IMFSE Master Thesis – what, why, how etc? part II

Can you tell me what is your master thesis topic and why did you choose it?

Darko: The title of my thesis is “Identification and characterization of design fires to be used in performance-based fire design of CERN facilities”.

After enjoying two fire dynamics courses during my first two IMFSE semesters – Fire Dynamics in Gent and consequently Advanced Fire Dynamics in Lund, I was quite interested in conducting a thesis related to that field. When I found out that Lund University collaborates with CERN, and that they proposed a topic related to fire dynamics, I decided to look no further.

Not too many places on earth are as mind-blowing as CERN 🌟

Kunsulu: My topic is “Numerical simulations of pressure effects in passive houses during fire”.  So, I will perform validation study of experimental data from a passive house facility built in Mons. Before choosing the topic, I decided that I will consider those related to FDS, as I wanted to develop my skills in using FDS. I chose this particular topic because currently there is a tendency towards energy efficient buildings and near-zero energy houses, and for me it was interesting to discover that there can be an evacuation problem and even a structural damage due to pressure rise during fire in well-insulated dwellings. Therefore, the topic seemed fascinating to me.

Melchior: I’m currently working on multiscale modelling of the accelerator tunnel located at CERN in Geneva. The accelerator ring is 27km long and impossible to model completely using 3D CFD. The mutliscale refers to the fact that I only model ± 500m using a 3D domain, while the other 26.5km is modelled using a much faster 1D approach.

I chose this topic as I’ve always been interested in fluid dynamics, and of course the opportunity to work with a world-renowned organization was a nice plus.

Julia: “Ignition of polymers under transient heating.”

I have always had a great passion for experiments and laboratory. Moreover, from the beginning I knew I preferred the focus of FSE at the University of Edinburgh. After a thorough view on the topics offered there, I decided that this would be the very exciting one!

Habib: The title of my thesis is “Validation and testing of full probabilistic quantitative risk assessment methodology for life safety in complex buildings”.

This thesis is more about testing an interpolation and response surface model to estimate FED simulation result, so the objective is developing a model that could estimate the simulation result in a certain case. In the end, we can reduce the computational time.

Arjan: My Topic is to make a cost benefit analysis of compartmentalization in logistic buildings, i.e. large floor areas. The analysis is made in essence for the private investor but also looks at the bigger picture of societal welfare and risk mitigation measures.

As I got the Rockwool sponsorship – half of my tuition fee was waivered and in return the thesis topic was based upon the company’s needs.


How has your thesis experience been so far?

Darko: Very interesting and unpredictable. Gathering and reading all the existing scientific papers on a certain matter and thereupon figuring what has not been done, and what should be done is a great challenge! Also, organizing custom experiments that include ordering and shipping a number of different gadgets and equipment from abroad can be quite tricky. I learned to always expect the unexpected.

Many thanks to Patrick and Dan from Lund University, but also to Oriol and Saverio from CERN for their selfless help whenever needed.

Kunsulu: It has been good so far. I am in the learning process: by trial and error (in writing codes especially); by asking questions and trying to find answers/explanation; with help and guidance of my supervisors. I just finished doing the study of leakage modelling methods in FDS where I considered how pressure and leakage flows can be affected by changing some parameters of leakage modelling in FDS. Now, I am starting the actual validation part.

Kunsulu pic
Kunsulu getting all the love and support needed for thesis from her dearest classmates 🙂 

Melchior: Not without the usual moments of wanting to throw my laptop out of the window, but overall more than positive. Also, thanks to the guidance both at CERN and LTH.

Julia: Some days are very productive and I manage to perform lots of experiments, whereas in other days several problems pop up and you cannot have a good progress. My experience has shown that activities usually take longer than we expect, mainly when working with experiments. Small problems will always come up, and most of them are unpredictable. However, finding solution for them is a very interesting part of the work.

Habib: It’s been very dynamic and demanding, but certainly enjoyable. Most of the time I run many fire and evacuation simulations, analyze the results, and then modify the prediction model.

Arjan: It took me a long time before I finally got started but once I got going the urge for more knowledge only grew. I can finally start writing the analysis tool now and I’m curious about the results it will give me.


What is your desired outcome from the thesis?

Darko: To produce a good paper that will be useful for CERN, but also that will comply with my expectations. To get satisfying results from my experiments and to have the test results obtained with different methodologies match!

Kunsulu: I want to deliver quality work and to be more confident user of FDS.

Melchior: To get a fully functional multiscale model, which will be able to resist experimental validation.

Melchior not sure whether he should laugh or cry after realizing what he needs to simulate in FDS 😅

Julia: The aim of the thesis is to find a trend that can lead to a new criterion – ignition criterion for transient heating conditions. Nevertheless, an important point is the knowledge that is acquired through the whole progression of the work, i.e. what I am learning with literature review, solving problems, understanding the uncertainties and limitations, and so on.

Habib: A prediction model, so that in the future if someone wants to obtain the result of similar case characteristic, they can use my model instead of running many simulations, so that they can save computational time.

Arjan: Finishing with a thesis that is up to standards, i.e. something of which I can be proud.


Do you see any interesting future research related to your thesis?

Darko: Yes! As CERN is such an enormous facility, this thesis produced design fires only for the major combustibles present in CERN (electrical cabinets, racks, magnets, klystrons, transportation vehicles). Further development of design fires for e.g. cables and cable trays, transformers etc. would allow more precise input for future FDS modeling.

Kunsulu: Yes, I do actually. I think there is a potential of future research both experimentally and numerically. For example, I think more tests could be conducted with solid fuels and realistic fire scenarios. Also, more work can be done in characterization of building air-tightness and ventilation configuration. To illustrate, sensitivity analysis of some default values in leakage modelling and quantification of uncertainties of ventilation flows can be useful.

Melchior: Yes, there is a lot to be improved upon, as the level of detail I’m currently using is relatively low. Also the implementation of current 1D modelling in FDS is not optimized yet.

Julia: Yes. The current ignition criteria are based on constant incident heat flux, which is not realistic in a real fire. If new criteria can be established based on transient heat flux, simulations will be more accurate and fire spread prediction will consequently be more precise and realistic.

Multitasking Brazilian 🔥

Habib: If we use the prediction model to predict other cases, it might result in high errors. So, the challenge is building a generic prediction model, so that we can apply it to many cases. It might be done by modifying the result model in this research.

Arjan: If the thesis is a success it should provide Rockwool with a tool they can use to advice/convince their future clients on installing compartmentalization.


Do you have any advice for future students on how to choose and conduct a good master thesis?

Darko: Make sure to choose a topic that seems interesting and useful. If you can’t decide between several topics, do not hesitate to contact the thesis promoters and further discuss the thesis possibilities. Contact with your potential future supervisor will probably solve all your doubts, and choice of a perfect topic will not be a problem anymore.

Start reading the literature related to your topic well on time and contact your supervisor occasionally with any questions. Ideally by the time when thesis semester starts (January), you will more or less have a clear picture of what you should do, and how will you make it happen!

Kunsulu: To choose the topic, I would recommend first to select areas you are most interested in or the ones you want to develop your knowledge in. Also, try to talk to the supervisors before selecting the topics and ask for more information. Regarding how to conduct a good master thesis, ask me in May-June after I will finish mine😊

Melchior:  Choose a topic that interests you, without looking at the location too much. You’ll have to work on the topic for several months and if it doesn’t interest you, those can become a really long four months.

Julia: It is important to firstly have an overview of the focus of each university. You also have to bear in mind what area you prefer (simulations, material burning, structures under fire, etc.).

Once your topic has been accepted, contact your supervisor. You can start reading papers and works related to what you will study and gather some ideas. When your last semester starts, try to get in touch with people involved in the same area, e.g. PhD students. And don’t forget, organize your activities! Before starting, set your objectives and keep in mind what you want to do and how much time you have for each step.

Habib: The most important thing that we need to take into account when we choose master thesis is our interest. By choosing the one that accords with out interest, it makes us more excited in doing the research, even though we face many difficulties during the research.

After that, we have to also understand the details concept of the research. We could discuss with the supervisor regarding the research topic if it is needed. So that we could prepare ourselves before doing the research and we are sure that we are able to complete the research.

Habib and his appealing model 👨🏻‍💻


  • Start as soon as possible
  • Don’t forget: it is/are your last year/months as a student. Enjoy it while it lasts, upgrade your social life and explore all the ins and outs of the city you’re in.
  • Don’t be afraid of your supervisor
  • Don’t think you don’t have enough material to set up a meeting with your supervisor, there’s always something to discuss
  • Study/work outside of your apartment. Go to the library or find some cool pub where there is an all-you-can-drink policy for tea or coffee, e.g. Caley Picture House Edinburgh.
  • Learn how to work efficiently with MS word a.s.a.p.
Our Belgian adrenaline junkie is getting inspired high above the ground 🛩️

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s