I don't work with Ls dyna, but if you have eigenvector data (numerical), two way existence for normalized in excel: 1- Normalized to the mass matrix and 2- Normalized to length. For mass matrix, in first step calculate mass modal matrix and for any mode (1/sqrt(Mi))*(eigenvector) and for length it's better for any mode, all data divided to degree freedom of the number of mode (mode 1 - DoF 1 , mode 2 - DoF 2 , ...).
If it's able for you, export these data from Ls dyna and after this procedure import this information to Ls dyna.