Deep beams are wall-like structural elements that behave mostly so, that under load, Euler-Bernoulli’s hypothesis of section planar behavior is not applicable, in the largest parts of their volume. When deep beams additionally contain openings, things get even more complicated. Because of the large number of possible combinations of unknowns, and their interdependence, it is sometimes very difficult to suppose which calculation model will give us a sufficiently “code-friendly” safe, and serviceable structure. The most used type of calculation are methods: #FEM, #StrutandTie, #StressFields, and #StringerPanel. All of them are implemented in numerous software. Off course, every one of them has its advantages and disadvantages, with varying levels of approximation.
What can help us with so many possibilities? Parametric design. Why? Because one of the best things about #parametricdesign is that we can very quickly change the variable from which design dependent is and see the repercussions of changes that we have made.
In the learning process, a good way is to use some benchmark examples from literature and examine how the changes in certain parameters affect on the behavior of the observed structural element
For this exercise, I have used @RHINO 3D GRASSHOPPER 3D again for geometry parameterization. For the first time, I have tried to visualize input in #HumanUI, so that exercise also can be comfortably practiced by non-grasshopper users. For parameterization process of structural design in FEM Method is used an example from standard Teddy files from SOFiSTiK AG. It was also made models based on Stress Fields, in IDEA StatiCa. In the end, I have tried to investigate a little bit about Stringer-Panel Method. For this purpose, I have used #SPanCad developed on TUDelft iGEM. For a Strut-and-Tie Method has used results from available literature (spec.J.Schlaich, A.Muttoni, Miguel Fernández Ruiz, Filip Niketić...).
Dispersion of results is 30-40 %. That show, or how complicated is to design this type of structure and how big an influence on results have every parameter and used type of design…or my current level of knowledge...:) Learning goes on…:)
23 сен 2024
