ERROR: Error on Jacobi solver: matrices are not positive definite with Declarations4.Solvers.EigenValuesSolver.Jacobi()

When starting the elastic analysis of a model with a dynamic load or when starting a model analysis , the error below appears:

Cause 1

The error can be caused by a lack of stiffness/too many degrees of freedom. The lack of stiffness can be in the cross-sections, the material(s), the supports or the internal boundary conditions (hinges).

• Verify the stiffness properties of the elements: check if the sections and material properties are well defined. If an element has no stiffness, it shouldn’t be modelled in Diamonds.
• If you want to run an modal analysis on a 2D structure, make sure the translation and rotation in the third direction are prevented.
• Regarding hinges: if N bars meet in a node (neglect tie rods), and you want no moment to be transferred in that node, you should apply a hinge to N-1 bars. Not to N bars. If you apply a hinge to N bar, you’re making a mechanisme (throught the eyes of Diamonds).
• The amount of eigen modes/ eigen frequencies that can be found equals the nomber od degrees of freedom in the structure. If you make a system with 8 degrees of freedom, you can calculate 8 eigen frequencies. If you’d ask 10, Diamonds will give an error.
  So what you could do, is ask less eigen frequencies. The minimum is 2.

Cause 2

This error can also be caused by a malfunction in the load groups.

• Remove all load groups besides Self-weight and Seismic.
• Regenerate the combinations and calculate the structure using only those two load groups.
• If this calculates, define the load groups again one by one. Regenerate the combinations and calculate the structure between each addition.

Cause 3

• If a model contains 2 (or more) independant structures, make a Diamonds-file for each structure.
  2 structures are independent if there is no connection between them. For example: the column and the porch below are not conntect to each other, thus independant. It is better to make one Diamonds-file for each structure if you want to do a modal analysis for them.
  

Cause 4

• In the analysis settings of the modal analysis, try with an concentrated (lumped) massmatrix.
  

Cause 5

• Diamonds doesn’t accept equal to zero for the self-weight. If you don’t want the self-weight to contribute to the mass for the modal analysis, set to 1 and the density of materials to 0.001kg/m³. That’s the lowest value for the density that Diamonds will accept.