[dhavalpatel1128]

Hello Everyone,

My name is Dhaval Patel. I am a senior mechanical engineering student at Temple University located in Philadelphia, PA. What brought me here to this thread is that I am taking a CFD course the current semester and my goal is to learn about CFD, its software, etc. but most importantly, I want to interact with people from all around the world who share the same interest as me. Please provide me few tips and tricks to be successful in this class, it could be theory or software or anything. I am not really smart with coding, so my question is if there is a lot of coding required in CFD? What type of coding is done? Does CFD have a good scope in the engineering field? My interest is in airplanes, jet engines and space transportation, and technology. One day, I hope to find a job in relative companies, so people who are experts, does having CFD knowledge give you an extra leverage in hiring decision?

[FMDenaro]

No one can give the right answer for you ... coding in CFD is relevant in academic environment and research field on the other hand nowday many CFD commercial code user has almost no knowledge of CFD basic and they work in industrial application....

[DarylMusashi]

Hello Dhaval,

Nowadays there are a lot of sophisticated and user-friendly CFD packages out there, so changes are good that you can mostly "avoid" coding if you are going to work in an industrial environment. But I am working in a CFD engineering office and we often code scripts for the post-processing of our projects (for example in Python or Shell/Bash scripts).

To dwell deeper into this interesting field it is necessary to write your own programs, I think. You can gain some useful insights of the behavior of different numerical schemes and turbulence models, for example. The AIAA offers some CFD courses for this purpose and might be interesting for you (Introduction to CFD, Advanced CFD, Turbulence modelling in CFD). During these courses FORTRAN is used. https://www.aiaa.org/CourseList/

Additionally, in the academic sector programming experience is more important in my opinion, because in this field non-commercial in-house CFD codes are more common. Even if you will rarely need to program it from scratch, you might need to adjust or further develop existing code.

[dhavalpatel1128]

That is really cool that you making a living from doing CFD. My goal is to get more in-depth knowledge in CFD and hopefully work for a firm or company modelling CFD on their products. I still want to work as an engineer, I like the design aspect of it, but CFD gives more mobility to work anywhere and anytime. Yes, I would love that. Any tips on how to grow my knowledge from beginner to an expert level?

To other users, please share your thoughts on CFD and how it plays a part in your life. What cool CFD models are you working on? Share and I plan on sharing what I find.

[mehakkar9999]

dhaval i am running a cfd group on whatsapp if you are intrested then ping me on +919033973127

[RobertB]

If you want to succeed focus on the product, how it works and what it needs to achieve. You are paid to design stuff and create products to sell.

Far too many analysts focus on the analysis as an end in itself. Just like experimentation, the analysis result is the beginning and not the end of the process.

Concentrate on the physical insights that the CFD gives you and focus on understanding the major drivers in the performance of the product.

Accept that the physics aren't perfect but, in many practical cases, CFD gives you the best chance of understanding how things work as it is so data rich. Many experiments are highly flawed both in geometry and physics. Just because someone measured something that doesn't make it right!

On the issue of accuracy, all tools are not perfectly accurate. It cracks me up when someone complains about CFD accuracy then takes the output of a 20 million cell model and boils it down to a number which they then use in a 0D equation to take their next guess.

In the case of general design (as opposed to calibrated design systems) complaining about the accuracy typically means that the person doesn't, or doesn't want to, understand what they are designing and how it functions and just wants the analysis to be a 'perfect' result.

Engineering is about using the best available tools to make the best possible decisions on a particular day.

[dhavalpatel1128]

Yes, you are absolutely right. People do tend to get the results and not care about how they are getting their results.

When you want to create flow analysis on, for example; entire fuselage of Airbus A380, How long does it take to run entire solution? If time is a factor of urgency, then don't you think it would be much better to make a small scale model and test it out inside wind tunnel?

I have heard solutions take 3 to 4 days to come out, what are your opinion and thoughts on this?

[ashishvshelke]

Time for simulation depends on your software used and obviously meshing.