A strain gauge sensor works by letting a machined body deform and reading that deformation electrically. Bonded foil gauges bridge across the strained metal, a Wheatstone circuit turns their resistance change into millivolts, and a certificate converts millivolts into engineering units — inch-pounds, in the case of a reaction torque sensor.
Every sensor described so far in this catalogue has one job and treats everything else as contamination: a torque sensor is supposed to ignore thrust, a load cell is supposed to ignore bending. It is worth being clear about how that ignoring is actually achieved, because the answer opens a door.
The metal does not ignore anything. A body subjected to torque, thrust, shear and bending at the same time strains in response to all of it, everywhere. What makes a single-axis sensor single-axis is not the material but the arrangement — gauges bonded at locations chosen so that the strain pattern from the wanted quantity adds up around the bridge while the patterns from the unwanted ones cancel. Rejection is a wiring decision.
Which means the discarded information is still there. Add a second, independently wired bridge to the same body, positioned to add up a different strain pattern, and the quantity you were rejecting becomes a second output. That is what a multi-axis sensor is: one spring element carrying several electrically isolated bridge circuits, each dedicated to one component and each producing its own signal.
On this series the second quantity is thrust. Torque about the axis and force along it, measured simultaneously by the same part. That combination is not arbitrary — a great many processes produce both at once. Driving a screw generates torque and axial feed. So does a drill, a press-fit, a valve actuator, a mixer working through a viscous batch, a rotating seal being run in. Measuring only the torque tells you half of what the process is doing.
The honest limitation is cross-talk. Apply a load precisely along one axis and you would like a signal from that channel alone. In practice a small signal appears on the others too, because spring elements have manufacturing tolerances and gauge sensitivities vary slightly. Published guidance on multi-axis sensors describes exactly this, and the important subtlety is how the specification is written: cross-talk is quoted as a percentage of the disturbed channel's full scale, not as a percentage of the reading. On a sensor whose two axes have very different capacities, a fraction of a percent on the larger channel can be a substantial absolute number, and it does not shrink when the load is small.
Sensors of this kind are used in industrial machinery and assembly verification, automotive engine, transmission and component testing, aerospace systems and propulsion work, material torsional characterisation, robotics and automation feedback, calibration and metrology, turbine and generator monitoring in power generation, medical device assembly, quality control and teaching.
Torque measurement here works exactly as on any reaction sensor, so the interesting material is elsewhere: two build options that change what the sensor fundamentally is, neither of which can be added after the fact.
Signal. Rated torque brings back 2 mV/V from a 350 ohm bridge on 10 VDC — call it twenty millivolts at the top of the range. Compensation and balance components sit behind the four conductors, and a ten foot mating cable arrives with the sensor.
A thrust bridge is a second sensor sharing one body. Specify it and the part arrives with an additional, electrically separate bridge measuring axial force. Two consequences follow immediately. It needs its own excitation and its own conditioning channel — two bridges means two measuring chains, not one instrument with a switch. And it has to be built in: the gauges are bonded and trimmed during manufacture, so a two-axis sensor is a decision made at the point of order rather than an upgrade fitted later.
Gauges can be relocated to the outside diameter. The standard arrangement places the sensing where the geometry suits a solid installation. Where a shaft, a spindle or a drawbar has to pass through the middle of the sensor, an option moves the gauge installation to the outside diameter instead, leaving the centre clear. That is a genuinely different mechanical proposition rather than a variant — it turns a solid reaction sensor into one that can be threaded onto an existing rotating assembly — and like the thrust bridge it is built rather than retrofitted.
The mechanical limits on this series are unusually comfortable. Maximum overhung moment, shear and thrust are published per model, and two things stand out reading down them. The shear and thrust figures are equal to one another on every model, which is not true of our smaller reaction sensors, where axial load is tolerated far better than side load. And both are large in absolute terms — thousands of pounds — because these are substantial flanged bodies rather than compact blocks. Fixtures that would be marginal on a small sensor are usually untroubling here, though the moment limit still sits at around half the torque rating on most models and deserves the same weigh-and-measure check as anywhere else.
Aluminium, and lighter than people expect. The bodies are anodized aluminium throughout, including at the top of the range, and the published weights run from well under a pound to a little over seven. For a sensor rated at a hundred thousand inch-pounds that is a remarkably light object, which matters when it is being fitted to something overhead or handled during a test setup.
Most of the decisions on this series are made once, at the point of order, and cannot be revisited afterwards. That is worth knowing before a purchase order is raised, and it is why a conversation with our application engineers is time well spent here.
Decide whether you need one axis or two before you order. Ask what your process actually does. If it produces torque and axial force together — driving, drilling, pressing, actuating, mixing — and you currently infer one from the other, a two-axis sensor replaces an assumption with a measurement. If your application is genuinely pure torque, a single-axis sensor is simpler, cheaper and has one less thing to condition. What you cannot do is defer the question, because the thrust bridge is installed during manufacture.
Then work out whether anything has to pass through the middle. A shaft, a drawbar, a hydraulic line or a spindle running through the sensor changes the configuration required, and the option that relocates the gauges to the outside diameter exists precisely for that. It is another build-time decision, so establish the mechanical arrangement before ordering rather than discovering the constraint during installation.
Size torque against the credible peak. Every accuracy figure is a proportion of rated output, so capacity chosen far above the working range costs resolution you could have had. Set against that, safe overload is a survival limit rather than an allowance, and rotating machinery produces starting and stalling torques well above its running figure. Establish the worst case the drive can actually generate, not the one the process is meant to see.
Check the fixture even though the limits are generous. Weigh what attaches and measure how far its centre of mass sits from the mounting face; the weight goes against the shear figure and weight times distance against the overhung moment. The margins on this series are comfortable, but a long or heavy fixture on a large sensor can still reach the moment limit, which sits at roughly half the torque rating on most models.
Plan two measuring chains if you specify two axes. Each bridge needs excitation and conditioning of its own, and the two channels should be labelled unambiguously at the instrument as well as at the sensor. It is also worth deciding early how the two will be recorded together, since the value of a two-axis measurement is largely in the relationship between the channels over time rather than in either number alone.
Then choose instrumentation to match. Budget one amplifier signal conditioner module per bridge if a controller or data system is consuming the output; where a person is watching the number or a limit has to trip something, a digital display is the better fit. Where several sensors are in rotation, Cal-Teds plug and play stores each unit's calibration data on an IEEE 1451.4 memory chip so an instrument picks it up on connection.
Describe the process, whether it produces thrust as well as torque, and whether anything needs to pass through the sensor, and we will help you specify the right configuration first time. Standard capacities are stocked, and there is a discount for academic and research purchases.
TRS Series Torque Sensor Applications.
The Transducer Techniques TRS Series reaction torque sensors are designed to provide precise torque measurements in a wide range of applications, offering extended operational life through a design with no rotating components.
- Industrial Machinery: TRS Series sensors verify component torque during manufacturing equipment assembly, helping maintain quality.
- Automotive Testing: These sensors measure engine, transmission and vehicle component torque to confirm proper performance.
- Aerospace Testing: TRS Series sensors are used in aircraft systems and component testing, including control surfaces and propulsion verification.
- Material Testing: These sensors characterize torsional properties of materials, including stiffness, elasticity and strength.
- Robotics and Automation: TRS Series sensors provide torque feedback in robotic systems for controlled gripping and fastening operations.
- Calibration and Metrology: These sensors serve as reference standards for calibrating other torque measurement instruments.
- Educational and Research Laboratories: TRS Series sensors are used for teaching and investigation in torque measurement and mechanical system studies.
- Energy and Power Generation: These sensors monitor turbine and generator torque to help assure operational efficiency.
- Medical Device Testing: TRS Series sensors verify component assembly, supporting safety and reliability standards.
- Quality Control: These sensors verify fasteners and components across electronics and consumer goods manufacturing.
Their design eliminates the need for high-maintenance components like slip rings and bearings, reducing operational costs and ensuring consistent and dependable torque measurements.
Frequently Asked Questions
What torque range does the TRS Series cover?
Eight capacities in inch-pounds: 500, 1,000, 2,000, 5,000, 10,000, 20,000, 50,000 and 100,000 in-lb. That top figure is over eight thousand foot-pounds, which puts the upper end of this series into heavy machinery, turbine and large drivetrain work. The range is built on three body diameters, so check which one your capacity falls into before designing a mounting — the bolt pattern and overall size step up with it.
What does it mean that the TRS is available as a two-axis sensor?
It means the same body can carry a second, electrically separate bridge that measures axial thrust alongside the torque. One part, two independent measurements: how hard something is being twisted and how hard it is being pushed or pulled along the same axis. That matters because a great many processes produce both together — driving a screw, drilling, pressing a fit, actuating a valve, running a seal in — and measuring only the torque leaves you inferring the other half rather than knowing it.
What is the thrust bridge option?
It is the second measuring circuit that makes the sensor two-axis. The gauges for it are bonded and trimmed during manufacture, so it is specified when the sensor is ordered rather than added afterwards. Because the two bridges are independent, each needs its own excitation and its own conditioning channel. Torque and thrust combinations vary a great deal between applications, so it is worth telling us what the process does rather than choosing from a table — the useful ratio between the two capacities depends entirely on your job.
What is cross-talk, and how should I read a two-axis specification?
Cross-talk is the small signal that appears on one channel when the load is applied to the other. It exists because spring elements have manufacturing tolerances and gauge sensitivities vary slightly, so the separation between axes is very good rather than perfect. The subtlety worth knowing is how it is specified: cross-talk is normally quoted as a percentage of the disturbed channel's full scale, not of the reading. On a sensor whose two axes have very different capacities, a fraction of a percent on the larger channel can be a meaningful absolute error, and it does not shrink as the applied load falls.
Do I need two instruments for a two-axis sensor?
Two measuring chains, yes. Each bridge needs excitation and conditioning of its own, so a two-axis sensor is wired and read as two sensors that happen to share a body. In practice that means either two conditioner modules, or a data acquisition arrangement with two bridge inputs. It is worth labelling the channels unambiguously at both ends, and worth deciding at the outset how the two will be logged together — most of the value in a two-axis measurement lies in the relationship between the channels rather than in either one alone.
Can the thrust bridge be added to a sensor I already own?
No. The additional gauges are installed on the body and trimmed as part of manufacture, and the sensor is then calibrated as a two-axis unit. Fitting them afterwards would mean rebuilding and recalibrating the sensor, which is not a field operation. The practical consequence is that this is a decision to make properly at the point of order. If there is any real chance your programme will want thrust data later, it is worth discussing now, because the cost of specifying it up front is far lower than the cost of discovering the need with a single-axis sensor installed.
What is the option that relocates the gauges to the outside diameter?
It is for through-hole applications — installations where a shaft, spindle, drawbar or hydraulic line has to pass through the middle of the sensor. Moving the gauge installation to the outside diameter leaves the centre clear so the sensor can be built around an existing assembly rather than needing to sit at the end of one. Like the thrust bridge it is a build option rather than a retrofit, so establish the mechanical arrangement before ordering. Tell us what has to pass through and how large it is and we will confirm what is possible.
How accurate is the TRS Series?
Nonlinearity and hysteresis are each 0.1% of rated output, nonrepeatability 0.05%, and zero balance 1.0%, with safe overload at 150%. Those describe the torque measurement under proper conditions. On a two-axis sensor there is a second figure to ask about, which is the cross-talk between channels — it is a separate specification from accuracy and, depending on the ratio between your two capacities, it can matter more. Ask us for the cross-talk figure applicable to the configuration you are considering.
What are the mechanical limits on the TRS?
Maximum overhung moment, shear and thrust are published per model, and they are comfortable by the standards of smaller reaction sensors — shear and thrust run into the thousands of pounds. Two observations. On this series the shear and thrust figures are equal to each other on every model, unlike our compact sensors where axial load is tolerated far better than side load. And the overhung moment limit still sits at roughly half the torque rating on most models, so a long or heavy fixture deserves the usual arithmetic: weight against the shear figure, weight times distance against the moment figure.
What is the TRS made from and what does it weigh?
Anodized aluminium throughout the range, including the largest capacities, with published weights running from well under a pound on the smallest bodies to a little over seven pounds on the largest. A sensor rated at a hundred thousand inch-pounds that weighs seven pounds is a genuinely light object, and that is worth knowing when it has to be fitted overhead, handled during a setup, or added to something whose own mass matters. Aluminium also machines cleanly into the shapes a reaction flexure needs, which is why it is used well beyond the capacities people expect it at.
Questions From The Field
Why would I want to measure thrust as well as torque?
Because in most rotating processes the two tell you different things about the same event, and one without the other is ambiguous. A driver that suddenly needs more torque might be meeting a tight thread or might simply be being pushed harder; the thrust channel distinguishes them. A drill that stops cutting shows it in feed force before it shows it in torque. A press-fit, a valve seating, a seal running in — each has a signature in both quantities. If you currently infer axial load from torque or from motor current, a second channel replaces an assumption with a measurement.
My thrust channel shows something when I apply pure torque. Is that a fault?
Almost certainly not — that is cross-talk, and every multi-axis sensor has some. Spring bodies carry manufacturing tolerances and gauge sensitivities vary slightly, so the separation between channels is very good rather than absolute. Two things to check before accepting it. Confirm the torque really is pure: a fixture applying a slight axial pull alongside the twist produces a genuine thrust reading, not cross-talk. And compare the magnitude against the cross-talk specification for your configuration, remembering that the figure is normally referenced to the disturbed channel's full scale rather than to the reading.
Can I mount the sensor so the fixture's own weight sits on the thrust axis?
You can, but think it through first, because on a two-axis sensor that weight is no longer merely an extraneous load — it is a permanent offset on a channel you are actually measuring. With the axis vertical, everything attached below the sensor is dead load on the thrust reading, and it comes out of the usable range as well as needing to be tared. Where thrust is the interesting measurement and the working forces are modest, it is often better to arrange the axis horizontally so the fixture's weight acts across the sensor rather than along it. Tell us the orientation you have in mind and what the fixture weighs.
Can I use the torque channel and ignore the thrust one?
Yes, and nothing is harmed by leaving the second bridge unread. But it is worth connecting it even if you do not plan to use it, because an unread channel is a wasted diagnostic. Axial load that you were not expecting is one of the most common causes of odd torque data, and a thrust reading tells you immediately whether the fixture is doing something you did not intend. Leave the second channel logged and ignored rather than unwired, and the first time a test looks strange you will have the answer already recorded.
How is a two-axis sensor calibrated?
Each channel is characterised in its own right, and on a multi-axis sensor the cross-talk between them is established as a separate part of the exercise. That is the important difference from a single-axis calibration: you are not only confirming what each channel reads when its own load is applied, but also what it reads when the other channel's load is applied. Where your results have to be defensible, ask what the calibration covers and whether cross-talk figures are included — a certificate that documents only the two axes independently leaves the question that matters most on a combined-loading application unanswered.
My shaft needs to run through the middle of the sensor. Is that possible?
It is, using the configuration that relocates the gauge installation to the outside diameter and leaves the centre clear. That turns the sensor into something you can build around an existing shaft or spindle rather than having to sit at the end of a drive. It has to be specified when the sensor is made, so it is a design-stage decision rather than a field one. Send us the shaft diameter, what else has to pass through, and how the reaction will be taken, and we will confirm what is achievable in your capacity.
Does adding a second bridge affect the torque measurement?
The two circuits are electrically independent, so the torque channel is read exactly as it would be on a single-axis sensor and its own accuracy figures apply. What the second bridge introduces is the cross-talk relationship — the torque channel may show a small response to applied thrust and vice versa. On an application where only torque matters and thrust is incidental, that residual is the thing to ask about, since it is the mechanism by which an axial load you are not interested in reaches a reading you are.
The largest model has no mechanical limits published. What should I assume?
Nothing — ask us. The mechanical properties table stops short of the largest capacity, and the honest answer is that you should not extrapolate the trend, because the body dimensions are shared with the model below it while the torque rating doubles. That is exactly the situation where the ratio between the mechanical limits and the torque rating changes, and guessing it is how sensors get damaged. Send us the capacity you need and the fixture you intend to hang on it and we will give you the figures.