200 kWh Roadster Pack: How is Tesla Pulling This Off?

[abasile]

If they do, or did, tweak the energy density/power density cell types, they would have gone to more power density. The point of the big battery wasn’t to have a 600 mile range (that’s fairly pointless in a 2+2 seater). The reason for the big battery was to have enough power for 1.9s launches.

Do we know that? With tweaks for power density, it's arguable that even a 100 kWh battery could be sufficient for 1.9s launches, particularly as the 2020 Roadster is a smaller car with likely more efficient motors.

My personal feeling is that Elon wanted to hit as many superlatives as possible. Range is a big deal with EVs. To be able to say that Tesla's top of the line vehicle hits 1000 km of rated range is just jaw dropping.

 

[mongo]

Guys, the prototype Roadster we saw Thursday night had a 200kwh (or whatever) pack, otherwise it could not have done 1.9s launches. And it cornered really well too. It is likely a double stack since people said they felt the seats were higher off the ground (which is actually a nice thing from a driving experience point of view). The only question is how they got it to corner so well in such a heavy car. One of the drivers in a demo video did say that the car was sprung very tight.

Quick numbers
Ground clearance 10 cm
Double pack height 14 cm
Structure 5 cm
Top of pack is at 29 cm, that is below the center of a P100D 265/35P21 wheel 28.3" diameter, radius of 35.8 cm.

Such a low CG means there is less rolling force in the curve. With a width of 1.9m (original Roadster) the angle is atan(29/950)=1.8 degrees. In useful terms, 1 G of side load produces 0.027 G of rotational force (I think)...

 

[Cosmacelf]

Also, the ridiculously wide tires in the Roadster helps a lot when cornering.

 

[mongo]

Also, the ridiculously wide tires in the Roadster helps a lot when cornering.

And they're good for bridging potholes

 

[Olle]

There shouldn't be a doubling of weight with two stacked packs if they are tailor made as one. Since you don't need two complete cases. From a quick google search you can deduct that there is 300 kg cells and perhaps 250 kg casing in a 100D pack. Let's say 600 kg cells in a 200kWh pack then. And that's with 18650! 2170 should be 15% more energy dense according Straubel in an earnings call last year. So now we are talking approximately 500 kg. I don't see a mystery.

In an expensive car like this with more exotic materials in integrated battery case/body frame wouldn't be awfully heavy. IMHO they're not planning on exotic batteries in this car, and pretty sure that the prototype was more or less the real deal with 200kWh worth of 2170 cells

 

[mikevbf]

Is no one going to mention the cornering advantage given to torque vectoring that can be accomplished by having a separate motor for each back wheel in addition to the front motor? I am much less of a car geek/engineer than most in these discussions, so please someone with more knowledge explain how significant or insignificant this could be in such a heavy car.

Also would't the extra motor be just as much about the power equation as the extra battery capacity? It's also possible that the 1.9's they were running in the prototype had larger more powerful motors and less batteries than the balance they will put in the final version. As battery density and power potential increases over the years, they might decrease motor size which would also increase trunk space in the rear space while still being able to do 1.9 launches. This would also give the the latitude for even more powerful models with perhaps less luggage and/or passenger space. Remember 1.9 launches is only a base spec!

 

[Saghost]

I'm guessing the New Roadster prototypes wern't possible until the Gigafactory's new 2170 cells were in production (for reasons of energy density and perhaps heat tolerance). So how are they pulling off 200kWh? Here are a few ideas:

(1) Switching to air cooling? By pulling out the liquid cooling hardware they could increase cell density by perhaps 40%, thus doubling the number of cells that would fit in the same space.

(2) Living with a lot of weight? In a way, weight is Tesla's secret weapon. It makes their cars safer, and increases tire traction for faster acceleration times. The traditional drawbacks are easily overcome with a low center of gravity, even distribution, and powerful electric motors.

(3) Using an odd pack shape that combines Tesla's traditional floor-mounted pack with an additional box somewhere else (perhaps under the rear seat)?

Actual air cooling would likely require more space than the current liquid cooling setup. What I think you meant is going to no heating or cooling of the pack, and I really doubt Tesla will do that, for safety reasons if nothing else.

On the other hand, we saw a string of patents a couple years ago for doing both contacts on the top of the cell with a flexible circuit board connection with designed in fusible bottlenecks. That makes the joining to the cell much simpler and leaves the bottom of the cell as a cooling surface along with the sides.

Tesla also patented a bottom of cell cooling with heat pipes as I recall - there was a lot of discussion about whether end on cooling would be more effective than having to cool through the depth of the insulators and electrolyte..

Put all together, they can likely do better cell heating and cooling than the current designs while packing the cells to the logical hexagonal limit in a pack that's still thinner and lighter than today's designs. I suspect the prototype has a pack like this, designed from the ground up as a double layer. Still a lot of weight, but you can't get away from that until the cell technology improves without big tradeoffs.

 

[Saghost]

Also would't the extra motor be just as much about the power equation as the extra battery capacity? It's also possible that the 1.9's they were running in the prototype had larger more powerful motors and less batteries than the balance they will put in the final version. As battery density and power potential increases over the years, they might decrease motor size which would also increase trunk space in the rear space while still being able to do 1.9 launches. This would also give the the latitude for even more powerful models with perhaps less luggage and/or passenger space. Remember 1.9 launches is only a base spec!

It's not motor size vs battery size. The motors and the batteries each impose constraints on the maximum output of the car, and the lowest limit becomes the car's maximum capacity.

Adding bigger motors to a car that's already battery constrained won't make it any faster (technically, it'll make it a little slower, since the bigger motors add more weight and are likely not quite as efficient at the lower output, but the difference will be small.)

 

[mikevbf]

It's not motor size vs battery size. The motors and the batteries each impose constraints on the maximum output of the car, and the lowest limit becomes the car's maximum capacity.

Adding bigger motors to a car that's already battery constrained won't make it any faster (technically, it'll make it a little slower, since the bigger motors add more weight and are likely not quite as efficient at the lower output, but the difference will be small.)

I get it on the power equation. I'm still working on the cornering angle. The torque vectoring advantage of 2 rear motors would somewhat explain how such a heavy car could corner well, right? The lower cog accomplished by the heavy low battery packs although great for handling to a degree have to have point at which the extra weight starts to become more of a disadvantage in cornering? I wonder if an upper stack of battery packs in the roadster is still low enough to still decrease the cog of the roadster, but less so than the lower stack. Or would the upper stack actually raise the cog?

 

[Saghost]

I get it on the power equation. I'm still working on the cornering angle. The torque vectoring advantage of 2 rear motors would somewhat explain how such a heavy car could corner well, right? The lower cog accomplished by the heavy low battery packs although great for handling to a degree have to have point at which the extra weight starts to become more of a disadvantage in cornering? I wonder if an upper stack of battery packs in the roadster is still low enough to still decrease the cog of the roadster, but less so than the lower stack. Or would the upper stack actually raise the cog?

I don't think we know much about the Roadster 2's cornering ability yet. Torque vectoring certainly will help with cornering some, but it's at least as much about feel and handling response at initial turn in as it is ultimate limits - at the limit, tire traction defines the boundary and vectoring can't expand it (though if the car's suspension design/tuning causes one tire to hit the limit before the others, vectoring can even that out.)

Not sure where you're going with the center of gravity question... from what we saw, anything that's under the seats will be below the centerline of the wheels, and below where most of the weight in a typical supercar is. Certainly putting all the weight in a single layer right along the bottom of the car produces the lowest CG, but we (the underinformed speculators on the web ) don't believe it's possible to put as much battery as Tesla has added to the car into that single layer - and having the power is more important than dropping the CG by an inch or two.

 

[abasile]

Put all together, they can likely do better cell heating and cooling than the current designs while packing the cells to the logical hexagonal limit in a pack that's still thinner and lighter than today's designs. I suspect the prototype has a pack like this, designed from the ground up as a double layer. Still a lot of weight, but you can't get away from that until the cell technology improves without big tradeoffs.

Yes. It's also worth pointing out that Tesla Semi would have provided motivation for this engineering work, given the need to stack battery cells. I'm truly impressed that, in about two years, the world will have a full-sized electric semi truck capable of hauling a full load for 500 miles on a single charge!

 

[Electroman]

And that's with 18650! 2170 should be 15% more energy dense according Straubel in an earnings call last year

It has been analysed and beaten to death in this forum - 2170 is no more energy dense than 18650.

That is why I think Musk is getting ahead of himself in the stated 200 kWh.

 

[Yggdrasill]

I've had some time to think about this now and I think the double-stack theory is correct. They may even have triple stacked under the seats, and some capacity may be in the frunk. I think it's doable (barely) without sacrificing the rear seats. I assume they're using the same module design as the Model 3 and/or powerwall, and probably the same chemistry as the Model 3. (Though they may have calculated some improvement before 2020, so the installed pack may be something like 180 kWh.)

Pack weight should be somewhere in the 900 kg/2000 lb area. And with basically all of it below the center of the wheels, the new Roadster likely has the lowest center of gravity of any car in history. That should lend itself to not only acceleration, but quite good handling as well.

 

[strider]

When the P85D Model S came out, and beat my Roadster's acceleration. I was envious, and figured I'd trade up to the Roadster 2 when it came out. But Tesla went overboard. Instead of building a better sports car they built a supercar, and priced it outside what almost anyone will pay. Yes, the super-rich will pay it. But this is no longer a car for upper-middle-class men having their mid-life crisis. This will be a car for race-car drivers in the 1%.

The v4 Roadster is exactly as I have been saying. They went straight after the top end 911 market, right down to the 2+2 seating, which I believe is the right target market for this car. From what I can gather with a quick web search, Porsche sells about 1,500 911 Turbo S's per year so there is a market there. A 911 Turbo S convertible starts at $203,000, hard top at $191,000 so they are right in there with WAY better performance (911 Turbo S is 2.8 to 60). Tesla needed/wanted a halo car and that's what they made.

I think the massive range is a byproduct of needing enough cells to flow the power required for 1.9 to 60, not a goal in and of itself.

I think the mid-life crisis car will be a Performance S or 3. It wouldn't take much to do a coupe version of the 3 with the existing SP100DL drivetrain (I'm assuming improvements in battery tech will allow them to put 100kWh in the 3) and voila, very quick but not stupidly expensive and still practical car.

 

[FlyingCookie]

I've had some time to think about this now and I think the double-stack theory is correct. They may even have triple stacked under the seats, and some capacity may be in the frunk. I think it's doable (barely) without sacrificing the rear seats. I assume they're using the same module design as the Model 3 and/or powerwall, and probably the same chemistry as the Model 3. (Though they may have calculated some improvement before 2020, so the installed pack may be something like 180 kWh.)

Pack weight should be somewhere in the 900 kg/2000 lb area. And with basically all of it below the center of the wheels, the new Roadster likely has the lowest center of gravity of any car in history. That should lend itself to not only acceleration, but quite good handling as well.

Until the whole car starts to slide, which is already the problem with the Model S. It'll turn without rolling, but you still hit a point where the tires just can't deal with the side-loading from the weight of the car pushing to the outside of the turn.

 

[12Pack]

Of course the real trick for an electric performance car is if it can routinely lap around a track where it’s continually being asked for hard accelerations. My Model S doesn’t make it through a single lap without power being limited due to battery heating.

BTW this is not a troll - I have a deposit on the roadster. But won’t go through with the purchase unless this is demonstrated to be not be an issue.

 

[Saghost]

Until the whole car starts to slide, which is already the problem with the Model S. It'll turn without rolling, but you still hit a point where the tires just can't deal with the side-loading from the weight of the car pushing to the outside of the turn.

Happens to every car, and should happen at the same level of sideways acceleration for a given tire and track surface regardless of weight, more or less.

 

[strider]

Of course the real trick for an electric performance car is if it can routinely lap around a track where it’s continually being asked for hard accelerations. My Model S doesn’t make it through a single lap without power being limited due to battery heating.

BTW this is not a troll - I have a deposit on the roadster. But won’t go through with the purchase unless this is demonstrated to be not be an issue.

In the current Roadster the limitation is the PEM (AC/DC Inverter) and motor itself, not the battery. Tesla has a patent for liquid cooling the rotor on the motor but hasn't implemented it. Perhaps it's been too expensive to do but they can justify it on this car since it's more expensive. Or perhaps they're going to use IPM motors instead of async AC which I have read are easier to cool. The 250+ mph claim leads to believe they think they have the cooling problem solved as it takes awhile to get to 250+ mph so the car would need to maintain peak output for much longer than any current Tesla model.

All of that being said, VERY few supercars/hypercars ever see a race track. They're usually driven on public roads by pudgy rich guys seeking attention. In that realm (public roads) the v4 Roadster will destroy any comer.

 

[mongo]

Happens to every car, and should happen at the same level of sideways acceleration for a given tire and track surface regardless of weight, more or less.

But can the active torque control of the Roadster keep the car in the traction circle better than others? i.e. corner to the mu of the rubber vs trying to accelerate / deccelerate and breaking free.

 

[Saghost]

But can the active torque control of the Roadster keep the car in the traction circle better than others? i.e. corner to the mu of the rubber vs trying to accelerate / deccelerate and breaking free.

I'm not entirely sure. It may be some advantage if the other cars are breaking loose before they hit the theoretical limit of the tires. There have been several cars with torque vectoring, maybe a look at how they handle the skid pad would be educational. Comparing the Rimac and the NSX to their rivals might be especially educational.