I ordered a 5/8 diameter 4140 steel rod which will be main pivot wheel axle on trebuchet (few thousand pounds of force going through it)
Would it be useful to try and harden the shaft? I don’t know how much that would improve its strength. Anyone have experience heating then oil quenching 4140?
Hardening it will make it more brittle and likelier to crack (although that depends on what the temper it currently is - might ask seller or look up ASTM spec referred to from where you bought it). But, IMHO, for safety reasons it’s better that it bends than breaks from being too brittle. Like wing spars on planes - better they are slightly weaker and bend than break off.
Something that has some thick padding or hard rubber where it hits will help dissipate the shock loading. Worse case if you’re worried about bending would be to weld on the BACKSIDE of the lever with say a thick triangluar shaped flange to reinforce it and spread the load while stiffing that area.
We don’t have a hardness tester. If someone at the Space has access to one at work, could you take a small piece of the drop-off get quick test done, say on Rockwell C scale. If this number is known then the engineer types can give you better evaluations of what to do with estimated strength and the math on the localized loading. Oh - reinforce whatever it is hitting also. Gusseting the welds that will take heavy shock loads probably would be a good idea.
Considering the things the machines of death and Darwinistic things you’ve built before and personally placed you body upon - who is this new Brandon talking safety FIRST. What have you done with him!
When it comes to hardening, you increase the hardness and tensile strength of the material at the cost of toughness and ductility.
It is extremely common to confuse the “strength” of a material with its hardness, but strength is an overly broad term. As you increase tensile strength and hardness in this case, you’re going to get a more brittle material which probably won’t fare well in an extremely dynamic loading situation (i.e. a pivot).
If you will look at this drawing: https://talk.dallasmakerspace.org/t/trebuchet-sig-meetings-sundays/12213/8
Brandon is talking about the axle represented by that small blue circle on the arm away from the uprights. In the ready-to-fire position, this axle and its two wheels are up in the air. As the weights (on the other axle) are released, the arm swings downward, pivoting around a support wheel at the end of the left crossmember until the main pivot wheels make contact with tracks attached to the top of the crossmembers.
This is the point of maximum stress on that axle as the downward angular momentum of the arm is converted into rotation about the pivot. There is no way to pad this sudden smashing impact.
To elaborate further about this axle, if you will look at the first picture in this post: https://talk.dallasmakerspace.org/t/trebuchet-sig-meetings-sundays/12213/88
That bar at the far end of the arm is where the weights will hang.
That cylindrical thing in the arm closer to the camera is a piece of the same material forming a bushing through which the 5/8" diameter rod axle Brandon mentions in the OP fits.
The main failure mode is that axle shearing at the ends of the bushing.
Should the bushings be slightly softer than the axle?
Currently we have 1018 rod for the axle, but when I saw it was only $12 to order a 5/8" 4140 rod on amazon I went ahead and clicked order, so the stronger rod will be here tomorrow.
I think these wheels or axle is the most likely failure point in the system, as all the force accelerating the arm is going through it.
This is the piece in question, mockup showing the wheels connected with the 1018 rod
The wheels with the bushing that the axle will go through
The 2.25" 1018 bushing welded into arm:
Shear may end up being your biggest problem, but there will still be plenty of flexure on that axle. This is where the FEA package on your CAD model comes in handy 
Shear strength loosely correlates with tensile strength (higher hardness, more tensile strength). Flexural strength is dependent on ductility.
Ehhhh, it’s not dependent on ductility, but I’m tired of typing
Why not just a grade 8 bolt? Also the yield strength for that 4140 at 5/8 is 18,469lbs
I actually did buy a 5/8" grade 8 bolt intending to use it for this, but ran into two issues, 1) ace didn’t have a 9" bolt so I got an 8" but that is a bit too short, 2) The clearance from the outside of the wheels to the uprights is about 1/4",not enough room for locking nut on one side
So my plan is to take the 5/8" rod and drill then super glue in a 3/32 hardened steel pin on each end to retain the wheels on the axle (I bought extra 3/32 axles for pinewood derby project along with a reamer for them, will come in useful here assuming I can manage to drill that small of a hole in the steel, will mill a small flat first)
My science may be off here, but I don’t think a cyanoacrylate glue (super glue) will do much against steel given the heterogenous materials being bonded and the (multiple-axial) stresses likely to be incurred during the drop. I would think to suggest JB Weld above a CA glue, but I would also think a washer tack welded to the end might be even stronger than the epoxy.
I guess at the end of the day I’m wondering about what kind of service life you need out of it. If this is only going to go through a few rounds of testing and competition before tear down, I would caution against over engineering, given a defined budget. A little flex here or there isn’t always a bad thing.
But hell, I’m still the new guy. Maybe I should swing by during the next build day. Is it a standing Sunday project?
Yeah we will be working on it sunday afternoon, there is lots left to do.
You could use an auto center punch without having to mill a flat. You could also use roll pins, hairpin, or lynchpin. Hell we could cut a custom washer out to tack on the end. Your side load shouldn’t be much if anything.
Machining required, but sounds like a job for a circlip…
Or a cotter pin… 1234567890
Be Brave. Be Confident. Weld it in place!
Continuing the saga of this axle, after the arm has completed the transition to pivoting about the axle, the weights continue to fall, whipping the end of the arm upward and downrange at higher speed. When the arm reaches its vertical position for the first time, any downward momentum left in the weights not converted into arm rotation will be borne by this axle and wheels against the track on the crossmembers. It will be supporting the entirety of the weights when the machine finally comes to a rest after the arm oscillates about the vertical.
I am hoping that this is not going to be a one-time thing. That we will continue to tune and improve it during the off-season to make it a contender year after year.
Turned out to be easier than expected to drill the holes in the 4140 rod for the dowel pins
Anyone want to place bets on if this assembly will survive 2-3 tons of force?
I went ahead and ordered some spare wheels of the same basic dimensions but different supplier as both backup and because I want to attempt reversing 4 jaw chuck jaws and trying to turn outside on the lathe to improve roundness.
