Method of input

The portal frame in this example was initially input as a single bay portal frame using the structure wizard. This allowed quick and easy generation of the basic structural geometry, restraints, section properties (including the haunch section properties) and material properties. If the extra column and rafter nodes were not required, it would then have been a simple matter to add the loads (graphically or using datasheet input) and then perform the analysis.

 

Node, member and plate numbering

In this example we wanted to match the node, member and plate numbering with the numbering used in the AISC example. Therefore, it was necessary to modify the geometry slightly so that the extra nodes were added and the nodes and members were re-numbered. This was done graphically by simply subdividing the members and then renumbering the structure with the extra nodes included. The rafter and haunch section properties were assigned to members 3 - 10 by graphically changing the section property numbers of members 5, 6, 7 and 8 to section 2, members 3 and 10 to section 3, and members 4 and 9 to section 4.

 

Node restraints

When the structural geometry was established, node restraints of FFFRFR were applied to support nodes 1 and 13, and restraints of RRFRRR were applied to rafter nodes 3, 6, 7, 8 and 11. The restraints on nodes 1 and 13 specified that the structure was pin-based, allowing rotation about both the X and Z axes. The standard 2D frame pin restraint of FFFFFR was not used in this case because it would have prevented rotation about the X-axis.

 

The rafter node restraints were applied to simulate the effect of wall and roof bracing that would prevent any out-of-plane (Z-axis) movements at those nodes. A general restraint of RRFRRR was not used in this case because it would have prevented the out-of-plane movements of nodes 2, 4, 5, 9, 10 and 12 which, in real life, would be free to move in that direction. Although no out-of-plane movements would occur in a static analysis (due to no loads in that direction), they could occur in a buckling analysis and, if restrained, could result in incorrect buckling load factors and effective lengths.

 

If no intermediate nodes were present that could move in the out-of-plane directions then a general restraint could have been used.

 

 Under normal circumstances it would not have been necessary to match the node and member numbering with the AISC example. This would have removed the necessity to subdivide the members, or change the member properties and node restraints as described above.

 

Loads

The node and member loads were applied graphically. Although there are many member loads, the graphical input facility made it very easy to input them en-masse. For most load cases, it was simply a matter of placing a window around the members and then specifying the load applied to them.

 

Self weight, combination load cases and load case titles were input using datasheets.

 

Input check

As a final check before the analysis was initiated, loading diagrams for each load case were viewed followed by an output report of the complete structural data. Any errors in the data were corrected and the model was then ready for analysis.