Alternating torque amplitude, defined as a periodic load that swings from a positive peak to a negative peak around zero, is not interchangeable with a static rated torque value, and the gap between the two numbers is where most under-spec'd drive and actuator selections fail.
The DKD-R 10-5 calibration guideline treats the end points of the alternating torque curve, the maximum clockwise and anti-clockwise peaks, as the values that bound a device's static rating, and any alternating application sized beyond those peaks is, by definition, no longer a static selection [S4].
Defining the Two Quantities
Static (reaction) torque is the torque measured with negligible angular acceleration of the rotating element, and is the number that appears on a motor nameplate, a brake rating, and a torque sensor calibration certificate [S3]. For an AC induction machine, static torque at standstill is also called holding or locked-rotor torque, and it is the maximum torque the motor develops at rated voltage and rated frequency without an abrupt drop in speed [S1]. In sizing, it is the single, repeatable scalar a designer works against.
Alternating torque, by contrast, is a time-varying signal with a positive peak, a negative peak, and a defined amplitude about a mean that is often zero. Static torque sensors calibrated under DKD-R 10-5 are explicitly evaluated on those two end points rather than on an RMS value, which is why the document's measurement rules are written as peak amplitudes against the static rated range, not as a derated average [S4]. When the peak-to-peak amplitude approaches the static rating, the device is operating at the edge of its qualified envelope and any further load belongs in a dynamic calibration chain.
Sizing Margins for Drives and Motors
VFD sizing practice on a constant-torque load reserves 150% to 200% of running torque for breakaway, because the drive must overcome static friction, material adhesion, and the first cycle of acceleration before the motor settles into its running curve [S7]. The Rockwell drives engineering handbook treats the same envelope as a torque threshold check, with motor flux held at rated value and the torque component held at or above that threshold across the full speed range, including the acceleration ramp to maximum speed at rated load [S2].
For an alternating-torque load that swings from a positive peak to a negative peak around zero, the relevant rule is not the 150% breakaway margin but the symmetric amplitude limit: the alternating peak amplitude should stay below the static rated torque, with the headroom sized by how often the cycle crosses zero and how much thermal loss the cyclic load deposits in the rotor and bearings. The Rockwell load taxonomy places these cyclic profiles in the variable-torque or constant-horsepower column, and the same constant-flux rule applies: rated flux at rated voltage, torque component at or above threshold [S2]. Stepper systems, which are the most explicit case in the engineering literature, draw a clean line: the static torque is the holding torque at zero steps with nominal current in one phase, and the dynamic torque is the running curve that drops as step frequency rises [S1]. The motor is selected on the running curve when the controller allows variable acceleration, and on the start-without-error curve when it does not, with a typical practical acceleration given as 50 Hz/ms [S1].
Actuator and Damper Sizing Rules
Valve and damper sizing should use the rated torque as the limiting value, not breakaway torque, and any safety factor should be applied to the rated number, not to the locked-rotor or unlatched value [S6]. On a control damper with a 1/2 inch W.C. static pressure differential and 100 fpm face velocity, the operational torque is 4.6 in-lb/sq.ft with seals and 3.1 in-lb/sq.ft without seals; raise the differential to 1 inch W.C. and the velocity to 1000 fpm and the same damper climbs to 6.0 in-lb/sq.ft with seals and 3.5 in-lb/sq.ft without [S5]. The same procedure on round dampers gives 5 lb-in minimum for a 4-8 inch diameter, 15 lb-in for 9-16 inch, and 25 lb-in for 17-22 inch at 1500 fpm closing velocity [S5].
The industry convention is a 0.80 safety factor on rated torque, applied to absorb misalignment, seal aging, and the unaccounted variables in any real installation [S5]. For modulating service, the actuator has to be sized for the full swing of the alternating load, because the same motor that moves the damper at 3.1 in-lb/sq.ft at low pressure has to move it at 6.0 in-lb/sq.ft at high pressure, and the higher number is the one that sets the actuator frame [S5]. A useful sanity check for the bench is the torque wrench tester family: the same static-versus-alternating logic that applies to a damper applies to a torque tool, where the applied joint load is an alternating signal about a target torque and the wrench rating is the static ceiling.
Comparison: Static Selection vs Alternating-Torque Sizing
Four criteria separate the two regimes. First, the limiting value: a static selection is bounded by the nameplate or calibration number, while an alternating-torque selection is bounded by the peak amplitude on each side of zero. Second, the safety factor: 0.80 of rated is a common rule on static-rated actuator duty [S5], while cyclic loads carrying a sign reversal usually need a derate of 40% to 50% on the static peak to keep the alternating amplitude inside the qualified envelope [S4]. Third, the verification artefact: a static selection is closed out by a static calibration against DKD-R 10-5 end points [S4], while an alternating-torque selection needs a dynamic calibration with a defined cycle rate and a confirmed thermal soak. Fourth, the data-acquisition path: a torque sensor on a static bench is read at rest, while the same sensor on an alternating load is read at cycle speed, with sample rate and bandwidth set to capture the peak without aliasing.
The torque transducer itself sits at the same junction. Interface Force's product taxonomy places reaction (static) torque transducers and rotary (dynamic) torque transducers in separate families, and the rule of thumb is that a sensor qualified for static duty should not be used on an alternating load whose peak amplitude is more than 40% to 100% of the static range, depending on cycle rate and duty [S3][S4]. Exceed that envelope and the device drifts out of its calibration class, which is exactly the failure mode DKD-R 10-5 was written to prevent [S4].
Failure Modes and Constraints
Three failure modes show up in the field when the alternating amplitude is sized as if it were static. First, the actuator stalls at end of stroke because the running torque was sized for the average, not the seal-induced peak [S6]. Second, the limit switch trips early because the gearbox backlash under cyclic load is wider than the static-spec window, and the contact opens before the mechanical stop is reached. Third, the limit switch box accumulates wear at a rate set by the peak-to-peak swing rather than the mean, and the mechanical life rating written against static cycles is consumed several times faster.
Actuator vendors publish a sizing guide that lists torques for clean, lubricating fluid with frequent operation and standard seat material, and call out that highly elevated torques require an alternative stem material or a larger frame, with the alternative-material path treated as a clean swap rather than a re-spec [S8]. The same vendor literature makes the static-versus-alternating distinction explicit: the listed torque is the rated static number, and the sizing procedure compares that rated number against the application's required running torque, with a safety margin that absorbs the alternating amplitude [S8].
Decision Matrix by Use Case
For a constant-torque VFD load such as a conveyor or extruder, the rule is rated running torque plus 150% to 200% breakaway headroom [S7]. For a constant-horsepower load such as a centre winder, the rule is rated torque at base speed with a power ceiling above base speed [S2]. For a damper actuator on modulating service, the rule is 0.80 of rated torque with the higher of the low-pressure and high-pressure operational numbers [S5]. For a stepper-driven index table, the rule is selection on the running curve when acceleration is allowed, and on the start-without-error curve when it is not, with 50 Hz/ms as a typical practical acceleration [S1].
For a torque sensor on a test bench, the rule is the DKD-R 10-5 envelope: alternating peak amplitude at or below the static range end points, with the cycle rate and duty disclosed on the calibration certificate [S4]. The same logic shows up in adjacent product categories: a fire-rated door closer is sized against the static closing torque, not the dynamic peak at latch, and a static-rated anti-static equipment wrist-strap tester is calibrated against a static discharge path, not an alternating one. The pattern is the same across the catalogue: pick the rated static number, then derate by the alternating amplitude, and verify with a calibration document that lists the end points explicitly.
Track the next change at two signals: the revision of DKD-R 10-5, which is currently Edition 01/2020 with Revision 1 and would re-open the alternating-amplitude envelope on the next cycle [S4]; and any VFD or actuator vendor datasheet that publishes a quantified alternating-to-static derate curve, which is the missing artefact in most current catalogues [S7][S8].
This topic is covered further in A706 vs A615 Rebar: Specifying Seismic-Grade Welded Reinforcement.