Calibrate a heat-treatment model from CSV and deliver a reviewable comparison

Establish what the measurements represent

Suppose thermocouple CSV files from cylindrical specimens will support a transfer-and-quench comparison. Record material batch, dimensions, sensor positions, time reference, units, and equipment conditions. Confirm that the selected material pack supports the calculation. Retain an unchanged source CSV and document cleaning and mapping separately.

Build the baseline before fitting

1. In HeatTreat Studio, select the material and slab, cylinder, or sphere geometry and define the thermal stages.
2. Run the uncalibrated baseline and inspect boundary conditions and surface/core histories.
3. Import thermocouple CSV and map time, temperature, position, units, batch, and uncertainty. Check HRC or Jominy data according to their own type.
4. Fit only parameters supported by evidence. Compare errors, identifiability, approximate intervals, and holdout results.
5. Save the calibration revision and compare scenarios under consistent assumptions.

A lower fitting error is not enough

If the surface fit improves while core or holdout behavior worsens, do not report only the average reduction. Recheck time zero, sensor depth, units, and parameter identifiability. Use mesh and time-step diagnostics to assess numerical influence. Do not add freedom solely to make curves coincide, or enable material tools without their required data.

Deliver a reproducible review package

For up to four scenarios, label exactly what changed and keep material, geometry, and reporting conventions comparable. Save the local project and export PDF, CSV, or Markdown with sources, mapping notes, baseline, calibration, holdout behavior, and limitations. Another engineer should understand which parameters changed and whether improvement extends beyond fitted observations.

Scope

Published version 1.0 supports on-device screening, calibration assistance, and review exports. It does not issue acceptance decisions, certification, or production guarantees. Bundled material packs are engineering seeds, not substitutes for licensed databases, plant trials, or equipment qualification. Phase and hardness tools depend on the selected pack’s actual data.

Source: Apple App Store: HeatTreat Studio 1.0.