[Gwagensteve]

Well after being told and thinking I "knew" this was impossible, I've just done an experiment and it's completely feasible to run 10mm drive flange bolts on a Sierra hub.

This may also be possible for Jimny/Vitara as well, but I won't comment on those cases.

Some background: I started making fully floating rear axlehousings for Sierras a long time ago... certainly by 2010 I had an completed axle under my Sierra.


For a number of reasons it's proved to be fiendishly difficult to keep drive flange bolts tight. Contributing factors are: The housings are often slightly bent, wheel bearings loosen, very heavy wheels and tyres with lots of traction, and high torque from reduction gearing.
Whilst M8 bolts are common in the industry for lots of cars running Aisin-Warner freewheeling hubs, that's typically on the front where their less heavily loaded. Additionally, some manufacturers (like Toyota) include dowels to improve the connection.

Broken rear drive flange bolts are a common problem with older landcruisers, which also run M8 bolts (but with dowels) and when the 100 Series land cruiser came out, Toyota upgraded the drive flange bolts to M10, still with dowels. This is a common upgrade for older cruisers now.

A few of us have gone to Nord-Loc washers, teamed with Unbrako Gr12.9 cap head bolts, and this seems to be fairly reliable, but I really liked the idea of getting an M10 bolts in there. ]

I'd always written it off as impossible as there was inadequate room.

However, I did some measuring last week and came up with a plan. \\

An M8 Unbrako has a cap diameter of ~13.12mm.

An M10 Gr. 12.9 unbrako has a cap diameter of ~16.12mm

So that's about 3.0mm bigger in diameter.

But, I have at least 2.0mm of clearance between the M8 unbkrako head and the centre bore of my rim.





Theoretically, then, an M10 unbrako will fit.

It was time for some testing.

I dragged a pair of old hubs out of the trash pile and tapped one out to 10X1.5. Initially, I have only done two opposed bolts, because if that fits, that's enough to prove it will work.

And you can see here - any rim that sits on the register of the hub will clear the M10 Unbrako



And here's a pretty clear image of how much more beef the M10 has.




Now to torque specs.

In a stock sierra (front) the drive flange bolts are torqued to spec at: 20-30Nm.

The maximum torque on a black Unbrako, M8 is 39Nm

Maximum torque on a block Unbrako, M10 is: 77Nm

The added friction between the bub and drive flange this provides should prevent the drive flange twisting on the hub, which loosens the bolts, ultimately puts the bolts into shear, at which point they fail.

I may get this done before tour, and I may get a set of M10 Nord-Loc washers to complete the solution.

[Gwagensteve]

Oh, just checked whether thy clear a stock WT rim, and they do, comfortably. A stock WT rim has about 2.0mm of play on the register of the hub, they're MILES from being hub-centric.

(oh, and this is why I'm non committal about this working for Vitaras and Jimny's - I know Vitara rotors are very tight on the vitara hub and I believe stock jimny rims are hub-centric in a stock jimny.

[Brenno]

Great tech Steve. If this becomes a problem in my build I'll have to check if it will work in my application, Vit hub and disc, XL7 rim.

Question; do you think it's more of a heavy tyre/wheel or more power/torque that causes the issue?

[alien]

My Sierra was full float rear on 31's with airlockers and daily driven, and the rear flanges/hubs never ever ever came loose from 2008 till I sold it last year. Must be related to the leverage of the tyres!

[Gwagensteve]

Cheers Brenno.

Alien - it's great you had a good run with yours. I'm not saying everyone has this problem, but interestingly, some years ago Joe (JML) came up with a ring that locked the next head bolts in position, preventing them loosening - so he was aware it was an issue for at least some people. I didn't like this solution because it did nothing to address the inadequacies of the connection - the bolts coming loose are a symptom of the problem, not the problem itself.

I think at least some of it is the nature of the driving we do and the weight of the tyre/wheel combination. I think most of us are running enough torque to overload the connection, I think that we oscillate between scrabbling and bouncing and then binding up under high load. The torque "spikes" must be very high, especially when there's near 50kg of tyre and wheel hanging off the hub. I don't think GregC has had any problems at all since he's dropped down to KR2's.

[Mike harris]

This is really cool, appreciate the write up, il certainly be doing this !

[BlueSuzy]

I've also done this on my "big sierra" with Hilux diffs - Hilux hubs

The locking hubs are the same pcd as your drive flanges. Hilux hubs are also doweled, could a little hard to dowel a sierra hub but it might be another option for strength.

The M10 fit nice through a locking hub without the cone washers. I don't have any loctite or washers with my capscrews and they've stayed tight. The dowels are probably helping that though. Nord-locks are good and would be my choice too as I've used them a lot in the workplace.

[royce]

Could you run into trouble with there not being enough meat around the outside of the threaded hole causing it to split?

Or an axle that is splined either end with a drive flange that slips onto the end and bolts on, give it a little bit more play for loose bearings etc?

[Gwagensteve]

The axles are splined at both ends and they aren't tight in the spline in the drive flange, well, they're not now my housing is straight.

If I can get the 77Nm without an issue I'll call it a success. The bolt is only in tension (as in there's nothing levering on it) so it only has maintain torque.

[Gwagensteve]

royce wrote:

Could you run into trouble with there not being enough meat around the outside of the threaded hole causing it to split?

Or an axle that is splined either end with a drive flange that slips onto the end and bolts on, give it a little bit more play for loose bearings etc?

Just torqued to spec- 77Nm. No issues with stability of thread.

[Gwagensteve]

And success!



Not without a small amount of struggle though, totally my own fault.

The bolts as M10X30. That is absolutely the longest bolt possible.... and here's how I know.

I ran an 8.5mm drill down the existing M8 threads, then tapped to M10 with a standard tap, then run a bottoming tap down. When I fitted everything up and went for torque, one bolt didn't pull up on the drive flange - so clearly one of the tapped holes was too short, maybe by a turn or so.

Now I had to get that bolt out, and it galled the thread as it bottomed out. Predictably, it was a bear to undo. It broke my hex driver, then, the bolt broke the last 8mm off. 10mm down in the hole.

Procedure for repair:

Cut a small length of 8mm hard fuel line - this fitted in the bore of the hole to protect the thread. Drilled a shallow hole with a 6mm drill. Grabbed a new 5mm drill and drilled through the broken bit of bolt. I then stepped up to a 7mm drill and ran that through, and that was enough to break up the remains of the bolt in the bore. Blew it out, ran the tap back down, and the bolt has tensioned properly.

My wife picked me up nord-lock washers yesterday, so today I have the passenger side to do then I can finish prep for Adventure Tour.

Very happy with the outcome. Not happy about clumsily breaking a bolt. I did have a spare hub I could have swapped in but that's a pain of replacing bearing races, setting up bearings etc.

[Mike harris]

That’s a really great upgrade, very impressed.

Keep us updated if they do come loose, I very much doubt they will as they met such a high torque !

Top work mate.

[Gwagensteve]

I’ve just given them hell for two days on some of the highest traction/steepest tracks I’ve driven. They’re still tight.

[Mike harris]

I’ve got to do this mod, il copy how you did it, hopefully achieve the same great results, I think it’s really cool you lads have so much info on Suzuki’s !

[Mike harris]

I just carried this out to a spare set worked great ! Can’t wait to try them out !

[Mike harris]

Done mine !

Il monitor them next weekend see how we get on !

[tanshi]

SO you are doing away with the tapered washers steve? Im struggling with unbrako m8s and dowels on the ultra4 coming loose. thinking m10 cap heads will be the go

[Gwagensteve]

Sierra drive flanges don’t have tapered washers. My new axles will be drilled to 10mm too. (I’ve already tapped the 60 series hubs to 10mm.

[tanshi]

Fair enough, Ive got some GV ones with and some GV ones without tapers, Didn't really want to try and cut tapers in rcv drive flanges if i could help it.
So much stuff comes loose on this buggy, 14mm Arp wheel studs with Nylocks come loose on the kingpins, must be stretching as theres no reason a nylock should come undone in that usage. 12x1.5 cruiser studs come loose, Hoping arp wheel studs will stop that issue

[Mike harris]

I’ve done three winch challenge events now in mine and they haven’t moved, really impressed.

I used to run arp studs but they always came loose and sheered off.

I run 38.5 inch tyres so there’s a fair bit of force on them I’d say, but yeah great upgrade for sure !

[Gwagensteve]

It's a design issue with the kingpins - your use is overpowering the connection and it's placing the studs in shear, where they're ineffective.

ARP material has effectively the same mechanical properties as a Gr.12.9 socket head cap screw. According to Unbrako, torque spec on an M14 unbrako is 215Nm. Is that what you're running them at?

Torque on an Unbrako M12 is 135Nm, about twice as tight as typical for a wheel stud. However, on an alloy wheel, I'd suggest the alloy will deform and you won't achieve peak clamping force. I'd consider steel inserts in the alloy rims.

[tanshi]

210nm i doubt it. I have chromo inserts in the 7075 alloy steering arms so i guess i could

[Gwagensteve]

If you’re not torquing to spec start there. That’s the reason I went to 10mm- to put the connection into tension not shear.

[Gwagensteve]

Sorry for anyone that's across this, but this is something that's often misunderstood about bolted connections so for others reading this thread I'll reiterate it as this is now a lengthy thread.

A bolted connection is a clamp. It works by generating enough friction between the two surfaces being clamped together that the connected parts don't move relative to each other.

If the bolts are placed into shear - i.e the flange moves under the bolts and attempts to shear the bolts off - the connection has already failed.
Yes, at this point the bolts will loosen, but that's just the result of the flange moving under the bolt heads. This is true for all highly loaded connections (like flywheel bolts, flex plate bolts, drive flanges etc)

I do not believe the dowels in the toyota hubs do anything substantial. Again, if the flange has slipped around so that the dowels are transferring torque, they'll break off - two 7mm dowels aren't going to drive the car if 6 torqued M8 arp studs/unbrakos aren't doing the job. I believe the dowels are basically an alignment/assembly aid.

The key is that with big tyres and reduction gearing, (especially with an automatic) we can overload the connection easily and there is just inadequate clamping force to stop the two surfaces slipping. M10's add ~45% or something of clamping force, which goes a long way to keep them clamped together.

Out of interest, I just went and had a look at my Dana 70 dually rear axle, and it has 8X 3/8" UNF studs. That's a substantial increase over what we see in predominantly Japanese 4WD's.

As an aside, I've looked at specs on premium fasteners to try and find the highest torque values. Basically, nothing exceeds the yield strength of a Gr12.9 unbrako. ARP and F911 fasteners match but do not exceed the mechanical properties of Gr.12.9. They're a lot dearer though and they're shiny #racecar

Here's another solution that's even stronger. Unbrako's are available in UNF (some of you may be aware that Unbrako's don't come in metric fine, only coarse.) Have a look at these numbers:

M8 Unbrako: 39Nm
M10 Unbrako: 77Nm
3/8" UNRF Unbrako: 96Nm

I wish there was a Gr.12.9 M10X1.25 (metric fine) available. I think that would exceed 100Nm.

I also make sure I'm mating clean steel to clean steel. I use a very fine smear of RTV but NO GASKET - that acts as a lubricant between the two surfaces.
Also note that oiled bolts have a LOWER TORQUE that dry bolts - don't get this wrong as you'll likely yield the bolt.

Lock wires (properly called "safety wire"), locking collars, nord loks etc are basically there for security, they don't improve the mechanical properties of the connection, so if the flange is moving under the bolts and taking them to shear, that will still happen. They loosen because the flange is moving under them due to a lack of clamping force, the flange doesn't come loose because the bolts undo.

I'd recommend anyone with some interest in this have a read of the unbrako engineering guide which is freely available on line. It's very informative, especially in relation to highly loaded and problem connections. Also, Caterpillar have apparently done some of the best engineering on massively loaded connections, as you'd imagine, but I imagine that's a bit harder to find.

[Gwagensteve]

For reference:

http://www.unbrako.com.au/images/downlo ... _guide.pdf