Residual tensile stresses

Date 07-14-2017 13:44

Hey waqasmalik,
Interesting question.
Definitely not the second pass since it's closest to the neutral axis.

I would guess that the top pass has the highest because it has the largest volume of weld metal contraction.

But..the first pass stress might be highest due to the additional stress caused by the contraction of the top pass. hmmm.

Tyrone

By 803056

Date 07-14-2017 14:18

One must recognize the adjacent base metal plays the greatest roll in the amount of distortion induced by welding.

The root bead is in tension assuming the two opposing members are fixed, i.e., rigid. As the weld cools, solidifies, and contracts, it is in tension in both the longitudinal and transverse directions. If the members are not fixed, i.e., rigid, the weld is free to contract as it cool in the transverse direction. The same cannot be said in the longitudinal direction since the members are rigid in that direction.

As successive weld beads are deposited, they cool, solidify, and contract. They are in tension in both the longitudinal and transverse direction. Now comes the fun, eventually, the root bead is placed into compression because as it cools to the equilibrium temperature, it no longer contracts, it is dimensionally stable for that temperature. Meanwhile, the weld bead just deposited is at a high temperature, cooling, contracting, and in a state of tension. Once the weld and adjacent base metal cools to about 800 degrees F, it has a tensile strength of about 50% of its room temperature yield strength. Cooling below 800 degrees F causes the tensile strength to rise quickly to it tensile strength at the equilibrium temperature or as we casually call it; the interpass temperature.
The last beads deposited are restrained by the preceding beads that have cooled and have reached a state of equilibrium. Thus, the last beads deposited are in tension, while the root bead is or can be in compression as evidenced by the fact that the distortion "cups" the members upward toward the last beads deposited.

All of this ignores the fact the majority of the distortion observed is the result of the base metal adjacent to the weld heating, expanding, reaching yield, deforming by plastic flow, cooling, contracting, and producing residual tensile forces once it cools to ambient temperature. Mathematically, any carbon steel that is restrained and heated to a delta T of about 220 degrees F above ambient temperature has experienced plastic flow, deformed, cooled, contracted, and produced residual tensile stresses equal to its yield strength. However, that wasn’t the question.

Got to go.

Al

By Northweldor (***)

Date 07-16-2017 12:14

I hope you are going to continue with this interesting reply!

By waqasmalik (**)

Date 07-16-2017 14:06

Thanx Al.

As always a fantasticly explained reply.

I would expect something more as Northwelder, when ever u are free

By waqasmalik (**)

Date 07-16-2017 15:28

What about the effects of heat of top passes on residual stresses in already deposited weld beads.

By Metarinka (****)

Date 08-11-2017 17:45

Al,

Really knocked it out of the park. The answer is driven almost entirely by restraint and geometry of the parts. When I was doing residual stress analysis the highest residual stress numbers always came from the root or face surfaces which makes sense as they are furthest from the neutral axis.

Beyond that It is hard tto give a concise example. For example a pipe will tend to have maximum stress on the face which is a in transverse and longitudinal tensile stress. On single pass square groove welds the root surface was in the most transverse stress. There is no simple answer and often times intuition was overruled by stress analysis techniques or unaccounted situations for example fitup and residual stress from previous welding operations in the area of the weld. During multipass operations residual stress is cumulative so more passes will put more residual stress and thermal energy puts in more residual stress.

I'm going in circles, my answer is: depends

By 803056

Date 07-16-2017 17:56 Edited 07-16-2017 18:03

Assuming groove is welded from one side using multiple layers the first few beads will be in compression once the weld is completed. How can this be? First, let’s assume the length of the joint is very short so we can ignore what is happening along the length of the groove weld.

Consider the two members of a butt joint. The root bead is deposited, cools, contacts, and pulls the two members toward each other. Think of how the root opening in an open root joint tends to close as the welder progresses from one end of the joint to the opposite end. To counter this, heavy tack welds are placed along the joint to keep the root from closing up, yet the tensile stresses are sufficient to close the root slightly, but not to the same degree experienced if there were no tack welds. If there are no tensile stresses developed as the weld cools, the root would not close, i.e., the two members of the butt joint would not move toward each other.

Once the root bead cools to some ambient temperature, it no longer contracts, the dimensions stabilize. Remember, initially the members were not fix and were free to move in the lateral direction as the weld contracted. We are assuming a groove length is very short so we can ignore what is going on along the length of the groove weld.

The root bead now fixes the dimensions of the assembly in the transverse direction. In other words, the root bead offers restraint to successive beads deposited on top of the root bead. As the second bead cools and contracts, it can no longer pull the two members closer together due to the restraint offered by the root bead. The previous bead, i.e., the root bead, keeps the two members from drawing closer toward each other. The result is, the second bead (which is also the second layer) is in tension. That tensile force is countered by an equal but opposite force offered by the root bead which is now in compression. The sum of the forces must be equal to zero.

Since the two members were not restrained when the root bead was deposited, they were pulled toward each other as the root bead cooled. That isn’t the case with the second layer. The root bead attempts to maintain the position of the two butting members until the second layer develops a tensile stress that is greater than the compressive strength of the root bead. Ultimately, the root bead yields slightly under the tensile load imposed by the second layer cooling and contracting. A point of equilibrium is reached where the compressive strength of the root bead counters the tensile forces developed by the second layer. If that wasn’t the case, the two members would simply draw together. The fact of the matter is, the tensile forces developed by successive layers are higher than the preceding layers once cooled. The root bead become the hinge or point of rotation that allows the plate to rotate upward toward the last weld layers deposited. Meaning, the initial weld beads forming the first couple of layers are subject to compression due to the tensile forces developed by the beads that follow.

If the groove is welded from both sides, assuming a top and bottom groove, the joint will rotate upward if the top is welded first. The plate will then rotate downward if the bottom groove is welded after the completion of the top groove. However, rotation cannot occur until the tensile forces developed by the bottom most beads overcome the strength of the existing weld. Both the uppermost beads of the top side groove are in tension due to the contraction of the weld beads and the bottom most beads of the bottom side groove must be in tension due to the contraction that occurs while those weld beads cool.

If the top side and the bottom side are welded at the same time by alternating from the top to the bottom groove, the butt joint will neither bow upward or downward because the tensile forces are balanced. However, the tensile forces have to be countered by compressive forces in order to maintain equilibrium. The sum of the forces must be equal to zero.

The root beads and immediate layers forming the mid thickness of the groove weld, acting as the hinge or point of rotation, are loaded in compression. The compressive strength of the weld at the mid thickness must be equal to the counteracting tensile forces of the top most and bottom most weld layers. The forces have to balance or the plates would have to move in response to the unbalanced force. It is the old premise learned in statics class that the sum of the forces must equal zero.

The sketch indicates the approximate distribution of tensile and compressive stress. However, I suspect my sketch is an oversimplification. In the sketch, the weld is deposited on opposite sides by alternating from one side to the next with each successive weld bead. The maximum magnitude of the stress is equal to the yield strength of the filler metal in either tension at the weld surface or the root in compression.

That's the best I can do.

Al