Temperature measurement in vacuum depends on the complete thermal path between sensor, surface, leads and surrounding radiation. With negligible gas coupling, a sensor with poor contact can report a value influenced by its wires and field of view rather than the intended surface. Sensor selection, mounting, channel planning and calibration must therefore be treated as one measurement system.
Why vacuum changes the measurement
At ambient pressure, surrounding gas provides an additional thermal path between a sensor and its environment. As pressure falls into the molecular-flow regime, that path becomes small and the sensor temperature is governed mainly by conduction through the contact point, conduction along the wires and radiation exchange with surrounding surfaces. Poor contact can therefore produce a plausible but biased reading that reflects the test item, harness and shroud in an unknown proportion.
The magnitude of the error is installation-specific. Even a calibrated sensor can produce misleading data if the junction is not well coupled, if the leads are not heat-sunk or if radiation from the shroud or heaters dominates. Mounting should be verified before closing the chamber and documented in the as-built configuration.
Thermocouples: Type T and Type K
Thermocouples are widely used where many compact channels are required. Type T is commonly selected for low and moderate TVAC temperatures; Type K covers a wider upper-temperature range. The usable range and uncertainty depend on wire type, tolerance class, reference junction, connectors, acquisition hardware and calibration. IEC and NIST data should be used for the selected sensor system rather than one generic accuracy value.
- Select thermocouple type and tolerance class from the full test range and required uncertainty.
- Keep compatible thermocouple materials through connectors and extension leads to the compensated input.
- Control reference-junction conditions and include acquisition-channel offsets in the uncertainty budget.
- Calibrate representative complete chains when qualification data or model correlation requires traceability.
PT100 and PT1000: precise reference points
Platinum resistance thermometers provide stable reference and control measurements when installed and read correctly. Four-wire measurement removes lead-resistance error and is normally preferred for traceable work. Sensor class, self-heating, wiring, readout current and calibration determine the resulting uncertainty.
- Use four-wire PT100 connections where lead resistance and traceability matter.
- Select PT100 or PT1000 from the readout architecture, range and uncertainty requirement.
- Limit excitation current and verify self-heating under the actual vacuum mounting condition.
- Use resistance sensors for control or reference points and thermocouples for broader mapping when that mix fits the test plan.
Mounting: contact quality decides everything
In vacuum, the mounting method is the measurement. The junction or element must be bonded or clamped to the surface with real conductive contact, the wires must be heat-sunk along an isothermal path, and the sensing point should be shielded from radiation that does not represent the surface being measured.
Proven mounting practice
- Use a vacuum-compatible attachment method whose thermal contact, cure state and contamination performance are approved for the project.
- Where adhesive is not permitted, use clamps, thermal washers or spring-loaded contacts with defined force.
- Heat-sink the first several centimetres of wire on the same surface, so the harness does not conduct heat into the junction.
- Cover exposed junctions with aluminium tape or foil against shroud and lamp radiation, without insulating them from the surface.
- Strain-relieve harnesses so pump-down movement and thermal cycling do not peel sensors off.
Placement, channel counts and feedthroughs
Sensor placement should follow the thermal model, control concept and safety analysis. Channel count ranges from a few points on a small assembly to hundreds on an instrumented spacecraft; it is not determined by article class alone. Feedthrough and acquisition capacity should be fixed after the channel list and redundancy concept are reviewed.
- Define the channel list early: sensor type, location, purpose (control, limit, model correlation) and connector pinout.
- Reserve spare channels and spare feedthrough pins – sensors fail, and opening the chamber to fix one costs a pump-down.
- Duplicate sensors at control and safety-critical points so a single failure does not end the test.
- Keep chamber-side and article-side harnessing separable at a defined connector plane to speed up integration.
Calibration, traceability and error sources
Qualification data is only defensible if the measurement chain is traceable: sensors, reference junctions and acquisition channels calibrated against references that link back to national standards, with records of when and how. A system-level calibration – sensor plus harness plus channel – catches offsets that individual component certificates miss.
Common error sources
- Poor contact: the dominant error in vacuum, often worth several kelvin.
- Harness conduction: wires that are not heat-sunk drag the junction toward harness temperature.
- Self-heating of resistance sensors from excitation current that cannot dissipate.
- Reference-junction and channel offsets from uncalibrated acquisition hardware.
- Ground loops and interference from heaters or motor drives corrupting millivolt signals.
Takeaway
Reliable TVAC temperature data depends on the complete measurement chain: sensor type, mounting, wire heat sinking, acquisition channel and traceable calibration. Channel planning should therefore be completed together with the thermal model and test procedure.
Frequently Asked Questions
Use the sensor types that fit the measurement role. Thermocouples can provide compact, economical mapping channels, while four-wire platinum resistance sensors suit stable control and reference points. The final choice depends on range, uncertainty, mounting, channel count and calibration requirements.
Mounting is a frequent cause, but not the only one. Poor contact, lead conduction, radiation, self-heating, reference-junction error and acquisition offsets can all bias the result. Diagnose the full measurement chain before changing the sensor type.
The channel count comes from the thermal model, control and limit points, redundancy and test objectives. It can range from a small set on a component to hundreds on a spacecraft. Plan feedthrough pins, acquisition channels and spares from the approved channel list.
Bond the junction with low-outgassing, thermally conductive adhesive, or clamp it with defined force using thermal washers where adhesive is prohibited. Heat-sink the first centimetres of wire on the same surface, cover the junction against radiation and strain-relieve the harness. Verify contact quality before closing the chamber – rework afterwards costs a pump-down.
Technical references
Primary standards, agency material and current Deepvac product pages used to verify the technical statements in this article.
- Thermocouple Calibration ServicesNational Institute of Standards and Technology (NIST)
- Thermometry Calibration ServicesNational Institute of Standards and Technology (NIST)
- IEC 60751:2022: Industrial platinum resistance thermometers and platinum temperature sensorsInternational Electrotechnical Commission (IEC)
- ECSS-E-ST-10-03C Rev.1: TestingEuropean Cooperation for Space Standardization (ECSS)