I'm not sure if this is the right spot for this but I would like to get some feedback from some pros. I'm currently in a bit of dilemma of deciding of going back to get a MS (or Phd) or try and get a job in the CFD arena.

I currently have no experience in CFD other than taking a few courses in numerical HT courses (soon will be getting into CFD) but would love to get into it. I have talked to professors and they say there is a demand for it. Is this true? I'm finding it difficult to find an entry level position just get my foot in the door and I thought that my lack of training might taken away if i go ahead and get my advanced degree. Does anybody have any suggestions?

Thanks Jason

(1). To answer your question about " Is it worth it? ", I have to say " No ". (2). This is based on several factors. I will explain it in just a few moment. (3). As for getting a job, I would like to share with you my son's comment. He used to say, " Even if the un-employment rate hit 10%, there are still 90% of people working ." (90% of working people working) So, if you decide to spend 4 more years to study CFD, it is also all right. (4). Now back to the real world. In the era before 90's, the major industries hiring CFD engineers are aerospace related companies and government laboratories, such as airframe companies, Boeing, Lockheed, Rockwell, Northrop, Grumann, General Dynamics, McDonnel-Douglas, TRW, etc.. and engine companies, such as GE, P&W, Lycoming, AlliedSignal,..etc and NASA/ Ames, Langley, Lewis,.. Now you can feel the size of the industries and government labs. These companies and labs all have their own dedicated department to do CFD. (5). Through 10 years of cut back, there are only two airframe companies remain. And with the government order for aircraft reduced, and with the global Asian economic crisis, even the remaining commercial aircraft company has to seek help from government for help in foreign competition. Now you are getting the tough picture. (6). So, even if there are positions in CFD available, the experienced CFD engineers will be able to get the job first, not mentioning the thousands of displaced experienced engineer looking for opportunities to come back. Now, the situation is really impossible. (7). In 90's, the trend has been for the CFD engineers to become their own boss, as consultants, consulting company, and CFD vendors,...or even as 3-D modellers for animation industries. And I am sure that some CFD developers or vendors were at some time in the past employee of aerospace companies. (8). Unless somehow US gets involved in the future world war 3, or somehow the global or Asian economy suddenly change the direction, the picture will remain the same. And the CFD job will be hard to find. (does not mean there will be no CFD jobs) (9). The other factor is because of the isolation of universities and the defense industries, the real contents of CFD in industries were not fully understood by the universities. This makes the matching difficult between the student and the industries. (professors did not know exactly what the engineers in industries were doing.) And even in the industries, there are two lines of thinking; one is the traditional and one is the CFD oriented. This was all ri;ght when the business or industries were booming. When the time is tough, the CFD part is usually the first to cut. (it is not mature and so it is not essential.) (10). Now with some commercial CFD codes becoming more mature, companies are hiring engineers just to model the problem and run the codes. (11). Well, I think, the Internet related services are the current and the future job market. In the capitalist countries, the need from the market is essential. (12). You have to learn how to survive first. If you love fluid dynamics and CFD, then you can ignore my comment completely. (I loved to build model airplanes and small rockets when I was a small child, long before the Appolo and Sputnik programs. CFD is a field for those who really like to work with computer programming, numerical methods, thermo/fluid dynamics. But the major driving force behind is long gone. It is much easier to get in the Internet related field because the support is everywhere.(both money and the problems)

So, CFD -code development is saturated. Industries are just wanting people who can run the existing codes and model their problems. Agreed universities are not aware of what's going on in industry.

But somewhere ( even without the world war III or Asian crisis..turning things upside down) ..'the already existing NEW techniques in academia' may seep into industry overnight. Example ..if lattice boltzmann techniques could Shape companies like EXA...how far are 'spectral elements'.. How long will it take 'lower dimensional models' to find a place in industry..?

I feel ..if nothing a good CFD specialist can at least end up at some convenient place in the 'SIMULATION' industry applying ( again at least) a part of his computational skills in his/her job.

So..why NOT CFD ? ( even in the worst case scenario..)

Prabhu

(1). It has nothing to do with CFD , or the potential of CFD, or even the plan to study CFD. There is nothing wrong with these issues. The issue is related to the investment in a MS or PhD degree and getting a job in CFD. (2). As is today, many number-one, number-two companies in the country, or in the whole world, are still struggling with 20 some years old codes, 1-D ,2-D or inviscid codes. (3). And so far up to now, I have not found any commercial CFD codes which can really make a positive contribution on the design of products or understanding of physical phenomenon relevant to the improvement of product design. (4). If any such fact exists, it is because of the engineer's insight of the problem and his understanding of the fluid dynamics. Yes, the CFD codes have been used for a long time in industries (aerospace industries), but they were carried out systematically along with endless testing of hardware. Not mentioning the hardware testings specifically designed to fine tune the CFD codes. (5). It was possible, because such support were available in those days. Without such support, it requires a magic word "love". " Love in CFD" is essential for survival, if the person understands the "love" part and the "CFD" part correctly. (6). For Internet field, even if a company is not making money, and even if everyone knows it, they still put the money in the Internet stocks. On the other hand, even if one has created a super-fast CFD code, he still has to convince a lot people to use the method in the CFD code. (7). Progress in CFD still will be made by some researchers ,regardless of the future job market need. (8). You can open the ASME magazine and do some research for the last five years in terms of the number of CFD jobs listed in the number-one mechanical engineering magazine. You would be very disappointed if you look into the AIAA's Aerospace magazine. (9). Anyway, there are still some CFD jobs around to do modeling in automotive industries, electronics equipment,...etc. But, these jobs are really very specific. The abilities of a young PhD to make a significant contributions in these areas are limited. (10). As a matter of fact, the published technology should be able to let the CFD to make a much bigger impact on many fields. Unfortunately, it requires a smart engineer with a good supporting environment to achieve this. When you have to worry about the monthly house mortgage payment, CFD simply does not exist at all in the daily activity list. ( A couple of years ago, a friend said to me that one of his friend was studying to be a dentist and same time helping an engine company as a summer job worker. My friend and his friend used to work for an engine company which in no longer on the map. (11). So, based on his friend's experience, dentist job seems to be a better choice in this unstable world. Is the number of CFD researchers larger than the population of whales?

I have possibly gotten a promotion in my current company doing R&D work. It is a very good opportunity for my career and COULD possibly lead into doing FEA in heat transfer. But then again it may not..I know everyone has heard that more Americans are not going to Grad school and my way of thinking leads me to believe that I will be more marketable in the long run if I did go. At what point does it make sense that a person should get an advanced degree.

A friend of mine went ahead and got an MSEE and is making about 20k more than I am... I also know several others that are getting more because they have advanced degrees even though I have been working in the industry instead of getting my advanced degree. I thought that the working engineer would make just as much (maybe a little lower) than the one who has an advanced degree.

It seems we are lied too in engineering school about pay and job markets. It took me 6 months to find a job and still got paid minimally when I did find one. I do agree that getting an MS/Phd is risky, especially with the economy the way it is jumping around, but aren't those with the higher degree usually last to get the axe because of their knowledge base?

(1).If you decide to go to the graduate school, you must do it as soon as possible. In the school, you also like to attend the one which will give you the degree as soon as possible. Some schools will take you 6 to seven years or even more to get a PhD degree. Also a degree has nothing to do with your actual knowledge at all. Actually, you can get a MS degree in a year and a half. And for those who take time to get a degree when working at the same time, normally do not get a salary raise at all after they obtain the degree. (2). There is probably a standard difference in salary between a BS and a MS engineer. But this is normally because the one with a MS degree must invest both the time and the money first, in order to get the degree. So, they have to get higher salary otherwise there is no incentive to get an advanced degree. (3). Going bach to school is not going to make you more intelligent or more experienced. The advantage of going back to school is that you will have the opportunity to learn some principles which you can use after graduation. In other words, it will open one more door for you. Whether you will succeed or not, it all depends on your effort. (4). In the information age, if you have learned how to get the necessary information and be able to digest it, normally you don't have to go back to school. This is why Bill Gates was able to succeed in the PC software field. Formally, he did not graduate from Harvard University many years ago. (that is my impression, I could be wrong.) (5). If you are interested in making money, you must get into Business field. That is the field where one is suppose to make a lot of money. (6). Beyond this point, it is difficult for me to give you any advice. My suggestion is : If you are interested in getting an advanced degree, then do it as soon as possible and finish it in the shortest possible time. A degree is a degree, it has nothing to do with your ability to solve problems, to make more money or become smarter. But if you make good use of the principles learned at school after graduation, it will help you to achieve more. ( but it is not free. I have a PhD degree, and I have been working hard day and night in order to survive in the rapid changing world. So, I actually put in twice amount of effort into the CFD related work. So, a slightly more in salary is fairly normal.) There is no direct relationship between a degree and his performance. The same is true between his performance and his salary. ( In social science, you can delay the time to go back to school. In math, physics, engineering, it is very important to get your degree as early as possible. )

A very good question.

What do you want in life? Education is about more than just making more money, it is about broadening the mind and the experiences which bring that about. Grad school can be a blast and a FUNDAMENTAL change in the way you think about problem solving or it can be the biggest waste of time in the world! I have had both experiences one was with a weak and narrow minded conformist moron the other (which I had to pay for myself) got me to work and study in Germany, I learned German made life-long friendships, had a blast as a teaching assistant, on and on!

The primary factors are supervisor, topic, and University.....and arguably only the supervisor is key becasue the good ones will have the other two figured out!

I don't agree AT ALL with John's comments the you should just finish as fast as possible and it does not make any difference what you do! True, these 8yr PhD's are a discrace! But come on now John what you are saying is paramount to saying: "Why not just lie about it on your resume and hope that nobody catches on"! Most really good profs will get you through in the 3-4 year range (that means really done, defended, corrected, etc not oh, now you can start writing up at 4 yrs...which will wind up to be 5) and 3-4 is probably how long it took them! But do look into any unusual time lengths and why...slave labour is common in grad schools!

good luck...............................Duane

(1). It was purely based on the financial consideration. (2). If you stay in the graduate school for too long, even with the salary increase after graduation, it is going to take a long time to pay off the loan.

First of all, you've got to appreciate John Chien's point of view. If you read many of his postings, you'll see a couple of repeated ideas 1) All comercial codes are worthless except in very specific applications 2) Anyone who uses computational codes without first having trudged through the creation of their own FORTRAN (or some other equally arcane language) programs can offer nothing to "real-world" industry, and 3) Current commercial codes offer nothing to the design cycle.

Well, everyone's welcome to their opinion.

I can tell you that I routinely use a very user-friendly, general commercial code, CFDesign, to save my company money and time. I design balanced airflow systems for electric vehicle battery packs. The airflow paths for these systems need to deliver equal cooling to all the batteries while minimizing pressure drops. I would guess that the use of CFD has saved 3 prototype iterations over the old "trial and error" system. That means less money on the experimental build floor. More importantly, that means a shorter design cycle. Believe me, any ignorant manager can be sold on CFD by those last two statements alone.

Here's another way to think of it: if commercial CFD isn't a good design tool, why are there so many commerical codes advertised in the header of this forum? Who's supporting this market?

By the way, I am a BSME. I've started my Masters on a part-time basis. I see no reason to rush through it. As for an advanced degree: I recomend you work for 1-2 years if you are unsure of a field of study. How pointless to get a degree in something you may end up not liking. Maybe you'll find something in Industry that you really love! If not, at least your paying off some of your undergrad loans along the way.

Sorry about the long-winded report. I'll stop now and cover my head in anticipation of John's reply.

Jeff W

(1). I am reading your comments, and I am reviewing the OpenGL programming guide, and I am watching this second part of the movie about passengers missing in an airplane. See, I am very busy. (2). So far, on this forum, there are only two persons like you who have claimed that the commercial codes are useful tools for their work. The other person was related to the thermal design of power supply with a fan and a couple of heat sinks. I am very happy to hear that because of your confidence. So, you have to continue on your good work. (3). I can only mention an example of real world example, that is about the story of day-trader in the stock market. For someone who is very cool, he can still make a lot of money through day-trading. That was the conculsion of the study. So, you and the other person have the same personality, and you don't rush into it. And that is main key to the success. So, I don't want you to change your mind about the usefulness of CFD. In this respect, you seem to have qualified under the principle of "love in CFD". The question I have is " as a user of this code, can you keep it alive?" "Can you keep the company alive?" " How many copies of code do they have to sell in order to survive?" If you continue on this process, pretty soon you will have the answer. "Do we really need the PC/Pentium III for Internet?, for word processing?, for e-mail?" Based on your logic, Billions of dollars can be saved if we stop using the computers. You can easily save over a thousand dollars by not buying the computer and the software. If you can convince one million people to stop using the PC, you can easily save everybody one Billion dollars. (4). The keybord I am using right now, the monitor I am watching right now, the laser printer I have been using are all purchased in 92 when I was working as a consultant in a foreign country. I did upgrade the mother board and disk drive a couple of times. But still, I am not a good consumer. Me alone can not support the PC industries. (5). So, in order to survive, the PC hardware vendors and PC software vendors must create the new market every year using different methods. And today, you can get a complete PC with 300, 400 MHZ cpu in it for less than 300 and 500 dollars. (6). The CFD code vendors are just like PC software vendors, they all try to survive. They all will come up with very creative methods to sell their products to the consumer. There is nothing wrong with it. (7). I think last year the German goverment passed a law which make the x-rated stuff illegal on Internet. If people think that they are happy and willing to spend money to buy computer and get on Internet, there is nothing wrong with it, right? (8). And even if you said you are getting good results from the commercial CFD code, I am not sure that the solutions you obtained are really accurate solutions. Can you say that your solutions are all accurate within a couple of percent when compared with the test data? Are you willing to provide such benchmark test results? (9). When I say useful CFD codes, I mean a code which can produce reliable and accurate solutions. It also must be repeatable. If a calculator is not repeatable, do you still want to use it? I think, it is likely that you would return it to the shop. (10). But then, how many people have return their slide rules, 30 years ago? I think, the fundamental problem or difference is: the same word or definition has different interpretation , it depends on the time, the location, and the person involved. What I have just said, or said before, are just virtual vehicles for someone to express his opinion.

John,

Are you being serious? You give a reply to almost every posting on this forum. Sometimes multiple replies in the same thread! Don't you think it's even a little untruthful to say that there are only 2 people on this forum to ever have claimed success with commercial CFD code? What's that crap about being a "Day Trader"? Maybe I'm lack the intelligence, but I don't follow your logic.

Jeff W

(1). The issue is not logic, the point of discussion is not truth. (2). The goal is to keep the cfd-online forum rolling. (3). In ERCOFTAC Seminar and Workshop on 3D Turbomachinery Flow Prediction V, COST Action F1 Workshop on 3D Navier-Stokes Codes, 6th-10th January 1997 Courchevel,France, in the Open Forum section, it printed:"This opened with a review of the Lectures and the Test Case Sessions by Hirsch. He added to his 'one-lines' from previous years! 1994:Nothing Works, 1996:There is Hope, 1997:Is there Hope? This was with reference to the discrepancies with data, and the difference between different codes with the same physics, e.g. turbulence model. Suggested reasons were: Numerical aspects(especially insufficient grid resolution); Physical modelling(turbulence/transition); Uncertainties in test cases with respect to specification of boundary conditions. Overall we must make progress, not go over old ground again! " (4). I think, in 1999: there is at least one satisfied commercial CFD code user. (5). The story behind the "day-trader" in stock market is: The suggestion from me, the suggestion from you, is just like the suggestion from a day-trader, is not applicable to anyone else. But the forum will provide some reference points, so that a reader can evaluate the suggestions and his background to make a decision ( or no decision at all). So, the forum is to provide " what's available ", not "what should be done". (6). There is an old saying that " a stone from other mountain, can be used to sharpen a file " ( A seemingly useless stone from a far away land, can become a very useful tool to correct the original design and improve the quality of the product.) My words here is just like the CFD code you use.

Dear Jeff,

Commercial codes have limited applications. But technology is advancing so rapidly and competitions among companies are accelerating so rapidly that no individual company can do monopoly business anymore. In order to stay in cutting edge shape, commercial codes are not enough. It has already been proven that if a company just only rely on commercial softwares, it is obvious that the individual company will be out of business slowly. This is a doggy dog world. One company is putting another company almost out of business to stay on the top. This is a war of technology and profit. The hiring managers are at a loss what to do. They have stress and pressure from the top management. The hiring managers want development as soon as possible to get the product in the market. So they mainly depend on commercial softwares. We need to unlock our door of thoughts. Look what is happening! Say for example, Diesel engine companies got penalty of one billion dollar for producing engines with high emissions problems. Now think how many employees the companies laid off to compensate this loss. Again think of the consequences of profit motivated General Motors. The results brought the company with a 4.9 billion dollar lawsuit. If GM has to pay this amount of money then three to five thousand high level employees have to get early retirements. Prognosis is not satisfactory even if GM survives this time. Everybody is trying to play it safe. In order to bridge the gap between profit and technology, both commercial softwares and advanced code development should proceed simultaneously to keep this earth environmently safe place for us and the next generations to come. Nobody can say what will be worthy and what will be not worthy in the long run. Thank you.

Heh, you forgot Hirsch's 1998 comment ... "there is progress". I don't know what he said 1999 though.

Note that the test cases used in this workshop are selected because they still pose a big challange to current CFD technologies. There are thousands of other cases where CFD of today works very well and where commercial codes produce very usefull results.

There will certainly be a demand for qualified CFD engineers in the future. I don't agree at all with the pessimistic comments from others here. The CFD market is growing steadily by approx 20% per year (anyone have any current figures on this? Otherwise, see the thread about this from last year) and as computers become cheaper CFD will find new application areas. I see no sign at all that we have reached the roof.

The strong cut in US defence and governmentally funded R&D that has occured over the last 10 years or so has lead to a temporary over-supply of CFD engineers in the US. It is not surpricing that this generates a lot of frustration and pessimistic forecasts from those who got hurt. In the last few years we have seen a new trend though... many of these NASA and Gov.Lab. guys that got laid off have started their own small companies and there are now many many small CFD companies that all live on commercial grounds. Both small and large companies have started hiring CFD people again.

Whether or not you should go into CFD of course depends on if you are truly interested in it.

Another thing that you might want to think about is if you are interested in numerics, coding and theory or if you are more interested in the application side of CFD. Don't let "the opportunity" make this choice for you.

(1). The point I was trying to make is: after nearly thirty years of CFD, there was this brave person willing to say that the CFD is not there yet! (2). The commercial CFD codes are about ten years old, they are still teenagers. For them, I think, it is appropriate to say that they have made significant contributions, especially in the areas of user-friendly graphic user interface (GUI), automatic un-structured mesh generation for complex geometry, interactive-adaptive mesh refinement, CAD interfaces ...etc... (3). The concern is that alone is not going to guarantee their survival in the highly competitive world. (4). The real solution is for the forum to create the trend, such that the industries and consumers are willing to invest in the development and applications of CFD.

Turn! Turn!! Turn!!!

Time has come to open our eyes. Time has come to unlock our door of thoughts and motivations for advancements. We don't want to be carried away anymore like a leaf in the air or by fear of the turbulent phenomena generated by mother nature called 'tornadoes'. We cannot expect to confine ourselves in limitations. God has given us the power to put our heads together to reveal his mysteries. Time is coming for us to even challange mother nature. Technology is summoning us to " Follow knowledge like the sinking stars,

Beyond the utmost bound of human thoughts."

(1)..... <CFD War> ..... (2). A solider in a battle field, (3). two soliders went in along, (4). three guns fired, and four canons groan, (5). all CFD soliders never come back home...... (6). CFD is a serious business. JC#

Mohammed,

I think I agree with most of what you said. Keep one thing in mind, though: All businesses are "profit motivated". GM is not some power hungry evil empire. It provides good vehicles at good prices to people all over the world. True, GM is facing a large penalty for a supposed design flaw. Never mind the fact that the guy who caused the accident was drunk. Even so, products are designed by Humans. Humans occasionally make mistakes. You will never be able to design for every possible situation. Most design regulations were brought about by flaws that surfaced. They were corrected and recorded to prevent future flaws of the same kind.

1) People need/want cars 2) GM makes cars 3) People pay GM for the cars 4) The profits flow to GM and all its suppliers 5) A lot of people get to eat 3 meals a day

Another thing: competition is good-- without free market competition, technology will grow slowly if at all.

Sorry, I don't subscribe to the communist/socialist idealogy of State (or some other power) controlled trade.

Thanks, Jeff W

(1). why do you think that people are buying Japanese cars? Technology, or economy? (2). Why do you think that companies are producing household products in communist country? To convert them to capitalists or to make money? (3). Do you know why based on the recent report, two US carriers were sent to Taiwan strait to prevent Chinese invasion of Taiwan in 1996? To protect US businessman in China or US security? (4). The answer is: US car manufactures were sleeping when Japanese were investing in car engine technology. The consumers know the difference. (5). The answer is: the companies can reduce the labor force here while making more money in communist country. (6). The answer is: Even if you don't like communist system, there is no way out. Face-to-face, you will have to know how to defend yourself. (7). Those are real-world competition. And we hope that, through this cfd-online forum, you will be able to get a clear picture and learn the real CFD skill. (if you can get your hands on a commercial code, your enemy also can.)