Hello! I need to extract the mass and stiffness matrices for a model with the following problem size:
P R O B L E M S I Z E
NUMBER OF ELEMENTS IS 249191
153326 linear line elements of type T3D2
84141 linear hexahedral elements of type C3D8R
102 linear line elements of type B31
11613 linear quadrilateral elements of type S4R
NUMBER OF NODES IS 267444
NUMBER OF NODES DEFINED BY THE USER 267240
NUMBER OF INTERNAL NODES GENERATED BY THE PROGRAM 204
TOTAL NUMBER OF VARIABLES IN THE MODEL 837207 (DEGREES OF FREEDOM PLUS MAX NO. OF ANY LAGRANGE MULTIPLIER VARIABLES. INCLUDE *PRINT,SOLVE=YES TO GET THE ACTUAL NUMBER.)
The properties are input as mass density, and I believe they will be used to generate a consistent mass matrix.
Here's the input file code I used: ** Global Mass and Stiffness matrix *Step, name=Export matrix *MATRIX GENERATE, STIFFNESS, MASS, VISCOUS DAMPING, STRUCTURAL DAMPING *MATRIX OUTPUT, STIFFNESS, MASS, VISCOUS DAMPING, STRUCTURAL DAMPING, FORMAT=coordinate
I have the following questions regarding my problem:
Dimensions of M and K matrices As indicated above, the number of degrees of freedom is 837,207, but the matrix dimensions are reduced to 354,231*354,231. Shouldn't the number of degrees of freedom match the matrix dimensions?Node numbering The model consists of 8 parts, and the nodes start from 1 for each part. However, when I extract the matrices using the FORMAT=matrix input option, a different node numbering system (1 to 241,751) is applied, making it difficult to match the entries to the actual model locations. How can I find the correspondence between the entries in the M and K matrices and the nodes in the model?In the coordinate format, I get 5,620,189 rows of data, while in the matrix input format, I get 2,987,210 rows of data. Shouldn't the number of data entries be the same in both cases?When using the matrix input format, the entries are extracted in the following format: 241751,3, 241751,3, 9.038200770026704e+00 Can I interpret the corresponding data as follows? 1: X (translational) 2: Y (translational) 3: Z (translational) 4: RX (rotational) 5: RY (rotational) 6: RZ (rotational)The modes obtained from modal analysis in ABAQUS CAE GUI and the eigenanalysis results obtained from extracting the M and K matrices and performing the Lanczos method in MATLAB do not match. Is there any way to reconcile them?
