What's wrong with the Tesla regen? Or my car? (Chart)

[tedsk]

Could it be because of inaccuracy in car speed being reported by software?
All manufacturers factor some inaccuracy in speedometer reading and usually it progresses with speed.
If inaccuracy is non-linear then it could be the answer.
From the graph I see that I made 0-60 in less than 5.5 seconds. I doubt I beat the specs of my MS 75. Rather readings were incorrect.
I'm tracking speed that car reports via CAN, no VBOX or similar equipment.

 

[Saghost]

Could it be because of inaccuracy in car speed being reported by software?
All manufacturers factor some inaccuracy in speedometer reading and usually it progresses with speed.
If inaccuracy is non-linear then it could be the answer.
From the graph I see that I made 0-60 in less than 5.5 seconds. I doubt I beat the specs of my MS 75. Rather readings were incorrect.
I'm tracking speed that car reports via CAN, no VBOX or similar equipment.

Are you asking if I think the increase in deceleration your graph shows as the speed falls is a product of CANBus reading inaccuracy?

If that's the question, then no, I don't. The car does slow faster as it gets to lower speeds, as your graphs show, and I believe it is because the car sets a regen energy limit rather than a torque limit, as I mentioned earlier.

Tesla's 0-60 times for the non-P cars are generally found to be conservative, even without the rollout discussion. I'm not surprised your car came up fractionally faster than rated.

 

[tedsk]

You are right, if the reading of speed on CAN factor increasing inaccuracy then rate slowdown should decrease not increase. So it is not the case here.
Car surely sets the power limit since it is limited by the AC-DC converter.

 

[tedsk]

BTW: Acceleration (below 42mph) is steadily increasing on my graph. What could it be?

 

[Saghost]

Car surely sets the power limit since it is limited by the AC-DC converter.

Maybe. Conventional wisdom is it's a decision to protect the battery pack, though during supercharging it sees ~50% higher levels. Could be to protect the drive inverters which are doing the AC->DC conversion instead, I don't think we know.

With the rear drive cars, there were driving dynamics concerns as regen was on the rear wheels only and the car could potentially get loose from the regen force. but Tesla didn't increase the regen much if any when they went to 4 wheel regen with the D cars.

 

[tedsk]

BTW: Acceleration (below 42mph) is steadily increasing on my graph. What could it be?

Never mind this one. It is averaging effect.

 

[Saghost]

BTW: Acceleration (below 42mph) is steadily increasing on my graph. What could it be?

Not quite sure. Most of the factors I can think of would seem to point the other way. Maybe Tesla programmed the curve that way, but I don't know why they would.

 

[tedsk]

Here is graph with 0.3sec averaging. Acceleration is flat.

 

[SageBrush]

Thanks to OP and all -- interesting topic!

Leaving aside the UI part of one pedal or two and the last 5 mph, aren't most drivers going to be most used to constant forces at constant pedal pressure ?

 

[AWDtsla]

Tesla recently sharply reduced regen under 10mph, tapering to 0 at 5 mph. It sucks.

Firmware 8.0

 

[tedsk]

Totally agree - constant force (deceleration) is what most drivers would expect if pedal position is not changing.

 

[Saghost]

Thanks to OP and all -- interesting topic!

Leaving aside the UI part of one pedal or two and the last 5 mph, aren't most drivers going to be most used to constant forces at constant pedal pressure ?

Not really. I do think that a constant acceleration rate for a given accelerator position would be great to have, and more intuitive than the alternative - but I've never had a car that came close to it. Actually, the Tesla is the closest I've had, even with the quirks shown here.

Typical cars respond differently depending on what gear they are in. Worse, many of them have widely varying torque curves as rpms change - so even in the same gear, acceleration at 1500 rpm and 2000 rpm may be widely different, or between 5000 and 6000 it may suddenly double the acceleration (turbo car getting on boost, VTEC hitting the fat cam.)

Several cars I've driven including my Eos had weird all or nothing responses off the line that make them almost impossible to start smoothly at anything except a slow drift.

GM programmed my Volt with a three different accelerator curves (Normal/Mountain/Hold, Sport, and Propulsion Power Reduced,) and none of them are anything like flat or predictable - the main profile does almost nothing for the first half of the pedal, then suddenly surges, while the Sport profile is nicely linear and very aggressive for the first half of the pedal - then does almost nothing for the rest.

I'm not saying there isn't room for improvement, but I actually think Tesla has done a great job with the accelerator - easy to modulate while driving either quickly or slowly, and much more predictable than most. I do wish they could give me predictable full regen with a cold battery for my daily commute in winter, though - the limited regen often trips me up now that I'm used to the full capability.

 

[SageBrush]

Totally agree - constant force (deceleration) is what most drivers would expect if pedal position is not changing.

I was just imagining pad pressure on the rotors. Does a constant application lead to a constant force ? And does a constant application lead to a constant rate of heat dissipation ? is dKE/dt constant ?

There is something goofy in my thinking because I am coming up with different conclusions here ..

 

[Saghost]

I was just imagining pad pressure on the rotors. Does a constant application lead to a constant force ? And does a constant application lead to a constant rate of heat dissipation ? is dKE/dt constant ?

There is something goofy in my thinking because I am coming up with different conclusions here ..

Consistent brake pedal application leads to consistent force on the pads - which mostly leads to consistent deceleration rates - coefficients of friction are relatively independent of rate and force, so you should get a linear force response through that system and a consistent torque on the wheel - until the pad heats up, reducing friction, or until you get to nearly stopped, where you're dealing with a different, higher coefficient (sliding coefficients of friction are lower than static coefficients.)

Remember how hard it was to avoid that sudden jerk at the end when coming to a stop when you were learning to drive, how you had to learn to ease off of the brake right at the end to make a perfect stop? This is why.

Brake disc/pad heating won't be constant - you're dissipating much more energy at higher speeds - and so brake performance may change over time at a given pressure level (the infamous "fade" that all the magazines test for and report on in reviews.)

 

[SageBrush]

Brake disc/pad heating won't be constant - you're dissipating much more energy at higher speeds - and so brake performance may change over time at a given pressure level (the infamous "fade" that all the magazines test for and report on in reviews.)

Thanks

Now that I think about it a bit more, is friction heat about proportional to car speed as the car slows down during a constant brake pedal application ?

 

[Saghost]

Thanks

Now that I think about it a bit more, is friction heat about proportional to car speed as the car slows down during a constant brake pedal application ?

I think heat energy added to the discs should be, yes. You're converting kinetic energy to heat, and to get the same deceleration at a higher speed you need to convert more energy linearly with velocity, based on the 1/2 m * v ^ 2 formula.

Disc temperature will depend on cooling and thermal mass as well as heat input.

 

[SageBrush]

I think heat energy added to the discs should be, yes. You're converting kinetic energy to heat, and to get the same deceleration at a higher speed you need to convert more energy linearly with velocity, based on the 1/2 m * v ^ 2 formula.

I was starting from the notion that Work = force * distance, and the distance the pads travel over the rotors is proportional to car speed.

Anyway, back to OP's comment and lament, I can see his point that a non constant deceleration would take getting used to.

 

[Saghost]

I was starting from the notion that Work = force * distance, and the distance the pads travel over the rotors is proportional to car speed.

Anyway, back to OP's comment and lament, I can see his point that a non constant deceleration would take getting used to.

Also a valid approach.

I thought you were talking about acceleration, hence my lecture about how different it is in most cars. It's true that Tesla follows a curve on deceleration- but also true that the curve is pretty predictable/consistent, and something your monkey brain seems to pick up on fairly quickly in my experience - I only get frustrated when my regen is reduced from cold/full battery.

 

[tedsk]

Not really. I do think that a constant acceleration rate for a given accelerator position would be great to have, and more intuitive than the alternative - but I've never had a car that came close to it. Actually, the Tesla is the closest I've had, even with the quirks shown here.

Typical cars respond differently depending on what gear they are in. Worse, many of them have widely varying torque curves as rpms change - so even in the same gear, acceleration at 1500 rpm and 2000 rpm may be widely different, or between 5000 and 6000 it may suddenly double the acceleration (turbo car getting on boost, VTEC hitting the fat cam.)

Several cars I've driven including my Eos had weird all or nothing responses off the line that make them almost impossible to start smoothly at anything except a slow drift.

GM programmed my Volt with a three different accelerator curves (Normal/Mountain/Hold, Sport, and Propulsion Power Reduced,) and none of them are anything like flat or predictable - the main profile does almost nothing for the first half of the pedal, then suddenly surges, while the Sport profile is nicely linear and very aggressive for the first half of the pedal - then does almost nothing for the rest.

I'm not saying there isn't room for improvement, but I actually think Tesla has done a great job with the accelerator - easy to modulate while driving either quickly or slowly, and much more predictable than most. I do wish they could give me predictable full regen with a cold battery for my daily commute in winter, though - the limited regen often trips me up now that I'm used to the full capability.

I meant regen braking. Nobody expect linear acceleration since there is no such experience from ICE cars. Braking is completely different - everybody got used to nearly linear deceleration that directly depends on the force applied to braking pedal. So one would naively expect same behavior from regen, right? I was unpleasantly surprised to find out that it is not the case with Tesla. At least when battery can't accept full regen power.

 

[tedsk]

Remember how hard it was to avoid that sudden jerk at the end when coming to a stop when you were learning to drive, how you had to learn to ease off of the brake right at the end to make a perfect stop? This is why.

This one is just partially true. I'm not sure to what extent actually. Yes indeed still friction force is higher than sliding but jerking at the end of deceleration comes from different source. Since your car's center of gravity is above ground there is a force that compresses front shock absorbers and relaxes rear ones. When car stops this force diminishes to zero but struts and shock absorbers still have potential energy that has to go somewhere. So front suspension goes up and rear down. Thats the source of jerking, not friction. If you gradually decrease braking force at the end then you are making shock absorbers less and less stressed so jerking is less noticeable.