Thanks to its open architecture, LARSA 4D has been evolving into something more than just a structural analysis and design package – perhaps, a structural engineering *platform*. Last week, I wanted to give a shot to see how easy it is to add a new finite element into LARSA 4D platform. Granted, most users will not be interested in adding a finite element formulation to LARSA; yet I see it as a good demonstration of how extensible LARSA 4D is.
There is no reason to reinvent the wheel, LARSA 4D already includes a software development system in the form of a plug-in for user-defined finite elements. If you are a serious finite element developer, then you should take a look at LARSA Element Development Kit, otherwise read on. What I want to do is to extend the Element Development Kit to make adding elements simpler. Given *only* a set of stiffness matrix terms of an element, can I write a plug-in to make it a fully functional finite element in LARSA 4D? The answer as expected is Yes, what is surprising is how easy it is to do so!
Getting custom elements into LARSA 4D involves spreadsheets to graphical view of the elements to analysis to result spreadsheets. That means when you create a custom element and provide the information for its stiffness matrix terms, the plug-in will need to create a spreadsheet in LARSA 4D for input. The elements will also be visible in LARSA 4D graphics window (select/unselect/hide/extrude etc.) You will be able to run analysis; the plug-in will handle global/local transformations, force recovery and all the other stuff that is required to make it look like a native LARSA 4D finite element formulation.
LARSA 4D Preview version 7.05.07 and up includes this plug-in (Tools > Custom Finite Element Definitions). The plug-in includes a neat way of defining input required by the element, and the ability to enter stiffness matrix terms. And, that is all you need to define a custom element. I already tried a simple one node spring element and a 3D beam column element.
There are some reasonable limitations, of course. The custom elements can only be linear (although can be used in nonlinear analysis or even as part of staged construction analysis). Only results that can be extracted out of a custom element are the local forces at joints. The elements can only access material and cross-sectional built-in properties of LARSA 4D, other data required by the element must be defined as custom fields. Elements can be 1, 2…7 or 8-node. One node elements are considered as grounded springs and follow the global axis coordinate system. 2-node elements are considered as similar to member elements and follow the same local axis rules. 3 to 7 node elements are considered as surface elements and follow the same coordinate system rules as LARSA 4D plate/shell elements. 8 node elements are considered as brick element and follow the same local axis rules as LARSA 4D solid element.
Here are some screenshots (click to enlarge):
Recent Comments