Welcome to the LS-DYNA Product Space

<h3 class="PDq2pG_selectionAnchorContainer" data-section-id="ww3lfh" data-start="227" data-end="241">Forum post</h3><p data-start="243" data-end="258">Hello everyone,</p><p data-start="260" data-end="560">I am developing an LS-DYNA model for <strong data-start="297" data-end="372">shearing a rectangular Cu-ETP copper wire with two rigid cutting blades. The purpose of the model is to simulate separation of the copper wire by shear deformation, similar to a shearing or blanking process. It is not a machining or chip-formation simulation.</p><p data-start="562" data-end="864">The copper wire is modeled with solid elements and a Johnson–Cook material model. Material failure is currently defined using Johnson–Cook damage and/or <code data-start="715" data-end="733">*MAT_ADD_EROSION. The blades are modeled as rigid bodies, and <code data-start="779" data-end="816">*CONTACT_ERODING_SURFACE_TO_SURFACE is used between the blades and the copper wire.</p><p data-start="866" data-end="1160">The calculation runs successfully, but I do not know whether the predicted separation is physically reasonable or whether it is mainly caused by numerical settings. I would therefore like to understand how the keyword parameters should be selected and improved based on post-processing results.</p><p data-start="1162" data-end="1270">My main problem is that I cannot find clear recommendations for the reasonable ranges of parameters such as:</p><ul data-start="1272" data-end="1688"><li data-section-id="9tub6x" data-start="1272" data-end="1323">failure strain or <code data-start="1292" data-end="1300">EFFEPS in <code data-start="1304" data-end="1322">*MAT_ADD_EROSION;<li data-section-id="q6a4nz" data-start="1324" data-end="1373">Johnson–Cook damage parameters <span class="katex"><span class="katex-mathml">D1D_1</span><span class="mord"><span class="mord mathnormal">D</span><span class="vlist-r"><span class="mord mtight">1</span><span class="vlist-s">​</span></span></span></span>–<span class="katex"><span class="katex-mathml">D5D_5</span><span class="mord"><span class="mord mathnormal">D</span><span class="vlist-r"><span class="mord mtight">5</span><span class="vlist-s">​</span></span></span></span>;<li data-section-id="b5o2ju" data-start="1374" data-end="1429">contact penalty stiffness scaling, especially <code data-start="1422" data-end="1428">SFSA;<li data-section-id="py57fx" data-start="1430" data-end="1487">static and dynamic friction coefficients <code data-start="1473" data-end="1477">FS and <code data-start="1482" data-end="1486">FD;<li data-section-id="1xxp1r8" data-start="1488" data-end="1519">hourglass control parameters;<li data-section-id="1a4nobf" data-start="1520" data-end="1562">element size in the expected shear zone;<li data-section-id="1bl6tf7" data-start="1563" data-end="1585">cutting-edge radius;<li data-section-id="1zs3i9" data-start="1586" data-end="1621">clearance between the two blades;<li data-section-id="1b0xmru" data-start="1622" data-end="1639">blade velocity;<li data-section-id="x6mtrn" data-start="1640" data-end="1688">workpiece constraints and clamping conditions.<p data-start="1690" data-end="1959">When I change these parameters, I obtain very different results. In some cases, the copper wire does not separate. In other cases, elements are deleted too early, the blades penetrate the copper wire, or the fracture surface appears to be controlled mainly by the mesh.</p><p data-start="1961" data-end="2063">I am therefore unsure which results represent real shearing and which results are numerical artifacts.</p><p data-start="2065" data-end="2131">At present, I am evaluating the following post-processing results:</p><ul data-start="2133" data-end="2432"><li data-section-id="1kajx3x" data-start="2133" data-end="2176">blade force versus time and displacement;<li data-section-id="59yzj4" data-start="2177" data-end="2246">internal energy, kinetic energy, hourglass energy and total energy;<li data-section-id="13bdu2e" data-start="2247" data-end="2269">contact penetration;<li data-section-id="1511y5" data-start="2270" data-end="2297">effective plastic strain;<li data-section-id="1avm5im" data-start="2298" data-end="2317">effective stress;<li data-section-id="1b5dn1b" data-start="2318" data-end="2357">damage variable and element deletion;<li data-section-id="oqswxv" data-start="2358" data-end="2390">development of the shear zone;<li data-section-id="3cq22b" data-start="2391" data-end="2432">final fracture and separation geometry.<p data-start="2434" data-end="2492">I would be grateful for advice on the following questions:</p><li data-section-id="1gsrlxe" data-start="2494" data-end="2588"><strong data-start="2497" data-end="2588">Which post-processing results are essential for validating a metal shearing simulation?<li data-section-id="1qdla3i" data-start="2590" data-end="2709"><strong data-start="2593" data-end="2709">What should physically reasonable force–displacement and energy curves look like during shearing and separation?<li data-section-id="pe9qr9" data-start="2711" data-end="2785"><strong data-start="2714" data-end="2785">How can excessive contact penetration be identified quantitatively?<li data-section-id="4gm6x2" data-start="2787" data-end="2902"><strong data-start="2790" data-end="2902">How can I distinguish physical shear failure from artificial separation caused by premature element erosion?<li data-section-id="1620pok" data-start="2904" data-end="3102"><strong data-start="2907" data-end="2962">How should I determine a reasonable <code data-start="2945" data-end="2953">EFFEPS value?<br data-start="2962" data-end="2965">Should it be obtained from tensile-test data, shear-test data, fracture tests or calibration against an experimental shearing process?<li data-section-id="9f5rgc" data-start="3104" data-end="3256"><strong data-start="3107" data-end="3256">If Johnson–Cook damage is used, should <code data-start="3148" data-end="3156">EFFEPS in <code data-start="3160" data-end="3178">*MAT_ADD_EROSION still be active, or should it only be used as a numerical backup criterion?<li data-section-id="12tl3u7" data-start="3258" data-end="3521"><strong data-start="3261" data-end="3349">Which parameters are physical process parameters and which are numerical parameters?<br data-start="3349" data-end="3352">For example, blade clearance, edge radius and friction are physical parameters, whereas contact penalty scaling and hourglass control are mainly numerical parameters.<li data-section-id="1xhs70i" data-start="3523" data-end="3693"><strong data-start="3526" data-end="3608">How should a mesh-convergence study be performed when element erosion is used?<br data-start="3608" data-end="3611">The fracture path and separation time change when the element size is modified.<li data-section-id="ixiusu" data-start="3695" data-end="3892"><strong data-start="3698" data-end="3763">How can I determine whether <code data-start="3728" data-end="3734">SFSA is too small or too large?<br data-start="3763" data-end="3766">In my model, changing <code data-start="3791" data-end="3797">SFSA significantly affects penetration, local plastic strain and whether the copper wire separates.<li data-section-id="1r9y7tc" data-start="3894" data-end="4018"><strong data-start="3898" data-end="4018">What is a reasonable relationship between hourglass energy and internal energy for this type of shearing simulation?<li data-section-id="1hj35fo" data-start="4020" data-end="4234"><strong data-start="4024" data-end="4111">Should the blade velocity be low enough to obtain a quasi-static shearing response?<br data-start="4111" data-end="4114">If so, which energy ratio or other criterion should be used to confirm that inertial effects are sufficiently small?<li data-section-id="zo9dgj" data-start="4236" data-end="4363"><strong data-start="4240" data-end="4363">Are there any LS-DYNA example models for wire shearing, sheet-metal shearing, blanking, punching or guillotine cutting?<p data-start="4365" data-end="4466">I am especially looking for an example that explains the complete modelling and validation procedure:</p><p data-start="4468" data-end="4657"><strong data-start="4468" data-end="4657">mesh generation → material model → damage and failure definition → blade motion → contact definition → workpiece constraints → post-processing → parameter sensitivity study → validation</p><p data-start="4659" data-end="4885">My goal is not to find one universal value for every parameter. I would like to establish a systematic method to determine whether the baseline model is reliable and then decide which keywords or parameters should be modified.</p><p data-start="4887" data-end="5046">Any example keyword files, papers, tutorials or recommendations for <strong data-start="4955" data-end="5023">copper-wire shearing, sheet-metal shearing would be very helpful.</p><p data-start="5048" data-end="5082" data-is-last-node="" data-is-only-node="">Thank you very much for your help</p>

<p>Below is what I see when I export from LSPP ICFD parts after converting mesh to MS Mesh in LSPP 4.13:  Also, the newer version of LSPP animation bar won't update correctly (like the bar wouldn't move),  also if I try to click on the animation bar from the bar on the bottom toolbar, it won't pop up.   None of these issues occurd in 4.10. </p><p>can Ansys employee provide some detail as to why LSPP is doing these things in the newer version?  Should I just stick wtih the older versions? </p><p></p>

<p data-pm-slice="0 0 []">Hello,</p><p>I encountered the following error while running my simulation on an HPC  using LS-DYNA R12:</p><p>*** Error 41434 (SOL+1434)</p><p>Error encountered in SPH neighborhood search routine.</p><p>Please check mes#### file for detailed information.</p><p>Based on another Q&A on this forum, I increased NMNEIGH in *CONTROL_SPH and ran the simulation again. However, the simulation terminated with the same error. In fact, the error occurred even earlier than before.First run: error occurred at 1.7 s of simulation time. Second run (after increasing NMNEIGH): error occurred at 1.2 s of simulation time. I am using LS-DYNA R12.</p><p>Could someone please help me understand: What is the most common cause of this error?</p><p>What is the solution?</p><p>Is there a way to prevent the simulation from terminating when this error occurs, or is the termination unavoidable?</p><p>Any suggestions would be greatly appreciated.</p>

<p>Hello,</p><p>I am currently learning LS-DYNA and trying to simulate an end milling operation with chip formation.</p><p>I have modelled the milling cutter as a rigid body and the workpiece as a deformable body. However, I am facing difficulties in defining the correct contact interaction required for material removal. The simulation also terminates with errors such as "Energy error too large" and "General solver error."</p><p>I have attached screenshots of my model setup, contact definition, and the solver errors.</p><p>Could anyone please guide me on:</p><p>The recommended workflow for end milling simulations in LS-DYNA.</p><p>The correct contact definition for chip formation.</p><p>The essential Control Cards, Database Cards, and Keywords required for this type of simulation.</p><p>Any official tutorials or benchmark examples for metal cutting using LS-DYNA.</p><p>Thank you for your time and guidance.</p>

<div data-index="122">Reviewing the System Coupling User's Guide (Release 2026 R1), the list of supported coupling participants includes Mechanical/Mechanical APDL, Fluent, CFX, Forte, Electronics Desktop/Maxwell, FMU, Thermal Desktop, Rocky, and Files - but LS-DYNA does not appear in this list.</div><div data-index="122">Therefore, if possible, I'd like to clarify the current status of coupling between Ansys Rocky (DEM) and LS-DYNA.</div><div data-index="122"> </div><div data-index="123">So far, I've found clear documentation for:</div><div data-index="123"> </div><div data-index="125">- Rocky ↔ Ansys Mechanical (1-way and 2-way, via System Coupling)</div><div data-index="126">- Rocky ↔ Ansys Fluent (CFD-DEM)</div><div data-index="127">- LS-DYNA participating in System Coupling with Fluent and MAPDL</div><div data-index="128">- A sequential/homogenization workflow linking Rocky → Multiscale.Sim → LS-DYNA (Cybernet's Multiscale.Sim add-in), which is one-way and does not preserve time-resolved feedback between the two solvers</div><div data-index="129"> </div><div data-index="130">My specific questions:</div><div data-index="131"> </div><li data-index="132">Is direct two-way co-simulation between Rocky and LS-DYNA officially supported today (2026 R1/R2)? If so, is it via System Coupling, or another mechanism?<li data-index="133">If supported, is LS-DYNA's participant status in System Coupling still a beta/experimental feature for this type of coupling, or has it reached production maturity?<li data-index="134">Are there known limitations regarding the time-step mismatch between Rocky's DEM solver and LS-DYNA's explicit time integration (microsecond-scale)? Is subcycling or another mechanism used to reconcile this?<li data-index="135">Is there a reference case, tutorial, or technical manual (beyond the general System Coupling User's Guide) that documents this specific Rocky↔LS-DYNA workflow?<li data-index="136">What licenses/executables are required (e.g., a special LS-DYNA build with SyC DLLs, as required for LS-DYNA↔Fluent coupling)?<div data-index="138">Any pointers to official documentation, technical support cases, or example projects would be greatly appreciated.</div><div data-index="139">Thanks in advance!</div>

<p>i am using ansys explicit dynamics to simulate blasting in concrete block . I want to calculate the eroded volume( vlume of material that has been blasted), how can i do so?</p>

<p>Hello, does anyone on the Ansys side know when the LSDYNA student version will be updated?  The current version will end soon (July 31st 2026). </p><p> </p><p>Thank you!</p>

<p>Hello all,</p><p>I am trying to write a UMAT that requires the total strain from the previous cycle as well as needing to store internal stress modes for the material model I am trying to implement. Setting IHYPER = 1 and LHV = (some number) in the keyword seems to wipe hsv(1) - hsv(9) at the beginning of every cycle.</p><p>How do I make use of the deformation gradient and save internal stress modes at the same time?</p><p>Many thanks,</p><p>K</p>

<p>Hi,</p><p>I perform a implicit static analysis and use *DATABASE_BINARY_D3MAX/*DATABASE_MAX_SOLID_SET to output the maximum stress of a target solid element set, and then perform post-processing in LSPP. </p><p>The issue I encounter is that, for the same set, the location and value of the maximum stress from d3max and d3plot are different:</p><p>- In d3max, the maximum stress is 853 MPa, with element ID = 19651.</p><p>- In d3plot, the maximum stress is 932 MPa, with element ID = 19114.</p><p>I would like to know what causes this discrepancy.</p><p>Thank you very much.</p><p></p><p></p>

<p>https://www.researchgate.net/publication/315973637_Development_of_a_Steel_Brace_with_Intentional_Eccentricity_and_Experimental_Validation<p class="isSelectedEnd">Hi everyone,</p><p class="isSelectedEnd">I am trying to numerically reproduce the experiment presented in the attached paper.</p><p class="isSelectedEnd">I first modeled the specimen in Abaqus as a 2D model, using constraints to represent the rigid elements. Now I am trying to build the same model in LS-DYNA, but I am having difficulties modeling the rigid elements correctly.</p><p class="isSelectedEnd">I have tried using *CONSTRAINED_NODAL_RIGID_BODY with pinned supports (free rotation), but the entire assembly behaves as a single rigid body. I also tried modeling the rigid elements with beam elements having a very high stiffness, but the results are still differ significantly from those obtained with Abaqus and the experimental data.</p><p class="isSelectedEnd">What is the recommended approach in LS-DYNA for modeling rigid elements between two nodes while maintaining pinned boundary conditions? Is *CONSTRAINED_NODAL_RIGID_BODY the appropriate choice, or is there a better approach?</p><p class="isSelectedEnd">Any suggestions would be greatly appreciated.</p><p>Thank you!</p></p>