a) Discuss the considerations when establishing a data acquisition system (DAS) for use in a condition monitoring (CM) system. You may assume that the CM system being establishing will be retrofitted to a high capital cost and highly utilised machine For the prototype DAS & CM system the CM research group have established the following: . There are 10 existing on-machine analog signals that may be utilised. A further 10 transducer with analog outputs have been added . There are 50+ digital on-machine signals that may be utilised. These are mainly the machine controller signals, along with some switchable machine operator inputs. These signals can be used to identify the current state and configuration of the machine during its programmed operations. A collection of such signals will form CM tests for which the data acquisition and associated signal processing needs to be easily re-configurable. Include the data acquisition aspects, such as sample rate, duration, triggering, validity checks, intervals between records in this section. You may then discuss signal processing aspects in part (b). [12 Marks] b) For the CM system, such as you have discussed in part (a) identify which signal processing methods might typically be made available as a signal processing toolki. For each method provide a brief description and data storage & processing/calculation complexity implications [8 Marks] c) Briefly discuss how recorded CM tests would be compared to the expected results for normal, fault free machine conditions. Where could the concepts of "fault sensitivity' and 'fault efficiency' then be used when refining the CM system design? Mackal Question 2 A condition monitoring (CM) research group are considering the design of a new generation of CM systems for use on their in-water testing of scale-model tidal stream turbines (TST) The researchers expect to obtain signals with significant frequency components at 25 Hz (matching the rotational speed), 5.0 Hz and 75 Hz (for the 3-bladed TST). For simplicity, they expect some relatively small random noise to exist on the experimental signals. a) From an earlier generation of such TST CM systems the experimentally generated TST thrust signals were captured via a data acquisition system with the following parameters: Sample frequency = 50 sample.s-1 Captured signal duration = 87 s Comment, and provide sketches wherever possible, on the amplitude frequency spectrum results. Particularly comment on the resolution and range of the frequency axis. Suggest a relatively simple modification to the captured signal duration [10 Marks] b) From the same earlier generation of such TST CM systems the experimentally generated TST torque signals were captured via a legacy data acquisition system with the following parameters: Sample frequency = 10 sample s-1 Captured signal duration = 30 s In this case the CM researchers wished to focus solely on the 7.5 Hz expected frequency component Comment, and provide sketches wherever possible, on the amplitude frequency spectrum results Particularly comment on any deficiencies for the stated setup c) When considering the development of (slowly) progressive TST blade faults, the CM researchers expect the following changes to the normal, fault free frequency spectrums. . The 2.5 Hz component will grow in amplitude as the fault progresses The 5.0 Hz component will not change significantly The 7.5 Hz component will reduce in amplitude, but less significantly than the increase in the 2.5 Hz component. Sketch and compare the predicted frequency spectrum plots, along with a residuals plot. For this case you may assume that the DAS provides sufficient and synchronised data for the Fast Fourier Transform (FET) processing Random noise at 5% of the signal component amplitudes is expected to be present on the signals. [8 Marks] 25 Marks a) Discuss the considerations when establishing a data acquisition system (DAS) for use in a condition monitoring (CM) system. You may assume that the CM system being establishing will be retrofitted to a high capital cost and highly utilised machine For the prototype DAS & CM system the CM research group have established the following: . There are 10 existing on-machine analog signals that may be utilised. A further 10 transducer with analog outputs have been added . There are 50+ digital on-machine signals that may be utilised. These are mainly the machine controller signals, along with some switchable machine operator inputs. These signals can be used to identify the current state and configuration of the machine during its programmed operations. A collection of such signals will form CM tests for which the data acquisition and associated signal processing needs to be easily re-configurable. Include the data acquisition aspects, such as sample rate, duration, triggering, validity checks, intervals between records in this section. You may then discuss signal processing aspects in part (b). [12 Marks] b) For the CM system, such as you have discussed in part (a) identify which signal processing methods might typically be made available as a signal processing toolki. For each method provide a brief description and data storage & processing/calculation complexity implications [8 Marks] c) Briefly discuss how recorded CM tests would be compared to the expected results for normal, fault free machine conditions. Where could the concepts of "fault sensitivity' and 'fault efficiency' then be used when refining the CM system design? Mackal Question 2 A condition monitoring (CM) research group are considering the design of a new generation of CM systems for use on their in-water testing of scale-model tidal stream turbines (TST) The researchers expect to obtain signals with significant frequency components at 25 Hz (matching the rotational speed), 5.0 Hz and 75 Hz (for the 3-bladed TST). For simplicity, they expect some relatively small random noise to exist on the experimental signals. a) From an earlier generation of such TST CM systems the experimentally generated TST thrust signals were captured via a data acquisition system with the following parameters: Sample frequency = 50 sample.s-1 Captured signal duration = 87 s Comment, and provide sketches wherever possible, on the amplitude frequency spectrum results. Particularly comment on the resolution and range of the frequency axis. Suggest a relatively simple modification to the captured signal duration [10 Marks] b) From the same earlier generation of such TST CM systems the experimentally generated TST torque signals were captured via a legacy data acquisition system with the following parameters: Sample frequency = 10 sample s-1 Captured signal duration = 30 s In this case the CM researchers wished to focus solely on the 7.5 Hz expected frequency component Comment, and provide sketches wherever possible, on the amplitude frequency spectrum results Particularly comment on any deficiencies for the stated setup c) When considering the development of (slowly) progressive TST blade faults, the CM researchers expect the following changes to the normal, fault free frequency spectrums. . The 2.5 Hz component will grow in amplitude as the fault progresses The 5.0 Hz component will not change significantly The 7.5 Hz component will reduce in amplitude, but less significantly than the increase in the 2.5 Hz component. Sketch and compare the predicted frequency spectrum plots, along with a residuals plot. For this case you may assume that the DAS provides sufficient and synchronised data for the Fast Fourier Transform (FET) processing Random noise at 5% of the signal component amplitudes is expected to be present on the signals. [8 Marks] 25 Marks