# Tickets fixed

## Ticket 38: Bad normalisation of the tensor when computing eigenvalues

Tensors are normalised when computing the eigenvalues. There was a mistake in the normalisation which can lead to severe numerical instabilities in some (rare) case.

For more details, see: https://sourceforge.net/p/tfel/tickets/38/

## Ticket #35: regression when using @CastemFiniteStrainStrategy or @CastemFiniteStrainStrategies in the castem interface

When definining finite strain strategies in the castem interface, a regression was introduced in version $$3.0.0$$, if only one finite strain strategy is defined: in this specific case, one define two behaviours (in fact the second is barely an alias for the first one):

• A first one having the base name of the behaviour (e.g. umatnorton).
• A second name having the base name of the behaviour following by the finite strain strategy (e.g. umatnorton_ss).

The regression was that the ModellingHypotheses symbols was only defined for the second behaviour.

For more details, see: https://sourceforge.net/p/tfel/tickets/35/