CoatingIQ
← Course index

Step 3 of 4

57%

Lesson 03·Running Forge Dies

Die-temperature management during the run: thermal profiling in plain English

Read a die-surface thermal trace, hold the H13 operating window, and use a contact pyrometer and an IR gun without lying to yourself about what either one is measuring.

7 min readLesson 3 of 13

Step 3 of 4Contact pyrometer versus IR gun

A contact pyrometer is a thermocouple in a handheld probe. Press the tip against the die surface, hold it until the reading stabilizes, read the temperature. It measures the actual surface temperature at one point, with no emissivity correction needed, because heat is transferring directly from the steel into the probe. The reading is trustworthy within the calibration of the instrument, typically plus or minus a few degrees. There are two caveats the operator has to manage. The first is stabilization. Contact pyrometers take time to settle, usually 3-15 seconds depending on probe mass and surface condition. A press-and-release on a hot die will under-read by tens of degrees because the probe never reached the surface temperature. The second is spatial coverage. The probe dwells at one point, so a contact reading is not a map of the die. It is a number for the specific spot where the probe touched.

An IR gun reads thermal infrared radiation from the surface and computes a temperature using an assumed emissivity. The output is a single number, but the underlying physics involves at least four variables the operator usually does not set correctly.

Emissivity is the main one. Polished steel reads around 0.3. Oxidized steel reads 0.7-0.9. Graphite-lubricant-coated steel reads somewhere between, and the value changes between hits as the lubricant layer builds, burns off, and refreshes. An IR gun set to emissivity 0.95 (the factory default on many guns) and pointed at polished steel will under-read by 100-200°C. The same gun pointed at oxidized scale will be roughly accurate. The same gun pointed at a lubricant-coated die surface is reading the lubricant film, not the steel underneath, and the relationship between the two depends on lubricant thickness at the moment of the reading.

The viewed area is the second variable. An IR gun has a distance-to-spot ratio, often noted as D:S on the spec sheet. A 12:1 gun aimed at a die surface from 12 inches away is averaging the temperature across a 1-inch circle. The hottest spot on a die face is rarely 1 inch across. The gun reads a weighted average, which is cooler than the peak. Angle is the third variable. IR emission falls off as the surface tilts away from the detector. A reading taken at 45 degrees off-axis is cooler than a reading taken normal to the surface, by a non-trivial amount on shiny steel. Distance is the fourth. An IR gun outside its rated D:S window is averaging over an area larger than intended, including cooler surroundings, and reading low.

The IR gun is not lying. It is reporting exactly what the sensor saw, given the assumptions it was set to. The operator who reads the number and acts on it without correcting for emissivity, angle, distance, and surface condition is the one doing the lying, mostly to themselves.

The right setup at the start of a run is a gut-check against contact. Heat a steel block (the die itself works if it is at temperature) to a known temperature read with the contact pyrometer. Point the IR gun at the same spot from the working distance and angle. Adjust the IR gun's emissivity setting until the gun reads the same as the contact pyrometer. Record the setting. That emissivity number is now valid for that surface condition. The moment the surface changes, oxidation thickens or lubricant builds, the calibration is approximate at best. Recalibrate at the start of each shift, or whenever the surface condition has visibly changed.

Quick check

Two operators on the same shift read the same point on a forge die with the same IR gun within a minute of each other. They disagree by 60°C. Both have been on the press for years. What is the most likely cause, and what does the disagreement do to the shift trend data?