I have a project that includes some standard AISC clevises on threaded rods in pure tension. I am trying to find guidance on how best to address the possibility that the rod could rotate with respect to the clevises over time, causing the clevis to eventually disconnect from the rod. My belief, and that of the engineer on the project, is that this is not a likely problem unless the rod was expected to ever see near-zero loads, and we have specified a jamb nut in any case, just to be conservative. All the same, I am wondering if there is any code or documented standard practice that would back us up for situations like this where pretension is not controlled.
Ideally, a code would answer this question in the general case, but some specifics of our application are as follows:
The material is ASTM A36 for both the rods and the clevises. The rods theoretically experience around 3 kips of tension under dead load only, and about 10 kips under the code maximum live load (controlling). They only experience a load reversal (go into compression) during a code-level seismic event. The threads on the rod are right hand at both ends (not turnbuckle style.) The clevis is AISC #3 size with a 1" DIA rod. Both clevises are themselves restrained against rotation. The axis of the rod is plumb. At installation, we will be able to verify that each rod is substantially loaded, but not necessarily that they are sharing the load equally.
I have looked at some other clevis assemblies like this used for diagonal bracing of frames, and they don't have even a jamb nut, but I doubt they experience the same sustained tension, and they aren't in the same orientation.