19 posts / 0 new
Last post
ScottR400D
Online
Last seen: 6 min 36 sec ago
Joined: 23/08/2015

That all makes sense James, you’ve clearly thought it through. 

Geoff Brown
Geoff Brown's picture
Offline
Last seen: 3 days 1 hour ago
Joined: 17/04/2014

Aerobod -  when the thread is updated post Titan rebuild & refit PM me. I will then struggle with the IT demons to move the info.

aerobod
aerobod's picture
Offline
Last seen: 4 hours 35 min ago
Joined: 17/04/2014

My rebuild kit arrived today, not bad, as it was only shipped yesterday from Caterham and it had over 7.000km to travel in it's delivery.

A few differences besides sintered vs carbon plates:

- No Belleville thrust springs for the side gears, instead a thicker static thrust washer.

- Side gear thrust washers are now flat instead of recessed on the side that faces the gear and steel instead of aluminium.

- Thinner shims for behind the side gear thrust washers (I will have to measure to see if they are needed).

- Plenty of shims that will allow 0.1mm increments on each side of the clutch pack from 0.3mm to 1.8mm with either one or two shims either side.

James

aerobod
aerobod's picture
Offline
Last seen: 4 hours 35 min ago
Joined: 17/04/2014

I added a drain plug in the lowest part of the casing today, Most of the work was getting the diff clamped down in a firm enough manner and ensuring both the internal and external pinion seals and bearings and any passages and blind threads were well covered to protect them from swarf.

I used an M12x1.75 magnetic plug that is normally a GM fit on a whole raft of vehicles, so is commonly available, it is from a company called Dorman, who produce decent quality aftermarket parts:


 

As I didn't clamp the diff down really rigidly (I would have had to make a jig), I didn't want to use a mill cutter on my general purpose bench top lathe / mill / drill press, so I just used a grinding wheel instead. I think anyone with a drill press where you can lock the height of the chuck would be able to use a similar technique to grind a flat spot by taking off about a quarter of a millimetre at a time. I moved the table with the diff clamped to it while keeping the drill head in the same spot:


 

I picked the thick part of the casing between the two lower holes that retain the cover. I was also looking at putting it further back, but any suitable relatively flat areas near the bottom are a bit close to the crown wheel and the casing would only be about 8mm thick after machining a flat, which is a bit thin for an M12x1.75 thread. After machining the flat, I drilled a 10.8mm hole (after centre drilling and pilot drilling 6mm first), then tapped it:


 

The drain plug I used has an integral rubber seal, but a crush washer would work, too. I torqued it to 30Nm, which is a common torque used by GM for this plug on a lot of cars:


 

The thread is exactly the right length for the casing thickness, leaving just the magnet protruding through the casing at the lowest point:

James

aerobod
aerobod's picture
Offline
Last seen: 4 hours 35 min ago
Joined: 17/04/2014

Today I did the LSD rebuild and re-install in the housing.

I took a few measurements of the new component thickness:

Belleville washers - 3.80mm uncompressed, 2.00mm fully compressed (material thickness, I didn't want to risk damaging them by compressing to 100%). Effective thickness as described below is 3.40mm, (x2 = 6.8mm).

Sintered friction plates - 2.08mm (x4 = 8.32mm)

Steel clutch plates - 2.00mm (x4 = 8.00mm)

Ramp "cups" with gears installed from face-to-face of the clutch surface on both cups - 56.00mm

Side gear thrust washer with backing shim - 3.50mm

LSD casing internal length at the maximum diameter - 79.00mm

LSD casing internal length between thrust washer recesses - 95.00mm

Minimum length between side gear thrust surfaces, end-to-end - 86.50mm

Side gear end float - 0.75mm each, 1.50mm total (no Belleville thrust washers, as these have been removed from the rebuild kit with thicker thrust washers instead).

The Belleville washers also sit within a "domed", "dished" or chamfered relief area in the casing ends, leading to a 0.4mm reduction in their effective height to 3.4mm, as can be seen in the highlighted area here:

I decided to start with 0.5mm shims either end (the shims used with the carbon plates were 0.45mm), leading to a total uncompressed pack length of (6.80 + 8.32 + 8.00 + 56.00 + 1.00) = 80.12mm. From some bench testing I did in measuring the compression of the Bellevelle washers, assuming a coefficient of friction of 0.2 in the bolt I used to compress the washer between two blocks, I came up with a spring rate of 3,800N per mm. This would mean that in compressing the pack 1.12mm, I would have a clamping pre-load force of about 2,130N (due to 2 spring washers in series having half the spring rate of the single washer).

Here is the set of components ready to assemble into the casing:

The thrust shim followed by the thrust washer with the 4 oil grooves facing out is first inserted into the casing recess:

Then a Belleville washer, clutch pack shim, clutch plate, friction plate, another clutch plate, another friction plate and the first ramp cup. Ensure the acceleration ramp, 30 degrees in my case, faces in the opposite direction to the forward rotation direction, if not swap the cups. My cups were marked "L/H" and "R/H", but when looking from the back of the car, L/H is on the right side, R/H is on the left side, so the marking doesn't seem to correlate to the side of the car. The side gear is them inserted through the pack. Ensure every component is liberally coated with your diff oil of choice as they are inserted:

The spider gears and cross shafts are then placed in the appropriate ramp positions (30/90 or 45/60), followed by the other side gear. Look through the holes in the casing to ensure that you place the other cup's ramps to match the one already in place, it will only align properly in one position of the 6 available:

Next repeat the friction disc, clutch plate, another friction disc, another clutch plate and then the shim followed by the Belleville washer. The thrust shim and thrust washer with the 4 oil grooves facing out are placed in the casing cap and should stay in place when turned upside down, with the right amount of oil on them:

The cap is then placed on the casing with the dots on each aligning:

Next, using the old bolts (the new ones will be used on final assembly after checking the preload), hand tighten in a triangular sequence until all slack is removed, then tighten a half turn at a time in the same pattern with a torque wrench until they are all tight at 40Nm.

Check the ramps are in the right direction, you can see the 30 degree acceleration ramp here, facing down when viewed from the back of the diff, the back of the crown wheel will move up when going forward:

I didn't manage to finish the install back in the car, as I found my RTV sealant had gone off, so will have to go to the shops to get some more before sealing the back cover on. I also didn't manage to find any new Nyloc nuts for the 1/2" bolts, although I did manage to get all the replacement metric ones.

The next bit will vary depending on how you removed the diff from the car. The aim is to have the LSD supported using wood or suitable material while inserted onto the driveshafts. I ensured by inserting wood under the LSD that it could not fall, then packed more around it to prevent it moving fore and aft. It does need to rotate, though, so avoid the teeth digging into the wood supporting it. Next apply an increasing torque on one axle (by using a torque wrench and socket on the end of the axle nut and rotating in a forward direction) while the other is restrained by stopping the wheel from rotating. Initially mine broke free at 65Nm, then settled at 55Nm with repeated clicks of the torque wrench, which correlates to 40.6lbft, about as close as I'm going to get to 40lbft preload.

I was lucky, guessing the right shim first time, in the current configuration the preload I have has the Belleville washers at 30% deflection (0.56mm of 1.80mm at 100% deflection).

If the preload is too high or low, remove the LSD, disassemble and swap shims as necessary:

Next, to disassemble if you need to swap shims, take out the 6 bolts, then use 3 of them in the threaded holes in the cap to lift it off by tightening each gradually until it frees from it's slight interference fit. I cleaned the threaded holes out and used blue thread lock to put the new bolts in, before torquing in sequence to 40Nm:

Once the preload is set, re-insert the LSD and crown wheel into the casing and slide the liberally oiled bearings in. Ensure the correct circlip is used on each side, based on your markings and/or recording of the circlip thickness from disassembly. The clip on the first side will generally completely expand into the slot, but the second one may need some tapping with a mallet and piece of soft metal such as aluminium or brass. ensure it is fully expanded and in the groove with the opening either side of the channel at the top:

The new seals should only be inserted until flush with the casing. Use a flat block to work around until the seal and machined surface are flush

The LSD is now fully installed in the casing:

Caterham calls for a 5mm bead of RTV sealant on the casing face to seal the cover. I used Permatex Ultra Gasket Maker:

Although each sealant will vary in the application instructions, this one says to apply a bead ensuring all bolt holes have a bead around the diameter, not just on one side. The bolts are then done up finger tight only until a bead of sealant is visible squeezed all around the periphery:

After one hour the 7 cover bolts are all torqued to spec - 48Nm. After leaving 24 hours to fully cure, 0.8 litres of diff oil is then added (Caterham says it should not reach the fill hole). I'm using Red Line 75w140 GL5 fully synthetic gear oil which already contains the appropriate friction modifiers for plate-type LSDs.

James

David_Long
David_Long's picture
Offline
Last seen: 1 hour 11 min ago
Joined: 17/04/2014

That's a great read, but what is now puzzling me is the difference between the Sierra and BMW. The Sierra one has no shims - it did have the correct pre-load when I put it together, but no adjustment if it didn't. You also mention that the Belleville sits in a 'dome', but the Sierra one is flat. 

It could be that the design has evolved (mine was one of the first) or the Carbon plate update required a more adjustable design. Perhaps it could be that the BMW diff just needs to be a different size and shape. How do you set the crown wheel position (backlash) on the BMW? 

I have a later Sierra one to do in the coming months so perhaps that will help answer some questions.

aerobod
aerobod's picture
Offline
Last seen: 4 hours 35 min ago
Joined: 17/04/2014

Hi David,

The shims aren't necessary if the preload reaches spec without any added. As Titan spec the preload on the "medium" LSD at 20 to 80 lbft, shims may only be necessary for the higher preloads. The "dome", "dish" or chamfered relief is quite subtle, I didn't initially notice it until I put a straight edge across the cap. It may have also been a running change. It allows the Belleville to centre and also potentially not reach 100% compression, as the inner diameter is 0.4mm deeper than the outer diameter in the casing and it's cap. If the ends of the casing were flat, then 0.4mm of shim could be removed from each end to give the same prefload if all other dimensions are the same. I believe this 0.4mm chamfered relief that the Belleville washer sits in will limit the washer to about 80% of deflection (3.8mm height of washer and 2mm material thickness, reduced to 2.4mm height when compressed against the chamfered relief).

The BMW backlash and the side bearing preload is set by the thickness of the side circlips that sit in an accurately machined groove about 5mm wide. I've seen them online in 0.02mm increments in the 3 to 4mm range, mine are marked 3.44mm for the left and 3.62 mm for the right. The pinion postition is set by a special crush washer if the bearings have to be changed or the pinion removed for any reason, before the crown wheel mesh and backlash is adjusted.

I measured my backlash at 0.14mm. I don't know what the exact spec is, but various BMW diffs are in the range of 0.05 to 0.2mm, so I think mine is fine:

 

James

aerobod
aerobod's picture
Offline
Last seen: 4 hours 35 min ago
Joined: 17/04/2014

I had the time to put the diff back in today, having left the RTV to cure for several days before filling it with oil. Due to the location of the filler, I prefer to fill it on the bench before lifting it into position. In terms of reassembly, I used the following torques on the components:

Fill plug - 40Nm

Carrier to diff casing M12 Allen bolts  - 110Nm with Schnorr washers and blue thread locker (Caterham specify 87Nm, but they are 12.9 bolts with a good thread depth that can take a higher torque and I have lost 2 on track before).

Carrier to diff cover M10 long bolt - 48Nm with new Nyloc nut and plain washers each end.

Carrier to chassis M12 Allen bolts - 87Nm with Schnorr washers (Caterham specify 60Nm, but I've had these bolts loosen before)

Carrier to chassis 1/2" UNF long bolt - 60Nm with the low profile serrated washers that should have been on each end of the bolt.

Hub and ear to deDion tube M10 bolts - 48Nm with new Nyloc nuts

Diff to propshaft M8 Allen bolts - 74Nm with blue thread locker.

Anti roll bar brackets M6 bolts - 14Nm with new Nyloc nuts

Brake caliper slider bolts M8 - 30Nm

A-Frame M12 and 1/2" UNF bolts - 81Nm with new Nyloc nuts

The assembly procedure is a reverse of the disassembly, but there are a few things to watch out for. First is to ensure the carrier is properly shimmed in the chassis bushes. My shims were rather beaten up so I measured the shims required to centre the diff squarely in the chassis (ensuring the drive flange is in the centre of the tunnel and the diff is square to the chassis). I needed 2.25mm shims on both lower bolts and a 2.5mm shim on the left of the top bolt, 3.5mm shim on the right of the top bolt. I measured the shim to just slide in under hand pressure, but not fall out when pushing the bolts through them.

The beaten up shims in the top of the photo, a nice custom one in the bottom:

I actually machined the shims from an old A-Frame bush that I had kept, the core had stripped out, so I just sanded off the remaining rubber, then sliced off 4 shims of the right thickness (measured with the diff in place using feeler gauges):

With everything square and properly shimmed, I found the long top bolt when covered liberally with copper slip only needed a few gentle taps from the mallet to be driven through.

Another thing to watch out for is using the correct number of washers on each side of the A-Frame to ensure the deDion is central in the chassis. Using a combination of normal washers at 2.5mm thick and fender washers at 1.5mm thick (I used ones similar to the normal washers in overall diameter and drilled them out to 13mm). you should be able to get the deDion within 0.5mm of absolute centre (up to 1mm different side-to-side).

Make sure when the diff is raised into place that the brake cables are not trapped above the carrier top mount:

To tighten up the propshaft bolts I ensured the brake cables and caliper were back in place and adjusted, so that the handbrake was functional again to hold the driveshafts and stop the propshaft rotating. Ensure that the propshaft and diff flange line up with the same hole positions as when removed. I have a paint mark on mine to line them up.

Finally, all back together and ready for next season!

 


 

James