Why is the modeling of non-metallic compensators different from that of metals? -stiffness and coefficient of friction are key
Anyone who has done pipeline stress analysis knows that there is a ready-made component library for metal bellows modeling in CAESAR II. Just enter the nominal diameter, wave number and stiffness. But when they come across non-metallic compensators, many people are stunned-why don't they even have a decent template?
The mechanical behavior of non-metallic compensators (which we often call fabric fiber expansion joints) is not the same as that of metals. Metal bellows rely on the elastic deformation of corrugations, and the stiffness is linear and predictable. What about non-metals? It is made of multi-layer flexible composite materials (glass fiber, PTFE, silicone rubber, etc.) pressed, and its deformation after stress is not pure elasticity, but also has obvious viscoelastic characteristics. More critically, non-metallic expansion joints generate greater frictional resistance when displaced axially-because there is sliding friction between fabric layers and between fabric and flange.
Therefore, when you model in CAESAR II, if you directly use the spring model of metal bellows to set non-metal, the calculated pipe load and the force on the equipment port will all deviate. Remember one sentence: the bending stiffness and axial stiffness of non-metallic compensators are two independent parameters, and the friction coefficient (friction damping) must be explicitly input, otherwise the calculation results will be useless at all.
You have to understand these data before modeling: JB/T 12235-2015 and product selection table
Don't rush to open CAESAR II with a click of the mouse. Before modeling, you should have three things on hand: one is the product standard, the other is the manufacturer's selection table, and the third is the geometric dimensions of the actual installation.
Product standards areJB/T 12235-2015 Non-Metallic Expansion JointsThis standard specifies the classification, technical requirements and test methods of non-metallic expansion joints. However, you should note that the standard gives the minimum performance requirements, and the actual product stiffness data depends on the selection table provided by the manufacturer. For example, for the non-metallic expansion joint (fabric fiber expansion joint) sold in our station, the thickness, allowable displacement, axial stiffness value (unit N/mm) and transverse stiffness of each layer will be clearly written in the selection table.
Also, don't forget to check the effective area. The effective area of the non-metallic compensator directly determines the pressure thrust, a force that is deadly in piping systems. How to calculate the effective area? Usually, according to the projected area of the middle surface of the corrugation, that is, the circular area (inner diameter + wave height), the specific value can be obtained directly in the selection table. No data? Call the manufacturer or estimate according to the formula in Appendix A of JB/T 12235-2015.
How to build a model in CAESAR II? -Component type, effective area and pre-compression must not be less
Okay, the data is ready, let's go. CAESAR II builds a non-metallic compensator. To put it bluntly, it uses aEXPANDING JOINTComponents to simulate.
-Insert a point in the pipeline model and select the component type "EXPANDING JOINT".
-Enter axial stiffness, transverse stiffness (bending stiffness), these two values are taken directly from the selection table.
-Enter the Effective Area, usually in in² (CAESAR II defaults to Imperial, remember to convert).
-Friction coefficient: For non-metallic compensators, it is recommended to take 0.1~0.3, depending on the number of fabric layers and surface treatment. The default is set to 0.2 basic security.
-Pre-compression: This is something many people ignore. When the non-metallic compensator is installed, there is usually a certain amount of PRECOMPRESSION (for example, absorbing 10mm axial thermal displacement, and compressing 5mm first during installation), which you need to reflect with COLD SPRING or PRECOMPRESSION parameters in the model. If not, the actual displacement window will be eaten by half, and the tube will directly exceed the limit as soon as it is heated.
The default assumption of the EXPANDING JOINT component in CAESAR II is axially free and laterally constrained? No, you need to set the constraint direction according to the actual structure. If the non-metallic compensator has a tie rod (such as a large tie rod expansion joint), rigid rod elements have to be added at both ends of the tie rod to simulate the limit.
Some of the most prone pits to rollover: deflector action, tie rod adjustment, and installation direction
Two days ago, I met an engineer who was doing a desulfurization project. He asked me, "Why is the pipeline thrust I calculated 30% greater than the actual one?" When I looked at the model, the problem was in the guide tube.
The first pit: the influence of the deflector.The non-metallic expansion joint usually has a guide tube (liner) inside, which acts to reduce the erosion of the corrugated fabric by the media flow and also prevent the accumulation of solid particles. But the presence of the guide tube changes the effective area and lateral stiffness of the compensator-because the guide tube itself is a rigid piece, it limits the radial deformation of the corrugation. When modeling, if the manufacturer's selection table includes a guide tube, the stiffness value you take must be the value corresponding to the guide tube. Don't use the elementary data to make up.
The second pit: how to adjust the tie rod nut.The role of the tie rod is to limit excessive elongation or compression of the compensator. There are two ways to simulate a tie rod in CAESAR II: one is to set a large stiffness with a spring unit (simulating a rigid limit), and the other is to add clearance with a rod unit. I recommend the latter, because the tie rod nut usually reserves a gap (for example, the nut is not screwed during installation, leaving a 5mm margin of activity). If this gap is not set in, the model becomes a rigid connection, and the pipeline stress will be calculated randomly. How to tune it? Build two nodes next to the EXPANDING JOINT, connect them with a ROD unit, fix one end, and set the GAP parameter at the other end.
Third Pit: Installation direction.Non-metallic compensators have a well-defined arrow direction (usually pointing to the direction of media flow) because the fabric stacking direction determines compression resistance. If it is installed backwards, it may flatten the ripples under high pressure. CAESAR II doesn't read arrows directly, but you need to make sure that the displacement direction in the model is the same as the actual installation. For example, in the smoke baffle door scene, the compensator is usually installed vertically and bears axial displacement, so you have to release the axial degree of freedom in the model and lock it laterally.
Take the scene of desulfurization flue gas baffle door as an example. What should we pay attention to in modeling?
A non-metallic compensator is usually installed before and after the desulfurization flue gas baffle door (that is, the desulfurization flue gas baffle door in our station), which is used to absorb the thermal expansion of the pipeline and the displacement caused by the opening and closing of the baffle door. This scenario has several particularities:
- High medium temperature (130-180 DEG C), strong acidic corrosion,Therefore, the inner layer of the non-metallic expansion joint must be lined with PTFE or F46. When selecting the model, the stiffness data should correspond to the high-temperature resistant model, and do not use the normal temperature data.
- The baffle door itself has an operating torque,If the axial stiffness of the compensator is too large, it will directly affect the opening and closing force of the baffle door. When modeling, accurate stiffness values must be used, otherwise the final calculated motor selection may be too small.
- The deflector must take into account,Desulfurization flue gas contains a large amount of gypsum slurry particles, and the guide tube is used for life-saving. The stiffness parameters with guide tube are selected in modeling, and the length of guide tube is simulated by rigid section in the model.
- Leave room for tie rod adjustment,The desulfurization site often needs to adjust the compensator preload on site, and the COLD SPRING value in the model should be consistent with the installation scheme. If the tie rod nut has an adjustment margin (such as ±10mm), don't forget to reflect it in the GAP parameter.
CAESAR II, a general-purpose stress analysis software, does not handle non-metallic compensators perfectly. Many fine behaviors (such as viscoelastic creep) cannot be calculated. However, as long as you get the four parameters of stiffness, effective area, friction coefficient and pre-compression accurately, the safety margin in engineering is enough. Don't be too troubled. Ask the manufacturer for a few measured stiffness curves, which is a hundred times better than guessing in the software.