Thesis (M.B.A.)--University of Hong Kong, 1985.

Analysis was performed to quantify and compare the reliability of several different notional shipboard DC distribution system topologies in serving their equipment loads. Further, the relationship between the relative placement of loads and generators within a distribution system and the system’s reliability was investigated, resulting in an algorithmically-derived optimal placement configuration in the system topology found to be the most reliable in the initial analysis. Using Markov models and fault-tree analysis, system reliability indices were derived from distribution system component reliability indices, and these values were compared between competing topologies and equipment configurations. A distribution system based on the breaker-and-a-half topology often used in terrestrial utility substations was found to be superior in terms of reliability to the currently-standard ring bus topology. Expected rates of service interruptions to equipment systems served by the breaker-and-a-half system were reduced overall, in some cases dropping dramatically to less than one expected interruption per 10,000 years. This improvement, however, came at the expense of requiring more circuit breakers in the distribution system’s construction. Within this breaker-and-a-half distribution system, an optimal placement of loads and generators was algorithmically derived, which further improved the reliability of the system. This improvement over the base case was marginal, but the optimized placement configuration was able to reduce the expected interruption rate of the ship’s radar system by over 40%. / text

published_or_final_version / Mechanical Engineering / Master / Master of Philosophy

No description available.

Naval submarines and surface ships are regularly subjected to a treatment called &quotdeperming&quot that seeks to design the vessel???s permanent magnetisation for optimal magnetic camouflage. A scaled model of a magnetic treatment facility (MTF) has been established as a valid system to simulate deperming and used to investigate various aspects of the deperm process including: magnetic anisotropy and demagnetising fields as factors in the physical modelling of magnetism in whole vessels; a comparison of current and alternative deperm procedures; the application of theoretical models of bulk magnetisation to calculate deperm outcomes in the physical model and in actual vessels. A &quotlaboratory MTF&quot was constructed to imitate the applied field geometry at a naval MTF. The system was calibrated and it was determined that the laboratory MTF could make magnetic measurements on a CU200T-G steel bar sample with an equivalent accuracy (error = ??5%) to that of standard magnetometric equipment. Experiments were conducted with emphasis on a holistic approach to modelling the deperm process and describing magnetisation changes in whole objects. The importance of the magnetic anisotropic changes to steel with cold rolling was confirmed. In CU200T-G steel sheet the initial susceptibility (ci) was found to increase by a factor of 3 ??0.1 in the rolling direction, from a value of ~ 110 in the un-rolled steel sheet (thickness dependent). ci in the rolled sheet transverse to the rolling direction was decreased by a factor of 0.94 ??0.09 to ci in the un-rolled sheet steel. Previous studies on hull steel have neglected to account for this transformation through cold work. The demonstration on mild steel here is expected to have an analogy in the final state of the hull sheet steel as it resides in a submarine pressure hull. Future studies either on hull material or on modelling whole vessels should include the same or similar magnetic anisotropic properties in the steel(s) under investigation. Hollow circular tubes made from CA2S-E and CU200T-G steel sheet were selected as models for vessels. It was shown that these steel tubes were a good choice in this regard: minimising the complexity of the experiment whilst maintaining the validity of a deperm simulation. During a deperm there was an excellent qualitative likeness in the permanent longitudinal magnetisation (PLM) for the steel tubes to PLM in both a submarine and a surface vessel. Permanent vertical magnetisation (PVM) deperm results from the tubes displayed a close qualitative match with PVM in a submarine but not in a surface vessel. A theoretical treatment for demagnetisation factors (Nd) in hollow ellipsoids was used in conjunction with a geometrical approximation to calculate Nd for finite hollow objects of revolution. Subsequent theoretical calculations correlated well with experimental results for measured effective ci (ceff) in hollow circular CU200T-G steel tubes of various lengths and aspect ratios. Using an estimate of 100 as ci for submarine hull steel, the same analysis produces Nd for the axial and transaxial directions in a submarine equal to 5.97??10-3 and 0.0142 respectively. Three items for potential improvement were identified in the current deperm protocol used on naval vessels (Flash-D): redundancy in the protocol; the duration of the deperm and a theoretical basis for predicting the final magnetisation or changes in magnetisation during a deperm. Simulations of a novel &quotanhysteretic deperm&quot method, designed to combat these issues, compared favourably to the Flash-D protocol. The standard deviation (s) of the final PVM from 30 Flash-D deperms on steel tubes was 206 A/m; for the final PVM from 30 anhysteretic deperms of the same duration, this was 60 A/m. The s for the final PLM for Flash-D and anhysteretic deperms of the same duration were 416 A/m and 670 A/m respectively. The conclusion is that adopting the anhysteretic deperm on actual vessels would improve the reliability of the PVM outcome. Though the procedure would demand the same duration as Flash-D, there is the advantage of saving time by not having to repeat deperms to obtain the desired result. Additionally the anhysteretic deperm is considerably more amenable to theoretical analysis. A modified version of Langevin???s equation was used to predict the final PLM and PVM results for anhysteretic deperms and to provide a useful analysis of the anhysteretic processes in the Flash-D procedure. Using a Preisach analysis of hysteresis, a mathematical description of bulk magnetic changes that occur to a specific object, within a deperm, has been developed. Theoretical calculations of PLM in a steel tube during and after both types of deperm are in excellent agreement with experimental data. The same theoretical approach was also used to retrospectively model PLM results from previous Flash-D deperms on a submarine with equal success. With this analysis it is proposed that anhysteretic deperm outcomes could be predicted a priori. The influence of magnetic cargo on hull magnetisation was demonstrated to be of significance during and after deperming. &quotSympathetic deperming&quot occurs where a magnetic source is located close to the hull during a deperm. It was found that a vessel or model vessel hull could still be demagnetised even when they contain magnetic cargo that would normally resist the direct application of the same magnetic fields. This was explained using the principles of demagnetising fields and anhysteretic magnetisation. A possible explanation was provided for a PVM measurement anomaly common to the model and vessel deperm results. From measurement, alternating longitudinal applied fields apparently induce corresponding changes in the PVM. This effect could be explained by the depermed object being offset longitudinally from the position expected by the measurement system. This offset could be estimated using an analysis of the changes to PLM and PVM after a longitudinal applied field. The offset displacements calculated for the vessels were too small to be verified experimentally (&gt 0.1m), but the predicted offset for the steel tubes coincided with the limit of precision for their placement in the laboratory MTF = 0.5mm The aim of this work was to look at the deperm process with reference to a system that demonstrated qualitative similarities to deperms on actual vessels. The laboratory MTF is a unique facility, permitting a useful practical analysis of deperming based on sound magnetostatic measurements The experimental and theoretical results gained here have direct application to future deperms on naval vessels with particular reference to submarines.

Naval submarines and surface ships are regularly subjected to a treatment called &quotdeperming&quot that seeks to design the vessel???s permanent magnetisation for optimal magnetic camouflage. A scaled model of a magnetic treatment facility (MTF) has been established as a valid system to simulate deperming and used to investigate various aspects of the deperm process including: magnetic anisotropy and demagnetising fields as factors in the physical modelling of magnetism in whole vessels; a comparison of current and alternative deperm procedures; the application of theoretical models of bulk magnetisation to calculate deperm outcomes in the physical model and in actual vessels. A &quotlaboratory MTF&quot was constructed to imitate the applied field geometry at a naval MTF. The system was calibrated and it was determined that the laboratory MTF could make magnetic measurements on a CU200T-G steel bar sample with an equivalent accuracy (error = ??5%) to that of standard magnetometric equipment. Experiments were conducted with emphasis on a holistic approach to modelling the deperm process and describing magnetisation changes in whole objects. The importance of the magnetic anisotropic changes to steel with cold rolling was confirmed. In CU200T-G steel sheet the initial susceptibility (ci) was found to increase by a factor of 3 ??0.1 in the rolling direction, from a value of ~ 110 in the un-rolled steel sheet (thickness dependent). ci in the rolled sheet transverse to the rolling direction was decreased by a factor of 0.94 ??0.09 to ci in the un-rolled sheet steel. Previous studies on hull steel have neglected to account for this transformation through cold work. The demonstration on mild steel here is expected to have an analogy in the final state of the hull sheet steel as it resides in a submarine pressure hull. Future studies either on hull material or on modelling whole vessels should include the same or similar magnetic anisotropic properties in the steel(s) under investigation. Hollow circular tubes made from CA2S-E and CU200T-G steel sheet were selected as models for vessels. It was shown that these steel tubes were a good choice in this regard: minimising the complexity of the experiment whilst maintaining the validity of a deperm simulation. During a deperm there was an excellent qualitative likeness in the permanent longitudinal magnetisation (PLM) for the steel tubes to PLM in both a submarine and a surface vessel. Permanent vertical magnetisation (PVM) deperm results from the tubes displayed a close qualitative match with PVM in a submarine but not in a surface vessel. A theoretical treatment for demagnetisation factors (Nd) in hollow ellipsoids was used in conjunction with a geometrical approximation to calculate Nd for finite hollow objects of revolution. Subsequent theoretical calculations correlated well with experimental results for measured effective ci (ceff) in hollow circular CU200T-G steel tubes of various lengths and aspect ratios. Using an estimate of 100 as ci for submarine hull steel, the same analysis produces Nd for the axial and transaxial directions in a submarine equal to 5.97??10-3 and 0.0142 respectively. Three items for potential improvement were identified in the current deperm protocol used on naval vessels (Flash-D): redundancy in the protocol; the duration of the deperm and a theoretical basis for predicting the final magnetisation or changes in magnetisation during a deperm. Simulations of a novel &quotanhysteretic deperm&quot method, designed to combat these issues, compared favourably to the Flash-D protocol. The standard deviation (s) of the final PVM from 30 Flash-D deperms on steel tubes was 206 A/m; for the final PVM from 30 anhysteretic deperms of the same duration, this was 60 A/m. The s for the final PLM for Flash-D and anhysteretic deperms of the same duration were 416 A/m and 670 A/m respectively. The conclusion is that adopting the anhysteretic deperm on actual vessels would improve the reliability of the PVM outcome. Though the procedure would demand the same duration as Flash-D, there is the advantage of saving time by not having to repeat deperms to obtain the desired result. Additionally the anhysteretic deperm is considerably more amenable to theoretical analysis. A modified version of Langevin???s equation was used to predict the final PLM and PVM results for anhysteretic deperms and to provide a useful analysis of the anhysteretic processes in the Flash-D procedure. Using a Preisach analysis of hysteresis, a mathematical description of bulk magnetic changes that occur to a specific object, within a deperm, has been developed. Theoretical calculations of PLM in a steel tube during and after both types of deperm are in excellent agreement with experimental data. The same theoretical approach was also used to retrospectively model PLM results from previous Flash-D deperms on a submarine with equal success. With this analysis it is proposed that anhysteretic deperm outcomes could be predicted a priori. The influence of magnetic cargo on hull magnetisation was demonstrated to be of significance during and after deperming. &quotSympathetic deperming&quot occurs where a magnetic source is located close to the hull during a deperm. It was found that a vessel or model vessel hull could still be demagnetised even when they contain magnetic cargo that would normally resist the direct application of the same magnetic fields. This was explained using the principles of demagnetising fields and anhysteretic magnetisation. A possible explanation was provided for a PVM measurement anomaly common to the model and vessel deperm results. From measurement, alternating longitudinal applied fields apparently induce corresponding changes in the PVM. This effect could be explained by the depermed object being offset longitudinally from the position expected by the measurement system. This offset could be estimated using an analysis of the changes to PLM and PVM after a longitudinal applied field. The offset displacements calculated for the vessels were too small to be verified experimentally (&gt 0.1m), but the predicted offset for the steel tubes coincided with the limit of precision for their placement in the laboratory MTF = 0.5mm The aim of this work was to look at the deperm process with reference to a system that demonstrated qualitative similarities to deperms on actual vessels. The laboratory MTF is a unique facility, permitting a useful practical analysis of deperming based on sound magnetostatic measurements The experimental and theoretical results gained here have direct application to future deperms on naval vessels with particular reference to submarines.

Naval submarines and surface ships are regularly subjected to a treatment called &quotdeperming&quot that seeks to design the vessel???s permanent magnetisation for optimal magnetic camouflage. A scaled model of a magnetic treatment facility (MTF) has been established as a valid system to simulate deperming and used to investigate various aspects of the deperm process including: magnetic anisotropy and demagnetising fields as factors in the physical modelling of magnetism in whole vessels; a comparison of current and alternative deperm procedures; the application of theoretical models of bulk magnetisation to calculate deperm outcomes in the physical model and in actual vessels. A &quotlaboratory MTF&quot was constructed to imitate the applied field geometry at a naval MTF. The system was calibrated and it was determined that the laboratory MTF could make magnetic measurements on a CU200T-G steel bar sample with an equivalent accuracy (error = ??5%) to that of standard magnetometric equipment. Experiments were conducted with emphasis on a holistic approach to modelling the deperm process and describing magnetisation changes in whole objects. The importance of the magnetic anisotropic changes to steel with cold rolling was confirmed. In CU200T-G steel sheet the initial susceptibility (ci) was found to increase by a factor of 3 ??0.1 in the rolling direction, from a value of ~ 110 in the un-rolled steel sheet (thickness dependent). ci in the rolled sheet transverse to the rolling direction was decreased by a factor of 0.94 ??0.09 to ci in the un-rolled sheet steel. Previous studies on hull steel have neglected to account for this transformation through cold work. The demonstration on mild steel here is expected to have an analogy in the final state of the hull sheet steel as it resides in a submarine pressure hull. Future studies either on hull material or on modelling whole vessels should include the same or similar magnetic anisotropic properties in the steel(s) under investigation. Hollow circular tubes made from CA2S-E and CU200T-G steel sheet were selected as models for vessels. It was shown that these steel tubes were a good choice in this regard: minimising the complexity of the experiment whilst maintaining the validity of a deperm simulation. During a deperm there was an excellent qualitative likeness in the permanent longitudinal magnetisation (PLM) for the steel tubes to PLM in both a submarine and a surface vessel. Permanent vertical magnetisation (PVM) deperm results from the tubes displayed a close qualitative match with PVM in a submarine but not in a surface vessel. A theoretical treatment for demagnetisation factors (Nd) in hollow ellipsoids was used in conjunction with a geometrical approximation to calculate Nd for finite hollow objects of revolution. Subsequent theoretical calculations correlated well with experimental results for measured effective ci (ceff) in hollow circular CU200T-G steel tubes of various lengths and aspect ratios. Using an estimate of 100 as ci for submarine hull steel, the same analysis produces Nd for the axial and transaxial directions in a submarine equal to 5.97??10-3 and 0.0142 respectively. Three items for potential improvement were identified in the current deperm protocol used on naval vessels (Flash-D): redundancy in the protocol; the duration of the deperm and a theoretical basis for predicting the final magnetisation or changes in magnetisation during a deperm. Simulations of a novel &quotanhysteretic deperm&quot method, designed to combat these issues, compared favourably to the Flash-D protocol. The standard deviation (s) of the final PVM from 30 Flash-D deperms on steel tubes was 206 A/m; for the final PVM from 30 anhysteretic deperms of the same duration, this was 60 A/m. The s for the final PLM for Flash-D and anhysteretic deperms of the same duration were 416 A/m and 670 A/m respectively. The conclusion is that adopting the anhysteretic deperm on actual vessels would improve the reliability of the PVM outcome. Though the procedure would demand the same duration as Flash-D, there is the advantage of saving time by not having to repeat deperms to obtain the desired result. Additionally the anhysteretic deperm is considerably more amenable to theoretical analysis. A modified version of Langevin???s equation was used to predict the final PLM and PVM results for anhysteretic deperms and to provide a useful analysis of the anhysteretic processes in the Flash-D procedure. Using a Preisach analysis of hysteresis, a mathematical description of bulk magnetic changes that occur to a specific object, within a deperm, has been developed. Theoretical calculations of PLM in a steel tube during and after both types of deperm are in excellent agreement with experimental data. The same theoretical approach was also used to retrospectively model PLM results from previous Flash-D deperms on a submarine with equal success. With this analysis it is proposed that anhysteretic deperm outcomes could be predicted a priori. The influence of magnetic cargo on hull magnetisation was demonstrated to be of significance during and after deperming. &quotSympathetic deperming&quot occurs where a magnetic source is located close to the hull during a deperm. It was found that a vessel or model vessel hull could still be demagnetised even when they contain magnetic cargo that would normally resist the direct application of the same magnetic fields. This was explained using the principles of demagnetising fields and anhysteretic magnetisation. A possible explanation was provided for a PVM measurement anomaly common to the model and vessel deperm results. From measurement, alternating longitudinal applied fields apparently induce corresponding changes in the PVM. This effect could be explained by the depermed object being offset longitudinally from the position expected by the measurement system. This offset could be estimated using an analysis of the changes to PLM and PVM after a longitudinal applied field. The offset displacements calculated for the vessels were too small to be verified experimentally (&gt 0.1m), but the predicted offset for the steel tubes coincided with the limit of precision for their placement in the laboratory MTF = 0.5mm The aim of this work was to look at the deperm process with reference to a system that demonstrated qualitative similarities to deperms on actual vessels. The laboratory MTF is a unique facility, permitting a useful practical analysis of deperming based on sound magnetostatic measurements The experimental and theoretical results gained here have direct application to future deperms on naval vessels with particular reference to submarines.

Naval submarines and surface ships are regularly subjected to a treatment called &quotdeperming&quot that seeks to design the vessel???s permanent magnetisation for optimal magnetic camouflage. A scaled model of a magnetic treatment facility (MTF) has been established as a valid system to simulate deperming and used to investigate various aspects of the deperm process including: magnetic anisotropy and demagnetising fields as factors in the physical modelling of magnetism in whole vessels; a comparison of current and alternative deperm procedures; the application of theoretical models of bulk magnetisation to calculate deperm outcomes in the physical model and in actual vessels. A &quotlaboratory MTF&quot was constructed to imitate the applied field geometry at a naval MTF. The system was calibrated and it was determined that the laboratory MTF could make magnetic measurements on a CU200T-G steel bar sample with an equivalent accuracy (error = ??5%) to that of standard magnetometric equipment. Experiments were conducted with emphasis on a holistic approach to modelling the deperm process and describing magnetisation changes in whole objects. The importance of the magnetic anisotropic changes to steel with cold rolling was confirmed. In CU200T-G steel sheet the initial susceptibility (ci) was found to increase by a factor of 3 ??0.1 in the rolling direction, from a value of ~ 110 in the un-rolled steel sheet (thickness dependent). ci in the rolled sheet transverse to the rolling direction was decreased by a factor of 0.94 ??0.09 to ci in the un-rolled sheet steel. Previous studies on hull steel have neglected to account for this transformation through cold work. The demonstration on mild steel here is expected to have an analogy in the final state of the hull sheet steel as it resides in a submarine pressure hull. Future studies either on hull material or on modelling whole vessels should include the same or similar magnetic anisotropic properties in the steel(s) under investigation. Hollow circular tubes made from CA2S-E and CU200T-G steel sheet were selected as models for vessels. It was shown that these steel tubes were a good choice in this regard: minimising the complexity of the experiment whilst maintaining the validity of a deperm simulation. During a deperm there was an excellent qualitative likeness in the permanent longitudinal magnetisation (PLM) for the steel tubes to PLM in both a submarine and a surface vessel. Permanent vertical magnetisation (PVM) deperm results from the tubes displayed a close qualitative match with PVM in a submarine but not in a surface vessel. A theoretical treatment for demagnetisation factors (Nd) in hollow ellipsoids was used in conjunction with a geometrical approximation to calculate Nd for finite hollow objects of revolution. Subsequent theoretical calculations correlated well with experimental results for measured effective ci (ceff) in hollow circular CU200T-G steel tubes of various lengths and aspect ratios. Using an estimate of 100 as ci for submarine hull steel, the same analysis produces Nd for the axial and transaxial directions in a submarine equal to 5.97??10-3 and 0.0142 respectively. Three items for potential improvement were identified in the current deperm protocol used on naval vessels (Flash-D): redundancy in the protocol; the duration of the deperm and a theoretical basis for predicting the final magnetisation or changes in magnetisation during a deperm. Simulations of a novel &quotanhysteretic deperm&quot method, designed to combat these issues, compared favourably to the Flash-D protocol. The standard deviation (s) of the final PVM from 30 Flash-D deperms on steel tubes was 206 A/m; for the final PVM from 30 anhysteretic deperms of the same duration, this was 60 A/m. The s for the final PLM for Flash-D and anhysteretic deperms of the same duration were 416 A/m and 670 A/m respectively. The conclusion is that adopting the anhysteretic deperm on actual vessels would improve the reliability of the PVM outcome. Though the procedure would demand the same duration as Flash-D, there is the advantage of saving time by not having to repeat deperms to obtain the desired result. Additionally the anhysteretic deperm is considerably more amenable to theoretical analysis. A modified version of Langevin???s equation was used to predict the final PLM and PVM results for anhysteretic deperms and to provide a useful analysis of the anhysteretic processes in the Flash-D procedure. Using a Preisach analysis of hysteresis, a mathematical description of bulk magnetic changes that occur to a specific object, within a deperm, has been developed. Theoretical calculations of PLM in a steel tube during and after both types of deperm are in excellent agreement with experimental data. The same theoretical approach was also used to retrospectively model PLM results from previous Flash-D deperms on a submarine with equal success. With this analysis it is proposed that anhysteretic deperm outcomes could be predicted a priori. The influence of magnetic cargo on hull magnetisation was demonstrated to be of significance during and after deperming. &quotSympathetic deperming&quot occurs where a magnetic source is located close to the hull during a deperm. It was found that a vessel or model vessel hull could still be demagnetised even when they contain magnetic cargo that would normally resist the direct application of the same magnetic fields. This was explained using the principles of demagnetising fields and anhysteretic magnetisation. A possible explanation was provided for a PVM measurement anomaly common to the model and vessel deperm results. From measurement, alternating longitudinal applied fields apparently induce corresponding changes in the PVM. This effect could be explained by the depermed object being offset longitudinally from the position expected by the measurement system. This offset could be estimated using an analysis of the changes to PLM and PVM after a longitudinal applied field. The offset displacements calculated for the vessels were too small to be verified experimentally (&gt 0.1m), but the predicted offset for the steel tubes coincided with the limit of precision for their placement in the laboratory MTF = 0.5mm The aim of this work was to look at the deperm process with reference to a system that demonstrated qualitative similarities to deperms on actual vessels. The laboratory MTF is a unique facility, permitting a useful practical analysis of deperming based on sound magnetostatic measurements The experimental and theoretical results gained here have direct application to future deperms on naval vessels with particular reference to submarines.

Naval submarines and surface ships are regularly subjected to a treatment called &quotdeperming&quot that seeks to design the vessel???s permanent magnetisation for optimal magnetic camouflage. A scaled model of a magnetic treatment facility (MTF) has been established as a valid system to simulate deperming and used to investigate various aspects of the deperm process including: magnetic anisotropy and demagnetising fields as factors in the physical modelling of magnetism in whole vessels; a comparison of current and alternative deperm procedures; the application of theoretical models of bulk magnetisation to calculate deperm outcomes in the physical model and in actual vessels. A &quotlaboratory MTF&quot was constructed to imitate the applied field geometry at a naval MTF. The system was calibrated and it was determined that the laboratory MTF could make magnetic measurements on a CU200T-G steel bar sample with an equivalent accuracy (error = ??5%) to that of standard magnetometric equipment. Experiments were conducted with emphasis on a holistic approach to modelling the deperm process and describing magnetisation changes in whole objects. The importance of the magnetic anisotropic changes to steel with cold rolling was confirmed. In CU200T-G steel sheet the initial susceptibility (ci) was found to increase by a factor of 3 ??0.1 in the rolling direction, from a value of ~ 110 in the un-rolled steel sheet (thickness dependent). ci in the rolled sheet transverse to the rolling direction was decreased by a factor of 0.94 ??0.09 to ci in the un-rolled sheet steel. Previous studies on hull steel have neglected to account for this transformation through cold work. The demonstration on mild steel here is expected to have an analogy in the final state of the hull sheet steel as it resides in a submarine pressure hull. Future studies either on hull material or on modelling whole vessels should include the same or similar magnetic anisotropic properties in the steel(s) under investigation. Hollow circular tubes made from CA2S-E and CU200T-G steel sheet were selected as models for vessels. It was shown that these steel tubes were a good choice in this regard: minimising the complexity of the experiment whilst maintaining the validity of a deperm simulation. During a deperm there was an excellent qualitative likeness in the permanent longitudinal magnetisation (PLM) for the steel tubes to PLM in both a submarine and a surface vessel. Permanent vertical magnetisation (PVM) deperm results from the tubes displayed a close qualitative match with PVM in a submarine but not in a surface vessel. A theoretical treatment for demagnetisation factors (Nd) in hollow ellipsoids was used in conjunction with a geometrical approximation to calculate Nd for finite hollow objects of revolution. Subsequent theoretical calculations correlated well with experimental results for measured effective ci (ceff) in hollow circular CU200T-G steel tubes of various lengths and aspect ratios. Using an estimate of 100 as ci for submarine hull steel, the same analysis produces Nd for the axial and transaxial directions in a submarine equal to 5.97??10-3 and 0.0142 respectively. Three items for potential improvement were identified in the current deperm protocol used on naval vessels (Flash-D): redundancy in the protocol; the duration of the deperm and a theoretical basis for predicting the final magnetisation or changes in magnetisation during a deperm. Simulations of a novel &quotanhysteretic deperm&quot method, designed to combat these issues, compared favourably to the Flash-D protocol. The standard deviation (s) of the final PVM from 30 Flash-D deperms on steel tubes was 206 A/m; for the final PVM from 30 anhysteretic deperms of the same duration, this was 60 A/m. The s for the final PLM for Flash-D and anhysteretic deperms of the same duration were 416 A/m and 670 A/m respectively. The conclusion is that adopting the anhysteretic deperm on actual vessels would improve the reliability of the PVM outcome. Though the procedure would demand the same duration as Flash-D, there is the advantage of saving time by not having to repeat deperms to obtain the desired result. Additionally the anhysteretic deperm is considerably more amenable to theoretical analysis. A modified version of Langevin???s equation was used to predict the final PLM and PVM results for anhysteretic deperms and to provide a useful analysis of the anhysteretic processes in the Flash-D procedure. Using a Preisach analysis of hysteresis, a mathematical description of bulk magnetic changes that occur to a specific object, within a deperm, has been developed. Theoretical calculations of PLM in a steel tube during and after both types of deperm are in excellent agreement with experimental data. The same theoretical approach was also used to retrospectively model PLM results from previous Flash-D deperms on a submarine with equal success. With this analysis it is proposed that anhysteretic deperm outcomes could be predicted a priori. The influence of magnetic cargo on hull magnetisation was demonstrated to be of significance during and after deperming. &quotSympathetic deperming&quot occurs where a magnetic source is located close to the hull during a deperm. It was found that a vessel or model vessel hull could still be demagnetised even when they contain magnetic cargo that would normally resist the direct application of the same magnetic fields. This was explained using the principles of demagnetising fields and anhysteretic magnetisation. A possible explanation was provided for a PVM measurement anomaly common to the model and vessel deperm results. From measurement, alternating longitudinal applied fields apparently induce corresponding changes in the PVM. This effect could be explained by the depermed object being offset longitudinally from the position expected by the measurement system. This offset could be estimated using an analysis of the changes to PLM and PVM after a longitudinal applied field. The offset displacements calculated for the vessels were too small to be verified experimentally (&gt 0.1m), but the predicted offset for the steel tubes coincided with the limit of precision for their placement in the laboratory MTF = 0.5mm The aim of this work was to look at the deperm process with reference to a system that demonstrated qualitative similarities to deperms on actual vessels. The laboratory MTF is a unique facility, permitting a useful practical analysis of deperming based on sound magnetostatic measurements The experimental and theoretical results gained here have direct application to future deperms on naval vessels with particular reference to submarines.

Thesis (M.S.)--Worcester Polytechnic Institute. / Keywords: Expert panel; analytical hierarchy process. Includes bibliographical references (leaf 104).