In the first part of this discussion, we learned that intermittent fillet welds are not always the lowest cost option, nor do they always reduce distortion.

In this second and final part, we'll examine one more factor that must be considered when specifying intermittent versus continuous fillet welds. Finally, we'll wrap this all up with a simple procedure that will ensure that you select the right fillet weld type.

When weldments are subject to repeated, cyclic loading, we must consider fatigue.

Fatigue cracking can initiate at stress concentrations where the localized stress exceeds the yield strength of the material. Even though the applied loads may result in elastic stresses, stress concentrations readily elevate the localized stresses into the inelastic range.

The toes of fillet welds and the ends of intermittent fillet welds are examples of such stress raisers.

Let's assume that the beam with the continuous fillet weld as shown in Figure 1 is subject to bending. In this configuration, the weld toes are not stress raisers — the toes are parallel to the stress field.

The ends of the welds don't create stress raisers — at the end of the beam, there is no stress. According to AWS D1.1 Table 2.4, these continuous fillet welds are Category B details. Category B details have the highest allowable stress range capability of any welded detail.

In Figure 2, an intermittent fillet weld has been specified. Each end of each segment creates a stress raiser. D1.1 classifies this detail as Category E. Category E details have the lowest allowable stress range capability.

It is important to note that the stress ranges for Category B and Category E details are for the stress in the base metal, not in the weld metal.

The weld metal in fillet welds is Category F, and this must be considered separately from the stress in the base metal. Fortunately, larger fillet welds can be used to ensure that Category F behavior is not limiting.

The simple point is this: Continuous fillet welds result in Category B behavior for the base metal — which is as good as you can get for weldments subject to fatigue. Intermittent fillet welds result in Category E behavior which is nearly the worst possible condition for the base metal.

In some cases, Category E behavior is not the limiting factor in the weldment's design. However, exchanging the best possible option — continuous fillet welds, Category B; for the worst option — intermittent fillet welds, Category E; should be considered only with caution. The potential savings in weld metal, or the reduction in distortion, or the reduction in fabrication time would come at a great expense if this is the case.

How can this mistake, as well as those identified in Part 1 be avoided? Here's a simple checklist:

Step 1: Divide the total force (F) transferred through the joint by the total joint length (l).

When this simple mathematical calculation is made with units of pounds for the force and the joint length in inches, the result will be the force per unit length (f), measured in units of pounds per inch.

Step 2: With the results from step 1, determine the required fillet weld leg size.

The fillet weld leg size can be determined from “look up tables”, or directly calculated, as follows:

This will yield the fillet weld leg size (ω) in inches.