Mercedes Power Unit Hardware & Software

[SiLo]

All the result of the lack of testing now I guess. Mercedes seem to be one of the few truly using the season as a testing platform alongside trying to win races and titles. Ferrari have been doing it this year as well which is refreshing to see from them.

Always playing the long game.

[KAIZEN]

The power was always there, but they never used it until now .....

This is my suspicion as well. They had this headroom from start of the season, but didn't know how far they can push it. Now we are gradually seeing more and more power being released, probably using bottas' car as a test bed, that's why he's having more reliability problems.

I suspect you're bang on, though I would speculate that the testing started a lot earlier in the season.

Austria.

Yes Merc ran a yacht sail as a rear wing, however this alone does not account for their lack of pace. I believe that Merc went into the double-header at Austria knowing they had reliability issues and were conservative with their power units.

Russell's power unit was run into the ground during the Styrian GP in an effort to assess where exactly the reliability issues were. Note the components that were replaced for the Austrian GP:

His power unit was returned to HPP in Brixworth, and after investigations it has been deemed that it suffered a gear drive failure. Russell's car will thus have a new ICE, MGU-H and turbo when Williams reserve driver Jack Aitken takes to the track in FP1 on Friday for the Styrian Grand Prix.

https://www.autosport.com/f1/news/russe ... 2/4981342/

The ICE, MGU-H and turbo are swapped out as a single piece and we now know that the issues are within the ICE.

The power unit introduced in Monza for Bottas would have come with reliability upgrades, the ICE was once again run hard to get a real world confirmation that the reliability upgrades were sufficient. These components were then retired and sent to Brixworth for analysis and Bottas got new components in Sochi, with Lewis only replacing the ICE in Turkey.

That's the conclusion that I come to when joining the dots.

This year is 2021.
The autosport quoted article is about 2020.

[TimW]

So the impression now is that Mercedes is running their power unit at a higher output at the cost of reliability. Since upgrades for reliability are allowed, and there is often a tradeoff between power and reliability, does that mean that the engine freeze does not have too much effect in reality?

[mstar]

I think Horner, implied that in the past Merc didnt need to run the engines hard through the weekend and race as they had a car advantage. So the engine issues is related to they now have to run the engines on max for a lot of races which is causing little issues arise which they never saw before.

Honda, with their new design i think mentioned they can run on higher power, longer so they dont have the same kind of issue.

[b2bL44]

This is my suspicion as well. They had this headroom from start of the season, but didn't know how far they can push it. Now we are gradually seeing more and more power being released, probably using bottas' car as a test bed, that's why he's having more reliability problems.

I suspect you're bang on, though I would speculate that the testing started a lot earlier in the season.

Austria.

Yes Merc ran a yacht sail as a rear wing, however this alone does not account for their lack of pace. I believe that Merc went into the double-header at Austria knowing they had reliability issues and were conservative with their power units.

Russell's power unit was run into the ground during the Styrian GP in an effort to assess where exactly the reliability issues were. Note the components that were replaced for the Austrian GP:

His power unit was returned to HPP in Brixworth, and after investigations it has been deemed that it suffered a gear drive failure. Russell's car will thus have a new ICE, MGU-H and turbo when Williams reserve driver Jack Aitken takes to the track in FP1 on Friday for the Styrian Grand Prix.

https://www.autosport.com/f1/news/russe ... 2/4981342/

The ICE, MGU-H and turbo are swapped out as a single piece and we now know that the issues are within the ICE.

The power unit introduced in Monza for Bottas would have come with reliability upgrades, the ICE was once again run hard to get a real world confirmation that the reliability upgrades were sufficient. These components were then retired and sent to Brixworth for analysis and Bottas got new components in Sochi, with Lewis only replacing the ICE in Turkey.

That's the conclusion that I come to when joining the dots.

This year is 2021.
The autosport quoted article is about 2020.

My apologies, deja vu!

Here's the correct article for 2021:

https://au.motorsport.com/f1/news/russe ... t/6615161/

Official FIA documentation for Russell for 3rd ICE, TC and MGU-H:

https://www.fia.com/sites/default/files ... 0Event.pdf

[velizare]

So the impression now is that Mercedes is running their power unit at a higher output at the cost of reliability. Since upgrades for reliability are allowed, and there is often a tradeoff between power and reliability, does that mean that the engine freeze does not have too much effect in reality?

in contrary. as journalist were convined honda provided reliability upgrades for the second engine this year, tanabe had to explain how this thing really works:

Then under the current PU regulations, we need to submit any changes, so only allowed to change for reliability, for cost reasons and logistics. Then we need to submit very detailed to the FIA first and the FIA approves those changes.

FIA distributes all the documents to the other PU manufacturers, so we need to have an approval from the other PU manufacturers to change any single parts, specification. Then, why we are doing such a very much detailed investigation is a long time ago some teams improved their performance to make a change, to improve their reliability, so we are very careful to change the performance and then it is not possible to improve the performance during the season. That’s my answer to that suspicion.

basically you'd need to share your technology with the others. no one is doing such things.

[El Scorchio]

So the impression now is that Mercedes is running their power unit at a higher output at the cost of reliability. Since upgrades for reliability are allowed, and there is often a tradeoff between power and reliability, does that mean that the engine freeze does not have too much effect in reality?

in contrary. as journalist were convined honda provided reliability upgrades for the second engine this year, tanabe had to explain how this thing really works:

Then under the current PU regulations, we need to submit any changes, so only allowed to change for reliability, for cost reasons and logistics. Then we need to submit very detailed to the FIA first and the FIA approves those changes.

FIA distributes all the documents to the other PU manufacturers, so we need to have an approval from the other PU manufacturers to change any single parts, specification. Then, why we are doing such a very much detailed investigation is a long time ago some teams improved their performance to make a change, to improve their reliability, so we are very careful to change the performance and then it is not possible to improve the performance during the season. That’s my answer to that suspicion.

basically you'd need to share your technology with the others. no one is doing such things.

Would be good to tell this to all the conspiracy theorists on here!

[KAIZEN]

According to Adauto Silva, the Mercedes ICE is an enhanced version with a 630g increase.

https://www.autoracing.com.br/exclusivo ... -mercedes/

[b2bL44]

According to Adauto Silva, the Mercedes ICE is an enhanced version with a 630g increase.

https://www.autoracing.com.br/exclusivo ... -mercedes/

Thank you for this article. How reliable is Adauto Silva for this information?

The information he provides here confirms my speculation above.

Translation:

630 grams. That could be a magic weight for Mercedes' internal combustion engine.

Explain. Since the British GP at Silverstone the Silver Ones have revised the mappings that were already ready for each track until the end of the season.

And because of the great performance of Red Bull and Honda, Mercedes have reworked their engine mappings making them much more aggressive. The PUs mappings introduced from 2014 are incomparably more complex than before, as the internal combustion engine has to work in perfect harmony not just with the turbo, but with the MGU-K, MGU-H, which in turn send electrical power to the batteries which, via the electronic control unit, returns it all in the form of mapping-controlled power to the car.

All this has to work like a link in a continuous chain and is done through extremely sophisticated software. When one link of this "chain" presents some problem, the whole chain is harmed by this weaker "link".

In the case of PU Mercedes, since the mappings were changed, the internal combustion engine became the weakest link, since its temperature and pressure increased and this started to lead to micro-leaks that increased with time. Each time in a different place of the ICE (internal combustion engine).

This caused all Mercedes drivers that were using the new mappings to have problems in their PUs to the point of having to change it ahead of schedule, besides hardly feeling the effect that the more aggressive mappings were supposed to do.

So, since the micro leaks were appearing in different places on the ICE, Mercedes decided to change all the gaskets, bolts, nuts, clamps, pipes, hoses and everything that connects the different parts of the ICE for others that were more reinforced and/or with different materials. The ICE became 630 grams heavier, an irrelevant weight increase if the problem was solved.

Valtteri Bottas was the "guinea pig" of this attempt to fix the problem in Russia. Mercedes didn't change his PU there because of race strategy, but rather an engine strategy, to see if what they did would work in practice. Toto Wolff even said that in a "cryptic" way that weekend...

Right after the Russian GP, 11 engineers from the Brixworth factory, which is where Mercedes AMG High Performance Powertrains are made, went with some equipment directly to the team's factory in Brackley to test and measure all parameters of the PU Bottas used in free practice for the Russian GP and the one he used on race Sunday, 630 grams heavier.

On the PU that Bottas used in the Russia race, no micro leaks were found and all measurements were within the standard they consider correct.

That's why Mercedes decided to change only Hamilton's ICE in Turkey and not the whole PU.

Tomorrow, Tuesday, the same engineers who thoroughly inspected Bottas' PU last week will be back at Brackley to start inspecting Bottas' PU again - now with two runs - and the one Hamilton used in Turkey.

If everything is up to standard, the team will continue using the revised mappings, otherwise they will revert to the 'original' mappings, which are far less aggressive.

[KAIZEN]

According to Adauto Silva, the Mercedes ICE is an enhanced version with a 630g increase.

https://www.autoracing.com.br/exclusivo ... -mercedes/

Thank you for this article. How reliable is Adauto Silva for this information?

The information he provides here confirms my speculation above.

Translation:

630 grams. That could be a magic weight for Mercedes' internal combustion engine.

Explain. Since the British GP at Silverstone the Silver Ones have revised the mappings that were already ready for each track until the end of the season.

And because of the great performance of Red Bull and Honda, Mercedes have reworked their engine mappings making them much more aggressive. The PUs mappings introduced from 2014 are incomparably more complex than before, as the internal combustion engine has to work in perfect harmony not just with the turbo, but with the MGU-K, MGU-H, which in turn send electrical power to the batteries which, via the electronic control unit, returns it all in the form of mapping-controlled power to the car.

All this has to work like a link in a continuous chain and is done through extremely sophisticated software. When one link of this "chain" presents some problem, the whole chain is harmed by this weaker "link".

In the case of PU Mercedes, since the mappings were changed, the internal combustion engine became the weakest link, since its temperature and pressure increased and this started to lead to micro-leaks that increased with time. Each time in a different place of the ICE (internal combustion engine).

This caused all Mercedes drivers that were using the new mappings to have problems in their PUs to the point of having to change it ahead of schedule, besides hardly feeling the effect that the more aggressive mappings were supposed to do.

So, since the micro leaks were appearing in different places on the ICE, Mercedes decided to change all the gaskets, bolts, nuts, clamps, pipes, hoses and everything that connects the different parts of the ICE for others that were more reinforced and/or with different materials. The ICE became 630 grams heavier, an irrelevant weight increase if the problem was solved.

Valtteri Bottas was the "guinea pig" of this attempt to fix the problem in Russia. Mercedes didn't change his PU there because of race strategy, but rather an engine strategy, to see if what they did would work in practice. Toto Wolff even said that in a "cryptic" way that weekend...

Right after the Russian GP, 11 engineers from the Brixworth factory, which is where Mercedes AMG High Performance Powertrains are made, went with some equipment directly to the team's factory in Brackley to test and measure all parameters of the PU Bottas used in free practice for the Russian GP and the one he used on race Sunday, 630 grams heavier.

On the PU that Bottas used in the Russia race, no micro leaks were found and all measurements were within the standard they consider correct.

That's why Mercedes decided to change only Hamilton's ICE in Turkey and not the whole PU.

Tomorrow, Tuesday, the same engineers who thoroughly inspected Bottas' PU last week will be back at Brackley to start inspecting Bottas' PU again - now with two runs - and the one Hamilton used in Turkey.

If everything is up to standard, the team will continue using the revised mappings, otherwise they will revert to the 'original' mappings, which are far less aggressive.

Look at the Adauto Silva column to determine reliability.
There was interaction with Mercedes engineers and Red Bull engineers, and the content obtained from chatting with them is very interesting.

[SiLo]

630g is quite a lot in the F1 world. Although I assume the cars are far underweight anyway and they run a decent amount of ballast. At least an increased to the weight of the ICE is relatively central to the entire car.

[Stu]

Are heat pipes allowed?? Whis is basically two phase cooling. Low boiling point water in the plenum walls, gets heated by intake, the vapour moves to a heat sink where it is condensed and water returns to the chamber walls again.

Pretty sure two phase cooling is not allowed.

You are aright. The use of latent heat is banned, even if vapour compression is not used.

7.5 Cooling systems
The cooling systems of the power unit, including that of the air destined for combustion,
must not intentionally make use of the latent heat of vaporisation of any fluid with the
exception of fuel for the normal purpose of combustion in the engine as described in Article
5.10.3.

The Adiabatic expansion you mention seems to be allowed. Since it doesn't use any significant latent heat.

Sub-cooling of the charge could be done using the Porsche method which is essentially a vapour compression technique - boost higher than you need - intercool to near ambient temp - expand to the desired boost pressure and enjoy the associated temperature reduction. Similar to air cycle refrigeration used in some aircraft - with the exception that the expansion is done in a turbine for even lower temperatures and some energy recovery.

Yes, that method is a good one also agree that all the teams are probably doing it. I suppose a large throttle could also be used to expand the air if they don't want to fiddle around with acoustics too much like Porsche (imagine the maps for the VLIM - could crazy), I think some balance must be struck between the compressor work, and inter-cooler size (water cooling really helps here in qualifying) etc.

I’ve been giving this some thought….
Considering the radical change in apparent volume of the old and new plenum chambers and the reported ‘out-of-corner-power-boost’ that has been reported; is it possible (or probable!) that they have, in effect, two pressure chambers within the ‘apparent’ plenum? If a second chamber held air at maximum boost pressure, but had a flexible wall, it could be used to influence the volume of the primary chamber during periods of low boost and increase the velocity of the inlet charge. At higher boost levels the volume in the primary chamber would increase as the relative pressure between the two chambers neared equilibrium. The whole thing could be controlled by a simple ‘sluice-gate-style’ valve between the two chambers. At, or nearing, maximum boost the two chambers would become a single large chamber.

Lateral thinking…?

[hurril]

Pretty sure two phase cooling is not allowed.

You are aright. The use of latent heat is banned, even if vapour compression is not used.

7.5 Cooling systems
The cooling systems of the power unit, including that of the air destined for combustion,
must not intentionally make use of the latent heat of vaporisation of any fluid with the
exception of fuel for the normal purpose of combustion in the engine as described in Article
5.10.3.

The Adiabatic expansion you mention seems to be allowed. Since it doesn't use any significant latent heat.

Sub-cooling of the charge could be done using the Porsche method which is essentially a vapour compression technique - boost higher than you need - intercool to near ambient temp - expand to the desired boost pressure and enjoy the associated temperature reduction. Similar to air cycle refrigeration used in some aircraft - with the exception that the expansion is done in a turbine for even lower temperatures and some energy recovery.

Yes, that method is a good one also agree that all the teams are probably doing it. I suppose a large throttle could also be used to expand the air if they don't want to fiddle around with acoustics too much like Porsche (imagine the maps for the VLIM - could crazy), I think some balance must be struck between the compressor work, and inter-cooler size (water cooling really helps here in qualifying) etc.

I’ve been giving this some thought….
Considering the radical change in apparent volume of the old and new plenum chambers and the reported ‘out-of-corner-power-boost’ that has been reported; is it possible (or probable!) that they have, in effect, two pressure chambers within the ‘apparent’ plenum? If a second chamber held air at maximum boost pressure, but had a flexible wall, it could be used to influence the volume of the primary chamber during periods of low boost and increase the velocity of the inlet charge. At higher boost levels the volume in the primary chamber would increase as the relative pressure between the two chambers neared equilibrium. The whole thing could be controlled by a simple ‘sluice-gate-style’ valve between the two chambers. At, or nearing, maximum boost the two chambers would become a single large chamber.

Lateral thinking…?

I don't think that could ever be enough air. Otoh, what about having a Koenigsegg-style compressed air-spool capability of the turbo driven this way?

Mercedes are also known for being quite deceptive so it might just as well be a different MGU-k setup that is much more capable in some way and we're all looking at their proverbial humps.

[PlatinumZealot]

Pretty sure two phase cooling is not allowed.

You are aright. The use of latent heat is banned, even if vapour compression is not used.

7.5 Cooling systems
The cooling systems of the power unit, including that of the air destined for combustion,
must not intentionally make use of the latent heat of vaporisation of any fluid with the
exception of fuel for the normal purpose of combustion in the engine as described in Article
5.10.3.

The Adiabatic expansion you mention seems to be allowed. Since it doesn't use any significant latent heat.

Sub-cooling of the charge could be done using the Porsche method which is essentially a vapour compression technique - boost higher than you need - intercool to near ambient temp - expand to the desired boost pressure and enjoy the associated temperature reduction. Similar to air cycle refrigeration used in some aircraft - with the exception that the expansion is done in a turbine for even lower temperatures and some energy recovery.

Yes, that method is a good one also agree that all the teams are probably doing it. I suppose a large throttle could also be used to expand the air if they don't want to fiddle around with acoustics too much like Porsche (imagine the maps for the VLIM - could crazy), I think some balance must be struck between the compressor work, and inter-cooler size (water cooling really helps here in qualifying) etc.

I’ve been giving this some thought….
Considering the radical change in apparent volume of the old and new plenum chambers and the reported ‘out-of-corner-power-boost’ that has been reported; is it possible (or probable!) that they have, in effect, two pressure chambers within the ‘apparent’ plenum? If a second chamber held air at maximum boost pressure, but had a flexible wall, it could be used to influence the volume of the primary chamber during periods of low boost and increase the velocity of the inlet charge. At higher boost levels the volume in the primary chamber would increase as the relative pressure between the two chambers neared equilibrium. The whole thing could be controlled by a simple ‘sluice-gate-style’ valve between the two chambers. At, or nearing, maximum boost the two chambers would become a single large chamber.

Lateral thinking…?

You mean basically a compressed air receiver that has a controlled release for when boost is needed. End result being it saves a bit more battery power by not having to maintain spool.

Two problems. One. You need a check valve to maintain this pressure or if not you still have to drive the compressor at this pressure all the time. So it cancels out the plan to save energy.

The other problem more valves and chockes adds pressure loss/ more work for compressor over the lap.

Come to think of it. I see why Porsche does this accoustically. To prevent complication and energy loss.

So i guess the theory of the the bigger chamber throttling into a smaller one is too expensive energy-wise. Precious battery charge is needed for the MGUK on the straights in qualifying.

[Stu]

You are aright. The use of latent heat is banned, even if vapour compression is not used.



The Adiabatic expansion you mention seems to be allowed. Since it doesn't use any significant latent heat.



Yes, that method is a good one also agree that all the teams are probably doing it. I suppose a large throttle could also be used to expand the air if they don't want to fiddle around with acoustics too much like Porsche (imagine the maps for the VLIM - could crazy), I think some balance must be struck between the compressor work, and inter-cooler size (water cooling really helps here in qualifying) etc.

I’ve been giving this some thought….
Considering the radical change in apparent volume of the old and new plenum chambers and the reported ‘out-of-corner-power-boost’ that has been reported; is it possible (or probable!) that they have, in effect, two pressure chambers within the ‘apparent’ plenum? If a second chamber held air at maximum boost pressure, but had a flexible wall, it could be used to influence the volume of the primary chamber during periods of low boost and increase the velocity of the inlet charge. At higher boost levels the volume in the primary chamber would increase as the relative pressure between the two chambers neared equilibrium. The whole thing could be controlled by a simple ‘sluice-gate-style’ valve between the two chambers. At, or nearing, maximum boost the two chambers would become a single large chamber.

Lateral thinking…?

You mean basically a compressed air receiver that has a controlled release for when boost is needed. End result being it saves a bit more battery power by not having to maintain spool.

Two problems. One. You need a check valve to maintain this pressure or if not you still have to drive the compressor at this pressure all the time. So it cancels out the plan to save energy.

The other problem more valves and chockes adds pressure loss/ more work for compressor over the lap.

Come to think of it. I see why Porsche does this accoustically. To prevent complication and energy loss.

So i guess the theory of the the bigger chamber throttling into a smaller one is too expensive energy-wise. Precious battery charge is needed for the MGUK on the straights in qualifying.

Not exactly that; if the inside of the plenum (which is much larger than previously) is split into two compartments with a flexible wall between them and a (probably hydraulic) control valve between the two chambers you could create a variable volume plenum chamber. When the engine is initially started the valve would be open and both ‘sides’ of the chamber see the same pressure, once the engine has been run up to full boost the valve will close, holding air at full boost inside it and expanding the flexible wall of the chamber into the remaining plenum space when boost pressure drops (under braking/zero throttle conditions maybe). Once the throttle is re-applied at corner exit boost can be built quickly because the plenum has a smaller volume so the engine ‘sees’ higher velocity air on valve opening than would be the case with a larger plenum. The chamber would then expand as revs and boost build until the pressures equalise in the two chambers, but the engine would effectively have a larger plenum than at low boost (more air at boost pressure in the plenum should give more power if the fuel is available) when the revs are higher. On stopping the engine the hydraulic valve could be set to re-open and discharge the air inside back to atmospheric.