Honda Power Unit Hardware & Software

[hollus]

I’ll leave the posts stay, as they include some arguments, but pretending to own the truth does not a forum make.
Please remain respectful as well.
Playing pigeon chess will not (is not) acceptable.

Ghost, just above, nailed it.

[hkbruin]

Here is a recent article/interview of HRC President Koji Watanabe on 06.14.2026 from Japan’s Autosport Web:
https://www.as-web.jp/f1/1325833

It will require translating to English but in a nutshell, Watanabe details their current situation at AMR/Honda and goes into why they are on the backfoot. The lack of continuity back in 2022-23 at Sakura when Honda decided to “return” to F1 showed how they essentially performed a hard reset with their research and development.

A small excerpt:

President Watanabe: "It is true that the new regulation was very complicated and difficult, but at the time when the new regulation was introduced, we teamed up with a new team and new fuel and I think it was a pretty high hurdle to return to F1 with new oil. In March 2022, when the 4th term of F1 activities ended and the power unit that was homorized to Red Bull and Alpha Tauri was handed over, HRC engineers from HRC Sakura to Honda Motor Corporation and Honda Technology Research Institute at their headquarters. I went back to the research institute. From there, for about a year until May 2023, when it announced that it would conclude a partnership with Aston Martin, Sakura continued only basic research and element research with a very small number of people. The delay between them was quite big. So I think it will take a little more time to reorganize the system.”

Apologies for any poor web translations.

[PlatinumZealot]

Power Unit driveability problems explained by Honda Cheif Engineer

Grabbed from the webs. It pretty much is what I expected. They just gotta keep working on the combustion.

Orihara:

"One of the challenges is preventing the rear wheels from locking up when the driver downshifts while braking. This requires control from the transmission, and the power unit must correctly deliver the necessary torque so that the engine speed matches the downshift and prevents the wheels from locking up."

"The difficulty lies in the fact that the 2026 model’s power unit will use a method of energy generation called partial regeneration more frequently than before.

"When entering a turn, even with the throttle off, the engine generates nearly maximum power (due to braking from the MGU-K). If the engine torque fluctuates even slightly, the driver will notice the rear end drifting inward during the turn. In Monaco, you have no choice but to drive with a safety margin, so the key is how to stabilize the car and enter the turn with the same engine torque—and that’s what we’re trying to achieve."

"This partial regeneration also occurs during acceleration. The internal combustion engine must generate stable torque to counteract the braking force of the MGU-K; otherwise, the driver cannot safely press the accelerator. The main cause of this lies in the internal combustion engine’s lack of power and its still-immature combustion settings."

"Unstable torque means unstable combustion. The problem was that combustion in our internal combustion engine was unstable, which resulted in inconsistent torque every time. Up until the Miami Grand Prix, we were still looking for a way to fix it, but after the Miami Grand Prix, we found a solution that would improve things, and we’ve been gradually fine-tuning in that direction, so the number of issues we need to resolve has decreased. It has decreased, but we haven’t been able to eliminate them completely yet, and the big difference is that, while staying within this range is acceptable at other circuits, Monaco requires a much higher standard (which has made the problems more evident)."

[mzso]

Here is a recent article/interview of HRC President Koji Watanabe on 06.14.2026 from Japan’s Autosport Web:
https://www.as-web.jp/f1/1325833

That verifies a thing or two. Particularly that they're not jet ready with the new spec ICE, he said "around the summer" according to the translation. So I guess barring further delays it will be August.

He was very vague about the engine and chassis being introduced together. So either he doesn't want to say that the new chassis is only for the new engine, or it doesn't matter.

[mzso]

"This partial regeneration also occurs during acceleration. The internal combustion engine must generate stable torque to counteract the braking force of the MGU-K; otherwise, the driver cannot safely press the accelerator. The main cause of this lies in the internal combustion engine’s lack of power and its still-immature combustion settings."

This part - regenerating during acceleration - doesn't make sense.

[Farnborough]

"This partial regeneration also occurs during acceleration. The internal combustion engine must generate stable torque to counteract the braking force of the MGU-K; otherwise, the driver cannot safely press the accelerator. The main cause of this lies in the internal combustion engine’s lack of power and its still-immature combustion settings."

This part - regenerating during acceleration - doesn't make sense.

Yes it does.

If the driver request with throttle is LESS than peak ICE output, then the torque demand can be delivered (as requested by driver) to rear wheels by the ecu control opening throttling further than needed, then torque "held" at the demand by ramping up regeneration with MGUK "request" the two (front and rear of crankshaft demand) blended to give invisible torque delivery "flow" ..... if they get the software right ..... in driving the rear axle.

Obviously not possible if deployment of ICE & E are called for simultaneously.

They've not got this right, obviously, from their published statement included on forum.

[gearboxtrouble]

"This partial regeneration also occurs during acceleration. The internal combustion engine must generate stable torque to counteract the braking force of the MGU-K; otherwise, the driver cannot safely press the accelerator. The main cause of this lies in the internal combustion engine’s lack of power and its still-immature combustion settings."

This part - regenerating during acceleration - doesn't make sense.

I did theorize this pre season. Charging the MGU-K during the very early phases of acceleration could act as a form of traction control. It allows the drivers to get on throttle early and the excess torque gets canceled out by the negative torque from the MGUK as it charges. Once the driver goes 100% throttle the MGUK would switch to deployment. It would help with corner rotation, traction management and harvesting all at once.

[mzso]

One would think in Monaco, particularly with the power limit there is certainly no need for part-throttle recovery.
Also it's my impression of other engine manufacturers, that they mostly rely on (super)clipping, rather than part-throttle harvesting for the best overall performance.

If the driver request with throttle is LESS than peak ICE output, then the torque demand can be delivered (as requested by driver) to rear wheels by the ecu control opening throttling further than needed, then torque "held" at the demand by ramping up regeneration with MGUK "request" the two (front and rear of crankshaft demand) blended to give invisible torque delivery "flow" ..... if they get the software right ..... in driving the rear axle.

Maybe they're the only one doing this because they're grubbing for every bit of energy, because they otherwise don't have enough because of lack of ICE power.

[Tommy Cookers]

I did theorize this pre season. Charging the MGU-K during the very early phases of acceleration could act as a form of traction control. It allows the drivers to get on throttle early and the excess torque gets canceled out by the negative torque from the MGUK as it charges. Once the driver goes 100% throttle the MGUK would switch to deployment. It would help with corner rotation, traction management and harvesting all at once.

notionally any time the ICE max torque is enough to spin the wheels generating is useful

but due to synchronicity eg generating torque can't necessarily be maintained at high rotational accelerations
(remember that the PU is rather close to the synchronicity danger zone at peak torque)
smell of driver aid ?

at other times also the 2026 PU won't behave as 2014-2025 PU torque mapping rules implied that 2014-2025 PUs did

[karana]

One would think in Monaco, particularly with the power limit there is certainly no need for part-throttle recovery.
Also it's my impression of other engine manufacturers, that they mostly rely on (super)clipping, rather than part-throttle harvesting for the best overall performance.

If the driver request with throttle is LESS than peak ICE output, then the torque demand can be delivered (as requested by driver) to rear wheels by the ecu control opening throttling further than needed, then torque "held" at the demand by ramping up regeneration with MGUK "request" the two (front and rear of crankshaft demand) blended to give invisible torque delivery "flow" ..... if they get the software right ..... in driving the rear axle.

Maybe they're the only one doing this because they're grubbing for every bit of energy, because they otherwise don't have enough because of lack of ICE power.

Partial load recovery is definitely preferable to superclipping. The latter actually slows the cars down while it's happening, the time gain in other parts of the track just outweigh the losses. The only downside of partial throttle recovery is increased fuel consumption (and apparently driveability at least for Honda).

This is also not new, Honda was already doing this in 2016. If you click on "read sample" in the following link you can see towards the end the telemetry of a q and a race lap at Austria from 2016 where you can see the MGU-K power being negative under partial throttle: https://www.amazon.co.jp/%E3%83%A2%E3%8 ... 2Y6RV63H2C

"This partial regeneration also occurs during acceleration. The internal combustion engine must generate stable torque to counteract the braking force of the MGU-K; otherwise, the driver cannot safely press the accelerator. The main cause of this lies in the internal combustion engine’s lack of power and its still-immature combustion settings."

This part - regenerating during acceleration - doesn't make sense.

I did theorize this pre season. Charging the MGU-K during the very early phases of acceleration could act as a form of traction control. It allows the drivers to get on throttle early and the excess torque gets canceled out by the negative torque from the MGUK as it charges. Once the driver goes 100% throttle the MGUK would switch to deployment. It would help with corner rotation, traction management and harvesting all at once.

The PU output must match the driver demand, so what you are proposing is not allowed, at least as a form of traction control.

[Tommy Cookers]

The PU output must match the driver demand .....

the FIA implies that it will always follow the functionality implemented in the ECU programme
but it won't eg rates of change of rpm from rear wheel locking or spinning will exceed rates possible by PU action

the FIA doesn't look into this because it can't - and can't look because it doesn't
that's for the best

[diffuser]

"This partial regeneration also occurs during acceleration. The internal combustion engine must generate stable torque to counteract the braking force of the MGU-K; otherwise, the driver cannot safely press the accelerator. The main cause of this lies in the internal combustion engine’s lack of power and its still-immature combustion settings."

This part - regenerating during acceleration - doesn't make sense.

They're allowed to regenerate under acceleration when throttle demanded is below max ICE output or something like that. I remember reading about a very complicated algorithm in these pages in preseason. That was when they could only recharge to a max rate of 250KW.

[diffuser]

One would think in Monaco, particularly with the power limit there is certainly no need for part-throttle recovery.
Also it's my impression of other engine manufacturers, that they mostly rely on (super)clipping, rather than part-throttle harvesting for the best overall performance.

If the driver request with throttle is LESS than peak ICE output, then the torque demand can be delivered (as requested by driver) to rear wheels by the ecu control opening throttling further than needed, then torque "held" at the demand by ramping up regeneration with MGUK "request" the two (front and rear of crankshaft demand) blended to give invisible torque delivery "flow" ..... if they get the software right ..... in driving the rear axle.

Maybe they're the only one doing this because they're grubbing for every bit of energy, because they otherwise don't have enough because of lack of ICE power.

Partial load recovery is definitely preferable to superclipping. The latter actually slows the cars down while it's happening, the time gain in other parts of the track just outweigh the losses. The only downside of partial throttle recovery is increased fuel consumption (and apparently driveability at least for Honda).

This is also not new, Honda was already doing this in 2016. If you click on "read sample" in the following link you can see towards the end the telemetry of a q and a race lap at Austria from 2016 where you can see the MGU-K power being negative under partial throttle: https://www.amazon.co.jp/%E3%83%A2%E3%8 ... 2Y6RV63H2C

This part - regenerating during acceleration - doesn't make sense.

I did theorize this pre season. Charging the MGU-K during the very early phases of acceleration could act as a form of traction control. It allows the drivers to get on throttle early and the excess torque gets canceled out by the negative torque from the MGUK as it charges. Once the driver goes 100% throttle the MGUK would switch to deployment. It would help with corner rotation, traction management and harvesting all at once.

The PU output must match the driver demand, so what you are proposing is not allowed, at least as a form of traction control.

For arguments sake let's say the ICE creates 50% of the power and the MGU-K, the other 50%. At 25% throttle the ICE is running at 100%, the MGU-K at -75% leaving 25% to drive the wheels. The PU matches the throttle.

https://www.gpfans.com/en/f1-news/10770 ... hatgpt.com

Recharge mode is when a driver refills the battery on their new cars using recovered energy, using a range of strategies to capture the energy and send it back to the battery, something previously referred to as 'harvesting'.This can be done a number of ways, including through braking, when partly on the throttle, when lifting and coasting or when a driver is super clipping, although the effectiveness of each strategy will depend on the engine map, the circuit profile, and the overall Recharge energy allowance for that circuit, which will vary from track to track.Most of the time, the Recharge will be automated by use of selectable Recharge maps and targets, which mean braking, part throttle and super clipping will all be automated functions controlled by the engine control unit (ECU).The only Recharge mode the driver will have control of personally will be the lift-off regen (just another term for lift and coast) which is when a driver lifts off the accelerator early before a braking zone, allowing their car to coast, something which allows the MGU-K to harvest the energy and recharge the battery.