When the expansion joint is thinner than the cylinder, is it necessary to trim the edge of the cylinder?

I had previously seen a similar container built by a colleague; the expansion joint is as shown in the figure below. The expansion joint itself consists of two layers of 5 mm thick steel plates, but the part that connects to the cylinder is 36 mm thick – the cylinder has not been thinned.

Conventional design should not thin the cylinder, unless specific calculations have been performed； Also, CS 16mm with SS 8 mm? Why does it seem to be so much worse? Could the CS corrosion margin be 8mm?

1. Since the corrosion on CS is so severe, it’s better to switch to stainless steel altogether. 2. I think trimming the edges is also fine; it’s not as if unequal thicknesses prevent welding. I think GB150 6.5.1 3 can be used as a reference. Simply use thick-walled expansion joints; however, finite element analysis is required, which is somewhat troublesome.

The cylinder body cannot be thinned; for example, if the calculated minimum wall thickness for a cylindrical body is 10 mm and the thickness of the expansion joint is 6 mm, thinning the cylinder body may result in insufficient strength.

When the thickness difference between the expansion joint and the cylinder is significant, and it is not permissible to thin the cylinder for strength calculations, then the expansion joint design shown in the diagram can be considered; this approach meets the requirements for compensation while also allowing for welding at the same thickness as the cylinder.

If the minimum wall thickness requirement (for heat exchangers) dictates a shell thickness of 16 mm, and the strength remains sufficient to meet the design criteria after thinning, then thinning is possible. If 16 mm is the wall thickness required for strength calculations, it cannot be thinned; instead, the thickness of the expansion joint must be increased to 14 mm (for carbon steel). The cost of this approach is likely not higher than that of using stainless steel! If the difference is too large, a structure for the second floor can also be a good option; the straight sections of the expansion joints should also be made longer appropriately, or the J6 structure as specified in HG/T20583 can be used.

If edge trimming is not allowed, how should clause 7.6.2.2 of the expansion joint standard GB/T 16749 be interpreted? As shown in the screenshot: Moreover, expansion joints also have straight edges, and if the strength of these straight edges is calculated using the formulas for cylinders, many of them will fail the strength tests

http://mp.wei*n.qq.com/s/If2LncB1GrSNp-_CkWlC7Q The original link is as above: Question: According to GB150, when connecting a cylinder with a spherical head, it is generally not necessary to thin out the cylinder section; instead, the head section is thinned so that the strength of both the cylinder section and the head section meets the required levels as determined by their respective calculation formulas. The straight sections of the expansion joint itself are also calculated, but the thinned portions of the cylinder are not. For example, if the cylinder is made of carbon steel with a corrosion allowance of 3 mm, the minimum design thickness required is 14 mm, while the nominal thickness is 16 mm. The expansion joint is made of stainless steel, 8mm. Will reducing the cylinder thickness to 8mm result in insufficient strength? If it isn’t thinned, how should it be connected? Can the thickness of the expansion joint be increased to reduce the trimming length? Answer: Regarding the issue of what to do when the wall thickness of the expansion joint differs from that of the vessel body, clause 7.6.2.2(b) of GB16749-1997 provides clear specifications: a single-sided or double-sided thinning transition should be carried out at an angle of 1:3 on one side of the vessel body. For this equipment, due to the significant difference in thickness between the two sides, it is recommended to thin both sides of the cylinder to facilitate a smoother transition and reduce the length of the thinned section of the cylinder. If the strength of the expansion joint meets the requirements, it is unreasonable to increase its thickness merely to address the connection issue between the expansion joint and the cylinder. This increases material consumption and reduces the compensating capacity of the expansion joint; sometimes, to meet the required compensatory amount, it may be necessary to increase the wave count as well. It is incorrect to think that the cylinder section adjacent to the expansion joint must be subjected to strength verification using the circumferential stress calculation formula for ordinary cylinders. The straight section of the expansion joint is also a very thin cylinder; how can it meet the strength requirements? This is because it is reinforced by the adjacent waveform segments, resulting in a reduction in hoop stress. The circumferential stress in the straight section of the expansion joint is calculated according to formula 6-1 in GB16749. For straight sections without reinforcing rings, this formula includes an additional coefficient k compared to the general formula for calculating circumferential stress in cylinders; this coefficient k is proportional to the length L4 of the straight section. As the length L4 of the straight segment increases, k also increases; only when k≥1 (note: when k>1, k is taken as 1) does the circumferential stress in the straight segment become equal to that of a regular cylinder. In fact, you can treat the thinned section of the cylinder as an extension of the straight section of the expansion joint; the sum of the length of this thinned section and the original straight section can be used as the length L4 of the straight section of the expansion joint. A new value for k can then be calculated. To be conservative, the thickness of the thinned section of the cylinder can be taken as the thickness of the expansion joint. The circumferential stress of the extended straight section can be calculated using formula 6-1 mentioned above, to determine whether this circumferential stress is less than the allowable stress for the cylinder. Typically, the value of k is small, and the calculation results are acceptable; this is why GB16749 allows the cylinder thickness to be reduced when it differs from that of the expansion joint. -------------------------------------------- Welcome to follow the original WeChat official account: VCAD001. We only post original articles

This post was last edited by Douwan on 2018-1-18 at 10:54. There are actually many questions regarding wave-shaped expansion joints; in particular, accidents such as leaks and cracks in these joints occur frequently during production and use, which has led to high expectations for the introduction of a new standard for such expansion joints. Regarding the current standard Chinese version 6-1, your explanation is reasonable based on the formulas. Have you noticed that although the calculations using formula 6-1 yield satisfactory results, the cylinder is not thinned out as shown in the figure below? Therefore, I believe it is not necessary to reduce the radius of the cylinder at this point. The requirement specified in 7.6.2.2 is only about meeting the tolerance for misalignment during manufacturing; it is not related to strength calculations. Therefore, I believe it is still necessary to meet the requirements regarding the wall thickness of the cylinder, and following the approach shown in the screenshot above (that is, the approach used on the second floor) is appropriate