Page 36 - 2024 - Q3 - Minerva in Focus

MINERVA IN FOCUS Marine Technology

Harnessing data-driven One key area where data-driven methods are making a notable
impact is in Torsional Vibration Analysis (TVA), a process that mon-
torsional vibration itors and mitigates harmful vibrations in ship propulsion systems.
By applying TVA early in the ship design process, shipbuilders can
analysis for enhanced optimise performance, extend the life of ship components, and
improve fuel efficiency.
shipbuilding Understanding torsional
vibrations in marine vessels

Torsional vibrations occur in a ship’s propulsion system due to the
internal combustion engine’s cyclical movement. These vibrations
result from the forces generated during the combustion process
and the movement of pistons within the engine cylinders. Over
time, unchecked vibrations can cause fatigue and failures in key
components like the crankshaft, potentially leading to costly repairs
or even catastrophic failures at sea.

by Angelos Laskaridis,
Sales Manager,
INJEGOV S.A.

Figure 1: Torsional vibration steel
spring damper

Traditionally, shipbuilders have relied on mechanical torsional
vibration dampers to address these issues. While effective, these
dampers are a reactive solution, designed to mitigate vibrations
after they occur. However, as modern propulsion systems become
more complex, it is increasingly important to take a proactive
approach. Data-driven TVA offers a solution that allows engineers
to monitor vibrations in real-time, identify patterns, and predict
potential problems before they escalate.
Leveraging machine learning
for predictive maintenance
The integration of machine learning into TVA represents a major
In recent years, the leap forward in shipbuilding. By analysing vast amounts of data
collected from sensors placed throughout the ship’s propulsion
shipbuilding industry has seen system, machine learning algorithms can predict operational con-
significant advancements through ditions and offer insights into potential issues. For example, the
the adoption of data-driven Geislinger Monitoring System (GMS) uses sensor data to track
torsional vibrations in real time and applies advanced algorithms
technologies. With the growing to correlate this data with engine performance metrics such as
complexity of marine vessels fuel consumption and engine load.
and increasing demands for With machine learning, these systems can detect subtle patterns
efficiency and sustainability, that might indicate the early stages of wear or failure. Shipbuilders
can use these insights to design vessels that are not only more
shipbuilders are turning to Big efficient but also equipped with predictive maintenance capabilities.
Data and advanced analytics to This means that potential failures can be identified and addressed
design safer, more robust ships. before they cause major disruptions, thus reducing downtime and
maintenance costs while extending the vessel’s operational life.

36 ISSUE 29 / Q3 2024