African Fusion

19 / 40

March 2016 AFRICAN FUSION 17
Welding of tubular k-joints
Figure 6: A simplified 3D computer model of a tubular K-joint.
Figure 7: The total distortion for different gaps between the braces.
Figure 8: The maximum principle stress of the sample welded at
270 A (right) and 320 A (left) is due to the larger gap between braces.
Figure 10: The residual stress along a measuring line normal to the
connecting line between the braces (welded at 270 A).
construction, it was important to generate a model that is as
realistic as possible and which is reliable. However, because
computing power is limited, meaningful simplifications had
to be made.
To reduce computation time, a simplifiedoutline of a single
K-joint of a tubular truss was created, as shown in Figure 6.
This simplification consists of a bearing under the chord, two
braces fixed by two rings and two clamps on the upper end of
the braces. The clamps symbolise the welded joint to another
chord. Some preliminary results of the simulations are covered
in this section.
Themain points of interest in the evaluation are distortion
and residual stresses. The residual stresses are comparedwith
the results from the hole drilling method, but for detecting
the distortion no measuring systemwas available. Due to the
strong relationship between residual stresses and distortion,
however, a comparison of the stresses gives information about
the levels of distortion.
Figure 7 shows an example of the variation of the gap be-
tween the two braces. It is apparent that distortion becomes
stronger with increasing distance between the two crown toes.
Themain reason is the relationship between residual stresses
and high heat input, because of the longer joint between the
braces.
The residual stresses are shown below. Figure 8 and Fig-
ure 9 show the maximum principle stress and the principle
stress minimum for two different welding procedures and
a varying gap. Limiting both colour legends for a range of
+500 MPa to -500 MPa in Figure 8, it is evident that the stress
around the weld seams does not differ. Disregarding the area
between the braces and only focusing on the crown toe, crown
heels and saddler clearly shows that the different welding
speeds and welding currents induce almost the same results.
Similar to Figure 8, Figure 9 validates the finding that both
welding procedures behave similarly.
Figure 9: A principle stress minimum of the sample welded at 270 A
(right) and 320 A (left) with different gaps.
To examine the calculated residual stresses, a measuring
line with five measuring points at almost the same position
as the hole drilling method (discussed in the next section)
was inserted. The results are shown in Figure 10. Maximum
principle stress (σ1) and principle stress minimum (σ 2) are
plotted against distance from the weld centre.
Welding experiments
A critical comparisonbetween calculatedandmeasuredvalues
can be carried out after manufacturing real tubular K-joints
with a robot welding system, as shown in Figure 11. A robot
welding system guaran-
tees constant conditions
for the whole test series.
Figure 11: A tubular K-Joint
after preparation for the
hole drilling method.