Skip to main content

Thermocouples and Furnace Pyrometry: Accuracy, Calibration, and Placement

Pyrometry is the discipline of measuring and controlling the temperature of a heat-treatment process, and the thermocouple is its primary sensor. UTEC Industrial provides in-house induction hardening, through-hardening, and quench-and-temper heat treating services for industrial components in the Pacific Northwest, with integrated CNC machining and reverse-engineering capability. The accuracy with which a furnace reaches and holds setpoint — and the accuracy of the record that documents the cycle to the customer — is bounded by the thermocouple system at each measurement point: the thermocouple wire itself, the extension cable, the reference junction compensation, the signal conditioning, and the calibration chain that traces each element back to a recognized temperature standard. A cycle that looks perfect on a programmable controller's display can still miss specification if the thermocouple is out of calibration, if it has drifted in service, if it is placed in a location that does not represent the load, or if the reference junction is uncontrolled. This article covers thermocouple type selection, accuracy and drift behavior, calibration practices, and placement strategy for industrial heat treatment — with specific attention to the practices that matter for code-required and specification-driven cycles.

What thermocouple types are used in industrial heat treatment, and how do they differ?​

Industrial heat-treatment thermocouples fall into two broad categories: base-metal types used for routine furnace control and most process measurement, and noble-metal types used when higher accuracy, higher operating temperature, or longer service life justifies the cost. Type K (chromel-alumel, nickel-chromium versus nickel-aluminum) is the workhorse of industrial heat-treatment — rated to 2,300 °F (1,260 °C) continuous service, with standard-class accuracy of ±2.2 °C or ±0.75% of reading (whichever is greater) and special-class accuracy of ±1.1 °C or ±0.4% of reading. Type N (Nicrosil-Nisil) is functionally similar to Type K at roughly the same cost but with improved drift stability in cycling service and less susceptibility to short-range ordering effects in the 500–1,100 °F range; it is increasingly the preferred base-metal choice in new furnace installations. Type J (iron-constantan) and Type T (copper-constantan) see use at lower temperatures (under 1,400 °F and 700 °F respectively) but are rarely used in steel heat treatment. Type S (platinum versus platinum-10% rhodium) and Type R (platinum versus platinum-13% rhodium) are the noble-metal standards for high-accuracy work and for service above 2,300 °F, with accuracies approaching ±0.5 °C in calibrated conditions — these are used for pyrometry surveys, reference calibrations, and in applications where Type K's drift over hundreds of hours of service would be unacceptable. Type B (platinum-6% rhodium versus platinum-30% rhodium) extends the platinum range to above 3,100 °F for specialty applications (ASM Handbook, Vol. 4B, ASM International, 2014; ASTM E230; Omega Engineering, Temperature Handbook reference practice).

What is AMS 2750, and what does it require for furnace pyrometry?​

AMS 2750 is the SAE Aerospace pyrometry standard that defines accuracy, calibration, and documentation requirements for heat-treatment furnaces used in aerospace work. Although most heavy-industrial heat treatment outside aerospace is not AMS 2750-governed by contract, the document's framework is widely adopted as an industry reference for defining what "good pyrometry" looks like in any context. AMS 2750 classifies furnaces into six classes (Class 1 through Class 6) based on allowable temperature uniformity across the working zone: Class 1 ±5 °F, Class 2 ±10 °F, Class 3 ±15 °F, Class 4 ±20 °F, Class 5 ±25 °F, Class 6 ±50 °F. Each class corresponds to a different heat-treatment application — Class 1 and 2 are used for critical aerospace alloy heat treatment (nickel superalloy age, titanium solution), Class 3–4 for standard alloy-steel quench-and-temper and PWHT, Class 5–6 for stress relief and other sub-critical processes where tight uniformity is not required. The standard also classifies instrumentation into Types A through E based on measurement accuracy, redundancy, and calibration interval. For a furnace to be certified to a particular class, it must pass a Temperature Uniformity Survey (TUS) per the defined procedure — typically nine temperature sensors placed in a 3×3 array throughout the working zone, cycled through the operating range, with all readings verified to remain within the class tolerance at every survey point. System Accuracy Tests (SAT) are performed at shorter intervals to verify that the installed control and recording thermocouples continue to read accurately against a calibrated test thermocouple inserted alongside. The standard prescribes calibration intervals, documentation retention, and corrective-action procedures when out-of-tolerance conditions are found. Heat treatment facilities not held to aerospace-level contractual requirements still benefit from adopting AMS 2750's survey-and-calibration framework as an internal quality standard (AMS 2750, current revision; ASM Handbook, Vol. 4B, ASM International, 2014).

How are control thermocouples different from load (part-mounted) thermocouples, and why does it matter?​

A furnace typically operates with at least three distinct thermocouple categories: control thermocouples, recording thermocouples, and load (part-mounted) thermocouples. The control thermocouple is mounted in the furnace wall or roof and feeds the programmable controller that runs the cycle — it is what the controller uses to hit the setpoint and maintain the soak temperature. The recording thermocouple is often a second wall-mounted sensor routed to an independent chart recorder or data logger that documents the cycle for customer records and quality compliance. The load thermocouple is physically attached to the part being heat-treated (or to a surrogate calibration block representing the thinnest or thermally slowest section of the load) and documents the temperature the part actually reached during the cycle, not just the furnace atmosphere temperature. The distinction matters because the furnace atmosphere and the part can be significantly different in temperature during the heating phase: the atmosphere reaches setpoint first, and the part warms to setpoint over the next 30 minutes to 2 hours depending on its thermal mass. If the control thermocouple alone is trusted, the soak time may be counted from the moment the atmosphere reached setpoint — but the part may not reach setpoint until an hour later, shortening the actual part-at-temperature soak by that hour. Most specifications require that the soak be timed from when the load thermocouple reads setpoint, not when the control thermocouple does. For code-required cycles (PWHT on pressure vessels, for example), a specific number of part-mounted thermocouples are required by the governing code (ASME Section VIII, AWS D1.1, and similar), with placement at the coolest expected location to guarantee every point in the load meets the soak requirement. UTEC Industrial's standard practice on heat treatment jobs is to tack part-mounted thermocouples directly to the load at coolest-point locations, routing the signal through feedthrough ports in the furnace door to the chart recorder that produces the documentation record (AMS 2750; ASME Section VIII Div 1, UW-40; ASM Handbook, Vol. 4B, ASM International, 2014).

What is thermocouple drift, and how is it detected and managed?​

Thermocouple drift is the progressive change in the output-versus-temperature relationship of an installed thermocouple over hundreds to thousands of hours of service at elevated temperature. Type K is particularly susceptible to two drift mechanisms: the "green rot" attack in low-oxygen atmospheres (reducing or vacuum environments in the 1,600–2,000 °F range), which selectively oxidizes the chromium in the chromel leg and shifts the calibration negative, and the "K-state" ordering phenomenon in the 500–1,100 °F range, which shifts the calibration by 1–3 °C after extended cycling through that temperature window. Type N was specifically developed to resist both mechanisms and is more stable than Type K over comparable service. Drift is detected by System Accuracy Test (SAT) — inserting a calibrated test thermocouple alongside the installed control or load thermocouple during a known-stable soak, reading both through equivalent instrumentation, and comparing. If the installed thermocouple reads within the allowed tolerance of the test thermocouple, it is considered still within calibration; if it exceeds the tolerance, it is replaced. The SAT interval depends on the pyrometry class and service severity: AMS 2750 Class 2 furnaces typically require monthly SATs on control thermocouples, while Class 5 and 6 furnaces have quarterly or semi-annual SAT intervals. Part-mounted thermocouples are single-use for the most part — the capacitor-discharge weld that attaches the junction to the part is cut off after the cycle, and a fresh thermocouple is welded on for the next job — so drift is less of a concern for load thermocouples than for the permanently installed control and recording thermocouples. The practice of replacing control thermocouples on a scheduled interval (typically every 6–12 months for Type K in heavy service) regardless of SAT results is a common preemptive management strategy in furnaces run at high duty cycle (ASM Handbook, Vol. 4B, ASM International, 2014; AMS 2750; ASTM E230).

Where should a load thermocouple be placed to produce a representative cycle record?​

The placement objective for a load thermocouple is to measure the temperature at the point in the load that reaches soak temperature last — so that when that point meets setpoint, every other point in the load has already reached or exceeded setpoint, and the whole load has soaked for at least the required minimum time. On a single part, this is typically the thickest cross-section or the interior of a large cored casting, where thermal mass delays warmup. On a mixed load, the slowest-warming piece is typically the largest or most thermally massive item, and the thermocouple is placed on it. Burner-blocking geometry matters: a part in the shadow of another part, or recessed below the level of the car bed in a zone where burner gas flow is restricted, may warm more slowly than a comparable part in direct gas-flow path, even if the two have similar thermal mass. Placement guidance for typical heat treatment loads: for a single large casting or weldment, one thermocouple at the thickest section and a second at the coolest-expected location (farthest from burner impingement, often near a corner or at a heavy mass intersection). For a stacked load of similar parts (crane wheel blanks, jaw plates), one thermocouple on the bottom-center part of the stack (the coolest expected location in a bottom-heated car-bottom furnace) and a second on a representative side part. For code-required PWHT, the specific thermocouple count and placement is dictated by the governing code — ASME Section VIII requires a minimum number based on vessel size and thickness, typically placed at the thickest section and at geometrically complex regions where thermal lag is expected. The physical attachment method for a load thermocouple is capacitor-discharge tack welding of the thermocouple junction directly to the part surface (for steel parts) or mechanical clamping to a drilled pocket for parts that cannot be tack-welded — both methods must place the junction in intimate thermal contact with the part, not in the surrounding atmosphere (AMS 2750; ASME Section VIII Div 1, Appendix R for local PWHT; ASM Handbook, Vol. 4B, ASM International, 2014).

What accuracy is achievable with properly calibrated industrial pyrometry, and what is reasonable to specify?​

A calibrated Type K or Type N thermocouple with documented System Accuracy Test results can reliably read within ±3 °F of true temperature across the 800–1,800 °F range typical of industrial heat treatment. Combined with a calibrated signal conditioner and programmable controller, the total system accuracy for a control loop in a well-maintained commercial heat-treatment furnace is typically ±5 to ±10 °F at the control thermocouple location, and ±10 to ±20 °F at representative load locations due to furnace uniformity effects. For specifications, requiring the cycle to hit a target temperature within ±25 °F is routine and achievable on any well-maintained industrial furnace; ±15 °F is achievable but requires careful furnace setup and a good uniformity survey result; ±5 °F is aerospace-class work that requires specialty furnace equipment, tight pyrometry class certification, and fresh thermocouples for each cycle. Specifications that call for unrealistically tight uniformity (±5 °F across a 50-ton load in a commercial car-bottom furnace) either cannot be met or require a specialty heat treater with different equipment — the right response is to negotiate a realistic tolerance or source the work to the appropriate facility. UTEC Industrial's car-bottom furnace, with calibrated Type K control and recording thermocouples and documented cycle records, routinely produces cycles within the ±25 °F tolerance typical of commercial stress relief and annealing work; tighter tolerances are discussed on a per-job basis and referenced against the furnace's most recent uniformity survey results. The specification of realistic accuracy requirements — matched to the actual heat-treatment process sensitivity — produces better outcomes than specifying the tightest possible number on every drawing (AMS 2750; ASM Handbook, Vol. 4B, ASM International, 2014; ASTM E230).

How is thermocouple data documented and delivered to the customer as part of the cycle record?​

The temperature record delivered to the customer after a heat treatment cycle includes the time-temperature trace from each recording thermocouple — both the furnace control thermocouple and the load thermocouples. The trace is typically delivered as a strip chart (older installations) or as a digital record printed to PDF and archived in machine-readable form (modern installations). The record shows the full cycle from ambient through the ramp, soak, and cool, with time on one axis and temperature on the other; the soak portion is annotated with the soak duration (the time the load thermocouple remained at or above the specified minimum), and any excursions outside tolerance are flagged. Accompanying the trace, the documentation package typically includes: the programmed cycle setpoints (the intended cycle); identification of the thermocouple type, class, and calibration status (date of last SAT and test result); identification of the furnace used; and a statement of cycle completion (whether the soak met the minimum time at minimum temperature across all load thermocouples). For customers whose contractual requirements include pyrometry compliance (aerospace, some pressure vessel, some defense), additional documentation of the pyrometry class certification and supporting SAT and TUS records may be appended. UTEC Industrial's standard documentation format includes the programmed cycle, the thermocouple trace, the cycle-completion statement, and the hardness verification results where hardness is part of the acceptance criterion — delivered as a PDF package with every job, with physical chart records available on request (AMS 2750; ASM Handbook, Vol. 4B, ASM International, 2014; ASME Section VIII Div 1, UW-40).

Related Articles

References​

  • ASM International. (2014). ASM Handbook, Volume 4B: Steel Heat Treating Technologies. ASM International.
  • AMS 2750: Pyrometry (current revision). SAE Aerospace.
  • ASTM E230: Standard Specification and Temperature-Electromotive Force (EMF) Tables for Standardized Thermocouples. ASTM International.
  • ASME Boiler and Pressure Vessel Code, Section VIII Division 1 (current edition). American Society of Mechanical Engineers. UW-40, Appendix R.
  • ASM International. (1995). Heat Treater's Guide: Practices and Procedures for Irons and Steels (2nd ed.). ASM International.
  • Totten, G.E. (ed.). (2006). Steel Heat Treatment Handbook (2nd ed.). CRC Press / Taylor & Francis.

Need In-House Heat Treating for Heavy Industrial Parts?​

UTEC Industrial operates a 6' × 10' × 17' car-bottom furnace (1,800 °F, 50-ton capacity), in-house induction hardening with per-part hardness verification, and automated vibratory stress relief at our Spokane, WA facility. Weldment stress relief, annealing, quench and temper, and induction hardening — all under one roof, with full documentation on every job.

Request a Quote →

Questions? Call (509) 922-1832 or email sales@utec.co

Last updated on