Several schemes are available to calculate torque in motion control, quality control, and test applications. Most are based on indirect measurements and are of limited value. Torque sensors, by contrast, measure torque directly. Their output signals are potentially more accurate, provided that the sensors are properly connected, especially in dynamic, changing motion situations.
Stop and go
Torque measurements are generally of two types: reaction and in-line. Reaction sensors measure torque where it is restrained from rotating, such as at a motor's mounting. Measuring reaction torque eliminates the need to electrically connect to a moving sensor. Reaction sensors, however, must often carry extraneous loads, such as motor weight or some of the driveline. This may limit their sensitivity because they are commonly oversized to handle these loads.
In-line sensors, on the other hand, incorporate directly into rotating drivelines, where they measure torque between components. Residing near the point of interest means that their output is accurate and free of parasitic torques from bearings, extraneous loads, and components with large rotational inertias.
An advantage of in-line torque sensors is their ability to spin with the systems they monitor. But this also creates a challenge when it comes to collecting sensor output signals. One common solution is to use a slip ring: a set of conductive rings that rotates with the sensor, while a series of brushes contacts the rings to transmit sensor signals.
Slip rings work well in numerous applications. However, their brushes and (to a lesser extent) rings, wear easily and do not withstand long-term operation or applications difficult to service regularly.
Using slip rings, the top speed for a medium-capacity torque sensor is about 5,000 rpm. At higher speeds, noise in the electrical connection between rings and brushes can degrade performance. Surface speed at the brush-ring interface determines the slip ring's maximum rotational speed. As a result, the maximum operating speed is less for larger, higher torque capacity sensors because slip rings must be larger in diameter.
Designers must account for the brush-ring interface contributing to drag torque. This is problematic in low-capacity measurements or applications where the driving torque cannot surpass brush drag.
Rotary transformers are another way to feed signals out. Here's how they work: An external instrument sends ac voltage via an excitation transformer to a strain-gage bridge that resides on the sensing element (or metal torsional rod). The bridge then drives a second rotary-transformer coil to pick up torque signals from the rotating sensor. The sensing element and strain gage together make up the sensor.
Free from wear and parasitic drag, rotary-transformer systems suit long-term torque monitoring. However, bearings and a frail transformer core limit the typical maximum speed to 10,000 rpm. The system is also prone to noise and errors induced by the transformer's primary-to-secondary coil alignment. Additionally, for acceptance by data acquisition systems, one must employ specialized signal conditioning, further raising costs.
Like rotary transformers, infrared (IR) sensors transfer torque signals from rotating sensors to the stationary world without any direct contact. And, like rotary transformer couplings, IR sensors also transfer power to the rotating sensor, but instead of directly exciting the strain-gage bridge, they power a circuit with heat. The circuit, in turn, provides voltage to the strain-gage bridge and digitizes the output signal. IR light then forwards the signal to stationary receiver diodes where another circuit checks for errors and converts back to analog voltage.
IR transmission of sensor signals is often the best connection for high-speed applications. Although expensive, the digital output makes IR sensors nearly immune to noise from electric motors and magnetic fields.
Unlike traditional rotary transformers, IR transducers can be configured either with or without bearings by mounting the sensor in the rotating driveline and locating stationary electronics near it. When configured without bearings — as a true, non-contact measurement system — they eliminate wear and can survive long-term testing rigs. More importantly, by eliminating bearings, operating speeds rise to 25,000 rpm and higher, even for high-capacity units.
Another way to connect rotating sensors to stationary systems is with a frequency-modulated (FM) transmitter. FM transmitters can unite any sensor to a remote data acquisition system by digitizing signals and transmitting them to an FM receiver for analog conversion.
In torque applications, FM transmitters are usually employed with application-specific sensors: applying strain gages directly to a vehicle's drive or half shaft, for example. FM transmitters require a power source (normally a 9-V battery) on the rotating sensor, which naturally limits test time. Nonetheless, they are easy to install (since they typically clamp to a gaged shaft) as well as reusable with multiple custom sensors.
Dynamic and static torque differences
Force — that product of mass and acceleration — can work on moving and stationary objects. The force acting to stop a moving car, for example, is considered dynamic because the car decelerates. On the other hand, if a moving car is the frame of reference, the force that the brake calipers exert is considered static because the calipers don't accelerate relative to the vehicle.
Torque, the force required to twist or swing a load through a distance, likewise acts on moving and stationary objects. Static torque is torque that remains constant or at rest, and does not accelerate (change rpm). Dynamic torque is that associated with changing rpm or acceleration.
Torque exerted by a clock spring, for instance, qualifies as static torque because there is no rotation and, hence, no angular acceleration. Similarly, torque transmitted through a car's drive axle as it cruises at constant speed constitutes rotating static torque because, though the axle rotates, its rpm remains constant.
Torque in a running car's engine may be static or dynamic, depending on the frame of reference. For instance, as each cylinder fires at the crankshaft, the resulting torque fluctuates greatly, qualifying it as dynamic torque. However, torque measured at the drive shaft appears static because the rotational inertia of the flywheel and transmission dampens torque fluctuations. Winding up car windows generates static torque because although acceleration and rotational inertia are present, the resulting dynamic torque is negligible compared to the frictional forces raising the window.
This last example illustrates that for most measurement applications, both static and dynamic torques come into play. But, when dynamic torque largely impacts the overall or torque of interest, special considerations are necessary to best measure it.