# how fluent works

 Register Blogs Members List Search Today's Posts Mark Forums Read

 September 30, 2006, 01:30 how fluent works #1 asim ibrahim Guest   Posts: n/a hello to every body can any body encompass the answer in five or six lines that " how fluent works" what will be ur answer if some lay man ask u this question.

 September 30, 2006, 04:37 Re: how fluent works #2 prapanj Guest   Posts: n/a Hi friend, I am a layman.Let me try to answer. U have ur calculator right? U ask what is sin43 the calculator gives u a value correct to ten decimals. now does it draw a triangle to find analytically the exact ratio? NO It has tailors series incorporated in it. It will find the value of a few terms of the series. and finally display an approximate value. Ur job is to feed in the input. Similarly u input an airplane's wing(an airfoil) and asks fluent to find the lift distribution.. There are so many numerical methods available. they are approximate, requires hell a lot of steps, but are easy to calculate in each step, and are far assertive that going for an analytical solution. Because with numerics its easy to incorporate multitudes of factors adding to the reality of a situation. got it? CFDians have programmed fluent (in c language i think) based on these numerical methods. And it finds the solution and plots it...(Very analogous to a program to find the sum of two numbers.. but here we do it for the gloriosus purpose of duplicating nature) now fluent is the calculator . hope it reached u. see only a layman can understand another layman's feelings. i am an undergraduate aero student . bye rajann_786 likes this.

 September 30, 2006, 05:07 Re: how fluent works #3 jimmy Guest   Posts: n/a To be a bit penantic: I guess calculators would rather use cordic's algorithm than implementing sin by tailor series . Fluent is (if I remember correctly) programmed in C and Fortran and uses finite volume methods. Nice explanation by the way!

 September 30, 2006, 05:26 Re: how fluent works #4 prapanj Guest   Posts: n/a thanks for the edification dude..

 September 30, 2006, 18:37 Re: how fluent works #5 Mani Guest   Posts: n/a to be honest I find the explanation a bit confusing and too generic (maybe because I am not exactly a lay man?). you could have explained anything from the analysis of electrical oscillators to weather prediction in exactly the same way (in other forums). does it satisfy you, asim, or is it too general? you must have some basic interest and/or understanding of aerodynamics, since you know about fluent. I suggest you tell us your own idea on what fluent is, what it does, and how it does it, and we can help you refine it.

 October 1, 2006, 01:37 Re: how fluent works #6 asim ibrahim Guest   Posts: n/a thanks for u all dudes for responding. mani u asked me that r u satisfied with this explanation...? friend what i all know is that if u want to model a system or physical situation than it will be defined by some differential equaiton but as it is not possible to have analytical solution of every problem so we go for the numerical techniques. thats what the fluent do it employ the numerical techniques and uses discritization equations and different schmes to solve thoese equations. and may be all that code is written in C++ thats what i already know. i put this question in the context of another question posted on this form i.e " what is the tendecy of CFD". in which people said that "A person who says that i am expert in FLUENT will be a liar" . and they compared the FLUENT (which is a CFD tool)with autombile which is a tool of tranpotation. to understand how FLUENT works compare these tools. to drive an automible 100 years back needs much experties of different fields but now a days even your mother can drive the car without knowing any details about engine and just manuplates with accelator and gear. it means that in past tool (i.e automobile) was not mature and tool user have to be an expert but now a days tool (i.e automobile) is matured and tool users are not experts (i.e your mother). similarly FLUENT is just like 100 years old automobile whoes user need to be an expert. ( thats all was discussed in the question "tendency of FLUENT" posted previously) keep all these discussion in ur mind. i group of people say that FLUENT is very wide tool and only experts can hadle it and other people just simply explains it that it employ the numerical techniques and give the solution. people of different fields and different areas of interest want to use the FLUENT. e.g a person related to aerodynamic or a person how want to do the combustion modeling. both people will be of different back ground and they want to be an expert in using FLUENT then from where they should start. who they can be expert. what is the suggestion for me what i should do because i think i am a lay man. rajann_786 likes this.

 October 2, 2006, 08:12 Re: how fluent works #7 Mani Guest   Posts: n/a I now see where you're going with this. The "car driving" analogy is good, but I prefer something more detailed. Let's look at wood carving as an analogy to CFD. A good wood carver can take a sharp knife and a block of wood and create a great work of art. What does it take to be a successful wood carver? He needs to know a lot about wood. That's the primary and most challenging requirement. Furthermore, he needs to know how to handle a knife, and needs to be familiar with wood carving techniques, i.e. how to work the wood with the knife. I know how to handle a knife, but that alone doesn't make me a wood carving artist. Fluent is nothing but a tool (just like the knife). Depending on how you use it, your results can be anywhere from good to completely useless, even damaging. In order to make good use of this tool, first and foremost you need to have a deep understanding of the physics you are trying to model (aerodynamics, combustion, heat transfer, whatever it is you are trying to do). That's like knowing the wood in wood carving: A fundamental unavoidable paramount requirement, an absolute neccessity for doing anything meaningful with any numerical tool such as Fluent. (I am putting do much emphasis on it, because, believe it or not, some people still don't realize that.) Knowledge of the physics is necessary and most important, but it's not enough: Furthermore, you need to know quite a bit about the numerical techniques that your tool applies. In particular, you should have a well grounded knowledge of the tool's limitations and underlying assumptions. To come back to wood carving, this is analogous to knowing how to work the wood with the tool, e.g. cutting with the grain and not against it. How often have we been asked the ultimate question on numerical methods "The computation does not converge! Why is that?" You have to be able to trouble-shoot your numerical tool (preferably without going long ways to get help from third-party support), and that's one of the reasons for understanding numerics. The other reason, of course, is the necessary ability to judge your results. For this you need to have a sound knowledge of physics and know the potential caveats of the numerical method. I put knowledge of numerics as a second priority behind knowledge of physics, because you are more likely to obtain useful support on application and trouble-shooting of the tool than on the underlying physics. Never forget that you are the artist using the tool, not the tool maker. However, there is one more thing: If you are really good at using your tools you'll get more out of them (more efficiency for example). For that you need to know quite a lot about how the tool works, i.e. you need to know about numerics and high-performance computing. Any good wood carver is also an expert in knife sharpening! There is really not that much to it, but it amazes me how often and consistently either one of the mentioned aspects of CFD are completely neglected. There is the numerics "expert" who doesn't understand that he cannot apply just any arbitrary boundary condition (because they happen to be known from experiments for example) and expect that the CFD problem is well posed. There is the fluid dynamics "expert" who wastes a lot of his time trying to get meaningful results out of CFD and doesn't understand why it's not working. And then there is the self-described Fluent "expert" who goes home early today, because he has finished his CFD problem in record time! Never mind that his problem is ill-posed and the results physically impossible: Hurrah, it's converged !!!

 October 2, 2006, 13:09 Re: how fluent works #8 HelpfulSoul Guest   Posts: n/a Magic ! or 'It doesn't'

 October 3, 2006, 00:08 that was really "edifying" #9 prapanj Guest   Posts: n/a thank u Mr. Mani, It would have been better if cfd-online.com introduces some free memberships scheme with a personal profile for each. that would have stopped me into poking nose between great brains. Ur explanation was great.And thanks a lot. As i am just a beginner, i will always keep that in mind. I dont plan to learn fluent well now. I am concentrating on physics and developing some codes myself. Hope that decision is right based on the advices u have given. Because then i will be safe from 'ill-posings'. thanks

 October 3, 2006, 13:10 Re: that was really "edifying" #10 asim ibrahim Guest   Posts: n/a mani u simply have a great mind. the anology u made of that wood carver was good and in the last paragraph u just take the words from my mouth. bye the way weather Mani is ur nick name or real name.

 October 4, 2006, 23:46 let mani be "mani" #11 prapanj Guest   Posts: n/a hey gusy let mani be mani itself. Mani, plz dont reveal ur identity now. its thrilling to get help from a mystic soul. keep it up.

 October 5, 2006, 04:55 Re: let mani be "mani" #12 diaw Guest   Posts: n/a Hi Prapanj, We really do try hard to not get into 'people issues' on cfd-online. There are a bunch of really smart guys out there who will help you when you ask a question. Perhaps it would be better to leave the names out of the posts? I love this site & have been coming here far a while. Can we try & be nice to each other? Thanks so much... diaw...

 October 5, 2006, 07:22 Re: let mani be "mani" #13 prapanj Guest   Posts: n/a Mr diaw, I hope u have misunderstood me. please reread my post. i never got any personal. i just wanted people to stay just the same and help each other. i was also askin the above person not to get personal just as u did. thank u

 October 6, 2006, 03:59 Re: i decided to let mani be "mani" #14 asim ibrahim Guest   Posts: n/a yes it is a interesting way to communicate and may be it is a blessing in disguise.

 Thread Tools Display Modes Linear Mode

 Posting Rules You may not post new threads You may not post replies You may not post attachments You may not edit your posts BB code is On Smilies are On [IMG] code is On HTML code is OffTrackbacks are On Pingbacks are On Refbacks are On Forum Rules

 Similar Threads Thread Thread Starter Forum Replies Last Post emmkell FLUENT 26 June 9, 2015 15:38 opehterinar81 FLUENT 0 August 19, 2011 11:55 ibex7 FLUENT 7 April 18, 2011 02:44 Freeman FLUENT 4 April 14, 2006 05:58 Jonas Larsson FLUENT 5 March 13, 2000 04:27

All times are GMT -4. The time now is 10:53.