Improving Load Balance in MPP-LSDYNA for ALE and SPH

In an earlier post on decomposition and region domains we saw how customized PFILE can be used to improve load balancing. Two new commands “aledist” and “sphdist” can now be included in “PFILE” that allows LS-DYNA to distribute the ALE and SPH information across all processors. This will greatly improve the load balancing when using…

January 17, 2008 | by

Modified Crushable Foam

LS-DYNA now offers a modified version of *MAT_CRUSHABLE_FOAM in the form of *MAT_MODIFIED_CRUSHABLE_FOAM . In the new material model, the strain rate effects can be included in the foam a *DEFINE_TABLE . To reduce or eliminate the noise in the evaulation of the new yield stress as a function of strain and strain-rate, LS-DYNA offers…

January 17, 2008 | by

D3VIEW 2.0 Beta Released

As of today, D3VIEW 2.0beta has been officially released in a closed environment. Update Due to heavy traffic, the site is currently down. Should be up shortly.

January 12, 2008 | by

Timestep Calculation

Attached is a simple Excel sheet that can be used to quickly evaluate the timestep calculated by LS-DYNA for solids, beams and shells. It must be noted that for solid elements, Q is based on the C1 (default = 0.06) only. timestep.xls Update For solid elements using MAT_SPOTWELD, LS-DYNA uses a slightly different method of…

January 10, 2008 | by

CONTACT_AUTO_MOVE to eliminate initial contact clearances

In many situations, initial clearances between two contact surfaces can cause convergence issues in IMPLICIT analyses while wasting simulation time in EXPLICIT analyses. To remove the initial contact clearance, LS-DYNA offers a capability using *CONTACT_AUTO_MOVE that helps to improve convergence in IMPLICIT analyses since contact is detected immediately. Contact AUTO_MOVE associates itself with another contact…

January 10, 2008 | by

Non-Structural-Mass (NSM) Lumping in LS-DYNA

LS-DYNA provides three ways to account for the non-structural-mass lumping on structures. 1. Traditional Node Based Mass Lumping using ELEMENT_MASS_{OPTION} This method relies on the preprocessors to compute the node-based additional mass that is to be lumped based on the total non-structural-mass that is to be included. The input (into a preprocessor) would simply be…

January 9, 2008 | by

Yield Stress, Strain-rate Dependency in MAT_024 and the significance of SIGY

*MAT_PIECEWISE_LINEAR_PLASTICITY (*MAT_024) is widely used material model for metals and in some cases plastics. Its popularity is widespread since it offers several plasticity models and can also be strain-rate dependent. One particle parameter, the Yield Stress, in the material card can appear in more than one place and can be sometimes confusing to know which…

December 20, 2007 | by

d3View v2.0beta – LS-DYNA Simulation Data Manager – Releasing Shortly

Since the first release of d3View, there have been many incremental but bold steps that were taken to tackle the problem of simulation data management especially related to LS-DYNA. I am proud to announce the next generation of simulation data manager that will be available shortly to worldwide users in the form of d3View v2.0…

December 17, 2007 | by

Force Deflection to Effective Stress vs Effective Strain

Attached is a simple code that I wrote a while back to convert a force-deflection/engineering-stress_strain/true_stress_strain curve to effective stress vs effective strain curve. The input is a simple LS-DYNA valid keyword file with *define_curve keyword. Its not an elegant code but works in most cases. A sample curve is also included. I am working on…

December 13, 2007 | by

Initializing Velocity

Initial velocity is always specified to a single node and can carry only one value. Multiple initial velocity definitions for the same node will always use the last encountered value. To illustrate this, if we have nodes N1,N2 and N3 and we define a initial velocity for all nodes to be ‘V1’ followed by N2…

December 13, 2007 | by