Selecting the right torque transducer
From Automotive Test Report, a T&MW newsletter, October 2002
Dan Romanchik, Editor, Automotive Test Report -- Test & Measurement World, 10/10/2002
Torque is one
of the most important parameters that an automotive test engineer must measure.
You measure torque when conducting engine performance tests, and you measure
torque transmitted to the wheels during road tests or dynamometer tests. You may
also need to measure the torque of electric motors that control such features as
electric windows or doors.
Figure 1 This
torque sensor mounts directly to a vehicle's wheel and transmits readings
wirelessly to a signal conditioner that outputs a high-level analog
signal.
Because torque is such a
fundamental parameter, it is important to know how to select the right
transducer. As noted in “Choosing the Right Torque Sensor” (Ref. 1), the first
thing to do when specifying a torque sensor is to calculate maximum average
running torque (MART) using the following equation:
MART (lb-in) = [maximum rated horsepower] x 63025/rpm
where the rpm value
that you use is the lowest speed at which the motor generates the maximum rated
horsepower.
Next, calculate the probable peak torque (PPT). Knowing
the PPT of your application will help you select a torque sensor with the
appropriate overload torque rating. The equation to use for
this
calculation
is:
PPT = MART x (LSF + DSF) (2)
where LSF is the load service factor and DSF
is the drive service factor. LSF = 1 for applications with smooth, constant
loads; LSF = 2 for applications with non-reversing, non-constant loads; LSF = 3
for applications with highly variable shock or light reversing loads; and LSF
> 4 for applications with full torque reversals. For most automotive
applications, use an LSF of 4 or more.
DSF = 0.5 for gasoline engines
with eight or more cylinders and diesel engines with 10 or more cylinders. DSF =
1 for gasoline engines with six cylinders and diesel engines with eight
cylinders. DSF = 1.5 for gasoline engines with four cylinders and diesel engines
with six cylinders.
Once you’ve calculated these values, you can begin to
specify your torque transducer. The transducer should have a full-scale load
capacity greater than or equal to MART and an overload rating greater than or
equal to twice the PPT. For example, Ford specifies an output power of 110 hp at
5000 rpm for the 2L, SOHC, four-cylinder engine it uses in the Ford Focus.
Plugging these values into Eq. 1
yields:
MART = 110 x 63025/5000 = 1386.6 lb-in.
Using a DSF of 1.5 and an
LSF of 4, you can calculate the PPT with Eq.
2:
PPT = 1386.6 x (4 + 1.5) = 7626.3 lb-in.
Another important
specification is the maximum rpm rating of the torque transducer. You must use a
transducer whose maximum rpm rating is greater than the maximum rpm at which you
plan to measure torque.
Determining Sensor
Accuracy
Three
specifications contribute to a torque transducer's inaccuracy: repeatability,
nonlinearity, and hysteresis. Repeatability is a measure of how much repeated
measurements of a stable input vary. Nonlinearity is a measure of how much
actual readings will deviate from a line drawn from the no-load output to the
full-load output. Hysteresis is a measure of how much a transducer's output will
vary depending on whether the load being measured is approached from the
positive direction or the negative direction.
Typically, you can calculate the
overall accuracy of a torque transducer with 
this equation:
For more information on determining
sensor accuracy, see “Transducer Considerations” (Ref. 2).
When
specifying a torque transducer, you also need to consider how you will power the
transducer and its associated instrumentation. For stationary applications, such
as an engine test stand, AC power will suffice, but for mobile applications,
such as road testing, you'll have to use either the vehicle battery or some
other battery power source.
The type of output is also important. Torque
transducers are available with mV/V, current loop, and ±5. VDC outputs.
Transducers with mV/V output require external signal conditioning, while those
with DC outputs have conditioning circuitry built in. Transducers with mV/V
outputs will generally cost less than those with DC outputs, but you'll have to
supply extra signal conditioning to use them.
Finally, consider the
environment in which you will operate the torque transducer. Extraneous loads,
the electrical environment, and bearing temperature will all effect your torque
measurements. If, for example, you plan to test electric motors using IGBT
adjustable speed drives or you plan to test in an environment that has other
radiating sources, consider using transducers that are shielded against
electrical noise to eliminate measurement errors.
When using a rotating torque sensor for long periods of
time, you should monitor the bearing temperatures by mounting thermocouples on
the transducer housing near the bearings. Typically, the maximum operating
temperature of a torque transducer is 200ºF. Operating a transducer above that
temperature will cause inaccurate readings and may damage the transducer.
References
1. Application Note, “Choosing
the Right Torque Sensor,” S. Himmelstein and Co., Hoffman Estates, IL.
847-843-3300. www.himmelstein.com.
2. Application Note, “Transducer
Considerations.” Lebow Products, Inc., Troy, MI. 800-803-1164.
www.lebow.com/titransducerconsiderations.html.
3. Application Note,
“Application and Selection of Force Transducers,” Al Brendel, Sensor
Developments. Orion, MI. 888-736-7671. www.sendev.com.
