High temperature is not as simple as 300℃
Two days ago, I met a customer who was doing power station maintenance. When I came up, I complained: the expansion joint of the same caliber, the next unit was fine after eight years of use, but it cracked in less than a year. When removed, the surface of the bellows is full of dense cracks, especially the trough. He asked me: Did the manufacturer cut corners?
Not necessarily.
The metal expansion joint of the high-temperature pipe looks like an iron lump, but it is actually a precision pressure-bearing component. Many people think that it is enough to look at the temperature when selecting the type-how much is the pipe? Isn't it enough to find a material with a higher temperature resistance grade? If it was that simple, there wouldn't be so many cases in the market that go bad in one year.
Temperature is just the starting point. What really determines the life of the expansion joint is the comprehensive working condition superimposed by three factors: temperature, pressure and medium. The requirements for expansion joints of 300℃ steam pipeline and 300℃ sulfur-containing flue gas pipeline are very different. The former mainly tests the durable strength of the material, while the latter has to deal with additional corrosion. How large is the amount of thermal displacement? How to balance internal pressure thrust? Are there chloride ions in the medium? If these problems are not clear, the chosen expansion joint is just a chance.
Materials are the watershed: from 304 to Inconel
In this material level, we can best see whether a manufacturer is carefully selecting or shipping a set of templates.
304 stainless steel begins to show obvious sensitization phenomenon above 400℃-carbide precipitates at the grain boundary, resulting in the material becoming brittle and the corrosion resistance dropping by a cliff. At 600℃, the creep rate rises sharply, and the bellows will slowly deform like plasticine. If you still use 304 at this time, you are not choosing an expansion joint, but laying a time bomb.
So there are improved versions of the 316L, 321, and 347. However, many people don't know that 316L has been in service for a long time above 500℃, and it also has a tendency to graphitize. When it really reaches above 600℃, or there are strong corrosive components in the medium, you have to change to superalloys or even nickel-based alloys, such as Inconel 625. This material can maintain good long-lasting strength at 800℃, and its ability to resist chloride ion stress corrosion far exceeds that of stainless steel.
Does it mean that everything will be fine if you go directly to the nickel-based alloy? Neither is it. Material upgrades mean doubling or even several times the cost, and the fatigue performance of nickel-based alloys may not be much better than that of high-quality stainless steel. Selecting materials is essentially a technical and economic problem, which has to be settled, and customers can't pay for unusable performance.
Structural design hides most of your life
With the right materials, the next step is to look at the structure. The bellows itself is the heart of the expansion joint, but no matter how good the heart is, it must have matching blood vessels and stents.
Let's start with wavenumber and wall thickness. Many people have a misunderstanding that the thicker the wall thickness, the more durable it is. In practice, the opposite is true-multi-layer thin-walled structures are much more resistant to fatigue than single-layer thick-walled structures under high temperature conditions. Because it is also subjected to internal pressure, the multilayer structure has less stress shared by each layer, and the interlayer slip can absorb a part of the deformation energy. This is why the corrugated expansion joints used in the power station industry are basically multi-layer structures. Of course, the number of layers is not better, and the interlayer friction and heat dissipation problems will worsen with the increase of the number of layers.
Let's talk about the deflector. This thing is often overlooked, but in high-temperature dusty airflow, the guide tube is the life-saving talisman of the bellows. Without the guide tube, the high-speed airflow directly washes the bellows, and coupled with the thermal stress caused by the temperature gradient, the crack is only a matter of time. With the guide tube installed, the airflow goes through the inner tube, and the bellows only undertakes displacement compensation, so the working conditions are much friendlier instantly.
As for the way of restraint, it is even more particular. High-temperature axial expansion joint is suitable for straight pipe section to absorb axial displacement, but the internal pressure thrust must be borne by the pipe support; If the space is limited, it is necessary to use the double hinge transverse expansion joint, so that the transverse displacement is absorbed by the hinge group; If the pipeline cannot withstand the blind plate force, it is necessary to choose a straight pipe pressure balanced expansion joint, so that the pressure thrust can balance itself inside the expansion joint. If you choose the wrong type of constraint, no matter how good the material is, the bellows will be pulled alive by extra force.
The failure scene does not lie
After working for so many years, my biggest experience is that failure analysis is the only criterion to test whether the selection is correct or not.
High temperature cracks generally appear first in troughs or peaks? The vast majority of cases are troughs. Because the curvature radius at the trough is the smallest, the stress concentration is the most serious, and in actual operation, the corrosive substances in the medium are easy to accumulate in the inside of the trough. When you look at the failed bellows, nine times out of ten the cracks start in the inner wall of the trough.
Another example is the difference between intergranular corrosion and chloride ion stress corrosion. The cracks of intergranular corrosion go along the grain boundary, and the fracture has no obvious plastic deformation, like being gnawed by insects; The cracks caused by chloride ion stress corrosion are typical dendritic bifurcations and usually occur where tensile stress is concentrated. The two countermeasures are completely different-the former should reduce the sensitization tendency of materials, while the latter should either change to chlorine-resistant materials or reduce the tensile stress level. There is also a type of failure that is purely pitted by the pipeline system. The fixed bracket is not strong, and the gravity of the pipeline is fully pressed on the expansion joint; Or the cold tightening was not done correctly, and the bellows kept working under over-limit displacement. In these cases, the design of the expansion joint itself is fine, but it is still short-lived.
Selection is not a set of parameter tables
When choosing metal expansion joints for high-temperature pipelines, the account to be calculated is far more than the column of temperature resistance grade.
Let's start with how to discount the allowed displacement. The rated displacement given on the manufacturer's sample is the data under normal temperature fatigue life. Every time the temperature increases by 100℃, the material creep increases, and the allowable number of cycles and displacement must be adjusted down accordingly. If it is not discounted, according to the normal temperature parameters, the actual life may only be a fraction of the design value.
Let's talk about installing pre-deformation. When the bellows leaves the factory, it is usually in a free state. If the thermal displacement direction of the pipeline during operation is stretching, and a certain amount is pre-compressed during installation, the bellows can work symmetrically between compression and stretching during operation, and the fatigue life can be improved a lot. But the amount of pre-deformation is not good-more or less is a matter.
The worst thing is the blind pursuit of "thicker". As mentioned earlier, thick-walled single-layer structures are more prone to fatigue cracking under high-temperature cycling conditions. Some users don't understand this truth, and think that enough materials are good goods. As a result, they spend twice as much money and their life span is cut in half. Type selection is really not something that can be solved by heaping parameters.
Installation is not finished
Many users think that the expansion joint will be installed once and for all, and they don't think of finding the manufacturer until it leaks. In fact, the maintenance rhythm of the expansion joint of high-temperature pipeline should follow the operation stage.
At the beginning of operation, the focus is on whether the tie rod nut is loose. After several hot and cold cycles, the thermal relaxation effect of the bolt will reduce the preloading force, and it has to be re-tightened at this time. During the full-load operation stage, regularly check whether the guide tube has been eroded and perforated, and whether there is "orange peel" or "bulge" on the surface of the bellows-these are precursors of creep. Once they appear, it means that the material is approaching its life limit. When the furnace is shut down for maintenance, drill in to see if there are any cracks in the inner wall of the bellows, especially the trough position, and it will be clear with a colored flaw detection photo.
Then when should I find a manufacturer for a comprehensive evaluation? My advice: don't wait for a leak to replace. If it is found that the bellows has obvious discoloration and deformation, or the expansion joint in the same position is damaged more than twice in a row, it is time to ask the manufacturer to make a comprehensive diagnosis. Spending thousands of dollars for evaluation is much more cost-effective than losing hundreds of thousands a day for shutdown and emergency repair.
The work of high-temperature pipelines is, in the final analysis, a detailed project. Materials, structure, selection, installation and maintenance, every link is in place, and ten years will be a problem; Which link is fooled, it is not unusual to scrap it in a year.