Skip to content

Commit

Permalink
Update BD_PARAM.md
Browse files Browse the repository at this point in the history
  • Loading branch information
@SteveDoyle2
SteveDoyle2 authored Apr 14, 2024
1 parent e5ac722 commit cf6f9fd
Showing 1 changed file with 43 additions and 51 deletions.
94 changes: 43 additions & 51 deletions dev_docs/BD_PARAM.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -21,13 +21,16 @@ values. Unless otherwise stated, only value V1 is used. The parameter name alway
field 2 and V1 always goes in field 3. When there is more than one Vi, the table explicitly states
in what fields the Vi go.

manual errors:
- CUSERIN spacing

| Parameter Name | Data Type | Function of Parameter NOTE: Default values of parameters are: N for Char, 0 for Int and 0.0 for real |
| --------------- | --------- | ---------------------------------------------------------------------------------------------------- |
| ARP_TOL | Real | Default = 1x10-6; Tolerance to use in Lanczos eigenvalue extraction method for convergence |
| ARP_TOL | Real | Default = 1x10-6 |
| | | Tolerance to use in Lanczos eigenvalue extraction method for convergence |
| ART_KED (for diff stiffness – not fully implemented) | Char | Field 3: ART_KED, default = N. If Y add artificial stiff to diag of KED stiff matrix |
| | | Field 4: ART_TRAN_MASS: value for translation degrees of freedom, default 1x10-6 |
| | | Field 5: ART_ROT_MASS: value for translation degrees of freedom, default 1x10-6 |

| ART_MASS | Char | Field 3: ART_MASS, default = N. If Y add artificial mass to diag of MGG mass matrix |
| | | Field 4: ART_TRAN_MASS: value for translation degrees of freedom, default 1x10-6 |
| | | Field 5: ART_ROT_MASS: value for rotation degrees of freedom, default 1x10-6 |
Expand All @@ -36,52 +39,41 @@ in what fields the Vi go.
| | Int | Field 5: AUTOSPC_NSET, default = 1 (see Section 3.4.1.1) |
| | Char | Field 6: AUTOSPC_INFO, default = N. If Y then print messages about the AUTOSPC’s |
| | Char | Field 7: AUTOSPC_SPCF, default = N. If Y print AUTOSPC forces of constraint |

BAILOUT Int Default = 1
If > 0 quit if a singularity in decomposing a matrix is detected.
If <= 0 do not quit
CBMIN3 Real Default = 2.0
CBMIN3 is the constant CB used in tuning the shear correction factor in Ref 3 for
the TRIA3 plate element. The default 2.0 is the value suggested by the author.
CBMIN4 Real Default = 3.6
CBMIN4 is the constant CB used in tuning the shear correction factor in Ref 4 for
the QUAD4 plate element (QUAD4TYP = ‘MIN4 ‘). See Ref 4
CBMIN4T Real Default = 3.6
CBMIN4T is the constant CB used in tuning the shear correction factor in Ref 4 for
the QUAD4 plate element (QUAD4TYP = ‘MIN4T’).
CHKGRDS Char Default = Y. If N do not check that all grids for all elements exist
CRS_CCS Char Default = CRS (compressed row storage of matrices). Also can be CCS
CUSERIN Char
Int
Int
If this parameter is present, Bulk Data entries for Craig-Bampton (CB) reduced
models will be written to the F06 file as a CUSERIN element (including grids, coord
sys, etc)
Field 3: element ID, default = 9999999

Int
Int
Char
Int
Field 4: property ID default = 9999999
Field 5: start index for SPOINT’s to represent modes of the CB model, default =
1001
Field 6: IN4 file # on the PUSERIN entry for this CUSERIN elem, default = 9999999
Field 7: Set-ID for CUSERIN elem (typically the “R”, or boundary, set), def is blank
field
Field 8: Format for how to write the comp numbers (1 thru 6) for each grid of the
CUSERIN elem. If 0, write them in compact form (e.g. 1356). If > 0 write them in
expanded form (1 3 56), default = 0
DARPACK Int Default = 2
how many extra modes to find above EIG_N2 on the EIGRL entry. These few
highest mode are not used due to difficulty with getting good GP force balance.
DELBAN Int Default 1. If equal to 1 delete the bandit output files on exit
EIGESTL Int Default 5000
For eigenvalue problems by the Lanczos method, if the number of L-set DOF’s
exceed EIGESTL the method for specifying the search range will be changed from
F1 to F2 to N (see EIGRL Bulk Data entry) to avoid excessive run times (since the
code to estimate the number of eigens in the F1 to F2 range can be excessive).
EIGNORM2 Char Default = N. if 'Y' then eigenvectors will be renormalized a last time by multiplying
by a set of scale factors (1 per eigenvector) supplied in a file with the same name as
the input file and extension 'EIN' (if it exists)

| BAILOUT | Int | Default = 1 |
| | | If > 0 quit if a singularity in decomposing a matrix is detected. |
| | | If <= 0 do not quit |
| CBMIN3 | Real | Default = 2.0 |
| | | CBMIN3 is the constant CB used in tuning the shear correction factor in Ref 3 for the TRIA3 plate element. The default 2.0 is the value suggested by the author. |
| CBMIN4 | Real | Default = 3.6 |
| | | CBMIN4 is the constant CB used in tuning the shear correction factor in Ref 4 for the QUAD4 plate element (QUAD4TYP = ‘MIN4 ‘). See Ref 4 |
| CBMIN4T | Real | Default = 3.6 |
| | | CBMIN4T is the constant CB used in tuning the shear correction factor in Ref 4 for the QUAD4 plate element (QUAD4TYP = ‘MIN4T’). |
| CHKGRDS | Char | Default = Y. If N do not check that all grids for all elements exist |
| CRS_CCS | Char | Default = CRS (compressed row storage of matrices). Also can be CCS |
| CUSERIN | Char | If this parameter is present, Bulk Data entries for Craig-Bampton (CB) reduced models will be written to the F06 file as a CUSERIN element (including grids, coord sys, etc) |
| | Int | Field 3: element ID, default = 9999999 |
| | Int | Field 4: property ID default = 9999999 |
| | Int | Field 5: start index for SPOINT’s to represent modes of the CB model, default = 1001 |
| | Int | Field 6: IN4 file # on the PUSERIN entry for this CUSERIN elem, default = 9999999 |
| | Char | Field 7: Set-ID for CUSERIN elem (typically the “R”, or boundary, set), def is blank field |
| | Int | Field 8: Format for how to write the comp numbers (1 thru 6) for each grid of the CUSERIN elem. If 0, write them in compact form (e.g. 1356). If > 0 write them in expanded form (1 3 56), default = 0 |
| DARPACK | Int | Default = 2 |
| | | how many extra modes to find above EIG_N2 on the EIGRL entry. These few highest mode are not used due to difficulty with getting good GP force balance. |
| DELBAN | Int | Default 1. If equal to 1 delete the bandit output files on exit |
| EIGESTL | Int | Default 5000 |
| | | For eigenvalue problems by the Lanczos method, if the number of L-set DOF’s exceed EIGESTL the method for specifying the search range will be changed from F1 to F2 to N (see EIGRL Bulk Data entry) to avoid excessive run times (since the code to estimate the number of eigens in the F1 to F2 range can be excessive). |
| EIGNORM2 | Char | Default = N. If 'Y' then eigenvectors will be renormalized a last time by multiplying by a set of scale factors (1 per eigenvector) supplied in a file with the same name as the input file and extension 'EIN' (if it exists) |
| ELFORCEN | Char | Default = GLOBAL |
| | | If ELFORCEN = GLOBAL, and nodal forces have been requested in Case Control, they will be output in the global coordinate system. |
| | | If ELFORCEN = BASIC, and nodal forces have been requested in Case Control, they will be output in the basic coordinate systeml. |
| | | If ELFORCEN = LOCAL, and nodal forces have been requested in Case Control, they will be output in the local element coordinate system. |
| EPSERR | Char | Default = Y. If N, do not calculate the NASTRAN like “epsilon error estimate” |
| EPSIL | Real | There are 3 EPSIL(i) values each of which requires a separate PAPAM EPSIL Bulk |
| | | Data entry with the index (i) in field 3 and EPSIL(i) value in field 4. These are small numbers used in MYSTRAN for the purposes indicated below:
| | | 1) EPSIL(1) (default = 1x10-15) is used in MYSTRAN such that, in any real number comparisons, any real number whose absolute magnitude is less than EPSIL(1) is considered to be zero. If no PARAM EPSIL 1 entry is in the data file then this value is reset (from the default) in LINK1 to a value based on machine precision calculated using LAPACK BLAS function DLAMCH. If the user has a PARAM EPSIL 1 entry, this value will be used for EPSIL(1) instead of the LAPACK machine precision. |
| | | 2) Currently not used |
| | | 3) EPSIL(3) is used in the Inverse Power method of eigenvalue extraction to test convergence of an eigenvalue. The default value (% change) is 1x10-5 % |
| | | 4) EPSIL(4) is used to calculate the maximum warp for quadrilateral plate elements, above which a warning message will be written. This maximum warp is EPSIL(2) times the average length of the quadrilateral’s two diagonals. The default for EPSIL(2) is 1.x10-1 |
| | | 5) EPSIL(5) (default 1.x10-6) is used in BAR and ROD margin of safety calculations. If a stress magnitude is less than EPSIL(5) a 1.x1010 margin of safety will printed out for that stress (in other words, an infinite margin of safety) |
| | | 6) EPSIL(6) (default 1.x10-15) is used in BAR margin of safety calculations
| EQCHECK | Int | Field 3: Default = 0 (basic origin) or reference grid to use in calculating the rigid body displacement matrix for the equilibrium check |

0 comments on commit cf6f9fd

Please sign in to comment.