I’m still curious as to how the viscous coupling goes into the hump state. My earlier posts with accompanying documents mention the hump state wherein 100% of torque is transferred from input to output shafts. Recall how the silicon fluid (some form of siloxane) thickens under shear stress and so begins to transfer torque. This shearing increases temperature and consequently pressure inside the coupling housing. VW’s own publicity literature states that the temperature rise is what causes the increase in fluid viscosity, but this research paper disproves that.
The problem I had was wondering what pushes the plates closer together to create the hump state. Just saying pressure increase does not cut it, for the coupling is in a sealed housing and pressure increase should be the same on both sides of the plates.
One thought I had was that there were localised asymmetric pressure increases, but I still had a problem with that as I thought pressure increases would be equalised quickly in a fluid.
Well I was half wrong, the following excerpt from a research paper by Mohan (2002) shows localised pressure increases forcing the plates together. Its a little like tilting your hand out the car window, or is it? Examine diagram closely and make up your own mind 🙂
Oh, STA, self induced torque amplification, is another term for the hump state.
So, as I see it, the progression goes like this:
-rotational difference between input and output shafts (ie front vs rear wheel speed difference, above the 5% or so allowed slippage) causes shear and increasing viscosity in the siloxane fluid.
– at some point, localised pressure differences between plates force the plates together, in effect coupling the input and output plates. Remember the plates are very close together, and one set is on a splined shaft and is free to move axially).
– then the standard story applies, when the plates are coupled there is no relative speed difference between the plates, the shear is gone, and the viscosity drops again, the plates separate and if the input and output speeds remain sufficiently different then the process repeats.
But one thing still puzzles me, in this paper (yes, same one referred to above)
the author notes the pressure increase inside the coupling during slippage, and wonders if some sort of pressure control on coupling could affect the hump state behaviour. I still haven’t resolved how overall internal pressure affects localised inter-plate pressure differences.
And to cap it all off, GKN Drivelines (heir to the Ferguson developers of the viscous coupling) published this bit of info:
See how the diagrams resemble Mohan’s diagrams above, But then they go on to say:
“The “Hump” mode is activated when the coupling achieves 100% filling due to fluid thermal expansion thereby amplifying a hydraulic throttling effect between the plates”
This statement is sort of misleading. Makes you think that the thermal expansion is general (which it is) and not localised (which it is too).
Oh and that reminds me, as you know the coupling is not filled 100% with fluid, there is a small amount of air left in (what is it, 7-12%?). This air ends up distributed, apparently, as small bubbles and acts as a moderating agent in how aggressively the coupling goes into the hump state. For instance, the less air left in, the more aggressive the coupling will be.
Call me an obsessed nerd if you will, but I wish the developers of the viscous coupling (Ferguson et al ?) would call me up and invite me over to explain all 🙂