dhanson865
Well-Known Member
Because FSD won't happen on the Semi for several years after they sell the first hundred thousand. No reason to discuss it now.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Because FSD won't happen on the Semi for several years after they sell the first hundred thousand. No reason to discuss it now.
dhanson865
Well-Known Member
I was assuming two pack capacity options.
I'm guessing the two trucks were just to show the adaptive bezels to cover the gaps.
dhanson865
Well-Known Member
dhanson865
Well-Known Member
He asked for the official video which is at Tesla Semi Unveil
The one you linked is not an official Tesla video.
To be fair, I edited my post as my driving isn't always typical. 6-6.5 isn't unheard of when grossed out. In the truck market aero is still huge, I remember when freightliner came out with their new design about 10 years ago and bragged about a 2-3% improvement in aerodynamics. The problem is you need radiators and an engine that can only go so many places so you're limited to the config.
Back to your numbers I'm not sure where you got 9kwh/gal of diesel. A kWh is 3,413 BTUs, a gallon of #2 Diesel is 129,500 BTUs, meaning diesel is 37.9kWh/gal. Lets call it 6mpg for normal driving that's 6.32 kWh/mile on diesel. @2.50/gal thats 6.58c per kwh(diesel) and 41.6c per mile(diesel)
If the Tesla really gets 2kwh/mile @ 7c per kwh that would be 14c/mile (electric) saving 27.6c/mile in fuel costs. Oil capacity on that ISX15 I mentioned earlier is 14 gallons, gallons. Changed every 30-50K miles.
Thinking about this issue more, the truck will probably have a large resistance heating unit in the cooling system. They can run that against the cooling system. 20kW would seem achivable without too much effort. The truck will have a decent sized radiator too.
The truck should almost never be at the top of a pass with an almost full battery. But AFAIK Tesla has to design the Truck so that regen always works.
6 kWh net per mile at 60 MPH is 360 KW to dissipate during the descent. If the pack is full, they could use externally mounted / louvered resistance elements like a diesel electric locomotive. Place behind the cab and adjust the wings ?
They are also going to have to figure out a way to get max regen braking at all times, regardless of battery charge level. you can't have a loaded semi with half regen going down the grapevine. Trucks use a combination of gearing, engine braking, and service brakes to control loads, with no engine or transmission brakes on an electric they can't rely on service brakes during times of high battery charge.
How do convoys compete with rail?
Elon gave three comparison on total cost per mile assuming 80k lbs at 60mpg, $2.50/gal diesel and 7c/kWh.
Diesel $1.51/mile
EV solo $1.26/mile
3 EV convoy $0.85/mile
Elon also claimed that this convoy cost per mile beat trains. Perhaps someone can challenge this.
Even at 2kWh/mile, energy makes up a very small portion of total EV cost, no more than $0.14/mile. Clearly energy efficiency comes nowhere close to explaining how a convoy can be competitive with rail.
I think the savings in a convoy come from autopilot elimination of 2 out of 3 drivers. The delta between EV solo and convoy is $0.41/mile. Some of this is energy savings, but a 20% savings would only account for about $0.03/mile. So let's say the remaining $0.38/mile is due to a 2/3 reduction in labor. This implies a labor cost of $0.57/mile or about $35/hour. I know other sources cite a much lower cost of labor around $0.36/mile ($22/hour in 60mph traffic). This seems a bit low to me. Perhaps it reflects wages only and not employee benefits, taxes and other overhead. Also wages can vary regionally. Regardless labor seems to be a big source of savings.
Am I overlooking other ways a convoy can save money besides labor and fuel?
Perhaps we can imagine ways convoys can create logistical advantages. For example, imagine convoys leaving the Fremont factory loaded with Next Gen Roadsters heads to eager customers along the east coast. The massive convoy stays together along I-80 till it gets to the Megacharger in Illinois. At this point a few drivers meet up with the convoy splitting it up into smaller convoys. One heads to the southeast while the other heads further east along I-80. Think of Megachargers as hubs where drivers switch trucks from one convoy to another. In fact a driver only needs to drive back and forth between two hubs. At each hub, trucks reassemble into new convoys headed to the next hub in its journey. Even if a driver wanted to go the whole journey with a single load, they could hook up with a convoy along the way and sleep in transit allowing some other driver to pilot the convoy between hubs. So the combination of Megachargers and convoys in an exclusive Tesla ecosystem creates some very interesting logistical options.
Elon gave three comparison on total cost per mile assuming 80k lbs at 60mpg, $2.50/gal diesel and 7c/kWh.
Diesel $1.51/mile
EV solo $1.26/mile
3 EV convoy $0.85/mile
Elon also claimed that this convoy cost per mile beat trains. Perhaps someone can challenge this.
Even at 2kWh/mile, energy makes up a very small portion of total EV cost, no more than $0.14/mile. Clearly energy efficiency comes nowhere close to explaining how a convoy can be competitive with rail.
I think the savings in a convoy come from autopilot elimination of 2 out of 3 drivers. The delta between EV solo and convoy is $0.41/mile. Some of this is energy savings, but a 20% savings would only account for about $0.03/mile. So let's say the remaining $0.38/mile is due to a 2/3 reduction in labor. This implies a labor cost of $0.57/mile or about $35/hour. I know other sources cite a much lower cost of labor around $0.36/mile ($22/hour in 60mph traffic). This seems a bit low to me. Perhaps it reflects wages only and not employee benefits, taxes and other overhead. Also wages can vary regionally. Regardless labor seems to be a big source of savings.
Am I overlooking other ways a convoy can save money besides labor and fuel?
Perhaps we can imagine ways convoys can create logistical advantages. For example, imagine convoys leaving the Fremont factory loaded with Next Gen Roadsters heads to eager customers along the east coast. The massive convoy stays together along I-80 till it gets to the Megacharger in Illinois. At this point a few drivers meet up with the convoy splitting it up into smaller convoys. One heads to the southeast while the other heads further east along I-80. Think of Megachargers as hubs where drivers switch trucks from one convoy to another. In fact a driver only needs to drive back and forth between two hubs. At each hub, trucks reassemble into new convoys headed to the next hub in its journey. Even if a driver wanted to go the whole journey with a single load, they could hook up with a convoy along the way and sleep in transit allowing some other driver to pilot the convoy between hubs. So the combination of Megachargers and convoys in an exclusive Tesla ecosystem creates some very interesting logistical options.
Ugh, he mentioned trains?
OK, so the problem with doing that is trains are measured as an efficiency of tons of freight moved one mile on one gallon of fuel. so even though a locomotive may suck down 10 gallons of fuel a mile they're moving 1,000 tons of cargo. Same with ships and their 'prime mover' engines they burn upwards of 350 tons of fuel per day but when you're moving 200,000tons who cares. Anyway, trains average 475 ton-miles per gallon. A truck gets about 120-150 ton-miles per gallon. So it really, really depends on how they calculated it. miles per gallon, ton-miles, et al.
OK, so the problem with doing that is trains are measured as an efficiency of tons of freight moved one mile on one gallon of fuel. so even though a locomotive may suck down 10 gallons of fuel a mile they're moving 1,000 tons of cargo. Same with ships and their 'prime mover' engines they burn upwards of 350 tons of fuel per day but when you're moving 200,000tons who cares. Anyway, trains average 475 ton-miles per gallon. A truck gets about 120-150 ton-miles per gallon. So it really, really depends on how they calculated it. miles per gallon, ton-miles, et al.
Although I have said I am skeptical about the potential for Tesla's success with the semi, I am impressed with the claimed performance, and I hope they can make the numbers work. more importantly, I hope that the truck manufacturers all look to electric propulsion. Truck operators will mostly care about the bottom line, of course. But we need to cut Diesel emissions for obvious reasons.
By coincidence, there is a large truck here at my condo today to deliver some new HVAC equipment for one of the buildings. It has been here for 6 hours so far, and the engine has been idling all that time! This is common practice despite the laws against it (legal limit in MA is 15 minutes, I think). All that wasted fuel and all that unnecessary air pollution! Using electric vehicles will help to reduce that type of totally unnecessary pollution, I hope....
By coincidence, there is a large truck here at my condo today to deliver some new HVAC equipment for one of the buildings. It has been here for 6 hours so far, and the engine has been idling all that time! This is common practice despite the laws against it (legal limit in MA is 15 minutes, I think). All that wasted fuel and all that unnecessary air pollution! Using electric vehicles will help to reduce that type of totally unnecessary pollution, I hope....
kingjamez
Member
This is actually fairly easy. You only charge to 70% of the battery. There are a number of reasons why this makes sense on a Semi with a 1,000,000 warranty. It solves the regen problem, prolongs the battery life, reduces the perceived degradation, and reduces the perceived charging time.
-Jim
SOC limits on high altitude charging stations, seems like a great option. Not a lot of plateaus to deal with.
I suppose it is possible for an unstabilized tractor to somehow get turned around while the trailer itself stays straight and aggravates it. I don’t think that is what Musk was suggesting he has solved. And all tractors today have ESC (electronic stability control) so this is not likely without an external trigger (an external lateral force applied to the front of the tractor, such as being hit by another truck). When a trailer jackknifes, it doesn’t necessarily stop at 90 degrees. It can keep right on going and penetrate the cab. It’s a big, dangerous deal. I’ve had the privilege of testing jackknife scenarios on the track (as the engineering passenger, I leave the test driving to the pros!). Trailers carry a tremendous amount of force, and can yank the tractor 180 degrees off kilter, or roll it if the trailer rolls (the king pin is mighty strong, and can lift the tractor off the ground when the trailer rolls). In our case, we had outriggers and jackknife chains to keep the rig from killing us!
Without sensing equipment in the trailer, I don’t see them implementing some magical new anti-jackknifing tech in the tractor. I could be wrong, but I just don’t see it unless they are supplying some new trailer tech. But that would be overkill anyway, because all that is necessary is simple ABS units on the trailer and dolly axles (which is becoming common today).
In practice, tractors pull many trailers. In fact, in a single day, a tractor may very well pull at least two different trailers. One can’t assume that a trailer will have special sensors installed. And installing sensors on the kingpin would be a significant mechanical challenge. Perhaps they’ll try to do something on the fifth wheel, but still the mechanical challenges are great. Even with that, I’m not sure the tractor could keep a trailer from jackknifing by its own maneuvers, at least assuming the tractor is constrained to it’s own lane.
His whole thing regarding jackknifing was just noise, as far as I can tell. But who knows, maybe Jerome has sprinkled in some powerful magic...
Wow, that had to be some crazy testing! I'd used 90 dress for brevity, it'll turn till it runs out of inertia.
Agree with pretty much everything. What I was suggesting doesn't require trailer mods. They must already have some sensing of trailer position for the side wings/ skirt adjustment. Trailer loading could be done by a set of strain gauges mounted at appropriate points on the 5th wheel (I'd incorrectly referenced kingpin).
Not sure what they are doing either, just spit balling what might be possible.
dhanson865
Well-Known Member
So it's like 4 charging cables in one port (2 pins per pack, 4 packs)
I think the fairing adjustment is done based on steering wheel position, not trailer sensing. But pulling the trailer in close and fairing it close is a great design. I like that part a lot.
Re: strain gauges - I’m not certain you would get sufficient usable data. There would be a lot of noise (the trailer yanks and jerks just based on road geometry, potholes and wind forces). I don’t know that you could get a detectable signal indicative of a tire or alignment problem from such a sensor, because I don’t think such signals would rise above the noise floor. I think more useful data will actually come from the rear view cameras.
Re: crazy testing. Definitely! And they made me wear a motorcycle helmet, for reasons I can’t fathom. A truck cab is so big my head couldn’t reach anything hard, and if we actually rolled or had a full trailer intrusion jackknife, I’d be crushed to putty (but maybe my head would have stayed pretty for the funeral...).
So it looks like they can split the overall combined pack into 4 parallel smaller, parallel packs to charge them at the same time (effectively 4x the charge rate versus leaving it as one giant pack). This is about what I expected, though the resulting charge connector is even more svelte than I would have guessed.
I imagine it's something like this in normal operation:
positive bus
-------------------------------
| | | | Charge connector
| | | | ---- pack 1 +
| | | | ---- pack 2 +
| | | | ---- pack 3 +
| | | | ---- pack 4 +
| | | |
contactor contactor contactor contactor
| | | |
pack1 pack2 pack3 pack4
| | | |
contactor contactor contactor contactor
| | | |
| | | | ---- pack 4 -
| | | | ---- pack 3 -
| | | | ---- pack 2 -
| | | | ---- pack 1 -
| | | |
-------------------------------
negative bus
And like this when charging:
positive bus
-------------------------------
Charge connector
/------------------------------------ pack 1 +
| /-------------------------- pack 2 +
| | /---------------- pack 3 +
| | | /------ pack 4 +
| | | |
contactor contactor contactor contactor
| | | |
pack1 pack2 pack3 pack4
| | | |
contactor contactor contactor contactor
| | | |
| | | \------ pack 4 -
| | \---------------- pack 3 -
| \-------------------------- pack 2 -
\------------------------------------ pack 1 -
-------------------------------
negative bus
Just need some heavy duty SPDT contactors to switch between connecting a given pack to the power bus or the charging interface. This way they could even keep everything at 400V nominal and share all their power electronics designs between platforms.
edit: well that's annoying, it lets me use courier fixed with fixed spacing in the editor but not when displaying on the page?
I imagine it's something like this in normal operation:
positive bus
-------------------------------
| | | | Charge connector
| | | | ---- pack 1 +
| | | | ---- pack 2 +
| | | | ---- pack 3 +
| | | | ---- pack 4 +
| | | |
contactor contactor contactor contactor
| | | |
pack1 pack2 pack3 pack4
| | | |
contactor contactor contactor contactor
| | | |
| | | | ---- pack 4 -
| | | | ---- pack 3 -
| | | | ---- pack 2 -
| | | | ---- pack 1 -
| | | |
-------------------------------
negative bus
And like this when charging:
positive bus
-------------------------------
Charge connector
/------------------------------------ pack 1 +
| /-------------------------- pack 2 +
| | /---------------- pack 3 +
| | | /------ pack 4 +
| | | |
contactor contactor contactor contactor
| | | |
pack1 pack2 pack3 pack4
| | | |
contactor contactor contactor contactor
| | | |
| | | \------ pack 4 -
| | \---------------- pack 3 -
| \-------------------------- pack 2 -
\------------------------------------ pack 1 -
-------------------------------
negative bus
Just need some heavy duty SPDT contactors to switch between connecting a given pack to the power bus or the charging interface. This way they could even keep everything at 400V nominal and share all their power electronics designs between platforms.
edit: well that's annoying, it lets me use courier fixed with fixed spacing in the editor but not when displaying on the page?
I agree. It looks like they can do all this with existing tech. It is reasonable to expect a truck driver to be able to handle a heavy cable. Two cables would be OK too.
Similar threads
