Transmissions and final drives

[Xoon]

Could anyone with a good understand of gearboxes explain how this steering differential works?

 

As far as I understand, the drive input powers the system, and when the steering differential's RPM equals zero, the torque is distributed equally.
When the steering input rotates clockwise, the right side drive gets the most torque, proportional to the steering input torque. 

When the steering input rotates counter clockwise, the left side drive gets the most torque, proportional to the steering input torque. 

 

When the drive input is put in neutral, the steering input can be used to pivot on the spot. When doing this, the steering input directly powers the side drives. 

 

The one thing I get is, how does the steering input control the torque of the drive side? Is it by the epicyclic gear? 
 

Why does it not transfer the torque back into the steering input and grind against the other epicyclic gear, or fight the steering input?

[Gripen287]

On 12/4/2019 at 1:13 PM, Xoon said:

Could anyone with a good understand of gearboxes explain how this steering differential works?

 

I'm no expert, but according to wikipedia , this drawing is incomplete. Wikipedia says this is, " [a] Differential steering mechanism, either double-differential minus the clutches, or triple-differential minus the brakes. 

 

More specifically, I believe they mean a Maybach double differential with clutches to disconnect the "slowed" (i.e., not sped up) sprocket when steering torque is applied. Alternatively, the diagram could depict an "ordinary" double differential transmission, but the diagram's "epicyclic gears" are not correctly drawn as differentials, as in the transmissions depicted here or here. To depict a triple differential transmission, the steering torque is input into the steering half shafts via yet another, third differential (the diagram is missing a carrier having idler(s) between the steering half shafts) and brakes and/or clutches on the steering half shafts selectively engage and/or disengage to effect a steer. If you haven't already, you might want to revisit Coli's link.

 

So, yes, you are correct in that the diagram does not include means for controlling the reactive torque applied to the drive input by the slowed sprocket when the steering input is driven. If the drive input and steering input were independently driven, you might be able to get away with that kind of set up for a while, maybe.... But the various multi-differential transmissions differ, more or less, in how to address the issues you have raised. 

[Xoon]

18 hours ago, Gripen287 said:

 

I'm no expert, but according to wikipedia , this drawing is incomplete. Wikipedia says this is, " [a] Differential steering mechanism, either double-differential minus the clutches, or triple-differential minus the brakes. 

 

More specifically, I believe they mean a Maybach double differential with clutches to disconnect the "slowed" (i.e., not sped up) sprocket when steering torque is applied. Alternatively, the diagram could depict an "ordinary" double differential transmission, but the diagram's "epicyclic gears" are not correctly drawn as differentials, as in the transmissions depicted here or here. To depict a triple differential transmission, the steering torque is input into the steering half shafts via yet another, third differential (the diagram is missing a carrier having idler(s) between the steering half shafts) and brakes and/or clutches on the steering half shafts selectively engage and/or disengage to effect a steer. If you haven't already, you might want to revisit Coli's link.

 

So, yes, you are correct in that the diagram does not include means for controlling the reactive torque applied to the drive input by the slowed sprocket when the steering input is driven. If the drive input and steering input were independently driven, you might be able to get away with that kind of set up for a while, maybe.... But the various multi-differential transmissions differ, more or less, in how to address the issues you have raised. 

Thank you, this cleared up a lot of things for me. 

 

I have been looking into torque vectoring transmissions and differential steering for electric and hybrid transmissions. 

 

I saw the transmission made by University of Munich:

 

 

 

 

Here is another one from Borg Warner:

 

 

 

Call me a idiot or something, but I can't for the love of god understand these. It can't be as simple as using a electric motor and a clutch to apply additional power to one of the two wheels right? 

 

I have also been pondering about the efficiency of several transmissions, and I honestly came to the conclusion  from rough estimates that a dual motor setup is the most optimal. 

 

Let's consider a few layouts:
-Single motor with a differential.

-Independent motor setup.

- Double differential with electric motors.

 

In a single motor setup with a transmission, one would think one motor would be more efficient right? Well, actually no, as seen in many hyper efficiency EVs.

 

Let's have a few basis numbers:
Electric motor(M): 95%

Motor controller(C): 95%

Differential(D): 93%

Reduction gear(R): 95%
Power(P): 100Kw

 

Single motor with differential:
Efficiency = (P*C*M*R*D)

79,93% = (100*0,95*0,95*0,95*0,93)

 

Total system efficiency is 79,93%.

Capable of basic torque vectoring with braking, greatly hurting the overall efficiency.

 

Independent motor setup:
Efficiency = ((P/2)*C*M*R)*2

85,74% = ((100/2)*0,95*0,95*0,95)*2

 

Total system efficiency is 85,74%.

 

Honestly, considering the complicated gearing, gearing of the double differential system with electric motors, I doubt it would come close to either above, so I will do any rough calculations. 

 

 

A independent motor setup also beats out a average transmission by 10 percentage points. (85% vs 75%). Factoring in the generator and rectifier (0,95 for generator and 0,95 for active rectifier) it still leads by 5% points (80% vs 75%). 

This might not seem like a big deal, but this is essentially 75hp in a 1500hp system.  And this does not even start to factor in regeneration, which in EV's extend their range by up to 30%, and should be just as effective in a tank that frequently sprints and stops. 
Not to mention the fuel saved by running the engine at a constant RPM, harvesting excess power from the turbocharger, and being able to automatically turn the engine on and off to save fuel while on standby, and the increased initial acceleration.  One could also saved weight, as there is no need for a starter, alternator or flywheel.  The internal generator could even be used to external equipment or high power equipment like a ETC gun.  1100Kw is a lot of power to have on standby. 

 

Other advantages of the independent  motor setup, hereby IMS, is the ability to "reverse" the transmission. You can literally drive the same speed in forward as in reverse, and you could reuse a MBT chassis as a IFV chassis by turning it around, vice versa. 

The motor can also be used for negative torque vectoring, recovering power when doing tighter brake turns. They can also produce peak power 3+ times their continues rating.  That means a 1000hp AFV would be able to do short sprints at 3000hp!

 

It has been noted that it would not be able to transfer torque between the motors, but when does a AFV actually need to apply more than 50% power on one track? As far as I have seen, it is a rare occurrence. Another issue some say would be that the motors would rotate at different speeds.  This would be very easy to rectify. Simply use the motor controllers current readings (Sinus curve corresponds to the position of the rotor magnets in relation to the poles on the motor) to track the motor's rotor position and estimate the speed from that , or at a rotary encoder for a simply speedometer. 

 

 

 

 

 

[Gripen287]

2 hours ago, Xoon said:

It can't be as simple as using a electric motor and a clutch to apply additional power to one of the two wheels right? 

 

Actually, it appears that that is the case in those two examples. I think the Borg Warner system IS described by the "incomplete" wiki figure, albeit laid out a bit differently (i.e., the balance shaft and idler gear perform the same function as the "steering input" bevel gears). The Borg Warner video is kind of confusing because it appears that the rotor of the TV motor rotates in the opposite direction of the "vectoring torque." As I understand it, the TV rotor drives a second, coaxial sun gear within the right wheel ring gear, the second sun gear driving a second carrier having planets that engage both a support member and the ring gear of the right wheel planetary gearbox. Accordingly, when the TV rotor spins CCW torque is vectored to the right wheel as depicted in the video by the CW "vectoring torque," and vice versa. The balance shaft and idler transfer and the reverse the direction of the "vectoring torque" with respect to the ring gear of the left wheel planetary gearbox. 

 

I think the clutch in the Borg Warner system is just there to improve efficiency when the propulsion motor isn't in use at all.

 

That MUTE system is a fair bit more complicated, but it appears to work using the same, simple principle. I couldn't really tell what was going on in the YouTube video, but I did find this paper. Figure 5, reproduced below, has a diagram of a further evolution of the system depicted in the YouTube video, which seem to mostly differ in how the "superimposing electric machine" is integrated.

 

In Figure 5, the left wheel is directly driven by the carrier in the spur gear differential, the carrier being driven by the planets depicted as hashed lines (I'll refer to these as the "primary left wheel planets") that interface with the ring gear, which driven the via the electric drive motor gear train. The spur gear differential planets represented by the solid lines (I'll call these the "primary right wheel planets") rotate with the carrier and in doing so drive the sun gear connected to the right wheel. The superimposing electric machine drives the interconnecting sun gear, and thus the primary left wheel planets and carrier) via the planets of the right-most carrier in the superimposing gear (I'll refer to these as the "secondary left wheel planets/carrier"). The planets of the left-most carrier in the superimposing gear (I'll refer to these as the "secondary right wheel planets/carrier") are connected to the secondary left wheel plants by an idling ring gear.

 

Because (i) the primary left wheel planets only interface with the "interconnecting sun gear" and ring gear and (ii) the primary right wheel planets only interface with the right wheel sun gear (although both are carried by the carrier directly driving the left wheel), the right and left wheels can rotate at the same speed when driven by only the electric drive motor and at different speeds when torque is added/subtracted via the superimposing electric machine. Torque can be vectored to the left wheel and away from the right wheel  by running the superimposing electric machine in a first direction and a vice versa by running in a second direction because the idling ring gear reverses the direction of the applied torque between the secondary left wheel planets/carrier and the secondary right wheel planets/carrier. 

 

I think/hope that makes sense.

 

I think passenger vehicles like these systems seem to be designed for can get away with the fatigue loads that these systems will likely place on the drive shafts, as opposed to isolating wheels/sprockets via multiple differentials/clutches, because the duty cycle and loads are probably relatively light. Trying to turn a 20+ ton tracked vehicle is an entirely different proposition. I'm sure you could do it with beefy enough components, but I can see why adopting a bit more complex and sophisticated system would start to become appealing. 

 

Standby for a post and document dump covering split torque transmissions as applied to helicopters as soon as I can get around to writing it up. I think moving away from planetary reduction gear boxes could open up some interesting possibilities as far as helicopters, particularly compound helicopters, are concerned.