Alibro,
So you have got some of the fluid?
If it is available I could maybe try to see if I could find someone here to do some viscosity vs temp tests
 
I am just thinking aloud, posing questions (even if I didn't put a ? after everything, perhaps there should be one), not stating facts :)

That is Polydimethylsiloxane (PDMS)

Says here in terms of shear rate (how fast you twist the plates with respect to each other) effect on viscosity they can be Newtonian, shear-thinning or engineered to be shear-thickening (dilatent)

The only data I found of viscosity vs temp (for pure PDMS) said that viscosity dropped with temp (see figure 12)

So how can you be sure that stuff does what it should do, even if its of more or less the right molecular mass?
The GKN or whoever original VCU fluid could be very different from the pure stuff via 'engineering' additives etc etc.

My thoughts:

When the VCU starts to engage I think it it does so because it gets thicker (more viscous)
I think that this could be because, (either or both):
(i) it gets thicker at higher temperatures caused by slippage between plates (as the graph from Mad Hatter's post shows a large step rise in viscosity at around 110C, GKN or whoever must have engineered the fluid to get this behaviour. Where does this plot come from by the way?)
(ii) it gets thicker as it is sheared between the plates at a higher rate (shear-thickening = dilatant)

The thing is, either of the two effects (i) or (ii) above could degrade without the ambient temperature &/or very slow shear rate behaviour (viscosity) being affected -
That is the one wheel off test could not show any difference but the fluid might not be working right at higher shear rates &/or temps

Then again the test might well do because if a sign of an ageing VCU is that it 'binds' in reverse on full lock then this indicates that the fluid is also too thick under the wheel up conditions when it shouldn't be?

So the thing to do would be to do a standardised one wheel up test (e.g. given weight, given lever arm and say only time from 45 degrees to the ground since the initial part would be at a very slow speed and so very sensitive t0 the starting angle) for some new VCU's and also for some suspected old and degraded fluid ones, and then compare the results of the two groups.
You could even do a very simple statistics test to allow for the inevitable scatter in the results to be sure (ish) that there is a difference between the new and old times, 'on average'.

But isnt that the whole point of this thread? Crowd-sourcing experimental results for the common good of mankind?
Did I miss the results?
Quite possibly, and maybe this conversation has been had before somewhere in the 49 pages :)
I could also be thinking along completely the wrong lines of course, wouldn't be the first time and I hope not the last.

PS I realise the VCU can fail by 'not gripping when it should do' as well, but I am just thinking about the other case at the moment
 
The tests have been done and the results are pretty conclusive. I've compiled the data from threads on here over a few years now. Although the test conditions have altered slightly over time, which is a pain.

I do know the if the OWUT is done with a 1.2m lever with 5 Kgs hung on the end. The time is the time taken for the lever to move from 45° to 90°.
A new genuine GKN VCU takes about 18 seconds.
The time taken for a home reconditioned VCO with 100,000 fluid is about 18 seconds. This is dependent on the fill level though. Less fluid and more air means lower times. More fluid or pressurised fluid slows the time down.
A Bell Engineering reconditioned VCU also times at 18 seconds.

A lower miles but well tept VCU times between 30 and 45 seconds.

A VCU that has over 100,000 well kept miles will generally show times of 45 to 60 seconds.

Abused VCUs ( incorrect tyres) generally show times well over a minute, some well over 2 minutes. It's these that will damage the IRD / diff in very short distances.
 
Thanks Nodge68!

That is very clear-cut, and practical

When I bought my 2001 Td4 the props were on but he assured me that it had been run with no props - he had only ever put them back on for MOT, and to let me test drive with it on.
It didn't seem to resist too much in reverse on lock (definitely no stall on tick over) but it had new Falken tyres on the front and old worn out Goodyears on the back, so I took the props straight off again until funds allow new tyres.
When I get them I will put the props back on and do the test to help me decide if I take it back off again or not
(And I'll post the result here)

Isn't there any way to pin that specific post?
(there are so many posts on VCU and tests etc that that info gets lost in the noise)
 
Thanks Nodge68!

That is very clear-cut, and practical

Isn't there any way to pin that specific post?
(there are so many posts on VCU and tests etc that that info gets lost in the noise)

There are a few threads on the VCU and some on home brew referbs.
But I've done no pinning of threads. I've just compiled loads of data, mostly in my head, ready for use some day.
You are correct, there's a lot of noise that needs filtering in the relevant threads.
 
So the thing to do would be to do a standardised one wheel up test (e.g. given weight, given lever arm and say only time from 45 degrees to the ground since the initial part would be at a very slow speed and so very sensitive t0 the starting angle) for some new VCU's and also for some suspected old and degraded fluid ones, and then compare the results of the two groups. From 45 to 135deg in a clockwise direction as stated previously
You could even do a very simple statistics test to allow for the inevitable scatter in the results to be sure (ish) that there is a difference between the new and old times, 'on average'. That was the intention.... too few results posted to correlate

But isnt that the whole point of this thread? yes
Did I miss the results? possibly
Quite possibly, and maybe this conversation has been had before somewhere in the 49 pages :) yes - a few times


PS I realise the VCU can fail by 'not gripping when it should do' as well, but I am just thinking about the other case at the moment yes - "Mondo" mode
@Bife
 
I am just thinking aloud, posing questions (even if I didn't put a ? after everything, perhaps there should be one), not stating facts :)

That is Polydimethylsiloxane (PDMS)

Says here in terms of shear rate (how fast you twist the plates with respect to each other) effect on viscosity they can be Newtonian, shear-thinning or engineered to be shear-thickening (dilatent)

The only data I found of viscosity vs temp (for pure PDMS) said that viscosity dropped with temp (see figure 12)

So how can you be sure that stuff does what it should do, even if its of more or less the right molecular mass?
The GKN or whoever original VCU fluid could be very different from the pure stuff via 'engineering' additives etc etc.

My thoughts:

When the VCU starts to engage I think it it does so because it gets thicker (more viscous)
I think that this could be because, (either or both):
(i) it gets thicker at higher temperatures caused by slippage between plates (as the graph from Mad Hatter's post shows a large step rise in viscosity at around 110C, GKN or whoever must have engineered the fluid to get this behaviour. Where does this plot come from by the way?)
(ii) it gets thicker as it is sheared between the plates at a higher rate (shear-thickening = dilatant)

The thing is, either of the two effects (i) or (ii) above could degrade without the ambient temperature &/or very slow shear rate behaviour (viscosity) being affected -
That is the one wheel off test could not show any difference but the fluid might not be working right at higher shear rates &/or temps

Then again the test might well do because if a sign of an ageing VCU is that it 'binds' in reverse on full lock then this indicates that the fluid is also too thick under the wheel up conditions when it shouldn't be?

So the thing to do would be to do a standardised one wheel up test (e.g. given weight, given lever arm and say only time from 45 degrees to the ground since the initial part would be at a very slow speed and so very sensitive t0 the starting angle) for some new VCU's and also for some suspected old and degraded fluid ones, and then compare the results of the two groups.
You could even do a very simple statistics test to allow for the inevitable scatter in the results to be sure (ish) that there is a difference between the new and old times, 'on average'.

But isnt that the whole point of this thread? Crowd-sourcing experimental results for the common good of mankind?
Did I miss the results?
Quite possibly, and maybe this conversation has been had before somewhere in the 49 pages :)
I could also be thinking along completely the wrong lines of course, wouldn't be the first time and I hope not the last.

PS I realise the VCU can fail by 'not gripping when it should do' as well, but I am just thinking about the other case at the moment
I am a not so sure about Non-Newtonian claims. Made usually because the belief is that as shear increases more torque is needed.

The VCU transfers an amount of torque when the slip (shear) rate is low and this amount drops as slip rises. This to me indicates that it is not thickening as shear rates increase, it is thinning (probably to avoid damage to transmission). It is only when the rate of slip reaches the tuned 'hump' point (ie when the engineers/designers believe the front wheels have lost traction) that things change dramatically. At hump point the VCU massively increases the amount of torque transmitted. This may be due to temperature, but it is pressure that is the main governing factor. When the fluid reaches that hump pressure its characteristics change dramatically - maybe due to the pressure increasing temperature, but without the pressure there would not be the temperature rise. That is why the air gap is so important as it tunes the point at which that pressure is reached.

I don't know if you've stumbled across links to this document, but its an interesting analysis of actually how a viscous coupling works...

http://www.easy2design.de/stuff/visco_sae.pdf

As for why VCUs fail, in part I'm sure it is because, over time, the plates wear and the particles become suspended in the fluid altering its characteristics.
 
@GrumpyGel

"the torque is proportional to the 4th power of the effective radius" therefor, to increase the torque capabilities, the effective diameter needs to be increased.... ergo a bigger VCU.


Pressures of up to 100bar may be experienced during operation.


Because of the seal types required, a basic friction of up to 30Nm may be found - ie no load friction.


Higher viscosity siloxanes deviate considerably from Newtonian principles...... ie become "non-Newtonian". It can be shown that they do not shear like Newtonian fluids but create "sausages" of fluid across the plate boundaries which then roll around the plate surfaces.

As the fluid temperature and pressure increases to the "hump mode", the coupling acts in a similar way to a Fottinger coupling - see video from Siemens




A decisive factor regarding the torque progression across the plates is the pressure which is generated. If the pressure cannot reach a critical magnitude, "hump" cannot be induced.


The air chamber is between 5% and 10% of the siloxane fill volume and disappears completely at the limit temperature, causing the pressure increase.


Plates which have been subjected to prolonged heavy loading may deform into a "cup" shape. with wear on one side and next to no wear on the opposite side of the same plate. To prevent this, the outside edge has several "slots" or "nicks" machined into it, to allow the plates to remain stable.

A decrease in stopping distance when breaking heavily of up to 23% was discovered with a properly functioning VCU against 2WD.

in tests, a difference of up to 5% due to front/rear tyre differences was found to have little or no affect on the VCU.
 
Last edited:
I don't think too deeply about it, I just go by results. I've refurbed 3 or 4 VCU's now using the same fluid and it works well enough for my needs.
The VCU is loose enough not to damage the drive train but still spins the rear wheels when the going gets slippy. Is the fluid the same as OEM? Will it transfer the same torque to the rear wheels as an OEM VCU? Probably not but I don't worry about it as it's good enough for me.
 
@GrumpyGel

"the torque is proportional to the 4th power of the effective radius" therefor, to increase the torque capabilities, the effective diameter needs to be increased.... ergo a bigger VCU.


Pressures of up to 100bar may be experienced during operation.


Because of the seal types required, a basic friction of up to 30Nm may be found - ie no load friction.


Higher viscosity siloxanes deviate considerably from Newtonian principles...... ie become "non-Newtonian". It can be shown that they do not shear like Newtonian fluids but create "sausages" of fluid across the plate boundaries which then roll around the plate surfaces.

As the fluid temperature and pressure increases to the "hump mode", the coupling acts in a similar way to a Fottinger coupling - see video from Siemens




A decisive factor regarding the torque progression across the plates is the pressure which is generated. If the pressure cannot reach a critical magnitude, "hump" cannot be induced.


The air chamber is between 5% and 10% of the siloxane fill volume and disappears completely at the limit temperature, causing the pressure increase.


Plates which have been subjected to prolonged heavy loading may deform into a "cup" shape. with wear on one side and next to no wear on the opposite side of the same plate. To prevent this, the outside edge has several "slots" or "nicks" machined into it, to allow the plates to remain stable.

A decrease in stopping distance when breaking heavily of up to 23% was discovered with a properly functioning VCU against 2WD.

in tests, a difference of up to 5% due to front/rear tyre differences was found to have little or no affect on the VCU.

Sorry my bad - the temp causes the fluid to expand and that creates the pressure. It is the pressure though that creates hump mode and 'locks up' the VCU - not the temp as the air gap can be tuned to bring in the hump mode at differing expansion and therefore temperature.

Interesting about the braking.
 
I don't think too deeply about it, I just go by results. I've refurbed 3 or 4 VCU's now using the same fluid and it works well enough for my needs.
The VCU is loose enough not to damage the drive train but still spins the rear wheels when the going gets slippy. Is the fluid the same as OEM? Will it transfer the same torque to the rear wheels as an OEM VCU? Probably not but I don't worry about it as it's good enough for me.
Absolutely, but the theory of what's going on is very interesting :)
 
It does it by magic! No doubt the fluid in the OEM gkn has addatives but we will never know what that are...set it as is right for you needs.
 
Will take a harder look at recent posts as they look very interesting but for now a point I thought of:

Newtonian just means that if you plot viscosity vs shear rate it isn't a straight line with a positive slope through the origin.
A Newtonian liquid in principle would also work because the fluid would get thicker as the VCU plates slip.
A shear-thickening (dilatant) fluid would just get thicker quicker
So the non-Newtonian bit is not really that important?
Edit: I should shut up until I have read that Peschke pdf :)

Alibro: Absolutely! I'm just a terminal research geek who enjoys thinking about these things, but am also a evangelical experimentalist - experimental results are the real world, and if it works then it works.
 
Last edited:
Will take a harder look at recent posts as they look very interesting but for now a point I thought of:

Newtonian just means that if you plot viscosity vs shear rate it isn't a straight line with a positive slope through the origin.
A Newtonian liquid in principle would also work because the fluid would get thicker as the VCU plates slip.
A shear-thickening (dilatant) fluid would just get thicker quicker
So the non-Newtonian bit is not really that important?
Edit: I should shut up until I have read that Peschke pdf :)

Alibro: Absolutely! I'm just a terminal research geek who enjoys thinking about these things, but am also a evangelical experimentalist - experimental results are the real world, and if it works then it works.

I've refurbed them both ways, by cutting the top off, cleaning every disk, welding the top back on then refilling and also by flushing with white spirit and refilling. Neither way is pleasant or quick and if I was doing it again I think I'd just give it a quick flush and stick some fluid in. Unless you intend using the car for off roading it doesn't need to be perfect, so long as it isn't over tight.
But I'm a bit jaded now by having done it too often. The first time I was full of enthusiasm but the last time it was a real chore. :(
Also I'm too thick to be a geek. :confused:
 
What we need is a spy in the GKN plant to take a picture of the label on the cans of fluid they fill the VCUs with.
For all we know they send the shop boy out to McDonalds for a can of used French fry oil.
 

Similar threads