Torque measurement system (TMS) with improved accuracy

Abstract

Disclosed is a torque measurement system (TMS) configured to determine produced torque using correction factors based on calibration of the torque measurement system and a method suitable for characterizing the behavior of the TMS when torque is applied at representative speeds.

Description

Torque Measurement System (TMS) with Improved AccuracyCROSS REFERENCE TO RELATED APPLICATIONS

[0001] The present application claims priority to U.S. Provisional Application No. 63 / 621,236 filed on January 16, 2024, which is incorporated herein.BACKGROUND

[0002] Modem turbine engines are capable of producing high torque values. Management of the torque output is essential to protecting gearboxes and aircraft structure from damage due to over-torque events. Additionally, modern engine management controls can utilize monitored torque values to enhance the efficient operations of high torque producing engines. Thus, improved accuracy in the measurement and monitoring of torque output will improve overall engine operation and enhance the longevity of associated equipment such as gearboxes and other driven devices.

[0003] At lower speeds, it is possible to measure the torque carried by a shaft and transmit a signal to a slip ring, but for high-speed applications, wear can shorten the life of a torque sensing system that relies on physical contact. Non-contact methods for torque measurement using variable reluctance (VR) sensors to measure twist across a shaft segment are well-known. Typically, a reference tube is used in conjunction with ferrous targets to assess twist across a length of shaft. Variable reluctance (VR) sensors are employed to measure changes in the timing of pulses produced by the passage of the ferrous targets. Twist in the shaft can be related to the relative change in pulse timing. Then, by knowing the torsional spring rate of the shaft, torque can be derived from twist.

[0004] The industry would benefit from the provision of a light weight and minimally invasive system capable of accurately measuring twist on a rotating shaft as well as multi-axis shaft motion.Monopole VR sensor-based solutions are light weight and minimally invasive but have limitations in terms of provided twist measurement accuracy. The best currently available systems have accuracies of 5-10% of full scale torque. They also use long (approximately 2 foot long) shafts (also called couplings or torque tubes). The present disclosure provides an improvement in accuracy of the torque measurement system by characterizing the behavior of the system through application of torque at representative speeds. In particular, use of a corrected torque function produced by calibration testing at representative torque and rotational speed conditions will improve torque measurement accuracy with improved integration (shorter torque tubes or couplings).SUMMARY

[0005] In one aspect, the present disclosure provides a torque measuring system which comprises a coupling shaft with a plurality of targets, at least one target sensor positioned a predetermined distance from the plurality of targets, at least one database, the database containing calibration data and a signal conditioning unit, the signal conditioning unit in data communication with the at least one target sensor. The signal conditioning unit includes programming configured to use data from the at least one target sensor to provide a calculated raw twist of the coupling shaft value and to provide a calculated coupling shaft rotational speed value. The programming is further configured to use the calibration data from the database, the calculated coupling shaft rotational speed value, and the calculated raw twist to provide a calculated corrected torque value for the coupling shaft to be output by the signal conditioning unit.

[0006] The present disclosure also provides a torque measuring system comprising a coupling shaft with a plurality of targets, at least two target sensors positioned a predetermined distance from the plurality of targets, at least one temperature sensor, at least one database, the databasecontaining calibration data and a signal conditioning unit. The signal conditioning unit is in data communication with the at least two target sensors and the at least one temperature sensor. The signal conditioning unit includes programming configured to use data from the at least two target sensors to provide a calculated raw twist of the coupling shaft, a calculated coupling shaft rotational speed value, and a radial motion parameter. Additionally, the programming is further configured to use the calibration data from the database, calculated raw twist of the coupling shaft, the calculated coupling shaft rotational speed value, the radial motion parameter, and temperature to provide a calculated corrected torque value for the coupling shaft to be output by the signal conditioning unit.

[0007] Additionally, the present disclosure provides a method for calculating the torque applied to a coupling shaft. The method comprises: providing a torque measuring system comprising: a coupling shaft, the coupling shaft having a plurality of interleaved targets; at least one target sensor; a signal conditioning unit; at least one database; storing calibration data within the at least one database; rotating the coupling shaft; positioning the at least one target sensor a predetermined distance from the targets; using a signal conditioning unit to monitor voltage signals received from the at least one target sensor; in response to the received voltage signals calculate a raw twist for the coupling shaft and a coupling rotational shaft speed for the coupling shaft;using programming within the signal conditioning unit and the calculated raw twist, the calculated coupling rotational shaft speed, and the stored calibration data to provide a calculated corrected torque value for the coupling shaft.

[0008] In another aspect, the present disclosure provides a method for calculating a corrected torque value for torque applied to a coupling shaft. The method comprises: providing a torque measuring system, the torque measuring system comprising: the coupling shaft; a plurality of interleaved targets carried by the coupling shaft; a first target sensor positioned to monitor the plurality of interleaved targets; providing a database, the database containing calibration data; providing a signal conditioning unit, the signal conditioning unit in data communication with the first target sensor, wherein the signal conditioning unit includes programming to: use data from the first target sensor to calculate a coupling shaft rotational speed value; use data from the first target sensor to calculate a raw twist value of the coupling shaft, use data from the calibration data from the database, the coupling shaft rotational speed value, and the raw twist value to calculate a corrected torque value; storing calibration data within the at least one database; upon rotation of the coupling shaft, the first target sensor generates a data signal; the signal conditioning unit receiving the data signal from the at least one target sensor;in response to the received data signal the signal conditioning unit uses the programming to calculate a raw twist value for the coupling shaft and a coupling shaft rotational speed value for the coupling shaft, the signal conditioning unit further uses the calibration data, the raw twist value and the rotational speed value to provide a calculated corrected torque value for the coupling shaft; the signal conditioning unit provides the calculated corrected torque value as an output value.BRIEF DESCRIPTION OF THE DRAWINGS

[0009] FIG. 1 provides a block diagram of a single channel torque measurement system (TMS).

[0010] FIG. 2 provides an example of a torque measurement system (TMS) rotating coupling and VR sensor components.

[0011] FIG. 3 illustrates an exploded view of the shaft assembly with twist section and cylinders with interleaved targets lined up for assembly.

[0012] FIG. 4 shows an example VR sensor output signal with logic level transitions at negative zero crossing and corresponding timer counts (vk).

[0013] FIG. 5 is a schematic for VR sensor signals being processed by the signal conditioning circuit and then read by the microcontroller.

[0014] FIGS. 6A and 6B illustrate the radial and tangential motion of a VR sensor relative to the interleaved targets.

[0015] FIG. 7 is a schematic showing the relative change in sensor position due to relative motion along the radial and tangential directions.

[0016] FIG. 8 is a flowchart showing steps applied by SCU programming for torque correction based on raw twist and speed, without radial motion and temperature correction, to provide a calculated corrected torque value for a TMS with single or dual VR sensors.

[0017] FIG. 9 is a flowchart showing steps applied by SCU programming for torque correction using a raw twist and speed using dual VR sensors and including temperature measurement and a radial motion parameter to calculate a corrected torque value.

[0018] FIG. 10 shows a block diagram for a torque calibration data generation apparatus using an aircraft engine or motor.

[0019] FIG. 11A provides an example of corrected torque function at a certain coupling temperature and radial motion correction value.

[0020] FIG 11B provides an example of a normalized coupling shaft stiffness versus temperature variation.

[0021] FIG. 12A is a plot showing the raw dual sensor torque measurements using the torque calibration data and the reference torque.

[0022] FIG. 12B provides a close-up of the final and actual data points for the region identified in FIG. 12 A.

[0023] FIG 13 A depicts the full-scale torque measurement error percentage over a range of shaft speeds and applied torques for dual sensor twist measurement with correction for actual twist but with no correction for speed, or radial motion.

[0024] FIG. 13B depicts a histogram of the error distribution and standard deviation for the values of FIG. 13 A.

[0025] FIG. 14A is a plot of the full-scale torque measurement error percentage over a range of shaft speeds and applied torques for dual sensor twist measurement with a calculated correctedtorque value based on the dual sensor raw twist and speed data but without radial motion correction.

[0026] FIG. 14B depicts a histogram of the error distribution and standard deviation for the values depicted in FIG. 14A.

[0027] FIG. 15A is a plot of the full-scale torque measurement error percentage over a range of shaft speeds and applied torques for a calculated corrected value based on dual sensor raw twist, speed, and a radial motion parameter.

[0028] FIG. 15B depicts a histogram of the corresponding error distribution and standard deviation for the values of FIG. 15 A.

[0029] FIG. 16A is a plot of the full-scale torque measurement error percentage over a range of coupling shaft speeds and applied torques for a calculated corrected torque value based on a single sensor raw twist and speed data but without a radial motion parameter.

[0030] FIG. 16B depicts a histogram of the corresponding error distribution and standard deviation for the values of FIG. 16A.DETAILED DESCRIPTION

[0031] The drawings included with this application illustrate certain aspects of the embodiments described herein. However, the drawings should not be viewed as exclusive embodiments. The subject matter disclosed is capable of considerable modifications, alterations, combinations, and equivalents in form and function, as will occur to those skilled in the art with the benefit of this disclosure.

[0032] The present disclosure may be understood more readily by reference to the following representative description. For simplicity and clarity of illustration, where appropriate, reference numerals may be repeated among the different figures to indicate corresponding or analogouselements. The following description is not to be considered as limiting the scope of the embodiments described herein. The drawings are not necessarily to scale and the proportions of certain parts may have been exaggerated to better illustrate details and features of the present disclosure. Also, the phraseology and terminology employed herein is for the purpose of description and should not be regarded as limiting except where indicated as such.

[0033] Throughout this disclosure, the terms “about”, “approximate”, and variations thereof, are used to indicate that a value includes the inherent variation or error for the device, system, or measuring method being employed as recognized by those skilled in the art.

[0034] The following abbreviations and terms are used in this disclosure.• AB = Twist Delta in degrees between the A (104) and B (105) target sets measured by a single VR sensor• ABra : single target sensor measured twist, raw.• CPR = Counts per Revolution• DAB = Dual VR sensor twist measurement (in degrees) between the A and B interleaved targets.• D ABraw: Dual target sensor measured twist, raw.• FADEC: Full Authority Digital Engine Controller• FPGA: Field Programmable Gate Array• fciock = Clock_Frequency or high speed counter rate (counts / sec).• FS = Full Scale• h = nominal air gap between toothed wheel and VR sensor in radial directionKmod: normalized torsional stiffness of the TMS coupling shaft 100 as a function of coupling temperature• m = slope of coupling shaft stiffness-temperature curve• N = number of VR targets (A-side and B-side combined) per revolution.• R = nominal radius to toothed wheel• RAD = Radial Angular Distortion Timing delta, converted to degrees, between consecutive twist-invariant passes of two targets (i.e., the time between when an individual target passes the VRi sensor and when another target from the same side of coupling shaft 100 passes the VR2 sensor on the next timestep)• SCU = Signal Conditioning Unit• SPD = Normalized coupling shaft 100 speed expressed in degrees per clock count.• T : temperature.• Tc : temperature at calibration• TMS = Torque Measurement System• VR = Variable Reluctance• ZCD = Zero Crossing Detector• Vj , V2 : timing counter values at negative slope zero-crossing for sensor 1 and 2 respectively• o : measured coupling shaft 100 speed.• 0raw : raw twist measurement calculated from zero cross timings (same as ABrawfor a single target sensor or DABraw for 2 target sensors)• T: torque at coupling shaft 100• Ax, Ay: change in position of VR sensor due to motion along x and y axes respectively.• Ah, Aq>: change in nominal air gap height and tangential angular position respectively due to radial motion along x and y axes of a VR sensor.

[0035] The present disclosure provides an improved torque measurement system 110 (TMS).TMS 110 incorporates improvements which enhance the accuracy of the torque calculations reported by TMS 110 as well as enable improved integration.

[0036] FIG. 1 shows a complete torque measurement system (TMS) 110 including a coupling 10 with interleaved targets 103, at least one target sensor 102, an SCU 109, and optionally or additionally at least one temperature sensor 112. TMS coupling 10 is the rotating component that transmits torque. TMS coupling 10 includes interleaved targes and has the target sensors 102 mounted around the interleaved targets 103. Types of target sensors 102, include but are not limited to, VR, eddy current, microwave, hall effect, optical including laser. Most embodiments of the described system will use VR sensors as target sensors 102. TMS 110 also includes a zero cross detection circuit 130 within an SCU 109. Zero cross detection circuit 130 is configured to receive electrical signals from the VR sensors 102. The SCU 109 also contains a programmable microcontroller 132 with programming which utilizes the zero cross timing data to calculate the raw twist 0raw (box 146) and the rotational shaft speed ® (box 147) of shaft 100. In another embodiment, SCU 109 can determine zero cross timing in microcontroller 132 by sampling the VR sensor 102 waveform and comparing to the microcontroller’s internal timing. Such programming will follow the steps outlined in either FIGS. 8 or 9 depending on the current system. Additionally, if two or more VR sensors 102 are present in the system, then microcontroller 132 will also include programming suitable for calculating a radial motion dependent parameter RAD (box 148) using the zero crossing timing data. The SCU 109 also calculates an estimate of the temperature (T) of the coupling shaft 100 by using the temperature sensor 112. The estimated value T can be used to compensate for changes in the stiffness of the coupling shaft resulting from changes in temperature. The configuration of SCU 109, zero cross detection circuit 130 andmicrocontroller 132 represent one embodiment of a system for carrying out the following methods. SCU 109 may be arranged in alternative configurations using differing circuitry, data storage arrangements as well as alternatives to microcontroller 132. Therefore, reference to the performance of certain steps by SCU 109 also includes functionality carried out using zero cross detection circuit 130 and / or microcontroller 132.

[0037] In the next step, microcontroller 132 uses stored torque calibration data 142 along with calculated shaft speed co and the calculated raw twist 0raw to calculate a corrected torque value 143. Additionally, when TMS 110 has two or more VR sensors 102, microcontroller 132 may utilize the radial motion parameter RAD (box 148) as an input to the corrected torque function 143 to produce a more accurate calculated corrected torque value. When TMS 110 includes a temperature sensor 112, the measured temperature can also serve as an input to the corrected torque function 143. The torque calibration data 142 represents the effects of shaft speed, and torque applied to the shaft 100. SCU 109 uses microcontroller 132 to output the calculated corrected torque value and / or rotational speed. When the stored calibration data also includes temperature effects on coupling stiffness calibration, microcontroller 132 may utilize this data along with data obtained from temperature sensor 112 to correct for the effect of temperature in the calculated corrected torque value. One skilled in the art will recognize that box 142 represents a database suitable for storing data. As known to those skilled in the art, databases included in TMS 110 may be configured to receive calibration data and mathematical operators, such as but not limited to polynomial coefficients or look-up tables, utilized by programming within microcontroller 132. Such databases may be located in non-volatile memory inside or outside of SCU 109.

[0038] FIGS. 2-3, show a rotatable TMS coupling 10 with a twist element 101, i.e., a compliant region of the coupling shaft 100, that undergoes enough twist under operational torque levels tobe detected by the following described method. VR sensors 102 are housed in a sensor cradle 1 11 structure, generically depicted in FIG. 10, which retains VR sensors 102 in a predetermined, i.e. a fixed nominal position in the non-rotating frame. Thus, sensor cradle 111 and VR sensors 102 do not rotate with coupling shaft 100. Interleaved shaft targets 103 are part of or are attached to the rotating coupling shaft 100 or a flange carried by rotating coupling shaft 100 such that they are not in the primary torque transmission path through coupling shaft 100. The A-side targets 104 are on one side of the twist element 101 and the B-side targets 105 are on the other side of shaft twist element 101. At least one non-contact variable reluctance sensor 102 is positioned such that the passing of targets 103 will induce an oscillating voltage waveform (see FIG. 4) with one voltage oscillation corresponding to each target 103 passing by the VR sensor 102. Thin- walled coupling shaft 100 with twist element 101 is commonly used in the art in connection with variable reluctance-based torque sensors 102. Rotation of coupling shaft 100 is around the z-axis defined by the coordinate system 108. In one embodiment, coupling shaft 100 carries input / output flanges or connection points 106 used to transmit torque from one drive train element to another.

[0039] An exemplary embodiment of TMS coupling 10 is illustrated in FIG. 3. In this example, A-side targets 104 are carried by a first cylinder element 150 and B-side targets 105 are carried by a second cylinder element 160. First cylinder element 150 is secured to mounting surface 152 and second cylinder element 160 is secured to mounting surface 162. Mounting surfaces 152 and 162 are on either side of twist element 101. Thus, as described above, interleaved targets 103 (A-side 104, B-side 105) are secured to opposing sides of twist element 101 .

[0040] FIG. 4 shows an example voltage signal produced by VR sensor 102 due to passage of targets 103 below it as represented in plot A. The change of magnetic flux density between a single ferrous target 103 followed by an air gap results in the oscillatory change in voltage.

[0041] A signal conditioning unit (SCU) 109, shown in FIG. 5, equipped with a zero-crossing detection circuit 130 reads the voltage signal (FIG. 4, plot A) and produces logic level one (FIG.4, plot B) whenever a negative (or falling) transition occurs at zero volts; the logic level resets to zero (FIG. 4, plot B) whenever an arming voltage level is reached. Note that the SCU 109 may be embedded within a larger electronic controller, for example, an engine controller or FADEC. A microprocessor, field programmable gate array (FPGA) or microcontroller 132, within SCU 109, reads the value of a free running counter whenever logic level transitions from zero to one. In a microcontroller, an “enhanced capture timer”, or ECAP reads the output of the zero-crossing detection circuit. For the purposes of this disclosure, the term microcontroller is considered to include any suitable programmable device capable of being incorporated into SCU 109 and performing the necessary calculations. This timer value is referred to as v as depicted in plot C of FIG. 4, where subscript n refers to the sensor number and k refers to the index of a specific timing measurement. In this instance, the specific timing event is the zero crossing event which corresponds to the period of time between two successive targets 103 passing a single VR sensor 102. The zero crossing event is indicated at each vertical line ZCD in FIG. 4, plot A. Thus, the zero crossing event is always on the negative slope of the waveform depicted in FIG. 4, plot A. A consecutive series of timer values from the first of two VR sensors 102 would be denoted v , v+1, Vi+2, etc. Note: while the zero cross detection 130 is shown as a separate circuit in FIGS. 1 and 5, parts or all of the circuit could also be implemented within microcontroller 132. The zero crossing timing values are then used to calculate the raw twist (0raw) of coupling 10; for a TMS with dual VR sensors, Qraw = DABraw, while for a TMS with a single VR sensor, 0raw=AB raw

[0042] With reference to FIGS. 1 , 8 and 9, microcontroller 132 calculates the speed of rotation of the TMS coupling 10 in two steps (box 147). The first step computes the counts per revolution (CPR) based on the timing count from a single sensor, say number 1, and the number of targets N as follows:CPR = vk- vk~NIn the second step, the speed SPD in degrees per clock count is calculated from CPR as follows:360SPD =~CPRThe speed co in revolutions per minute (RPM) is calculated using the system timing clock rate, fdock, and CPR as follows:60 x fcioci a> —CPRAn example value of fdock could be 200x10A6 (i.e. 200 MHz).

[0043] When the signals from two sensors 102 are spaced such that one target from the A-side 104 passes a first VR sensor 102 at approximately the same time as a target from the B-side 105 passes the other VR sensor 102, the twist can be characterized with a speed-independent dualsensor A-to-B (DAB) twist parameter. Before normalization, the A-to-B timing difference in counts, dabk, is: dabk= ( lip)(—l)k(vk— vf) where vkis the timing count from a first VR sensor 102 at timing step k, and vkis the timing count from a second VR sensor 102 at timing step k, and flip is a 1 or -1 which is used to rectify the signal if the mean dabkvalue of the previous two revolutions is below zero. If the waveform pulses are not registered in a way so that the A-side and B-side target positions are known for a given time step, k, the sign of the twist (i.e., whether it is positive or negative twist) is unknown. This can be addressed by machining targets 103 such that there is a twist offset under zero torquegreater than the maximum designed negative twist. In that case, the A-side and B-side position is known according to the sign, positive or negative, of dabk, and signal rectification can provide the appropriate twist value once the offset is accounted for. The speed-normalized twist, DABraw, averaged over the entire set of N targets and in degrees is:

[0044] For single sensor-based approach, twist measurement is made by comparing the timing difference between when the A-side targets pass a sensor to when the B-side targets pass the same sensor. Before normalization, the A-to-B timing difference in counts for the first VR sensor 102, abk, is:where vkis the timing count from first VR sensor 102 at timing step k, and vk~kis the timing count from the same first VR sensor 102 at timing step k-1. The absolute speed-normalized twist, ABraw, measured by sensor 1 and averaged over the entire set of A' targets and in degrees is:

[0045] The waveform varies in reaction to changes in the relationship of targets 104 and 105. During application of torque to a thin-walled twist element 100, twist can be identified by the shift in targets 104 and 105 relative to one another. The small deflections result in a small timing change between sequential vkvalues.

[0046] FIGS. 6A and 6B show views of TMS 110 with twist element 100, interleaved targets103, and VR sensors 102. As represented in FIG. 2 and FIGS. 6 A, 6B, VR sensors 102 are in a single axial plane. FIG. 6A also provides a section view of the coupling with variable reluctance sensors 102 mounted in proximity to interleaved targets 104 and 105. It also shows radial motion114 that causes a change in the air gap 123 between the interleaved targets 104, 105 and sensors 102. FIG. 6B shows a different view of TMS coupling 10 with relative motion between sensors 102 and interleaved targets 104, 105 in the tangential direction 115. During operation of the engine 310, heat, vibration, and other stresses can cause shifting of the components in TMS 110. Relative movement between the components can occur due to differences in thermal expansion between different metals used throughout an aircraft drivetrain. Manufacturing tolerances also provide a source of position variation between parts. Each TMS 110 will have slight differences from another TMS 110 due to stacking of tolerances as the various parts in the final assembled TMS 110 will have slightly different sizes within the accepted range of tolerances per manufacturing requirements. In fact, the stack-up of tolerances may result in substantial measurements in the radial 114 and tangential 115 directions between unique assemblies. Additionally, structural flexing may contribute to relative movement in the radial 114 and tangential 115 directions. Thus, the construction limitations of TMS 110 will also produce tangential and radial shifts of variable reluctance sensors 102 with respect to coupling 10. In FIG. 6A, shift of variable reluctance sensors 102 in the radial direction is reflected by arrows 114 and in FIG. 6B, shift of variable reluctance sensors 102 in the tangential direction is reflected by arrows 115.

[0047] FIG. 7 illustrates how relative motion of the VR sensors 102 with respect to TMS coupling 10 can produce a change in the radial air gap 123 as well as the tangential location of sensor(s) 102. Fig. 7 shows the nominal position 120 of a VR sensor 102 which is located at the sum of the nominal radius (R) 122 and nominal air gap (h) 123 from the centerline. Due to movement of the sensor in the local x-direction by Ax (124) and in the local y-direction by Ay(125), the final position 121 of sensor 102 is different from the nominal position 120. The final position 121 has a different air gap equal to the sum of the original air gap, h 123 and the changein the gap Ah 126, and new position tangentially shifted by Acp, 127. This geometric analysis illustrates the need for including radial motion correction factors (RAD) to calculate an accurate value of torque.

[0048] Changes to the radial air gap 123 and the angular spacing of the VR sensors 102 change the voltage waveforms of the VR sensors 102 and therefore change the timing of zero crosses which is reflected in the vkvalues illustrated in FIG. 4. To compensate for these changes, a speed- normalized radial angular distortion (RAD) parameter is defined as follows (box 148):

[0049] FIG. 8 is a flowchart showing the program steps incorporated into microcontroller 132 and used by microcontroller 132 to determine the shaft torque (T) from the VR sensor signals. Using SCU 109, the process starts with the detection of a series of negative zero crossings of the two VR sensor signals by the SCU 109. Microcontroller 132 uses the resulting crossing event timer values, vk, through vk N, to calculate the dual sensor shaft speed co and raw twist, DABraw, of coupling shaft 100. Alternatively, when using a single VR sensor signal, microcontroller 132 calculates the shaft speed and single sensor raw twist, ABraw. SCU 109 then applies the calculated corrected torque function to determine a calculated corrected torque value, T = fi(co, DABraw) or T = f2(c , ABraw) using the calculated raw twist, calculated shaft speed, and stored calibration data that is stored in the database. The numerical calculations can be performed using software written in a computing language suitable for the microcontroller 132 or SCU 109, such as C or Simulink.

[0050] FIG. 9 is a flowchart showing the program steps incorporated into microcontroller 132 to use radial motion correction parameter (RAD) and the temperature measurement from temperature sensor 112 to produce the calculated corrected torque value. Using SCU 109, the process starts with the detection of a series of negative zero crossings of the two VR sensor signalsby SCU 109. Microcontroller 132 uses those crossing event timer values, vk, through vk'N, to calculate shaft speed co, raw twist DABraw, and radial motion parameter, RAD. The corrected torque, T, is calculated by microcontroller 132 using the torque calibration function T = f3(co, DABraw, RAD, T) which is a function of the shaft speed, co, the raw twist, DABraw, the stored calibration data from the database, and temperature, T, and the radial motion parameter, RAD.

[0051] FIG. 10 depicts an exemplary apparatus used during end-of-line testing of an engine or motor incorporating TMS 110 with SCU 109 to generate the calculated corrected torque functions used for calibration and stored in the database 142. Typically, the end-of-line testing occurs at the final stage of manufacturing after TMS 110 is built and assembled but before it is used on a vehicle. The calculated corrected torque functions created during end of line testing are then stored in the SCU 109 for use in service on an engine. The setup includes an engine or drive motor 310, corresponding engine or motor controller 311, a calibrated dynamometer or brake 312 as the load, and the torque measurement system 110, all mounted on the same shaft 307. A reference torque transducer 303 may also be used to provide actual torque measurement. A control processor 206 is used to control the engine or drive motor 310 and dynamometer or brake 312 as well as received torque measurement information from the SCU 109 and optional reference torque transducer 303. Control processor 206 typically resides on a test computer (e.g., Windows personal computer). Engine or motor controller 311 is controlled by the control processor 206 to run the drivetrain at a commanded rotational speed co. The shaft torque T is measured by TMS 110 using SCU 109. The control processor 206 runs tests at different speed and torque combinations to generate a calculated corrected torque function, e.g., T = fi(co, DABraw) for dual sensor correction without motion compensation or T = f2(co, ABraw) for single sensor correction without motion compensation. An example corrected torque function, T = file), DABraw, T, RAD), generated by this test apparatus ata certain value of RAD and shaft temperature T is shown in FIG. 11 A. Note: the test setup from FIG 10 is the nominal setup on the final engine or vehicle with the addition of the test equipment used by the Control processor 206 during the calibration of TMS 110. The additional equipment used during calibration includes a Reference torque transducer 303 supported by a shaft 307 and a calibrated dynamometer if brake 312 also supported by shaft 307. The identified additional test equipment would not be present during normal operation of TMS 10.

[0052] TMS coupling 10 stiffness versus temperature variation may optionally be included as a component of the torque calibration function. This variation can be determined through changing the ambient temperature during the on-engine calibration or via a static calibration of a coupling. Figure 11B shows an example stiffness variation versus temperature for a coupling. The temperature compensation of the shaft can be accomplished through the following equation:In the above equation, Kmod is the normalized shaft stiffness compensated for temperature, m is the slope (% change in stiffness per temperature), T is measured temperature from sensor 112, and Tc is the calibrated temperature value. Use of temperature sensor 112 can provide an adequate estimate of the coupling 10 temperature for these purposes. The values m and Tc would be stored in the database as calibration data for the coupling shaft.

[0053] The calculated corrected torque function can have the following polynomial form for simple and flexible implementation in an embedded microcontroller: / (x1, x2, x3, x4) = a0+where xi, X2, x.3, X4 are independent variables and ao through ai4 are constant coefficients. For dual sensor calculated corrected torque functions with radial correction, the independent variables would be xi = co, X2 = DABraw, X3 = T, and X4= RAD. For a dual sensor calculated corrected torque function without radial compensation, the independent variables would be xi = co and X2 = DABraw with coefficients aio, an, ai2 , ai3 and ai4 set to zero. For a single sensor calculated corrected torque function, the independent variables would be xi = CD and X2 = ABra with coefficients aio, an, ai2 , ai3 and ai4 set to zero. The polynomial coefficients are then stored within the SCU 109 calibration data (box 142) for use in the torque correction process described in FIG. 8 and FIG. 9. The coefficients ai through aio would be stored in the database as calibration data.

[0054] After all the test points have been completed, a suitable fitting algorithm, e.g. the Levenberg-Marquardt non-linear least squares method, may be used to compute the coefficients of the calculated corrected torque function by fitting a mathematical function to the raw twist data. Levenberg-Marquardt is a popular alternative to the Gauss-Newton method of finding the minimum of a function F(x) that is a sum of squares of nonlinear functions,Let the Jacobian of '(x) be denoted J / (x), then the Levenberg-Marquardt method searches in the direction given by the solution p to the equations,where Xk are nonnegative scalars and I is the identity matrix. The method has the nice property that, for some scalar A related to Xk, the vector pk is the solution of the constrained subproblem ofminimizingsubject to ||p||2 < A. (Gill, P. R; Murray, W.; and Wright, M. H."The Levenberg-Marquardt Method. " §4. 7.3 in Practical Optimization. London: Academic Press, pp. 136-137, 1987).

[0055] An alternate format for generating and storing the calculated corrected torque function is a look-up table. In this format, for a dual sensor scheme with radial motion correction, the calculated corrected torque values are stored within the SCU 109 calibration data (in box 142) along with the corresponding independent values of (DABraw), shaft speed (co), radial motion parameter (RAD), and / or temperature (T). Microcontroller 132 would then use linear interpolation to calculate the corrected torque values corresponding to intermediate measured values of (DABraw), shaft speed (co), radial motion parameter (RAD), and / or temperature (T).

[0056] FIG. 12A shows the measured (reference), raw, and calculated corrected torque values from a dual sensor TMS 110 at various levels of applied torque and at different shaft speeds. The raw torque values, calculated by multiplying the raw twist by the coupling stiffness, is shown with circle symbols in FIG 12A. Note that the raw torque values in FIG 12A are for comparison purposes only. The reference torque is shown using x-symbols. The calculated corrected torque values are shown using squares in FIG 12A and then again zoomed in in FIG 12B.

[0057] FIG. 13 A shows the torque measurement accuracy for a dual VR sensor system with no speed, radial motion or temperature compensation. The measurement errors can be up to ±7% and there is significant variation with applied torque levels; the standard deviation for measured errors is 3.42%, as seen in FIG. 13B.

[0058] FIG. 14A shows the torque measurement accuracy for a dual VR sensor system using a calculated corrected torque function with speed compensation but without radial motion and temperature effects based on the flowchart shown in FIG. 8. The mathematical form for thecalculated corrected torque function is T = fi(®, DABraw). The maximum measurement errors can be up to 3.8% with a standard deviation of 1.36%, as seen in FIG. 14B. This step is an improvement over the results shown in FIG. 13A and corresponds to using block 143 but not block 148 from FIG. 1.

[0059] FIG. 15A illustrates the torque measurement accuracy when using a dual VR-sensor based TMS with a calculated corrected torque function (box 143) including the additional radial motion correction RAD (box 148), based on the flowchart shown in FIG. 9 . The mathematical form for the calculated corrected torque function is T = f5(co , DABraw, RAD). The error distributions in FIGS. 15A and 15B show that by using the calculated corrected torque function including a radial motion input, most of the torque measurement error is limited to ±1% with a standard deviation of 0.45%, which is more accurate than the other configurations, thus showing the significance of including all correction steps.

[0060] FIG. 16A shows the torque measurement accuracy for a single VR sensor system with a calculated corrected torque function using speed and raw twist compensation but without radial motion or temperature effects. The mathematical form for the calculated corrected torque function is r = f2(co, ABraw). The maximum measurement errors can be up to 4.4% with a standard deviation of 1.72%, as seen in FIG. 16B. Since the parameter RAD is defined as timing difference from two sensors, it cannot be calculated and thereby is not applied for a single sensor TMS configuration.

[0061] Note that the twist at full-scale torque is approximately 0.6 degrees. However, this system will function equally well with full-scale twist as low as 0.2 deg. This is much smaller than the typical twist of existing torque systems that is 2 degrees or higher. This invention achieves high torque accuracy at low twist, which results in low overall shaft length (e.g. 0.4 feet long versus 2 feet or longer). However, shaft lengths between 0.25 feet and 2 feet may also be used.

[0062] Prior to installing an engine which incorporates TMS 110, as described above, the engine will be operated at various torques and speeds under various conditions to establish calibration data. During calibration operation, a control processor 206 will manage the operation of the engine to establish calibration correction factors necessary for the calculation of the corrected shaft torque. The resulting calculated corrected torque functions are stored in memory in SCU 109 for use during normal engine operation. Subsequently, during operation of the engine on an aircraft or other device, TMS 110 using SCU 109, will monitor torque produced by the engine and apply the appropriate calculated corrected torque function to provide the corrected torque value for the current operational conditions.

[0063] Other embodiments of the present invention will be apparent to one skilled in the art. As such, the foregoing description merely enables and describes the general uses and methods of the present invention. Accordingly, the following claims define the true scope of the present invention.

Claims

We claim:

1. A torque measuring system comprising: a coupling shaft with a plurality of targets; at least one target sensor positioned a predetermined distance from the plurality of targets; at least one database, the database containing calibration data; a signal conditioning unit, the signal conditioning unit in data communication with the at least one target sensor, the signal conditioning unit including; programming configured to use data from the at least one target sensor to provide a calculated raw twist of the coupling shaft value and to provide a calculated coupling shaft rotational speed value; wherein the programming is further configured to use the calibration data from the database, the calculated coupling shaft rotational speed value and the calculated raw twist to provide a calculated corrected torque value for the coupling shaft to be output by the signal conditioning unit.

2. The torque measuring system of claim 1, wherein the targets on the coupling shaft are interleaved.

3. The torque measuring system of claim 1, further comprising a temperature sensor and the signal conditioning unit further including programming configured to use data received from the temperature sensor, wherein the programming is configured to use the temperature sensor data in the calculation of the corrected torque value for the coupling shaft.

4. The torque measuring system of claim 3, wherein the programming of the signal conditioning unit is configured to use the temperature sensor to compensate for changes in the coupling shaft stiffness in the corrected torque value for the coupling shaft.

5. The torque measuring system of claims 1 , wherein the at least one target sensor is a variable reluctance sensor, eddy current sensor, microwave, hall effect, or optical including laser sensor.

6. The torque measuring system of claim 1, wherein the torque measuring system has at least two sensors positioned a predetermined distance from the plurality of targets in a single axial plane.

7. The torque measuring system of claim 4, wherein the signal conditioning unit further includes programming configured to calculate a radial motion parameter between the coupling shaft and the at least two target sensors and wherein the programming is configured to use the radial motion parameter in the calculation of the corrected torque value for the coupling shaft.

8. The torque measuring system of claim 1, wherein the database is contained within the signal conditioning unit.

9. The torque measuring system of claim 1 wherein the database is contained within a microcontroller.

10. The torque measuring system of claim 1, wherein the database is contained external to the signal conditioning unit.

11. The torque measuring system of claim 1, further comprising: a zero crossing detection circuit configured to receive a voltage signal from the at least one target sensor; and, wherein the signal conditioning unit programming is configured to receive data from the zero crossing detection circuit.

12. The torque measuring system of claim 1 , wherein the signal conditioning unit circuitry and programming is contained within an engine controller or FADEC.

13. The torque measuring system of claim 1, wherein the signal conditioning unit includes a microcontroller and wherein the signal conditioning unit programming for performing thecalculations to provide a calculated corrected torque value for the coupling shaft is stored in the microcontroller and wherein the microcontroller is in data communication with the database.

14. The system of claim 1, where the calculated corrected torque value for the coupling shaft is calculated from a polynomial or look-up table of the calculated raw twist and the calculated coupling shaft rotational speed.

16. The torque measuring system of claim 1, wherein the microcontroller programming could be implemented in at least one field programmable gate array (FPGA) or microprocessor.

17. A torque measuring system comprising: a coupling shaft with a plurality of targets; at least two target sensors positioned a predetermined distance from the plurality of targets; at least one temperature sensor; at least one database, the database containing calibration data; a signal conditioning unit, the signal conditioning unit in data communication with the at least two target sensors and the at least one temperature sensor, the signal conditioning unit including: programming configured to use data from the at least two target sensors to provide a calculated raw twist of the coupling shaft, a calculated coupling shaft rotational speed value, and a radial motion parameter; wherein the programming is further configured to use the calibration data from the database, calculated raw twist of the coupling shaft, the calculated coupling shaft rotational speed value, the radial motion parameter, and temperature to provide a calculated corrected torque value for the coupling shaft to be output by the signal conditioning unit.

18. The torque measuring system of claim 17, wherein the programming of the signal conditioning unit is configured to use the temperature sensor to compensate for changes in the coupling shaft stiffness in the corrected torque value for the coupling shaft.

19. The torque measuring system of claim 17, further comprising: a zero crossing detection circuit configured to receive a voltage signal from the at least one target sensor; and, wherein the signal conditioning unit programming is configured to receive data from the zero crossing detection circuit.

20. The torque measuring system of claim 1, wherein the signal conditioning unit includes a microcontroller and wherein the signal conditioning unit programming for performing the calculations to provide a calculated corrected torque value for the coupling shaft is stored in the microcontroller and wherein the microcontroller is in data communication with the database.

21. A method for calculating a corrected torque value for torque applied to a coupling shaft, the method comprising: providing a torque measuring system, the torque measuring system comprising: the coupling shaft; a plurality of interleaved targets carried by the coupling shaft; a first target sensor positioned to monitor the plurality of interleaved targets; providing a database, the database containing calibration data; providing a signal conditioning unit, the signal conditioning unit in data communication with the first target sensor, wherein the signal conditioning unit includes programming to:use data from the first target sensor to calculate a coupling shaft rotational speed value; use data from the first target sensor to calculate a raw twist value of the coupling shaft, use data from the calibration data from the database, the coupling shaft rotational speed value, and the raw twist value to calculate a corrected torque value; storing calibration data within the at least one database; upon rotation of the coupling shaft, the first target sensor generates a data signal; the signal conditioning unit receiving the data signal from the at least one target sensor; in response to the received data signal the signal conditioning unit uses the programming to calculate a raw twist value for the coupling shaft and a coupling shaft rotational speed value for the coupling shaft, the signal conditioning unit further uses the calibration data, the raw twist value and the rotational speed value to provide a calculated corrected torque value for the coupling shaft; the signal conditioning unit provides the calculated corrected torque value as an output value.

22. The method of claim 21 for calculating the corrected torque value for torque applied to a coupling shaft, wherein: the torque measuring system further includes a temperature sensor and wherein the calibration data also includes temperature compensation data for the coupling shaft; wherein the programming to provide the calculated corrected torque value also uses the temperature compensation data to provide the calculated corrected torque value.

23. The method of claim 21 for calculating the corrected torque value for torque applied to a coupling shaft, wherein:the torque measuring system further includes a second target sensor positioned to monitor the plurality of interleaved targets, the second target sensor in data communication with the signal conditioning unit; upon rotation of the shaft, the second target sensor generates a second data signal; wherein the signal conditioning unit further includes programming which uses the data signals from the first and second target sensors to calculate a radial motion parameter between the coupling shaft and the at first and second target sensors; wherein the programming to provide the calculated shaft torque value for the coupling shaft also uses the radial motion parameter to provide the calculated corrected torque value.

24. The method of claim 21 for calculating the corrected torque value for torque applied to a coupling shaft wherein: the torque measuring system further includes a temperature sensor in data communication with the signal conditioning unit and a second target sensor positioned to monitor the plurality of interleaved targets, the second target sensor in data communication with the signal conditioning unit; upon rotation of the shaft, the second target sensor generates a second data signal; providing data from the temperature sensor to the signal conditioning unit; wherein the calibration data also includes temperature compensation data ; wherein the programming for calculating the raw twist value of the coupling shaft uses data from the first and second target sensors; wherein the signal conditioning unit further includes programming which uses the temperature compensation data, the data from the first and second target sensors to calculate a radial motion parameter between the coupling shaft and the first and second target sensors; and,wherein the programming also uses the temperature data, the radial motion parameter between the coupling shaft and the first and second target sensors, and the stored calibration data from the database to provide the calculated corrected torque value.

Images