Testing Torque-Out of heat-set and molded-in inserts

First, make sure you’ve read this page on heat-set inserts: Heat Set Inserts . There are a bunch of PDFs there that should help you understand the failure modes (pull-out, torque-out), and various installation processes.

For the fuse bolt-down locations, we are planning on using an inserted thread so that we only need to use 1 tool to install them in the battery box, and we still need to be able to tighten them to the proper torque.

I printed out 2 different versions of inserted threads, both with 100% infill. The first one was a regular heat set insert with a slight taper in the hole, and the other was for a molded-in insert with epoxy. They were all M6 inserts from Spaenaur.
Note that we can assumed that the molded-in inserts will be a little better than the heat-set inserts in epoxy as they have deeper knurls.

Test Setup

I used a socket wrench and a crane scale to measure the torque (Force * Distance) at failure. The printed pieces were clamped in the vise on the lower part, and was not clamping on the sides where the heat set insert was (the insert did not go all the way to the bottom and the vise clamped on the part where the insert did not reach).

The printed parts were printed at 100% infill with Orange PETG from DigitMakers on a Prusa Mini.

The heat set inserts were installed per the recommendations here: Heat Set Inserts

For the molded in inserts, we used MG Chemicals 9200 Structural Epoxy Adhesive, and let it sit for 48 hours at room temp to cure. It would have been good to test the cure at elevated temperatures (above room temperature) and test multiple different adhesives (3M DP420 for example).

Open

Failure Methods

The 1st test sample with the het seat insert failed by stripping the brass threads of the heat-set insert.

Open

The 2nd heat set insert sample failed by cracking the plastic. For this failure method, a larger plastic edge around the heat set insert would increase the strength of the plastic. Note that with both the heat-set inserts, the fastener tightened to a max torque and then took another almost full turn to break/fail.

Both the molded in inserts failed by stripping the epoxy - the insert just turned freely inside epoxy.

Test Results

This points to some possible better results with a stronger epoxy and possible better cure times and proper molded-in inserts.

But what load to we really need?

Open

The above PDF and the previous research into the tightening torques for the series connections (Electrical Contact Resistance for Module to Module Series Connections ), as well as the recommended tightening torque for our EV200HAANA contactors (8.8 - 11Nm of M8 nut for 500A rating), leads me spec a tightening torque of 8Nm for these joints, which can safely be reached by the 100% infill regular heat set inserts and greatly simplify the manufacturing process. Care should be taken to not tighten the bolts when the plastic is cold, as the plastic is brittle in cold temperatures. We may print the parts from Polycarbonate instead of PETG for the final version.

This was a less intensive research / torque specification process than I would have liked to do (ideally do the tests with other epoxy, get molded in inserts, test with polycarbonate, and test the actual electrical contact resistance) but it should be good enough. I don’t foresee major issues with this as long as all the proper procedures are followed during manufacturing/assembly/maintenance.