Experimental Study on Restraint Intensities of Fillet Welded Joints
Akio Itoh, Noboru Watanabe
This paper presents the results of the following experimental study: the intensity of tensile restraint (RF) of tack-welded joints and the intensity of bending restraint (KB) caused by side fillet welding performed as onesided welding, using I-section girders and hollow cylinders as structures.
In this paper, RF values are defined as the product of the intensity of tensile restraint coefficient (p) and the plate width (h). After tack-welding, p values were measured. Subsequently, welding was conducted: after cooling off, we observed the occurrence with or without cold cracking in the welded joints. The effect of KF on RB values and the relationship with the weldment cracking parameter (Pw), etc. were then determined. The major findings of this investigation are as follows.
1) The coefficient p for girders and cylinders is influenced by the geometrical mean thickness among the vertical and horizontal plates; it is influenced not so much by the number of joints as by the location along the wled line.
2) The effect of KB of girders and cylinders on the coefficient p was determined. In other words, the coefficient p in both kinds of models showed a tendency to decrease under the influence of the bending restraint. However, it was observed that even as KB increases, the p values for the girders and cylinders converge approximately on 8 and 4-5 kN/mm.mm/mm, respectively.
3) The occurrence of cold cracking of girders can be determined by using the relation between Pw and Tcr of butt joints (graph) as RF increases, even in the case of high KB values. On the other hand, for cylindrical models, cracks were sometimes observed in the welded joints with relatively small KB and RF values; thus, making it very difficult to judge cold cracking from relational graph mentioned above.