LocostUSA.com

Hi everyone, after returning from a major adventure, driving this guy from Baltimore to Minneapolis and back:



I decided to stop planning and just dive into a Locost build, and this is the start of the log. There are some important questions about moving forward at the end of this post, but first some pictures.

Here's the donor, a 1994 MX-5 with Torsen LSD that I'll be bringing down from PA this month:



I bought enough 16-gauge A500 steel tubing for the Vodou plans, and just started chopping and welding. Here's the current state of the frame:



Working in a tiny row house basement, and instead of a full assembly table, I tack-welded the lower rails while clamped to 4-foot pieces of I-bar, and use an abrasive chop saw and a digital angle-finder to cut and fit the pieces:



Some weld close-ups, starting with one of the earliest:



A basic flat join (most of the outer welds are like this):



The open edges of a tube:



An obtuse angle:



And probably most importantly going forward, my current version of a decent fillet weld versus not-at-all-decent:




As you can see, I'm not much of a welder! The flat outer welds are easy enough, but the rest can get pretty messy, and I've been grinding them off, making multiple passes, etc pretty freely.

Before pushing ahead with the diagonal bracing and beyond I wanted to be sure I'm not too far off-track, because I have no sense of the acceptable tolerances, weld quality, etc. None of the paired angles (RHS to LHS) differ by more than about half a degree (according to my 0.3-degree-accurate angle-finder, so maybe just say "all are within a degree"), generally centered on the "correct" Vodou angle. From a rear corner to the opposite front corner, it twists maybe 3/16" (i.e. if I push down on one corner, that's how much the other corner rises). Finally, as you can see from these pictures, there's a lot of variety in the weld quality (in the first image, I doubt I can weld in that extremely-acute corner at all):

I'm using this project to learn the basics, so perfection's never been on the table, but given these angles/twists/welds, can it become a decent build? My hope is to correct for geometric issues when attaching brackets, positioning the differential mounts, etc. For the welds, I think I'm getting better pretty quickly and can go back over the rougher points, but at what point does heat cycling become the problem? Do these frames have some redundancy such that regularly inspecting for any cracks will catch them before serious failure?

Thanks for any and all feedback! I'll keep updating this thread as things progress, and perhaps create an external (full) gallery at some point.

tilppy
I'm not questing your sanity but how are you going to get this completed vehicle thru the basement door!
I would recommend that you weld the frame in the free-state and on a rotisserie once the frame is completely tack welded together. Once the frame is tacked together it will hold its alignment, if weld in a systematic order. In the initial tack welding process, every tube really needs to be clamped down in place. After the complete frame has been tack welded together, you are ready for the final welding. One should try to balance all of your welding to minimize distortion. Weld the outside of the joint then move to the inside 180* opposite and then weld that side of the joint. Next move to the opposite side of the frame and do the same process on the other side "mirror side" so the weld distortion is balance from side to side.
Never completely weld all of the tubing joints on one side of the frame at one time and than flip it over and do the other side. Your frame will end up looking like a banana. Just take time to think how the weld will shrink and draw the tube to the side that the weld is place on.
The individual builder sets his/her own dimensional standards during the build process, but the biggest impact relative to dimensional inaccuracies in the frame will be in the suspension system during construction. With that said you need to realize that this type of building process and limited fixturing allows for quite a few dimensional errors to stack up in both the frame and suspension arms. With that said; if you lay out the pick up points for the suspension accuracy even with some frame errors, things like ball joint centers etc., will still be OK. Davew

Thanks Dave!



There's a cellar door right behind it, with about 29" width, so without a roll bar it shouldn't be a problem.



I clamped as best I could when tacking, and have been switching sides for the final welds, but it's certainly been sub-optimal! My plan (based on intuition not experience so open to suggestions) is to go ahead and put the braces in, and finish all the welds as best I can. Then, to try and account for dimensional errors, I'll carefully orient and secure the frame using a digital level, and use the same level and various straight-edges/squares to lay out the suspension brackets and differential/transmission/engine mounts.

I've got some clear pyrex gas lenses and thinner filler rods coming tomorrow, hopefully that will help the weld quality in some of those hard-to-see corners.

If you can locate the suspension MOUNTING HOLES within a 2mm range you should be as good or better than most. Once you trail fit the engine/trans, consider adding a lowered crossmember, either welded or bolted in, at the rear of the engine bay. You need it to complete the shear plane. It will make a big time improvement in torsional stiffness. Davew

When we did our frames, we drilled a whole bunch of 1" holes in the build table, then used the holes to pass clamps through, and clamped the entire thing down to the table. It was VERY solid. Even so, we alternated from side to side until the entire thing was tacked at every corner where tubes met. We ended up with frames that were closer than 1/8" identical from corner to corner diagonally.

For those who haven't tried this (we did and were shocked!), take a spare piece of 1" square tube, maybe a foot or so long, and put one good, full-penetration tack at one corner of either end. Wait for it to cool, and take another look at the tube. In our test, the tube warped over 1/4" over its length, just from that one, single tack. THAT'S why we decided to clamp the whole kit 'n kaboodle down like it had to survive a hurricane.

It worked.

_________________
Scratch building, at continental-drift speed, a custom McSoreley-design framed, dual-Weber 45DCOE carburated, Zetec-engined, ridiculously fast money pit.

http://zetec7.webs.com/

Thanks! It's good to hear what scale the tolerances are: I expect I'll take it slowly, over the winter, and that'll be a lot of time to focus on really nailing the alignment.

I like that first pic. Did they ask if you were dropping off the art?

_________________
Miata UBJ: ES-2074R('70s maz pickup)
Ford IFS viewtopic.php?f=5&t=13225&p=134742
Simple Spring select viewtopic.php?f=5&t=11815
LxWxHt
360LA 442E: 134.5x46x15
Lotus7:115x39x7.25
Tiger Avon:114x40x13.3-12.6
Champion/Book:114x42x11
Gibbs/Haynes:122x42x14
VoDou:113x44x14
McSorley 442:122x46x14
Collins 241:127x46x12