At the Model S Plaid delivery event, Elon Musk gave us some pretty juicy details, including the whole motor map with a comparison to the old P100D motor, a .208 drag coefficient, a new, more efficient heat pump, and a battery without many details.

The fact that Elon did not tell us the battery was the new 4680 cells leads us to believe that Tesla has chosen to stick with the 18650 cells. Our opinion agrees with both Electrek and InsideEVs. Using what he has told us and feeding these numbers through our performance and heat transfer models, we can make an educated guess on the battery specifics and the new heat pump configuration.

Our performance model closely duplicated what Elon gave us for acceleration times. We can estimate gear ratios and max motor RPM with our performance model. We estimate Tesla’s new carbon sleeved motor has a 23,308 RPM redline@200 MPH compared to the P100D redline of 18000 RPM@155 MPH, as well as gear ratios the same as P100D.





New Carbon Sleeved Rotor in Motor

Based on our performance model we are predicting Tesla will use three 250 kW motors. Two in the back and one in the front. The two motors in the back will allow torque vectoring. Only one motor needs to be used for commonality like the Model 3.

The most incredible thing about this motor is the super high redline and extremely flat power curve. Tesla’s older motors had a peak power point after which power began to taper off. This power curve allows the new Plaid to pull well all the way to the 23,308 redline at 200 MPH.

Here are a few other interesting tibbits from Elon as of June 14, 2021: The motor is small enough to be picked up by one person. “The sleeve is actually a carbon fiber that is wrapped around the motor by a special machine made by Tesla automation. The fiber wrap put the rotor in compression and limits expansion allowing the higher RPM. Main advantage is a much stronger EM field compared to a rotor held together by metal.”

Perhaps this stronger EM field is what gives us the flat power curve?

Musk also stated that they “have a few ideas for increasing the torque and max RPM even further for the new Roadster.”

Battery using 18650 cells?

As mentioned, Elon didn’t say a whole lot about the battery but we can infer some things based on our performance model. We think Tesla may be using a higher power cell in the new Plaid compared to P100D. We are calculating a 7.5C discharging rate. That’s pretty high. In order to get the approx 400-mile range, we estimate 100 kWh usable energy.

We also ran our thermo-electric model, and, based on that we think, Tesla will use a battery configuration similar to Model 3 where the cells are glued directly to the cooling tubes. P100D cells were not glued to the cooling tubes, with the new gluing technique, we estimate 2X the heat transfer from the pack.



New high-efficiency heat pump

This super high C rate during acceleration and charging results in a pretty high amount of heat to get rid of. We are estimating that the new heat pump will have twice the heat removal capacity of the old P100D.


Based on Tesla’s photo of the new heat pump, it appears that the condenser will NOT be a glycol heat exchanger like the Model 3's. Remember that the Model 3 used a separate glycol loop to carry the heat from the heat pump to the front of the car. It looks like the condenser in the Model S Plaid will be a direct refrigerant to air heat exchanger at the front of the car. This setup is more effective for rejecting high heat loads.


Detailed Discussion

The following is from our heat transfer consultant Keith Ritter.

A detailed list of assumptions we used in the performance model is shown in figure 9


Motor map: 

  • We pretty much nailed the maps for all previous versions. Our performance model assumed a short flat peak HP section after hitting the peak power point, then a linear HP drop to redline.
  • The Plaid motor has some sort of special magic in it, as it almost holds peak HP to redline. I find that hard to believe, as you get increasingly higher gear and windage friction losses, plus the magnetic hysteresis, winding impedance, and harmonics all build at higher RPM/frequencies. That is why EVERY motor has a peak power-point at a specific speed and then HP starts to sag. Perhaps they are over-amping at higher speeds and getting away with it because of better cooling capacity? 

Rotor sleeve:

  • The carbon fiber rotor sleeve is something that is high-end, but not a "breakthrough" for high-performance PM motors. There are several vendors that specialize in making the sleeves and/or rotors. They reduce rotating mass plus increase sleeve strength to keep the magnets in place under the extreme centripetal forces at high RPMs.


IEVs says 18650 cells, based on the fact Tesla did NOT say the refreshed S will have anything else. I think they are right. Musk is not one to go stealth on implementing the 4680. Here is what I think the pack has:

  • 18650 cells. Perhaps with a slightly different higher-power-density chemistry. Tesla needs to keep the pack absolutely as flat as possible. The 4680 is better for the Model Y, Cybertruck, and Semi. Note that the new pack does not have the "double-decker" modules that the P100D has.
  • Note that the 18650 cells also have better inherent thermal capacity to remove heat than the 2170 cells. Their median radius is 15% less, which means the mean effective internal thermal "R" to conduct heat out of the cell core to the can is 15% less.  
  • Additionally, I believe Tesla is using cooling snakes similar to the Model 3. Glued directly to the cells, as opposed to the "fingers" thermal conductive pad. This pad was a thermal bottleneck. I re-ran our pack thermal analysis model for the P100D, substituting the thermal conductance and thickness of the M3 snake and increasing the snake height to 90% of cell height. Also, I changed the snake glycol design to make each snake just single-pass, which reduces the number of cells each snake has to cool. Combined, these changes more than doubled the heat removal capacity. Finally, the net cell spacing will go down, improving pack density and reducing weight. This could explain how Tesla is eliminating two modules
  • This is more speculative, but (channeling my "if I were a Tesla engineer thinking") I bet the new modules are built more like how the new 4680 pack will be built. All cells oriented the same way (not flipped + and - every other cell group). Perhaps indexed into the bottom plate, using buttons or something on the cells. No "bandolier assembly hell."  Also, I bet they did NOT use the wire-weld connections to the cells with either the fingers type M3 current collector or the single punched-plate type current collector of the MS. I bet they used the multi-layered sheet type plus and negative current collector like Munro hypothesized for the 4680 pack. This both improves current handling capacity and assembly simplicity.  

TMS/heat pump:

  • The article says the "radiator" has twice the cooling capacity. First, The "radiator" shown in the image is not a glycol radiator. It is an air-cooled refrigerant condenser. Look closely at piping connecting the condenser to the HVAC equipment and the manifolds along the side of the condenser. One is a small pipe, the other a large one. The large one is the hot gas manifold entering the condenser. The small one is the condensed refrigerant liquid connection leaving. If it were glycol, both manifolds and connecting tube connections would be the same size. It appears they have manifolds on both sides, as I see some dangling tube connections on the other side. For cooling purposes, direct refrigerant condensers have more capacity than an equivalent-sized liquid glycol radiator. The heat transfer per unit area is a lot higher for DX condensing than for liquid forced convection. 
  • Second, twice the capacity means probably instead of 3-4 tons of cooling, it has 6-7 tons of cooling. The Plaid pack thermal model using my assumptions, at 250 kW max charging, calculates out to 6.6 tons of cooling. This also supports having a DX condenser. That is the only way they can fit 6 tons of condensing under the Model S hood.
  • This appears to be an entirely different heat pump design. It appears to incorporate a similar printed circuit type glycol and refrigerant manifold, but turned vertical. I see the refrigerant accumulator + the stack chiller. It appears there are two sets of refrigerant lines feeding the cabin.

Tesla's engineers are on fire these days. This is brilliant work. 

George Bower BSME

Keith Ritter PE

Engineered Compliance 

Energy Analysis

Got a tip for us? Email: