Sponsored

Skysurfer

Active Member
Joined
Aug 7, 2024
Threads
3
Messages
41
Reaction score
79
Location
Oregon
Vehicles
'25 R1S, '18 Model 3, '87 XL600R, '86 IROC-Z


Edmunds just released their video comparing a Gen2 dual motor R1T to a Tacoma TRD Pro. Was surprising to see how poorly the brake torque vectoring was performing on the R1T, causing it to be unable to climb several obstacles. At 15:45 you can see the unloaded wheels spinning up quite quickly as the system is failing to transfer the torque effectively.

This hasn't been my experience with my dual motor R1S, but maybe I just haven't had the same types of technical challenges or could there be something not working correctly with the Edmunds R1T? Any other dual motor R1 owners have similar experiences off road as Edmunds did in this video?
Sponsored

 

CarGuyCarl

Active Member
Joined
Sep 10, 2024
Threads
0
Messages
27
Reaction score
47
Location
Chicago
Vehicles
GX470, S2000
If an old early 2000s Toyota can do this (speaking from experience with my GX) with ATRAC and no lockers, you'd think the dual motor should have no trouble with far easier torque control versus an ICE powerplant.

Still a big head scratcher that they didn't integrate lockers. Seems like a no brainer and super simple, especially on a dual. But then the dual would probably do better off roading than tris and quads.... Oh marketing.
 

R1Thor

Well-Known Member
First Name
Joe
Joined
Aug 9, 2023
Threads
7
Messages
1,179
Reaction score
1,752
Location
Lancaster, PA
Vehicles
23QM R1T, Limestone + Ocean Coast, 21" & UBS
Occupation
Mechanical Engineering Lead
Clubs
 
The dual motor vehicles, for better or worse, are open differential. That's going to be a pretty big limitation. There will be no 'transfer of power' because of this. You need *some* form of differential to transfer force. The idea being that torque vectoring would use the brakes to imply a false torque (traction) back to the 'loose' wheel should transfer power back to the wheel that IS maintaining traction is what torque vectoring does.

However, in practice, there is a huge efficiency loss and it's not super easy to accomplish this in all scenarios. Typically it's helpful for a moment, but technically the only way the system can 'detect' it's regained traction, it has to disable the brake to detect whether or not the wheel is turning the same speed as the wheel with traction. And it's sort of a loop logic at this point: Slipping? - > Lock brake and hope traction engaged wheel moves you out of the spot you're 'stuck in' -> release brake -> detect slippage (or not) -> re-apply torque vectoring if slipping -> re-engage nominal driveline characteristics if both wheels appear to no longer be traction limited (turning at same rate).

If the vehicles had a limited slip differential (Quaiff, DCCD, Torsen, Viscous Coupling, etc), they'd have less difficulties putting power to the traction positive wheel by 'borrowing it' from the traciton limited wheel.

If you're on a surface that requires a LOT of intervention or is going to prove to be difficult to maintain traction, having power to all 4 wheels is advantageous. I'm not going to argue for/against lockers--I'm sure someone's going to bring that up (deadhorse.gif), but the basic idea here is: get power to where it's needed when it's needed. And that's going to be difficult when you have open differentials and a drivetrain dependent upon traction dynamics detection through the system.
 

Dark-Fx

Well-Known Member
First Name
Brian
Joined
Jul 15, 2020
Threads
127
Messages
11,601
Reaction score
22,991
Location
Michigan
Vehicles
R1T, R1S, Livewire One, Fisker Ocean, Sierra EV
Occupation
Engineering
Clubs
 
Highest mode isn't the best for articulation, more meant for instances where you actually need the clearance. I suspect on the dual motor, that you need to manually provide the brake torque for situations where you're just barely moving. If I were trying to get my quad up this rock at such a low speed, it'd probably stall and put me into turtle jail until a reboot/key cycle.
 

zefram47

Well-Known Member
First Name
Aaron
Joined
Feb 6, 2022
Threads
16
Messages
2,410
Reaction score
3,919
Location
Denver, CO
Vehicles
Rivian R1T, Alfa Romeo 4C
Occupation
Software Engineer
If an old early 2000s Toyota can do this (speaking from experience with my GX) with ATRAC and no lockers, you'd think the dual motor should have no trouble with far easier torque control versus an ICE powerplant.

Still a big head scratcher that they didn't integrate lockers. Seems like a no brainer and super simple, especially on a dual. But then the dual would probably do better off roading than tris and quads.... Oh marketing.
Yup...but the Rivian also weighs nearly 2000 lbs more than the Taco and it of course has a rear locker that the R1T doesn't have.
 

Sponsored

R1Tom

Well-Known Member
First Name
Tom
Joined
May 19, 2022
Threads
37
Messages
3,621
Reaction score
4,839
Location
Wisc
Vehicles
Riv R1T
Occupation
Sales
Highest mode isn't the best for articulation, more meant for instances where you actually need the clearance. I suspect on the dual motor, that you need to manually provide the brake torque for situations where you're just barely moving. If I were trying to get my quad up this rock at such a low speed, it'd probably stall and put me into turtle jail until a reboot/key cycle.
Does the Rivian allow manual brake torque vectoring? I haven't wheeled my Rivian so curious. In All Purpose or Sport (know for launch use) it won't provide power if brake is pressed. Maybe the offroad modes do?
 

-0-coach-0-

Well-Known Member
First Name
Greg
Joined
Jan 22, 2024
Threads
24
Messages
84
Reaction score
118
Location
Orange County, CA
Website
x.com
Vehicles
R1T
Occupation
Educator
Clubs
 
Can't Rivian do what Tesla did with the Cybertruck and create a software-enabled locker for better control in off-road situations?

I did think the efficiency comparison was good to see. When they combined the road and off-road efficiency, the two were nearly identical. With that being said, Tacos are notorious for terrible gas mileage.
 

ElGuano

Well-Known Member
Joined
Oct 9, 2024
Threads
13
Messages
371
Reaction score
629
Location
Cali
Vehicles
R1T Trimax - Storm Blue, Driftwood, Sport Dark
Occupation
darkweb peddler
Can't Rivian do what Tesla did with the Cybertruck and create a software-enabled locker for better control in off-road situations?

I did think the efficiency comparison was good to see. When they combined the road and off-road efficiency, the two were nearly identical. With that being said, Tacos are notorious for terrible gas mileage.
The dual motor Cybertruck (one motor to two wheels) has mechanical lockers at each of the front and back axles.

The tri-motor cybertruck has one motor for the front wheels, which also has a mechanical locker. The rear has one motor per wheel, which is how it achieves its virtual rear locker (but the DM Rivian only has one motor, so it can't do it the same way).
 
OP
OP
Skysurfer

Skysurfer

Active Member
Joined
Aug 7, 2024
Threads
3
Messages
41
Reaction score
79
Location
Oregon
Vehicles
'25 R1S, '18 Model 3, '87 XL600R, '86 IROC-Z
The dual motor vehicles, for better or worse, are open differential. That's going to be a pretty big limitation. There will be no 'transfer of power' because of this. You need *some* form of differential to transfer force. The idea being that torque vectoring would use the brakes to imply a false torque (traction) back to the 'loose' wheel should transfer power back to the wheel that IS maintaining traction is what torque vectoring does.

However, in practice, there is a huge efficiency loss and it's not super easy to accomplish this in all scenarios. Typically it's helpful for a moment, but technically the only way the system can 'detect' it's regained traction, it has to disable the brake to detect whether or not the wheel is turning the same speed as the wheel with traction. And it's sort of a loop logic at this point: Slipping? - > Lock brake and hope traction engaged wheel moves you out of the spot you're 'stuck in' -> release brake -> detect slippage (or not) -> re-apply torque vectoring if slipping -> re-engage nominal driveline characteristics if both wheels appear to no longer be traction limited (turning at same rate).

If the vehicles had a limited slip differential (Quaiff, DCCD, Torsen, Viscous Coupling, etc), they'd have less difficulties putting power to the traction positive wheel by 'borrowing it' from the traciton limited wheel.

If you're on a surface that requires a LOT of intervention or is going to prove to be difficult to maintain traction, having power to all 4 wheels is advantageous. I'm not going to argue for/against lockers--I'm sure someone's going to bring that up (deadhorse.gif), but the basic idea here is: get power to where it's needed when it's needed. And that's going to be difficult when you have open differentials and a drivetrain dependent upon traction dynamics detection through the system.
If you look at the situation where the left rear wheel is on the ground with the majority of traction and the right rear is in the air with no traction, applying full braking to the right rear would cause 100% of the motor torque to be seen at the left rear wheel. Only when the brake is slipping would there be some power loss in the form of heat in the slipping brake. This is a common system used fairly effectively for decades in open differential vehicles (see Toyota's ATRAC as a prominent example). This has the advantage that at least 50% of total torque can, in theory, be applied to a given wheel with a dual motor setup versus 25% for a quad motor and the dual motor can allow the motor to start spinning before transferring torque, getting out of the lower stall torque condition. Certainly no replacement for locking differentials, since you can't beat a direct link between the wheels, but should be able to perform better then what was seen here.

That being said, as you pointed out, it is a tricky controls problem determining what brake to apply and how forcefully. Seems the R1T was not doing a great job with the brake biasing in this case. Will be interesting to see how the new quad motors do and if they have better stall torque compared to the previous Bosch units.
 

drichan

Active Member
First Name
Dean
Joined
Aug 30, 2023
Threads
2
Messages
29
Reaction score
40
Location
El Dorado Hills
Vehicles
R1T
Occupation
Software
Clubs
 
Emme Hall is a pro and has great experience with Rivian but I was surprised she didn’t air down. Seems that would have helped with traction in both climbing difficulty examples. I wonder if Rivian will consider adding lockers. At least in the dual motors.
 

Sponsored

evdriver2016

Member
Joined
Dec 8, 2024
Threads
2
Messages
16
Reaction score
20
Location
US
Vehicles
Model X and S going to Rivian R1S and Macan 4
Edmunds is owned by Carmax who needs to sell lots of used ICE trucks.
They are always looking for ways to downplay EV while pumping up ICE.
 
 





Top