1. Introduction to Oasys D3PLOT and T/HIS

Oasys D3PLOT and T/HIS are two complementary postprocessors developed by Arup for the Oasys LS-DYNA Environment. They are used across the CAE community to inspect, quantify, and communicate results from explicit and implicit finite element analyses. D3PLOT is the graphical results viewer: it renders the model geometry, contour plots, animations, and section cuts, and it lets you interrogate individual nodes, elements, and parts. T/HIS is the time-history processor: it reads ASCII and binary curve data, applies mathematical operations to curves, and produces publication-quality XY plots. Understanding where each tool sits in the postprocessing chain—and how they exchange data—is the foundation for every workflow that follows in this course.

Two tools, two questions

Postprocessing answers two fundamentally different questions. The first is spatial: where in the structure is the stress highest, how does the deformation evolve over time, which part absorbs the most energy? D3PLOT answers this class of question. It works on the model database (typically d3plot states written by LS-DYNA) and gives you the full 3D context: fringe plots, vector plots, deformed and undeformed overlays, cross-sections, and animated state sequences.

The second question is temporal and quantitative: what is the peak acceleration at a specific node, how does the force–displacement curve compare between two design variants, when does the contact force exceed a threshold? T/HIS answers this. It operates on curves—binout channels, nodout/elout ASCII files, rcforc contact data, or any two-column text file—and provides filtering, integration, differentiation, FFT, and curve arithmetic.

The distinction matters in practice because the two tools have different data models. D3PLOT's native currency is the state database; T/HIS's native currency is the curve. A complete postprocessing session usually moves between them: identify a hot spot in D3PLOT, extract its time history, then analyse that history in T/HIS.

Typical end-to-end workflow

A conventional session follows a predictable arc, though the details vary by analysis type. The steps below describe a crashworthiness-style example; the same skeleton applies to drop tests, forming, and blast analyses.

1. Load the model. In D3PLOT, open the d3plot family. The reader reconstructs the mesh, part and material assignments, and the state sequence. Confirm the number of states and the time step between them—this determines the temporal resolution available for animation and for any history you extract.
2. Orient and sanity-check. Set a consistent view, check that the model is complete (no missing parts), and verify units. A common early error is a unit mismatch between the solver deck and the postprocessing expectations, which shows up as absurd fringe magnitudes.
3. Scan the global response. Animate the full event, then step through key states. Use fringe plots of von Mises stress, plastic strain, and resultant displacement to locate hot spots and the timing of peak deformation.
4. Quantify locally. Pick specific nodes or elements at the hot spot. D3PLOT can write their time histories directly, or you can identify the node IDs and read the corresponding channels from binout or nodout in T/HIS.
5. Analyse curves in T/HIS. Apply the operations the question demands: a CFC filter for acceleration, integration for velocity and displacement, or a moving average to remove numerical noise. Compare against test data or a baseline design.
6. Report. Export annotated images, animations, and XY plots. For repeatable studies, record the session as a script so the same views and curves can be regenerated after a design change.

How the tools integrate with solvers

D3PLOT and T/HIS are solver-agnostic in principle but LS-DYNA-native in practice. The d3plot format is LS-DYNA's own state database, and D3PLOT reads it directly with no translation step. The binout file is LS-DYNA's binary time-history container; T/HIS reads it natively and exposes each channel (e.g. glstat energy, rcforc contact force, nodout nodal acceleration) as a selectable curve.

For other solvers, the path is usually through ASCII. PAM-CRASH, Radioss, and Abaqus results can be converted to LS-DYNA-format databases with the Oasys translators, or their time histories can be exported as two-column text and read directly into T/HIS. The practical consequence is that the same postprocessing skills transfer across solvers, provided the data is brought into a format the tools understand.

What "good" postprocessing looks like

The quality of a postprocessing result is not measured by how attractive the fringe plot is, but by whether the extracted quantities answer the engineering question with a defensible level of confidence. Three habits distinguish effective use of these tools:

• Know the state spacing. If states are written every 1 ms and the event of interest lasts 5 ms, you have five samples—enough for a trend, not for a peak. Request denser output in the solver deck when peak values matter.
• Filter deliberately. Raw nodal acceleration from an explicit run is noisy. A CFC 60 or SAE 1000 filter is standard for occupant and structural acceleration respectively; state the filter in the report so the result is reproducible.
• Cross-check spatial and temporal views. A peak stress in D3PLOT should correspond to a peak in the T/HIS curve for the same element. If they disagree, suspect a unit error, a wrong state, or a misidentified element ID.

These habits are developed throughout the course. For now, the key mental model is simple: D3PLOT shows you where and when in space; T/HIS tells you how much and how it compares. The two views are most powerful when used together, and the rest of this course builds the workflows that connect them.