When the metal expansion joint cracks, the first reaction is often "repair welding". Wait a minute, put down the welding torch first. The crack is not simply "cracked", there are reasons behind it, and the location is also particular. Fix it or not, how to fix it, read this article before starting.
Cracks are not simply "cracked": four common triggers
The most common cracks in metal expansion joints are not "bad quality" at all, but these four situations are at fault.
Alternating fatigue.The pipeline expands and contracts repeatedly, and every time the bellows expands and contracts, the peaks and valleys experience a stress cycle. Over time, the microscopic cracks slowly expand into penetrating cracks. There is usually more than one such crack and it tends to be distributed in parallel near the trough.
Stress corrosion.In the medium containing chloride or sulfide, the surface passivation film of stainless steel corrugated pipe is destroyed, and the crack will drill along the grain boundary as soon as the tensile stress is superimposed. The crack forked like a tree root. It looked fine to the naked eye, but it was actually crisp.
Media flushing.When high-speed airflow or dust-containing medium directly impacts the bellows wall, especially when the guide tube fails, the local wall thickness decreases and then cracks are worn through. Such cracks have irregular edges and are often accompanied by obvious thinning zones.
Installation flaw.Forced matching, cold drawing exceeding the limit, and improper support dismantling lead to the bellows bearing additional forces outside the design. This kind of crack often appears in the joint or straight edge section, and is characterized by a tear caused by "one-time force" without traces of fatigue growth.
You see, the reasons are different, and the way it is handled is completely different. Repairing welding without checking it out is equivalent to taking medicine without seeing a doctor.
Whether it can be repaired depends not on the size of the crack, but on where the crack grows
The crack is only 5mm, can it be welded? Truth be told, the problem isn't length, it's location. The size of the crack determines "how to repair it", and the location of the crack determines "whether it can be repaired".
At the crest:The radius of peak curvature is large and the stress is relatively low. If it is a slight surface crack, it is feasible to repair welding after polishing it off. However, the premise is that the grinding depth does not exceed 30% of the wall thickness, and PT inspection has to be done after welding.
At the trough:That's the trickiest part, tsk. The trough is the stress concentration zone, where every stretch and contraction exerts force. Surfacing here is equivalent to patching a repeatedly bent sheet of iron, forming new stress concentrations around the solder joints, and it won't take long to crack from the edges again. Therefore, the consensus in the industry is that direct surfacing is not recommended, but it is better to change the bellows.
Straight edge segment:The straight edge section does not participate in the corrugation deformation, and the stress is mainly axial. As long as the crack does not extend to the trough, the success rate of repair welding is high. Groove before welding, polish smooth after welding, and don't save the heat treatment that should be done.
Joint weld:If it is a weld crack between the joint and the bellows, it is necessary to first judge whether it is a welding quality problem or excessive force. The former can be cleaned and re-welded, while the latter has to check the pipeline support and cold drawing amount, otherwise it will crack after welding.
Correct posture for field repair welding
If it can be repaired after evaluation, repair welding is not something that can be done by just finding a welder. The wall thickness of bellows is usually only 0.5~2mm, which is as thin as paper, and it will be worn as soon as it is welded. Even if you don't wear it, the heat input is large, the shape of the ripples changes, and the flexibility is wasted.
Controlling the heat input is the first priority. Use tungsten argon arc welding, small current, segmented jump welding, each section does not exceed 20mm, after welding one section, wait for the temperature to drop to normal temperature before welding the next section. Conditionally, the back is filled with argon protection to prevent oxidation.
The welding material must match the base metal. Use 304 welding wire for 304 bellows, 316L for 316L, don't mix. Some manufacturers use ordinary welding electrodes in order to save trouble. As a result, the corrosion resistance of the weld is much lower than that of the base metal, and it doesn't take long to corrode and crack.
Direct surfacing is generally not recommended at the trough. When it is really necessary to weld, it is necessary to obtain design consent, and use low-strength matching welding materials to make the weld seam appropriately "soft" and transfer the force to the base metal. But this is only a temporary measure. It is counted as a year for one year, and it is still necessary to arrange a replacement.
Acceptance and life reduction after restoration
Repair welding is not finished after welding. Stress testing has to be done. Perform the hydraulic test at 1.5 times the design pressure, or test with 1.1 times the air pressure (pay attention to the safe distance). Before the test, the weld should be polished to be flush with the base metal to facilitate flaw detection.
In terms of flaw detection, PT (penetration detection) is enough for surface cracks, and MT or UT is added for suspected internal defects. But I recommend doing PT at least once to make sure no microcracks remain.
The account of life reduction should be clearly calculated. It is conservative to estimate the fatigue life of repair welded bellows according to 30% ~50% of the new pipe. If the original design life is 10,000 cycles, it should be reviewed if it has passed 3000 times after repair welding. Don't expect a weld for all, the repair just delays the replacement time, not cancels the replacement.
Therefore, before repairing, ask yourself: How long can this expansion joint last? If the whole system is about to be overhauled, it is no problem to temporarily repair welding it. If you have to fight for another five years, you might as well replace it with a new one. Calculating downtime losses, new waves are cheaper than repairs.
Source control is more cost-effective than repair
Finally, let's be honest.
When selecting, don't only focus on 304 for bellows material. 316L or 254SMO for flue gas desulfurization environment, Inconel625 or special heat resistant steel for high temperature steam. In terms of waveform design, multi-wave number is more resistant to fatigue than small-wave number. Although it is a little more expensive, the life may be twice as long.
The key points should be paid to the inspection during operation. Focus on whether there are rust spots, pits and bulges on the surface of the bellows, especially the troughs. Measure the wall thickness regularly with a thickness gauge, record the data, find the thinning trend and deal with it in time. If possible, install a displacement indicator to observe whether the actual expansion and contraction of the bellows is within the design range.
Two days ago, I met a customer, saying that the expansion joint cracks every year, and every time I find a welder to fix it. Later, we looked at the scene and found that the pipe bracket was crooked, and the thermal displacement was fully pressed on the bellows. Once the bracket was adjusted, it didn't leak again for three years. You say, is this repair or governance?
Crack repair is never a matter of torch touch. Judge whether to repair or replace, evaluate the crack location, control the quality of repair welding, reduce the acceptance life, and then plug the leak at the source-every step is done in place, so that the system can run steadily.