[wolfkinara]

Hello all,

I am doing my masters about sustainable energy/PEM fuel cells. I studied mechanical engineering in my bachelor and was very interested in fluid mechanics. I think I am a good ANSYS-Fluent user and I am also proficient on ANSYS-ICEM CFD, I co-authored 2 SCI papers about CFD modelling of PEM fuel cells via Fluent during my bachelor. In this semester I took CFD and turbulence theory courses as an electives, however the problem is that the courses are full of theory and only deals with 1-D and 2-D simulations with MATLAB. So I dropped them. However, still I am not sure because in the future I want to integrate my lab. work with simulation. I believe that even if I do very good job with those courses I wouldn't be able to simulate real world examples without commercial CFD softwares. So I decided to take courses about physics and materials instead, and studying COMSOL on my own. What do you think? I need suggestions

[BlnPhoenix]

"I am doing my masters about sustainable energy/PEM fuel cells."
Good.

"I studied mechanical engineering in my bachelor and was very interested in fluid mechanics. I think I am a good ANSYS-Fluent user and I am also proficient on ANSYS-ICEM CFD, I co-authored 2 SCI papers about CFD modelling of PEM fuel cells via Fluent during my bachelor."

Good. Some students do not even start with CFD in bachelor.

"In this semester I took CFD and turbulence theory courses as an electives, however the problem is that the courses are full of theory and only deals with 1-D and 2-D simulations with MATLAB. So I dropped them."

Ok, so you dropped a course that you felt is not close enough to real world problems. No biggie. Look for an interesting alternative that relates to CFD and real world application.

"However, still I am not sure because in the future I want to integrate my lab. work with simulation. I believe that even if I do very good job with those courses I wouldn't be able to simulate real world examples without commercial CFD softwares. So I decided to take courses about physics and materials instead, and studying COMSOL on my own. What do you think? I need suggestions."
Very good, you already did what i suggested above.

Try to do what you think helps you solve real world engineering problems with CFD. Also try to do something CFD/industry related in your Thesis and or internships. Good luck

[sbaffini]

This is really a 1 Million dollar question.

For sure, it is quite uncommon, if ever happened at all, that someone with the skills to actually develop industrial level CFD software then ends up doing something different. This is both because you can't do it efficiently as a part-time job and because the effort required to get there somehow implies that you really don't want to do anything else than CFD.

So, in your case, I think we can safely exclude the acquisition of top level knowledge for CFD. But, do you need, at least, a little to safely use a commercial CFD package?

Your mileage here may vary as a function of the underlying CFD package. Most companies nowadays advertise their product (or one of their products) as "democratic", hassle free, etc., implying that you don't need CFD knowledge at all to use them.

Besides my personal disagreement with these statements (if they were really so democratic, these companies would not need to provide support anymore), I consider myself a CFD developer and when I think about this I feel pain for those using a CFD code without a minimum, well organized, knowledge. The same pain I felt when I started.

You don't probably need to remember the equations or be very precise, but how can you choose an algorithm, a wall function or any other model without having a minimum understanding of how it works and its limitations? Because this is what we are talking about, using a tool that has known limitations.

You wouldn't allow a 10 year old to drive a F1 car, nor would allow a non expert to make the structural verification of your building, or less than ever a novice fly a 747.

The only reason CFD is allowed for non experts is that it has no consequences. If it has no consequences it is not used correctly and is just a waste of money.

This has nothing to do with the courses you mentioned. They might still turn out as obsolete or useless. Only you can know that.

[LuckyTran]

Well put it this way. Not many people take an advanced theoretical calculus class and are able to prove the fundamental theorem of calculus, or any theorem in calculus for that matter. But most people do know how to use calculus.

The underlying science of a field is distinct (but not necessarily mutually distinct) from any utility that it provides. If your goal is only to do real world CFD without any understanding of how it works, then learning the theory doesn't contribute to that goal and you probably won't like it.

You can use a hammer to make a house without knowing any material science or any laws of physics. If your goal is to build real world houses and make real world money, then going to school and getting any degree doesn't really help you earn any more money. I like hammers & houses doesn't mean you'll like learning about material science, structural mechanics, continuum mechanics, etc.

Figure out what you want and pursue it. You do need a drive to understand the principles of of how stuff works to enjoy learning theory.

[sbaffini]

I still have to decide if this was sarcastic, but in the meanwhile let me politely disagree.

Originally Posted by LuckyTran

Well put it this way. Not many people take an advanced theoretical calculus class and are able to prove the fundamental theorem of calculus, or any theorem in calculus for that matter. But most people do know how to use calculus.

Besides the fact that my calculus classes were just made of both parts, a classical counter example is that of people trying to integrate/differentiate stuff that is not integrable/differentiable, a thing that can have several names depending on the specific operation and field. When the integration/differentiation is numerical the failure is just spectacular.

Originally Posted by LuckyTran

If your goal is only to do real world CFD without any understanding of how it works, then learning the theory doesn't contribute to that goal and you probably won't like it.

This is so wrong, but let's put it this way: everything you don't know is , soon or later, going to waste your time trying to figure out stuff and/or calling the support. Somebody paid for support, but what about the time?

Originally Posted by LuckyTran

You can use a hammer to make a house without knowing any material science or any laws of physics. If your goal is to build real world houses and make real world money, then going to school and getting any degree doesn't really help you earn any more money. I like hammers & houses doesn't mean you'll like learning about material science, structural mechanics, continuum mechanics, etc.

The only way to do that without a degree granting path is learning the same stuff with some other path, which is ok. Unless we also disagree on the "real world" measure. Don't take me wrong, but a strong renovation of a building in the historical center of Rome or the construction of a 3+-storey building is not the same as building a wood house that will be taken away by the next storm.

Originally Posted by LuckyTran

Figure out what you want and pursue it. You do need a drive to understand the principles of of how stuff works to enjoy learning theory.

Ok, I actually agree with this.