VCUs - The sweet spot

I don't think the temperature of the VCU provides any meaningful info. We know that as the fluid warms up the VCU will transmit less torque to the rear wheels, so if they're warm they will be doing less harm to the transmission than when cold. Once the fluid reaches that magical temp where expansion results in 'hump mode' would be good to know - but we will never know what that is. It is possible that all brand new GKN units achieve hump mode at the same temp (although I doubt it very much) but we know VCUs degrade at differing rates and most VCUs are now recons - so its a rather pointless exercise. Measuring the outside temp of the VCU is even more meaningless.

Similarly, I'm not sure that differences in prop speeds gives the information you need - it won't tell you if the VCU is to tight.

What is needed is to know is how rapidly the transmission is being worn and what is causing it.

The only way I can think of measuring rate of wear is to see how hot the bearings in the IRD (and rear diff) pinion are running. In this case, I would think it is OK to measure the temp on the outside of the casings housing the bearings. Once again, we don't necessarily know what is the temp of an acceptable rate of wear (ie the VCU & tyres are OK) and what is an unacceptable temp. What we do know though is that if it is getting hotter then wear is increasing. Therefore if we take the temp of a known OK transmission and the temp consistently/continuously increases above this - then something is wrong.

If something is wrong, then it is because the VCU is to tight or because the tyres are mismatched. We 'can' measure the tyres by tapping into the ABS monitors. If 1 of the wheels is showing a difference in rotation then the tyres need addressing. If the ABS monitors show the wheels are rotating OK, then the VCU is presumably to tight.

The only spanner in this works would be if, as described in a post above, the VCU would not let a hub rotate at the speed the tyre wants to (due to its size) and scrubs the tyre. In this case, the bearing monitors will show something is up - but the ABS monitor will be looking OK - so the VCU is wrongly accused of the fault. If this is the case though - the important thing is that we have determined that there is a problem (rather than knowing because the IRD has disintegrated) and its down to the owner to determine what the problem is.

Incidentally, when my IRD went, it was due I believe to a tyre down on pressure. I believe I may have driven up to 400 miles over 2 days like this. When the IRD finally went bang, the tyre in question was completely buggered. Most of the rubber had been worn off it, down to wire/steel in places, and it was very hot. I'm not sure if this would happen normally on a tyre down on pressure or whether the VCU did it by dragging it round at a speed that the tyre didn't want to rotate at.

GrumpyGel said:

I don't think the temperature of the VCU provides any meaningful info. We know that as the fluid warms up the VCU will transmit less torque to the rear wheels, so if they're warm they will be doing less harm to the transmission than when cold. Once the fluid reaches that magical temp where expansion results in 'hump mode' would be good to know - but we will never know what that is. It is possible that all brand new GKN units achieve hump mode at the same temp (although I doubt it very much) but we know VCUs degrade at differing rates and most VCUs are now recons - so its a rather pointless exercise. Measuring the outside temp of the VCU is even more meaningless.

Similarly, I'm not sure that differences in prop speeds gives the information you need - it won't tell you if the VCU is to tight.

What is needed is to know is how rapidly the transmission is being worn and what is causing it.

The only way I can think of measuring rate of wear is to see how hot the bearings in the IRD (and rear diff) pinion are running. In this case, I would think it is OK to measure the temp on the outside of the casings housing the bearings. Once again, we don't necessarily know what is the temp of an acceptable rate of wear (ie the VCU & tyres are OK) and what is an unacceptable temp. What we do know though is that if it is getting hotter then wear is increasing. Therefore if we take the temp of a known OK transmission and the temp consistently/continuously increases above this - then something is wrong.

If something is wrong, then it is because the VCU is to tight or because the tyres are mismatched. We 'can' measure the tyres by tapping into the ABS monitors. If 1 of the wheels is showing a difference in rotation then the tyres need addressing. If the ABS monitors show the wheels are rotating OK, then the VCU is presumably to tight.

The only spanner in this works would be if, as described in a post above, the VCU would not let a hub rotate at the speed the tyre wants to (due to its size) and scrubs the tyre. In this case, the bearing monitors will show something is up - but the ABS monitor will be looking OK - so the VCU is wrongly accused of the fault. If this is the case though - the important thing is that we have determined that there is a problem (rather than knowing because the IRD has disintegrated) and its down to the owner to determine what the problem is.

Incidentally, when my IRD went, it was due I believe to a tyre down on pressure. I believe I may have driven up to 400 miles over 2 days like this. When the IRD finally went bang, the tyre in question was completely buggered. Most of the rubber had been worn off it, down to wire/steel in places, and it was very hot. I'm not sure if this would happen normally on a tyre down on pressure or whether the VCU did it by dragging it round at a speed that the tyre didn't want to rotate at.

Click to expand...

Yes I think 1) a simple regular turnip test or even just a cursory touch the prop after a run - 2) regular tyre pressure checks and 3) wear checks - then 4) a three point temperature check as above - and 5) an occasional wheel up test - and armed with a little knowledge of what to expect from these threads, most guys should know when to change the VCU.

Let's keep collecting the data

GrumpyGel said:

I don't think the temperature of the VCU provides any meaningful info. We know that as the fluid warms up the VCU will transmit less torque to the rear wheels, so if they're warm they will be doing less harm to the transmission than when cold. Once the fluid reaches that magical temp where expansion results in 'hump mode' would be good to know - but we will never know what that is.

Click to expand...

Nom 100C - see QFAC page

jamesmartin said:

rrc and p38 v/cs stiffen over time, though it takes longer from my experience but then rr are equally driven front and rear. lr supplied us with test prodecure for v/c for their recon borgwarners but it was a very commonly replaced part because you needed to think of the length of life of the recon unit

Click to expand...

James - I am not familiar with the P38s etc - do they by any chance require different tyre pressures front to rear to cope with the loaded rolling radius question - and in that way aim for a 1/1 ratio front to back without undue stress on the transmission?- bearing in mind that LR say 30psi FRONT AND BACK for Freelanders

GrumpyGel said:

I don't think the temperature of the VCU provides any meaningful info. We know that as the fluid warms up the VCU will transmit less torque to the rear wheels, so if they're warm they will be doing less harm to the transmission than when cold. Once the fluid reaches that magical temp where expansion results in 'hump mode' would be good to know - but we will never know what that is. It is possible that all brand new GKN units achieve hump mode at the same temp (although I doubt it very much) but we know VCUs degrade at differing rates and most VCUs are now recons - so its a rather pointless exercise. Measuring the outside temp of the VCU is even more meaningless.

Similarly, I'm not sure that differences in prop speeds gives the information you need - it won't tell you if the VCU is to tight.

What is needed is to know is how rapidly the transmission is being worn and what is causing it.

The only way I can think of measuring rate of wear is to see how hot the bearings in the IRD (and rear diff) pinion are running. In this case, I would think it is OK to measure the temp on the outside of the casings housing the bearings. Once again, we don't necessarily know what is the temp of an acceptable rate of wear (ie the VCU & tyres are OK) and what is an unacceptable temp. What we do know though is that if it is getting hotter then wear is increasing. Therefore if we take the temp of a known OK transmission and the temp consistently/continuously increases above this - then something is wrong.

If something is wrong, then it is because the VCU is to tight or because the tyres are mismatched. We 'can' measure the tyres by tapping into the ABS monitors. If 1 of the wheels is showing a difference in rotation then the tyres need addressing. If the ABS monitors show the wheels are rotating OK, then the VCU is presumably to tight.

The only spanner in this works would be if, as described in a post above, the VCU would not let a hub rotate at the speed the tyre wants to (due to its size) and scrubs the tyre. In this case, the bearing monitors will show something is up - but the ABS monitor will be looking OK - so the VCU is wrongly accused of the fault. If this is the case though - the important thing is that we have determined that there is a problem (rather than knowing because the IRD has disintegrated) and its down to the owner to determine what the problem is.

Incidentally, when my IRD went, it was due I believe to a tyre down on pressure. I believe I may have driven up to 400 miles over 2 days like this. When the IRD finally went bang, the tyre in question was completely buggered. Most of the rubber had been worn off it, down to wire/steel in places, and it was very hot. I'm not sure if this would happen normally on a tyre down on pressure or whether the VCU did it by dragging it round at a speed that the tyre didn't want to rotate at.

Click to expand...

Through testing we have found the temp of the vcu does give us some indication of how it's working. Higher the temp... the more it is being pushed to slip. Sheering effect generates heat which is averaged into the metal to give us an average temp of the case. They take a while to cool so this would be at the higher end of the average. This temp is higher when travelling faster, when compared to the same distance at a lower speed. A cold vcu isn't being pushed to slip which could be ok but it can equally be the case it's too tight so it won't slip. I don't trust the document which refers to the hump mode as it's years old and newer info doesn't list this. The vcu finds an equilibrium state of slipping as per the doc mhm put up previously.

Difference in prop speed tells you what is actually happening, which is based on the state of the vcu (if it can slip) and if yer FL is pushing it to slip or not. The greater the difference in speed between props the greater the force the FL wants to apply to the vcu.

Why do you doubt original vcu's would not be built to a spec which makes them all the same, or very near the same. Same fluid and known quantity put in would achieve a good tolerance in manufacture.

The ird is hot anyway as the coolant from the engine flows through the plate on the end. The rear diff gets hotter as the vcu it working/fighting harder but from testing it seems this is the point where damage to gears has already started, so too late as an indicator. It's a reaction to a stiff vcu rather than a guide.

Hippo said:

Through testing we have found the temp of the vcu does give us some indication of how it's working. Higher the temp... the more it is being pushed to slip. Sheering effect generates heat which is averaged into the metal to give us an average temp of the case. They take a while to cool so this would be at the higher end of the average. This temp is higher when travelling faster, when compared to the same distance at a lower speed. A cold vcu isn't being pushed to slip which could be ok but it can equally be the case it's too tight so it won't slip. I don't trust the document which refers to the hump mode as it's years old and newer info doesn't list this. The vcu finds an equilibrium state of slipping as per the doc mhm put up previously.

Click to expand...

The "equilibrium state" is the "hump mode". This is theoretically fine in brand new VCUs, as @TheManWithTheHats says 'nom 100C', but at 'nom' 99C the VCU is providing the least torque it will transmit. So you can not say its in 'hump mode' until it clicks over from 99C to 100C. However, 1) these figures are 'nom', so it may be when it clicks over from 98C to 99C and we've just said 99C there is minimal torque - its not progressive. 2) As soon as the VCU drives off the forecourt or off your front drive after being installed, its characteristics are changing, so even if we knew what it was when new, we don't after (nom) 1 month of use. 3) I'm sure each one has differences in operating spec from new, and definitely a recon will have differences to a new GKN unit.

Hippo said:

Difference in prop speed tells you what is actually happening, which is based on the state of the vcu (if it can slip) and if yer FL is pushing it to slip or not. The greater the difference in speed between props the greater the force the FL wants to apply to the vcu.

Click to expand...

I can't see how measuring rotation speed of the front & rear props tells us anything other than their speed and how much difference there is. It doesn't tell us whether the VCU has reached hump mode and it doesn't tell us if the wheels actually want to turn at a different rate to that - ie the VCU is doing its job and speeding up the rear prop/axle or the VCU is to tight and is speeding up/slowing down and axle when it should not. So what is "actually happening" to the transmission has nothing to do with differences in prop speed. The differences will tell you if you have an under/over inflated tyre - but the ABS sensors can tell us this already.

Hippo said:

Why do you doubt original vcu's would not be built to a spec which makes them all the same, or very near the same. Same fluid and known quantity put in would achieve a good tolerance in manufacture.

Click to expand...

I don't know what they are, but I'm sure there are factors in the design and/or manufacture that make VCUs operate differently even when new. My brother had a brand new F1, he said that if he turned right into his road and then right onto his drive the car was almost stalling. That obviously should not occur. There were also reports of VCUs failing very early which could very well mean a difference from new. Given all the things I've seen/read, I'm pretty confident that there are differences from new.

Hippo said:

The ird is hot anyway as the coolant from the engine flows through the plate on the end. The rear diff gets hotter as the vcu it working/fighting harder but from testing it seems this is the point where damage to gears has already started, so too late as an indicator. It's a reaction to a stiff vcu rather than a guide.

Click to expand...

Good point about the coolant heating the IRD. The coolant will dampen the temp of the VCU to around the temp of the engine - ie just over 100C. I'm no engineer, so I don't know what temps bearings run at when under the correct levels of stress and when they are over stressed. If 100C is a bearing under to much stress then my suggestions/thoughts would not work. If a bearing under excessive stress runs at temps greater than 100C then it is a valid test. We need an engineer to tell us - maybe wammers could provide some constructive input (but somehow I doubt it!).

GrumpyGel said:

The "equilibrium state" is the "hump mode". This is theoretically fine in brand new VCUs, as @TheManWithTheHats says 'nom 100C', but at 'nom' 99C the VCU is providing the least torque it will transmit. So you can not say its in 'hump mode' until it clicks over from 99C to 100C. However, 1) these figures are 'nom', so it may be when it clicks over from 98C to 99C and we've just said 99C there is minimal torque - its not progressive. 2) As soon as the VCU drives off the forecourt or off your front drive after being installed, its characteristics are changing, so even if we knew what it was when new, we don't after (nom) 1 month of use. 3) I'm sure each one has differences in operating spec from new, and definitely a recon will have differences to a new GKN unit.

I can't see how measuring rotation speed of the front & rear props tells us anything other than their speed and how much difference there is. It doesn't tell us whether the VCU has reached hump mode and it doesn't tell us if the wheels actually want to turn at a different rate to that - ie the VCU is doing its job and speeding up the rear prop/axle or the VCU is to tight and is speeding up/slowing down and axle when it should not. So what is "actually happening" to the transmission has nothing to do with differences in prop speed. The differences will tell you if you have an under/over inflated tyre - but the ABS sensors can tell us this already.

I don't know what they are, but I'm sure there are factors in the design and/or manufacture that make VCUs operate differently even when new. My brother had a brand new F1, he said that if he turned right into his road and then right onto his drive the car was almost stalling. That obviously should not occur. There were also reports of VCUs failing very early which could very well mean a difference from new. Given all the things I've seen/read, I'm pretty confident that there are differences from new.

Good point about the coolant heating the IRD. The coolant will dampen the temp of the VCU to around the temp of the engine - ie just over 100C. I'm no engineer, so I don't know what temps bearings run at when under the correct levels of stress and when they are over stressed. If 100C is a bearing under to much stress then my suggestions/thoughts would not work. If a bearing under excessive stress runs at temps greater than 100C then it is a valid test. We need an engineer to tell us - maybe wammers could provide some constructive input (but somehow I doubt it!).

Click to expand...

If we don't use prop speed as an indicator then we don't have much else to go by. Hot vcu's are a warning and sometimes past their best if they get too hot. Differing prop speed tells you the FL wants it's props to turn at differing speed and the vcu will allow this. How much they differ depends on factors like tyres, tread pattern (enforces more grip), inflation, size... all the usual things we've become accustomed to checking. Being able to spot props not turning at differing speeds or suddenly changing from differing speed to same speed would advise the vcu has activated. It's not a precise science and those above are going to have some fun getting it working [evil laffin] but it is a way to investigate what's going on at higher speed. That's the difficulty with most vcu stuff we do. Minor slipping at jogging pace is one thing but the greater effects of the same amount of slip per mile, but done so at higher speed (hence in a smaller amount of time), is what we're not sure of. By using heat as a measurement we know that doing the same trip at a higher speed, will heat the vcu case more, than the same trip at a lower speed. I'm talking of a vcu possibly passing the one wheel up test ok but struggling at speed when it's essentially being turned the same way (be it when mounted in the transmission) but it's requested to slip faster due to higher speed and constant operation. You don't get a kick or feel the vcu activate. I get the feeling the graph isn't as sharp to chance by 1 degree and then it activates. If it does then there's some magic in the fluid to have a knee response like that. I still believe it varies slightly as it goes from slipping, to fighting more, then activated. I'm trying to prove this with me new test but it's difficult to catch on film with controlled speed increase in prop difference. I will continue trying. I hope to be able to find the point at which we see a change in revs which makes the vcu change from slipping to activated. I can now apply a constant force to a vcu as if it were one wheel up tested continuously but faster. Ideal outcome would be to say a new (and nearly new) vcu will allow slip of N revs. Then compare this to aged vcu's to see if there's a bench mark we could work too. Failing that it's the nearest car park and some builders gasps while moving back and forth when turning.

A new vcu is a new vcu. I have one and tested it when new, and after hundreds of miles. The bench test and one wheel up test gave the same results when filmed and played back slowly to get betterer timing. Yes there is a degrade over time as it ages by factors we're not fully sure of (you can kill it with hi temp quickly from failures we've seen), but the original manufacturing solution must have a tolerance when made. It's not going to be anywhere near say 25% but it would be a lot smaller tolerance. Metal is easy to chop up and the fluid will have a mix that suits the application. The build spec must be able to cover operating tolerances comfortably or they would cause trouble much earlier in their lives.

Prop speed and temp are the way forward from the testing I've done. Real life test if you like. Being able to say it must be able to turn N times a minute as a minimum or not hit a certain temp while doing so, is a possible outcome. Ideally we would like something which defines if a vcu is good or beyond good service. That was a question posed years ago. We may never get this, but the current testing I'm doing will be the last as I can't see any other way forward other than this. Results on one wheel up testing are just getting to the volume we should of had 4-5 years ago.

Regarding bearing temp. Can you ask the peeps on the engineering forum. I never bothered to join. Just read what they put as a guest. they may know or point us to theory. You could ask the a hole but you'll have to polish his ego first.

If it is easier to add speed sensors to the prop shafts rather than reading the existing ABS sensors then that's fine - they will both give the same info about VCU inputs and axle speeds.

I agree you could measure speed differences between the axles/props and warn if the difference reduces. You'd probably need some sort of steering position sensor or (sidewise) gravity sensor as well to know that you are checking in a straight line or applying the correct algorithm depending on turn rate. Unless you are just looking for a sudden stop in difference, it would need to recalibrate itself constantly due to differing loads on different journeys and altering responses as tyres warm etc. You would also need a noticable difference in wheel diameter unless you were taking into account turn angle - and you could then only test on turns.

I agree you could try it, I don't know if it will work though! Unless the props do suddenly stop rotating at different rates I think there's so much that's unknown.

So I've moved a bit in my thought that measuring wheel/axle/prop rotation rates/differences might help, but I still think the underlying test must be how much stress/overload it is putting into the IRD/diff.

The easier place to measure is the abs sensors as there's several "marker" points for each turn of the wheel. I measured counts per time period. The period kept restarting so I received a totalised count for each time period. We're talking periods of seconds but it was adjustable. Measuring the wheels means you can tell if yer not steering straight as the wheels travel at differing speeds so the number of markers counted to each wheel on the same axle pair will be different. There's probably interesting info there which is valid but the time to mess with the data to understand it and abstract what we want is considerable. I didn't see the pattern I was expecting when wheel spinning on grass which confused me so I put this to one side for a bit.

My interest has always been to be able to pin point a vcu being considered good or bad. Theory being we could have a bench mark value from a test that others could try themselves. Reconners have a vested interest in selling recons so when peeps talk to them when looking for a recon they're always told our work is pointless and rubbish. But there has to be something of value ere.

From what I see the vcu has 2 states of operation:

• Activated = when it's in it's seized or betterer put it's equilibrium state of passing the majority of the power through it.
• Resistive = when it's not activated (when it's sat doing nothing or not turning fast enough to activate)

So far the One Wheel Up Test has measured how fast the vcu turns with a known torque applied. This is measuring the vcu during the resistive phase. What I describe as the resistive phase is the bit that leads up to the activated phase, but stops being resistive as it activates. We can calculate the force in Nm from the weight and length of weight to pivot. From what I see the heavier the weight, the onger the bar from pivot point, the greater the force applied, and the faster the vcu will turn. It's to my understanding that we could continue putting heavier weights on as we go further up the resistive phase until the vcu is being turned fast enough for it to activate. I think the vcu will continue to resist the force applied right up to the activated state. With this in mind I ask the question:

• Does the required force applied to the vcu to make it turn by the same amount, increase as the vcu deteriorates? From OWUT results we think it does as we see vcu's become more resistive to turning across themselves when testing failed vcu's. Hence test times increase if the vcu has degraded.
• Does the required force applied to the vcu to make it turn fast enough for the vcu to activate, lower as the vcu deteriorates? Hence for eggsample a vcu needs to be turned at N revs per minute or greater to activate. If we could find the figure N revs and measure a duff vcu to compare... we would see if the activation phase happened sooner. Hence applying more pressure to the transmission if activating sooner than it should whilst fitted to yer FL1.

From my thinking there's 2 forms of stress in the FL1 transmission:

• Stress from the active phase of the vcu where a lot of force is passing through the transmission to push/force the rear wheels to turn when powered by the engine/gearbox/ird.
• Stress during the resistive phase where the FL1 happily drives around normally when the vcu doesn't activate as it doesn't get enough turns across itself to make it. It's during this phase the vcu resistance varies depending on driving straight/turning, normal weight or when heavier due to loading, or when tyres are different due to make/model/size/pressure not being correct and the same.

It's often said a vcu is causing more stress in the transmission due to activating sooner as something is wrong (tyre problems etc), or the vcu has degraded and therefore activating sooner than it would if it were in good condition. It would be nice to know if this is really true. If so then I would expect the number of revs required across the vcu to activate it to reduce as the vcu degrades. Hence activating sooner, causing more stress than average, which will blowed something up if left undetected. It may just be a simple case the typical resistance across the vcu during the resistive phase is much greater if the vcu has degraded, and this provides enough additional transmission stress to cause problems without the vcu activating.

It's during the resistive phase we see the vcu stiffen if the vcu has degraded. Hence when stiffened the vcu is applying a lot more transmission wind up than normal, which can't be good for the transmission. I'm wondering if in this degraded state... does the point at which the vcu activates move. For eggsample a good vcu activates at T Nm and a bad one at 75% of T or less. Hence less torque required to activate if degraded. Or similarly less turns of the vcu across itself to activate if degraded. That's where my current testing is going. I can now apply a constant force to a vcu which is greater than what we've done with the OWUT test so far. I need to re-film it again but I think I'm very close to seeing how many revs I can turn a new/good vcu before it's got enough turns across itself to activate it. Mad I know but I get the feeling I may be at the frontier of the workings of the vcu and able to report back a measured value which represents where the activation phase is with video proof,








or a video of it going horribly wrong we can laugh at.

I don't like to throw a spanner in the works, especially with all the testing that you have done Hippo. However I'm a little concerned that you might be looking for a figure that simply isn't there to find. I'll try to explain how I perceive the VCU functions. The fluid is a special fluid that hardens as it is agitated. This is call a non-Newtonian fluid. This fluid hardens locally between alternating plates inside the VCU. Now from the research I've done over the last few years, the VCU doesn't lock or "activate" in the same way as a clutch for instance. It works more like an auto gearbox torque converter. Essentially the more torque that is applied the less it slips so the output torque increase. This only happens up until the torque / slip equilibrium is reached. So in effect, a correctly working VCU will always allow an amount of slip, even if it's taking plenty of torque. It has to have some slip for the fluid to be agitated. If there is little agitation, drive transfer can't take place. It's these very properties that make the VCU so good at what it does. I believe that the fluid only has a limited life as a non-Newtonian fluid before it's properties start to degrade. As fluid degradation takes place, the fluid starts to thicken, transferring more torque than it was designed to do. So basically it's slip/ torque equilibrium shrinks causing increased torque transfer. This is where the Freelander's design foibles come into play. As we know, LR in there wisdom geared the front and back slightly differently. This forces the rear half of the VCU to over run, or drive the front half. As the VCU ages, it's less likely to allow for this gearing mismatch, so starts overloading the gears in the IRD and rear diff. Eventually something it the drive train goes bang due to the sustained added stress.
There appears to be only 2 reasonably reliable tests for the VCU, ideally used together for improved accuracy. The OWUT and VCU temperature measurement. The OWUT loads the VCU while monitoring the slip rate, this is a valid test as far as I can tell. Meanwhile measuring the temperature is also going to give an indication of the VCUs ability to slip. If it stays cold, no or at least little slippage is occurring and the drive train is taking a pounding. The VCU can't actually seize solid or the wheels would skip and something would break very quickly. Wheel skipping does happen when turning on a tight lock with a failed VCU.
I hope I've not ruined all your hard work but this is how I perceive the VCU function and how we can check it at home.
I could be completely wrong of course.

We're all thinking here, but the science of VCUs shows that as they heat up (through shear), the amount of torque transmitted reduces. This is confirmed by MHM's (original?) graph taken from a VCU on a dyno. It was on the thread https://www.landyzone.co.uk/land-rover/the-definitive-freelander-vcu-testing-thread.99163/ but is no longer showing. It is only once heated sufficiently to increase pressure to a tipping point that the amount of torque transmitted rapidly increases (so rapidly that it in effect locks up). Although MHM's graph isn't available any more, this link concludes an identical result, see fig 8.1 http://www.easy2design.de/stuff/visco_sae.pdf

If we go back to GKN's doc on VCUs http://www.gkn.com/driveline/about-us/Documents/datasheets/Viscous-engl.pdf although it doesn't say much, it talks about "degressive locking characteristics (viscous mode)" and "hump mode" - akin to what Hippo describes as "resistive" and "active" - but the resistive is "degressive".

I think we should take the above as fact - there are a lot more people who have put a lot more time and scientific effort into concluding this than us - and their results all agree.

Our problem unfortunately is that these people have not studied degrading VCUs and how their characteristics change. I tend to agree with Nodge, degraded fluid, and I believe contamination from worn plates is as bigger influence, will make the viscous mode transfer to much torque, generate less heat, be less degressive and possibly never reach hump mode because its already transmitting to much torque - that's if the degraded fluid does still have a hump mode characteristic.

Going by that synopsis, the 1 wheel up test is probably as good as we are going to get at testing the VCU. We don't really know what temps to expect and how the outside temp of the VCU relates to the inside.

My thoughts are still that if there is better testing methods to be created, then it is of the impact of the VCU on the IRD & rear diff. But, that's just my opinion.

Thinking about it, measuring the temp of the tyres might give you a good indication of wind up. @dfossil talked about tread moving around to compensate for forces and I've watched enough F1 to know that means the tyres get hot

Nodge68 said:

I don't like to throw a spanner in the works, especially with all the testing that you have done Hippo. However I'm a little concerned that you might be looking for a figure that simply isn't there to find. I'll try to explain how I perceive the VCU functions. The fluid is a special fluid that hardens as it is agitated. This is call a non-Newtonian fluid. This fluid hardens locally between alternating plates inside the VCU. Now from the research I've done over the last few years, the VCU doesn't lock or "activate" in the same way as a clutch for instance. It works more like an auto gearbox torque converter. Essentially the more torque that is applied the less it slips so the output torque increase. This only happens up until the torque / slip equilibrium is reached. So in effect, a correctly working VCU will always allow an amount of slip, even if it's taking plenty of torque. It has to have some slip for the fluid to be agitated. If there is little agitation, drive transfer can't take place. It's these very properties that make the VCU so good at what it does. I believe that the fluid only has a limited life as a non-Newtonian fluid before it's properties start to degrade. As fluid degradation takes place, the fluid starts to thicken, transferring more torque than it was designed to do. So basically it's slip/ torque equilibrium shrinks causing increased torque transfer. This is where the Freelander's design foibles come into play. As we know, LR in there wisdom geared the front and back slightly differently. This forces the rear half of the VCU to over run, or drive the front half. As the VCU ages, it's less likely to allow for this gearing mismatch, so starts overloading the gears in the IRD and rear diff. Eventually something it the drive train goes bang due to the sustained added stress.
There appears to be only 2 reasonably reliable tests for the VCU, ideally used together for improved accuracy. The OWUT and VCU temperature measurement. The OWUT loads the VCU while monitoring the slip rate, this is a valid test as far as I can tell. Meanwhile measuring the temperature is also going to give an indication of the VCUs ability to slip. If it stays cold, no or at least little slippage is occurring and the drive train is taking a pounding. The VCU can't actually seize solid or the wheels would skip and something would break very quickly. Wheel skipping does happen when turning on a tight lock with a failed VCU.
I hope I've not ruined all your hard work but this is how I perceive the VCU function and how we can check it at home.
I could be completely wrong of course.

Click to expand...

No worries about questioning me thoughts or what I've tried/done. All input is good input but don't tell the ass hole.

Some time ago I wrote this which sort of says they same thing as you above.

Hippo said:

...
As the ratio of speed of front/rear prop increases, the sheering effect in the vcu become greater. The differing plates rotate at differing speeds, and this speed increases as the ratio increases. The sheering effect creates heat which has an effect on the fluid in the vcu. The greater the heat the more the fluid inside the vcu wants to grab the plates and reduce the differing speed at which they turn at, in comparison to each other. Hence the vcu starts to seize, or betterer put, the resistance across the vcu becomes greater and the vcu seizes momentarily if the ratio is great enough to cause it too. When the vcu is in this state the lack of differing speed of the differing plates means the sheering effect reduces and the fluid cools. This cooling effect allows the vcu to slip more as the resistance across itself reduces. This causes the speed difference between the props (and hence the inner vcu plates) to increase again. The vcu's plate don't fully release their grip as the fuild in the vcu will cool until the ratio of differing speed between the props (and hence the plate) increases to an extent where the sheering effect causes the fluid to heat up again, and the resistance across the vcu increases.
It's this varying resistance across the vcu which is a reaction of the fluids ability to rapidly warm up and cool down inside the vcu, which is why the vcu works so well. If you can mentally visualise the rapid alternating from "some resistance" to "more resistance" to near or seized, and back again due to the expected drop in resistance thereafter (due to cooling when the plates get closer to a 1:1 ratio), then you can appreciate why the vcu has the ability to vary the resistance across itself. It's the rapid heating/cooling which is proportunal to the changing ratio applied to the vcu, which varies the resistance (and therefore drive) across it. It's this ability of varying resistance which is key to the vcu reacting to conditions (ratio of drive across it) to achieve a state of equilibrium resistance, with respect to the ratio across it. This equilibrium state varies as the ratio across it varies. Hence the vcu varies the resistance across itself, whilst monitoring/reacting to change of ratio across itself. On an "as and when required" basis. It's when this goes wrong the vcu feks up.

Click to expand...

When looking at me graph below of me OWUT results you can see the results vary with weight, or betterer put force. No big surprise there. I always wanted to be able to apply more force to see what happens well past 10kg (workmate holding one end stationary gambols and the weight falls oft). Reason for this is the graph starts to straighten at 8+ kg onwards. Looking at it now I can't help thinking most peeps tend to measure with less than 8kg. It's in that section of the graph the value of time/weight varies far more, compared to the 8+ kg area where it's nearly proportional. I'm starting to think there may be too much variation in the lower section of the graph for results to be reliable on comparison, but can't confirm this as we don't have enough test data. It may just be a case there's a wider tolerance in the lower area which can be ignored as it's "early days" on vcu resistance. I still think there's a cut oft point when the bar may stop turning or rapidly slow down if the vcu starts to activate.

GrumpyGel said:

...
Going by that synopsis, the 1 wheel up test is probably as good as we are going to get at testing the VCU. We don't really know what temps to expect and how the outside temp of the VCU relates to the inside...

Click to expand...

This is essentially the crux of the problem. In early days of trying to do the OWUT not many took part. Over the years those that did have left so getting them to redo the test after some years to see if there's a change (degrade) is difficult. Far too many arguments were had about the quality of data retrievable from the results, coupled (pun) with peeps trying to tell us there were too many factors stopping the test from being reliable. In reality it's the best we have so far.

The Turnip Test (named after vagrent) measures the heat of the outer case. I would agree this isn't the most reliable heat measurement but I contemplate the following:

The vcu creative heat inside when in normal use. If used for a period of time that heat will eventually spread to the outer case. The average temp around the case will be about the same. The case doesn't really have any cooling ability other than convection to air. This is speeded up when driving due to air flow but when driving it's propbably heating up faster or certainly enough to keep it warm anyway.

It's often been said in the past a vcu at 70 degrees on the outside is dead, duff or highly suspect. 70 on the outside is probably lower than the inner temp which is assumed to be way above that of the fluid would allow, hence causing the fluid to fail. So 70 degrees was said to be the limit. When working the vcu is probably ok to generate enough heat to overcome cooling so it will stay warm. Hence after a period of use (3 miles is enough) the outer temp of the vcu will be proportional to the inner temp. If not the same then more likely a % of the heat inside. I think this is representative of the heat inside. Enough to be considered reliable. It's easy to change the tyre pressure to get the vcu to warm up more as it reacts to a greater diffing of prop speed but I don't advise peeps try this.

Hippo said:

Click to expand...

When I was calculating the equivalent OWUT on your Measure Your wheels thread, I loaded your data of time against weight from your YouTube vid into Excel. Graphed, it is a very consistent curve as in your above example. I then did a search to see if I could turn that data/curve into an equation. If I had an equation I could then calculate the OWUT weight from X number of VCU slips per minute. I failed to find a way of converting it into an equation - Google was not my friend. However, all you have to do is some simple comparisons on your data and it quite quickly becomes obvious that for every doubling in weight, the time taken reduces by 32%. That's why the curve flattens, ie from 4kg to 8kg it reduces by 32%, as does 8kg to 16kg and 16kg to 32kg.

If you put enough weight on the bar to achieve hump mode, the bar would definitely not stop - hump mode is activated by pressure due to temperature generated from shear, therefore if the bar stopped there would be no shear and the fluid would cool and the bar move again. Its possible any additional weight above this point might result in the same times (ie the curve becomes a straight line), or (more likely) the times then start to increase again and the curve starts going up.

Hi Grumpy

Excel is your friend here - it allows to do a correlation calculation using a predefined curve shape (linear, square, log etc)

Or you convert the curve to a square (root maybe), log etc and do a linear regression

digby said:

Hi Grumpy

Excel is your friend here - it allows to do a correlation calculation using a predefined curve shape (linear, square, log etc)

Or you convert the curve to a square (root maybe), log etc and do a linear regression

Click to expand...

G'day my Anglo French travelling buddy

Summer's starting here, weather people are warning of an El Nino summer of hot temps and droughts... 8 degrees and drizzle today!

I thought Excel would be able to do it - but not knowing the terms I was stuck - I think that little Microsoft Dog that used to pop up with help must have gone to the pet cemetery! So I turned to Google searching for stuff like "convert data to algorithm" etc - but failed miserably. If Hippo posts some more data, I'll know where to start looking now though, ta

if you cant find what you are looking for then email me your spreadsheet

Autumn is starting here - i'll send you a pic

pic on the way

That's a very interesting graph... can be modelled using an exponential if someone can really be bothered.

I would theorise that the VCU was never meant to completely lock, just carry on through that curve until it hits the point of eventually flatlining. I have a feeling that regardless of how much weight you put on it it would stabilise at about 6-7 seconds ish because at the end of the day you're still dealing with fluid between two plates. It would have to turn to concrete to get a full lock and it'll never get there, even when it's a non Newtonian fluid.

We need to get @Alibro to replicate this graph with his completely rebuilt VCU using only silicone fluid - if the results come out similar I'd bet good money on there being no real witchcraft involved. Purely basic fluid dynamics...

But I agree with @Nodge68 in that this is a very touchy-feely kind of diagnosis - I don't think we'll ever nail it down to more than a 5kg:
30-45s all good
45-60s think about changing soon
60s+ daaaaaaanger zooooonee

There are just too many factors involved, as we saw in your other thread, VCU slip varies massively car to car even when people are being responsible with their tyres keeping them well inflated and of the same make and tread.