Differential bushing bolts: do you preload them?

[sapote]

In some cases, the diff clunk is caused by the diff housing wiggling up/down or rotating CCW during shifting gear and heavy acceleration. Under heavy acceleration, the diff wants to lift the front up and rotate CCW (rear view). The front bushing and bolt have some gap in between, and the same for the two rear bolts. The new updated rear 2 bolts have larger shank to reduce the gap between bolt and the bushing center core, which helps minimize the wiggling. But I think by preload the bolts, the diff will have lower chances of movement between the bolts and the bushing center cores.
Hand threaded all bolts + blue Loctite, but the final turn, then proceed in this order:
1. Front bolt: jack up the front of the diff to take out the backlash between bolt and bushing core, then do the final torque 200 Nm
2. pull the left side of diff rear cover down (using a lever or something), then torque bolt to 20Nm + 90 degrees
3. jack up the right side of the rear cover to take out the backlash, then torque to 20Nm + 90 deg.

Ideally, the bolts should have no gap with the bushing core center hole. Any gap in between will lead to movement under heavy acc, then this worn down the bolt washer and the bush thrust end, leading to loosen bolts and clunking.

[Tbonem3]

Ya that could help. I didn't, but I have no clunk or rocking/knocking anyway.

or machine some sleeves for the bolts

[Tones]

My clunking is from my diff wearing and being sloppy.

Sent from my Pixel 3 XL using Tapatalk

[sapote]

or machine some sleeves for the bolts

I would think machining a custom bolt is easier than making that precision thin sleeve.
Also, to avoid twisting the rubber bushing during torqueing, especially the more compliant OEM parts, the bolt head and the washer under the head should be lubed with grease. Not the threads though.

Since the final torque on the axles is 3x more than the diff input shaft, and so the torque that lifts the diff front up and pushes the diff rear down is 3x more than the torque that turns the rear right bush up and rear left bush down. So I would preload both rear bushings down and push the front bushing up before torqueing.

[03m3vrt]

This is a cool idea. Technically a bolted joint has “failed” if it slips but i may try this out

[sapote]

This is a cool idea. Technically a bolted joint has “failed” if it slips but i may try this out

Yeah, this is a tough engineering problem to solve. The bush/bolt location is relatively short distance from the diff rotation center (bisect point of the input shaft and the output shafts), and so the torque creates a large force on the bolt, flexing it up/down during accel/decel. Eventually the washer between the diff and the bushing center core will wear out along with the parts on both sides of it. This starts the clunk noise even the bolt is not lose due to Loctited. So the bolt needs to be removed for inspection, then torqued to 150 ft-lbs with Loctite and preload.

Here is an idea: coating the bolt shank and the bushing center hole with JB weld, and the inserted bolt has no air gap in the bushing. Let it seat for 24 hrs.

It would be much better if the diff has a 24" long arm extended forward for mounting to the bushing, instead of the threaded hole on the side of the diff housing.

[sapote]

At least for the front diff bushing, they should have used the vertical bushing method (similar to the diff subframe bush) and not horizontal. I believe Porsche have used vertical bush since 1965 for similar application, and the bolt has no or very little side force to shake it loose.