Lets work out the Tesla Semi-Truck Technical Specs

[mongo]

While I expect their upgraded tire inflation system to be better than current systems the fact is they already exist on tractor and trailer. The current systems are able to detect leaks and keep the tires inflated without sensors. This is because a semi always keeps air pressure in tanks for friction braking around 120psi. Semi tires are inflated to roughly 95 (application dependent) with a one way valve keeping pressure from those air brake tanks flowing to the tires at all times. We have had them on the majority of our trailers for the past 5 years.

The Semi doesn't need fancy extras to be a game changer. The cuts to maintenance and fuel are more than sufficient on their own especially in slip seat operations where getting equipment into the shop for preventative maintenance is near impossible because the trucks never stop.

Interesting, if I might pick your brain.
How does the tractor know the tires are being inflated? Wouldn't there still be a problem at a road side safety check, or any situation where the tire is leaking and the engine is of (or tractor disconnected)? How does the tire pressure get adjusted downward based on ambient temp and/ or altitude? What if the valve sticks?

 

[omgwtfbyobbq]

Older 18650's were faster, except for the newer 75 strangely.

Tesla claims both the lowest cost (now questionable) and highest energy density. If they go all green and stuff on cobalt reduction, are competitors not even trying?
Again, where is an annual improvement of more than 1-2%? Tesla's greatest range impacting development has been cramming more cells into the 100 pack.

Elon and JB have been speaking at significant density improvements at the chemistry level. These did not make it to Model 3 then, as even the more efficient form factor couldn't make density measurably better.

Do you have a reference for Tesla not making any density improvements? Their EPA application for the Model S 75 and up lists the batteries at more than 500kg, while the application for the 3 lists the battery at less than 500kg, which to me implies the 3 has a more energy dense chemistry per unit weight, especially since Tesla has more ancillary stuff in the 3 pack than they do in the S pack.

https://iaspub.epa.gov/otaqpub/display_file.jsp?docid=39834&flag=1
https://iaspub.epa.gov/otaqpub/display_file.jsp?docid=40001&flag=1

 

[arnis]

Not so sure... comment at 9.02s.

>> Current SC's are not limited to 330A, it's the plug/socket connection that can't safely transfer more.

So... the boilerplate on the SC is BS?

>> Model 3 has the same socket, therefore it will be unable to receive more.

Internal cables are thicker, fish it out on the you tube.

>> 525A came out of nowhere.

EPA official filling is "nowhere"?

Fudsters should try harder to not end up looking so stupid...

Half a year later and my expectations appear to be correct. Even that 525A current EPA magically made up.

So somebody did something real and actually has data to prove that Model 3 with best battery (LR) will hardly ever charge
faster than what SC can offer today. 120kW charge rate is near maximum no matter what charger Model 3 will charge.
Be it even European Ionity 350kW charger.
http://www.roperld.com/Science/TM3LR_SCChargingCurves.pdf
Curve clearly shows that tapering happens few minutes after reaching maximum. And in another source (video) same results are clearly seen. Up to 10 minutes hovering at maximum 116kW. Keeping in mind within those 10 minutes pack voltage must go up and amps must go down. Therefore if Model 3 is charged at "unlimited" charger it will still not be able to pass 116kW for any meaningful period.

So, how much extra can be gained within 30 minutes of charging?
Made a tweak to screenshots to visually see the "surface area" of current SC and theoretical maximum (for example SC v3):

And in case of any longer charging session, gains are diluted even more.
Oversized cables help with efficiency and heat buildup.
PS: Model 3 gets different socket for the rest of the world (not US nor Canada). 330A limit will likely stay exactly the same.

I'm not aFraid nor am I excessively Uncertain; highly Doubt I shall be called FUDster therefore I say. #Yoda.

 

[mongo]

Curve clearly shows that tapering happens few minutes after reaching maximum. And in another source (video) same results are clearly seen. Up to 10 minutes hovering at maximum 116kW. Keeping in mind within those 10 minutes pack voltage must go up and amps must go down. Therefore if Model 3 is charged at "unlimited" charger it will still not be able to pass 116kW for any meaningful period.

Due to the pilot signal, of course an unlimited charger would not supply more power to the car, if the initial throttling is done by the vehicle. If that is what the vehicle is requesting, the charge curve is likely tuned for conservative pack life and Tesla SCs, that does not mean the pack is maxed out at 116 (120kW).

Potential ambiguity could be eliminated by watching the pilot signal traffic (and pack V and T) to see what the vehicle is requesting. That would isolate power loss in charging cable (116 vs 120kW), any SC based throttling (only a factor at the high power end), pack temperature limitations, and limitations due to pack vs SC voltage headroom.

 

[arnis]

So when is "pack maxed out"? Well.. on most EV's, including Tesla, charge curve is almost always limited by the highest cellstack reading.

So, for example, Leaf has a limitation of 4.14V. Tesla has 4.20V. And looking at those two graphs, when charging power starts to drop.. one of the cellstacks reached 4.20V on Model3.
If we apply more current, lets say 400A.. one of the cellstacks will reach 4.2 either immediately or within a minute or two, making the whole over-dimensioning of all components excessively pointless (waste of money, weight etc).
This all boils down to "LR model 3 will not charge any faster on DC charger that is more capable" as it is fundamentally "maxed out".
There will likely be another pack made in 3-5 years that will likely have extra 10-20% capacity and might have more capable chemistry.
That doesn't mean vehicle will charge at faster rate in those few early minutes. It might, though it will have less impact. Rather the fact that overall charging curve will stay at around 120kW for MUCH MUCH longer.
Many other EV's have a limit way lower than what battery can handle (and have a flat charging curve from 5% up to 50 or even 80%).
If I had to find a sweet spot, I would consider ideal curve to be pretty much something that it is right now (red line):

Full consistent speed between 10-50% in ideal conditions.
This keeps some safety margin for components other than just cells and hardly influence minutes to stay at SC as it is
impossible to arrive to SC at exactly sweet spot (10%).

 

[mongo]

So when is "pack maxed out"? Well.. on most EV's, including Tesla, charge curve is almost always limited by the highest cellstack reading.

There are two charge limits. The first (at lower SOC) is maximum charge current (power = Max_I * pack_V). The second is the maximum cell (or pack if well balanced) charge voltage (power = charge_I * max_pack_V).

Depending on the pack and charger, you can also have a third limit ((charger voltage) - (pack voltage)) / (pack equivalent resistance). Along with charge max current and power limits.

While possibly not the case, a charger that can put out 120kW, but is undersized in terms of max voltage would also create a tapering charge curve.

In general, a flat to start 120kW limited curve is harder on the pack at the lower voltage (SOC) region due to higher charge current. If the pack can handle 120kW at a really low SOC, then it can take more than 120kW at higher voltages (while staying within cell max voltage limit, of course). However, it makes sense to have a pack with a higher rate than the charger so the entire flat portion is well within the pack's capabilities.

 

[J1mbo]

So somebody did something real and actually has data to prove that Model 3 with best battery (LR) will hardly ever charge
faster than what SC can offer today. 120kW charge rate is near maximum no matter what charger Model 3 will charge.
Be it even European Ionity 350kW charger.
http://www.roperld.com/Science/TM3LR_SCChargingCurves.pdf
Curve clearly shows that tapering happens few minutes after reaching maximum. And in another source (video) same results are clearly seen. Up to 10 minutes hovering at maximum 116kW. Keeping in mind within those 10 minutes pack voltage must go up and amps must go down. Therefore if Model 3 is charged at "unlimited" charger it will still not be able to pass 116kW for any meaningful period.

Nobody outside of Tesla has charged a CCS-enabled M3: this analysis is purely based on US Supercharger performance (plus, I assume a pre-CCS version of the firmware).

Not really appropriate to mention the 350kw Ionity charger here.

 

[transpondster]

There are two charge limits. The first (at lower SOC) is maximum charge current (power = Max_I * pack_V). The second is the maximum cell (or pack if well balanced) charge voltage (power = charge_I * max_pack_V).

Depending on the pack and charger, you can also have a third limit ((charger voltage) - (pack voltage)) / (pack equivalent resistance). Along with charge max current and power limits.

While possibly not the case, a charger that can put out 120kW, but is undersized in terms of max voltage would also create a tapering charge curve.

In general, a flat to start 120kW limited curve is harder on the pack at the lower voltage (SOC) region due to higher charge current. If the pack can handle 120kW at a really low SOC, then it can take more than 120kW at higher voltages (while staying within cell max voltage limit, of course). However, it makes sense to have a pack with a higher rate than the charger so the entire flat portion is well within the pack's capabilities.

Tesla don't do simple CC-CV, they reduces current before reaching full voltage with some slope

 

[mongo]

Tesla don't do simple CC-CV, they reduces current before reaching full voltage with some slope

Which may be a power limit, the point I was responding to was that max individual cell voltage is not necessarily the limiting factor.

 

[arnis]

It appears we all agree that individual cell voltage doesn't have to be the limiting factor at every moment.

I assume a pre-CCS version of the firmware

Can't change individual cell voltage limitation. Plug has almost never been a limiting factor. European M3 will have the
same chemistry. Handshake will be different and communication protocol, but not "terms agreed"
So post-CCS firmware in EU and US will not change maximum agreed charging rate.

 

[J1mbo]

It appears we all agree that individual cell voltage doesn't have to be the limiting factor at every moment.


Can't change individual cell voltage limitation. Plug has almost never been a limiting factor. European M3 will have the
same chemistry. Handshake will be different and communication protocol, but not "terms agreed"
So post-CCS firmware in EU and US will not change maximum agreed charging rate.

You are assuming the "cell voltage limitation" matches the US Supercharger max rate. Personally, I'll wait and see.

FWIW, I am pretty sure that the hard-core cooling used by "Track mode" was not built into the system just for track mode.

 

[arnis]

You are assuming the "cell voltage limitation" matches the US Supercharger max rate. Personally, I'll wait and see.

FWIW, I am pretty sure that the hard-core cooling used by "Track mode" was not built into the system just for track mode.

Track mode changes cooling logic, doesn't change capability. Model S/X-s cooling capacity can be easily maxed out at SC.
Problem is not cooling (battery will not overheat when it goes from 10% to 30% where somewhat faster charging is plausible).
Problem is internal resistance. 4.2V limit can be reached even at 20% SOC. Especially at slightly colder pack (like 104F).

 

[Tobyw]

As far as I understood, tractor has 4 Model 3 permanent magnet motors with no differentials, just reduction gears.
And ratio will definitely be way different. Truck max nominal speed will likely be 70mph.
This allows very different ratio compared to Model 3.
Also different reduction ratios between second and third axle for efficiency optimisation at different speeds.

Onboard vehicle, I believe drivetrain will run at the same voltage. 800V motors/inverters are hardly any better.
Nominally, each motor will be under comparable load while truck is cruising at 60mph compared to car
cruising at 60mph. Diesel cars have fuel consumption 5-7l/100km, trucks 30-40l/100km. Divide that load to 4 motors and
we get comparable nominal loads (especially with drag coefficient 0.36).
But while charging, 800V arrangement will help with charging current/conductors. AFAIK, It is possible to use 400V
electronics to charge 800V battery. Positive pin at +400V DC, negative not at 0V DC, but rather -400V DC.

Musk mentioned, that truck's drivetrain is redundant. Therefore I believe one pack failure (4 of them total likely, graphics
from event) will not result in breakdown on the side of the road. Same with motor. It's reasonable to believe that
odd packs serve left motors, even packs serve right motors. Or something similar.

BAMF will likely not have Model 3 battery packs. But it should be reasonable to use Model 3 pack modules.
Vehicle gross weight matters very little in terms of range. If fully loaded BAMF can do 500 miles, then with the same trailer
completely empty it might do 550 miles.

Megachargers we have been shown are actually just stalls. Charger is nearby. That same charger will likely be
able to charge passenger vehicles as well. Just not at the same stall. Trucks will have drive-through stalls with different plug(s).
Truck stalls will have priority over car stalls. Similarly like first Model S arrived per stall pair has priority over the second one today.
They will need to have powerpacks as load from the grid during peak time is not acceptable. Solar will just be there for statement,
but it will not be sufficient. I would remind the tweet when Elon hinted about 350kW as "child's toy". This is what he was talking about.
Like I said, cars could be charged at megacharger sites. Just not at the same bays. There is no reason one megacharger couldn't
serve 4 cars simultaneously.

I would like to get some estimations as well. How much energy is required for 0.36 Cd truck to drive 400 miles on flat road?

Drag is calculated by multiplying frontal area by drag coefficient. So the drag and fuel consumption should be proportional to the relative drag coefficients and energy consumption of each.

 

[Jeff N]

This is the first time I’ve come across a photo of the Semi charging connector (I’ve previously seen photos of the truck inlet).

I just posted the article linked below:

https://electricrevs.com/2019/02/15/how-to-supercharge-a-tesla-semi/

Several Tesla Semi truck prototypes have been seen during the past year on public highways making long-distance trips around the United States but the actual process of charging the Semi at a Supercharger site has been shrouded in a bit of mystery....

....Tesla employees have commonly asked onlookers not to photograph the adapter arrangement that allows the truck to charge at today’s existing Superchargers....

 

[mongo]

The motors are in front and behind the axle and there is a big differential/transmission in between them. Nothing eliminated there.

View attachment 407591

There is no differential nor transmission there. Each drive unit (likely) contains the end bearing block and bull drive gear of the half axle shaft. The size of the former pumpkin area is to allow the two units to fit.

 

[mongo]

1500 lbs is reasonable for 4 motors, gears, inverter, cooling system and charging cables.

The engine is not that heavy compared to the 80000lb weight of the total vehicle so the frame is not going to get lighter, specially that the frame still has to carry the batteries.

If you are including all of that too, I can see it being closer.

It is not a issue of the engine weigh, it is an issue of reacting all the torque from the engine/ trans combination to the motor mounts. Watch a truck twist on acceleration from a stop.

It's not a differential, it's a single gear transmission on both wheels. I just call it diff/tranny because I'm talking about weight. These single gears should have similar weights than the diesel's differential.

Maybe for the part, but you are removing the pinion, planitaries, and drive shafts. Just two modified ring gears with bearing.

 

[Silent Ludicrosy]

Tesla Semi won’t twist like this

 

[Remus]

1500 lbs is reasonable for 4 motors, gears, inverter, cooling system and charging cables.

The engine is not that heavy compared to the 80000lb weight of the total vehicle so the frame is not going to get lighter, specially that the frame still has to carry the batteries.



It's not a differential, it's a single gear transmission on both wheels. I just call it diff/tranny because I'm talking about weight. These single gears should have similar weights than the diesel's differential.

This is too much of hand waving and guessing going on.

How do you know it's reasonable? How do you compare the cooling system between electric drive train and diesel ones when the efficiency are drastically different? I am also guessing because that's what you came out first based on the model s drive unit weight and you must stick to it?

80000lb is the maximum allowed weight including the trailer and full load. And @mongo was talking about the diesel engine, being the single source of power to pull all that weight, has to twist itself against the output axle, which put tremendous stress on the engine mounts and frame, that's just Newton's third law!. You don't have that single source of massive twisting in the Tesla semi when forces are distributed and very closed to the wheel. Much less loss on the transmission and differential too.

Edit: in fact, you can have two motors on the same side of the tractor spin at opposite direction to cancel out some of the twisting, although it may not be needed at all.

All in all, considering the evidences you provided, you showed hefty confidence in your post.

 

[acoste]

This is too much of hand waving and guessing going on.

How do you know it's reasonable? How do you compare the cooling system between electric drive train and diesel ones when the efficiency are drastically different? I am also guessing because that's what you came out first based on the model s drive unit weight and you must stick to it?

80000lb is the maximum allowed weight including the trailer and full load. And @mongo was talking about the diesel engine, being the single source of power to pull all that weight, has to twist itself against the output axle, which put tremendous stress on the engine mounts and frame, that's just Newton's third law!. You don't have that single source of massive twisting in the Tesla semi when forces are distributed and very closed to the wheel. Much less loss on the transmission and differential too.

Edit: in fact, you can have two motors on the same side of the tractor spin at opposite direction to cancel out some of the twisting, although it may not be needed at all.

All in all, considering the evidences you provided, you showed hefty confidence in your post.

I feel my numbers are reasonable, but I have no evidence.
The motor will be larger than the Model 3 motor and the gears need to be sturdier. The reason is that the Model 3 motors are used only for short periods at high performance but these truck motors produce high power for extended time. To reach 1M miles these need to be larger. (I don't think Model 3 drive train is designed for 1M miles, that is just marketing).
Also the cables. This truck will have massive charging cables, that's heavy stuff. Cooling system is light I assume.

Yes, 80000 is full weight but the frame has to pull 60000lbs and keep it straight (torsional forces) so I don't see weight reduction there. Tesla however might use lightweight materials. And they also might leave out some functionality as mentioned. I have the impression tha their truck is shorter.

One data I would say is pessimistic in my calculation is the 7x energy consumption compared to a Model S sized car. 7x is based on an average 18 wheeler. Assuming Tesla will improve their efficiency (air drag for example) this number may be around 6.

 

[acoste]

The rear drive axle with the gears looks massive. (at 5:00)