Mercedes Power Unit Hardware & Software

[MrPotatoHead]

re "hole in plenum entry". The pipe feeding the plenum would be tapered, the idea being to gradually reduce the velocity as the air approaches the plenum. Ideal condition in the plenum itself is zero velocity and the transition from very high velocity at the compressor wheel tips to near-zero in the plenum should be as gradual as possible.

I can show you plenty of CFD that will show you near zero velocity in the plenum is nearly impossible to achieve.
Especially on a forced induction application.

But yes the Variable Length hardware is inside the plenum itself.
That's how it works - you vary the total intake runner length.

[63l8qrrfy6]

I don't think you would get any helmholtz resonance from a inlet tube within a tube. There will not be enough change in volume at the mouth of the runner.

You don't need a large change. Have you seen how small the step is on an exhaust manifold ?

There doesn't seem to be any visible mechanism for operating variable length runners.

Look again at the pics hurrill has posted - there's a big ass rail running along the crank centreline with little wings extending on both sides towards the plenum. Want to hazard a guess what it does ?

[Zynerji]

Interesting story by Matthew Carter (former team Lotus F1 CEO) on Missed Apex Podcast about Spa 2015, when Mercedes "gave Lotus a different engine mode" in order to have Grosjean finish ahead of Vettel. According the Carter, Grosjean told him after the race that the car had nevert felt like that (good) before. Lotus were neven 'given' that engine mode again.

You can watch the whole clip here: https://youtu.be/xBQA9eABkpc?t=27m10s

What I wonder is how Mercedes can 'give' or unlock an engine mode for the driver to use, when pit-to-car telemetry is forbidden? I understand how Mercedes can keep certain engine modes from their customer teams, but I'm a bit puzzled as to how they can choose to unlock them for a driver to use during a race when the car is on track.

This is by far the most interesting question... hidden switch combinations? Should the FIA 2021 rules include all software and disclosure of modes to the customer teams as well?

This seems to be a crazy advantage to overcome. And provably why Ron Dennis dropped Mercedes in favor of the factory Honda deal.

[stevesingo]

I don't think you would get any helmholtz resonance from a inlet tube within a tube. There will not be enough change in volume at the mouth of the runner.

You don't need a large change. Have you seen how small the step is on an exhaust manifold ?

Perhaps I should have worded that differently. I don't think the best benefit of VLI would be having a small transition.

There doesn't seem to be any visible mechanism for operating variable length runners.

Look again at the pics hurrill has posted - there's a big ass rail running along the crank centreline with little wings extending on both sides towards the plenum. Want to hazard a guess what it does ?

I could guess, but without seeing the the opposite side of the plenum being handled and how it may connect, it would be difficult to be sure.

Unlike yourself, which judging by the tone of your response, you seem very sure.

[PlatinumZealot]

I don't think you would get any helmholtz resonance from a inlet tube within a tube. There will not be enough change in volume at the mouth of the runner.

You don't need a large change. Have you seen how small the step is on an exhaust manifold ?

There doesn't seem to be any visible mechanism for operating variable length runners.

Look again at the pics hurrill has posted - there's a big ass rail running along the crank centreline with little wings extending on both sides towards the plenum. Want to hazard a guess what it does ?

From the photo it looks like bumps on a stick so we need more angles.

Btw. It is crazy to have the VLIM so close to the head. The VLIM has to go at the end of the runner. That is the only way you will get the big reflection from that big ole' plenum they put there "for whatever reason."

[gruntguru]

re "hole in plenum entry". The pipe feeding the plenum would be tapered, the idea being to gradually reduce the velocity as the air approaches the plenum. Ideal condition in the plenum itself is zero velocity and the transition from very high velocity at the compressor wheel tips to near-zero in the plenum should be as gradual as possible.

I can show you plenty of CFD that will show you near zero velocity in the plenum is nearly impossible to achieve.

No need for CFD to show that a 100mm diameter plenum feeding one bank of an F1 will have an average velocity of about 10 m/s. OK - it's not "near-zero" - it's "low velocity". Peak intake runner velocity would be an order of magnitude higher so kinetic energy will be two orders of magnitude higher. On this basis, the velocity in the plenum is "very low".

Especially on a forced induction application.

Boost makes no difference to velocity. It's the air density that changes.

[MrPotatoHead]

re "hole in plenum entry". The pipe feeding the plenum would be tapered, the idea being to gradually reduce the velocity as the air approaches the plenum. Ideal condition in the plenum itself is zero velocity and the transition from very high velocity at the compressor wheel tips to near-zero in the plenum should be as gradual as possible.

I can show you plenty of CFD that will show you near zero velocity in the plenum is nearly impossible to achieve.

No need for CFD to show that a 100mm diameter plenum feeding one bank of an F1 will have an average velocity of about 10 m/s. OK - it's not "near-zero" - it's "low velocity". Peak intake runner velocity would be an order of magnitude higher so kinetic energy will be two orders of magnitude higher. On this basis, the velocity in the plenum is "very low".

Especially on a forced induction application.

Boost makes no difference to velocity. It's the air density that changes.

Things are not quite that simple though when it comes to the behaviour in an intake and I think you will find the average velocity will be far higher than 10 m/s in the plenum.

[gruntguru]

I can show you plenty of CFD that will show you near zero velocity in the plenum is nearly impossible to achieve.

No need for CFD to show that a 100mm diameter plenum feeding one bank of an F1 will have an average velocity of about 10 m/s. OK - it's not "near-zero" - it's "low velocity". Peak intake runner velocity would be an order of magnitude higher so kinetic energy will be two orders of magnitude higher. On this basis, the velocity in the plenum is "very low".

Especially on a forced induction application.

Boost makes no difference to velocity. It's the air density that changes.

Things are not quite that simple though when it comes to the behaviour in an intake and I think you will find the average velocity will be far higher than 10 m/s in the plenum.

erm no. I did the calc for 100mm diameter, 12,000 rpm, 100% VE. The average velocity perpendicular to a 100mm diameter cross section is close to 10 m/s - fact. Sure there will be velocity components in the plane of the cross section and velocity peaks well above (and below) the average velocity but a fundamental goal of the design will be to minimise these. High velocity = high losses.

So - if there happen local velocities of the order of 20 m/s in said 100mm plenum my basic argument still holds.

[hurril]

From the photo it looks like bumps on a stick so we need more angles.

Btw. It is crazy to have the VLIM so close to the head. The VLIM has to go at the end of the runner. That is the only way you will get the big reflection from that big ole' plenum they put there "for whatever reason."

But they are at the end of the runners and they [the runners] have got to be at 50cm for sure, judging from the photos.

[stevesingo]

Using this calculator http://www.exx.se/techinfo/runners/runners.html with the following

Bore: 80mm
Stroke: 53mm
inlet cam duration: 300deg

6000rpm 52.7cm length (1st order)
8000rpm gives a 39.6cm
10500rpm gives 30.3cm
12000rpm gives 26.6cm, although I don't see the point in going beyond 10500rpm due to fuel limit.

Looks like that in the meat of the operating range a 10cm rang of length would be required.

[henry]

Using this calculator http://www.exx.se/techinfo/runners/runners.html with the following

Bore: 80mm
Stroke: 53mm
inlet cam duration: 300deg

6000rpm 52.7cm length (1st order)
8000rpm gives a 39.6cm
10500rpm gives 30.3cm
12000rpm gives 26.6cm, although I don't see the point in going beyond 10500rpm due to fuel limit.

Looks like that in the meat of the operating range a 10cm rang of length would be required.

You go beyond 10500 rpm because by increasing the efficiency of the induction system you reduce the compressor load which increases the amount of energy available to the MGU-H. So there is no increase in instantaneous ICE power but the PU can deliver power for longer.

[hurril]

Using this calculator http://www.exx.se/techinfo/runners/runners.html with the following

Bore: 80mm
Stroke: 53mm
inlet cam duration: 300deg

6000rpm 52.7cm length (1st order)
8000rpm gives a 39.6cm
10500rpm gives 30.3cm
12000rpm gives 26.6cm, although I don't see the point in going beyond 10500rpm due to fuel limit.

Looks like that in the meat of the operating range a 10cm rang of length would be required.

You go beyond 10500 rpm because by increasing the efficiency of the induction system you reduce the compressor load which increases the amount of energy available to the MGU-H. So there is no increase in instantaneous ICE power but the PU can deliver power for longer.

This is a non-sequiteur. What is the relation between upping the revs and the efficiency of the induction system?

You want to rev past 10,500 rpm because as you change gears, you'll want to have the "peak power segment" align perfectly with the "torque requirement."

[henry]

Using this calculator http://www.exx.se/techinfo/runners/runners.html with the following

Bore: 80mm
Stroke: 53mm
inlet cam duration: 300deg

6000rpm 52.7cm length (1st order)
8000rpm gives a 39.6cm
10500rpm gives 30.3cm
12000rpm gives 26.6cm, although I don't see the point in going beyond 10500rpm due to fuel limit.

Looks like that in the meat of the operating range a 10cm rang of length would be required.

You go beyond 10500 rpm because by increasing the efficiency of the induction system you reduce the compressor load which increases the amount of energy available to the MGU-H. So there is no increase in instantaneous ICE power but the PU can deliver power for longer.

This is a non-sequiteur. What is the relation between upping the revs and the efficiency of the induction system?

You want to rev past 10,500 rpm because as you change gears, you'll want to have the "peak power segment" align perfectly with the "torque requirement."

I think the question was “why would you vary the induction tract length beyond 10500”. If you don’t change the length of the induction tract the Volumetric Efficiency will reduce.

So that is the relationship between upping revs and the efficiency.

I suspect one of us is misinterpreting the question.

[stevesingo]

That makes sense to the extent that anything which increases VE will reduce compressor load and therefore more turbine work available for MHU-H.

I expect it is a fine line, from what energy gained from increased MGU-H harvest and what is lost through increased friction.

[MrPotatoHead]

No need for CFD to show that a 100mm diameter plenum feeding one bank of an F1 will have an average velocity of about 10 m/s. OK - it's not "near-zero" - it's "low velocity". Peak intake runner velocity would be an order of magnitude higher so kinetic energy will be two orders of magnitude higher. On this basis, the velocity in the plenum is "very low".
Boost makes no difference to velocity. It's the air density that changes.

Things are not quite that simple though when it comes to the behaviour in an intake and I think you will find the average velocity will be far higher than 10 m/s in the plenum.

erm no. I did the calc for 100mm diameter, 12,000 rpm, 100% VE. The average velocity perpendicular to a 100mm diameter cross section is close to 10 m/s - fact. Sure there will be velocity components in the plane of the cross section and velocity peaks well above (and below) the average velocity but a fundamental goal of the design will be to minimise these. High velocity = high losses.

So - if there happen local velocities of the order of 20 m/s in said 100mm plenum my basic argument still holds.

I would love to see your math on those calculations please