Repeatability verification confirms that temperature uniformity surveys of horizontal vacuum furnaces deliver stable, trustworthy results across consecutive test runs. This article explains the applicable standards, instrumentation, test loads, procedures, statistical evaluation methods, and acceptance criteria used by accredited third-party laboratories and in-house pyrometry teams. Understanding this process helps manufacturers maintain compliance with AMS 2750, GB/T 9452, GB/T 30825, and related heat-treatment pyrometry requirements.
Technical Overview
Horizontal vacuum furnaces are the workhorses of modern heat treatment, performing vacuum hardening, tempering, annealing, brazing, and sintering of high-value components made of tool steels, stainless steels, titanium alloys, superalloys, and cemented carbides. Because these processes are sensitive to even small thermal gradients, the temperature uniformity of the effective work zone must be qualified and continuously monitored.
Temperature uniformity testing, commonly called a temperature uniformity survey (TUS), maps the deviation between the furnace control setpoint and the actual temperature at multiple points inside the work zone. For horizontal vacuum furnaces, the survey is typically performed with nine calibrated sensors placed at the eight corners and the geometric center of the effective heating zone, with additional sensors added for multi-zone equipment.
Repeatability verification takes the survey one step further. Instead of relying on a single run, the same survey is repeated several times under strictly identical conditions — same setpoints, same sensor positions, same load state, same pressure regime — and the run-to-run dispersion of the measured uniformity values is statistically evaluated. This confirms that the survey results are reproducible, strengthens the measurement uncertainty budget, and demonstrates that the furnace can consistently hold its specified uniformity class over time.
- AMS 2750 defines furnace uniformity classes from ±3 °C (Class 1) up to ±28 °C (Class 6) and governs TUS frequency, sensor types, and instrumentation.
- GB/T 9452 specifies the determination of effective heating zones, with uniformity grades typically ranging from ±3 °C to ±25 °C.
- GB/T 30825 covers temperature measurement requirements for heat treatment, including sensor calibration and system accuracy.
- GB/T 10066.4 provides test methods for indirect resistance-heated vacuum furnaces.
- ISO/IEC 17025 governs the competence of laboratories performing the measurements, while ISO 5725 and the ISO/IEC Guide 98-3 (GUM) provide the framework for repeatability and uncertainty evaluation.
Testing Samples
The test object is the furnace system itself. Repeatability verification of temperature uniformity is applicable to the following categories of horizontal vacuum furnace equipment.
Horizontal Single-Chamber Vacuum Furnaces
Single-chamber batch furnaces with graphite or metallic heating elements and graphite felt or molybdenum radiation shielding, with hot zone dimensions commonly ranging from 300 × 300 × 450 mm to 1200 × 1200 × 1500 mm and maximum operating temperatures of 1100 °C to 1350 °C. These furnaces are surveyed empty, with a dummy load, or with a representative production load depending on the qualification stage.
Horizontal Multi-Chamber and Continuous Vacuum Furnaces
Multi-chamber lines that combine heating, soaking, and gas-quenching modules require each heating station to be surveyed independently, with inter-chamber door cycles and transfer sequences included so that the thermal disturbance caused by load transfer is captured in the uniformity record.
Horizontal Vacuum Brazing and Sintering Furnaces
Furnaces used for aluminum or nickel-based brazing, metal-injection-molding (MIM) debinding and sintering, and ceramic-to-metal joining. Because fixturing mass and part loading strongly influence radiant heat transfer in vacuum, repeatability verification for these furnaces is normally performed with the actual brazing fixtures or sintering setters in place.
Testing Items
A complete repeatability verification program for a horizontal vacuum furnace typically includes the following test items.
- Effective heating zone temperature uniformity at multiple setpoints spanning the production range, for example 650 °C, 850 °C, 1050 °C, and 1150 °C for a hardening furnace, or up to 1300 °C for brazing and sintering applications.
- Uniformity determination under empty-furnace, dummy-load, and production-load conditions.
- Uniformity verification under the pressure regime used in production, either high vacuum (typically 1 × 10⁻³ Pa to 1 × 10⁻¹ Pa) or partial pressure of high-purity nitrogen or argon (typically 100 Pa to 800 Pa).
- Temperature stabilization time from setpoint reach to within-tolerance soak at each survey temperature.
- Run-to-run repeatability of measured per-sensor deviations across consecutive survey runs, expressed as range, standard deviation, and relative standard deviation.
- Type A uncertainty contribution derived from the repeated runs, combined with Type B components from thermocouple calibration, data-acquisition accuracy, and cold-junction compensation.
- Verification of control and over-temperature instrumentation consistency with the survey results, complementing the separate system accuracy test (SAT).
Testing Methods
Instrumentation and Sensor Deployment
The survey uses a multi-channel calibrated data acquisition system with measurement uncertainty typically not exceeding ±0.5 °C, traceable to national measurement standards. Survey thermocouples are selected according to temperature range: calibrated Type K or Type N base-metal thermocouples for surveys up to approximately 1200 °C, and Type S, Type R, or Type B noble-metal thermocouples for higher temperatures. Best practice for repeatability studies is to use sensors from the same calibration batch for all runs, attached to a lightweight, geometrically reproducible sensor rack that fixes each measurement point at the exact corner and center coordinates of the qualified work zone.
Test Conditions and Procedure
Each survey run follows an identical thermal cycle. The furnace is evacuated or backfilled to the specified pressure, ramped to setpoint, and soaked until every survey sensor indicates within a defined stabilization tolerance — commonly 3 °C to 5 °C — of the setpoint. The survey period then begins, during which all channels are recorded at fixed intervals of one to five minutes for at least 30 minutes. If several temperatures are to be qualified, the furnace is stepped from the lowest to the highest setpoint, and a full survey period is recorded at each step. Radiation-dominated high-vacuum conditions and convection-assisted partial-pressure conditions can produce measurably different uniformity behavior, so the pressure regime must be documented and held constant across all repeatability runs.
Repeatability Runs and Statistical Evaluation
Repeatability verification consists of a series of consecutive survey runs — typically three to five — executed without any change to sensor layout, load configuration, setpoints, or pressure. The runs may additionally be distributed across different shifts or days to capture realistic operator and environmental variability. For each sensor position, the deviation from setpoint is averaged across the survey period, and the per-position results are compared across runs. The evaluation determines the range and standard deviation of the run-to-run results at each position and computes the Type A standard uncertainty of the mean in accordance with the GUM. Acceptance is based on criteria such as the following.
- Every individual run must satisfy the specified uniformity band, for example ±6 °C for AMS 2750 Class 2 or ±8 °C for Class 3, with the relevant GB/T 9452 grade applied where required.
- The run-to-run spread of the uniformity index is normally limited to one third of the tolerance band, for example no more than ±2 °C variation between runs for a ±6 °C class furnace.
- The relative standard deviation of repeated measurements at each sensor position should remain within the limit agreed in the test plan, commonly 5% or less of the tolerance band.
- Measured deviations should keep a demonstrable safety margin from the tolerance limit after the expanded measurement uncertainty (coverage factor k = 2) is applied.
Applications
Repeatability verification of horizontal vacuum furnace temperature uniformity is applied wherever thermal-process capability directly affects product quality and regulatory compliance.
- Aerospace heat treatment of structural and engine components, where AMS 2750 pyrometry compliance and furnace-class stability are audit-critical.
- Medical device manufacturing of implant-grade titanium and cobalt-chromium alloys under ISO 13485 quality systems.
- Automotive powertrain and bearing production aligned with IATF 16949 and CQI-9 heat-treatment system assessment requirements.
- Vacuum brazing of heat exchangers, cold plates, and sensor assemblies, where joint quality depends on uniform soak temperatures across complex fixtures.
- Tool and die hardening of high-speed steels and powder-metallurgy tool steels requiring tight, repeatable uniformity at austenitizing temperatures.
- MIM and cemented-carbide sintering, where long high-temperature cycles make run-to-run stability essential for dimensional and metallurgical consistency.
FAQ
How many survey runs are required for repeatability verification?
Three runs are the practical minimum, and four to five runs are recommended for initial qualification, after major repairs, or when qualifying a new load configuration. The number should be defined in the test plan based on the desired statistical confidence and the tolerance band of the furnace class.
Should the repeatability survey be run empty or with a load?
Both conditions have a place. Empty-furnace surveys qualify the intrinsic thermal performance of the hot zone, while dummy-load or production-load surveys qualify the process condition actually used in manufacturing. For repeatability verification, the load state must be identical in every run, because changing the load between runs measures load sensitivity rather than repeatability.
Why do vacuum and partial-pressure surveys give different results?
In high vacuum, heat transfer is almost purely radiative, so shielding geometry and element condition dominate the temperature field. Under partial pressure, convection through the inert gas homogenizes the temperature and typically narrows the spread. Because production may use either regime, repeatability must be verified under the exact pressure condition specified for the process.
What are the common causes of poor repeatability?
Frequent causes include inconsistent sensor positioning between runs, reuse of degraded thermocouples, unstable vacuum or partial-pressure control, heating element or insulation aging, control thermocouple drift, non-reproducible load placement, and insufficient stabilization time before the survey window begins.
What happens if repeatability verification fails?
A failed verification triggers investigation and corrective action: sensor placement and calibration are rechecked, the hot zone and element condition are inspected, the control loop tuning is reviewed, and the survey is repeated. The furnace cannot be released to its uniformity class until all runs of the verification sequence pass the acceptance criteria.
How often should the verification be repeated?
A full repeatability verification is performed at initial qualification, after relocation or major repair of the furnace or its hot zone, after control system changes, and whenever drift is observed in routine surveys. Routine single TUS runs continue at the frequency required by AMS 2750 or GB/T 9452, typically monthly to semi-annually depending on furnace class and sensor type, and these routine records provide ongoing evidence of run-to-run stability.