"Bi-directional" Torque Converter?

[aaqubit]

Torque converters in autos are driven one way only: from the impeller to the turbine. The strength of the coupling is mainly a function of the impeller speed only, since this variable is directly responsible for the centrifugal force accelerating the fluid through the turbine blades to provide coupling strength.

My question is, what effects are there when the impeller is trying to force the turbine clockwise, but the turbine, at t=0, is spinning counterclockwise with: 1. low speed? 2. medium speed? 3. high speed? How would these three cases compare with a stationary turbine?

I think that with the turbine spinning opposite to the force applied by the impeller, the coupling strength would be slightly enhanced, but the the effect would be quite small. In other words, case 1 below would give a stronger coupling than case 2, which would be stronger than case 3:
1: Impeller 1000rpm, turbine (at t=0) 0rpm
2: Impeller 500rpm, turbine (at t=0) -500rpm
3: Impeller 0rpm, turbine (at t=0) -1000rpm

If the above is true, any suggestions on making the above three cases give the same coupling strength? If the above is not true, then could someone explain to me?

 

[dkozloski]

The original GM Hydramatic transmission that dates back to the late 1930's was bidirectional in that there was no angle to the two element coupling blades either the drive or driven side. Because torque multipication was small 1st gear was about a 4:1 ratio.

 

[aaqubit]

does this mean that an impeller spinning at 1000rpm would apply the same force to a stationary turbine as a 500rpm impeller would to a -500rpm turbine?

To clarify myself, I guess the real question is this:

if the function f gives the magnitude of force applied by the impeller to the turbine, can this function be written as:

f(x,y) = g ( x - y )

where x = impeller velocity (note the vectorial nature), y = turbine velocity.