Design of a life prediction tool for high-speed diesel engines

The objective of this design project is to develop a prognosis tool that will promote the use of physics based predictive maintenance models in the maritime industry and aid in their development and application.
Preventive or condition-based maintenance is generally in use in ships. These are generally based on operation hours suggested by equipment manufacturers or by predefined alarm thresholds. In addition to that, present maintenance programmes rely mostly on previous experiences and expert knowledge and do not consider the actual condition of the asset. Sometimes unexpected failures occur, which incurs operational risk, disruptions in supply chains and higher operational costs. To increase the reliability and dependability of assets, maintenance actions are presently being conducted in a more conservative way. These actions affect the flexibility of operations and the associated logistics process. Competitive pressure now forces companies to increase the availability of their equipment. It is therefore important to plan maintenance actions at the appropriate time.
In practice it is not always feasible to conduct maintenance based only on the actual condition of the asset. In order to do so, acquisition systems need to be used to collect real-time data, about the condition of the asset. Predictive maintenance (PM) aims to inform asset owners of the current and future state of their assets. Although predictive maintenance offers various benefits, the adoption in practice lags behind the theoretical understanding of it. The issue is that systems are built of many subsystems, which are built of many components (or even more levels of subsystems). Monitoring these systems by monitoring all components is neither feasible nor useful. A solution is to select the most suitable component having the highest asset benefits and best potential for implementation. Having identified it, the dominant material degradation parameters can be determined in relation to the component’s operation and load conditions.
To address these issues, a prototype prognosis tool is developed in this work. The critical components that is representative of the maintenance cluster is selected by means of a fourquadrant analysis. The four-quadrant analysis helps to focus only the most promising candidates, namely those with a low frequency of failure and a high associated failure consequence. Based on this method, cylinder liner and air inlet & outlet valves are identified as the critical maintenance driving components. After selecting the component preponderant failure modes are analysed and it is found that the degradation mechanism for both of these components is wear. Subsequently wear prediction algorithms are developed for valves and cylinder liners, and prognosis tools are developed based on these methods.
This prognosis tool is applicable for various propulsion systems, such as diesel electric, waterjet, and engine-driven controllable pitch propeller (CPP). Once the usage profiles of the ships are known, the tool is capable of predicting the component’s remaining useful life time. Its ability is demonstrated by subjecting this tool to various usage profiles that the ships encounter in their life time. In all these simulations, the effects of ship operating environment and the influence of propulsion systems on the expected life time of components agreed with the general observations in the maritime industry.
As the results are promising, the prognosis tool – which is in prototype form – can be developed further and integrated in the present maintenance practices. The presented framework enables the shipping companies to carry out maintenance in an effective way and to improve the reliability and availability of the ships.