Computational Fluid Dynamics (CFD) is a fascinating area of study, and some university degrees are available that focus either exclusively on CFD or on CFD applied to a specific problem area. In this article, we look at reasons for and against specialisation and whether it is worth paying for it through a university degree. Finally, we look at some places where you can get a degree which specialises in CFD in one way or another.
In this article
- Introduction
- So, should you get a degree?
- Places to study CFD
- Cranfield University (UK): MSc in Computational Fluid Dynamics
- Imperial College London (UK), MSc in Advanced Computational Methods for Aeronautics, Flow Management and Fluid-Structure Interaction
- Politecnico Milano (Italy), LM (equivalent to MSc) in Mathematical modelling for engineering
- Swansea University (UK), MSc in Computational Engineering
- University of Southampton (UK), MSc in Aerodynamics and Computations
- University of Manchester (UK), MSc in Thermal Power & Fluid Engineering
- Summary
Introduction
Computational Fluid Dynamics, or CFD, is a fascinating area to study. The first time CFD was performed dates back to the 20th of May 1910, when Lewis Fry Richardson computed the weather 6 hours ahead, some 36 years before the first electronic computer was put together (I don’t say invented to avoid an angry mob of computer science historians unhelpfully pointing out the early work on computers by Charles Babbage).
Richardson used human computers arranged in a way we would recognise as a computational grid nowadays. The human computers calculated the solution of pressure, velocity, and energy for their current location based on information provided by their neighbouring counterparts. This is still how CFD is done these days, except that we have replaced the sub 1 FLOPS (floating point operations per second) human computers with processors capable of performing 3-4 billion FLOPS.
Ultimately, Richardson failed to predict the weather accurately, but only since stability criteria such as the CFL stability condition were not considered (the CFL condition was published in 1928, 18 years after the work of Richardson). Post-analysis of his work revealed, though, that his approach was correct, and with the appropriate settings, his calculation would have been fairly accurate.
CFD was first seriously developed in the 1950s, with early works by Peter Lax, Burton Wendroff, and Sergei Godunov. Godunov died in 2023 at the age of 93, while both Lax and Wendroff are still alive (as of May 2024) at the crisp ages of 98 and 94, respectively. If this is anything to go by, a career in CFD promises at least longevity (but don’t tell my pension manager)!
In the 1970s, when CFD produced the first usable results, it became increasingly important in the aerospace industry, which was the main driver behind the early development of CFD for commercial applications. Other industries soon followed, and nowadays, you will find CFD being used in any conceivable engineering industry (and beyond!).
I have seen people performing multiphase and solidification studies on the melting of chocolate to optimise taste and production, testing the combustibility of human faeces for energy production in remote areas, predicting stock prices as well as interest and exchange rates (an area known as computational finance which heavily relies on methods developed for CFD), and the prediction of traffic jams in UK roundabouts at peak traffic times, to name but a few exotic applications of CFD.
I love CFD because it is so versatile. While I tend to do most of my work in external aerodynamics (both for aerospace and motorsport applications), I enjoy exploring other areas that take me away from these rather conventional applications (most recently, the combination of CFD and machine learning, which seems to be an area with huge potential, but also lots of frustration and false promises!). I’m sure we’ll come back to this at some point on this website.
So, should you get a degree?
So then, the question is, should you do a degree in CFD? Is it worth specialising in this area, or should you become a generalist where you apply CFD occasionally but don’t need an in-depth knowledge of CFD? I think this is a question we all need to ask ourselves; the answer will depend on your personal circumstances.
Let’s take myself as a case study. In 2011, I first worked on a CFD problem in my undergraduate studies and wanted to learn more about this field. So, I looked for places to specialise in CFD and got my (MSc) degree in this area. Before I joined the MSc course, I wanted to work in the aerospace industry; I had no interest in writing my own codes or having an academic career. I considered doing a PhD, but that was never a serious thought back then.
After I finished my MSc degree, I realised that my true interest was in coding, that a PhD is necessary to really understand CFD from the ground up (and the only way to do this was to write my own solvers), and so I enrolled as a PhD student. Afterwards, I never really thought of going into academia for teaching (as I had no background in it), but I liked the idea of staying in academia for at least some time to do some research. However, there were no interesting positions to do a post-doc in CFD at the time, so I decided to work in a research institute where I was working on a commercial CFD solver.
Luck had it that my previous supervisor contacted me, asking me to apply for a position at my previous university. I did, and I got the job. This was the right call, and I realised I am passionate about teaching and supervision. Every day, I get up and look forward to the work ahead; this is the best part of my job. I wouldn’t have reached this point without my specialisation and MSc in CFD.
But who knows, had I just become a generalist with no particular specialisation in my MSc studies, I may have ended up in the aerospace industry and continued my undergraduate work on wingtip designs, and you would now be reading about it on my website www.wingtip.university, who knows.
The point is that, for me, specialisation was the right choice. I only discovered my true passion by really immersing myself within the field of CFD, exploring different concepts through assignments, group projects, and individual research projects, and finding what really interested me, which prompted further specialisation. I think this is the strength of specialisation and why I would prefer to specialise, rather than generalise, in my postgraduate (MSc) studies.
However, that doesn’t mean everyone must specialise. If you are gunning for positions where CFD plays a role but not a dominant role, a CFD specialisation may not be the right choice. Having said that, I have had students who were on their second or third MSc degree, in which case a specialisation in different areas may make sense again. This is a question for you to answer: Do you want to work with CFD most of the time in your next position, or do you anticipate it will play a smaller role in your day-to-day activities?
Here is something else to consider. Positions with specialisations typically receive fewer applicants, and it is easier to stand out among the crowd if you have a specialisation. I recently chatted with the head of Aerodynamics and CFD of a prominent Formula 1 team, and he said that a typical junior aerodynamics job posting will attract 300 – 400 applications, while a junior CFD job posting will attract 100 – 200 applications.
Sure, Aerodynamics is a specialisation, but I would argue that most aerospace-orientated undergraduate courses will have some modules on aerodynamics, but only a few will explore CFD. And even if they do, the coverage of CFD is typically less than that of aerodynamics. Most CFD modules are placed in a post-graduate degree program, where they are explored at different levels of depth. My point is, with a specialisation, you can increase your employability (in the case above, by 200%).
So, let’s examine the pros and cons of getting a (specialised) university degree further and also look at the cost of specialisation.
Pros of getting a degree
In-depth education
The first thing to point out is the in-depth knowledge you will gain. CFD lives at the intersection of physics, mathematics, and computer science, and this is just a very broad definition. I have found myself reading up on computational geometry books for some very specific problems (mainly related to geometry and meshing problems). Case and point: Space-Filling Curves: An Introduction with Applications in Scientific Computing by Michael Bader. Ah, yes, what a page-turner!
You will need to gain expertise in all three areas, at least to some degree. If you are studying CFD on the side, you’ll likely concentrate on the fun bits (i.e. running simulations), which means you can follow tutorials but not much else. When you have to set up your own simulations or encounter problems with them, you won’t have any intuition to go by and are at the peril of Google and co to figure out what’s wrong.
It is not difficult these days to run simulations and get results; solvers like Ansys Fluent have made it look easy to get results, but making sure the results are trustworthy and correct is an entirely different aspect, and while not everyone gets excited about validation and verification, it is the bread and butter of a serious CFD engineer. Quantifying the level of accuracy is difficult, but if you can do that, you have a solid foundation to become an expert in the field. To get there, you will, again, need to know how the areas of physics, mathematics, and computer science influence your simulations.
A degree in CFD should expose you to exactly these areas: you should be programming, running simulations, critically assessing each simulation and being able to say whether it is correct or not and how accurate it is. You should also have a good foundation in mathematics and at least be able to derive and modify the Navier-Stokes equations to tailor them to specific situations (there are countless permutations of the equations). If you can do that, you can assess the limitations of the set of implemented equations.
For example, do you know what limitations the Euler equations have in the context of incompressible and compressible flows? What about Stokes flows? When can we reliably use them? At what point do we need averaged Navier-Stokes equations, using either RANS or FAVRE averaging? Can we be so crude and use potential flow equations? (the answer is yes, and solvers like OpenFOAM and Ansys Fluent use them!). Without knowledge in these areas, following the implementations in the solver you are using will be difficult.
Or let’s take computer science. You are probably thinking about programming, but it is simply a tool or a means to get you to an implementation. For me, computer science is a heavily underrepresented discipline in CFD (and partly the reason why I spend so much time on it on this website), and it has more to do with how we can write performant software. Performance is one of the most important aspects of CFD, given how power-hungry it is, and we need to make an effort to write efficient code if we want to run large-scale applications.
For example, how much environmental pollution is released by a single large-eddy simulation (LES)? I did the calculation for one of my simulations, and the CO2 released into the atmosphere from the energy production to run this simulation was the same as me driving a compact-sized car from Paris to Madrid (or about 1000km (or about 600 miles for my non-SI-savvy readers)). This is why we need to ensure our code is as performant as possible; we can’t afford wastage here. Computer science teaches us how to do that.
Supervision and support
This is probably the most important point after the in-depth education argument! Anyone can learn about a new area, but if you want to become proficient quickly, you need quick feedback cycles to understand where you are going wrong and why. Learning requires constant supervision and guidance from someone who understands your field of study and can point you in the right direction should you go wrong.
A good supervisor will take a personal interest in your development and ensure that you get the best possible learning trajectory out of your studies. However, most supervisors are overloaded with too many other tasks, so the best you can hope for is somewhat regular meetings. I try to be of the former category, but I leave judgement to my previous students.
Another point to consider is the use of reference letters. I personally find them extremely useful, at least in theory, but the way that academics write reference letters (i.e., write them once and give them to each student who asks for them without modifications apart from the name) is not helpful and defeats the purpose of a reference letter.
A good reference letter should be personal, reflect the student’s work honestly, and highlight both positive and areas of development to give it more credibility. I have never asked my students to write their own reference letters and always write a unique one for each student. Why? Because I have seen this to make the biggest difference during a selection process. I have had companies come back to me saying that they selected my former student based on the reference letter, which happened numerous times. This is the difference the supervisor can make.
Access to specialised resources
This point perfectly continues my previous paragraph: If you want to run a large-eddy simulation, you won’t do it on your laptop! You will not just need access to a high-performance cluster, but also training in how to use one. You can’t just build your own cluster to get a feeling for how it all works (although there are some nice DIY projects doing that with raspberry pies, though they are horrible for CFD purposes), you will need the real deal.
Cloud computing is becoming more popular and is a decent solution for small- to medium-sized companies. Still, larger companies and universities will have their own clusters, which are slightly different from operating a cloud-based solution (i.e., you will need to understand how a scheduler like PBS works). Supercomputing facilities operate similarly, so you must at least understand the basics here. Cloud computing doesn’t scale well regarding cost here, so it is a good starting point but not the final solution.
A good university should give you access to a decent cluster, which means you should also have a basic understanding of UNIX operating systems. The majority will likely use Red Hat, but I always advocate that you make the fastest progress with Ubuntu. I still prefer it as my UNIX system of choice (and I have probably now installed and tested 5-10 different flavours of UNIX on my PC, but I always gravitate back to Ubuntu).
Access to a decent cluster is one thing, but you also need the right software to run on the cluster. If your University only offers Ansys Student licenses, the cluster is not good, as the student version is too restrictive. You need full Ansys licenses (or alternative software) to really make the most of your compute time on the cluster. If you only work with OpenFOAM or other open-source software, you are also missing out, as the learning curve with OpenFOAM is so steep that you likely only ever run simple geometries and test cases.
Finally, specialised resources, to some extent, also include the staff themselves. Even if you have access to a decent cluster and software, if no one knows how to operate either of them, then you can’t really make any progress here. You want to ensure that the staff is competent (which likely is difficult to check before you start your studies) and that they use the same tools they teach regularly. I stay updated with ANSYS releases and go through tutorials when new features are available to ensure I know what’s new (and the same is true for OpenFOAM).
Networking opportunities
Let’s face it—you want a degree to get certain benefits. For most, this is a quick route into employment. Universities offer possibilities to network with their partners, be it through informal or formal events or through the university staff directly (which can put students in contact with people in the industry).
You may be working on a thesis topic together with an industrial partner, or you may be on an internship and placed directly within a company, or you may attend a seminar given by one industrial partner, and you get into contact. Perhaps you have a background in the area they work on through previous employment or university projects. These are all great ways to get into contact with potential future employers.
In a post-COVID world, though, I have found engagement with industrial partners somewhat on the decline, but that doesn’t mean that the networking opportunities aren’t there anymore. What I have observed is that students are really good at initiating contact with companies themselves and getting them to discuss topics that they might be interested in working on, for example, in a thesis-related context.
Finding the right person may be difficult, although if you look through academic papers and for company affiliations (a tool like Scopus can do this really well if you have access to it), you do get a person, an email address, and a topic they have previously published in, so contacting them about this research will already get you a higher response rate and they are likely more open to academic projects anyways. I have found this to be a great starting point for finding useful contacts and have established a few of them myself through this route.
Credibility
The last paragraph also leads me straight into the credibility argument. Everyone can read through textbooks and get the same knowledge as someone without a degree. Still, someone with a degree has gone through assessments and demonstrated that they had obtained a certain knowledge (and you get a rough idea of how much based on the grade), while someone going through textbooks hasn’t. They may have the same theoretical knowledge, but what about practical skills? Will they have the same knowledge of CFD software and how to run it on a cluster?
A university degree gives you the credibility you’ll need to get your foot in the door, and offers far more credibility than any massive open and online courses such as provided by coursera, edx, and the like (though they are perfect to boost your skills in specific areas which can be a bonus, but not a degree replacement, in your CV).
Some courses also undergo a lengthy accreditation process, especially in the UK, which signals to future employers that this course meets the professional’s body assessment to be considered a high-quality course and one in which relevant skills are taught and assessed. This doesn’t mean that you have to study on a course with accreditation; many countries will not have a similar system in place, but it offers you certain advantages as well (such as gaining chartered engineering status easier if this is what you are looking for).
The university ranking can play a role, but in my view and experience, the aptitude of the student is far more important than the rank of the university they come from. The only advantage a high-ranking university will give you is to get a foot in the door; the rest is down to you and your skills and knowledge.
Cons of getting a degree
Time and financial investment
If you have chosen to specialise in an area and go to university, there is a financial and time-wise cost. If you go straight to your post-graduate studies after your undergraduate studies, you’ll likely not perceive the time investment as significant, but if you decide to do a degree after getting a job and some experience, you’ll notice that the time investment will become significant and you have to make changes in your life to accommodate your studies.
The cost may also be a significant factor depending on where you want to study. In the English-speaking world, university degrees typically require taking on a student loan to pay for tuition alone, which can be off-putting for some. In other parts of the world, you will find places where no tuition fee is taken, regardless of your nationality, and some places even pay you during your post-graduate studies (but as with jobs, these places may be highly competitive).
Time and financial investment are likely the biggest obstacles people will face and likely the deciding factors in your decision. If you can’t overcome this hurdle, you’ll likely decide not to take up studies. However, since the cost is becoming disproportionally more important as a deciding factor, I have addressed this point further below.
Availability of alternative learning resources
If you did see my list of resources to get started with CFD, you’ll find many good books at reasonable prices and even some completely free resources to use at your disposal. So why should you pay for tuition if you can get the knowledge for free (or at a much lower cost) these days? With additional alternatives such as free online courses (usually provided by universities themselves), material on YouTube, and other self-study material on the internet, justifying the cost of a university degree may become more and more difficult.
If it is just the knowledge you are after, I would argue you don’t need a degree; you can get that elsewhere. However, the flipside is that if you want to gain expertise quickly, a university is designed to get you to that stage in a predefined time, and likely you won’t be able to get to that point in the same time by just consuming self-study material. You won’t set yourself deadlines as you see in universities, and you won’t study material that is difficult to the extent that you get sick of it, but all of this is required to gain an overall well-rounded view.
Furthermore, you’ll also miss out on supervision, guidance, and potentially some networking opportunities. A supervisor will be able to point you in the right direction should you go into the wrong one (oh, I have so many juicy supervision stories I’d like to share, but likely get in trouble for if I do here; perhaps that’s best left for the email list …), and this direction can help you to the path of expertise quicker.
So yes, plenty of spaces are available to study material for free. If you only want to have the knowledge, a university may not necessarily be the best place. For some this is perfectly acceptable, some may already be in employment and just need to have a bit more foundation in a specialisation and get the rest through experience on the job, so whether you see the alternative learning resources that are available as a con will largely depend on your personal circumstances.
Rapidly evolving field
Universities are slow to adapt, and this may result in some degrees quickly becoming outdated and out of touch with reality compared to others. For example, when I started teaching, I inherited the lecture material from the person I was replacing. I realised that the material taught was based on where the field was about 20-30 years ago. Granted, not much has changed in this field of theory, but software tools have moved on, and with outdated lecture material, it is difficult to link lectures and computer labs.
I did update the lecture material and brought it into the 21st century. I did this because initially, I had nothing else to do at university: no course to manage, no research students to supervise, no internal university committees to sit on, and the list goes on. I still regularly update my material, but I know that it takes exponentially more effort now to do so, and this may result in some academics putting this off for later and never getting around to doing that.
It helps if an academic course doesn’t exist in a vacuum, i.e. if there are external bodies that check if the course still teaches relevant skills. Accreditation is one thing. Another is a panel of industrial experts meeting once a year to discuss the course and things to include/retire from the curriculum. This is an ideal scenario, but not many universities will go to this length.
If you want to know whether a course is updated regularly, use a tool like the Wayback machine (which allows you to browse a website at a previous time). See how the course was structured 10 years ago and where it is now. If there are no updates, the programme may not have been updated. This is not a foolproof method, though, as sometimes the modules get updated only without changing the course structure. I guess common sense applies here. If in doubt, contact the course team and ask them directly.
Practical experience vs. theoretical knowledge
With all things being equal, an applicant with more relevant working experience than another with the same academic background will likely be picked for a job first. When I did my undergraduate studies and looked at the job market, it can be quite frustrating, and I remember a lot of memes circulating on Facebook (yes, that’s what we used back then) similar to this one:
An academic degree may be required for the positions you want to apply for, but once you have started to work, your experience will become increasingly important and the deciding factor in landing a job. If I were to apply again for positions, I don’t think my MSc or even BEng studies would be considered important. My PhD degree would likely be the most deciding factor in terms of my academic background, but still not as important as my past work history.
The conundrum remains: you’ll need experience, even for entry-level jobs. Some universities counteract that by requiring you to do an internship for 6 months as part of your studies, which will get you some initial experience. But even if a university doesn’t have that requirement, most will allow (and encourage) their students to seek out opportunities for themselves and contact companies directly. These are excellent opportunities to fill that gap in a CV and hit the ground running once applying for jobs.
What about the cost? Is it worth paying for a degree?
I probably have a somewhat different view on the cost of tuition now than I did a few years ago. I’ll share that with you and let you decide whether you agree. But let’s look at a less controversial statement: Time is money. I mentioned the time commitment above and want to elaborate on that a bit first.
I hate mowing my lawn. It probably takes me about 20 minutes to do, yet I can’t find the motivation to do it every other week. I’ll very happily fork over some money for someone else to do it, and if I do, I essentially trade my time for someone else’s and compensate them with money. This is a very basic example of the time and money idiom, but it applies to many more situations, and a degree is the same. If you want to get the expertise in a short amount of time, then it will cost you in one way or another.
If we view a university degree through this prism, then the only question we need to ask ourselves is, how quickly do we want to finish the degree? The shortest degree you can get is probably around a year, so if we invest a year of our time and some money on top, are we going to get where we want to be in a year’s time? If the answer is yes, then it is probably worth considering.
But let’s get my more controversial view. First of all, when I was a student, I viewed education as a right, not a privilege, and applying this view means tuition fees should not be collected. In Germany, where I got my first degree, the government is funding universities for their day-to-day activities (and research findings can be obtained on top through various means), but this means the university does not have an incentive to increase student numbers. Academics are probably very happy if student numbers decrease, as they have less work to mark.
In countries like the UK, where the government does not cover the baseline cost of universities, tuition fees need to be raised to cover expenditures and salaries. While this is an economic essential to survive, it also means that education is available only to those who can afford it or obtain a student loan, but many will be blocked out. While I understand that tuition fees pay for my salary, in an ideal world, I’d rather work in a place where every student has an equal opportunity to study.
However, there is a flipside to this as well: Looking back at my studies in Germany, when we started our studies, we were about 40 students. At the end of our studies (after the 3.5 years it took to complete the degree), we were only 5 people or so left. A few took one or two additional semesters to finish their study, and while I don’t have an exact number for how many people finished the course in the end, I would guestimate that a total of 10-15 students finished their study, less than half.
If you think this is quite a low success rate, it is very common in Germany. Since the government is paying the bills and salary, there is no incentive to keep fee-paying students. So they design the curriculum to get the students the knowledge they need, and if there are one or two tough modules in between, which filter out weaker students, that is perfectly acceptable and, in fact, common practice. The reality is that many courses are oversubscribed, and factor in that after a year or so, many students will have left the course.
A common phrase you hear during the first semester from one of the professors is: “Look to your left and look to your right; statistically speaking, both will be gone next year” (obviously, there is a logical flaw in this statement, but I think this is on purpose, in any case, it is true. After a year more than half of the people will have left their studies). The reason is that many didn’t anticipate how much work is involved in studying or realise that the course is not for them; either way, quite a few quit very quickly.
So, while you may not have to pay any tuition fees (or a fraction of what you would have to pay elsewhere), you may find yourself out of university. Statistically speaking, there is a high chance of failure. So you won’t pay with your money, but with lost time (in which you could have gone on to work and earn money instead).
Let’s look at a place like the UK, where tuition fees are raised and likely require a loan of some description. If you want to study a course that likely will see you repaying a student loan for multiple years, you will ensure that you will study the right course. You look at the website, attend an open day or webinar, talk to course staff or see what former students say about the course. You do a lot more research to make sure this is the right course for you.
The bottom line is that once you commit to paying (and then repay) the tuition fee, you have a strong motivation to finish the course, and this means you will push yourself to a greater extent than you would if the course was offered for free. There is a good complimentary example in the book Poor Economics: A Radical Rethinking of the Way to Fight Global Poverty by Abhijit V. Banerjee and Esther Duflo.
The authors discuss how some people pass up the opportunity to get free immunisation, even if they know that it will prevent them from contracting a disease that is widely spread in their region (and potentially life-threatening). However, if the same immunisation was offered at a price, which was a considerable amount of their monthly income, interest suddenly spiked as it was seen as more valuable.
This phenomenon is well captured in the book $100M Offers, where the author argues that the simplest strategy to enhance the perceived value of a product (which can be something you buy, or immunisation, or a university education) is by simply increasing the price. Granted, you can’t arbitrarily increase the price, but if the buyer (patient, prospect student, etc.) perceives the value to be equivalent to the price tag, you have a good offer.
This brings me to my conclusion: if you think that you will get a financial benefit out of a degree (e.g., an expected salary at the end of your studies through a job where the skills you learn are essential) or if you value the shortened time it takes to get expertise in your field, then yes, I think paying for your studies (tuition fees) is worth it, and they will keep you motivated to finish your studies and do well.
The students I interact with every year are some of the most motivated individuals I have seen. If I compare their motivation to mine (and that of my peers) during my undergraduate studies in Germany, I can comfortably state that we were not that motivated to do well. Sure, we all loved what we studied, got along well, and studied together for exams, but when there were no exams, you would not find any of us studying in the evening and certainly not on the weekend.
When I looked for specialisation courses in Germany after my undergraduate studies, I couldn’t find anything specialist enough in CFD, so I went to the UK, which offered far more specialisations than anywhere else. But if there had been something comparable, I probably wouldn’t have left the comfort of the motherland and still be having pizza Friday instead of oily fish and chips, with mashed peas, what even are those? Anyway, I digress.
In the end, you have to decide whether it is worth the investment and if you are able to cover the cost (personally or through a student loan). If the answer is yes to both, this is a no-brainer. And if you have found a course that really piques your interest, so much so that you lose motivation to study any other course (this was the case for me), then trying to find a way to pay for your studies is usually the right choice, even if that may take some time.
At this point, you may be accusing me of trying to get people into the university I work for. Sure, I see your point, but if that were the case, I wouldn’t give you a list of options where you can study CFD; I would just tell you how great our university is and how bad the competition is. You will find a list of alternatives below where you can study CFD, and these are all good options; the only question is, which option is the right one for you?
As I said, the choice is yours. You need to clarify your thoughts, and once you have, you will probably know what the right decision is!
Places to study CFD
Let’s look at some places where you can study CFD or specialise in it. The only requirement that I have is that the course is offered in English, which is the language you and I use to communicate on this website, so I can assume that there won’t be a language barrier. The list reflects my personal opinions based on the descriptions available on the course websites. I don’t have in-depth knowledge of all of the courses, but the curriculum should be the most important factor in my view.
I have reviewed quite a few websites and courses, but I haven’t included them all here. If your favourite course is not on the list, it is either because some compulsory modules are not offered in English or because the main focus is not on CFD. In one case, I am aware of one course trying to copy another, and thus, I have excluded that as well.
You will find that most specialisation degrees are found in the UK, which seems to be a trend I observed as well when I was looking for a Master’s degree programme some time ago. Most universities don’t commit to one field of study but rather give you a very broad specialisation (case and point: MSc in Mechanical Engineering, which could specialise in anything; really, there isn’t a specific industrial sector targeted here). To me, that defeats the point of a post-graduate study; there should be specialisation, not just in the MSc thesis.
So, if you are serious about CFD, a move to the UK, at least for one year, may be on the horizon for you. This is why I came to the UK as well, thinking it was a temporary arrangement to study for a year, but somehow, I am still here. It’s a great place to study, but don’t come here for the food or weather.
Cranfield University (UK): MSc in Computational Fluid Dynamics
Let’s get the obvious choice out of the way first. A few paragraphs ago, I stated that I am not trying to get students to secretly apply for our course, yet I start this list with the course for which I am the course director. I have gone through this course myself as a student, and I have the most knowledge on this, so it would be silly for me not to write about it, but for transparency, I have to mention this. With that out of the way, let’s look at the course.
If you want to specialise in CFD, this course is for you. You get an overall well-rounded view of the theory, from basic fluid mechanics to discretisation and writing your own solvers, should you wish to do so. But that doesn’t mean that it is all about coding; writing code is a necessary evil to grasp concepts and build up an intuition, but the course is balanced in that you are exposed to writing some code but learning software (open-source and commercial alike) that is used in research and industry at the same time.
Modules include pre- and post-processing (grid generation and data analytics), turbulence models (from RANS to URANS, transitional RANS, DES, LES, SAS, DNS), incompressible and compressible flows, and discretisation and parallelisation. Students have access to full commercial licenses and enough computing power on a dedicated high-performance computing (HPC) cluster to run DES and LES studies, with a maximum of 12 288 CPU hours per simulation.
There are no exams, and all assessments are report-based (working on a practical problem, be it programming or performing CFD simulations). Graduates go on to work in the aerospace, automotive, motorsport (F1), and related industries, and this is the only course on this list specialising in CFD, without focusing on a particular application area (meaning you get to go deeper into the fundamentals than you would elsewhere).
The course is accredited by the Institution of Mechanical Engineers (IMechE) and the Royal Aeronautical Society (RAeS). It has an industrial advisory board in place as well that meets once a year to propose changes and modifications to the course based on current industrial needs. You can find more information about the course on the course’s website.
Imperial College London (UK), MSc in Advanced Computational Methods for Aeronautics, Flow Management and Fluid-Structure Interaction
This MSc degree is a hybrid between CFD and areas important to aeronautical engineering. If you are interested in CFD and aerospace, this is a worthy contender; if you aren’t interested, you may miss a few modules here.
The course covers the basics of CFD, mathematics and programming, which is good to see. There seems to be a gentle introduction to CFD, focusing on 1D problems, and then the complexity is increased. There is a strong focus on fluid-structure interaction in the course and an area gaining increased interest in the industry from what I can see (and projects I have worked on). I’d argue that is not only relevant for aeronautical applications but increasingly so for other disciplines as well.
If you are purely interested in CFD and want to gain expertise in this area, then you may notice that there are no modules on incompressible flows Furthermore, turbulence modelling and high-performance computing are optional modules, which I personally find strange (especially if you focus on (aeronautical) fluid-structure interactions, which are turbulent and require an HPC cluster, as well as knowledge on how to use it).
There are optional modules on artificial intelligence and innovation management, though, which are an added bonus if you want to explore some new research areas or consider commercialising your ideas. They provide a fresh perspective.
The course is accredited by the Royal Aeronautical Society (RAeS). You can find more information about the course on the course’s website.
Politecnico Milano (Italy), LM (equivalent to MSc) in Mathematical modelling for engineering
This is a course not necessarily focusing just on computational fluid dynamics (in fact, you won’t find any modules on incompressible, compressible, and turbulent flows, but rather one module on computational fluid dynamics, which supposedly covers all of that), but rather it focuses on the theory quite heavily, in particular the mathematics (partial differential equations) and programming (including high-performance computing). As a foundation, this course is great, and you get some flexibility in selecting additional modules you have an interest in.
The main reason I have included this course here is that I have had students from this university, and they are usually among the strongest, as they have studied the right background. They are not necessarily very strong on using software, something which isn’t taught it seems extensive, but this is something which is much easier to pick up, compared to being strong with CFD software and then having to pick up programming and mathematics.
If you want to explore the programming side of things more and specialise in that, or if you are interested, in theory, in CFD but you don’t want to commit just yet, this is a great course to consider. You can find more information about the course on the course’s website.
Swansea University (UK), MSc in Computational Engineering
On this MSc course, you get the basics of CFD and then have a range of optional modules to choose from. Again, there are no modules for turbulence modelling or high-performance computing. Still, they offer deep specialisation in big data and data-driven tools such as artificial intelligence and machine learning, which is an area of ever-increasing importance.
Their course focuses heavily on finite elements, which makes sense if you are interested in combining fluid mechanics with elements of structural mechanics. There seems to be a focus on geotechnical engineering (e.g. earthquakes), and so this explains the focus on finite elements here, which you are unlikely to find elsewhere (CFD is typically dominated by finite volume codes).
Having said that, if you are more interested in environmental applications of CFD, then this MSc may be worth consideration. I like the different optional modules and the fact that you can pick and match the courses that interest you the most, though there doesn’t seem to be a strong focus on CFD itself, but rather CFD in the context of other applications.
The course is accredited by the Joint Board of Moderators (JBM). You can find more information about the course on the course’s website.
University of Southampton (UK), MSc in Aerodynamics and Computations
This is another great course if you want to focus not just on CFD but also aerodynamics. The core modules cover essential parts of CFD, and you will be able to dive deep into its fundamentals while learning more about aerodynamics. There is an additional focus on machine learning that is a prominent trend in other courses as well.
With a heritage in race car engineering, or should I say, having a prominent graduate in the area of race car engineering, there is an additional option to focus on this in the MSc, which may give you some additional insights in this area. There is an additional optional module on experimental methods, which is probably overlooked most of the time, but having an appreciation for how experiments are conducted, and potential sources of mismatch between CFD and experiments is a great skill every CFD engineer must acquire at some point in their career.
The course is accredited by the Royal Aeronautical Society (RAeS) and the Institution of Mechanical Engineers (IMechE). You can find more information about the course on the course’s website.
University of Manchester (UK), MSc in Thermal Power & Fluid Engineering
Last but not least, let’s talk about Manchester as well. They have a rich tradition in fluid mechanics, starting with the work of Osborn Reynolds in the 1890s to contemporary achievements in the field of turbulence modelling (especially RANS turbulence models).
Like other courses, there are no dedicated modules on turbulence and high performance computing (which seems to be a sticking point for me), but elements of that are included in other modules (you just won’t be able to go into depth here). They have chosen to combine CFD here with thermal power, and if you are interested in power generation, engines, or just heat-transfer-related problems, then this MSc may be a good option as well.
The course is accredited by the Institution of Mechanical Engineers (IMechE). You can find more information about the course on the course’s website.
Summary
So then, the question is, should you specialise in CFD, yes or no? In this article, we looked at reasons for and against specialisation, and as always, the decision is down to you if you are willing to invest the time and money to specialise in this field. We looked at places to study CFD as well, and apart from one outlier, the UK is dominating the market when it comes to specialised degrees in CFD.
In the end, it comes down to two questions in my view: do you want to specialise in the first place (do you find CFD interesting enough?), and can you afford it (or are you able to secure a student loan or attract a sponsor)? If the answer is yes to both, this should be a no-brainer. With tuition fees high, this can be a daunting decision, which I took once myself, but having done so, it turned out to be one of the best decisions I’ve made.
I can comfortably state that I get up each day and look forward to my work. If I won the lottery now, I would continue to work, just for the fun of it. It is one of the best things that can happen to you in life, and without my specialisation, I probably would not have been able to achieve the same (though I may have found something else which I would have equally enjoyed, I would rate my chances lower for that).
I can’t make that decision for you, but I can only provide encouragement; CFD is a fascinating area to study, and if you agree, I’m sure choosing to specialise in this field will provide you with very high job satisfaction in the future (and, apparently, longevity as well!)