The first NEI models of a SNR X-ray emission can be found in Itoh
(1979), Cox & Anderson (1982) and Hamilton, Sarazin and Chevalier
(1983). They solve the NEI system of equations for each element in
several concentric shells covering almost the whole remnant, and then
integrate the emission over all shells to obtain the predicted NEI
spectrum to be compared with observations, once folded with the
instrumental response. Our data call for a more detailed approach for
two main reasons: first, the clumpy nature of the emission from Vela
SNR suggests expansion in a non uniform medium, which in turn requires
spectral analysis to be performed on a small angular scale. Second, the
high spatial resolution and high counting statistics of our observation
allow us to split the 
radius image in
many spatial bins, and to perform separate spectral analysis on each of
them. Because of the under-ionization of the plasma, the spatially
resolved spectra of the
shock region are expected to differ considerably from either
the spectra of the central region or the overall integrated spectrum of
the SNR (Brinkmann, 1988).