Tesla P85D does sub 9 minute BTG run at Nurburgring

[Ingineer]

There is no chance that the system hasn't maximized its efforts to cool the car down by running the compressor full tilt.

You know this for sure? Based on the performance I've seen on the code inside that HVAC controller, it's full of bugs and can't even manage to properly implement a decent cabin temp control algorithm.

I think the electric compressor Tesla uses has a 7.4kW cooling capacity. That adds some, but the airflow though the center radiator should be able to handle a lot more than that, especially with the airflow present on the straight-away speeds on the track. It might be hobbled by plumbing or pump capacity though.

 

[Saghost]

Yeah, getting the inverter, rotor and stator temps would help a lot. While we're at it, add in the pump RPM's, A/C compressor target, valve positions, etc. Seeing these would help us to devise a system to keep them cooler for track use. Tesla probably doesn't care about such things, but maybe there's an Engineer or two that would be interested enough to help off-hours. I bet Elon would be ok with this, even though I don't see him officially tasking engineering with such a project.

You mean maybe a center console screen showing all of the temperatures and pump speeds? Maybe something like this?

Sometimes Tesla frustrates me. They created a screen full of useful information, with pretty animations even - and buried it behind a Tesla SC password. If we could have this screen as a selectable option, it would let us answer so many questions...

(Edit: Not my picture. I saw it on a thread a year or two ago, found it again with Google, and had to download/re-upload it because the image wasn't displaying right. Original thread here.)
Walter

 

[Dbitter1]

You mean maybe a center console screen showing all of the temperatures and pump speeds? Maybe something like this?

Wait Whut? That's like the single coolest screen I've seen from the MS yet!!!!1!!eleven!

I wholeheartedly second the frustration to any of the actual Tesla employees reading this...

I came from the nuclear power industry, and I LOVE screens like that (reminds me of old home?). It will save us lots of time trying to reverse engineer it from CANBUS or whatever is used (I haven't even tried in my MS yet, too busy launching myself silly).

 

[Ingineer]

Yes, I've seen this screen too. There are many other cool things in the service menu too! If you hold down the "T" symbol for about 5 seconds then release, it will ask for a code. That code will get you into the service menu. Unfortunately (supposedly) the code is not only different for each car, but it also changes every 4 hours. They use a cryptographic algorithm to generate it from the VIN. You have to have that algorithm to get in, and apparently only the mothership does.

They would have to give independent mechanics access to this menu for general repair functions, just like service manuals. But they only give the service manuals to residents of MA because they have a law forcing them to do so. But apparently the software clause doesn't apply until 2018. Nobody has yet sued Tesla over this practice, probably because most of the cars are still under warranty. Someone definitely will once the cars come off warranty. Of course by then, they'll probably dumb-down the service menu to where it's not very useful.

 

[Saghost]

Yes, I've seen this screen too. There are many other cool things in the service menu too! If you hold down the "T" symbol for about 5 seconds then release, it will ask for a code. That code will get you into the service menu. Unfortunately (supposedly) the code is not only different for each car, but it also changes every 4 hours. They use a cryptographic algorithm to generate it from the VIN. You have to have that algorithm to get in, and apparently only the mothership does.

They would have to give independent mechanics access to this menu for general repair functions, just like service manuals. But they only give the service manuals to residents of MA because they have a law forcing them to do so. But apparently the software clause doesn't apply until 2018. Nobody has yet sued Tesla over this practice, probably because most of the cars are still under warranty. Someone definitely will once the cars come off warranty. Of course by then, they'll probably dumb-down the service menu to where it's not very useful.

A lot more details than I knew about the service menu.

I can see not wanting to release the algorithm, but they really should give owners access to their own cars. Maybe they could set up a web portal for it, where owners have to authenticate (to MyTesla?) and supply the VIN, and it gives them their own code for the moment (which they can then give to the shop for a few hours if a shop needs it.)
Walter

 

[FredTMC]

Great vid! Thx for posting. I was sad to see the car was power limited for almost the entire run. Ugh

 

[Ingineer]

A lot more details than I knew about the service menu.

I can see not wanting to release the algorithm, but they really should give owners access to their own cars. Maybe they could set up a web portal for it, where owners have to authenticate (to MyTesla?) and supply the VIN, and it gives them their own code for the moment (which they can then give to the shop for a few hours if a shop needs it.)
Walter

That's just it, they don't want owners in there. (or anyone but them for that matter) The lawyers are piled on thick at Tesla.

 

[Saghost]

That's just it, they don't want owners in there. (or anyone but them for that matter) The lawyers are piled on thick at Tesla.

Well, hopefully that will change by the time I can afford one.

 

[thegruf]

He clearly is running a lot of regen. I also see the traction/yaw control constantly triggering. I wonder if it's good enough to help or hinder on a track like this? I bet it can be disabled with a simple fuse pull.

Hmm, would adding a beefier pump with more flow help? Seems like with the size of the radiator and A/C there should be enough heat transfer to get rid of it all. Maybe it's not even kicking in the A/C?

Yeah, getting the inverter, rotor and stator temps would help a lot. While we're at it, add in the pump RPM's, A/C compressor target, valve positions, etc. Seeing these would help us to devise a system to keep them cooler for track use. Tesla probably doesn't care about such things, but maybe there's an Engineer or two that would be interested enough to help off-hours. I bet Elon would be ok with this, even though I don't see him officially tasking engineering with such a project.

I devised a hack to the LEAF's HVAC controller to allow direct control of the system's cabin fluid heater. This was before the addition of the heat pump. Their HVAC team had (badly) engineered the system to turn on the heater anytime the glycol loop was below a certain temp, which led to a LOT of wasted energy consumption. When you only have about 21kWh usable, that ~5kW heater load is a serious issue. Luckily Nissan listened to us when we asked for a heat-pump in 2011. They added it in 2013, which is pretty awesome. Now it only Tesla would do it...

I have thought for a while that the simplest way to improve current Model S for the track is to fit big brakes and switch off/low regen. An entirely "normal" performance mod that wouldnt upset any of the electronics and not outlandish costs either, and presumably could be perfectly useable for daily driving.

Losing the heat generated by regen in the motors and battery would I assume significantly extend the track capabilities of the Model S.

Somewhat surprised nobody done this yet?

 

[ItsNotAboutTheMoney]

I have thought for a while that the simplest way to improve current Model S for the track is to fit big brakes and switch off/low regen. An entirely "normal" performance mod that wouldnt upset any of the electronics and not outlandish costs either, and presumably could be perfectly useable for daily driving.

Losing the heat generated by regen in the motors and battery would I assume significantly extend the track capabilities of the Model S.

Somewhat surprised nobody done this yet?

Improve the capabilities but reduce the range. Better to work on improved capability, I think.

 

[thegruf]

nope. Regen still works on the street so range intact.

For the track, okay range suffers but most amateur races are short or one lap specials eg Nurburgring but thermals improve, possibly at least get some laps at full power

Agreed overall improved cooling capability is the best answer but assuming the rads are well enough sized it is more likely local heating within motors and battery cells that limit improvement.
It might be that larger rads, bigger cooling pumps improve matters, however I suspect that Tesla already have had to work pretty hard on this.

 

[Saghost]

nope. Regen still works on the street so range intact.

For the track, okay range suffers but most amateur races are short or one lap specials eg Nurburgring but thermals improve, possibly at least get some laps at full power

Agreed overall improved cooling capability is the best answer but assuming the rads are well enough sized it is more likely local heating within motors and battery cells that limit improvement.
It might be that larger rads, bigger cooling pumps improve matters, however I suspect that Tesla already have had to work pretty hard on this.

I'm pretty sure you're right, it is local heating in the inverter and/or the drive motor rotor that leads to power limiting.

That doesn't mean there isn't room for improvement with the current hardware, though. Tesla has it tuned for minimal power consumption, using the radiators and ambient air to cool when it can.

However, the system is already designed to be able to use the A/C to cool the battery when needed - and it can interconnect the two halves and cool the motors and battery on one circuit.

That means that you could have a track mode that throws efficiency out the window and precools the battery and motors well below ambient, and then continues to feed the motors supercooled coolant from the A/C while using the battery as a heat sump, with minor programming changes.

Since the heat transfer in cooling is a function of the temperature difference, cold coolant will transfer a lot more heat out.

The A/C will probably slowly lose ground - if you figure the motors are ~90% efficient and the A/C is 6 kW and ~2:1 (moves twice as much heat as it consumes,) it can handle about 120 kW worth of continuous power - but the cooled battery will give you some time, and once the coolant gets warm enough you can use the radiators, too.

There's a lot of opportunity if we could mess with the code just a little.
Walter

 

[JER]

I reckon rotor cooling is the limiting factor. It seems (in my estimation) to be relatively easy to efficiently liquid cool the inverter IGBTs, motor windings and battery cells compared to cooling a spinning copper drum being heated by eddy currents inside a sealed metal oven.

I did read somewhere that the new drive unit has liquid cooling of the rotor; not sure if this is true, or if so how it is accomplished. Fluid could only enter through the shaft, whereas it's the copper outer that needs to be cooled.

 

[tkizzy]

love seeing Model S' doing laps!

 

[Saghost]

I reckon rotor cooling is the limiting factor. It seems (in my estimation) to be relatively easy to efficiently liquid cool the inverter IGBTs, motor windings and battery cells compared to cooling a spinning copper drum being heated by eddy currents inside a sealed metal oven.

I did read somewhere that the new drive unit has liquid cooling of the rotor; not sure if this is true, or if so how it is accomplished. Fluid could only enter through the shaft, whereas it's the copper outer that needs to be cooled.

I'm not 100% certain, but from some of the things I've read I believe that all Model S drive motors are liquid cooled and have hollow drive shafts with coolant pumped through.

 

[brianman]

fit big brakes and switch off/low regen.
...
Losing the heat generated by regen in the motors and battery would I assume significantly extend the track capabilities of the Model S. ...

I do the latter for lapping events. I tends to keep the limiter a while a "little" longer but not much. "Significantly extend" - no.

 

[JER]

I'm not 100% certain, but from some of the things I've read I believe that all Model S drive motors are liquid cooled and have hollow drive shafts with coolant pumped through.

A hollow driveshaft was exactly what I was suggesting above; there isn't really another way of doing it for the rotor. Shafts would be hollow anyway for mass optimisation.

There is what looks like a manifold on the outer end of the old motor that might be drawing fluid through the shaft. I've not had a good look at the new ones.

Quite how flow would be arranged in the rotor to get good thermal exchange and get fluid both in and out of the rotor, I'm not sure. You couldn’t just run it down the middle and expect it to do much.

 

[Ingineer]

The rotor is definitely liquid cooled:

Here's the full patent: Patent US7489057 - Liquid cooled rotor assembly - Google Patents

 

[Saghost]

A hollow driveshaft was exactly what I was suggesting above; there isn't really another way of doing it for the rotor. Shafts would be hollow anyway for mass optimisation.

There is what looks like a manifold on the outer end of the old motor that might be drawing fluid through the shaft. I've not had a good look at the new ones.

Quite how flow would be arranged in the rotor to get good thermal exchange and get fluid both in and out of the rotor, I'm not sure. You couldn’t just run it down the middle and expect it to do much.

The other obvious option for rotor cooling (which I believe GM uses for the Volt?) is to keep it in a coolant bath (transmission fluid in the Volt's case,) possibly with channels through the rotor from side to side. You probably get better overall cooling, but you pay a price in increased spinning losses from rotating the more viscous fluid.
Walter

 

[JER]

The other obvious option for rotor cooling (which I believe GM uses for the Volt?) is to keep it in a coolant bath (transmission fluid in the Volt's case,) possibly with channels through the rotor from side to side. You probably get better overall cooling, but you pay a price in increased spinning losses from rotating the more viscous fluid.
Walter

Not going to happen at 15k RPM. You could maybe do a spray system. That might not play nice with the stator though.

The rotor is definitely liquid cooled:

Here's the full patent:

So (assuming they're actually using that patent) they're doing the twinwall thing, in and out at one end. Probably the best option.

Also, fluid only goes through the shaft in the illustration. (Probably the only way to manufacture it where sealing doesn't become a huge problem.) That means a rather small area for thermal transfer to the fluid. Honestly, I'm not surprised it can't keep up at the motor's maximum power.

I also note that the other end of the shaft is open:

youtu.be/cHRMDYXBJdc?t=8m57s

There must be uncut metal (or a bung) closing the tube somewhere.


On a slightly different note, maybe there's another, potentially better way to make thermal contact with the rotor?

Sandia Labs demonstrated an air bearing that acted as a thermal junction a few years back; last I heard it was being considered for use in a more efficient CPU cooler. Here's a more recent article I found about it: spectrum.ieee.org/tech-talk/consumer-electronics/gadgets/coolchips-kinetic-cooling-engine-combines-fan-and-heat-sink-into-awesome

In the CPU cooler application, the rotating part is a finned heatsink and the static part the heat source, but I suppose it could work in reverse in a motor: Stationary fluid jacket(s) in thermal contact with end(s) of the rotor via an air bearing. I guess it would have to be lifted off and "parked" away from the rotor when it wasn't rotating fast enough to maintain the air bearing.