This option is used to perform eigenvalue extraction to calculate the natural frequencies and corresponding mode shapes of a system.
Products: Abaqus/Standard Abaqus/CAE Abaqus/AMS
Type: History data
Level: Step
Abaqus/CAE: Step module
For the AMS eigensolver and Lanczos eigensolver, set ACOUSTIC COUPLING=ON to include the effect of acoustic-structural coupling during the natural frequency extraction procedure in models with acoustic and structural elements coupled using the *TIE option or in models with ASI-type elements. This is the default option for the Lanczos eigensolver.
For the AMS eigensolver and Lanczos eigensolver based on the SIM architecture, set ACOUSTIC COUPLING=PROJECTION to extract the uncoupled acoustic and structural modes and project the acoustic-structural coupling operator during the natural frequency extraction procedure in models with acoustic and structural elements coupled using the *TIE option. This is the default option for the AMS eigensolver.
Set ACOUSTIC COUPLING=OFF to omit the projection of the acoustic-structural coupling operator and to ignore the effect of acoustic-structural coupling during natural frequency extraction in models with acoustic and structural elements coupled using the *TIE option or in models with ASI-type elements.
This parameter is not relevant for the subspace iteration eigensolver.
This parameter is relevant only for the AMS eigensolver or for the Lanczos eigensolver used in conjunction with the SIM parameter.
Set DAMPING PROJECTION=ON (default) to project the viscous and structural damping operators during the natural frequency extraction procedure. If there is no damping defined in the model, the projection is not performed.
Set DAMPING PROJECTION=OFF to omit the projection of damping operators.
Set EIGENSOLVER=LANCZOS (default) to invoke the Lanczos eigensolver.
Set EIGENSOLVER=AMS to invoke the automatic multi-level substructuring (AMS) eigensolver.
Set EIGENSOLVER=SUBSPACE to invoke the subspace iteration eigensolver.
Set NORMALIZATION=DISPLACEMENT to normalize the eigenvectors so that the largest displacement, rotation, or acoustic pressure (in coupled acoustic-structural extractions) entry in each vector is unity. Displacement normalization is the default for both the subspace iteration eignensolver and for the Lanczos eigensolver when they are used without the SIM parameter.
Set NORMALIZATION=MASS to normalize the eigenvectors with respect to the structure's mass matrix (the eigenvectors are scaled so that the generalized mass for each vector is unity). Mass normalization is the default and only available option for the AMS eigensolver. Mass normalization is switched on for both the Lanczos eigensolver and the subspace iteration eigensolver when they are used in conjunction with the SIM parameter.
Set this parameter equal to the frequency at which to evaluate frequency-dependent properties for viscoelasticity, springs, and dashpots during the eigenvalue extraction. If this parameter is omitted, Abaqus/Standard will evaluate the stiffness associated with frequency-dependent springs and dashpots at zero frequency and will not consider the stiffness contributions from frequency domain viscoelasticity in the *FREQUENCY step.
This parameter is relevant only for the Lanczos and AMS eigensolvers.
Include this parameter to indicate that residual modes are to be computed.
This parameter is relevant only for the Lanczos and subspace iteration eigensolvers.
Set the value of this parameter equal to NO if the non-SIM architecture is required for the Lanczos or subspace iteration eigensolvers.
Set the value of this parameter equal to YES (default) if the SIM architecture is required.
The SIM architecture is used by default if the AMS eigensolver is activated.
Set this parameter equal to the name of the node set or include the parameter with no value to allow Abaqus/Standard to automatically select the nodes at which eigenvectors will be computed. If this parameter is omitted, eigenvectors will be computed at all nodes.
First (and only) line:
Number of eigenvalues to be calculated. This field can be left blank if the maximum frequency of interest is provided and the evaluation of all the eigenvalues in the given range is desired. The number of requested eigenmodes must be provided in a cyclic symmetry analysis or if the analysis includes more than one natural frequency extraction step.
Minimum frequency of interest, in cycles/time. If this field is left blank, no minimum is set.
Maximum frequency of interest, in cycles/time. If this field is left blank, no maximum is set. This value is required if the first field was left blank.
Shift point, in squared cycles per time (positive or negative). The eigenvalues closest to this point will be extracted.
Block size. If this entry is omitted, a default value, which is usually appropriate, is created.
Maximum number of block Lanczos steps within each Lanczos run. If this entry is omitted, a default value, which is usually appropriate, is created.
Acoustic range factor. This factor applies only to structural-acoustic problems and is used to set the maximum frequency for the acoustic stage of the uncoupled eigenproblem as a multiple of the nominal maximum frequency of interest. This factor is supported only when using the SIM architecture, and the maximum frequency of interest is provided. The acoustic range factor must be greater than 0. The default value is 1.0.
First line:
Number of eigenvalues to be calculated. If this field is left blank, Abaqus evaluates all the eigenvalues from the minimum frequency of interest up to the maximum frequency of interest.
Minimum frequency of interest, in cycles/time. If this field is left blank, no minimum is set.
Maximum frequency of interest, in cycles/time.
, the first AMS parameter. is a cutoff frequency for substructure eigenproblems, defined as a multiplier of the maximum frequency of interest. The default value is 5.
, the second AMS parameter. is the first cutoff frequency used to define a starting subspace in the reduced eigensolution phase, defined as a multiplier of the maximum frequency of interest. . The default value is 1.7.
, the third AMS parameter. is the second cutoff frequency used to define a starting subspace in the reduced eigensolution phase, defined as a multiplier of the maximum frequency of interest.. The default value is 1.1.
Acoustic range factor. This factor applies only to structural-acoustic problems and is used to set the maximum frequency for the acoustic stage of the uncoupled eigenproblem as a multiple of the nominal maximum frequency of interest. The acoustic range factor must be greater than 0. The default value is 1.0.
No additional data lines are needed if default residual modes are sufficient or residual modes are not requested. Otherwise, subsequent lines:
Node number or node set label.
First degree of freedom for which residual modes are requested.
Last degree of freedom for which residual modes are requested. This field can be left blank if residual modes for only one degree of freedom are requested.
Repeat this line as often as necessary to request residual modes.
First (and only) line:
Number of eigenvalues to be calculated.
Maximum frequency of interest, in cycles/time. This user-specified maximum frequency is increased automatically by 12.5% to help capture closely-spaced modes. Abaqus/Standard will also report all eigenvalues that converge in the same iteration as those in the specified range, even if their frequencies are more than 12.5% above the maximum frequency specified by the user. If this field is left blank, no maximum is set.
Abaqus/Standard will extract frequencies until either of the above limits is reached.
Shift point, in squared cycles per time (positive or negative). The eigenvalues closest to this point will be extracted.
Number of vectors used in the iteration. If this entry is omitted, a default value, which is usually appropriate, is created. The default number of vectors used is the minimum of (n+ 8, 2n), where n is the number of eigenvalues requested (the first data item on this data line). In general, the convergence is more rapid with more vectors, but the memory requirement is also larger. Thus, if the user knows that a particular type of eigenproblem converges slowly, providing more vectors by using this option might reduce the analysis cost.
Maximum number of iterations. The default is 30.