Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Semi gear reduction ratios

This site may earn commission on affiliate links.

scaesare

Well-Known Member
Mar 14, 2013
10,780
25,103
NoVA
I haven't seen this discussed yet in several of the treads musing about the Semi design.

In KMan's video of the Semi drivetrain components I note that the axles have differing gear reduction ratios. The fore axle is 23:1 reduction, and the aft axle is 15:1.

I assume this is done to optimize the power delivery and efficiency through the RPM range the motors are capable of, seeing as there's isn't a multi-gear transmission.

This would seem similar to the dual-motor Model S, which has a ~9.73:1 rear reduction ratio (at least for the "large drive units), and a slightly "taller" gearing for the front motor (~9.34:1). This allows the car to cruise primarily with the front motor operating at a lower RPM range closer to it's efficiency sweet spot.

With the semi having a ~53% difference between the axles, I assume that's for the low-end torque it takes to get 80K lbs moving, and the system will be biased toward the taller geared axle at highway speeds.

Any other thoughts or implications about this design?
 
I haven't seen this discussed yet in several of the treads musing about the Semi design.

In KMan's video of the Semi drivetrain components I note that the axles have differing gear reduction ratios. The fore axle is 23:1 reduction, and the aft axle is 15:1.

I assume this is done to optimize the power delivery and efficiency through the RPM range the motors are capable of, seeing as there's isn't a multi-gear transmission.

This would seem similar to the dual-motor Model S, which has a ~9.73:1 rear reduction ratio (at least for the "large drive units), and a slightly "taller" gearing for the front motor (~9.34:1). This allows the car to cruise primarily with the front motor operating at a lower RPM range closer to it's efficiency sweet spot.

With the semi having a ~53% difference between the axles, I assume that's for the low-end torque it takes to get 80K lbs moving, and the system will be biased toward the taller geared axle at highway speeds.

Any other thoughts or implications about this design?

Mentioned briefly in the Semi Tech Spec thread:

Reportedly, fore rear axle 15:1, aft rear axle 23:1.

The lower gear ratio (15:1) allows more motor torque at a given road speed (lower motor RPM). The higher ratio (23:1) allows more wheel torque at a given road speed (greater motor torque multiplication).

At higher speeds, back EMF limits torque, so it could be a split to hit both ends of the performance spectrum. If the PM efficiency is fairly flat for RPM/Power, then all 4 motor can share the normal driving load to reduce losses due to current.
 
Mentioned briefly in the Semi Tech Spec thread:



The lower gear ratio (15:1) allows more motor torque at a given road speed (lower motor RPM). The higher ratio (23:1) allows more wheel torque at a given road speed (greater motor torque multiplication).

At higher speeds, back EMF limits torque, so it could be a split to hit both ends of the performance spectrum. If the PM efficiency is fairly flat for RPM/Power, then all 4 motor can share the normal driving load to reduce losses due to current.

Yup... one axle will provide more low-end torque, the other better high-rpm performance. The degree of difference between the two is pretty significant, however... it will be interesting to see how that performs.

Are we assuming that's the ONLY reduction in the drivetrain? Or is there a second reduction in the axle housing? Note that the gear boxes with the ratio labels are what the motors are mounted to, and they in turn is are attached to the axles.

If so, then it's interesting to note is the axle ratio vs tire size. A Model 3 outer tire diameter is ~26". From what I can find, a Super Single truck tire diameter is typically about 42". That's about a 62% difference. A similar factor applied to the 15:1 ratio leads to a ~9.3:1 result which is probably in the ballpark for what the Model 3 final drive ratio is. That leads me to believe the 15:1 ration axle motors will operate in a similar efficiency zone as they would on an Model 3.

The other pair of motors making 500+HP and geared at 23:1... it appears that typical semi's can have effective ratios of ~46:1 in low gear... but then again they aren't many with 1000+ HP either...

(it appears you have the axle ratios reversed. As per the video I linked, it's fore axle at 23:1 and aft is 15:1.)
 
Are we assuming that's the ONLY reduction in the drivetrain? Or is there a second reduction in the axle housing? Note that the gear boxes with the ratio labels are what the motors are mounted to, and they in turn is are attached to the axles.

I think that is total reduction ratio. From the Twitter video, it looks like the drive axles are inserted in the drive units after DUs are mounted to the assembly. Similar to standard differential carriers.

The other pair of motors making 500+HP and geared at 23:1... it appears that typical semi's can have effective ratios of ~46:1 in low gear... but then again they aren't many with 1000+ HP either...

Yeah, I'm still iffy on the tug of war spec v.s. a geared down tractor. Given the electric motor's ability to produce maximum torque at 0 RPM, 2 at 23 and 2 at 15 should be plenty good enough. Good point on tire diameter!

(it appears you have the axle ratios reversed. As per the video I linked, it's fore axle at 23:1 and aft is 15:1.)

Huh, I pulled that from the KMan after reveal report (linked in linked post), and thought I got it right. At least ratios were correct.
 
  • Like
Reactions: scaesare
The casing looks like it might have two stages of reduction. Perhaps it has stock 3 gears as a first stage and then the extra reduction after?

I'm not quote sure I'm following. I presume you are talking about the gear reduction casings with the Tesla labels on them?

They are interesting shape. By "stock 3 gears" ae you referring to a planetary-type reduction setup, or a typical 3.x single gear reduction ratio?
 
Ah cool, thanks.. hadn't remembered that.

So at highway speeds, it's possible that a pair of Model 3 motors operating at similar overall effective wheel torque could be primarily responsible for moving 80K lbs down the road.

Flat at 60 MPH 80k with low resistance tires would be around 100kW continuous draw. Split 4 ways, 25 kW per motor or 34.5 HP each. 350 V nominal pack would be around 72A each.


I'm not quote sure I'm following. I presume you are talking about the gear reduction casings with the Tesla labels on them?

They are interesting shape. By "stock 3 gears" ae you referring to a planetary-type reduction setup, or a typical 3.x single gear reduction ratio?

Yeah, the gear reduction drive unit things. From the side profile of the rendered video, it looked like there could be an intermediate shaft besides the motor and drive axle. Doing a full 23:1 step down in one stage seems (totally unfounded) like it would reduce torque capability. So I was throwing out there the idea that the drive unit starts with the same motor and gear as the 3, but that gear then feeds another reduction stage which meshes with the final gear (with internal splines for the axle shaft). The intermediate input gear could also be the same as the 3, although changing it would allow two lesser jumps in ratio say 4:1 and 5.75:1 v.s. 9:1 and 2.6:1. Custom gears probably make the most sense...

This two stage approach would also assist with the two ratios. The motor gear and final gear could stay the same, and only the intermediate shaft and its gears change, along with the mount points. Or only the output of the intermediate changes. Lots of options!

I'm guessing they don't use a planetary since things are not concentric/ colinear and they are running a fixed ratio.

All wild guessing on my part!:D
 
  • Like
Reactions: Zoomit
mongo said:
Flat at 60 MPH 80k with low resistance tires would be around 100kW continuous draw. Split 4 ways, 25 kW per motor or 34.5 HP each. 350 V nominal pack would be around 72A each

But assuming the bias is is similar to the dual motor sedans, the motors in the "RPM sweet spot" (i.e. the 15:1 set) are likely to provide the majority of the power. So could we be looking at those motors running at 30-40KW continuous during highway cruising?

mongo said:
Yeah, the gear reduction drive unit things. From the side profile of the rendered video, it looked like there could be an intermediate shaft besides the motor and drive axle. Doing a full 23:1 step down in one stage seems (totally unfounded) like it would reduce torque capability. So I was throwing out there the idea that the drive unit starts with the same motor and gear as the 3, but that gear then feeds another reduction stage which meshes with the final gear (with internal splines for the axle shaft). The intermediate input gear could also be the same as the 3, although changing it would allow two lesser jumps in ratio say 4:1 and 5.75:1 v.s. 9:1 and 2.6:1. Custom gears probably make the most sense...

This two stage approach would also assist with the two ratios. The motor gear and final gear could stay the same, and only the intermediate shaft and its gears change, along with the mount points. Or only the output of the intermediate changes. Lots of options!

OK I see what you are saying... combining the initial 9:1 reduction present in the Model 3 with an additional reduction ratio for the Semi.

I do indeed suspect you are right in that it's likely multi stage... but have no real idea if there's any Model 3 geartrain reuse either... but I'd have imagine it may be beefier than a passenger car usage.

It would also be interesting to see what that connect in to the axle setup is. I assume no differential, as each motor drives a wheel separately.
 
But assuming the bias is is similar to the dual motor sedans, the motors in the "RPM sweet spot" (i.e. the 15:1 set) are likely to provide the majority of the power. So could we be looking at those motors running at 30-40KW continuous during highway cruising?



OK I see what you are saying... combining the initial 9:1 reduction present in the Model 3 with an additional reduction ratio for the Semi.

I do indeed suspect you are right in that it's likely multi stage... but have no real idea if there's any Model 3 geartrain reuse either... but I'd have imagine it may be beefier than a passenger car usage.

It would also be interesting to see what that connect in to the axle setup is. I assume no differential, as each motor drives a wheel separately.

PM motors have less of a sweet spot.
What’s the Difference between AC, DC, and EC Motors?
ECDRIVE DC Brushless Motor - LORENTZ

I'm also going on the assumption (yeah I know) that each motor/pack is independent. Putting those two things together, I'm thinking all 4 put out the same power during cruise. Assuming they don't hit the 100% inverter limit, they could all share across the operating band. One set would produce higher torque, but the other would be running at higher RPM.

Maybe they reuse the 3 motor gear, but I doubt it (thus contradicting myself). A little research I did says 1:10 is a general max for a set of gears. Splitting the ratio into squarish roots would be a reasonable thing.

Rear axle connection guess:
Shaft with final gear milled in. Inner shaft is splined for axles to slide into. Bearings on either side of gear. Shaft bolted to carrier via half shell pieces over bearings.
Similar to How To Rebuild A GM 8.5” 10-Bolt Rear Differential but with no differential gears, pinion and hypoid replaced with helical.

Edit: More likely that final drive gear bolts onto a flange on the shaft rather than being integral to it. Also improves commonality.
 
Last edited:
PM motors have less of a sweet spot.
What’s the Difference between AC, DC, and EC Motors?
ECDRIVE DC Brushless Motor - LORENTZ

I'm also going on the assumption (yeah I know) that each motor/pack is independent. Putting those two things together, I'm thinking all 4 put out the same power during cruise. Assuming they don't hit the 100% inverter limit, they could all share across the operating band. One set would produce higher torque, but the other would be running at higher RPM.

While PM motors can be more efficient than AC induction motors at a given RPM range (by a few percentage points), PM motors are not without back-EMF at higher RPM, I believe. And with there being >50% difference in RPM for the 23:1 set of motors, I would have to expect they are going to be in the range where they are generating quite a bit of back EMF.

If my earlier SWAG that the 15:1 motor set will see the same effective gearing/tire ratio as a model 3, then at 65 MPH the motor-set on the 23:1 axles would be turning at the equivalent of 100MPH.

I wonder if the power delivery would be split evenly to both pairs in that scenario?

Maybe they reuse the 3 motor gear, but I doubt it (thus contradicting myself). A little research I did says 1:10 is a general max for a set of gears. Splitting the ratio into squarish roots would be a reasonable thing.

Rear axle connection guess:
Shaft with final gear milled in. Inner shaft is splined for axles to slide into. Bearings on either side of gear. Shaft bolted to carrier via half shell pieces over bearings.
Similar to How To Rebuild A GM 8.5” 10-Bolt Rear Differential but with no differential gears, pinion and hypoid replaced with helical.

Edit: More likely that final drive gear bolts onto a flange on the shaft rather than being integral to it. Also improves commonality.

That description of axle design would seem to make sense... thanks.
 
While PM motors can be more efficient than AC induction motors at a given RPM range (by a few percentage points), PM motors are not without back-EMF at higher RPM, I believe. And with there being >50% difference in RPM for the 23:1 set of motors, I would have to expect they are going to be in the range where they are generating quite a bit of back EMF.

If my earlier SWAG that the 15:1 motor set will see the same effective gearing/tire ratio as a model 3, then at 65 MPH the motor-set on the 23:1 axles would be turning at the equivalent of 100MPH.

I wonder if the power delivery would be split evenly to both pairs in that scenario?

Back EMF is linear to RPM and based on motor winding and rotor magnets. So the 23:1 will have 50% more back voltage and need 50% more voltage headroom for the same motor torque output. What the torque per amp and volts per RPM are for their motor is a big ????

At cruise, I can see them able to share equally.Power needed on the 5% grade is less than the 0-60 in 20 seconds metric, so it could be able to share in that case. If they don't share equally at the 5% grade scenario, and the pack/motors are independent, then the packs will get unbalanced to each other. This can be corrected by shifting the drive power once back on flat land (or during the descent).

All based on numbers we don't have. :mad:
 
The casing looks like it might have two stages of reduction. Perhaps it has stock 3 gears as a first stage and then the extra reduction after?

The highest feasible ratio for a single stage of helical gears on parallel shafts is around 4:1, so you can guarantee that these are both 2-stage designs (same as Model S and probably Model 3) or 3-stage designs (for 23:1).
 
The highest feasible ratio for a single stage of helical gears on parallel shafts is around 4:1, so you can guarantee that these are both 2-stage designs (same as Model S and probably Model 3) or 3-stage designs (for 23:1).

Do you have a link to a tear down of a drive unit? All I've found are exterior views which don't seem like they have features for an intermediate shaft. So I've been assuming one stage of reduction. Are they running a planetary and a helical?

A few sites I've found give the 10:1 Max ratio, I'd be happy to be pointed to a more conservative source. Comparison of Gear Efficiencies - Spur, Helical, Bevel, Worm, Hypoid, Cycloid
 
I've seen static teardown pics before, but I can't find them right now. You can see in the "How it's Made" video (start at 8:45) that there's two stages:

It's a bit hard to see, but the primary motor shaft is behind the clear plastic protecting the motor lugs. It's driving that intermediate gear behind it, which in turn drives the final gear on the differential assembly he installs...
 
  • Informative
Reactions: mongo
I've seen static teardown pics before, but I can't find them right now. You can see in the "How it's Made" video (start at 8:45) that there's two stages:

It's a bit hard to see, but the primary motor shaft is behind the clear plastic protecting the motor lugs. It's driving that intermediate gear behind it, which in turn drives the final gear on the differential assembly he installs...

Ah, yes at 8:53 (site prevent URLs with time embedded). Thank you for that. Makes a lot more sense after looking at what the gear diameters would need.

Also looks like a spur gear differential unit.
 
  • Like
Reactions: scaesare
A problem I see with the different ratios would be breaking traction on the wheels with the higher ratio in slippery conditions. Are these units capable of manipulating the inverters so they can provide the exact same torque output to each wheel?
 
A problem I see with the different ratios would be breaking traction on the wheels with the higher ratio in slippery conditions. Are these units capable of manipulating the inverters so they can provide the exact same torque output to each wheel?

Quite definitely. That is how ABS/ traction control works now. Should be even easier with PM motors each driving a single wheel.
 
In addition to what @mongo said, I'm not sure if big rigs today employ limited slip diffs (I would think so in at least some cases), but with the Tesla drivetrain layout you should get all the advantages of full lockers, with the driving ease of fully open diffs.

I actually think it would perform better than full lockers...
 
  • Like
Reactions: mongo