By Spencer Erling,
Education Director, SAISC
It may come as a surprise to engineers and fabricators alike that tube to tube welds are not just a simple ‘all welds 6mm continuous fillet weld’ that we so often see on drawings.
The fact of the matter is that when we weld a ‘branch member’ (a truss internal lacing) to a ‘main member’ (a truss chord), then depending upon where you are on the connection i.e. toe, heel or sides you will have different weld requirements and sizes to conform with an AWS prequalified weld (see figures 1 and 2).
This is also strongly influence by the angle between the branch member and the chord. For instance, consider a weld required in the toe region. For an angle in excess of 120° it will not be possible to do a fillet weld at all. It therefore becomes necessary to cut away the tip of the steel and to do a prepared weld (see figure 3).
Along the sides of such a weld, this is the simple part of the weld, and then it is usually possible to do a fillet weld of the given leg size.
However, in the heel area, AWS assumes that for small (acute) angles that it will not be possible to get a welding torch or rod right into the root of the weld and so it ignores the first 3mm (called the Z zone). It also specifies that to achieve the equivalent of a given leg length’s weld strength, the weld needs to be 1.5 x the given size.
As mentioned in the article about Jeff Packer’s design course, as clearly depicted in the ‘side (box)’ sketch in figure 3, it is very important to land a rectangular lacing on the square portion of the chord box so that the welds do not get caught up in a radiused area and hence become a ‘flared groove weld’ which also has special requirements. But on the other hand, to eliminate secondary bending effects we would like the lacings to be as big as possible, but not to get close to the radius.
A matched set up (figure 4) would only be used for Vierendeel trusses (moment frames)
Table 1 summarises the requirements by AWS, by the size of the angle, for the increased leg length of welds to ensure the correct effective weld sizes are achieved. E is the effective weld size.
AWS does cover the requirements for complete joint penetration (full penetration) welds as well. In this instance, the weld in the heel area will always be a large fillet weld, as it is physically impossible to do a complete joint penetration weld.
AWS has specific requirements for fatigue related welds which emphasize the need for concave ‘flowing surface’ welds (figure 5).
There are of course going to be transition zones between the sizes and the toe or heel areas. The requirements for these are well detailed in AWS.
WHAT ABOUT OVERLAP WELDS FOR TELESCOPIC TYPE CONNECTIONS?
I have only once in my career seen this called up on a drawing. It was a German design which provided for close tolerance fit between the smaller and the larger tubes, which turned out to be seriously problematic with our South African range of profiles.
But if you are ever faced with such a situation, AWS has been there before…(figure 6).
JOINING TUBES OF DIFFERING WALL THICKNESSES TOGETHER
This is most likely to happen in long members where a change in wall thickness makes sense (figure 7). This should not occur more frequently than at 6.0 metre centres (standard tube lengths) but can go out to 9 metre centres (usually also available).
So now all you engineers and contractors out there, unless you can get a welding engineer to produce a more user friendly (qualified) weld procedure, if you joining highly stressed members together, please consult AWS to ensure that you achieve your effective weld size.
You have been warned!