Hubba, Hubba

[ron parola]

Here are the pix of the cores I made up for the rear hangers, and here are the pixs of my new two piece stub axle. It WAS one piece for a couple of days. I did machine .050 off the weld area but I guess the guy building these didn't under cut and REALLY weld them. They are going to be pinned AND welded.... this time for SURE, maybe. ronp

[EuropaTC]

If I'm interpreting that second picture correctly it seems very strange to hold the hub in place with a keyway slot and then rely on a fillet weld at the UJ couple. More so since unless you do a J prep weld detail there's not much room to get anything like the equivalent load bearing area to what the shaft is capable of. It's almost like a built in "snap here first" option. Is there a keyway on the other side of that UJ section ?

Is that what the OEM ones are made like ? If so I'm amazed there's not been more failures.  As an aside, I'm not convinced that heat treatment would sort that problem out unless you're talking of pre/post weld heat treatment for the actual weld itself to prevent cracks in the weld zones.

Brian

[BDA]

Are you pinning the u-joint to the stub axle? I thought that's were you were going, but you didn't show that. I agree with Brian's concerns.

The original ones appear to be one piece. I took a look at the ones on eBay and he attaches your yokes to his stub axle but doesn't say how (for a lot more money, he will make one-piece axle & u-joint pieces).

[ron parola]

This IS an ebay one; the axle stub is fitted into the ujoint (the original one bored out I guess) via the turned down stub. There is no key nor is it a press fit, it just slides in, it appears to be welded at the junction of the inner bearing area and to the FACE of the cut down ujoint flange, with just a token weld at the other end where the stub sticks out of the flange ( where the ujoint resides) So my plan is to taper the inner bearing area and taper the step where the seal rides and weld it there. Also planing to drill and tap three holes, parallel to the axle at the joint of the stub and ujoint flange and inserting three hardened screws. And THEN cutting a taper at that joint (cutting the ends of the screws off) and welding the end of the stub into the ujoint area.  I will take some pix as I go along. I'm open to any other suggestions, and he didn't mention that he had single piece ones available, but that would be a LOT of machineing to make one. Cheers rp   

[EuropaTC]

I'm not trying to be a smartass and I really don't want to offend anyone Ron, but if that was mine and I've interpreted everything correctly then I'd be sending it back and asking for a refund.

Like many others on here I've got some background in engineering (retired metallurgist) and I realize that it's quite feasible to have hollow driveshafts with all the driving shear force going through welds, but as you've described that fabrication, it just looks insufficient.  Well, it's more than "looks insufficient" because you've broken it first time out without really trying !

Ok, I'll admit some prejudice in that I'm always suspicious of fillet welds because it's just too easy to make something look right but have internal problems, so that's where my alarm bells start ringing.   As an idea it might be better with something to transmit the rotating drive and the welds just holding it together, like maybe a key in the shaft,  but I'm heading out of my comfort zone in suggesting such things. 

I know there are more knowledgeable guys on here in this area so let's hope they chip in with ideas.

Brian

Edit to add....   Re-reading my post I realise I'm being too negative here - blame my background, we're famous for being over conservative and liking belt'n'braces engineering. 

I'm guessing your tapped bolts will be a sort of keyway to transmit torque loads ? If so, I think one would do the job. I've also attached a sketch (can't sketch, I admit it) of the type of weld I'd do if I really, really had to fix that part.  It's a modified J prep and at first it looks like you're machining away good metal just to replace with more weld metal, but the idea is to give your welder access clearance to get good root fusion at the base. I'd be going for a TIG root if possible, MMA fill.

The depth (B) depends on how much metal you have to play with but my first guess would be for something in the 4-6mm region. I stress I have NO idea here, it's a guess and someone who can do the maths will have better ideas.  The fillet at A will need machining into a radius for the bearing land & fatigue resistance. The internal fillet inside the UJ arm section is just to stop it flopping about and putting stress on the weld root, that would be whatever I could get in there.

I don't know the materials involved but given the thickness and unknown alloy I think I'd be tempted to give the assembly some pre-heat before welding. If it's anything clever in terms of steel it might need post weld heat treatment, if it's mild steel then I'd just insulate after welding and allow it to cool slowly.

[ron parola]

Yeah I'd send them back but it took somewhere around 6 months for them to show up the first time AND I had removed some metal from them, AND I'm pushing way more torque than original through them, AND with this oddball hobby oneoff  type stuff I really don't expect much. Being in the trade and being forced to use aftermarket products on older vehicles, I'm used to it. So here's pix on what I've done, cut a taper on the shaft and joint flange for somewhere for a weld to reside, drilled and tapped  three holes and screwed in set screws at the joint, and off to the welder to have it stuck together and stress relieved.  I have machined off the screwheads and put a taper so that end can be welded . But first I've got to CUT the other side apart cuz I certainly don't trust it! ronp

[BDA]

I would certainly defer to Brian. I went through mechanical engineering but I went into computers after a few years so I'm more than rusty about all that. Welds are generally strong in shear (IIRC) - which is what you have, but as you've found out, not strong enough for your application (and mods). I think I would have at least pinned it in addition to the weld but I like your idea about the axial bolts. A set screw would do the same thing and you wouldn't have to grind off the bolt head.

The machinist/engine builder for Team Highball (back in the day, they were the factory Mazda IMSA team which won their class in the Dayton 24 hour four times, so I think he knows what he's doing - you can still look them up on youtube, etc) helped me with my half shafts. I don't remember why I brought them to him in the first place but he did something similar with the nut. Rather than using that D (lock) washer, he drilled and taped an axial hole through the threads of the nut/stub axle after he tightened the nut, then locktited a set screw in there. It ain't coming loose!

[EuropaTC]

Well Ron, you certainly have my respect for "getting on with it", that's a great attitude to have.  I can't see those screws failing at all ! My only suggestion would be to widen the V you've machined to give your weld better access and ensure full fusion.  I think I understand why you want to check the other side 

BDA - I know it comes across as negative, but I've always been wary of fillet welds on loaded components. From the fracture Ron shows it looks to me as if there's been no weld penetration into the shaft material and it's been relying on two fillet welds for all the torque.  Given this is supposed to be a more robust solution to the OEM splines, I'm not convinced by what I see there.

I know driveshafts are hollow piping designed with butt welds and as you say, most metals are great in shear.  At the risk of going into geek mode or insulting anyone's knowledge, the butt welds are generally the same throat thickness as the parent tube material so there's no strength loss. With a fillet you get less load bearing metal than it appears (typically 0.7x the leg length) so what looks like a 6mm fillet only has just over 4mm of load bearing metal and that's assuming perfect root fusion. The latter should be a given but not assumed with relatively thick components, fillet welds generally lose heat quicker than butt welds, so fusion can be an issue. (/anorak mode off)

It's compounded on this item because the diameter at the weld looks to be less than that on the intermediate driveshafts, hence less overall load bearing metal - a "please snap here" point, as Ron has managed to show. I'm just surprised at what I'm seeing I guess.

Brian

[Bainford]

Yes, agree with everything Brian says above (and well stated, Brian). Your solution looks good, Ron. That should be very robust assembly.

[ron parola]

Ok, off to the welder this AM. After CUTTING the other side apart; it had a deeper and larger weld, and it was knurled where it went into the ujoint flange, it may have been fine. BUT the issue now may be; the ujoint flange stub was not concentric with the axle bearing surface, I guess they partially machined these then welded them together and then finish machined them. Unfortunately I didn't mark the flange so it may or may not be true to the shaft, off by .010". I don't have any good way to measure squareness of ujoint mountings. But I've seen joints with no needles, rusted AND squeaking but no vibration, so fingers crossed ronp

[BDA]

Boy, what a pain! I hope everything goes together ok! I sure would have expected better build quality from that guy - especially for the price he charges.

[jbcollier]

Driveline shops will have jigs for that.

[Lotusjps]

I'm a firm believer in simplicity.  Here's a picture of the stub axes I made (designed and had Dutchman Axles machine), where the hub and axle are one piece and the u-joint yoke bolts on.

[BDA]

Those are pretty! I understand that splines are expensive to machine. If you don't mind telling us, how much did they set you back and can they make another set?

[Lotusjps]

I designed them back in 2008, after I had 2 year old stub axle break (see pic) during an autocross (fortunately a very wet one).  In order to bring the price down I tried to find others interested from the yahoo group, though when it came time to pony up money only 2 out of 12 came through (though I went a head and made 4 sets in order to bring down the price.  At the time they were around $1,300 per set (2 axles/hub, 2 yokes, 2 radius savers, 2 custom nuts).  I sold them at cost, I was mainly interested in having a set for my own use than making anything from them.  I'm sure they'd be a fair amount more to make them now days, which would probably price them out of most peoples comfort zone.