BMW/Titan LSD Viability

[Wrightpayne]

It was a number of years ago now and I expect the majority affected have been replaced now. Basically the lower shock mount peeled out of the dedion tube. I just changed mine for peace of mind. There is quite a bit in the archives and always worth a check while you are underneath. The upgraded tubes have a web welded between the main tube and shock mount.

Regards

Ian

 

[ScottR400D]

I did have correspondence with CC on the subject of LSD and expected life. They were pretty non committal but confirmed they need regular servicing to avoid failure. 

By regular servicing, they mean changing the plates and washers, not just oil, so basically a rebuild. I understand that will be £5/600 but if there is any risk of bearing damage or they are changed during the service it’s £900+. As I know to my cost!

That's if you avoid CWP damage of course, then you can double that cost.........

[ScottR400D]

#47 David, I for one would be very interested in your findings if you overhaul your diff, especially if it is still functioning and has appreciable pre load before you strip it. 

One point that’s not been majored on is the Belleville washers. I’m told that in the Titan they operate from very little deflection to 100% deflected. I’m not sure how accurate that is but it would explain the apparently  common failure of these components. 

It’s accepted good practice that Bellevilles should operate between 15% and 75% of their full deflection. If as I’m told they’re operating up to 100% (ie flattened) they aren’t going to have a long life. 

[David_Long]

I measured it this morning and it was around the 28 ft lbs mark so by the figures on here, it's near its 'overhaul' point.  The Bellevilles spring deflection is interesting point as my understanding of how this diff works is that it effectively locks solid when high torque is applied. That would almost certainly result in 100% defection. But that brings me to one of the underlying themes of this thread that I am uncomfortable with: the design is really intended for a motorsport environment where the plates spend much of the time clamped together under full throttle high torque conditions. If these diffs are in a road car environment, then they spend much more of their time operating at the preloaded levels that a) don't provide much LSD function, and b) the wear is likely to be very high.

[CtrMint]

Please correct me if I'm wrong, but I'm understanding from the thread;

• every 2 years
• approx 3000 miles, 
• the Titan diff will require a rebuild at;
• approx 800GBP

If I'm wrong could you correct the figures to what people are seeing and repost.  Thanks.

[ScottR400D]

Mark, I think it's hard to be exact on mileage, which I would think is more relevant than time. Clearly some people get their diffs to last to much higher miles than others. I guess it's a function of power and torque and where and how the car's driven.

I would think most 200 bhp + cars would fall in the area of  5-8000 miles though some will wear out before 5000 and others will last a lot longer. All this is why regular checking of the preload is important and is why I'd love to know what David finds out when he strips a diff with a still reasonable preload.

I stand to be corrected but I think RRT charge about £200 + VAT + a rebuild kit at about the same sort of money, so £500 for a basic service. If you need bearings it adds to that. I would think bearings come into the equation once bits have started falling off plates and washers. Mine needed all new bearings, hence over £900.

[ScottR400D]

#54 David, again I see what you mean but I don't think the Bellvilles need to be flat to provide a full lock up. If they're sized correctly they should be able to provide enough force to lock both output shafts without being flat, shouldn't they?

The washers just need to supply enough force to cause the plates to lock together. 

[aerobod - near CYYC]

What is needed in the design is a limiting spacer on either the inside or outside of the Belleville washer (depending if there is any room) to stop the washer from being fully flattened by stopping the plate pack when the washer reaches about 75% flex. I will have a look in my diff when I do the rebuild this winter to see if there is any way to do this.

I also plan to switch from carbon to sintered plates and install a ferrite disc magnet somewhere low on the back of the cover with a small aluminium holder screwed into the cover to hold the magnet in place, or alternatively I may look at inserting a magnetic drain plug if there is enough meat to thread the cover in the right place. This will allow debris collection as it settles within the oil, as opposed to the collection from oil splash on a magnetic fill plug.

[7 wonders of the world]

That's one of the design flaws in the Titan the Bellevilles are permitted to move beyond their deisgned limit, they are designed to work and exert the design pressure between 25 - 75 % of total deflection.

This also means the ramp blocks are also moving further than they need to, so when the Bellevies finalaly fracture the blocks bang like a barn door in the wind.

The Bellevilles only provide the preload, the torque applied to the cross pins forces the ramp blocks out and apples load to the clutch packs which limits the slip.

[aerobod - near CYYC]

If there are shims needed to reach the initial preload, it might be possible to remove some of the shim thickness and insert a 1mm thick top hat behind the Belleville washer with a boss that is 2 or 3mm thick that extends through the Belleville bore and contacts the plates when the Belleville reaches 75% deflection. Depends on the size of the Belleville centre hole vs the diameter of the side gear shaft, plus leaving an allowance for the Belleville washer centre bore to contract when it is compressed.

Another possibility is just to place a thin ring around the casing bore behind the Belleville washer, so that it is blocked from reaching full deflection, as long as the washer is a fairly close fit to the casing outer diameter.

[7 wonders of the world]

if the LSD is correctly designed and built that wont be needed.

The problem is when you start adding and subtracting shims you also alter the travel of the ramp blocks.... which can lead to over working the Bellevilles.

You need to look at changing the Bellevilles for springs with the correct rate, and use shims to control the movement in the ramp blocks.

 

[ChrisC]

Remember that's based on carbon plates, fitted post 2014/2015(is) to early 2018.  Your 420R doesn't have carbon plates, that's why your diff tends to bang, according to RRT, but can be improved that with their special blend oil.  Your 620R, I believe, doesn't have a BMW Titan LSD. I think 620Rs are fitted with Ford based Titan LSDs, but I don't know what plates were fitted.   

If you check and monitor your diff's health, and rebuild when the preload is no longer within specification, you will minimise the rebuild cost.       

[7 wonders of the world]

The design of the Titan is the same irrespective which final drive it has been adapted to fit into.

The problem is the LSD not the final drive.

[ChrisC]

I believe the internals are the same (based on the Titan website), and maybe its because the week carbon plates only started post 2014, based on post #34, so long after the swap from Ford to BMW diff by CC.  What I haven't ever seen is complained about the Ford Titan LSD unit.  

[aerobod - near CYYC]

Re #61, the shim thickness plus any Belleville limiting device thickness would stay the same as the original shim thickness, so wouldn't have any effect on the ramp block travel. The total thickness of anything inserted into the casing to alter the Belleville preload would be adjusted to meet the spec (which is 40lbft for a Caterham, I believe. Based on the Titan specifications doc, they say 20-80lbft is used, depending on application).

[7 wonders of the world]

Understand that James however if the correct spec of Belleville is selected you don't need a limiting device as the travel is limited when the ramp blocks contact the clutch plates. 

[CtrMint]

Yeah the 620R and 420R diffs are like chalk and cheese, makes you wonder why bother with two specs.  620 has mild noise, the sort I'd expect, unlike the 420R which is like a hammer on anvil!  Surely the 620 would be more brutal.  I've given up looking for logic with these cars.

I keep meaning to change the oil Pete convinced me to do so, it's just on the list of things to do.

[aerobod - near CYYC]

Re #66, I'm not saying that putting in a Belleville that has a greater travel range isn't the best solution, but would probably require a new housing with an LSD redesign. With the constraint that more space for a larger Belleville isn't available and producing one may be difficult, too, a limiting device that is easily machined and can be fitted to the diff by anyone with a few handtools is what I'm suggesting, this is what I'm going to investigate when I do my diff rebuild, it may be as simple as cutting a thin ring out of a sheet of aluminium and putting it between the Belleville washer and the housing.

What I will probably do is create a couple of adapters to fit into the drive splines so that I can take the LSD and support it out of the diff via one adapter in a vice and apply torque to the other adapter (after sealing the holes and filling with enough oil to coat the plates). I can then play around with shims and limiters to see if I can get a decent lockup from the initial 40lbft preload up to when the limiter stops the travel, rotating the casing to open the ramp side rings to increase the plate pressure.

BTW, for those of you who think it is beyond your capabilities, these diffs are very easy to work on and it is easy to get at the LSD to disassemble. It is more of a chore to remove the diff from the car. You just need two new side seals, large circlip pliers for the side clips (don't mix them up, as they come in various thicknesses to set the backlash and mesh position, but don't need changing if you are only servicing the LSD) and a torque wrench with a few different sizes of socket.

[ScottR400D]

I'd love to know what LSD is in the 620. I'm told it's the same Titan as the other cars but I'm not convinced. 

Firstly, from my experience the same Titan LSD as used in the BMW housing wouldn't last two minutes with the torque of a 620.

Secondly, they can't be the same because the crown wheel is different so there must be two, and Titan do sell another LSD for the Sierra diff, what they call the traction Master. Images of it look completely different to what is fitted in the Caterham BMW housing.

Who knows? Not me, but it wouldn't surprise me if the Titan we're all discussing is fitted into the BMW because the one that fits in the Sierra housing is too big for the BMW assembly and the whole Ford assembly with a different LSD is used on the 620 because the BMW assembly is no good at all for that car. 

 

[CtrMint]

I'm decat'in the 420 tomorrow so will be the garage.  I'll shoot some video of the diff housing, maybe there some visual differences. 

[7 wonders of the world]

Hi James understand the concept but I think you will fnd its not that simple, as by adding a shim you will be also limiting the ultimate travel on the ramp blocks and heir ability to clamp the clutch plates as these must only be receiveing full compression when the belleville is flat otherwise they wouldn't be failing.

Problem is the titan allows the ramp block to move circa 2.5mm !

Belleville in the UK were very helpfull in doing rate calcs for me on the OE ZF springs compared to the ones I got from them, I ended up using lighter springs in stacked in parallel to achive my desired preload.

I bought some packs of steel shims in the correct OD for my ZF from ebay from in 0.1, 0.3, 0.5 mm thickness, cheaper than getting them cut.

 

[ScottR400D]

#70. Visual differences 420 to 620? Stevie Wonder should spot those, Mark! One should be Ford the other BMW! 

[aerobod - near CYYC]

Problem is the titan allows the ramp block to move circa 2.5mm !

If they are moving that far with only 4 clutch plates, I can't see how a 40lbft preload would translate into not locked enough until a further 2.5mm of pack compression has taken place with the sintered plates, unless a lot of wear has occurred and the preload has been lost. This is perhaps why the carbon plate packs have been failing if there is too much compressibility in the friction material combined with excessive wear, leading to over flexing of the Belleville.

Perhaps the issues will be solved by less compressibility of the sintered plates with less wear and the right preload, leading to less range of movement in the Belleville washer that prevents premature failure.

[aerobod - near CYYC]

#69 Peter, The BMW LSD from Titan is also called Traction Master. They come in 3 sizes (med, large and extra large) and otherwise have the same design, with the only options being a range of ramp angles and Sintrak vs Carbon XT clutch plates. The Titan rebuild kit is the same for the BMW 168 and Sierra 7" installation, so it seems both use the same medium sized Traction Master, the only difference being the splines in the side gears and/or the crown wheel bolt pattern and spacing.

[ScottR400D]

Thanks James.