where RF is the vector of reactions caused by the given load, which is
equal to the vector of nodal loads taken with the opposite sign.
The total load on the node is defined as the sum of the loads from the elements adjacent to the node. Since the non-nodal load is replaced by
the equivalent nodal load in the directions Z , then the vector of rections
RF R1F , R2F ,..., RnF T F ,
where
F F1 , F2 ,..., Fn T is the vector of nodal loads.
After solving the system of equations (15.20), the displacements Z of nodes in the general coordinate system become known. To calculate the forces in a finite element, it is convenient to first find the nodes displace-
ment vector Z in the local coordinate system, and then determine the reactions at the ends of the FE using the stiffness matrix R . For-mulas for the corresponding transformations are given in sections 15.9 and 15.10.
This form of calculation corresponds to the FEM variant “in displacements”. It is the most common form.
Another approach to solving the problem with the help of FEM is also possible. The stress-strain state of the FE must be described by a finite set of generalized nodal forces, and then establish their relationship with the load. This form of calculation corresponds to the FEM “in efforts”.
15.9. Bar Stiffness Matrix in the Local Coordinate System
There are several ways to obtain stiffness matrices of separate bars. One of the simplest is a method based on the known conditions of the displacement method.
Each end of the bar adjacent to the rigid node has three degrees of freedom: linear displacements in the horizontal and vertical directions and the angle of rotation. The force factors corresponding to these dis-
placements are the reactive forces R1 , R2 , R4 , R5 and the moments R3, R6 , located at the FE edges. (Displacements, reactions, stiffness
matrix and its elements in the local coordinate system are indicated by letters with strokes). The stiffness matrix (the matrix of unit reactions) converts the displacement vector