Blast Explicit FEA — Mesh, Energy & Solution Controls

Numerical Stability Is Not the Same as Physical Accuracy

An explicit model can run to completion while producing a poor structural prediction. The central verification question is whether mesh, timestep, contact and numerical controls change the response quantities used for acceptance. Solver stability is necessary, but convergence of physically meaningful outputs is the stronger requirement.

Mesh the Deformation Mechanism, Not the Picture

Element size should resolve expected curvature, plastic hinges, local contact and gradients that influence the governing limit state. Uniformly refining the entire model is inefficient. Transition zones should be gradual enough to avoid spurious wave reflection. Shell elements are often appropriate for thin panels, while local solids may be needed around thick joints, contact or through-thickness stress features.

Local Strain and Failure Are Mesh Sensitive

Once softening, damage or element deletion begins, localisation can make failure strain and dissipated energy strongly element-size dependent. Mesh refinement may therefore reduce apparent ductility instead of converging. Use regularised material models where available, calibrate to representative element size, or base acceptance on less mesh-sensitive structural quantities when detailed fracture is not required.

Stable Timestep and Mass Scaling

The explicit stable timestep is controlled by the smallest element dimension and wave speed. Small local elements can dominate runtime. Mass scaling may be acceptable if added mass is limited and does not alter the inertial response controlling blast demand. Assess both global added mass and its location; a small total mass change concentrated in a critical panel can still matter.

Hourglass and Artificial Energy

Reduced-integration elements can develop zero-energy deformation modes. Hourglass controls suppress these modes but introduce artificial stiffness or energy. The acceptable level is problem-dependent; the important point is that artificial contributions remain small enough not to drive the result and that deformation shapes are physically credible. Do not use a single universal percentage without considering the model formulation.

Contact Energy and Penalty Stiffness

Contact algorithms may add or remove numerical energy depending on implementation. Large contact energy, excessive penetration or chatter can indicate poor penalty stiffness, mesh mismatch or unrealistic initial conditions. Review interface force histories and visual contact sequence together with energy terms.

Artificial Damping and Bulk Viscosity

Numerical damping can suppress high-frequency noise and shock oscillation, but it can also remove real structural response. Use default or tuned controls only with an understanding of which frequency content is being affected. Compare key displacements, reactions and accelerations with reduced damping to establish that conclusions are not control-dependent.

Energy Balance Is a Diagnostic, Not a Pass/Fail Number

Kinetic and internal energy histories should make sense for the identified response regime: applied work first accelerates and deforms the structure, then energy transfers between kinetic and strain/plastic forms and may be dissipated through damage or contact. Unexpected jumps often reveal load, contact or deletion problems. The precise accounting varies by solver, so interpret definitions rather than copying generic thresholds.

Convergence of Decision-Driving Outputs

Perform mesh and control sensitivity on peak/permanent displacement, support rotation, major reactions, acceleration bands and failure extent as relevant. Local peak strain may not converge in a conventional sense near a discontinuity. The verification plan should distinguish quantities expected to converge from those used only qualitatively.

Engineering Outcome

A defensible explicit-analysis package records mesh strategy, timestep, mass scaling, artificial energies, contact settings, material regularisation and sensitivity results. The aim is to show that the engineering conclusion survives reasonable numerical changes, not merely that the solver completed without error.

For explicit blast FEA, convergence means stability of the engineering decision under reasonable numerical refinement — not simply a smooth animation or a completed job.

Mesh Transition and Wave Reflection

Abrupt element-size changes can reflect stress waves and introduce local oscillation. Transition gradually between fine blast-response regions and coarser global structure, particularly where wave propagation influences support reaction. Compare reaction histories before and after transition refinement to ensure numerical reflection is not controlling the result.

Shell Thickness Integration and Through-Thickness Response

Shell models require enough integration points to represent bending plasticity and reversal. Too few points can distort plastic hinge development or residual curvature. If through-thickness stress, delamination or local indentation is important, shells may no longer be sufficient and local solid modelling should be considered.

Deletion, Mass Loss and Momentum

When elements are deleted, the treatment of their mass and momentum varies by solver. This can affect subsequent motion, especially if a substantial panel area erodes. Review deleted mass and energy histories and test whether the global response changes materially when deletion is delayed or replaced by a non-eroding damage cap.

Automation and Reproducibility

Explicit blast studies often involve many load cases and sensitivity runs. Scripted extraction of peak displacement, permanent set, reactions, energies and model diagnostics reduces manual error and makes comparisons reproducible. Automated checks should flag unusual energy balance, mass scaling or contact penetration rather than simply collect headline maxima.

Precision and Output Sampling

Explicit solvers may use single or double precision and write output at intervals much larger than the internal timestep. Coarse output can miss short force peaks even when the solver captures them. Choose output sampling from the bandwidth of the engineering quantity, and use higher-precision runs when accumulated contact or energy error is suspected to affect a marginal result.

Initial Penetration and Contact Setup

Small initial penetrations or tied-contact conflicts can inject force at time zero. Review initial contact diagnostics and early energy histories before the blast pulse arrives. Correct geometry or contact definitions rather than relying on damping to suppress artificial start-up motion.

Quality Dashboard

A concise explicit-analysis dashboard can report total/added mass, stable timestep, major energy terms, peak contact penetration, pressure resultant, reaction balance and key response sensitivities. Standardising these checks across blast cases makes numerical quality easier to review and prevents visually plausible but poorly controlled runs from entering the substantiation set.

Key takeaways

• Design the mesh around the physical deformation and failure mechanisms.
• Bound mass scaling, artificial energy and contact effects against decision-driving outputs.
• Demonstrate sensitivity of displacement, reaction and damage measures rather than relying on solver completion.