XANSYS Message: 34769 [Go back to message list] [bookmark on del.icio.us]
No rating yet Subject: Re: Regarding substructuring Author: Christopher Wright Date: 1970-01-01 00:59:00> where does the border lie between a "do a >superelement and then contact" and "do the contact only" strategies? The expected number of iterations has always been the biggest part of the trade-off. If you're doing a time history problem, substructure the linear parts, otherwise re-forming the stiffness matrix for each time step will eat you alive. The only real down side of substructuring is having to do stress passes and keep track of what you know. Without a good understanding of the problem, people end up doing stress passes on hundreds of load steps which can be overwhelming.
You can also use substructuring to provide realistic boundary conditions for modal analysis of a piece of a structure. For example if you were to design an additional clump for an existing widget having about the same mass and stiffness. You'd substructure the widget with masters at the connection points, then model the clump and couple the clump supports to the substructured widget. The combined model would give you proper support stiffness, and if the widget has any vibration modes that affect the response of the clump, their effect is included. And whenever the design of the clump gets changed, you don't need to re-model the widget.
Christopher Wright P.E. |"They couldn't hit an elephant from chrisw@s... | this distance" (last words of Gen. ___________________________| John Sedgwick, Spotsylvania 1864) http://www.skypoint.com/~chrisw