First thing to address is the bent wishbone. I pulled it earlier this week to get measurements off of it. To reiterate, the plan was to fully plate this, but that was obviously too far on the back burner since it got overloaded quickly.
The plan was to make a new one from scratch, and reuse the bungs and maybe some of the tubing. I decided instead of doing a plated version like my very original plan, but to do a trussed one. Since I would inevitably be using tubes anyways, may as well use the tubing in the most structurally efficient way. Because I had to go around the shocks (thus the reason for not just a normal 4-link with some straight links on the top), I would follow a similar overall geometry as the original layout but turn the bends into nodes. I figured I'd throw a little engineering at this, but still keep it pretty lazy. I started off by calculating the buckling limit of a normal 4-link that would use 1.75x.120 wall tubes, and the new wishbone would then be designed to handle those loads. With the length between joints, that is a buckling load of 20k pounds, and from the angle of the links that breaks down into a lateral vector of 10k which is one link going into compression, and a frontal vector of 35.5k which would be two (both) links going into compression. And that is then the loads I designed the wish bone for, a lateral load of 10k, a frontal load of 35.5k, and a combined loading of the two simultaneously. The actual frontal load the links experience even in low range with a much larger engine is a fair bit lower than 35.5k, but it was close enough that I decided to stick with the higher load since then I know if a 4 link can handle it then so can the wishbone and ultimately the stresses between the two loads was marginally different. I started out with a truss looking like this:
The reason I'm calling this a truss and why it's setup this way is if you assume each node (joint) can't react a moment (torque) then it will only take axial loads which is ideal. This is why triangles are preferable, everything is in purely compression or tension, since bending is usually bad. In this case it's essentially two triangles that connect at that middle horizontal element, and then that is mirrored over to the other side. So it's really three triangles that all share a common element in the middle. Because none of the tubes (very simplistically) are in bending, there is no need to add plates to stiffen anything up. This is opposed to my original idea where everything is plated, and is essentially a brute force approach to react all of the bending loads--where as the truss is reacting everything the same way that a normal 4-link would where the links are only in axial compression or tension.
However, after my first run through FEA I realized the center horizontal tube had essentially no load going through it, so I removed it. While I'm not a huge fan of this, the reason it works is because this is obviously not a simple truss where each node is free from transferring moment. All the nodes are welded and can transfer moment, so getting rid of the three overlapping triangles for a "two triangles and a diamond" didn't change anything and just saved mass. With that said, I would've stilled added the horizontal tube but what really turned me off was figuring out how to actually fabricate it. It would have to get added before at least one of the longer tubes that intersects it, and all of those tubes are primary load path and I wouldn't be able to get a full weld where it really mattered. If I could've added the horizontal tube after the three main tubes were welded, then I would've added the horizontal just because it feels right, but since I couldn't do that and would have to compromise some other weld on a part that would be taking load, I decided to delete that tube and go with this configuration:
And as I said, since I'm super lazy and not getting paid to do this, I only did FEA of the three loading conditions:
Looking up the strength of DOM tubing which is what I was using, the yield is 60ksi so that's what I aimed to stay below. The highest stresses go beyond that, but the highest stress is usually an outlier of some tight corner or the mesh isn't fine enough or something else. What really matters is what does the majority of the stress distribution look like. As you can see, the combined loading has a nominal stress right around yield, what I call nominal is whatever the highest stress that is commonly found in the model (essentially the highest REAL stress to be concerned about). As you can see from the clipped view in the bottom left, at 60ksi (so the steel is brought right up to yield) there is really only the top left tube that is kissing on yield, and in reality that section of that tube is in compression so probably isn't yielding anyways. On top of that, this is at YIELD, meaning the structure will begin to permanently deform but has not catastrophically failed so it can handle a lot more and make it home fine, and this is at the loads that would buckle a traditional 4-link. Going back to what I was saying how this isn't as simple as a basic truss, you can see the right side of the part is all straight and happy since it's in tension, but the left since since there's some eccentric load paths due to the tubes having to join together and attach to the rod end in a way that isn't a straight shot, results in some bending which is then highlighted by the two tube elements that are deflecting and have extremely high stresses.
Moving right along since this is probably boring people, I converted the tubes into flat pattern layouts and printed them off at 1:1 scale so I could trace them onto the tubes to notch them. This was needed due to the extreme mating angles of some of the tubes, which were outside the range of my tube notcher.
It turned out I didn't have enough 1.75 DOM but did have some smaller and larger DOM, and some close 1.65 x .135 HREW tubing. The issue is HREW only has a yield of 40ksi, and at the end of the day no matter how I convinced myself to lower the loads or how much stress was acceptable or what would happen if it started to yield, the obvious answer was just go find more 1.75 (I literally needed two 30" sections!). I called up Fabn801 who is the guy I originally bought my chassis from 5 years ago, and out of coincidence lives the next town down the highway, so I went over there and he gave me the extra tubing I needed which worked out great!
Fast forward all of yesterday and we ended the day here:
I traced the flat patterns, cut out the bulk on the band saw, grinded the rest out, did a bunch of fitting and sanding, and had welded on some temporary holding parts to make sure everything was as square?)---triangular as possible. I reused the axle side tube and rod end bung, which is 2.0 x .25 DOM, and then cut the other bungs off the original and turned them down on the lathe back to their original dimensions to reuse them which worked out great. I decided to MIG weld everything since I didn't feel like using the TIG, and after swapping new lenses into my helmet and being able to actually see again, I think it turned out pretty good!
Today I installed it into the buggy. There were two issues, first up was my measurements were off by about 0.25" so I had to go in and cut the chassis side bungs down by 0.25". And then the mufflers were hitting at full droop, so I pizza cut the exhaust tube right near the muffler to angle them down as well as outboard (they were pointing right at the shocks before).
I think it turned out great, it's super simple and elegant. It's not light, but it might be the lightest option, and at the end of the day I think it's only a couple of pounds heavier than a 4-link setup since it essentially has two extra half length tubes, but then one less rod end. And I am not advocating that this is better than a 4-link, this is just the only option that works for how my suspension is setup. If I had planned ahead better originally this would've been a triangulated 4-link but it is what it is.
Congratulations on getting it out on the trail!
