Residual Stresses and Distortion in Welds
Earlier in this chapter stress was likened to an internal load existing inside of the metal. Stresses in metal are a result of external loads and when the external load is removed, the stresses are relieved.
Stresses inside metals can also result from other causes, such as cold working, machining, grinding, heat treating, casting, and welding. Since there are no external loads to remove, these stresses cannot readily be relieved and they remain locked up inside the metal. These stresses are called residual stresses. Residual stresses can be relieved by heat treatment, by physically removing a section of the metal, or by yielding (distortion) of the metal.
Residual stresses always react against other residual stresses inside a metal in order to achieve a balance. No attempt will be made here to show exactly how these stresses are distributed inside the metal, but rather to show their cause and effect.
An edge weld is shown in Fig. 11-1 BA. Assume that this weld bead could be separated from the base metal while it is still very hot, perhaps just below the solidification temperature. Of course, the metal bar adjacent to the weld bead would also be very hot however, for simplicity, let it be assumed that it is at room tempera
ture. This condition is shown in Fig. 11-18 B, where the length of the bead is equal to the length of the block (L*).
Let the weld bead cool to room temperature. When this happens the weld bead shrinks in length and, as shown in Fig. 11-І8С, the length of the weld bead, L2, is less than the length of the block, Lj.
In order to attach the weld bead to the block so that the ends of the bead are flush with the ends of the block, it is necessary to pull on the bead in order to stretch it, and to push on the block to compress it, Fig. 11*18D. Pulling on the bead results in a tensile stress in the metal inside of the bead. Similarly, a compressive stress is set up inside the block in the region adjacent to the bead.
In Fig. 11-18E, assume that the weld bead and the block are now firmly joined together. The block, having been compressed, wants to stretch out again because of the elasticity of the metal. In doing so it
Fig. 1 1-18. Schematic drawing showing why residual stresses exist in wetds and their effect in causing weld distortion. |
pulls on the weld bead thereby setting up a permanent residual tensile stress in the metal inside the weld bead. The weld bead, in turn, was stretched in order to join it to the block. The elasticity of the weld metal causes it to want to pull together, thereby exerting a permanent compressive force on the metai in the block that is adjacent to the bead.
The end result is that a permanent residua! tensile stress remains in the weld bead and a permanent residual compressive stress remains in the metal in the block that is adjacent to the bead. In actual practice, the situation is much more complex; however, this provides a model for visualization.
The reaction of the weld bead on the block in compressing the metal adjacent to the bead tends to bend the block, as shown in Fig. і 1-18F. The degree of actual bending that might occur will depend upon many factors such as the size of the block, the size of the weld, and the restraints upon movement placed on the block by other members to which it may be attached.
Similarly, it can be shown (Fig. 11-19) that a butt weld has residual tensile stresses in the weld bead and residual compressive stresses in the metal within the plates that are adjacent to the weld. In this case, however, the two plates react against each other, if they are of approximately the same width, and they will not bend as shown in Fig. 11-18E.
However, if the plates are unrestrained, as in Fig. И-20А, the weld will tend to offset them at an angle “X,"’ as shown in Fig.
11- 20B. The reason for this is the shrinkage of the metal in the weld bead upon cooling to room temperature. As can be seen, there is
Fig. I Ы9. Generalization of the action of residual stresses in the lengthwise direction in a butt weld. |
more weld metal at the top of the bead than at the bottom; therefore, the top will shrink a greater distance than the bottom. This action results in the movement of the plates as shown.
If these plates are prevented from moving, as in Fig. 11-20C, the metal in the weld is prevented from shrinking. While the temperature is such that the metal in the weld bead is soft and plastic, it will deform; but when it has cooled down enough to obtain strength it will be stretched elastically, thereby setting up a permanent residual tensile stress in the weld bead.
The residual tensile stress in the weld bead tends to pull on the plates to which it is attached, and in doing this sets up residual
Fig. I і -20. Generalization of the action of residual stresses perpendicular to the lengthwise direction in a butt weld. |
tensile stresses in the plates, Fig, 11-20D. Again, the actual stress pattern in the weld is more complicated; however, this example helps to visualize why distortions occur and how residual stresses within the weld occur. Without going into the reasons why at this time, the welded plates in Fig. 11-20С would tend to distort in the general direction of the plates in Fig. 11-20B.
In Fig. 11-21 A, the fillet weld causes the plates to move at an angle, in a manner similar to the butt weld in Fig. 11-20. If the plates in Fig, 11-21 are restrained from movement, residual stresses will be set up in the weld bead and in the plates being joined.
When a fillet weld is made on relatively thin plates, where the temperature rises to nearly a red heat at the bottom of the lower plate, the metal may be upset, as shown in Fig. 11-2IB and C.
The metal in the lower plate is heated in the region of the weld and wants to expand. However, it is prevented from doing so by the surrounding colder metal and the restraints. As a result, the relatively weak hot metal near the weld is upset by compression and possibly bent. During cooling the plates were not sufficiently rigid to iron out the bend.