I am trying to model a maintenance bulkhead structure consisting of several steel beam members, upstream and downstream girders, connected by diagonal bracing members (the structure resembles a truss laid horizontal on its side). The model also utilizes plate members connected to the upstream girder beams to model the skinplate of the bulkhead.
My question pertains to the use of beam offsets. Is it correct to assume that the main function of offsetting the beams is to achieve the appropriate member lengths (since STAAD connects members at their centerlines, which is not where the members are physically connected), and offsetting does not directly effect the way load is transferred between members, since STAAD still sees them as connected at the nodes.
Specifically, in my model, the skin plate members, when viewed in 3D rendering, appear attached to the centerline of the webs between the upper and lower girders. In reality, the skinplate will be welded to the outside of the flanges of these girders. Since in my model the loading is hydrostatic load applied to the plates, does the location of the plates matter? I.e., will offsetting the girders so that the flanges line up with the skinplates effect the way the structure is loaded?
Also, the bracing members between the upstream and downstream girders are modeled as being connected to the centerlines of the webs of the girders. In reality, these members will be welded to the outside of the flanges of these girders. Am I correct to assume, since these are compression members, that it is conservative to not use offsets?
Thanks in advance for any help,
In general, I say don't use offsets unless you have to. The offset creates a rigid link introducing eccentricity in the system, potentially causing instability and moments where you might not expect them.
Using rigid offsets to shorten a beam to the face of a support is a common application of offsets. It reduces the beam length in the stiffness matrix giving you a stiffer beam which may be realistic. Be careful using large offsets when the shortened member is loaded with a line load. In some applications the line load will also get shortened, leaving you will less total load (other applications carry the forces back to the rigid link back to the node).
Using offsets to align beams and girders at the top flange is impractical in my opinion, especially where the intent of the beams is to be simply supported.
As I understand it, loads that are applied at the nodes are still applied at the nodes, i.e. the original, user defined coordinates. Meanwhile, loads that are applied to the members are still applied to the member but in the offset position.
I would love to hear comments from other users of the community regardless what application you are using, because there is some brainstorming about offsets currently happening in the RAM Structural System camp.
In general the designer may not like to apply member offset to the structure but I have observed that self weight and member load do get considerably increased if member offset is not applied to simulate the real geometry of the structure. As a result the support reaction becomes higher to the extent of nearly 10% of the actual load.or even more in some cases.