Truth is confirmed by inspection and delay; falsehood by haste and uncertainty : Tacitus

roon
444
Joined: Sat Dec 17, 2016 6:04 pm

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

saviour stivala wrote:
Mon May 20, 2019 6:22 am
Unlike some on here, Whiting’s position couldn’t allow him to just blab blab blab when explaining FIA technical rules.
Sadly you'll never be able to appreciate the irony of your statement, ss.

subcritical71
71
Joined: Tue Jul 17, 2018 7:04 pm
Location: USA-Florida

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

roon wrote:
Mon May 20, 2019 2:03 pm
saviour stivala wrote:
Mon May 20, 2019 6:22 am
Unlike some on here, Whiting’s position couldn’t allow him to just blab blab blab when explaining FIA technical rules.
Sadly you'll never be able to appreciate the irony of your statement, ss.
I also think Charlie’s quote would be different if talking to a reporter vs. technical delegate from a team. He was talking to a reporter here and did not go into the intricate detail of every case the light would turn on.

saviour stivala
-13
Joined: Wed Apr 25, 2018 11:54 am

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

I have no problem with choosing a Japanese F1 magazine take on a formula one power unit matter under discussion over that of the official formula one site and the official FIA spokesman on the matter.

nzjrs
59
Joined: Wed Jan 07, 2015 10:21 am
Location: Austria

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

saviour stivala wrote:
Mon May 20, 2019 7:46 pm
I have no problem with choosing a Japanese F1 magazine take on a formula one power unit matter under discussion over that of the official formula one site and the official FIA spokesman on the matter.
At least three threads dedicated to you, several warnings by the mods, engagement by a large number of posters, data directly demonstrating your misunderstanding, insulting people in Italian, 150 negative votes and it finally ends with "fake news".

ian_s
14
Joined: Tue Feb 03, 2009 1:44 pm
Location: Medway Towns

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

nzjrs wrote:
Mon May 20, 2019 7:59 pm
saviour stivala wrote:
Mon May 20, 2019 7:46 pm
I have no problem with choosing a Japanese F1 magazine take on a formula one power unit matter under discussion over that of the official formula one site and the official FIA spokesman on the matter.
At least three threads dedicated to you, several warnings by the mods, engagement by a large number of posters, data directly demonstrating your misunderstanding, insulting people in Italian, 150 negative votes and it finally ends with "fake news".
if the data plots in this article were false, another team would have raised a protest at some point claiming Honda are doing something wrong. That hasnt happened as far as we know, so to me the best assumption is that they *all* harvest while under throttle at some point. it's unlikely to be at high power demand, more likely at lower demand in my mind.
take these figures for an example:
MGU-k output = 120kw
ICE output = 880kw
total output = 1000kw ( nice round number for simplicity)

the driver demands 20 Kw at 20% throttle. Ice actually runs at 140 Kw and K harvests at 120Kw: total output 20Kw
the driver demands 40 Kw at 40% throttle. Ice actually runs at 160 Kw and K harvests at 120Kw: total output 40Kw

It's never going to the this simple as 0% throttle should be equal to engine braking so less than 0 Kw output, but it demonstrates that you can power the ICE higher than the demanded 'Torque' but the actual torque is reduced by the MGU-K

henry
242
Joined: Mon Feb 23, 2004 7:49 pm
Location: England

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

ian_s wrote:
Mon May 20, 2019 9:41 pm
nzjrs wrote:
Mon May 20, 2019 7:59 pm
saviour stivala wrote:
Mon May 20, 2019 7:46 pm
I have no problem with choosing a Japanese F1 magazine take on a formula one power unit matter under discussion over that of the official formula one site and the official FIA spokesman on the matter.
At least three threads dedicated to you, several warnings by the mods, engagement by a large number of posters, data directly demonstrating your misunderstanding, insulting people in Italian, 150 negative votes and it finally ends with "fake news".
if the data plots in this article were false, another team would have raised a protest at some point claiming Honda are doing something wrong. That hasnt happened as far as we know, so to me the best assumption is that they *all* harvest while under throttle at some point. it's unlikely to be at high power demand, more likely at lower demand in my mind.
take these figures for an example:
MGU-k output = 120kw
ICE output = 880kw
total output = 1000kw ( nice round number for simplicity)

the driver demands 20 Kw at 20% throttle. Ice actually runs at 140 Kw and K harvests at 120Kw: total output 20Kw
the driver demands 40 Kw at 40% throttle. Ice actually runs at 160 Kw and K harvests at 120Kw: total output 40Kw

It's never going to the this simple as 0% throttle should be equal to engine braking so less than 0 Kw output, but it demonstrates that you can power the ICE higher than the demanded 'Torque' but the actual torque is reduced by the MGU-K
I think you’re right and driving against the K mainly happens at part throttle.

However, last year Amus published some data comparing Mercedes and Ferrari at the end of straight at Hochenheim. When I wrote a simulation the only way I could match the Ferrari data was to have both K and H generating simultaneously at the max throttle at the very end of the straight. So I hold the opinion that it is possible that sometimes it may happen that a team might choose that option.
Fortune favours the prepared; she has no favourites and takes no sides.
Truth is confirmed by inspection and delay; falsehood by haste and uncertainty : Tacitus

gruntguru
437
Joined: Sat Feb 21, 2009 6:43 am

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

saviour stivala wrote:
Mon May 20, 2019 7:46 pm
I have no problem with choosing a Japanese F1 magazine take on a formula one power unit matter under discussion over that of the official formula one site and the official FIA spokesman on the matter.
Ian_s has already alluded to this point but let me try to make it a little more clear.

You don't need to believe the text of the Japanese magazine article to appreciate Henry's point. You only need to look at the data plot and follow Henry's description.

There is only one way that Henry can be wrong here and that is if the data plot is a complete fabrication. If you believe that, you need to ask yourself:

Why did the magazine fabricate the plot? The article was about "extra harvest" and the data Henry points to is not pertinent to the "extra harvest" function so why fabricate that?
je suis charlie

Tzk
20
Joined: Sat Jul 28, 2018 11:49 am

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

henry wrote:However, last year Amus published some data comparing Mercedes and Ferrari at the end of straight at Hochenheim. When I wrote a simulation the only way I could match the Ferrari data was to have both K and H generating simultaneously at the max throttle at the very end of the straight. So I hold the opinion that it is possible that sometimes it may happen that a team might choose that option.
This might happen when the engine can harvest enough to not run out of battery on the next straight. Remember that the rules state that the batteries state of charge may differ by 4mj between empty and full.

So lets assume that you would overcharge the battery on the next
Braking zone, why wouldn‘t you use the mgu-k for the full straight right before said zone?!

henry
242
Joined: Mon Feb 23, 2004 7:49 pm
Location: England

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

Tzk wrote:
Tue May 21, 2019 9:09 pm
henry wrote:However, last year Amus published some data comparing Mercedes and Ferrari at the end of straight at Hochenheim. When I wrote a simulation the only way I could match the Ferrari data was to have both K and H generating simultaneously at the max throttle at the very end of the straight. So I hold the opinion that it is possible that sometimes it may happen that a team might choose that option.
This might happen when the engine can harvest enough to not run out of battery on the next straight. Remember that the rules state that the batteries state of charge may differ by 4mj between empty and full.

So lets assume that you would overcharge the battery on the next
Braking zone, why wouldn‘t you use the mgu-k for the full straight right before said zone?!
I didn’t mention that the MGU-K was used to drive the Ferrari a little longer than the Mercedes, but you’re right that this was in fact the operating mode I simulated.

I don’t think this would be influenced by SOC. On the straight the the PU will have been run for a short period in E-boost, electric supercharger, drawing the ES at maybe 180kW or so, 120 to the K and 60 to the H to drive the compressor. It will have then run the rest of the straight at 120kW. So the SOC will have been depleted considerably more than could be recuperated in the braking zone.

An implication of Ferrari’s strategy was that they could afford to charge the ES at a very high rate, 180kW maybe more. This is in comparison to Mercedes’ more conservative approach where the ES is charged only by the H at perhaps 60 or 70kW.
Fortune favours the prepared; she has no favourites and takes no sides.
Truth is confirmed by inspection and delay; falsehood by haste and uncertainty : Tacitus

saviour stivala
-13
Joined: Wed Apr 25, 2018 11:54 am

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

Since the introduction of the hybrid power unit to this day there was no change in regulations as regards electrical energy flow. For the first three or so years it was being said that teams were struggling to even harvest the permitted 2mj per lap by the MGU-K under braking at least on some tracks and so a fully charged ES was hard to maintain for lap after lap, at that time MGU-H turbo combination was still not as developed as it is today. What was back than preventing teams harvesting by the MGU-K when not using the brakes?. Nowadays it is said that thanks to MGU-H turbo combination development progress teams are able to deploy electrical power virtually for the whole lap. And that ability is not due to MGU-K harvesting but all due to MGU-H turbo combination development progress.

Dr. Acula
42
Joined: Sat Jul 28, 2018 12:23 pm

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

saviour stivala wrote:
Wed May 22, 2019 1:47 pm
What was back than preventing teams harvesting by the MGU-K when not using the brakes?
Well, maybe the fact that it would be incredibly inefficent under race conditions to do this when the cars are basically energy limited because of the limited amount of fuel?
If you harvest under off throttle conditions, regardless if braking or just lift and coast, you gain overall system efficiency because otherwise the energy would be wasted either by the brakes or simply by aerodynamic drag.
If you use the ICE directly to shove some energy to the MGU-K you already start down at somewhere around 40-45% thermal efficency and it only gets worse from there. So you lose overall efficiency.

Also let us do here a short calculation. How much speed would a F1 at car at 300km/h and 800kg mass actually lose when the driver does lift and coast for 1 second and the MGU-K would start to harvest with 120kW?

$\frac{800*(300/3.6)^2}{2}=2.777MJ$

$2.777-(0.12*1)=2.657MJ$

$\sqrt{\frac{2.657*2}{800}}*3.6=293.45km/h$

So it would lose about 6.5km/h...that's not very massive. Also if a car behind would keep its speed of 300km/h during this 1 second it wouldn't even gain a full meter. But the car in front has won one second of MGU-K boost which it can use very efficently when it accelerates out of a slow corner.

henry
242
Joined: Mon Feb 23, 2004 7:49 pm
Location: England

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

Dr. Acula wrote:
Wed May 22, 2019 3:09 pm
saviour stivala wrote:
Wed May 22, 2019 1:47 pm
What was back than preventing teams harvesting by the MGU-K when not using the brakes?
Well, maybe the fact that it would be incredibly inefficent under race conditions to do this when the cars are basically energy limited because of the limited amount of fuel?
If you harvest under off throttle conditions, regardless if braking or just lift and coast, you gain overall system efficiency because otherwise the energy would be wasted either by the brakes or simply by aerodynamic drag.
If you use the ICE directly to shove some energy to the MGU-K you already start down at somewhere around 40-45% thermal efficency and it only gets worse from there. So you lose overall efficiency.

Also let us do here a short calculation. How much speed would a F1 at car at 300km/h and 800kg mass actually lose when the driver does lift and coast for 1 second and the MGU-K would start to harvest with 120kW?

$\frac{800*(300/3.6)^2}{2}=2.777MJ$

$2.777-(0.12*1)=2.657MJ$

$\sqrt{\frac{2.657*2}{800}}*3.6=293.45km/h$

So it would lose about 6.5km/h...that's not very massive. Also if a car behind would keep its speed of 300km/h during this 1 second it wouldn't even gain a full meter. But the car in front has won one second of MGU-K boost which it can use very efficently when it accelerates out of a slow corner.
Good post.

I’d take issue with the phrase “incredibly inefficient”.

As you point out later the goal is lap time efficiency. This involves shifting energy from places where it doesn’t do much good, such as end of straight as you point out, to the acceleration zone at the beginning of the straight.

@tommy cookers has pointed out a couple of times that the ICE efficiency at demand+120kW may be higher than demand by enough that the round trip electrical energy, K>ES>K is essentially free. A further gain is that running the ICE harder will increase H output.

All this is wholly dependant on the actual figures for outputs and efficiencies of the component parts of the PU. I don’t think we know these so we can only speculate on mechanisms that might be exploited until we have evidence, such as the article I quoted earlier in this thread, that confirm or add weight to, or otherwise, some of our hypotheses.
Fortune favours the prepared; she has no favourites and takes no sides.
Truth is confirmed by inspection and delay; falsehood by haste and uncertainty : Tacitus

saviour stivala
-13
Joined: Wed Apr 25, 2018 11:54 am

### Re: Can the MGU-K harvest or deploy under its own control? And related questions.

Incredible and a pity that nobody from the lot of the three power unit manufacturers including the research work Profs Limebeer did for FERRARI back than could hit-on or consider these here presented mathematical equations and their resultant gains by having the MGU-K harvesting when off throttle but still not when braking. If any of the above lot did, the old days difficulties of keeping a topped-up ES for lap after lap by capturing kinetic/heat that would otherwise goes to waste only. Would have been solved.