Richman,
This is an excerpt from a paper on distortion by The Welding Institute in Britain.
Check Item 13.
Welds going bananas?
Fabricators frequently ask: 'How much should I allow for distortion caused by welding?'. The formulae for calculating the shrinkage that will occur in an arc welded fabrication can be applied only to very simple cases. Distortion of arc welded components is generally caused by two factors: shrinkage of the cooling weld metal and local expansion and contraction of the plate.
Longitudinal shrinkage shortens the weld, transverse shrinkage decreases the width, and angular distortion causes rotation of the plates. Apart from these simple effects of shrinkage, longitudinal contraction of a weld may cause components to bow in a direction depending on the location of the weld in relation to the neutral axis of the component. The middle of a length of weld will bow towards the neutral axis.
Some values for shrinkage quoted in TWI's booklet Control of distortion in welded fabrications (available from Woodhead Publishing) are as follows:
Transverse shrinkage
Fillet welds: 0.8mm per weld where the leg length of the weld does not exceed 0.75 x the plate thickness.
Butt welds: 1.5-3mm per weld for 60° V joints, depending on the number of runs per weld.
Longitudinal shrinkage
Fillet Welds: 0.8mm/3m of weld.
Butt Welds: 3mm/3m of weld.
However, it may not always be necessary to make allowances for shrinkage at the assembly stage. This is particularly true when only one set of dimensions is important, for example, that which relates to the overall length of the fabrication. In this case, it is often possible to leave the fitting of one member, such as an endplate, until the remainder of the fabrication has been welded and most of the shrinkage has taken place. The remaining item is then fitted and welded in its correct position.
The above allowances for distortion apply to welded joints that are free to move; in practice, the restraint built up during the fabrication will determine the distortion. TWI welding engineers have, over the years, built up a wealth of experience available to Industrial Members to help them avoid distortion. Rectification of distortion is possible by the use of mechanical force or judiciously applied heating, but the cost of correction is generally at least ten times that of making the job to the required dimensional tolerances in the first place.
It has also become increasingly possible in recent years to use computer-based modelling to predict likely distortion and develop fabrication procedures to minimise it.
Rules for minimising distortion during welding
1. Design fabrications so that welds are balanced each side of the neutral axis.
2. Do not over specify fillet weld sizes.
3. Use double sided welds rather than single sided, and minimum bevel angles, to reduce the amount of weld metal.
4. Use minimum gap sizes.
5. In non-fatigue sensitive areas use intermittent fillet welds where possible.
6. Use double fillet welds where possible, rather than full penetration T butt welds.
7. Use clamps, strongbacks, jigs or fixtures.
8. Use welding positioners so that welding can be carried out in the flat or horizontal-vertical positions with high deposition rates. 9. Deposit a few weld runs alternately on each side of the joint in double V butt welds. 10. Weld a large construction from the centre outwards.
11. Use high speed welding processes where possible, eg, iron powder MMA electrodes, MIG welding or mechanised welding.
12. Use frequent tacking.
13. Balance welding on each side of the neutral axis, i.e. do not weld all one side before starting the other.
14. Weld fabrications clamped back to back and preset if possible; alternatively stress relieve before releasing from the clamps.
15. Use block welding to prevent movement.
16. When block welding thick plate, butter the sides of the preparation and build up the buttering progressively towards the centre of the joint, so that most of the joint can contract transversely before the joint is bridged.
17. Weld first the joints that cause the most contraction.
18. Make use of sub-assemblies.
19. Make frequent dimensional checks during welding, and if distortion is evident change the welding sequence or the clamping arrangements accordingly.
Have you checked the straightness of your piping ? Is there any angular distortion ?
As for the properties of your weld I would think hardness testing would tell you if you have put too much heat into one area. Others on here are more knowledgeable with the metallurgical aspect and may be able to advise,
Regards,
Shane
Block welding results in very high residual weld stresses at the tie-ins. The block results in a highly restrained condition that limits transverse shrinkage at the tie in. This can cause transverse weld cracking. If the joint thickness is greater than 3/4"-1", the risk would be much greater than in thinner material. In stainless steel, the high residual stress may be an issue in service conditions where stress corrosion cracking is a concern. There can also be problems with complete fusion or trapped slag if the ends are not prepared properly.
Interesting conversation.
I have a major fabricator that requires block welding of butt joints when long seams are involved. One purpose is to minimize distortion by using the block technique to tie the structure together with partial welds (not partial joint penetration) along the joint to provide structural integrity for alignment purposes and to minimize transverse shrinkage.
The block technique can be utilized to minimize transverse shrinkage/contraction. Each block increases the rigidity of the joint and prevents further transverse shrinkage and contraction as additional "blocks" are welded to completion. The alternative is the cascade technique, but increased NDT of the starts and stops are required to ensure weld soundness in comparison with the block technique. Each thermal cycle, i.e., each weld pass, increases the sum of the distortion. For simplicity, the distortion can be considered to be additive for each additional thermal cycle, i.e, weld pass. The block technique is limited to a relatively small area of the total joint length and it does incur transverse shrinkage (as the weld undergoes phase transformation, i.e., liquid to solid, and then contraction as it cools to room temperature and transforms from body centered to face centered cubic and back to a body centered cubic crystalline lattice structures). As each block is welded, the preceding block (weld), now cooled to the preheat temperature or ambient temperature, increases the joint rigidity and resist the transverse shrinkage and contraction of the block being welded.
As each block is completed, the ends (starts and stops) are ground to remove the inevitable areas of incomplete fusion bound to occur as the welder initiates the arc and the crater areas where there may be some tendency to crack where the weld are terminated. NDT in the form of magnetic particle testing or penetrant testing can be performed to verify weld soundness before proceeding to the next area to be block welded. Areas between adjacent blocks may be several feet apart and then additional blocks are welded in between existing blocks until the entire joint is completed.
Those of you that weld pipe using the technique where the weld is started at the 6:00 O'clock position progressing toward the 12:00 O'clock position are fully aware of the tendency for the root to close up as the weld progresses to it termination point. A better sequence is to weld in quarters, i.e., 6:00 to 9:00, 3:00 to 12:00, 9:00 to 12:00, and 6:00 to 3:00. The residual stresses are more uniformly balanced and the tendency for the root to close up is reduced.
The block technique described is similar to welding pipe in quadrants in the later paragraph.
Best regards - Al
Good stuff Al.