METHOD AND DEVICE FOR CONDITION MONITORING

Description

TECHNICAL FIELD

[0001] The invention concerns condition monitoring of a transmission part of rotating equipment such as a torque transmission part in a system between an electric motor and a load during operation.BACKGROUND

[0002] Traditionally condition monitoring of transmission parts of rotating equipment / machines is done by checking that the load is rotating, turning, or moving. This is usually done by means of one or more sensors reacting to one or more triggering devices. The triggering devices may be optical, physical, or wireless. Either the sensor(s) or the triggering device(s) are mounted on the load, thus the sensor(s) generating a signal every time they pass each other.

[0003] A disadvantage of such a system is that a part of the system has to be mounted on the load, possibly negatively impacting the system and / or the load as to reliability, accuracy, influence, etc. There seems to be room for improvement.

[0004] DE 11 2017 005 650 discloses a power transmission mechanism anomaly diagnosis device that comprises a monitoring and diagnosis unit for determining an anomaly in a power transmission mechanism, as well as a current detector connected to a power supply of an electric motor. The monitoring and diagnosis unit includes an analysis unit for analyzing the current transmitted by the current detector and an anomaly determination unit for determining an anomaly in the power transmission mechanism based on the analysis results from the analysis unit.

[0005] JP 2017-181437 discloses a rotary machine system abnormality detection method based on the operation time current signal of a three-phase induction motor, an abnormality monitoring method using the abnormality detection method, and an abnormality monitoring device using the abnormality monitoring method. This rotary machine system abnormality detection method carries out simple diagnosis that includes: a first step of executing high-speed Fourier transformation for the operation time current signal of a three-phase induction motor; a second step of calculating current spectrum peak values of one or both of side band waves by extracting, from acquired current spectrums, with a current spectrum peak of power supply frequency as the center, the side band waves of the current spectrum peaks of power supply frequencies present at frequency positions respectively separated to a low-frequency side and a high-frequency side by feature frequencies relating to the abnormality of a rotary machine system; a third step of calculating deterioration parameters of the rotary machine system from the current spectrum peak value of the side band waves and the power supply frequencies; and a fourth step of detecting the abnormality of the rotary machine system by recording the values of deterioration parameters and comparing them with a deterioration determination reference value.SUMMARY

[0006] An object of the invention is to define a condition monitoring method and device of a torque transfer part between an electrical motor and a load, without negatively influencing the load, the motor, or the system as such.

[0007] The present invention is defined by the appended claims only, in particular by the scope of appended independent claims. Reference(s) to "embodiments" throughout the description which are not under the scope of the appended claims merely represents possible exemplary executions and are therefore not part of the present invention.

[0008] The aforementioned object is achieved by a condition monitoring method and device that samples torque values, such as a Q-current in a FOC of an electric motor, in relation to motor rotor rotation. Thereafter performing a Fourier transformation of the sampled torque values / Q-current samples from a predefined number of motor rotor rotations for analysis to thereby determine the existence of load variations in sync with the rotation of the load or not. Theoretically there are perfect loads that do not induce load variation over a full turn of the load, however, in practice, loads, especially larger loads, will have some unbalance that can be detected by the torque variations needed to keep a constant rotational speed of the load. If there are no load variations correlated to the revolutions of the load, then it is determined that the torque transfer part is broken or otherwise damaged to properly transfer torque as intended.

[0009] The aforementioned object is also achieved according to the invention by a method of condition monitoring a torque transfer part by means of measuring torque values / Q current of a field oriented controlled electric motor. According to the invention the method further comprises measuring, Fourier transforming, analyzing, correlating and determining. Measuring involves measuring the torque values / Q-current in relation to a rotor of the electric motor for a predetermined number of turns of the rotor. Fourier transforming involves Fourier transforming measured torque / Q-current samples over the predetermined number of rotor turns. Analyzing involves analyzing the Fourier transformed content as to a condition of the torque transfer part by correlating the transformed content with a full revolution of the load. Finally, the determining involves determining the condition of the torque transfer part based on the analysis of the Fourier transformed content.

[0010] The aforementioned object is also achieved according to the invention by a method of condition monitoring a torque transmission part transferring torque between an electric motor and a load. The method comprises measuring a signal representative of a torque value of the electric motor. According to the invention the method further comprises a first step of determining, a second step of determining, a third step of Fourier transforming, a fourth step of analyzing and a fifth step of determining. The first step of determining, determines the torque value at a motor rotor position of the electric motor at a predetermined rotor revolution interval of the electric motor. The second step of determining then determines the torque value for a predetermined number of rotor revolutions of the electric motor. The third step of Fourier transforming, Fourier transforms the determined torque values for the predetermined number of rotor revolutions of the electric motor at the motor rotor position at the predetermined rotor revolution interval of the electric motor. Then in the fourth step of analyzing, analyzes the Fourier transformed content as to a condition of the torque transmission part. Then finally in the fifth step of determining, a condition of the torque transmission part is determined based on one or more analysis of one or more Fourier transformed content in the fourth step.

[0011] In some versions the method of condition monitoring is intended to function where the electric motor is controlled by a Field Oriented Controller (FOC). The signal representative of a torque value of the electric motor is then a Q-current signal of the Field Oriented Controller. Suitably the Field Oriented Controller is a sensorless Field Oriented Controller, which can then provide a signal representative of the motor rotor position. In other versions the method of condition monitoring is intended to function with only the parts of a sensorless Field Oriented Controller that generates a Q-current and a signal representative of the motor rotor position, and where the control of the electric motor is done in an alternative manner.

[0012] The determining of the torque value can suitably range from once every plurality of motor rotor revolutions to a plurality of torque value determining per motor rotor revolution. In some versions the number of determined torque values for the predetermined number of rotor revolutions, are 2 N< , where N is an integer. Sometimes it is advantageous if sampling the torque value is done at a higher rate than the rate of determining the torque value, suitably then the determining of the torque value is based on an average of the sampled torque values done between determining the torque value.

[0013] The motor rotor position of where torque values are determined are suitably one or more predetermined intervals or motor rotor positions around a rotor rotation. Suitably the analyzing of the Fourier transformed content is done by looking at values at cycles corresponding to a load cycle, to see if there are torque variations. The fifth step of determining a condition of the torque transmission part is dependent on the existence or not of torque variations at a load cycle, one load rotation. If there are no load variations correlated to the revolutions of the load, then it is determined that the torque transfer part is broken or otherwise damaged to properly transfer torque as intended.

[0014] A load may be a rotating heat exchanger that is mainly used to recover energy from exhaust air and suitably configured to heat incoming colder air using the recovered energy. Such a load could rotate at a relatively slow speed, especially in relation to a rotational speed of an electric motor used to drive it. A load can alternatively be a part of a fan, a pump, a compressor, a driving drum of a conveyor belt system, an escalator, a centrifuge, an agitator system, a saw, a belt or rotary sander, a planer, a stone crusher, or other devices and systems comprising a rotating load or indirectly a linear load. A torque transmission part from an electric motor to a load, can be a belt, such as a round belt, a toothed belt, a flat transmission belt, a V-belt, a hexagonal transmission belt, a synchronous belt, or the like such as a chain.

[0015] The different additional enhancements of the method of condition monitoring a machine part according to the invention can be combined in any desired manner as long as no conflicting features are combined.

[0016] The aforementioned object is also achieved according to the invention by a condition monitoring device to monitor a condition of a torque transmission part transferring torque between an electric motor and a load. The device comprises a torque sampler that samples a signal representative of a torque value of the electric motor. According to the invention the torque sampler determines the torque value at a motor rotor position of the electric motor at a predetermined rotor revolution interval of the electric motor. The torque sampler further determines the torque value for a predetermined number of rotor revolutions of the electric motor. The condition monitoring device further comprises a signal processor, an analyzer and a decision unit. The signal processor is arranged to Fourier transform the determined torque values for the predetermined number of rotor revolutions of the electric motor at the motor rotor position or positions at the predetermined rotor revolution interval of the electric motor. The analyzer is then arranged to analyze the Fourier transformed content as to a condition of the torque transmission part. Finally, the decision unit is arranged to determine a condition of the torque transfer part based on one or more analysis of one or more Fourier transformed contents. If there are no load variations correlated to the revolutions of the load, then it is determined that the torque transfer part is broken or otherwise damaged to properly transfer torque as intended.

[0017] In some embodiments the device further is arranged to cooperate with an electric motor that is controlled by a Field Oriented Controller (FOC). The signal representative of a torque value of the electric motor is then a Q-current of the Field Oriented Controller. Suitably the Field Oriented Controller is a sensorless Field Oriented Controller, also providing a signal representative of the motor rotor position. In other embodiments the device only further comprises the parts of a sensorless Field Oriented Controller that generates a Q-current and a signal representative of the motor rotor position, and where the control of the electric motor is done in an alternative manner.

[0018] In some embodiments the torque sampler can suitably determine the torque value in different ranges from once every plurality of motor rotor revolutions to a plurality of torque value determinations per motor rotor revolution. Advantageously the number of determined torque values the torque sampler samples during the predetermined rotor revolutions, are 2 N< , where N is an integer, to thereby facilitate the DFT / FFT that follows. In some embodiments the torque sampler samples the torque value at a higher rate than the rate of determining the torque value, the determining of the torque value is then suitably based on an average of the sampled torque values done between determining the torque value.

[0019] The positions / intervals around a motor rotor of where torque values are determined can be one or more predetermined positions / intervals around a rotor rotation. Suitably the analyzer analyzes the Fourier transformed content by looking at values at cycles corresponding to a load cycle, a load rotation, to see if there are torque variations or not. The decision unit can preferably determine a condition of the torque transmission part in dependence on the existence or not of torque variations at a load cycle.

[0020] A load may be a rotating heat exchanger. A rotating heat exchanger is mainly used to recover energy from exhaust air. The recovered energy is then used to heat incoming colder air. Such a load could rotate at a relatively slow speed, especially in relation to a rotational speed of an electric motor driving it. A load can alternatively be a part of a fan, a pump, a compressor, a driving drum of a conveyor belt system, an escalator, a centrifuge, an agitator system, a saw, a belt or rotary sander, a planer, a stone crusher, and other devices and systems comprising a rotating load or indirectly a linear load, and a torque transmission part from an electric motor to a load. A torque transmission part may be a belt, such as a round belt, a toothed belt, a flat transmission belt, a V-belt, a hexagonal transmission belt, a synchronous belt, or the like such as a chain.

[0021] The different additional enhancements of the condition monitoring device according to the invention can be combined in any desired manner as long as no conflicting features are combined.

[0022] Other advantages of this invention will become apparent from the detailed description.BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The invention will now be described in more detail for explanatory, and in no sense limiting, purposes, with reference to the following figures, in which Fig. 1illustrates a schematic diagram of the invention in use, Fig. 2illustrates an example of a sensorless Field Oriented Control (FOC) for Permanent Magnet Synchronous Motors (PMSM) suitable to be used in some embodiments of the invention, Fig. 3illustrates schematically a signal processing flow according to the invention, Fig. 4illustrates a flowchart of the signal processing according to the invention, Fig. 5illustrates a flowchart of a forced condition monitoring according to one embodiment according to the invention. DETAILED DESCRIPTION

[0024] In order to clarify the method and device according to the invention, some examples of its use will now be described in connection with Figures 1 to 5.

[0025] Figure 1 illustrates a schematic diagram of the invention in use. The system comprises an electric motor 101, coupled with a belt pulley 103 and via a torque transmission part 105 coupled to a load 107. In this example the illustrated load 107 is a heat exchanger rotor in a rotary heat exchanger mainly used to recover energy from exhaust air and thereby heat incoming air. These rotate relatively slow and a relationship of 64:1 or more between motor rotor revolutions and load revolutions is not unusual. The load can also be a part of a fan, a pump, a compressor, a driving drum of a conveyor belt system, an escalator, a centrifuge, an agitator system, a saw, a belt or rotary sander, a planer, a stone crusher, and other devices and systems comprising a rotating load and / or indirectly a linear load. The load has to have some irregularity or be provided with one, that will cause / require torque variations from the electric motor to keep the load at a constant desired speed around a revolution / cycle of the load. The system is also suitable for very large loads that rotate relatively slow in relation to an electric motor driving the load. Large loads, such as larger heat exchanger rotors are relatively weak constructions in relation to their size, and are therefore very difficult to manufacture without unbalances and with perfect roundness. These imperfections will cause a motor controller to have to vary a torque to keep the imperfect load at a constant desired speed. These torque variations are repeated synchronously with the load cycle, one complete load revolution. The torque transmission part 105 can for example be a belt, a round belt, a toothed belt, a flat transmission belt, a V-belt, a hexagonal transmission belt, a synchronous belt, a chain, a toothed coupling, a direct coupling between the motor and the load or the like.

[0026] The electric motor 101 is driven by a control unit 110 by means of a motor drive output 156. In addition to an electric motor controller, the control unit also comprises a condition monitoring unit according to the invention. The condition monitoring unit monitors the condition of the torque transmission part 105 and provides a condition monitoring output 150 from the control unit. The condition monitoring output 150 can be an indicator light in addition to a connection with a supervisor system that will determine if the system should be shut down and an alarm given or not. This can also be completely or in part built into the condition monitoring unit. The condition monitoring unit is dependent on a signal representing the electric motor torque, which signal can be from a sensor or as in a preferred embodiment from an available Q-current signal from a preferred electric motor controller embodiment based on a sensorless Field Oriented Control (FOC). The control unit 110 further comprise a desired motor speed input 154 and a power input 152. A motor control using FOC needs a minimum rotational speed so that a sufficient back-emf is generated. This back-emf is necessary for FOC to properly function. Some implementations of the invention require that the electric motor is controllable also below the minimum rotational speed, and in such implementations the motor controller is able to switch between FOC and an alternative low RPM alternative such as traditional frequency controlled sinus waveforms.

[0027] Figure 2 illustrates an example of a sensorless Field Oriented Control (FOC) for Permanent Magnet Synchronous Motors (PMSM) suitable to be used in some embodiments of the invention. The sensorless FOC can be seen as having a desired speed input 242 and a resulting three phase inverter 231 correctly driving 234 an electric motor 201 at that speed. Connected thereto is a control unit 211 that comprises a condition monitoring unit according to the invention. The condition monitoring unit in this case requires the Q-current 276 as a measure of the electric motor torque that is required to drive the load at the desired speed 242. The condition monitoring unit further requires the rotor position 284, to be able to sample the Q-current at the same position around the rotor of the motor 201. A sensorless FOC comprises a Position and Speed estimator 280 that uses V α 262, V β 264, i α 272, and i β 274, to be able to calculate rotor speed 282 and rotor position 284. The raw rotor position signal 284 that is provided is the electrical position that has to be converted to mechanical position, or at least taken into account. Further details about the different blocks can be found below and in other literature.

[0028] Figure 3 illustrates schematically a signal processing flow according to the invention. The signal processing needs both the motor rotor torque and the motor rotor position to be able to sample the torque in relation to the position of the motor rotor. A sensorless FOC control 312 is a suitable motor controller as it can both generate a torque related signal, a Q-current 361, and also the required rotor position 362. The Q-current 361 is then low pass filtered 320, generating a low pass filtered Q-current signal 363. Thereafter sampling and possibly down-sampling 322 of the torque / Q-current, depending on type of sampling, correlation of the sampling with the motor rotor rotations using the position signal 362. Individual samples 365 are then collected into a vector in the next stage 324. Sampling / down sampling 324 of a for the embodiment, adequate number of samples for an adequate number of motor rotor revolutions. In an example there is a rotational ratio of 64:1 between the motor and the load. The Q-current can then suitably be sampled twice per motor rotor revolution, for a total of 64 revolutions, thus resulting in 128 samples 367. The set of samples 367 are then put through a Fourier transformation 326 in relation to motor rotor position, thus making it speed independent. In the current example we would get a data set 369 of 64 data points representing frequency of load variations at different rotational ratios 328. The data points represent load variation frequencies at different rotational ratios in relationship to the motor rotor. The first gives a DC value, an average. The following are the rotational relationships to the motor rotor 1 / N, 2 / N, 3 / N ..... (N-1) / N, where N is the number of data points. In our example we have 64 resulting data points. 1 / N is then the frequency of load variations having a rotational relationship with the motor rotor as 1 / 64, which is one rotation of the load since the motor to load rotational relationship is 64:1 in this simplified example. What is looked for is if there are load variations having a repeatability equal to one cycle, one rotation, of the load. If there are load variations that have the same cycle as a load cycle, then there is a high probability that the torque transfer part works as it should mirroring back to the motor control any load variations during a revolution of the load. If there is a lack of load variations being cyclic with a load rotation / cycle, then there can be a cause for worry that the torque transfer part is completely or partially broken / damaged.

[0029] The motor to load relationship might be known and programmable / preprogrammed into the motor control / condition monitoring unit. In other implementation there might be an automatic or on demand seek function in the software that identifies the relationship. In still further embodiments the condition monitoring unit might be set up to accept a range of relationships between motor and load rotations, for example 200:1 to 10:1, thus examining a wider range of bins 328, resulting data points 369, to search for cyclic load variations.

[0030] Figure 4 illustrates a more traditional flowchart of a signal processing flow according to the invention following the torque / Q-current. First there is a low pass filtering 410 of the Q-current. The Q-current is sampled 420 at determined motor rotor location(s) / position(s), this can be at one or more positions on a motor rotor rotation, or at one position every two or three motor rotor rotations. There has to be enough sampling done, this is tested 430. If there are not enough then process is returned to step 410. If there are enough samples then the process continues with step 440, that Fourier transforms the sampled set of Q-current samples. Suitably by at least DFT, preferably by FFT, or some other suitable method. When the Fourier transform has been done, then there is first an analysis 450 of the result of the FFT. This analysis is to find the relevant bins with an indication that there are torque variations appearing in relation to the load cycle, or multiples / divisions thereof. In a final step 460 there is a determination of the implications of the analysis 450 and suitably a communication of this to one or more users and / or to a database / further processing.

[0031] As mentioned previously, a FOC controlled electric motor cannot be regulated to very low RPM:s where there is not enough reverse EMF for the FOC to be able to function properly. If the condition monitoring is done in a sensorless FOC environment, then there cannot be any condition monitoring done if there is no Q-current to sample. Figure 5 illustrates a flowchart of forced condition monitoring according to one embodiment according to the invention. A forced condition monitoring is done when there has not been any condition monitoring done within a predetermined time interval. The process starts in a first step 500. Variables and timers are appropriately set and reset, such as timer t=0 in a second step 510. The timer t is increased after a predetermined time interval, t=t+1, in a third step 520. It is tested in a fourth step 530 if the timer value t is greater than a preset time t1. If no, then the process returns to the third step 520 to keep on increasing the timer t after the predetermined time interval. If the time t value is greater than a preset timer value t1, then the process continues with a fifth step 540. The fifth step 540 tests if the motor control is in FOC mode or in a in another mode, such as conventional frequency controlled sinus wave mode? If the motor control is in FOC mode, then torque, Q-current, is available and the process then continues with a seventh step 560. On the other hand, If the motor control is in another mode, such as a frequency controlled sinus wave mode, then the process will continue with step 550. Step 550 increases speed / motor rotor revolutions per minute for the purpose of forcing the motor control to enter FOC mode. After a motor speed increase the process goes back to step 540 to see if the motor control has entered FOC mode or if a further increase in motor speed is necessary.

[0032] Once it has been established that the motor control is in FOC mode, then a condition monitoring check is done with the help of the now available Q-current as a signal indicative of the motor torque in step 560. That is a condition monitoring of the torque transfer part is done by a complete measurement of Q-current during a predetermined number of motor rotor revolutions, FFT, analyzes and then a determination if the torque transmission part is functioning or not. If in step 570 it is determined that the torque transmission part is functioning, then the process will continue with step 580. If it is determined in step 570 that the torque transmission part is not functioning as expected, then the process continues with step 575. Step 575 keeps track of how many condition monitoring failures there are, if condition monitoring failures consecutive or if there are condition monitoring passes in-between, the time between failures etc. The decision process is in many embodiments handled elsewhere where it is determined if a shutdown is motivated or not depending many factors. In some embodiments these decisions are taken by pre-programmed routines in the condition monitoring unit. Finally, in step 580, the condition monitoring is finished for this time and the motor speed is returned to a set desired speed value and the process is then continued with step 510.

[0033] The invention is not restricted to the above-described embodiments, but may be varied within the scope of the following claims.

[0034] Figure 1 illustrates a schematic diagram of the invention in use: 101An electric motor, 103A belt pulley, toothed pulley, rubber wheel and the like coupled to the electric motor 101, 105A torque transmission part from electric motor to a load, such as a belt, a toothed belt, a flat transmission belt, a V-belt, a hexagonal transmission belt, a synchronous belt, a chain, a toothed coupling, a direct coupling between the motor and the load or the like, 107A load, such as a heat exchanger rotor in a rotary heat exchanger mainly used to recover energy from exhaust air and thereby heat incoming air, the load can also be a part of a fan, a pump, a compressor, a driving drum of a conveyor belt system, an escalator, a centrifuge, an agitator system, a saw, a belt or rotary sander, a planer, a stone crusher, and other devices and systems comprising a rotating load or indirectly a linear load, 110A control unit comprising the condition monitoring unit according to the invention and an electric motor controller comprising an available Q-current, in a user friendly embodiment the control unit comprises power input, a desired speed input, a motor drive output, and a condition monitoring output, 150A condition monitoring output from the control unit, 152A power input to the control unit, 154A desired motor speed input to the control unit, 156A motor drive output from the control unit.

[0035] Figure 2 illustrates an example of a sensorless Field Oriented Control (FOC) for Permanent Magnet Synchronous Motors (PMSM) suitable to be used in some embodiments of the invention: 201An electric motor, 211A control unit comprising the condition monitoring unit according to the invention, inputs are calculated speed 242 and Q current, i q , and the output is condition monitoring, 230A Pulse Width Modulation (PWM) unit for motor control, 231A three phase inverter, 232Control signals from PWM to the three phase inverter, 234Drive connections to the electric motor, 236Measurement of i a , from the drive connections to the electric motor, 238Measurement of i b , from the drive connections to the electric motor, with two known, i a and i b , then the third i c can be calculated, 240A speed PID controller, inputs are the desired speed 242 and the speed 242 as calculated in this embodiment, output i a the desired i q (desired Q current) 244, 242A desired speed, 243D current, i d , PID controller, inputs are desired i d 249, that is set to zero, and a measured i d 278, an output is V d 248, 244A desired Q current, i q , 245A Q-current, i q , PID controller, inputs are a desired i q 244 and a measured i q 276, output is V q 246, 246V q , 248V d , 249A desired D current, i d , which is set to zero in some embodiments as the one shown, in some embodiments a small amount of D current gives some torque margins and can contribute to stabilize the motor, 250A condition monitoring output from the control unit, 260An inverse Park transformation, from V q and V d , to V α and V β , inputs are V q 246, V d 248 and rotor position 284, outputs V α 262 and V β 264, 262V α , 264V β , 270A Clarke transformation, inputs are i a 236 and i b 238, outputs are i α 272 and i β 274, 272i α , 274i β , 275A Park transformation, inputs are i α 272, i β 274 and rotor position 284, outputs are i d 278 and i q 276 276i q , 278i d , 280Position and Speed estimator, inputs V α 262, V β 264, i α 272, and i β 274, outputs are calculated speed 282 and calculated rotor position 284, 282Calculated speed, 284Calculated rotor position.

[0036] Figure 3 illustrates schematically a signal processing flow according to the invention: 312Suitably a sensorless FOC control generating a Q-current, 320Low pass filtering of the Q-current, 322Sampling and possibly down-sampling of Q-current, depending on type of sampling, correlation of the sampling with the motor rotor rotations using a rotor position signal, 324Sampling / down sampling of a for the embodiment, adequate number of samples for an adequate number of motor rotor revolutions, for example two samples per revolution, for 64 revolutions, 326Fourier transform, such as FFT, DFT or similar Fourier transformation, 328Binning and analysis of the result of the FFT, 361Q-current, 362motor rotor position, 363Filtered Q-current, 365Q-current samples, 367Q-current samples ready for Fourier transformation, 369Result of Fourier transform, frequency of load variations.

[0037] Figure 4 illustrates a flowchart of a signal processing flow according to the invention: 410Low pass filtering of Q-current, 420Sampling of Q-current, 430Is there enough samples been acquired over enough motor rotor rotations for the FFT of the current embodiment?? If No, then return to 410, if Yes, then continue with 440, 440Fourier transform, such as FFT, DFT or similar Fourier transform, 450Analysis of the result of the FFT, 460Determination of the implications of the analysis and communication of this to one or more users and / or to a database / further processing.

[0038] Figure 5 illustrates a flowchart of forced condition monitoring according to one embodiment according to the invention: 500Start 510Reset timer t=0, 520Increase timer t=t+1, 530Is the timer value t greater than a preset time t 1 ? If No, then return to step 520 to keep on increasing the timer t, if Yes, then continue with step 540, 540Is the motor control in FOC mode or in a sinus mode? If in FOC mode, then continue with step 560, If in sinus mode, then continue with step 550, 550Increase speed / motor rotor revolutions per minute, then go to step 540, 560Make a complete measurement of Q-current during a predetermined number of motor rotor revolutions, FFT, analyzed and determine if the torque transmission part is functioning or not, 570Is it determined that the torque transmission part is functioning, then continue with step 580, If it is determined that the torque transmission part is not functioning as expected, continue with step 575, 575Add one to the alarm count, this is handled elsewhere where it is determined if a shutdown is motivated or not depending on the alarm count within a predetermined time interval and / or if a predetermined number of consecutive alarms is determined, 580The condition monitoring is finished for this time and the motor speed is returned to the desired speed value and the process is continued with step 510.

Claims

1. A method of condition monitoring a torque transmission part (105) transferring torque between an electric motor (101) and a load (107), the method comprises measuring a signal representative of a torque value of the electric motor (101), determining the torque value at a position along a motor rotor of the electric motor (101) at a predetermined rotor revolution interval of the electric motor (101), characterized in that the method further comprises: - determining the torque value for a predetermined number of rotor revolutions of the electric motor (101), - Fourier transforming the determined torque values in relation to the motor rotor position for the predetermined number of rotor revolutions of the electric motor (101) at the position along the motor rotor at the predetermined rotor revolution interval of the electric motor (101), - analyzing the Fourier transformed content as to a condition of the torque transmission part (105), - determining a condition of the torque transmission part (105) based on one or more analysis of one or more Fourier transformed content, and wherein the analyzing of the Fourier transformed content is done by looking at values at cycles corresponding to a load cycle, to see if there are torque variations.

2. The method of condition monitoring according to claim 1, wherein a load (107) is a rotating heat exchanger mainly used to recover energy from exhaust air and thereby heat incoming colder air, such a load could rotate at a slow speed in relation to a rotational speed of an electric motor (101), and wherein a torque transmission part (105) from an electric motor (101) to a load (107), comprises one of a belt, a round belt, a toothed belt, a flat transmission belt, a V-belt, a hexagonal transmission belt, a synchronous belt, a chain, or the like.

3. The method of condition monitoring according to claim 1 or 2, wherein the electric motor (101) is controlled by a Field Oriented Controller (FOC) and that the signal representative of a torque value of the electric motor (101) is a Q-current of the Field Oriented Controller, and / or where the Field Oriented Controller is a sensorless Field Oriented Controller, providing a signal representative of the motor rotor position.

4. The method of condition monitoring according to any one of claim 1 to 3, wherein the determining of the torque value ranges from once every plurality of motor rotor revolutions to a plurality of torque value determining per motor rotor revolution, and / or where the determined torque values for the predetermined number of rotor revolutions, are 2N, where N is an integer.

5. The method of condition monitoring according to any previous claim, wherein sampling the torque value is done at a rate higher rate than the rate of determining the torque value, the determining of the torque value is based on an average of the sampled torque values done between determining the torque value, and / or where the position along the motor rotor of where torque values are determined are one or more predetermined intervals or positions along the motor rotor around the rotor.

6. The method of condition monitoring according to any previous claim, wherein determining that if there are no torque variations correlated to the load cycles, then the torque transfer part is broken or otherwise damaged to properly transfer torque as intended.

7. The method of condition monitoring according to any previous claim, wherein the determining a condition of the torque transmission part (105) is dependent on the existence of torque variations at a load cycle.

8. A condition monitoring device to monitor a condition of a torque transmission part (105) transferring torque between an electric motor (101) and a load (107), the device comprising a torque sampler, sampling a signal representative of a torque value of the electric motor (101), characterized in that the torque sampler is arranged to determine the torque value at a motor rotor position of the electric motor (101) at a predetermined rotor revolution interval of the electric motor (101), and where the torque sampler further is adapted to determine the torque value for a predetermined number of rotor revolutions of the electric motor(101), and in that the condition monitoring device further comprises: - a signal processor arranged to Fourier transform the determined torque values in relation to the motor rotor position for the predetermined number of rotor revolutions of the electric motor (101) at the motor rotor position or positions at the predetermined rotor revolution interval of the electric motor (101), - an analyzer arranged to analyze the Fourier transformed content as to a condition of the torque transmission part (105), - a decision unit arranged to determine a condition of the torque transfer part based on one or more analysis of one or more Fourier transformed content wherein the analyzer analyzes the Fourier transformed content by looking at values at cycles corresponding to a load cycle.

9. The condition monitoring device according to claim 8, wherein a load (107) is a rotating heat exchanger mainly used to recover energy from exhaust air and thereby heat incoming colder air, such a load could rotate at a slow speed in relation to a rotational speed of an electric motor (101), and wherein a torque transmission part (105) from an electric motor (101) to a load (107), comprises one of a belt, a round belt, a toothed belt, a flat transmission belt, a V-belt, a hexagonal transmission belt, a synchronous belt, a chain, or the like.

10. The condition monitoring device according to claim 8 or 9, wherein the electric motor (101) is arranged to be controlled by a Field Oriented Controller (FOC) and that the signal representative of a torque value of the electric motor (101) is a Q-current of the Field Oriented Controller, and / or the Field Oriented Controller is a sensorless Field Oriented Controller, providing a signal representative of the motor rotor position.

11. The condition monitoring device according to any one of claim 8 to 10, wherein the torque sampler is arranged to determine the torque value in ranges from once every plurality of motor rotor revolutions to a plurality of torque value determinations per motor rotor revolution, and / or the determined torque values for the predetermined number of rotor revolutions, are 2N, where N is an integer.

12. The condition monitoring device according to any one of claims 8 to 11, wherein the torque sampler is arranged to sample the torque value at a higher rate than the rate of determining the torque value, the determining of the torque value is based on an average of the sampled torque values done between determining the torque value.

13. The condition monitoring device according to any one of claims 8 to 12, wherein the position along the motor rotor of where torque values are determined are one or more predetermined intervals or motor rotor positions around a rotor rotation.

14. The condition monitoring device according to any one of claims 8 to 13, wherein the decision unit is arranged to determine a condition of the torque transmission part (105) in dependence on the existence of torque variations at a load cycle.

15. The condition monitoring device according to any of the claims 8 to 14, wherein the decision unit is arranged to determine that if there are no torque variations correlated to the load cycles, then the torque transfer part is broken or otherwise damaged to properly transfer torque as intended.