Skip to content
QuenchLogic

← Historical schematic · Preserved prior solver runs · separate computed cylinder references

MATERIALS / COMPUTATIONAL EXPERIMENT 01

Quench.
Matter in motion.

Cooling steel changes more than its temperature. Follow heat flow, crystal transformations and the strains that reshape a material.

Real solver output. Traceable reference cases. A first step toward explaining the historical curvature of the Japanese sword.

FINITE ELEMENT RESULTSThermal–metallurgical referenceDrag to orbit · scroll to zoom
Temperature°C
01
Axisymmetric calculation · revolved for 3D display
Inspect a display sample

Click the surface, or enter an index. Coordinates refer to the undeformed meridional model; this is a resampled display point, not a raw FE node.

0.0s
STARTStored solver steps · no live recalculationEND
LOWEST TEMPERATURE—
HIGHEST TEMPERATURE—
AUSTENITE—
MARTENSITE—

WHAT THIS CALCULATION ESTABLISHES

Following the transformation.

34 SECONDS / TWO PRESERVED REFERENCE CASES

Watch the transformation.

Chapter 1: temperature, bainite and martensite in MTLP103A. Chapter 2: growth, contraction and axial stress in the separate HSNV101B case. Axisymmetric fields are reconstructed in 3D; displacement is magnified ×5 in chapter 2. The two cases are separate calculations. Cooling uses prescribed boundaries; no CFD solution is shown.

THE HISTORICAL QUESTION

Why does the katana curve?

01 / HEAT

A boundary that changes.

A clay-based coating changes the blade’s interaction with the quenching liquid. It is more complicated than insulation: the original experiments found that thin clay could increase heat transfer in a high-temperature regime.

02 / PHASE

The crystal structure changes.

Different cooling histories produce different phase fractions. Transformation changes the material’s natural dimensions, while temperature changes its thermal contraction and mechanical response.

03 / STRAIN

Compatibility leaves a memory.

The connected material cannot follow every region’s preferred expansion independently. Elastic strain, plastic strain and transformation plasticity evolve together, leaving permanent shape and residual stresses.

The historical terms are mune for the back or spine and ha for the cutting edge. The reference specimens above are not sword models. Their results must not be transferred to a blade geometry or steel without a separate material model and validation.

TOWARD THE COMBINED MODEL

Every coupling must earn its place.

The first reference calculations check individual parts of the chain. A quantitatively accurate historical simulation needs compatible data and verification across the full chain.

  1. Heat → phase fractionsReference steel transformation calculation
  2. Temperature + phases → mechanicsSeparate constitutive benchmark with transformation plasticity
  3. Transformations → thermal feedbackExplicit phase-dependent latent heat remains a development requirement
  4. Coating + boiling → heat fluxCalibrated nucleate / transition / film boundary data still required
  5. Historical geometry → curvatureRequires a separate literature-grounded case; no fracture prediction here
Sources, verification and data handling +

Benchmark checks are being assembled.

Three.js displays precomputed fields. The 3D surface is a visual reconstruction of an axisymmetric calculation, not a claim that the underlying solver used a full 3D mesh. Nodal averaging and display resampling are recorded with the data. Colours use a fixed scale across the entire selected case. Animation advances through stored solver frames.

Original explanatory design; no historical paper figures reused. Reference input attribution: EDF, Code_Aster, GPL-3.0-or-later. Software notices.