In this era of commercial codes, colorful fluid dynamics and quick results it is nice to sometimes sit back and reflect a bit about where CFD stands today and where we will be in 5 or 10 years. I'll give my opinion here. I appreciate any feedback.

Today I see a very strong trend towards commercial codes. It has been happening for a few years, but this last year, with the release of a couple of new unsctructured codes, I think the commercial market has taken a big step forward. I personally know of several companies that now intentionally have put all development on their "in-house" codes on hold and instead they are switching to commercial codes. The same trend has also started in academia. First comes the more applied fields like turbomachinery, combustion engines, etc. and soon we'll probably see more and more pure fluid dynamics guys that use commercial codes. I don't know if this is good or bad, but I think it is inevitable. Hand in hand with this commercialization goes also a more widespread acceptence of CFD in industry, and we now see CFD being used in areas where noone even knew what CFD stood for 5 years ago.

Enough said about commerial codes.

Then we have the never ending problem in CFD - turbulence modeling. 5 or 10 years ago everyone were optimistic and eager to tackle this problem. Or perhaps it was just me beeing younger and more innocent ;-) I don't know. Anyway, people had just started to do serious LES simulations, Germano and others came with interesting new ideas. LES would save the world. At the same time we also began to see some promising Reynolds stress models published by Launder and others. These models were very interesting and relied on a sound foundation. In a few years everyone would have switched to RSM some people argued. But then the years went by and not much happened. LES still has a lot of unsolved problems, and it is still too expensive for most applications. RSM models proved to be very difficult to use and they aren't as general as people hoped. In an attempt to take the best from two worlds there was a short new interest in non-liner eddy-viscosity models a couple of year ago, but that soon faded away. Now recently we've seen a similar sudden interest in explicit algebraic RSM models, but that also seems to fade away. Instead people are now going back to use one-equation models in the boundary layers. We've sort of walked one turn around the problem without coming any closer to a solution. I don't know where we'll head next. Perhaps a new interest in LES? Perhaps a completely new approach based on lattice-gas / cellular-automata methods. A couple of years ago I listened to a very interesting discussion about turbulence models between Prof. Piomelli (LES guru) and Prof. Launder (RSM, EVM guru). Interesting enough Pimelli didn't belive that LES would save the world, he was more positive about the new RSM/EVM models developed by Launder and others. Launder on the other hand, had the opposite view, or at least that was what he said then. He was optimistic about LES and thought that it would become more and more useful. It's not very encouraging when scientists working on one approach belives more in another approach. Perhaps we've come to a dead end here. Where do you think turbulence modeling will head next?

I see a big danger in the commercialization of CFD in relation to the turbulence modeling problem. As everyone who has worked with turbulence modeling knows it is extremly important how you implement your turbulence model in your code. With commercial codes this becomes a well guarded secret, and we might end up in a citation where noone knows which equations they are really solving. There is no danger yet, academia is still very much in the front line of turbulence modeling, but in 5 year it might look different. Or what do you think?

Then we have the other side, numerics, I haven't done any research in this area lately, so I migth be a bit out-of date with this. If I write something wrong someone please correct me. My impression from this side is a bit similar to my view of turbulence modeling. 10 years ago things looked bright. We had multigrid that would speed up things, we had new FEM methods that could perhaps give us a-posteriori error control, we had new discr. schemes and implicit methods that were very promising.... and computer power doubled every 18th month. Looking back at this the area of numerics has had much more success than turbulence modeling. Although a lot of problems remains to be solved, we today we have robust methods that give good results for most types of applications. However, it still takes a lot of experience and knowledge to use a CFD code, or perhaps I should say to interpret results and to improve predictions. Perhaps someone in-tune with the latest developments on the side of numerics could give his view on what will happen now?

You might interepret me as if I'm sceptic about the future of CFD. This is not the case though, I'm just saying that we still have a lot of problems to solve. However, I see one very positive thing happening now, and that is that people from different areas are beginning to talk to eachother. 10 years ago the turbulence guys were using their own small boundary layer solvers and the numerics guys were using very simplified turbulence models, and noone talked to the guy who actually had a problem to solve. Today people are much more aware of that in order to advance from this point we have to collaborate. There is no general solution, and what happens in one field very much influences other fields. I'm speaking in general terms here, but I'm sure you can think of a number of examples yourself. This new era of "global thinking" is very positive!

Opps, this is what happens when you're bored with writing on your thesis. I'd better get back to work now. Please feel free to add your view about CFD today and in the future. I've left out a lot of things that should be said... before I end I will just give my guess of where CFD will be in 10 years from now:

o Turbulence modeling is still a *major* problem

o Commercial codes are completely dominating, however, they have become much more modular and it is easy to add your own parts here and there. Also we will see more "networked" computing with people hiring a "code+computer+consultant" directly over the internet. Visalization has become a real 3D thing (not only flat screens)

o Numerics is a more mature field, less problems left to solve here, but we still haven't got any good error-control.

o New problems in focus - multiphase flows, combustions, multi-physics simulations.

o Perhaps a completely new approach - lattice gas or something like that.

Do you share my views? Did I forget something important?

Computational Fluid Dynamics is similar to Experimental Fluid Dynamics. In Experimental Fluid Dynamics, you define a problem ( such as a new fighter external aerodynamic design, a new inlet concept for low RCS, a redesign of a HP turbine blade, a low emission combustor, etc..), then you create some concepts of design. You define the parameters for your problem, such as flow conditions, configurations such that you can derive the optimum solution for your problem through the experimental testing procedure, by varying the combination of problem parameters. To accomplish this, you need to: build a test facility ( a wind tunnel, engine test facility, combustor test facility,....), build various test models ( these models must be designed, fabricated, checked out for accuracy ), sensors instrumented, model installation in the test facility, check-out runs, hundred and thousand hours of testing, data reduction, design review of results, decision for the selection of the optimum design , or the input to the next round of design change. So there are many steps involved. The goal naturally is to find the best solution to your original problem. In Computational Fluid Dynamics, the processes involved in deriving the best soultion are identical to those involved in Experimental Fluid Dynamics, except that the hardware model is now replaced by the mesh representation of the model, and the wind tunnel test facility is replaced by CFD simulation codes. The model fabrication, instrumentation, installation , check-out and testing steps are now replaced by activities on computers. You no longer have to touch the hardware part of the process. In this way, you maybe able to save the "cost" and the "time" in the hardware testing approach. ( It is very important to know that "cost" and "time" are relative and can change from time to time.) No one has demonstrated that by doing so Computational Fluid Dynamics will save you time and cost in deriving the best solution to your problem. In fact, you still have to use Experimental Fluid Dynamics to verify your design. It is too risky to bypass the testing approach. So, what was the difficulties involved in preventing you to get the optimum solution using the Computational Fluid Dynamics ? The first limitation is the speed of a computer. The cost of a computer maybe relatively low now, but it is still very slow. This in turn will force you to build a smaller model with fewer mesh points. ( A model with limited mesh points will always kill you sooner or later .) The second limitation is the modeling of the physical processes. This includes the turbulence modeling, numerical solution of Navier-Stokes equations, etc.. This is a tough job to do. And you are hoping that someone somewhere will do it for you. In reality, you will have to deal with it yourself. Because someone else is already taking care of the computer speed problem for you. So, you don't have to worry about creating a fast computer first. ( In some parts of the world, you may have to worry about the computer problem.) At this point, it becomes clear that there are two items sitting in front of you : 1) turbulence modeling, and 2) numerical solution of Navier-Stokes equations. And the problem is that very few people have any knowledge about these two subjects. Can you go to a school and take a course in turbulence modeling, and be able to model the turbulence in your problem ,current and future? Can you go to a school and take a course in numerical solution of Navier-Stokes equations, and be able to write a code for your problem,and obtain the right solution ? Then, you hope that someone else will take care of that for you , so you can simply click click click to run a code and you will have the solution right there. The first Low Reynolds number two equation model was published more than 25 years ago. When you are using a turbulence model, do you really know how it was derived ? and how to use it ? The fact is : we know almost nothing about these two subjects. Our only knowledge about the turbulence was mainly derived from the experimental data. What we are doing is nothing but data fitting or curve fitting. Can you say you really know how to solve the incompressible Navier-Stokes Equations ? The state-of -the-art of Computational Fluid Dynamics today is nothing but "Experimental". The future of the Computational Fluid Dynamics is still wide open in those two subjects I just mentioned. ( Remember that a faster computer can only give you the wrong answer faster. For the right answer, you would use a slower computer, so that you can charge your customer more. just a joke.) As for the commercial codes, I think it depends on the market need. The contents of a commerical code is not important at all. The most important factor is whether you will be able to satisfy your customer's need in providing the commercial code. Therefore, the ability to difine the customer's need will be the most important issue. Unfortunately, it is outside the scope of those two subjects I mentioned earlier.

I agree fully with you. CFD is still very much an experimental exercise... "lets try it and see if it works". And for the last 10 years the major problems have been and still are turbulence modeling and numerics. The interesting question is, where do we go from here? I'll ask a few retoric questions.

Will CFD remain a big academic subject?

Will everything be based on Navier-Stokes also in the future?

Will turbulence modeling ever take a big leap forward? This subject desperately needs a new "Einstein"! Do you see any hopes on the horizon?

Which trends do you see in numerics? FVM/unstructured methods a'la Fluent and others?, FEM/multiphysics a'la Centric/Spectrum?, Lattice Gas a'la Exa? Will we see only unstructured/grid-"simple" methods or will people still spend weeks generating grids ten years from now?

As computers and methods improve which new application areas will we see? Perhaps CFD will split into several separate subjects related to each application field?

We need the kind of people like you. I'll answer your questions one by one. As I said before, there's nothing wrong with CFD. Coupled with other tools, CFD will be the way to design better products. I'm sure that even right at this moment, computers in government lab's, academic institutions, aircraft companies, engine companies, private companies, and even the individuals are running CFD codes. The basic problem is that the job market has changed. The budget cut back and the company merging have the serious impact on the development of this critical technology ( many 3-D CFD codes for advanced applications are under export-control). Because I have worked for many large and small companies, government and private, I know what they are doing in the CFD area. They are running codes everyday, but they don't have resources to improve the existing codes or to develop new codes. You need people to do that, and it costs money ( it's created by the different way of thinking in 90's. You didn't see it in 70's or 80's ). I'll get back to the main issues later on, I have to worry about my phone bill first.

I tend to agree with your points. However, I think it is a mistake to understimate the importance of LES in industry in the future. Companies that are faced with intrinsically rotating flows such as aircraft engine companies (Snecma-GE here in France) have started to sub-contract LES studies to government research agencies because all existing turbulence models gave them unsatisfactory results. By the way, I am not familiar with lattice gas methods. Could someone tell me more about them ? thanks

Yes, I have also noticed a renewed interest in LES lately. But I haven't noticed any new promising ideas... but then I'm not an LES guy, so I might just have missed it. My feeling is that we're heading for a new "turn around the problem". Anyone see any promising new ideas around the corner in LES simulations?

About lattice gas /cellular automata methods. This is definitely not my area, but I can tell you what I know. The basic idea is that instead of starting from Navier-Stokes and trying to solve those equations you start from a fluid-model based on discrete fluid "particles" and you have models that dictate how these "particles" interact and collide - a bit like the molecules in a gas. If your model is correct these "rules of interaction" will in the limit reproduce the Navier-Stokes equations. The advantage with this is that the "particles" have discrete states and that simulations can be done in pure integer on massively parallell machines (very fast). In the old days these methods had a lot of problems - models were not Galilean invariant, noise and problems with viscosity made high-Re simulations virtually impossible, ...

Lately there has been a renewed interest in these lattice gas methods and I think that some of the problems with the old models have been solved. Perhaps someone else can tell you more. If you want references check out papers by "Chris Teixeira". For a background on cellular automata stuff Stephen Wolfram "Collected Works on Cellular Automata" is unbeatable.

(I can't spend too much time online bacause I've to pay my phone bill. It's taking too long to type, so, I decided to write my manuscript on paper first. At least,in this way,I'll be able to save some money. And at the same time I'll be able to talk with you.) As I said before, I'll cover all of your questions, but not in one day. It may take a while. Giving out too much information at once will cause"indigestion". "information Indigestion" is a common sympton of today's common people. The problem has reached a point that"We don't know where we are going." This is a serious disorder, sickness. "Where do we go from here?" The answer is:"We are going to war." The second world war created the need to design nuclear devices, and the computational fluid dynamics simulation was born at Los Alamos National Lab. The space race between U.S. and U.S.S.R. during the cold war era created the need to model missiles, fighters, bombers and space shuttle. This inturn made DOD,NASA,seven airframe companies very busy. And in order to have a piece of pie, a defense contractor had to invest in the CFD to win a contract. The change of world political environment in 90's has shifted the war to the common people. Recently, a newspaper report estimated that nearly two hundred thousands people were displaced from aerospace related industries in southern California. And the trend is still continuing today. The job posting on CFD-online also has the similar effect of creating war between common CFD people, at even higher levels of intensity. It becomes obvious that the right solution to relieve the suffering from the common CFD people is to shift the war back to the political level. That is to remove the "War" from the common CFD people and put it at the organizational level. With this CFD-War principle, you can easily understand why a racing car designer is trying to use the CFD to come up with a better design of airfoil, induction system,and external aerodynamics. The name of the game is:"War". Looking from this CFD-War principle for a while, and you will suddenly see the light from the other end of the tunnel. "Which way to go " will become very clear to you. ( If they know that by using a commerical code on a super-computer will solve their design problem and thus win a contract competition, they sure will try very hard to do it). For the commercial code developers, if they can not help their clients to win the "War", sooner or later, they will cease to exist. And those companies don't invest enough in CFD will also cease to exist. The same principle applies to individuals also. In 80's, the CFD development was mainly controlled by the computer hardware (super-computer,etc..). In 90's, it's being replaced by the workstations. And in the next century ( a couple of years from now), it will be replaced by PC workstations. The CFD war will be everywhere. It is up to you to decide whether to adopt the CFD-War principle to survive, or to ignore and vanish. But don't worry about the financial resources,because the money accumulated in 90's through the merging and reorganization of companies, will have to be invested into the CFD development at that time. They will be forced to invest or they will simply die. So the conclusion for this part of answer is: " Countries, companies and people will soon be forced to adopt CFD-War principle in order to survive, otherwise, they will simply die." ( a Taiwanese computer company CEO recently said that manufacturing PC hardware is just like biting bones there is no beef, no profit.) The beef is in the software. For an engine company to compete in the global market, they will be forced to invest a lot in CFD software.( and you can only get that from an well trained engineer, not from a CD- disk ).

Too many questions at once, let me add my two bits on this one:

"Which trends do you see in numerics? FVM/unstructured methods a'la Fluent and others?, FEM/multiphysics a'la Centric/Spectrum?, Lattice Gas a'la Exa? Will we see only unstructured/grid-"simple" methods or will people still spend weeks generating grids ten years from now?"

I spent about 10 years working with spectral methods (pseudospectral to be exact), and I believe that we will see some interesting things happening in the field. However, these changes will probably be evolutionary rather than revolutionaly (no Exa type quick fix).

First, let me explain, why we need these weird spectral methods at all. In my opinion, many failures in applying modern turbulence models originate in people attempts to build their sophisticated turbulence models on top of rather primitive numerical approximation procedures. Many engineers working with finite difference and finite element methods do not realize how rude these methods actually are.

In most cases, applying turbulence models for 100x100x100 finite difference, or finite element grid is a waste of time, since numerical approximation error, normally in the form of 'numerical viscosity' (engineers like to use upwind schemes to hide negative density from their bosses and customers), will be considerably bigger than 'turbulent viscosity'. The same grid used for collocation in pseudospectral method will produce a solution, which is order of magnitude more accurate, and it will leave us some degrees of freedom to experiment with adding some 'turbulence'. In fact, working with (pseudo)spectral methods we can correct the spectrum of, say, velocity component, or pressure, rather than add some artificial terms to the Navier-Stokes equations.

The reason why finite difference/element methods dominate in CFD now is that spectral methods are still too complicated technically to apply, especially for complex geometries (which is probably 95% of engineering problems). People prefer to have quick and dirty solution now, rather than nice and accurate one 2 years later.

In my opinion, in the next few years, spectral methods will gradually merge with finite element methods, producing what they sometimes call 'spectral element' methods. Instead of producing a sophisticated grid of a few million nodes, we can split the computational domain into several dozens of subdomains, and use (pseudo)spectral method(s) in each of these subdomains. Each of the subdomains may contain a few millions of collocation nodes, but it may be 'simple' structured grid, which makes our life easier. In general, it is going to be harder than the methods they use in commercial solvers now, but, I guess, it is the only way if we are serious about solving problems involving turbulence rather than pretending that we are solving them.

CFD-War Principle::Identify_CFD_Target; One of the most important function under the CFD-War Principle is the Target Identification. What is CFD target ? Where is CFD target? How do you define CFD target? The need to address this question is real. People ( at management level ) have talked straight to me that they do not understand CFD. Most people at management understand money, but they normally do not understand CFD. It is very hard to sell to the management that you need to do research to study the mesh generation, or to find a new model for the near wall region. The problem is mainly created by the missing of the CFD target identification. Without the CFD target identification, it is very hard to link your part of the job to the bigger management picture. Therefore, CFD researchers must look at this CFD target identification everyday. How do you identify the CFD target? The easiest way to do is to find the CFD target from the real world applications. In this was, you can see it, feel it, associate value to it, communicate your idea with management,iet.. If you look around, you can easily identify these CFD targets. Here are some examples: automobile external aerodynamic design, racing car aerodynamic design, engine intake manifold design, fuel injection and combustion chamber design, under-hood air cooling design, engine block water cooling design, air conditioning ducting design, subsonic aircraft design, transonic aircraft design, supersonic aircraft design, aircraft turbine engine components design ( fan, compressor, diffuser, combustor, turbine,nozzle etc.), engine-aircraft integration design, launch vehicle external aerodynamic design, solid rocket internal aerodynamic design, nozzle design, liquid rocket design, turbo-pump design, missile external aerodynamic and control design, film cooling design, turbine blade internal cooling passage design, etc., etc.. Almost all of the fluid dynamics related product design can be improved by using the CFD approach, and thus resulted in quality improvement. So the first step into the next century, CFD researchers must think in this direction, before he/she starts the daily CFD activities. ( Without this, one person will tell you that you must do a good job in mesh generation, while the other person selling the code will tell you that the mesh is automatic, you don't have to do anything. The management sure will be very confused. He does not know which way to go. To Do or NOt To Do must be settled before the next century.) I hope that by focusing on the real applications as the first step in CFD, you will have a better life in the next century. With this, you should be able to take care of your boss. I'll cover the next subject soon. Thank you for your time.

After looking at these messages, I would like to share some information with all of you guys.

In the past few years, I was trying to learn how to do CFD by spectral method in Clarkson University. After, coming to Brown University, I find out that people in here are expanding numerical computation to a new area, Computation Electromagnetics(CEM) in time domain, which is solving Maxwell equations by spectral method.

Currently, the CEM time domain methods are based on a sceond order finite difference scheme proposed by K.S. Yee in 1966. However, the time domain methods were not so popular until 80's. During that time lots of centers and labs were established and people started to work on electromagnetic problems by developing new schemes and methods.

Maybe you guys will ask what does CEM related to the people in CFD? It is the numerical computational level that CFD people can do on the CEM such are constructing new schemes based on the technology developed from CFD and also learn something from CEM such as PML(perfectly matched layer) method, one kind of absorbing boundary condition, which can be applied on accoustic problem.

For people who are interesting about CEM, one may try to search the net by using the key word computational electromagnetics or finite difference time domain methods.

I think the Computational ElectroMagnetics is very important to the design of stealth vehicles. I hope you are not saying that many people at Brown University are working on defense projects or technology. As a matter of fact, some CFD researchers had already started working in this area back in 80's. This is sure a very important and sensitive field strongly related to the aerodynamic design.

Heh, I like this CFD-war and target identification theory. I fully agree with it. I was trying to say something similar in my first post in this. Focusing on what problems need to be solved for a particular application and a closer interaction between numerical analysts, turbulence modelers and application engineers (and bosses?!) is what will be the driving force in the future. This is one of the few new and really positive developments that I see in CFD today... of course Moore's law is also always on our side (computing power double every 18th month).

Well, scattering problem is an area to work on. However , that is not the only place to play with. For example, transmission line modeling(TLM), wave guides in optics, microwave technology and communication system, these fields are starting to use finite difference to solve the Maxwell equations.

There is a center of fluid mechanics but not center of electromagnetics in brown university, which means there are not so many people as you think working on these problems yet compared to CFD in here.

CFD-War Principle::Know_your_CFD_Target; Everyday,when you get up in the morning, you look at your CFD_Target, after lunch, you look at the target again, from different angle. In the evening, the target disappears into the sunset. After a while, you become curious about your CFD target. You decide to find out more about your target. You talk to your friends about it. You read books at the city library. You read magazines, and you study textbooks. Then you start surfing the Internet, asking smart questions, hoping someone will give you a straight answer. On and on, your target becomes part of your life. This normally takes three years, regardless of your global locations. Well, this may have something to do with your brain structure. ( This rule does not apply to monkeys, because they maybe smarter than you.) Let's just say it takes three years to know your CFD Target. In the first couple of weeks, you think your target is there. It may take another three months to realize that the target is really there. If you are lukcy, you will be given six months to dig into library network to find out what's available out there. The more you read, the less you know about your CFD target. A year and a half later, you suddenly realized that there is really very little available on your CFD target, and you also know very little about the target. After three years, you finally developed some kind of respect to your CFD target. ( If you still are not convinced, then you should just forget about your target chasing and go to the nearby mall and play the arcade games. Or you can watch Olympic Winter figure skating. Based on my intuition, three years is typical for a person to get some "feeling" about the joystick and the ice.) At that point, you are well into the second stage of CFD_War. If you don't follow this principle, you don't get a chance to fight the CFD_War at all. I'll cover another subject soon. Thank you very much for your time.

It is really nice to know that CFD reseachers have the ability to invade other field. In general, Computational ElectroMagnetics should be part of Computational Physics. Well, that is because the finite-difference approach is a mathematical approach. It is not unique to the computational fluid dynamics. Don't you think that people specialized in commercial code applications would have a harder time to work in CEM?

CFD-War Principle:efine_Your_CFD_Goal; Now that you have identified your CFD Target, and you also have some feeling about it, the question is "What are you going to do with it?" The goal is something you can set on your own, it is not something you are forced upon. Do you like to work as a nurse aide? Do you like to work as a licensed practical nurse (LPN)? Do you like to work as a licensed registered nurse (RN)? Do you really think you like to be in the nursing business? How do you prepare for this CFD job? Can you study at home? Can you study by yourself by reading a self study book? Do you really understand what was covered in the book? Do you have to actually run a machine in the hospital in order to learn how to use it? Do you have to practice on patients to give them a shot? Can you just go to a hospital and apply for a job? Can you find your job only through the CFD-online? What are my chances of getting a CFD job through CFD-online? Do you need to take a license examination? What are you trying to achieve in your CFD goal? Do you think you need to join a CFD union before you can get a CFD job? Do you need a license to join a CFD union? In old days, when we run a CFD code using a card deck, we normally have an engineering aide to prepare the key-punch cards, the JCL cards, and submit the deck at the computer room. She will check the output box from time to time and prepare the plots from the printout. In 21th century, can you still find this job? Do you like it if it's still available? Where is your CFD Goal and where is your CFD Job? Before you rush into the 21th century CFD Battle Field, you sure would like to study your CFD Goal first. I'll cover technical aspects of CFD next time soon. Thank you for your time.

CFD-War Principle::Polish_your_CFD_Technology;Somehow through this self screening process, you have become a potential CFD soldier. What are you going to do next? Go out and buy a commercial CFD code and pretending that you are a GI Joe or GI Jane. You think your enemy will spend a lot of money and creat a commercial code, so that you can use it to develop a better product on the market to get rid of him? The right way to do is to develop your own CFD technology. This is because once your enemy knows that you are using a particular commercial code, he can easily come up with a better product. He can easily find out the limitations of the commercial code you are using and develop a better codes or methods. The CFD technology can be classified in several different ways. I'll give you some examples so that you can dig into each area and develop your own skill. 1). incompressible flows vs compressible flows. 2). 2-D vs 3-D problems. 3). laminar flows vs turbulent flows. 4). steady-state formulation vs un-steady state formulation. 5). non-reacting flow vs reacting flow. 6). sub-sonic flow vs supersonic flow. 7).hypersonic flows. 8). boundary layer formulation vs Navier-Stokes formulation. 9). explicit method vs implicit method. 10). S.O.R. (successive over-relaxation ) method vs A.D.I (alternating direction implicit) method. 11). elliptic equation vs parabolic, hyperbolic equations. 12). primitive variable formualtion vs vorticity function formulation. 13). free-surface problem formulation. 14). fixed boundary vs moving boundary problem. 15).fixed reference frame formulation vs rotating reference frame formulation. 16). Cartesian coordinates vs general coordinates. 17). external flows vs internal flows. 18). pressure-based formulation 19). central, upwind-wind, exponential law, power-law formulations. 20).body-fitted coordinate system. 21). algebraic mesh generation vs numerical mesh generation. 22). adaptive mesh generation. 23). finite-difference vs finite-volume method. 24). finite-element approach. 25). eddy-viscosity vs stresses equation turbulence modeling. 26). 1-equation vs 2-equation turbulence modeling. 27). high Reynolds number vs low Reynolds number turbulence modeling. 28). numerical consistency, stability and convergence. 29). grid-independent solution. 30). grid stretching. 31). structured mesh vs un-structured mesh. 32). general parametric curves, surfaces, patches. 33). surface mesh vs volume mesh. 34). multi-phase flows. 35). combined structure and fluid interaction. 36). graphic display of geometry, mesh, solution and animation. 37). speed-up schemes. 38). code development and validation. 39). user-friendly interface. 40). size, speed, and accuracy of the code. etc...etc... There is no way to cover all these areas in your PhD study. So, you say you are giving up. You are going to use commercial codes or someone else's codes to save the trouble. This approach may give you time to re-think, but eventually, if you don't have the skill, you will be killed ! ( using a commercial code is different from getting help from commercial code company, in the latter case, you are getting help from a human being. The code in this case is just a vehicle.) Next time, when I am not busy, I'll cover some of these topics. Thank you for your time.