We analyzed the ASCA data of five supernova remnants (SNRs): 3C 391 (G31.9+0.0), 3C 400.2 (G53.6- 2.2), G69.4+1.2, HB 21 (G89.0+4.7) and Kes 27 (G327.4+0.4). Among them, we identified G69.4+1.2 as a new SNR which has counterparts at radio and optical wavelengths. Although 3C 400.2 has an unusual morphology at radio wavelengths suggesting two SNRs superposed along the line of sight, our spectral analysis of 3C 400.2 supports that it is a single SNR. The morphology of 3C 400.2 can be explained with a supernova explosion occurred near the edge of an interstellar cloud, similar to the case of 3C 391. ASCA images of the SNRs confirm that all five remnants show a centrally peaked X-ray morphology, in spite of their shell-like morphologies at radio wavelengths. We find that the remnants have thin thermal X-ray emission with the elemental abundances close to or less than cosmic values. SNRs showing these properties are often referred to as "mixed-morphology SNRs", which are distinguished from the well-known SNR groups: shell-type, composite and Crab-like SNRs. The properties of mixed-morphology SNRs cannot be explained with a standard model such as the Sedov solution. Two models are generally considered for the X-ray emission from mixed-morphology SNRs: the model of evaporating cloudlets inside the remnant, and the model of the X-ray emission from a radiative remnant with a cooled shell.
We compared the physical parameters derived from the spectral analysis between mixed-morphology and shell-type SNRs or Crab-like SNRs. We find that the thermal pressure PX derived from the X-ray observation is correlated with the radius of the radio shell, R. It can be well described by PX ∝ R-3 not only for shell-type SNRs but also for mixed-morphology SNRs. Since the value of PX R-3 corresponds to the thermal energy contained inside each remnant, our result suggests that the total amount of the thermal energy is similar to each other, and it has been almost kept constant through the SNR evolution. Furthermore, the relation of PX ∝ R-3 is predicted by the Sedov solution in which PX R-3 is proportional to the initial explosion energy. If we can neglect the radiative loss or a fraction of the kinetic energy of the plasma, the fact that PX R-3 is constant for all SNRs suggests the similar values of the initial explosion energy among SNRs.