A heat-treat shop can run perfect thermal cycles for years, with temperature uniformity surveys (TUS) that pass every time. Yet, a Nadcap auditor can still issue a major nonconformance that puts the shop's merit status at risk. This happens frequently, and the root cause is rarely the furnace itself. The finding often comes down to something far more subtle: a thermocouple log that can't prove a sensor was within its usage limits, or a calibration certificate that lacks an unbroken chain of traceability to NIST .
Compliance with AMS 2750, the foundational specification for pyrometry, is a discipline of instrumentation, testing, and, above all, documentation. Failures are almost always procedural. They happen when teams misunderstand the specification's detailed requirements for how temperature is measured, controlled, and verified.
With the release of AMS 2750 H in June 2023, several of these requirements have become more stringent. Shops still operating under the assumptions of Revision G are exposed to new audit risks. This guide provides a practitioner's walkthrough of the requirements that matter most for staying audit-ready. We will cover what AMS 2750 governs, what changed from Rev G to H, and the critical details of furnace classification, instrumentation types, TUS, SAT, thermocouple calibration, and record retention.
AMS 2750 is the SAE International specification that defines the pyrometric requirements for thermal processing equipment used in the heat treatment of metallic materials. In simple terms, it governs how temperature is measured, controlled, recorded, and verified. It is the rulebook for ensuring the temperature you think you have in your furnace is the temperature you actually have and that you can prove it.
The most important distinction for any quality manager to understand is the specification's scope. AMS 2750 governs pyrometry , not the heat-treat process itself. It ensures the integrity of the temperature measurement system. The actual time-at-temperature recipes and material-specific requirements come from separate process specifications, such as:
When a customer questionnaire asks if you are "AMS 2750 compliant," they are asking about your pyrometry program. Confusing this with process compliance is a common source of audit findings.
Maintained by SAE International, the specification is currently on Revision H. While its origins are in aerospace, its requirements are now flowed down by prime contractors across defense, medical device, and energy sectors. It is also often used alongside the AIAG's CQI-9 standard in the automotive industry.
Read more: Heat Treat Calibration Compliance | BAC 5602 & CQI-9 | ISO/IEC 17025
AMS 2750 H officially replaced Revision G in June 2023. For any organization audited by Nadcap or serving customers who require pyrometry compliance, operating under Rev H is not optional. The transition period has passed, and auditors now expect to see procedures and records that align with the current standard.
For a quality manager or calibration coordinator, the critical question is: what actually changed, and where are the new compliance gaps? While the revision includes dozens of clarifications, a few key changes have a direct operational impact on most heat-treat shops.
| Requirement Area | AMS 2750 G (Previous) | AMS 2750 H (Current) | Practical Impact |
|---|---|---|---|
| Expendable Thermocouples | Allowed limited reuse with some ambiguity on tracking. | More explicit limits on reuse. Requires documentation of each thermal cycle or time at temperature for each use. | Shops must implement rigorous tracking for every expendable sensor used more than once. Simple logs are no longer sufficient. |
| Correction Factors | Allowed for SAT and instrument calibration with broader limits. | Tighter limits on the magnitude of allowable correction factors. Requires more detailed documentation for their application. | Procedures that relied on large offsets to pass SAT may now be noncompliant. More instruments may require recalibration or replacement. |
| System Accuracy Tests (SAT) | Defined procedures and intervals. | Clarified requirements for testing all relevant instrument channels and documenting the test setup more thoroughly. | SAT procedures must be reviewed to ensure every required channel (e.g., recording, over-temp) is being tested, not just the control loop. |
| TUS Sensor Placement | Required sensor placement to monitor the work zone. | Requires documented rationale for sensor placement, justifying why the chosen locations represent the extremes of the work volume. | Shops must now be able to defend their TUS sensor map with a written justification, which auditors will review. |
| Digital Data Integrity | General requirements for record retention. | Introduced more explicit language around electronic record security, audit trails, and preventing data alteration. | Unsecured spreadsheets or simple scanned PDFs for TUS/SAT records may no longer be sufficient. Systems need to show data integrity. |
The transition from G to H is fundamentally about closing loopholes and increasing the burden of proof. Practices that may have passed an audit under Rev G due to ambiguity in the specification are now explicitly defined and noncompliant if not followed.
AMS 2750 H tightens five critical requirement areas over Rev G.
AMS 2750 defines six furnace classes, each corresponding to a specific temperature uniformity tolerance that must be maintained within the qualified work zone. This tolerance is the allowable temperature variation measured during a Temperature Uniformity Survey.
| Furnace Class | Temperature Uniformity Tolerance |
|---|---|
| Class 1 | ±5°F (±3°C) |
| Class 2 | ±10°F (±6°C) |
| Class 3 | ±15°F (±8°C) |
| Class 4 | ±20°F (±11°C) |
| Class 5 | ±25°F (±14°C) |
| Class 6 | ±50°F (±28°C) |
A common mistake is to view furnace class as a label you choose for convenience. In reality, the required class is determined by the material process specification for the parts you are heat treating. You cannot process Class 2 work in a furnace that has only been qualified to Class 3 standards.
Consider this scenario: a shop processes a batch of aerospace components requiring a Class 2 (±10°F) uniformity. The furnace they use has only been surveyed and qualified as Class 3 (±15°F). Even if the heat-treat cycle runs perfectly according to the recipe, every part in that batch is technically nonconforming. The pyrometry qualification of the equipment does not support the precision required by the process. This is a classic audit finding and can lead to costly customer rejections or scrapped material. Verifying the required furnace class from the customer flowdown or process specification is the first step in ensuring pyrometry compliance.
AMS 2750 requirements define six furnace classes by uniformity tolerance.
Alongside furnace class, AMS 2750 defines five instrumentation types (A, B, C, D, and E). These types specify the combination of control, recording, and monitoring sensors that must be installed on the thermal processing equipment. The instrumentation types are not about accuracy; they define what hardware must be present in the measurement chain.
| Sensor/Instrument | Type A | Type B | Type C | Type D | Type E |
|---|---|---|---|---|---|
| Control Instrument | Required | Required | Required | Required | N/A |
| Control Thermocouple | Required | Required | Required | Required | N/A |
| Recording Instrument | Required | Required | Required | Required | Required |
| Hottest & Coldest Load TCs | Required | N/A | N/A | N/A | N/A |
| Recording Thermocouple | N/A | Required | Required | N/A | Required |
| Load Thermocouple(s) | Required | Required | Optional | Optional | N/A |
| Over-temperature Instrument | Optional | Optional | Optional | Optional | N/A |
Like furnace class, the required instrumentation type is not a choice. It is dictated by the process specification, customer contract, or Nadcap accreditation scope. A frequent error is assuming a higher type is "better." For instance, Type A provides the most data by requiring sensors at the hottest and coldest locations of the load, but it also carries the highest calibration and maintenance burden. A shop that commits to Type A without understanding this workload can easily fall out of compliance on sensor calibration intervals.
Imagine a shop operating with Type C instrumentation discovers during a Nadcap audit that a key customer's process specification actually requires Type B. This means they need additional load thermocouples connected to their recording instrument that they do not have installed. The finding is immediate, and production for that customer halts until the furnace can be retrofitted and re-qualified. Verifying the required instrumentation type against all customer flowdowns is a critical risk-reduction step.
A Temperature Uniformity Survey (TUS) is a periodic physical test that maps the temperature variations inside a furnace's qualified work zone. It is the primary evidence that a furnace is capable of heating parts uniformly within the tolerance defined by its furnace class.
Under AMS 2750 H, TUS frequency is not fixed; it depends on the furnace class and its historical performance. A furnace with a consistent history of passing surveys may qualify for extended intervals (e.g., annually for Class 1-2), while a furnace with a recent failure may be required to perform surveys more frequently (e.g., quarterly).
The TUS also establishes the furnace's qualified operating range . This is the temperature range over which the furnace has been proven to perform. Processing parts at a temperature outside this documented range without conducting a new TUS is a clear nonconformance. Furthermore, AMS 2750 H requires a documented rationale for the placement of survey thermocouples, forcing shops to justify that their sensor map truly represents the extremes of the work volume.
And this, of course, always seems to happen on a Friday afternoon: a TUS is run, and the data shows one corner of the furnace is running 5°F outside the soak band. The furnace is now noncompliant at that temperature. What happens to the parts processed in that furnace all week? They are now suspect and must be quarantined pending disposition. This is why a TUS is not just a checkbox; it is the core validation of your process capability.
A TUS failure is not a maintenance event it is a quality event with product-impact implications. A failed survey immediately triggers a nonconformance process. The furnace must be taken out of service for processing at the failed temperature range.
The quality team must then:
Nadcap auditors look specifically for the documentation trail proving this process was followed. A missing TUS failure disposition is one of the most common and serious pyrometry-related findings.
A TUS failure triggers a mandatory four-step quality disposition process.
While a TUS verifies the furnace heats uniformly, a System Accuracy Test (SAT) verifies that the instruments reading and recording the temperature are accurate. The two tests are often confused but serve fundamentally different purposes. A passing SAT does not guarantee a passing TUS, or vice versa.
During a SAT, a calibrated field test instrument with a test thermocouple is placed next to the furnace's resident control or load thermocouple. The readings from the furnace's installed instrumentation (control and recording instruments) are then compared to the reading from the independent test instrument.
The acceptance criteria are defined by AMS 2750 and depend on the instrumentation type. The difference between the installed instrument and the test instrument must be within ±2°F (±1.1°C) or the published accuracy of the thermocouple, whichever is greater.
When the difference, or SAT offset , is within limits but not zero, it must be documented. For example, a recording instrument that reads 3°F higher than the field test instrument may be within the allowable tolerance. Under Rev H, the rules for applying correction factors to compensate for these offsets have been tightened. A shop that previously relied on large correction factors to pass a SAT may now need to recalibrate or replace the instrument. The decision to apply a correction factor versus taking other action is a critical judgment call for the calibration coordinator.
Thermocouple management is arguably the single greatest source of AMS 2750-related Nadcap findings. The issues are rarely about having thermocouples; they are about proving those thermocouples were calibrated, used within their limits, and replaced on schedule. During a Nadcap pre-audit review I once conducted, the shop's log for expendable sensors showed an initial calibration date but no usage-hour tracking and no documented pull date. They had no way to prove compliance with AMS 2750's maximum use-life requirements. The auditor wrote four findings on thermocouple management alone, and the shop's merit status was suspended.
AMS 2750 specifies requirements for two main categories:
The practical problem most shops underestimate is Type K drift . A new, calibrated Type K thermocouple may be perfectly accurate, but after dozens of thermal cycles, its output can shift significantly. A sensor that reads correctly during a SAT can be out of tolerance a month later. This is why AMS 2750 H emphasizes usage limits and tracking so heavily. Without a log proving a thermocouple was within its documented life during a production run, an auditor can invalidate that entire run.
Ultimately, AMS 2750 compliance is a documentation discipline. A shop can have the best furnaces and a perfect testing record, but if the documentation is incomplete, inaccessible, or lacks traceability, the audit finding is the same as if the tests were never performed. We've all seen it: a perfect set of calibration records that falls apart under one simple question from the auditor.
AMS 2750 H requires the retention of specific records, typically for a minimum of five years plus the current year. Auditors will request:
A prime contractor flowdown of AMS 2750 H puts Tier 2 and Tier 3 suppliers under scrutiny, and these suppliers often face the steepest learning curve. A common finding is a shop presenting a valid calibration certificate, only for the auditor to discover the calibrating laboratory's ISO/IEC 17025 accreditation had lapsed at the time of service. This breaks the traceability chain and invalidates the certificate. Your calibration records are only as good as the verifiable accreditation of the lab that issued them.
AMS 2750 compliance hinges on the accuracy, traceability, and documentation of your pyrometry instrumentation. Most nonconformances originate not from furnace hardware but from gaps in the calibration and record-keeping chain. This creates a clear need for a calibration partner that understands the specific demands of heat-treat pyrometry, not just general instrument service.
As an ISO/IEC 17025-accredited calibration provider, CTPM has deep experience supporting heat-treat compliance programs across the Midwest. Our temperature calibration capabilities and consultative approach help you verify that your records, thermocouple management, and traceability documentation meet AMS 2750 H requirements before an auditor arrives.
With both lab-based and on-site calibration services , we work with your team to maintain the unbroken chain of traceability that Nadcap auditors demand, minimizing downtime and reducing audit risk. We function as an extension of your quality team, ensuring your measurement systems are audit-ready.
Talk to CTPM about your heat-treat calibration and compliance needs.
Compliance with AMS 2750 is a continuous discipline, not a one-time furnace qualification event. The specification defines in rigorous detail what must be measured, how accurately, how often, and with what documented evidence. The release of Revision H has only raised the bar, demanding greater rigor in sensor management, correction factor application, and record traceability.
Shops that treat pyrometry as an ongoing quality program diligently managing their instrumentation, testing, and documentation are the ones that pass audits without findings. They are the ones who can confidently prove the integrity of their thermal processes, keeping production moving and customers satisfied.
Is AMS 2750 H mandatory for all heat treat shops or only Nadcap-accredited facilities?
AMS 2750 H is mandatory wherever a customer contract, process specification, or purchase order references it which includes most aerospace and defense work regardless of Nadcap status. Nadcap-accredited shops are audited directly against it, but non-Nadcap shops serving aerospace primes are typically required to comply through flowdown requirements in their contracts.
Can correction factors still be applied to thermocouples under AMS 2750 H?
Yes, but Rev H tightened the limits on allowable correction factors and added documentation requirements for how and when they are applied. Shops that previously relied on large correction factors to bring instruments into SAT compliance may find that their existing practices no longer meet the specification. Review the correction factor limits for your instrumentation type before your next SAT.
How does AMS 2750 interact with CQI-9 for shops serving both aerospace and automotive customers?
AMS 2750 and CQI-9, the AIAG heat treat system assessment, overlap in areas like furnace temperature uniformity and instrument calibration, but they are not interchangeable. A shop compliant with AMS 2750 H is not automatically CQI-9 compliant, and vice versa. Shops serving both sectors must map the requirements of each standard to their pyrometry program and identify where the more stringent requirement applies.
Does AMS 2750 H apply differently to vacuum furnaces versus atmosphere furnaces?
The core pyrometry requirements instrumentation types, TUS, SAT, thermocouple calibration apply to both. However, vacuum furnaces present unique challenges for TUS sensor placement and thermocouple routing that require specific documentation of the survey setup. The specification does not exempt vacuum furnaces from any major requirement, though some details may be handled differently in practice.
What are the most common Nadcap findings related to AMS 2750 noncompliance?
The most frequently cited findings involve thermocouple usage beyond documented life limits, incomplete or missing TUS failure disposition records, SAT correction factors applied outside allowable limits, calibration certificates lacking traceable accreditation verification, and insufficient documentation of sensor placement rationale during TUS. Most of these are documentation and tracking failures, not equipment failures.