Torsen Differential, How it works ?

The working of Torsen differential is elaborately explained in this video with help of animation.

[trueblue]

It’s quite simple really, the engine applies torque to the input; *’witchcraft happens’*, and the wheels turn. See very simple.

[techdaemn]

It is totally false to state (as this video does) that a Torsen diff will "lock" when one side has traction and the other does not. This has caused much confusion. As many of the commenters who actually have Torsen (also known as torque biasing) diffs have confirmed, it still acts like an open diff when the traction on the low traction side approaches zero. A Torsen diff *never* locks. The angle of the teeth on the worm gear and wheel establish a torque multiplier (torque bias ratio) greater than 1. In a conventional open diff, the multiplier is 1 (torque bias ratio of 1:1). This means the high traction side always gets exactly the same amount of torque as the low traction side. When traction on one side drops to zero, it takes effectively zero torque to spin that tire, so the open diff sends zero torque to to the high traction side, and you go nowhere. A Torsen diff may have, for example, a TBR of 3 to 1. In a low (but not zero) traction situation, the higher traction side gets 3 times the torque of the low traction side. If it takes 200 N*m to break loose the tire on the low traction side, the Torsen will send 600 N*m to the high traction side, giving you a total of 800 N*m to the ground. But being a multiplier, if the low traction side takes close to zero torque to induce wheelspin, then the high traction side gets 3 time 0 = 0, and you still go nowhere, just like an open diff.

The above is the primary reason why Torsen differentials are not used for all applications. They are *not* noisy (as was claimed by +Learn Engineering), but they are non-adjustable (can't change torque bias ratio on the fly), can't disengage when used in the center to reduce drivetrain losses and they never actually lock at all, making them mostly unusable for extreme offroad use.
The original Hummer uses Torsen diffs because they're mechanically strong and zero maintenance and the vehicle was designed for desert and other off highway usage, while keeping all four tires on the ground, not rock crawling (where you may loft a tire off the ground). When you do loft a single tire off the ground on the Hummer, it can get stuck. The trick is, while continuing to apply power, to lightly apply the brakes which increases the load on the previously zero traction side of the diff, and the Torsen diff then multiplies the torque required to overcome that braking force by its torque biasing ratio to give you more torque to the high traction side as explained above. This same trick works a little bit with an open diff.
Any Audi quattro made in the last 15-20 years or so does the same trick electronically by selectively applying the ABS brakes only to the spinning tire on the zero traction side and Torsen then sends that torque times the TBR ratio to the high traction side. It's quite effective.

On a side note. Understanding torque transfer can be extremely confusing. For instance, it is quite common to state that when a locking diff is locked, that it has a 50/50 torque split, but that is not true. An open diff has a 50/50 torque split, 100% of the time. A locked diff, where one side has no load (zero traction) and the other side has traction, now has a 100/0 torque split. 100% of the available torque goes to the high traction side, and virtually nothing to the zero traction side, despite the fact that both sides of the diff are rotating at the same speed. Relative rotation rates have almost nothing to do with applied torque.

[mikevez]

The graphics made to explain this are amazing, something like this would be impossible to explain otherwise. Great Work!

[lejink]

"here comes the tricky part"    I was struggling long before we got to that point lol

[felixnauta]

Fabolous grafics, and without any distrubing musical background...much better.
Thanks for sharing.

[GERntleMAN]

That is the best explanation I've seen so far in the last 20 years. I would recommend this video to anyone who did not fully understand the locking or the correspondence of the worm wheels with the worm gears.

[teemukeskinen]

I have to say, when my university teacher tried to explain this with only technical CAD drawings and such, I had absolutely no idea how that thing worked. This was very clear explanation of a quite tricky subject! Thanks!

[akrounds]

This is the clearest explanation I have seen about the principle of the Torsen differential. You only need 5 minutes to fully understand.👍👍👍👍👍👍

[SkypowerwithKarl]

They work great. Had them installed 30 years ago in a truck. You could have one tire in the air and the one on the ground would still pull you. Both tires will turn about the same speed and one won’t spin wildly. In turns on ice and snow you could still steer unlike a locker or most posi’s. An outside force (road) can cause the wheels to differentiate.

[Guywithcrazyideas]

I need a desk model that I can put my fingers in and play with.

[martincraig]

Truly genius. I always was mechanically inclined, but gears (and rope knots) was something I still can't comprehend fully.
Although this doesn't lock the wheels completely in a zero traction situation.
The simplicity and robustness is truly amazing.

[CombraStudios]

That's the most easy-to-understand presentation of an amazingly simple solution to a long time technical problem. I'd never be able to invent such a thing.

[CXGTiTurbo]

Best Torsen dif explanation ever! Perhaps also mentioning the downsides would be great in the future.

[ladamyre]

I had a Torsen T2 installed in my 2004 Mercury Marauder and could not be more pleased. It cost me $2500 and has been worth every penny.

For street driving, HARD driving, there is no better solution. Many times before the upgrade I would have places where I would be negotiating a tight turn while applying high torque to the drive wheels, such as a 4-2 downshift coming out of the turn. Before, with the stock Traction-Loc differential, the two rear wheels would lock together and I would lose traction on the inside wheel because it would be spinning at the same speed as the outside wheel. But with the T2 that never happens and I can make those 4-2 downshifts coming out of the turns with confidence and full traction on both rear wheels. She's like a roller skate on a rail!

[ianhinkle]

Things like the Torsen differential and VTEC and variable compression engines just make me in awe of how a person even thought of the idea, let alone implement it effectively

[rotorblade]

I’ve installed an aftermarket (quaife) one on my sports car and it greatly enhanced the driving experience. You can put more power through corners, when you reach the limit the rear starts to slip gradually but the limit is higher. Before it would not do anything then snap suddenly because it would spin the inner wheel with the other not spinning then it would suddenly spin the other wheel if pushed further.

[benvincent]

First patented in 1958 by Vernon Gleasman. He built them himself. He had to modify cutting machines to be able to make the gears. It was first marketed as the Dual Drive. It was later manufactured by Triple-D Inc of Detroit. Gleasman then went to Gleason Works in Rochester, NY when demand increased. Over the years there have been several owners of the patents.

[karthikkrishna]

Just watched 90 seconds, and I located my car blend door and just tapped it, it started working . You saved me 300$ just today . Thank you .

[mahxylim]

I love the animation! I learn by images and this really helps!

[jackt]

Read this and techdaemon's comment BEFORE watching the video. The Torsen does not work the way the video narrative suggests.

1. The video begins a worm gear can turn the worm wheel, but the worm wheel cannot turn the worm gear. That statement is correct. What is not correct, is that the Torsen uses worm gears and worm wheels. Next he calls his worm wheel, a spur gear, which cannot mesh with a worm gear. Then, notice how the gears magically transform from worm gears to 90 degree helical gears when the animation assembles them into the differential case. 90 degree helical gears work much like the ring and pinion gears on cars and transmit power in both directions. Moreover, with worm gears and worm wheels, the worm gear is always the much smaller gear. This shows the opposite. Since the differential pinions (which he calls the worm wheel) is smaller in diameter to the axle side gear (which he calls the worm gear), it is actually easier for them to turn the axle side gear than the reverse. This invalidates the basis of his worm-gear-based premise. However, the narrative continues based on the false worm gear premise, even though it is obvious to all there are no worm gears and worm wheels, and thus the remainder of his explanation of how it works is also false. I will prove in the following that differential pinion gears driving the axle side gears is central to how the Torsen works in a low-traction environment.

2. Gear types: Spur gears are the most efficient because all of the rotational thrust is applied between two gears in mesh is in their direction of rotation, which is 90 degrees to the shaft they spin on. Spur gears are never in contact with more than one tooth of the opposing gear at a time. The drawbacks are, since gears must have clearance to work, they generate noise and vibration when transferring from one tooth to the next. They are also not as strong because there is only one tooth in contact with the opposing gear. Helical gears are quieter and stronger because more than one tooth is engaged at all times as they slide in and out of contact, which is why they are used in transmissions. The downside of helix cut gears is gears sliding in and out of contact and the thrust vectors are not all in the direction of rotation of the gears, because the helix creates an inclined plane, thereby causing some thrust to be in line with the shaft, they rotate on. Optimizing this downside is what allows the Torsen bias its torque. One could argue that condition exists with 45 degree spur gears. That is true, but to a much smaller degree.

3. Open Differential: With a standard open differential, the axle side gears and the differential pinion gear that spans the distance between the two axle side gears is a highly efficient 45 degree bevel spur gear. The only time the differential pinion would turn relative to the differential case is if the axles were not going the same speed, either due to a turn, or loss of traction on one of the wheels. Therefore, when traction is good, and the vehicle is traveling straight down the road, the differential pinion gears are not rotating relative to the differential carrier, and since the differential pinion gears are in mesh with the axle shafts side gears, the teeth of the differential pinion gears simply push on the axle side gears through their teeth, with no movement between the the gear teeth of the gears. With no gears moving relative to each other, the effect is as though the axles were simply splined to the differential carrier because the axles rotate at the same speed as the differential carrier. Assume the differential carrier is rotating at 2 rpm, and you go around a corner that reduces the inner wheel speed to 1 rpm, then the outer wheel speed must turn at 3 rpm. Next, assume we have 100 rpm and 100 lbf of torque applied to the differential carrier. 50 lbf torque would go to each wheel when there is adequate traction at both wheels. Assume we are stopped, we add power, and the right-side tire starts to spin at 10 lbf of torque but the left tire doesn't move. As the carrier continues to turn at 100 rpm, the torque to the right-side wheel drops to 10 lbf of torque because that is all that it can do. However, because the carrier is still rotating about the stationary left-side axle, the left-side axle drives differential pinion gear, which in turn drives the right-side axle gear, thereby doubling the rpm of the right-side axle. The force generated to turn the additional 100 rpm cannot exceed the 10 lbf maximum imposed by the right-side axle, but the opposing force that drives the right-side axle through the differential pinion is on the left-side axle side gear, so 10 lbf is exerted added to the total tractive force for a total of 20 lbf, 10 lbf to each wheel, even though only right-side axle is turning. An open diff has a 50/50 torque split, 100% of the time.

4. Torsen Differential: While going straight down the road and axles turning the same speed, the Torsen differential behaves similarly to the open differential. If you simply go around a corner, and one wheel is going proportionately faster than the other, there is no torque being applied through the turning differential pinions because their axle side gears and rotating differential carrier are both perfectly coordinated. If the left-side wheel is on dry pavement, and the right-side wheel is off the ground, the differential carrier revolves around the left-side fixed axle side gear, and as with the open differential, transfers the left-side rpm through the differential pinions to the right-side axle side gear. Because the wheel is off the ground, very little additional torque against the left-side axle side gear is required to transfer the rpm from the left-side axle to the right-side, and thus negligible torque on the left-side axle gear is available to move the vehicle. If the right-hand wheel is on ice, and it can generate 10 lbf of torque, the differential carrier again revolves around the left-side fixed axle side gear, but this time it requires significantly more than 10 lbf of torque is required due to the thrust vectors of the helical gears to transfer the the rpm through the differential pinions to the right-side axle side gear, and arrive with 10 lbf of torque. The extra effort of torque as a ratio required at the left-side axle side gear is the design Torque Bias Ratio (TBR). If the designed TBR can achieve 3:1, then there will be can be up to 30 lbf applied to the left-side axle side gear, 10 lbf to the right-side axle side gear for a total of 40 lbf. If 40 lbf, is not sufficient to move the vehicle, the left wheel will remain stationary, and the right wheel will continue spinning at twice the differential carrier rpm, just like with an open differential. If there were a worm gear and worm wheel used here, it would be instant lockup, and there could be no TBR.

5. Torsen Implications:
- When turning a corner, it is not effectively different than an open differential.
- When the differential carrier rpm is not perfectly coordinated between the axles, as is the case when there is a traction problem, it biases torque to the wheel with the lowest rpm.
- If there is 0 lbf required to turn one wheel, there is 3 x 0 = 0 lbf transferred to the stationary wheel.
- If there is 10 lbf force on the spinning wheel, with a TBR of 3:1, there is a maximum of tractive force of 40 lbf available. If that is not enough to move the vehicle, it will still behave like an open differential with one wheels spinning
- If 40 lbf was not available to move the vehicle, you could drag the brakes. By adding 10 lbf to both wheels, you would increase the tractive lbf at the stationary wheels to. (10) + (30 + 30 - 10 = 50) = 60 lbf of total tractive force. Of course with one wheel off the ground being at 0 lbf for one wheel, you could also get moving by applying the brakes and gas.
- If I can drag the brakes on a Torsen and gain more traction, then why doesn't that work with an open differential with its 45 degree spur gears? It does to a lesser degree. Computers and traction control these days can also apply the brakes to the spinning wheel to bias torque perfectly to meet conditions.

 Again, techdaemon's comment explains accurately how this device functions, which is critical to its application and usage.