NLPARM (Bulk Data Entry)

Bulk Data Entry

NLPARM – Parameters for Nonlinear Static Analysis, Nonlinear Direct Transient Analysis, and Heat Transfer Analysis

Description

The NLPARM bulk data entry defines parameters for Nonlinear Static Analysis, Nonlinear Direct Transient Analysis, and Heat Transfer Analysis solution control.

Format

(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
NLPARM	ID	NINC	DT		KSTEP	MAXITER	CONV
	EPSU	EPSP	EPSW			MAXLS		LSTOL
		TTERM

(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
NLPARM	99	5

Field	Contents
ID	Each NLPARM bulk data card must have a unique ID. No default (Integer > 0)
NINC	Number of implicit load sub-increments. See comments 2 and 3. Default = 1 (no increments) for ANALYSIS = NLSTAT and ANALYSIS=NLHEAT (Integer > 0)
DT	Initial load increment. See comment 7. No default (Real ≥ 0.0)
KSTEP	Number of iterations before stiffness update. See comment 3. Default = 6 for ANALYSIS = BCS solver Default = 3 for ANALYSIS = PCG solver (Integer > 0)
MAXITER	Limit on number of implicit iterations for each load increment. If reached, the solution is terminated (ANALYSIS = NLSTAT and ANALYSIS=NLHEAT). Default = 25 (Integer > 0)
CONV	Flags to select implicit convergence criteria. See comment 6. Default = UPW for ANALYSIS = NLSTAT Default = PW for ANALYSIS = NLGEOM (Any combination of U, P and W)
EPSU	Error tolerance for displacement (U) criterion. See comment 6. Default = 1.0E-3 for ANALYSIS = NLSTAT Default = 1.0E-2 for ANALYSIS = NLGEOM (Real > 0.0)
EPSP	Error tolerance for load (P) criterion. See comment 6. Default = 1.0E-3 for ANALYSIS = NLSTAT and ANALYSIS=NLHEAT Default = 1.0E-2 for ANALYSIS = NLGEOM (Real > 0.0)
EPSW	Error tolerance for work (W) criterion. See comment 6. Default = 1.0E-7 for ANALYSIS = NLSTAT and ANALYSIS=NLHEAT Default = 1.0E-3 for ANALYSIS = NLGEOM (Real > 0.0)
MAXLS	Maximum number of line searches allowed for each iteration. Default = 0 for ANALYSIS = NLSTAT Default = 20 for ANALYSIS = NLGEOM (Integer > 0)
LSTOL	Line search tolerance. Default = 1.0E-3 (Real > 0.0)
TTERM	Termination time. See comment 7. Default = 1.0 (Real ≥ 0.0)

Comments

1.	The NLPARM bulk data entry is selected by the Subcase Information command NLPARM=option. Each subcase for which nonlinear analysis is desired requires an NLPARM command.

The solution method for quasi-static nonlinear analysis (ANALYSIS = NLSTAT) is full Newton. The stiffness matrix is updated at each iteration. NINC > 0 represents the number of equal subdivisions that the total load in a given subcase will be divided into. If NINC is blank, the entire load for a given subcase is applied at once. The Newton method will be applied to consecutive load levels until the final load is reached.

3.	Additional control for geometric nonlinear implicit static solution schemes (ANALYSIS = NLGEOM) can be defined using the NLPARMX bulk data entry. Defaults will be used if NLPARMX is not present.

4.	The solution method for geometric nonlinear implicit analysis (ANALYSIS = NLGEOM) is modified or Quasi-Newton. The frequency of stiffness matrix updates is controlled by KSTEP. For highly nonlinear problems, it is recommended to reduce KSTEP for better performance. KSTEP = 1 means full Newton.

If the loading is defined using NLOAD, the termination time TTERM must defined by a TTERM subcase entry (this is different from the TTERM field on this NLPARM entry). The initial implicit time step is TTERMS/NINC with TTERMS = TTERM – T0. All subsequent time steps will be determined automatically. In a simulation with multiple nonlinear subcases, T0 is the end time of the previous load step. If there is only a single nonlinear subcase, T0 = 0.0.

If the loading is defined using LOAD, TTERM is not mandatory. These loads are treated as linear ramp-up. If TTERM is defined, the load ramps up from the end time of the previous subcase to TTERM. If TTERM is absent, it will be determined from the subcase sequence such that the duration of each subcase TTERMS = 1.0. In this case, the initial time step is 1.0/NINC.

For Nastran compatibility, NLPCI is imported if present. Only the fields ID, TYPE are interpreted. With NLPCI present, the default for NLPARMX, SACC will be reset to RIKS. TYPE will be translated into CTYPE; all other entries are set to default. A warning will be issued. NLPCI and NLPARMX cannot be used simultaneously. It is recommended to remove NLPCI and use NLPARMX with the appropriate definitions.

U, P, and W are convergence criteria for nonlinear solutions. U is the Displacement-based convergence criterion, P is the Load-based convergence criterion, and W is the Work/Strain-based convergence criterion. For increased accuracy, all three convergence criteria are set together as the default for NLSTAT analysis. If one or more convergence criteria are excluded, the result accuracy may vary. For additional information regarding convergence criteria, see the Nonlinear Convergence Criteria section.

The DT and TTERM fields allow for an alternate method (instead of NINC) to define the loading curve for Nonlinear Static Analysis. DT and TTERM, if specified together, override the value defined on the NINC field. The DLOAD Subcase Entry can be used in a NLSTAT subcase to reference the TLOAD1 Bulk Data Entry. This allows for the specification of a user-defined loading curve via the TID field on the TLOAD1 entry. The TID field references the Time (X) vs Load Scale Factors (Y) that scales the loading specified on the EXCITEID field. If DT is not specified, the default initial load increment is equal to TTERM/NINC. If DT, TTERM, and NINC are not specified, the default initial load increment is equal to 0.1 (that is, 1.0/10).

8.	This card is represented as a loadcollector in HyperMesh.