Inertia tensor of an assembly for dynamic analysis-Inconsistency from Solidworks

[andueza]

Hello for all.
When I calculate the mass property of an assembly, solidworks calculates inconsistent answers.
The assembly is composed by sub-assemblies and components.
The answer of the inertia tensor is inconsistent because of the following reasons:
1) The eigenvalues of the inertia tensor around the center of mass obtained from mathcad using the tensor supplied by solidworks are not equal of the principal moments of inertia given by solidworks;
2) The principal directions of the inertia tensor forms a rotation matrix (arranged by columns). The following mathmatical rule applies: RT*(inertia tensor)*R must result in a diagonal tensor, where RT is the transpose of the rotation matrix.The resulting diagonal corresponds to the principals moments of inertia. This rule fails when checked with the values supplied by solidworks;

I also performed the same checks with Pro-engineer and everything works according to the mathmatical rules in the same model.

Have anyone faced similar problems?

Many thanks in advance.
Alejandro.