Thruster Driveline Digital Twin : Bearing and Shaft Fatigue Life Prediction
Leppänen, Taneli (2021)
Leppänen, Taneli
2021
Konetekniikan DI-ohjelma - Master's Programme in Mechanical Engineering
Tekniikan ja luonnontieteiden tiedekunta - Faculty of Engineering and Natural Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2021-06-15
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202105245370
https://urn.fi/URN:NBN:fi:tuni-202105245370
Tiivistelmä
In the most basic form, drivelines consist of combination of shafts, gearboxes and clutches to transit and transform the torque and of bearing for support. Continuous and predictable operation is key for profitable operation of most commercial assets, ships and their thrusters included. Lifetime predictions and maintenance schedules are of based on calculation made in the design stages of the asset or on general recommendation from the manufacturer. These design stage calculations usually rely on conservative estimation of the usage profile or past experiences, if those are available, resulting in the best case in premature replacement of components or in the worst case in unexpected failure, both of which can be costly.
One way to alleviate this mismatch between expectations and real operation is to setup a condition monitoring scheme which can be further enhanced with real time calculations based on the measured data to predict component failures. This kind of real time data based calculation system can be collegially referred as a digital twin. With our load based lifetime predictions we can setup a condition based maintenance scheme, where maintenance can planed according to actual component condition.
For the scope of the thesis bearing and shaft calculation packages were chosen to be implemented in Python 3. The calculations were based on DIN 743 for shaft, and ISO 281 and ISO TS 16281 for bearings. The approach of using standardized methodologies was chosen for two reasons. Firstly, Kongsberg has experience using all of these standards and secondly, many of the maritime classification societies recognize these standards, which makes classification of the digital twin system easier.
With the shaft calculation package, generating valid for the standard calculation proved difficult due to the method requiring the load to be inputted in three different directions where the rainflow counting routines usually only can analyze a signal from a single direction. To account for this a synchronization scheme was implemented, but a more specialized multidirectional counting routine should be further investigated. The shaft stress calculation match well with the reference commercial implantation, but fatigue damage results should be further verified since the reference program only outputs safety factors.
For bearing calculations, the computation on of the contact stresses between the rolling element and the bearing ring proved most difficult. The difficulties came down to unavailability of the bearing inner geometries, which manufactures consider proprietary information. Due to lack of information verification of the bearing module proved difficult since separation between errors introduced by incorrect inputs and possible errors in the implementation is not possible. To circumvent this issue, generation of the pressure distributions by the manufactures to generate a ROM model should be investigated.
One way to alleviate this mismatch between expectations and real operation is to setup a condition monitoring scheme which can be further enhanced with real time calculations based on the measured data to predict component failures. This kind of real time data based calculation system can be collegially referred as a digital twin. With our load based lifetime predictions we can setup a condition based maintenance scheme, where maintenance can planed according to actual component condition.
For the scope of the thesis bearing and shaft calculation packages were chosen to be implemented in Python 3. The calculations were based on DIN 743 for shaft, and ISO 281 and ISO TS 16281 for bearings. The approach of using standardized methodologies was chosen for two reasons. Firstly, Kongsberg has experience using all of these standards and secondly, many of the maritime classification societies recognize these standards, which makes classification of the digital twin system easier.
With the shaft calculation package, generating valid for the standard calculation proved difficult due to the method requiring the load to be inputted in three different directions where the rainflow counting routines usually only can analyze a signal from a single direction. To account for this a synchronization scheme was implemented, but a more specialized multidirectional counting routine should be further investigated. The shaft stress calculation match well with the reference commercial implantation, but fatigue damage results should be further verified since the reference program only outputs safety factors.
For bearing calculations, the computation on of the contact stresses between the rolling element and the bearing ring proved most difficult. The difficulties came down to unavailability of the bearing inner geometries, which manufactures consider proprietary information. Due to lack of information verification of the bearing module proved difficult since separation between errors introduced by incorrect inputs and possible errors in the implementation is not possible. To circumvent this issue, generation of the pressure distributions by the manufactures to generate a ROM model should be investigated.