Coupling Two Shafts to A+X RPM with Fluid Coupling

[one_raven]

Hi.
I am trying to figure out how to couple two seperately driven shafts so that the speed of shaft B is always at X RPM greater than the speed of shaft A regardless of what speed A and B are driven at.
(X is a constant, but the value depends on some still unfinished math, we will say 500 for now.)

A--C--B

A is a variable speed (1000 - 7000 RPM) 500 HP motor.
B is a variable speed (2400 - 16800 RPM) 200 HP motor.
B is always spinning significantly faster (anywhere from 1400-9800 RPM faster) than A in the same direction.
C is the coupling device.
Regardless of how fast A is spinning, I want C to limit the speed of B's shaft to X RPM greater than the shaft of A.
Can this be accomplished with C being some type of a Fluid Coupling device?

This is the configuration I was thinking of...
(sorry for the poor diagram)

A--C--D--B
E-----|

A is connected to C with a direct shaft.
B is connected to a differential gear set (D) with output 1 connected to C and output 2 serving as a secondary power path (E).
I want output 1 of D to always rotate at RPM(A)+X.

For example:
(if X is 500)
If A is spinning at 2000 RPM and B is spinning at 4800 RPM:
I want output 1 of D to be spinning at 2500 RPM.
If A is increased to 3000 RPM and B is increased to 6000 RPM:
I want output 1 of D to be spinning at 3500 RPM.
C will be effectively acting as a speed limiter for B.


Is there some type of Fluid Coupling that would accomplish this?
If not, do you have any suggestions regarding what configuration I can use?

Thanks for any help you can offer.

 

[enigma]

Unfortunately, I can't remember the specifics of how it worked (or what it was called. It was a year and a half ago, and I haven't thought of it till now.), but one of the labs at my school has something that may do the trick.

The basic gist was that the driver shaft had a thing attached to the end of it which was spring mounted. As the shaft sped up, the attachment expanded radially. Somehow, a variable length belt was connected to the outside of the shaft attachment and to the slave shaft.

I think that with calibration, a grouping of springs with different strengths could couple the two shafts together in the method you're describing.

Alternately (if it's an option) you could simply place an RPM sensor on shaft A, send that value to a microcontroller, and have the microcontroller output to the second shaft what it's desired operating RPM should be. If it's an option, this would be the way I'd go. There would be much fewer moving parts, and much less to go wrong.

 

[one_raven]

That first scenario is clever, but it won't work for my application.

How wide the device is open is determined by position of the spring loaded flanges.
Positioning of the spring loaded flanges is determined solely by the RPM of the shaft they are attached to, irrespective of the other shaft.
What I need is something that will keep B running @ 500 RPM greater than A irrespective of overall RPM.

So, if you set them to open at 2500 RPM, for example, at 2500 RPM they would be open.

Besides, I doubt a variable length belt system would work at the horsepower numbers I am dealing with without a great deal of slippage.

As for the second idea.
That would work, but I want to do this solely mechanically.
No computers.
No electronic parts.

 

[Bystander]

You can't run "A" and "X" as inputs to a differential and take "B" out?

 

[one_raven]

X isn't a motive device.

X is just the number I am looking for as the difference in RPM between the shafts (A and B) of the two motors.

 

[enigma]

Do you need the ratio to be consistantly +X, or can it be discretized and vary by a fraction of X?

 

[one_raven]

It depends on how much it varies, I suppose.

Ideally, it wouldn't vary at all.

Why?
What did you have in mind?

 

[enigma]

Some big nasty gearshift type thing which would run along the length of the axle. As the axle speeds up, it'd push itself down the row and hit larger and larger gears on the slave axle. It'd definitely need a clutch or a transmission to operate, otherwise it'd grind itself to pieces.

I don't know if there are any feasible fluid linkage scenarios like you asked about. Speculating would be far outside of my area of expertise.

Aside from that, I can't really envision anything non-computer controlled which could be constructed for a reasonable cost or size.

I'll keep it on the back-burner though.

Maybe the Mechies here could give it a go? *hint, guys!*

 

[wimms]

one_raven, what are you trying to build it for? (just curious)

What you effectively want to find is device that limits its single output shaft to exactly X rpms, no less, no more. Imagine box - with single output shaft. You attach that box to A shaft, and yes, it would be your C coupler. Anything with gears seems to lead to gear and rpm ratios, not fixed difference.
Not to forget that limiters sense torque, not rpm. rpm difference is a result. Thus you need torque sensing device with very sharp characteristic of going from not resistence to major resistence.

How to limit rpm to a fixed value? To me, limited slip diffs from cars come to mind. Viscous coupler could be done that would fix the rpm diff to given value. But it wouldn't be terribly precise, and working by heat expansion, has its downsides.
So, depends on what you really want to do.
You may also want to take a look at Haldex clutch principle. Its microcontrolled, but if you need fixed setup, could be done purely mechanically too I guess.
Of course, you could make rpm limiter at any suitable rpm range, and then gear down or up to get needed torque/rpm parameters. Makes me wonder of gyros at huge rpms with perhaps some fancy wormgearing.

 

[enigma]

Don't know if you're still working on this problem, or not...

Is the limiter device allowed to contain another motor?

Have a motor attached to the end of shaft A. Provide power to it with brushes and rings. Have it output a constant 500 rpm, and have it driving a chain which is connected to shaft B.

 

[one_raven]

Originally posted by enigma
Don't know if you're still working on this problem, or not...

Is the limiter device allowed to contain another motor?

I am still working on it.
Unfortunately, that is not an option.
But thanks for the input.

Originally posted by wimms
You may also want to take a look at Haldex clutch principle.

I hadn't heard of this before.
I looked it up, and it is very interesting.

I have to find more documents about it.
Thanks.

 

[wimms]

And btw, have you looked into Torque Converter used in automatic tranny? Afaik they could be tuned to have suitable lockup rpm difference, and they offer you torque addition from faster rpm engine.

 

[one_raven]

Originally posted by wimms
And btw, have you looked into Torque Converter used in automatic tranny? Afaik they could be tuned to have suitable lockup rpm difference, and they offer you torque addition from faster rpm engine.

Yes I have, somewhat.
That was my original thought.
In the diagram in the first post in this thread, I was thinking that E could be an adjuatable fluid coupling of some sort whose lockup speed would be adjusted with a hydraulic pump attached to either A or B so that when the RPM of the motor increases, the lockup speed of the coupling would increase.
I am just not sure of the feesibility/accuracy of a setup like that.
Plus it would be a pretty complex setup.

The other thing I was thinking about was C being a fluid coupling with a lockup speed of 500 RPM but one that was independant of the RPM of the whole unit.
I would attach the input to motor B and the output to motor A, so that when B tries to spin @ 500 RPM > the relative speed of the output shaft, the coupling would lockup and keep it at 500 RPM over.
The problem with that is that I am not aware of any fluid coupling that would operate in that same way and lockup at the same RPM regardless of rotation of the unit.
Do you?
That would be ideal.

I was just about to submit a post asking about the operating dynamics of fluid couplings. I am not as versed in them as I would like to be.
Maybe someone can provide me with some links that will help me learn more about them.

 

[wimms]

Originally posted by one_raven
The other thing I was thinking about was C being a fluid coupling with a lockup speed of 500 RPM but one that was independant of the RPM of the whole unit.
I would attach the input to motor B and the output to motor A, so that when B tries to spin @ 500 RPM > the relative speed of the output shaft, the coupling would lockup and keep it at 500 RPM over.
The problem with that is that I am not aware of any fluid coupling that would operate in that same way and lockup at the same RPM regardless of rotation of the unit.
Do you?

That isn't really your problem here. Torque converter lockup isn't what you are after. It gains ability to lockup with unit rpms. Whether it locks or starts working as motor depends on design.
LSD clutches work on principle of fluid becoming stiff with heat. They have 2 plates that when rotate at different rpms, generate frictional heat, due to that fluid stiffens, and locks up clutch. They do not depend on RPMs of the unit, only on relative rpms.
Something similar to TC, if you can redesign it.
But both these solutions aren't quite good for you, because you need constant rpm difference under torque. Translate that into power at your rpms and torque, and you'll know exactly how much heat it generates, that is effectively loss. I bet its in tens of KW for your engines. Any frictional device will do that.
I don't have any idea how to limit rpm difference to a constant without friction, so I'd suggest to look into ways to avoid a need for such a device.