A twist on triangulation

[kreb]

I was looking at the Goblin site and noticed how the diagonal tubes were stopping just shy of the traditional termination point. Can see how it would facilitate fabrication, but how much strength is lost when you do this:?

[a.moore]

Quite a bit. You're putting that short section of tube in bending and it is going to flex more and fail in bending sooner than if you ran the diagonal to the node.

viewtopic.php?f=39&t=2224&start=155

[Driven5]

It helps when you're running 1.5x.095 chassis rails.

For those that aren't though, you could still bring the triangulation tubes close-but-not-intersecting, and then reinforce the space between with a gusset/doubler of some sort.

[kreb]

Not to diss on the Goblin, which I respect, but I find it interesting that they could have tightened the clearances quite a bit without making it any harder to fit and weld.

[KB58]

When it's a business, it's all about balancing time and cost and quality. Yes they probably could have done a bit less of that, but if it's over-designed to start with, it'll be just fine.

[hfmaxi]

Rollover protection from FIA 253 Appendix J generally allows for 100mm max gap for tubes to connect to junctions. Not sure how the FIA made that determination but they are generally who I look to when thinking about safety gear and requirements. Needs more gussets on some of those high angle joints though.

[TRX]

Last time the subject came up, the concensus was that it was an acceptable compromise, if not actually superior to "big wad of weld and slivers of chamfered tubes".

[kreb]

Live and learn. Thanks guys.

[Driven5]

Not to diss on the Goblin, which I respect, but I find it interesting that they could have tightened the clearances quite a bit without making it any harder to fit and weld.

I too brought my triangulation tubes in adjacent to the node, rather than fully into it, with the triangulations tubes offset to create a 0.5" gap on the tight angle side. My intention is to also go back and reinforce the gap with fully wrapped (3-sided) 'taco' gussets, made from the same square chassis tube with one side cut off. While my design has definitely allowed better welding access (and subsequently better welds) than than I otherwise would have been able to achieve if bringing the triangulation tube directly into the node, I can also tell you that the 0.5" gap is still too tight from a manufacturability standpoint. I would guess to continue seeing manufacturability improvements continuing easily through a 1" gap, and probably at least a bit beyond, but without further experimentation I can't really say how far out the 'sweet spot' might actually end up either. Like Kurt said, it's all a balancing act when designing and building for production.

I'd also clarify that the takeaway from FIA Appendix J is simply that the larger and heavier the tubing is, the larger the gap may be...Not that a nearly 4" gap is an acceptable general design practice throughout a spaceframe chassis. Which goes back to the first statement in my previous post though. Along with that, increasing stiffness is a game of diminishing returns, and there has to be a point where it becomes 'stiff enough'. And while I don't know how stiff their chassis is, but I wouldn't be surprised if it is in fact quite a bit stiffer than most book based Locost chassis...Which also still tend to perform more than adequately for most enthusiasts.

TLDR: From a technical standpoint, I agree with you completely. But from a manufacturability standpoint, I also agree with them...At least up to a point.

[seattletom]

I understand the manufacturability and cost arguments in favor of offset diagonals, but from a low production standpoint it's not that hard to properly fit the diagonal tightly into the joint.

Just fully weld the main tube joint and then cope and fit the diagonal. Some care in coping to the right length is needed and yes, one of the fish-mouth "cheeks" often has to be cut off to insert the diagonal (and the cheek then welded back in place.) Lots of software programs will generate the needed coping pattern for the joint, including the diagonal (I use Digital Pipe Fitter.)

Stronger (perhaps minimally) and IMO better looking. And my build definitely qualifies as a low production operation.

[Bobber]

In truss design, the centerlines of the tubes should meet at a point. Small eccentricities are considered to be resisted in tube shear rather than tube bending.

[KB58]

The problem isn't tube fabrication, it's allowing (easier) access to the entire circumference of every tube when welding. If the head of the TIG/MIG can't be maneuvered into position, it's hard to know whether weld strength is sufficient.

[davew]

These are custom frames. It's not like they are building 40 or 50 frames an hour. It would take 10 or 15 minutes longer to weld all the tight fitting coped tubes? It makes you wonder what other short cuts have been taken?
Davew

[kreb]

x

[rx7locost]

If I were building a Locost, I would stick with the well proven designs and construction methods. If buying a Goblin, I wouldn't worry about the tubes that don't meet so-called "best practices" as BP's are based on ideal situations, not specific designs. Many steel trusses do not follow those BP rules either. That doesn't make them unsuitable for their intended use.