Selection of Test Criteria

One of the most important factors in joint development testing is ensuring that the correct test criteria are utilized and that the test methods used to assess that criteria are fully capable of doing so. For example, the function desired of a threaded fastener is not always the same. In some cases it acts purely as a shear pin, in other cases the clamp load generated by bolt tension is required to prevent the separation of components under axial loading or slipping under shear loading. Although these requirements are different, in many joint development programs the same test criteria might be used in each case.

The most common test for threaded fasteners is to assemble them in the actual joint and torque them to failure. The installation torque is then set at a percentage of that failure torque. For applications where installation is the most critical element, like the tapping screw discussed in the previous section, this might be an appropriate test. However when threaded fasteners are used in structural application where the need to maintain a clamp load is usually required, failure torque should not be the primary test criteria. Figure 3 shows the how the “right” answer can change depending on the criteria utilized. This table shows the results of testing three different means of creating a nut function for use in a structural joint. The component containing the nut member is 16 ga. galvanized steel, while the clamped member is the same material in 12 ga. Tension was measured though the use of ultrasonics.

	Nut Member	Failure Torque, (ft-lb)	Rank	Tension at 18 ft-lb target torque, (lb)	Rank
	Extruded Hole	41.1	1st	768 lb	3rd
	Pierce Nut	40.2	2nd	870 lb	2nd
	Clinch Nut	35.0	3rd	1053 lb	1st

Figure 3 – Average torque-to-failure and residual tension of 5/16 – 18 Grade 5 fasteners into various nut members.

Perhaps the most important point to take away from Figure 3 is the small percent of the expected tension that is generally retained in threaded fasteners with very short grip lengths. In this case the “textbook” expectation of tension was just over 3300 lb. As that bolt will only have stretched about 0.0003 inches to achieve that tension with the short grip length of a12 ga. sheet, even minute levels of embedment after tightening will result in dramatic reduction in clamp load. The other common factor leading to tension degradation is where clamp members are not in intimate contact. The means of creating the nut feature in the extruded hole and the pierce nut leaves these areas that will relax during tightening. These areas are highlighted by arrows in Figure 3. Figure 4 shows a portion of the torque-angle plots of the torque-to-failure testing.

Figure 4. – A portion of a typical torque-angle plot for the 5/16 – 18 joints, showing the relative stability of the joints during tightening.

A review of these plots would loosely predict the relative tension achieved in the three methods based on the relative proportion of linear torque/tension relationship, as superimposed on the trace. Non-linear portions indicate relaxation in the joint during tightening, leading to greater rotation for a given increase in torque. Having the facility to record this type of data is critical to understanding the “personality” of the joint being tested. Simply reporting the torque at failure is does not provide any insight, and its relationship to maximum torque and torque at yield is not consistent, varying in particular with the fastener’s grade. Although secondary locking methods are often successfully employed to combat low residual torque conditions, working to establish a joint geometry that is capable of maintaining the required tension is always desirable.