Is there an error in the Harwell Boeing dump (HBMAT) of a superelement in 11.0?
I understand that matrices from SUB files are dumped completely, regardless of whether the matrix is symmetric or unsymmetric or whether there are zero-values. I also understand that the offset of 1 in the column pointer array at the beginning of the file has been fixed in ANSYS 11.0. However, I think that there still is a bug: when I try to dump for example a 6x6 superelement matrix, the column pointer array should read as follows: 1, 7, 13, 19, 25, 31, 37 with the corresponding row indices 1, 2, 3, 4, 5, 6 for each column. The column pointer array ANSYS produces is instead 1, 7, 13, 19, 25, 31, 32. Therefore, any read-routine that has implemented the HB format correctly, will fail to read in the correct matrix.
Yes, you are correct. The error has been fixed in 12.0. |
||
|