The following is an excerpt.
Maximal expertise on how and where nuts and bolts fail can obviously be of critical importance. Here’s an abstract of a presentation by PhD candidate Erik Løhre Grimsmo in one of SFI CASA’s internal seminars recently.
“Thread failure of bolt and nut assemblies is generally undesired, as it is a more brittle failure mode than bolt breaking (fracture through the threaded shank of the bolt). Furthermore, failure of the threads due to over-tightening is difficult to detect during installation. Thus, determining the causes of thread failure is appropriate.
One factor that can determine the failure mode of bolt and nut assemblies, but that has gained limited attention in literature, is the number of the threads in the grip length (the distance from the head of the bolt to the nut). This is particularly relevant for partially threaded bolts, where typically only a few threads are within the grip.
To investigate this, pure tension tests on bolt and nut assemblies with different grip lengths were performed. The tests showed that four or less threads in the grip gave thread failure, whereas eight or more threads caused bolt breaking. For the tests with five to seven threads in the grip, both failure modes occurred. Finite element simulations of the tests were carried out to study the mechanisms occurring during failure. It is observed that necking of the bolt takes place at the engaged threads when there are few threads in the grip, which further induced thread failure.
Seemingly not considered
The current standards do not seem to consider this effect, and poor choice of fasteners can produce thread failure even though bolt breaking is the intended, and expected, failure mode. Some practical approaches to reduce the probability of thread failure are, for instance, to use fully threaded bolts rather than partially threaded bolts and to use high nuts rather than regular nuts.”