TPi January 2019

Applying trusted engineering theories and benchmarks: gain confidence in your pipe stress analysis By Phil Senior BEng, CEng, FIMechE, principal application engineer, Bentley Systems

also require the assumptions made during the modelling, which is impractical. Even converting one model from one program to another cannot guarantee the same results on both sides. Perfect translation is not possible due to differences in how both programs approach modelling and analysis. Although the basic theory and assumptions behind the two applications may be the same, as the products evolve the different components may start to diverge slightly based on selected options. For example, if you run very simple cantilever models in both the applications with a tip force applied, you will get the same results. These results could tell us that both the applications are using the same beam theory with shear deformation included. However, when we depart from simplistic models, some assumptions/implementations might come into play, including: 1. The beam theory employed by the application 2. Rigid beam and boundary support elements’ stiffness values 3. Restraint and support behaviour 4. Special elements formulation like bends, valves, soil elements and flexible joints 5. Internal pressure affects implementation, such as pressure stiffening and thrust load 6. Modal analysis theory and mass discretisation 7. Applied loads and load vector formulation 8. Post-processing assumptions and options 9. Non-linear solution engine differences, load sequencing On an individual level, there might only be slight differences in these formulations. However, when the entire system is analysed, the differences may get compounded due to interactions of these individual differences. Users can trust any application if they understand the assumptions involved in the solution and understand that the results are a fair approximation of the actual system.

When using an analytical program to perform engineering analysis, you need to trust the results. For reliable results, it is important to provide the correct set of inputs, select the appropriate program options, and understand the assumptions made by the program. If you cannot understand the program correctly and simply use it as a ‘black box’, then the validity of the analysis would be in question and the results could be disastrous. Validating the results after each analysis is a time- consuming business. Many software users rely on the vendors to have robust verification processes with hand calculations to back up the validation of the program. Experienced pipe stress analysts can often predict the correct result and will have a feel for when something does not ‘add up’ from a particular analysis. However, even they need to have confidence that the software analysis application is working correctly. Analysis program differences How can the analysts know if their results are satisfactory? The simple answer has already been given: rely on the vendor’s verification, that they follow the correct procedures and thoroughly test the program. You should also independently validate the results, either through hand calculation or comparison with other software. With these methods, you will know that the program is producing the correct results, increasing your confidence in it. When comparing results with similar analytical programs, we are often asked why there are differences in the results. The answer to this question, and any other comparison questions, would be generic. To provide a specific answer to even slightly complex piping system differences would require the full details of the program’s methods and calculations. It would

52

TUBE PRODUCTS INTERNATIONAL January 2019

www.read-tpi.com