How to select, position, install and test RTDs and thermocouples in fluid-film journal and thrust bearings.
An embedded RTD or thermocouple reports temperature at its sensing element. The value depends on circumferential or radial position, axial position, depth below the running surface, thermal contact, backing material and response time. A sensor can provide effective alarm and trend protection only when its exact installation and the associated limits are known.
Local heat is generated and conducted into the lining and backing.
The RTD or thermocouple responds at one installed position.
Wire resistance, polarity, shielding and terminations affect the signal.
The input converts, displays, trends, alarms and trips from the signal.
A loop simulation can prove stages 3 and 4 from the injection point onward. It does not prove element accuracy, installation contact or placement in stages 1 and 2.
Both can protect bearings when the complete sensor and monitoring system is correctly specified.
A resistance temperature detector changes electrical resistance with temperature. Platinum Pt100 elements are common in industrial bearing monitoring.
Two dissimilar conductors generate a small voltage related to the temperature difference between sensing and reference junctions.
| Factor | Platinum RTD | Thermocouple |
|---|---|---|
| Primary signal | Resistance in ohms | Small EMF in millivolts |
| Common bearing forms | Pt100; two-, three- or four-wire; single or dual element | Specified letter type such as J, K or T; grounded or ungrounded junction |
| Useful characteristics | Stable, repeatable and well suited to normal bearing-metal temperature ranges | Compact, rugged, responsive and tolerant of a broad temperature range |
| Important errors | Lead resistance, wrong element curve, self-heating, damaged insulation or poor thermal contact | Wrong type, reversed polarity, incorrect extension cable, parasitic junctions, noise or cold-junction error |
| Basic electrical check | Element resistance, lead balance and insulation where applicable | Continuity, polarity and insulation only for an ungrounded junction |
| True sensor calibration | Controlled temperature source plus resistance measurement and reference thermometer | Controlled source plus EMF measurement with defined reference-junction compensation |
Both lead resistances add directly to the indicated value unless compensated elsewhere. Suitable only where the specified error remains acceptable.
The monitor estimates lead resistance, normally assuming the matching leads have similar resistance. Unequal leads introduce residual error.
Kelvin measurement largely removes lead resistance from the result when connected to a compatible four-wire input.
A Pt100 has nominal resistance of 100 Ω at 0°C. An IEC 60751 element with the common 0.00385 coefficient is approximately 109.7 Ω at 25°C, before two-wire lead resistance. Confirm the actual element curve and measured temperature.
The approved OEM or bearing drawing takes precedence. These locations explain common practice, not universal drilling dimensions.
Place the element in the predicted high-temperature loaded region. Where shaft or housing misalignment can cause edge loading, sensors near both axial ends can reveal a temperature split that one centreline sensor would miss.
A common location is in the loaded pad toward the trailing-edge high-temperature region, often around 75% of pad arc from its leading edge. Axial centreline suits some self-aligning pads; edge sensors may be required for non-aligning pads.
For unidirectional service, 75/75 commonly means 75% of circumferential length from the leading edge and 75% of radial width outward from the inner diameter. Bidirectional bearings require a different position defined by the drawing.
Depth: a deeper sensor usually reads lower and responds more slowly. Field sensors are commonly fitted in the backing just beneath the lining; laboratory research may embed very small elements closer to the working surface.
Coverage: one sensor reports one location. Multiple pads, dual elements or axial pairs may be needed to detect unequal load sharing, reverse rotation or edge loading.
Rotation: leading and trailing edges reverse when shaft direction changes. Offset pivots, oil feeds and sensor positions must be reviewed for every permitted direction.
Machine the diameter, depth and flat-bottom geometry specified for the sensor case. Spring loading, a retaining ring or qualified potting can maintain contact and permit replacement.
Do not distort or break through the finished lining. Remove burrs and swarf, then verify the remaining depth to the running surface against the drawing.
Use controlled grooves and bend radii. Keep leads clear of pads, pivots, split faces and rotating parts, and prevent fretting against sharp housing edges.
Prevent lubricant tracking along leads or braid through the housing feedthrough. The sealing method must suit oil, temperature, pressure and hazardous-area requirements.
Specify grounded or isolated construction. Insulated bearings may require electrically isolated sensors so instrumentation does not create an uncontrolled shaft-current path.
Mark bearing side, pad number, element A/B, sensor type, wire configuration, location and rotation reference at both bearing and panel terminations.
RTD material, nominal resistance, coefficient and accuracy class; or thermocouple type, tolerance and grounded/ungrounded junction.
Two-, three- or four-wire RTD; single or dual element; thermocouple polarity; shield termination and isolation requirements.
Pad or shell identity, circumferential/radial and axial coordinates, depth, rotation direction and drawing datum.
Insulation, braid, cable length, bend radius, oil blocking, extension cable, feedthrough, connector and terminal box.
Temperature, vibration, oil chemistry, ingress, EMC, electrical isolation and hazardous-area certification.
Continuity, insulation, calibration or response test, documentation, channel identification and as-built resistance values.
Only trained personnel should test protection circuits. Follow the machine isolation procedure and prevent an unintended trip, start or loss of protection.
Disconnect the sensor from monitoring electronics before resistance or insulation measurements. Do not apply a high-voltage insulation tester unless the sensor, cable, barriers and connected equipment are specifically approved for that test.
Continuity, resistance balance, insulation and visual condition find opens, shorts and damaged leads. They do not prove temperature accuracy.
Controlled heating with a reference checks that the element responds correctly. An installed bearing may have gradients that limit calibration accuracy.
A calibrated resistance or millivolt source checks transmitter scaling, display, alarms, trips and logic. It bypasses the physical sensor.
A documented test combines sensor evidence, wiring checks, channel identity, monitor response and alarm/trip action under controlled conditions.
| Observed symptom | Possible sensor or circuit cause | Mechanical condition to exclude |
|---|---|---|
| Sudden open-circuit or upscale indication | Broken element, lead or terminal; input burnout direction | Confirm no actual rapid over-temperature event occurred first |
| Implausibly low or fixed reading | Shorted leads, wrong input type, failed transmitter or poor thermal contact | Low load, cold oil, unloaded pad or changed operating condition |
| One channel consistently offset | Wrong RTD curve, lead resistance, thermocouple type/polarity, junction or calibration error | Real pad load imbalance, misalignment, oil distribution or sensor-position difference |
| Noisy or intermittent trend | Loose terminal, damaged braid, ground loop, EMI, fretting lead or moisture ingress | Actual vibration, pad flutter, unstable oil supply or transient load |
| Axial pair diverges | Different depths, swapped channels or damaged element | Edge loading, shaft slope, housing distortion or uneven clearance |
| All pads rise together | Common reference, input or scaling fault | High inlet temperature, low flow, viscosity loss, speed or load change |
Use the bearing temperature limits guide to distinguish supply oil, drain oil, embedded metal and predicted film temperatures before changing alarm settings.
Not universally. Pt100 RTDs offer stable, repeatable measurement; thermocouples are compact and rugged. The input system, response, wiring, vibration, certification and redundancy requirements decide.
It depends on bearing geometry, load, speed, oil supply and distortion. The maximum film or surface temperature may not occur at the embedded sensor. Use thermal analysis and approved drawing locations.
For a unidirectional thrust pad it conventionally identifies 75% of pad length from the leading edge and 75% of radial width outward from the inner diameter. State the datums and rotation on the drawing.
A room-temperature resistance check can identify an open, short or gross error. Calibration requires known temperatures, a reference thermometer and comparison with the correct resistance curve.
Not as an isolated element-to-case circuit: its junction is intentionally grounded. Confirm construction before interpreting continuity to the case.
No. It checks the circuit from the injection point onward. It cannot prove that the sensor is at the correct depth, on the correct pad or in good thermal contact.
Sensor definitions and testing were cross-checked against IEC 60751:2022 for industrial platinum resistance sensors, IEC 60584-1:2013 for thermocouple EMF and tolerances, Minco's bearing temperature sensor replacement criteria and spring-and-ring installation guidance, and Fluke's RTD calibration procedure. Placement was cross-checked against Kingsbury guidance on fluid-film bearing temperature measurement and tilting-pad temperature patterns. The bearing drawing, sensor manufacturer and machinery protection specification govern each installation.
Send the bearing drawing, rotation, load cases, sensor specification and monitoring-system details for an instrumentation review.