If you want to make serious assemblies out of 3D printed parts, you’ll eventually need to deal with threading. The easiest way is to make a nut trap that you can either insert a standard nut into after printing or even during printing. However, there are limitations to this method. If you want a real threaded part you can print the thread, cut the thread with a tap or bolt, or use a threaded insert. [Stefan] ran some tests to see how each of those methods held up to real use. (YouTube, embedded below.) He used fifty test parts to generate data for comparison.
We like the threaded insert method where a brass insert is pushed into the plastic while hot. Special features in the insert cause the brass part to grab the plastic, making it difficult to pull the insert out or twist it within the hole. Another thing we liked was that the tests used holes printed in the horizontal and vertical plane. You can clearly see that the orientation does alter how the holes work and fail to work.
The test holes were M5. [Stephan] noted that you can’t tighten a bolt quite to the recommended torque with these methods, but you can get pretty close. It makes sense that the threads were not usually the part that failed, but surrounding material would crack or delaminate.
Although the printed threads performed somewhat worse than the other methods, they still worked surprisingly well. Of course, you would imagine the threads would wear faster than the brass inserts if you plan on disassembling and reassembling the part multiple times.
As a final trick, [Stephan] shows how he selectively increases the shell thickness around holes to increase the strength. This works differently depending on the slicer you use, but most modern slicers can change parameters within a specified volume of your print.
We’ve seen custom inserts made with nuts that ought to work pretty well even though they don’t have the features to grab the plastic. If you are making a sizable run, you might want to cobble up a thermal press first.