DEEN

Institut für Prozess- und Partikeltechnik

Studienarbeiten

Fractionation of Fibre Suspensions (PAID MASTER THESIS)

Current trends in paper and pulp production aim on product diversification covering new markets, e.g., fibre-plastic compounds. Separating fibres by length may become a crucial process step in future. Answering to this future need, we developed a novel fractionation device in a collaborative project with industry.

The master student will prepare construction drawings using CAD, preferably SolidWorks. Prior skills from a technical high-school (HTL) are of advantage, but not required. The master student will receive training in the handling of the fractionator, image recording and post-processing with our existing high-speed camera.

We offer

· high industrial and scientific relevance (i.e., a novel separation process which will be applied in “real-world” trials at a paper mill)

· bleeding edge high-speed camera equipment and image post-processing routines

· support from the project team at IPPT and IPZ

· desk and office space

· Remuneration: 6 months á 440€.

For details CLICK HERE.

Simulation tool for pharmaceutical bioreactors: Adaptive grid refinement

Background
The Institute of Process and Particle Engineering is a world leader in the development of simulation tools for industrial-scale bioprocessing units. For example, our current code can model processes in large-scale bioreactors, up to 200m³.

The simulation code currently uses a regular grid. For small structures, i.e. small stirrers, the grid around the structure may be too coarse to resolve fine details, e.g. the fluid jet coming from the stirrer. Hence, the objective of this thesis is to employ a new grid addressing scheme and advanced interpolation methods to be able to define areas with a finer grid efficiently within the simulation code.

The coding is done in C++ in combination with the CUDA library on high-performance graphic cards.

Tasks
      • Find and implement a new grid addressing scheme
      • Research and choose interpolation methods
      • Literature study on test cases to validate the program module
      • Include the validation in the test harness of the code

Requirements
      • Background in chemical or bioprocess engineering, physics or similar
      • Basic biotechnological knowledge
      • Being familiar with thermodynamics, simulation and modeling

What we offer
      • Integration in an internationally leading team
      • Opportunity to be part of a commercialization project
      • Paid master thesis
      • Start of a future career in software creation

Start: Summer/Fall 2019

Contact
Dr. Christian Witz
0316 873 30416
christian.witznoSpam@tugraz.at

Advanced heat transfer algorithms in pharmaceutical bioreactors

Background

The Institute of Process and Particle Engineering is a
world leader in the development of simulation tools for
industrial-scale bioprocessing units. For example, our
current code can model processes in large-scale
bioreactors, up to 200m³.

We are therefore looking for a bachelor or master student
with engineering background (chemical engineering,
bioprocess engineering, physics or similar) interested in
extending the simulation tool by adding highly efficient
and fast models to calculate the heat transfer from
heat jackets, heat exchanger or air bubbles to the fluid
in the reactor by employing dimensionless relations like
the Nusselt number. An algorithm for the convective heat
transport is already included in the code.

The coding is done in C++ in combination with the CUDA library on high-performance graphic cards.

Tasks
      • Implement the respective heat transfer by the heat exchanger, heat jacket or air
        bubbles to the fluid as sources in the convective heat
      • Literature study on test cases to validate the program module
      • Include the validation in the test harness of the code

Start: Summer/Fall 2019

Contact
Dr. Christian Witz
0316 873 30416
christian.witznoSpam@tugraz.at

Automation of the setup and the post processing for pharmaceutical bioreactors simulations

Background

The Institute of Process and Particle Engineering is a world leader in the development of simulation tools for industrial-scale bioprocessing units. For example, our current code can model processes in large-scale bioreactors, up to 200m³.

We are therefore looking for design exercise (Konstruktionsübung) student with engineering background (chemical engineering, bioprocess engineering, physics or similar) interested in extending the simulation tool by adding algorithms for the automation of the setup and the post processing of the simulation.

The simulation code currently uses a JSON text file as interface to the user to define the simulation parameters. The task of this exercise is to extend a HTML page with Java Script to create this JSON file. A working example already exists.

The second task is to create Python scripts to control the software Paraview, which is currently used to post process the simulation results (e.g. create videos, images and plots automatically).

Tasks
• Expand the existing HTML/Java Script page
• Write the Python scripts necessary to create videos, images, plots with Paraview automatically

Requirements
• Background in chemical or bioprocess engineering, physics or similar
• Being familiar with programming, simulation and modeling

What we offer
      • Integration in an internationally leading team
      • Opportunity to be part of a commercialization project
      • Paid design exercise
      • Start of a future career in software creation

Start: Summer/Fall 2019

Contact
Dr. Christian Witz
0316 873 30416
christian.witznoSpam@tugraz.at

CFD Simulations of Gas/Liquid and Liquid/Liquid Flow in Mixing Elements (2 Master Theses)

The ultimate goal of these Master Theses are database containing geometric models for gas/liquid (G/L) and liquid/liquid (L/L) reactor mixing elements. These geometric models should be characterized via CFD simulations, such that their selection according to predefined synthesis conditions is possible. Additionally, an approximate model that describes the flow (e.g., the mean speed in the device) in these mixing elements should be established.

Based on the results of the CFD simulation studies, advanced G/L and L/L reactor elements will be prepared, and evaluated in the laboratory (this experimental work is not part of the theses). Based on the outcome of these evaluations, additional simulations should be performed, and results should be compared with experimental findings.

The objectives of this work are:

Performing benchmark simulations of gas/liquid flow and evaluation of the results against literature data (literature survey partially available)

Evaluation of existing G/L and L/L mixing elements (based on simulation) with respect to their suitability for different flow regimes

Design of new, advanced G/L and L/L mixing structures to be fed into the model database

Comparison of the simulation output with experimental data for selected geometries

Start: May 2018
Contact: Stefan Radl (radlnoSpam@tugraz.at, 0316 873 30412)

Modelling of a perfusion reactor

Perfusion reactors are used to host microbial cells, which are able to produce antibiotics, potent drugs substances for cancer therapy or other active pharmaceutical ingredients. During the reactor’s operation, the produced drug molecules have to be extracted continuously. This is currently done via alternating tangential flow filtration (ATF).
A part of the solution in the reactor is sucked through a fiber filter element with a diaphragm pump. The concentrated cell solution is pumped back in the reactor by the diaphragm and the cell-free filtrate is pumped to the next stage to extract the active pharmaceutical ingredient.

Tasks

Literature study on the topic of perfusion reactors
Modelling of the fluid flow through the filter fiber and the filter process
Study on the influence of the microbial cells and rheological properties of the solution on the filtration process
Development of suggestions for improvements for the filtration process

We offer

Opportunity to work on an industrially relevant task
Contact to a leading pharmaceutical company
Paid master thesis
Start of a future career in modelling and simulation

PDF OF THIS MASTER THESIS SOLICITATION

Contact:
Dr. Christian Witz
christian.witz@tugraz.at

Statin synthesis via heterogeneous (bio)catalysis (Paid Master Thesis)

Statins are the active pharmaceutical ingredient (API) of many cholesterol lowering drugs. Their structure consists of the typical statin side-chain possessing two chiral alcohols linked to a heterocyclic core. This side-chain can be synthesized from simple and inexpensive starting materials via a two-step aldol condensation catalyzed by an enzyme called DERA (2-deoxyribose-5-phosphate aldolase). The side-chain can either be directly built at the core of the molecule or linked to the heterocyclic core subsequently via a C-C coupling reaction catalyzed by Palladium.

The goal of this work is to investigate the biocatalytic step in this synthetic route. A number of substrates, such as acetaldehyde, chloroacetaldehyd, benzaldehyde and cinnamaldehyde, will be testes as acceptors in the aldol condensation. The obtained product will be characterized and evaluated according to their potential for serving as intermediate in the synthetic route of statins. Further the enzyme (enclosed in E. coli cells) will be immobilized in order to apply it in a continuous process.

The results of this thesis will serve in the development of an integrated multistep process for the synthesis of statins consisting of a biocatalytic and a metal-catalyzed step.

The objectives of this work are:

Substrate screening in batch
Immobilization of the enzyme/cells for the application in a continuous process
Purification and characterization of the products (NMR)

We offer:

payment according to the FWF-rate (€440/month)
a comprehensive introduction to the research topic
access to novel experimental and analytical devices
individual assistance for an efficient realization of the thesis

Start: March 2018
Contact: Bianca Grabner (b.grabnernoSpam@tugraz.at, 0316 873 30409)

2 Bachelor-Projects - Chemical Engineering (Verfahrenstechnik)

Experimental Evaluation of a Particle Fractionator

Current trends in paper and pulp production aim on product diversification covering new markets, i.e. fibre-plastic compounds. Separating fibres and fines (i.e., particles smaller than 200μm) may become a crucial process step in future. Answering to this future need, we developed a novel fractionation device in a collaborative project with industry.

First studies revealed a dependence of the separation performance on key flow parameters, and was investigated by means of high-speed imaging (Figure 1, right panel).

Bachelor projects will aim on detail investigations of how the fibre network formation affects separation performance. The bachelor students will receive training in the handling of the fractionator, image recording and post-processing with our existing high-speed camera.

We offer

high industrial and scientific relevance (i.e., a novel separation process which will be applied in “real-world” trials at a paper mill)
bleeding edge high-speed camera equipment and image post-processing routines
desk and office space

Contact

Stefan Radl, radl(at)tugraz.at; 0316 873 30412

The bachelor thesis projects can be started earliest in summer 2018

Exploiting Artificial Intelligence in in Gas-Particle Flow Simulations

Current trends in advanced multiphase flow prediction aim at the usage of machine learning algorithms and deep neural networks to improve the accuracy and to speed-up numerical simulations. Overall, it can be expected that these tools (and artificial intelligence, AI, in a wider context) will become an important part of numerical modelling used in chemical engineering applications.

The overarching goal of this Master Thesis project is to increase the speed of gas-particle flow simulations (see the right panel in the Figure below) by using an AI-powered prediction algorithm.

Your tasks will include (i) a literature review on the usage of deep learning in numerical simulations, (ii) an investigation related to key parameters of a neural net structure to speed up a widely-used gas-particle flow simulator, and (iii) application of the improved simulator to a use case. The machine learning and deep neural nets will be based on existing open source AI platforms (e.g., Tensorflow), for which expert knowledge is already available at our institute.

Deep Learning Algorithms

Qualification

Interest in Computational Fluid Dynamics (CFD), Discrete Element Method (DEM)
Basics programming skills (Matlab/octave, Python, C/C++ or other), and interest to refine your programming skills during the Master Thesis project

We offer

Extremely high scientific and industrial relevance
Introduction to the leading open-source gas-particle simulation tools OpenFOAM® and CFDEM®. Support with the AI platform, as well as with respect to programming.
Computer power, desk and office space
Possibility to publish the results and findings in a scientific journal

Contact

Josef Tausendschön: josef.tausendschoennoSpam@tugraz.at, Stefan Radl: radlnoSpam@tugraz.at

Nach oben

Kontakt

Michaela Cibulka
Mag.phil.

Institut für Prozess- und Partikeltechnik
Inffeldgasse 13
8010 Graz

Tel.: +43 (316) 873 - 30403
Fax: +43 (316) 873 - 1030403

Sprechstunden: nach Vereinbarung

m.cibulka@tugraz.at

Visitenkarte