Abschlussarbeit am Centre for Ocean Energy Research

Die Stelle für Angebote und Gesuche von Studien- u. Diplomarbeiten, Praktika, Studentenjobs (Themen ohne weitere Beiträge nach 250 Tagen werden automatisch gelöscht.)

Moderator: (M) Mod.-Team Allgemein

Antworten
christianwindt
TalkING. Newbie
TalkING. Newbie
Beiträge: 2
Registriert: Mo, 09. Apr. 18, 16:17
Kontaktdaten:

Abschlussarbeit am Centre for Ocean Energy Research

Beitrag von christianwindt » Mo, 21. Mai. 18, 10:59

++++++++++++++++++++++++++++ UPDATE: Stellen sind immer noch zu vergeben!! +++++++++++++++++++++++++++++++++

Im Zuge des Projekts "Development of Energy Maximizing Control Strategies for Wave Energy Systems" hat das Centre for Ocean Energy Research (COER) der Maynooth University, Irland, zwei Abschlussarbeiten (Master) zu vergeben.

1) Implementation of a non-linear Power Take-Off system in the OpenFOAM environment

2) Development of a Smoothed Particle Hydrodynamics-based numerical wave tank for ocean wave energy systems

Projektbeschreibungen und Anforderungen siehe unten.

Schöne Grüße,
Christian
Implementation of a non-linear Power Take-Off system in the OpenFOAM Environment

This project requires advanced C++ knowledge.
Throughout the development of wave energy converters (WEC), engineers heavily rely on small scale experiments and numerical tools. Within the
latter category, an increased interest in high-delity non-linear numerical modelling of ocean wave energy systems by the means of Computational Fluid Dynamics (CFD) can be observed. These high-delity models are e.g. employed for performance estimation, structural analysis, pure hydrodynamic modelling or the investigation of control strategies for
different types WECs.

Avoiding linearisation, as applied in widely used low- and mid-fidelity models, CFD based numerical wave tanks (CNWTs) are able to capture all occurring hydrodynamic non-linearities. The gain in fidelity however, comes at the cost of hugely increased computational burden, compared to linearised models.

In order to justify the additional cost of CNWT experiments, a representation of the entire WEC system with highest possible fidelity is desired. However, most CFD models currently only include single subsystems of a WEC. A first step towards a more holistic numerical model includes the coupling between the non-linear hydrodynamic and a non-linear power take-off (PTO) model is desired. Both models are readily available as independent in-house development.

A first coupling has been implemented, between an OpenFOAM based CFD model and a Matlab based PTO model. However, in order to increase computational speed, avoid license purchase and provide the PTO model as standalone OpenFOAM toolbox, a transfer of the Matlab model to the OpenFOAM C++ environment is required. This translation forms the core of the proposed project. Furthermore, code verication and validation will be included in the project.

For further information, please contact:
Christian Windt christian.windt.2017@mumail.ie
Dr. Josh Davidson josh.davidson@mu.ie
Prof. John V. Ringwood john.ringwood@mu.ie
Development of a Smoothed Particle Hydrodynamics-based Numerical Wave Tank for Ocean Wave Energy Systems

In recent years, industry and academia have progressed on many techniques to harvest ocean wave energy resources. In order to design cost competitive and durable wave energy converters (WECs), engineers heavily rely on small scale experiments and numerical tools. Within the latter category, an increased interest in high-delity non-linear numerical modelling of WECs by the means of Computational Fluid Dynamics (CFD) can be observed.

Avoiding linearisation, as applied in widely used low- and mid-fidelity models, CFD based numerical wave tanks (CNWTs) are able to capture all occurring hydrodynamic non-linearities. These high-fidelity models are employed for a range of applications such as: performance estimation, structural analysis, pure hydrodynamic modelling or the investigation of control strategies for different WEC types.

In the most commonly used solution methodology, the finite volume method (FVM), the governing Navier-Stokes equations are solved on a numerical grid. This method is inherently prone to discretisation errors such as the insufficient resolution of turbulent fow characteristics. Furthermore, FVMs struggle to handle the large amplitude mesh deformations required when modeling typical WEC motions. More recently, so called meshless approaches are made available to a wide user-community. Here, the governing equations are solved for particles rather then on (mesh) nodes or volumes. Using smoothing kernels, quantities can be extracted at particular points in the domain. One of these new methods are the smoothed particle hydrodynamic (SPH) solvers, e.g. the open-source code DualSPHysics. Due to the nature of the solution method, SPH models avoid discretisation errors, can capture turbulent effects in high resolution and are not limited in terms of structure motion. However, these benefits come at even higher computational costs compared to classical CFD methods.

This project aims at the development of a SPH-based Numerical Wave Tank (NWT) and comparison of its performance to the on-site CNWT. This includes the setup on a simple initial model for the analysis of simple wave structure interaction problems. Hereafter, the model complexity should gradually be increased by the means of PTO and mooring
models.

For a successful project completion, experience in computational fluid dynamics as well as scripting and coding languages (e.g. Python, C++) are desirable. The time frame of this project is set to minimum 5 months.

For further information, please contact:
Christian Windt christian.windt.2017@mumail.ie
Dr. Josh Davidson josh.davidson@mu.ie
Prof. John V. Ringwood john.ringwood@mu.ie
Zuletzt geändert von christianwindt am Di, 10. Jul. 18, 16:58, insgesamt 1-mal geändert.
Christian Windt, M.Sc.
PhD Student
Centre for Ocean Energy Research
Dept. of Electronic Engineering
Maynooth University
Maynooth, Co. Kildare, Ireland

Benutzeravatar
Dennis Worry
Moderator
Moderator
Beiträge: 721
Registriert: So, 14. Okt. 07, 15:42

Beitrag von Dennis Worry » Di, 22. Mai. 18, 14:43

Very interesting topic ! So sad that the guy I know would fit 110% is just finishing his Master thesis using the same technologies :{
Zur Vereinfachung ist das Skalarprodukt des zu untersuchenden Vektorraumes als Flächenintegral zweier unbekannter Funktionen definiert.
Hellgate Harburg (tm)
http://rs85.rapidshare.com/files/917478 ... LA1_Dl.pdf

Antworten