Applies To
	Product(s):	STAAD.Pro
	Version(s):	All
	Environment:	N/A
	Area:	Postprocessing workflow
	Subarea:	N/A
	Original Author:	Bentley Technical Support Group

I have a model where supports are defined at the nodes of some of the plate elements in the structure. If I divide the support reaction values by the thickness, length etc., of the side of the elements adjacent to the support, shouldn't the values match the ELEMENT NODAL STRESSES? I am aware of the fact that element stresses are in the local axis system of the element, and support reactions are in the global axis system, and am making the required transformations before making the comparison.

The element nodal stresses are obtained as the value of the stress polynomial at the coordinates of those joints. Stresses in an element are most accurately determined only at the center of the element (in the middle of the joint displacement locations used in calculating that stress). The stress values calculated at the nodes will only be approximate (only the displacements of the joints from this one element are used in calculating the stress). Stresses at a joint would be improved if the stresses from the other elements at the joint (on the same surface) were averaged. Consequently, the comparison you suggest is not feasible.

A better alternative would be to compare the forces at the node rather than the stresses at the node.

The output for the command

PRINT ELEMENT FORCES

consists of the 3 forces and 3 moments at each of the nodes of the elements, reported in the global axis system. Thus, the output will consist of FX,FY,FZ,MX,MY,MZ with the 3 forces having units of force (not stress) and the 3 moments have units of moment (not moment per unit width). If you add up the values at the nodes of those elements which are connected to the support, those values must be equal to the support reaction.

Another consideration is the way in which element loads are evaluated and used. Staad computes the equivalent forces at the corner joints (same total force, center of force, and direction). The remainder of the analysis and results are as if you had applied the loads as joint loads rather than as element loads. Two exceptions, temperature loads are applied internally to the element and plate releases will affect the load distribution to the joints.

Say you have a wall with uniform pressure. Half of the load on the elements along the base will be applied directly to the base, the other half is applied to the line of joints at the top of these elements. So the internal transverse shears are too high at the top of the element. The transverse shears are OK at the center and too small at the base. The same will be true for the element force output of transverse forces. However, the reactions will have the entire force. A finer mesh in general, and near the base in particular, will improve the element stress and load distribution.