Initial results of rivnut pullout testing: (Due to budget constraints, only doing 10-32 steel thin-wall large flange inserts this year; if we're happy with the results, we'll expand our options next year).
I drilled a 19/64" hole using a drill press and inserted an SKL 10-32-130 steel thin nut with large flange into each of three pieces of aluminium:
- 3/4" x 1/8" bar (Ace Hardware, looks like 6061, pictured above)
- 0.1" wall Vex 1"x1" tubing (I used a factory hole as a pilot)
- ~1/16" wall channel (sold as 1/2" plywood edging, also pictured)
I then centered a 1/2" hex nut over the rivnut (large enough hole for the rivet to pass freely), added three steel 1/16" thick fender washers, and inserted a 1" stainless 10-32 cap screw (allen drive). I did not have a 3/8" to 1/4" adapter handy, so I could not use my torque wrench, so I used my shortest 1/4" ratchet driver. In each case, I tightened with my index finger and thumb against the adapter socket. In each case, I was using roughly the same amount of torque (to within perhaps 50%) at failure. I shall repeat these tests with an adapter so I can measure torque, and with the longer rivnuts in coming days.
Each setup reached normal (tee-handle) torque with no noticeable deformation or stress relief (that is, yielding or breakage) of any part. On applying additional torque to failure, each case had a somewhat different failure mode. Consult the attached image for clarification.
1/8" bar - the bar did not noticeably deform on a large scale, but the small crimped area of the rivnut reamed its way through the aluminum to a depth of about 0.1" (I was able to put the versaframe tubing wall between the stock and rivet flange with no difficulty) when the machine screw sheared at the head end of the threads. Because the failure was through reaming not bending, I expect that if the force had not been applied normally to the surface, far less force would have been required to pull the insert from the material. I definitely intend that we shall use the longer .225 rivets for aluminum of this thickness. In steel, it is entirely possible that the stock would have held the rivnut to a rather greater pull out force.
0.1" Vex tubing: The crimped area of the rivnut was adequate to hold in the aluminum. However, the wall of the tubing was deformed (volcano style) until the hole was large enough for the insert to be pulled completely clear of the tubing. It appears that a longer rivnut would provide minimal additional capacity, though I shall test this when the longer inserts arrive.
1/16" plywood edging: The aluminum yielded "volcano style" to a height of 1/16" on one side and 7/64" on the other, then the machine screw broke at the top of the treads on the insert, presumably from the bending torque.
Curiously, the VF thick wall tubing deformed much farther than the hardware store grade aluminum bar and plywood edging.
Edit/Addition:
Quote:
Originally Posted by Fusion_Clint
Ok,
Having never used a rivnut before, I started looking on youtube and came across this. https://youtu.be/3Og-tHIR_0I?t=66
For those of you that have rivnut experience, what are the cons of using this method? |
Quote:
Originally Posted by RoboChair
Installation is time consuming and rather annoying. Anything larger than a #10 rivnut takes some considerable muscle and/or the use of a cheater bar(s).
|
I will add testing this method (with 10-32 thin wall steel inserts) to my next round of tests. It seems to me that you can get better torque with hex head bolts 1/4"-20 and larger than with any normal #12 and smaller heads. I am also concerned that applying torque to the threaded part of the insert during the crimp might result in a weaker/twisted crimp.