Honda Power Unit Hardware & Software

[Tommy Cookers]

... As far as I understand resonance requires external vibration frequency matching the resonant frequency of an object. So how can resonance be an issue if the vibrations in a PU constantly and very quickly change....

resonance is only a vibration response bigger than the forced vibration response
ie resonance can occur without even a complete forcing cycle
(consider a spring balance or NASA's Challenger nonsense and Airbus Queens NY crash lies)

a small amount of damping will greatly reduce the amplitude of resonance (and somewhat detune it)
high-tech composites have very little inherent damping


the resonator can vibrate at multiple frequencies (simultaneously?) eg crankshaft torsional (& bending?) vibrations

[Holm86]

I have a hard time believing Honda isn't capable of simulating drivetrain harmonics via software, and if the MGU-K and it's gearbox is the problem, wouldn't it have shown up on engine dynos? And couldn't they have run the tests with the MGU-K decoupled, just to get some more running time??

[Tommy Cookers]

With LS statement "like being electrocuted in a chair " with UK power supply @ 50Hz, then that would be wave form harmonic from MGU-K (given it's rotation at peak load or near to it) and possibly causing resonance/echo in the crankshaft
Very much lower you'd get "whitefinger" :D from something like a British twin motorcycle in 360 degree crank arrangement.

'white finger' is/was called Froelich's disease (Dr Frolich with an umlaut)
damaging to chainsaw users & also to men having a sitting contact with vibrating stuff

the MGU-K rotates at 1000 revs/sec ie 1 msec/rev
nominally there is no torque ripple (only the torque from individual phase windings varies with rotation)
torque is reviewed at about 1 kHz but the rate of change of torque is limited

yes the gearing to the crankshaft might have some non-ideal effect

[diffuser]

I have a hard time believing Honda isn't capable of simulating drivetrain harmonics via software, and if the MGU-K and it's gearbox is the problem, wouldn't it have shown up on engine dynos? And couldn't they have run the tests with the MGU-K decoupled, just to get some more running time??

Didn't they just say they spent the last 2 weeks working on the full car dyno to prevent the vibrations effecting the battery? Sounds like they can replicate it.

[Martin Keene]

Which begs the question, why didn’t they find it until the car went on circuit.

[diffuser]

Hm. Maybe Honda decided that the pistons should move in unison.

A truly remarkable press conference from Newey. My sense is that he's resistant to the idea that his chassis could be the cause of the vibrations and is putting the onus of fixing it squarely on Honda. I can't see how something in the way the engine interacts with the mounts or the transmission is not the primary cause of these issues which were clearly absent from the test bench. I wonder if the shorter battery lets them move the engine up closer to the driver and this is a contributor to the increased shaking. I'm not sure the blame for this is Honda's alone.

But where can I watch or read the whole thing?
How could be the car the source? It's the engine that vibrates. But why does it vibrate too much? Again I'm reminded of the 919 issues where they had to re-design it with a new crankshaft.

A 90-degree V6 is inherently more difficult to balance. For smooth operation, the engine requires evenly spaced combustion events every 120 degrees of crankshaft rotation. However, this cannot be achieved if opposing cylinders share the same crank journal. To correct this, the crankpins are split and offset, allowing the engine to simulate the firing geometry of a 60-degree V6. This produces evenly spaced 120° firing intervals and significantly improves balance and smoothness.

But Honda has done that before... I doudt this is it.

[Holm86]

Hm. Maybe Honda decided that the pistons should move in unison.

A truly remarkable press conference from Newey. My sense is that he's resistant to the idea that his chassis could be the cause of the vibrations and is putting the onus of fixing it squarely on Honda. I can't see how something in the way the engine interacts with the mounts or the transmission is not the primary cause of these issues which were clearly absent from the test bench. I wonder if the shorter battery lets them move the engine up closer to the driver and this is a contributor to the increased shaking. I'm not sure the blame for this is Honda's alone.

But where can I watch or read the whole thing?
How could be the car the source? It's the engine that vibrates. But why does it vibrate too much? Again I'm reminded of the 919 issues where they had to re-design it with a new crankshaft.

A 90-degree V6 is inherently more difficult to balance. For smooth operation, the engine requires evenly spaced combustion events every 120 degrees of crankshaft rotation. However, this cannot be achieved if opposing cylinders share the same crank journal. To correct this, the crankpins are split and offset, allowing the engine to simulate the firing geometry of a 60-degree V6. This produces evenly spaced 120° firing intervals and significantly improves balance and smoothness.

But Honda has done that before... I doudt this is it.

These engines are odd firing V6's and has been since 2014. Only one crankshaft design is legal, and that's 3 throws with 120° spacing. Offset crankpins are not allowed. So there's nothing to redesign

[diffuser]

Hm. Maybe Honda decided that the pistons should move in unison.


But where can I watch or read the whole thing?
How could be the car the source? It's the engine that vibrates. But why does it vibrate too much? Again I'm reminded of the 919 issues where they had to re-design it with a new crankshaft.

A 90-degree V6 is inherently more difficult to balance. For smooth operation, the engine requires evenly spaced combustion events every 120 degrees of crankshaft rotation. However, this cannot be achieved if opposing cylinders share the same crank journal. To correct this, the crankpins are split and offset, allowing the engine to simulate the firing geometry of a 60-degree V6. This produces evenly spaced 120° firing intervals and significantly improves balance and smoothness.

But Honda has done that before... I doudt this is it.

These engines are odd firing V6's and has been since 2014. Only one crankshaft design is legal, and that's 3 throws with 120° spacing. Offset crankpins are not allowed. So there's nothing to redesign

I don't see any crankpin phasing restrictions in the regs.

[diffuser]

A 90-degree V6 is inherently more difficult to balance. For smooth operation, the engine requires evenly spaced combustion events every 120 degrees of crankshaft rotation. However, this cannot be achieved if opposing cylinders share the same crank journal. To correct this, the crankpins are split and offset, allowing the engine to simulate the firing geometry of a 60-degree V6. This produces evenly spaced 120° firing intervals and significantly improves balance and smoothness.

But Honda has done that before... I doudt this is it.

These engines are odd firing V6's and has been since 2014. Only one crankshaft design is legal, and that's 3 throws with 120° spacing. Offset crankpins are not allowed. So there's nothing to redesign

I don't see any crankpin phasing restrictions in the regs.

I see it would weaken the crank to use splitpins.....Instead of using the V-angle to determine timing, they design the crank like this:
- 3 crankpins
- 120° separation
- cylinders from opposite banks share pins.

This gives perfect 120° firing without split pins despite the 90° block.

So the firing sequence works out evenly across the cycle.

0° cyl 1
120° cyl 4
240° cyl 2
360° cyl 5
480° cyl 3
600° cyl 6

?

[Vortex37]

the MGU-K rotates at 1000 revs/sec ie 1 msec/rev
nominally there is no torque ripple (only the torque from individual phase windings varies with rotation)
torque is reviewed at about 1 kHz but the rate of change of torque is limited

yes the gearing to the crankshaft might have some non-ideal effect

@TC

Looking at the picture of the 2026 ICE, I am assuming (?) that the MGU-K is the device next to the battery pack. You quite rightly mentioned torque ripple as a potential problem. This would certainly vibrate the battery cells. A clue comes from Ant Davidson(SKY) at the side of the track when Stroll had that weird lock-up and spin in testing. He said that the engine seemed to stop completely and restart. A de-sync between K and ICE? K stalling during harvesting when he braked? If the MGU-K "stalls" or shuts down during braking, the car loses its regenerative braking torque instantly. This is like someone suddenly cutting the brake lines to the rear wheels while you're at the limit. The Brake-By-Wire system can't compensate fast enough for a total electrical blackout, leading to the rear wheels either locking up or losing all "engine braking" stability—sending the car into the gravel. I would also look at the possibility of a fundamental problem in the K itself. They are probably using 3D printed coils and stators. If coil switching in either harvest or motor mode is ‘off’, then horrific vibrations/shocks would result. If the 3D-printed stator geometry hasn't accounted for these specific harmonic resonances, the motor could act like a tuning fork, vibrating until something, maybe the battery or the inverter either breaks or more likely goes into a safe mode. What is the energy store software programmed to do with out of phase pulses? It would be interesting to know what battery chemistry they are using. The vibrating battery cells problem is an easy fix.

[Tommy Cookers]

it's a bit of a late hour to have any ideas but fwiw ....

I've often said that also ....
successful occasional departures from the synchronous might amount to discreet ABS-like and TC-like behaviours ...
hard to police - all a matter between the so-called CU and the MG
and anyway the teams need some freedom to make things look good

[mzso]

Point of order on symmetry.
Gears don't need to be circular, they can be oval, still mesh and still work off the same sorts of shafts circular gears work off.
In such an arrangement, torque/speed would look like a waveform instead of a straight line on a graph, and this could be used for several reasons.
For example to overcome certain types of pulsed input from things like electric motors, or to allow breakthrough torque where a linear approach is insufficient.

TL;DR don't assume symmetry just because it's normal.

Okay. Can you refer to an actual automotive application of such wacky gears?

[mzso]

resonance is only a vibration response bigger than the forced vibration response
ie resonance can occur without even a complete forcing cycle
(consider a spring balance or NASA's Challenger nonsense and Airbus Queens NY crash lies)

a small amount of damping will greatly reduce the amplitude of resonance (and somewhat detune it)
high-tech composites have very little inherent damping


the resonator can vibrate at multiple frequencies (simultaneously?) eg crankshaft torsional (& bending?) vibrations

Why add wacky conspiracy theories? The airbus lost its rudder because the pilot kept stomping the pedals and over-stressed it. The challenger got destroyed because of flawed sealings that got rigid in the cold.

[mzso]

A 90-degree V6 is inherently more difficult to balance. For smooth operation, the engine requires evenly spaced combustion events every 120 degrees of crankshaft rotation. However, this cannot be achieved if opposing cylinders share the same crank journal. To correct this, the crankpins are split and offset, allowing the engine to simulate the firing geometry of a 60-degree V6. This produces evenly spaced 120° firing intervals and significantly improves balance and smoothness.

But Honda has done that before... I doudt this is it.

Not that I have the skill for it, but it should be possible to tell how the engine fires by the sound of it, shouldn't it?

[gearboxtrouble]

I very much doubt we're looking at traditional ICE rotational balancing issues here. Nobody would get that wrong and the rules are fairly restrictive in that regard. I expect the problem is something to do with either the mounts themselves resonating to what would other wise be normal levels of engine vibrations or something driven by the MGU-K itself under harvesting from the transmission (because it didn't happen on the bench). Neither would be a quick or easy fix.