Hello everybody,
I am currently working with Hyperworks and Nastran as my solver and finding my way into superelements.
The objective of my work is the dynamic reduction of a gearbox into a superelement. The creation and integration works now. However, I do have some discrepancies when performing a modal analysis and comparing them against the reference model (by performing a MAC, FRF, modal stiffness and eigenfrequencies etc).
The gearbox is supported by 7 struts (4 in z, 2 in x and 1 in y). Each strut (except of the y strut) is connected to the gearbox via: RBE2(on strut)-CBUSH-RBE2(on gearbox). I chose the CBUSH to be part of the superelement and the RBE2 on the strut as part of the residual.
When defining the ASET, I looked into the properties of the CBUSH and defined the DOFs (for example: Stiffness of CBUSH: 1000000,100000,100000,0,0,0)—-(ASET 123456(id) 123(dof)).
When doing that, I get a lot of weird modes.
By choosing all dofs of the CBUSH as 1000000 and 123456 in the ASET, I do get better results, but still some discrepencies.
The y strut is sharing two independent nodes with the gearbox and gearbox and strut is connected via an RBE2 to this node.
Does anybody have an idea why I do get discrepencies? I am aware that the dynamic reduction is an approximation, however I do expect better results.
Unfortunately, I am not able to upload any pictures.
Thanks everybody in advance
