 to the brief overview of 5-day IST workshop for Coordination. First of all, I would like to thank Professor Gayathonde for sending us an email to the department faculty and encouraging to take part in this workshop. I would also like to thank Professor Upen Bandarkar for the nice introduction of the course. So, our course is on Competition Fluid Dynamics under the National Mission of Education through Information and Communication Technology. We are planning the first part of the workshop, which is for coordinator in the last week of February or first week of March next year. Along with myself, Professor Balchandar Puranik is also the coordinator for this course. I would also like to thank Professor Patat, who is the project coordinator for giving opportunity to our course as a part of this project. I would also like to thank Dr. Mukta Atre for helping us in organizing this course. I will begin this overview with what is CFD, Computational Fluid Dynamics. It is a methodology for computer simulation of fluid mechanics and heat transfer problems. In simple language, I can say that CFD is a subject in which you learn to develop a tool, which is like a video camera. You take a video camera and capture a movie, you get a movie, which has a pictorial representation. In computational fluid dynamics, we create a movie, which has a fluid dynamic representation in terms of the flow properties like velocities, pressures and temperatures. You not only create movies in terms of fluid mechanics and heat transfer, but you also calculate engineering parameters such as if you take an example of flow across a car, the drag force is one of the important parameter. If you take a flow across an aeroplane, the lift force is a parameter and if you take a flow in a pipe, the pressure drop is an important engineering parameter. So, the CFD results in the flow field, in the domain of interest and engineering parameter, which are useful in the design and optimization of the processes and equipments. It is an open-ended application of undergraduate course on fluid mechanics and heat transfer. CFD reduces the time and cost of designing and analyzing the engineering system and it is becoming a part of a bigger subject, which we call as a computer aided engineering. In earlier days, the computational solid mechanics and computational fluid dynamics were two different subjects, but slowly there are a lot of problems, where we need knowledge of both solid as well as fluid mechanics. So, this is a subject, which is coming up, which is called as a computer aided engineering, which includes both finite elemental assets for solid mechanics and finite volume method for fluid dynamics. Now, let us go to the need for the course. In academics, if you see, this course is taught in various branches of engineering like aerospace, civil, chemical, mechanical and metallurgy. In industry, it is developing lot of interest in various sectors of the industry like aerospace, automobile, process industry, biomedical and electronic cooling and so on. It is an interest developing on CFD software development, application and analysis in Indian industry and research organization and there is a lack of trained manpower in the subject. So, this is the important of this course. However, there is a lack of trained teachers for this course. So, this is the biggest motivation for us to take up this course. I will start with the brief outline of the course, although we will discuss the detailed course contained later on that although we mention later on that there are certain prerequisite for this course, but in this course, what are very essential for this course will be teaching. So, the very essential part of the fluid dynamics and heat transfer and numerical methods will start with our course with this topics and then we will move on to two discretization methods, which have been used commonly in computational fluid dynamics. The older one, which is a finite difference method with which the subject started, will discuss this method for the model equations. The finite volume method, which is a more commonly used method nowadays will be discussed for the governing fluid mechanics and heat transfer equations. There are mainly two objectives of this course, the development and application. So, the first objective is to develop an appreciation of theory behind the computer screen, so that you learn how a CFD software is developed. The second part is to appreciate the application of CFD through computer simulation of carefully designed exercise problem. To give you an example, here we give emphasis on development because we believe it is interesting and challenging to learn how to develop a car, then how to use run a car. So, with this two pronged approach of theory and exercise problem, we believe that the participant will be firmly set on the path of becoming a CFD expert. As far as prerequisite for this course is concerned, we expect an undergraduate course on fluid mechanics and heat transfer and some introduction of numerical methods and computer programming. However, I would like to emphasize that some essential part of this prerequisites will be taken in the beginning of the course. Now, I will leave the mic to Professor Balchendra Puranik who will be taking the first part of the course content, which is shown in this slide. Hello everyone, I am Balchendra Puranik and I will be teaching the first part of this course, which is presented right now on the screen. So, what we will begin with is a discussion of essentials of fluid mechanics and heat transfer, where initially what we will go through is a complete derivation of the governing partial differential equations, which involve the derivation of the mass governing equation, continuity equation, momentum equations as well as the energy equation. Along with these partial differential equations, we will discuss all the requisite boundary conditions and then what sort of governing and engineering parameters that result from these equations. So, this is what we call essentials of fluid mechanics and heat transfer, which is absolutely necessary to understand before you get on to the numerical simulation part. Later on, once hopefully we understand the basics in terms of these equations, we will move to discussions of the essentials of numerical method as we use them in this computational fluid dynamics. So, this will typically involve discussion on how to do numerical differentiation and integration, as well as some straightforward iterative solution techniques for linear algebraic equations. The reason is because in case of CFD applications, what we end up doing is we generate a system of linear algebraic equations, which needs to be solved and the typically most commonly used techniques for solving these linear algebraic equations as far as at least the level of this course is concerned would be what we call a Gauss-Seidel algorithm and a Thomas algorithm or a tridiagonal matrix algorithm. So, these two techniques will be discussed and understood completely before we move on to the formal CFD discussion. The first part in the formal CFD discussion would be the finite difference method for model equations. So, what we mean by model equations is that we will not consider the full governing equations for fluid dynamics and heat transfer, but we will come up with some sort of a subset of those equations, which will be called as model equation. And these are typically what we call either a wave equation or a diffusion equation or a heat equation or a Laplace equation, which fall under these elliptic parabolic hyperbolic type equation. And we will use this finite difference method to solve some of these equations subjected to certain boundary conditions and initial conditions. Along with this, there is a small topic of stability analysis, which we will also cover. So, this part which is presently projected on the screen is roughly 40 percent of the course and we call it some sort of a prerequisite that you need to know before you go on to the more modern techniques of finite volume method, which now Professor Sharma will again talk about. Thank you. I would like to emphasize that this course concerned, which we had tentatively proposed, which is tentatively finalized is has come up by looking into the syllabus of the course of CFD, not only in the IITs or ISEs, but to the various colleges in the country. Although the CFD course is presently not being taught in most of the colleges in the country, however, if you compare other computational courses like computer aided design or finite element methods, they are more common or popular, because we have teachers to teach these courses and the biggest of motivation and objective of this course is to train teachers to take up this course as an undergraduate course. So, the second part of the course contained, which I will be taking is on finite volume method, which is being used in most of the CFD software, today's CFD software and here I will be taking up for two dimensional unsteady state equations in for fluid flow and heat transfer. And I will basically I divided the topic into three parts, starting with pure diffusion to a convection diffusion to a fluid dynamics equation. So, the first topic is on computational heat conduction and the second topic is on computational convection, which is basically an advection diffusion equation. Although I call it as a heat convection, but in this in this topic the way we will be solving heat convection is that we will assume that we have a prescribed flow, so that we do not have to solve the flow equations and we just need to solve the advection diffusion equation. So, the combination of diffusion and advection, which we will learn from the first and the second topic respectively will be used to teach the computational fluid dynamics part, which is the third topic. In all these three topics, we first start with the finite volume method of discretization and then go on to the approximations, which are involved in the finite volume method. We will discuss the explicit and implicit method. One of the most difficult part of the course is that once you learn how to develop algebraic equation, the most challenging part is how to write up a program. So, you need to know certain implementation details, so that you can easily write up a program. So, for each of this topic, the implementation details to convert the formulation into a proper working program will be discussed in detail. Solution algorithm will also be discussed and there are some special topics, which will be taken up in conduction like multi solid heat conduction and an all linear heat conduction. Each of this topic will be supported by some example problems. In fluid dynamics and heat transfer, we will discuss both staggered as well as collocated grid to handle the pressure velocity decoupling. I would like to mention that the finite volume method, which will be taught here is slightly different in what you see in most of the present CFD books. If you in your under graduate fluid mechanics and heat transfer, what we do basically is that we start from a control volume and learn how to obtain the governing partial differential equation. This is what is taught in under graduate course, but what you see in most of the CFD books is that they start from the governing partial differential equation and they end up with an algebraic equation. By teaching this course for quite a long time, I had realized that when I start from a governing partial differential equation, when I start from mathematics and then go to the algebraic equation, I had felt that the student do not get the feel of the approximations, which are involved or the methodology finite volume method. So, what I do, which I call as a present novel finite volume method is that I start from the control volume. So, instead of starting from the mathematical equation, I start from a picture, which is a control volume, which they see in their under graduate course in the beginning of the under graduate course lecture. And from there, I use certain level two levels of approximation and end up to the same set of algebraic equation using the similar level of approximation, so that if you solve this set of algebraic equation, you get the same results. So, this makes the understanding of finite volume method much easier. These are the reference book for course Tanneil, Anderson and Fletcher, which will mainly be used for finite difference method. There is a book by edited by Professor Murli Dharan Sundarajan. This will be used for finite difference as well as finite volume method. There is a good book on computational fluid dynamics by one of the professor of our department, Professor Darte, which will also be used. There are some very good example problems in this book. This is one of the classic book in computational fluid dynamics by Patankar. So, these are the reference books. The methodology of instruction is that this course involves lectures as well as laboratory. We are typically planning right now about 4 hour of lectures and 2 hours of lab sessions. So, the mode of lecture will be using board as well as power point slides and the complete course material would be put up on the web immediately after the course. In the case of laboratory, we are developing programs which will be provided to the participants and they are expected to solve the CFD problems for various governing parameters. Like we will tell you to vary the Reynolds number, you may tell you to vary the geometry of the problem and to run the program which will be providing it. Right now, we are thinking that we will develop program using open source software like Sylab. It is easy to learn and because the long term objective of this course is that you should be able to develop your own program. So, in this course we will be giving you programs, but in long term we expect that you learn how the program is written. In the laboratory, we expect you to run certain example problems which will be given to you, plot and analyze the results. After the course, it is expected that the participant should be able to derive the conservation equation, develop program for numerical differentiation and integration, should be able to convert a differential equation to set up a library equation by finite difference and finite volume method, develop program for simulation of fluid dynamics and it runs the problem. So, with this we had come to this end of this overview session. We would like to thank all of you for your attention and welcome for any questions, comments or suggestions. Yeah, please ask a question. N83G, over to you. A method, up to that I learned and previously I was using MATLAB and MATLAB is taking much time. Please suggest me any programming language which is available at free of cost and I am using Windows 7 OS and for that is it good OS to execute the programs and what should be the RAM size. I am using 4GB RAM. Can you suggest me in that way? Now I am in learning stage. Thank you. Over to you. Yeah, first I will repeat the question. There are two parts of the question. First part is that he has developed a program in MATLAB and he says that it takes lot of computational time. Yes, it is correct. If you the CFD programs, if you take the real world problems and if you use MATLAB as a programming language, it takes lot of computational time because it does compiling as well as execution simultaneously. So, it is better is that you use develop your programs using traditional programming languages like C++ or Fortran. And the second part of the question is that operating system, although Linux is used as a common platform, but Windows also works well. So, I am not sure about it. I hope I have answered your question. If you have any other question, I will be happy to answer. Sir, please repeat the last part of the explanation OS regarding OS. Your question is that as far as the running program in MATLAB in Windows or Linux, which one is faster, I am not sure about it. I do not have idea about it. So, I cannot answer this question. Sorry. Over to you. Sir, I am having Linux Mint also and in Linux Mint, which programming language is readily available, freely available. Over to you. Fortran C++, this programming language, there are three ways of this in Linux. So, you can get it easily. Thank you. Sir, please ask your question. Over to you. Sir, this is Lugma from PhD College of Technology, Coimbatore. And we have this computational fluid dynamics course in UG as well as in PG. And we follow the Danderson book, Computational Fluid Dynamics. Can you suggest some book for the beginners, especially in this computational fluid dynamics, so that they can understand and they can go for the next level of book. Over to you. Sir. Yeah, I agree with you that you are looking forward to books, which is slightly an advanced level of Anderson, but the explanation is the way Anderson book has done. But unfortunately, the one which I can think of close to it is, if you can go through book of Computational Fluid Dynamics by Prof. A.W. Date, which will be a good book for a beginner. It is an introduction to Computational Fluid Dynamics by Prof. A.W. Date had given in my reference book list. Over to you. A.W. Date, sir, introduction to Computational Fluid Dynamics. Is it so? Yes, yes. Any other question, please? Okay, sir. Thank you, sir. Over to you, sir. Yeah, please go ahead and ask your question. K.J. Somayya. Please guide regarding computer programming requirement related to CFD. I mean, what should be the prerequisite regarding computer programming related to computational fluid dynamics? Yeah. Yes, it is not that you need to know lot of programming before you can attend and understand this course. What we expect is that if you know the basic syntax of one language such as say Fortran or C++, that should be more than enough. What we are planning to do is we will explain you the complete algorithm of how to code a certain or how to code for a certain problem. And once you know the syntax of a given language such as say Fortran or C++, you should be able to go ahead and write the program. So, as far as my opinion is concerned, you really do not require very high level of proficiency in terms of computer programming. As long as you have basic working knowledge of the syntax of a language and if you have gone through a basic numerical methods course earlier, that in my opinion should be sufficient. Over. Thank you. Over to you. Government college Salem, if you can hear me, please go ahead and ask your question. Sir, good afternoon. Sir, this is a question regarding there is a lab session in your CFD course. Am I correct? Yes, there is. Sir, in this lab session we do not have option for the software using in our college. Sir, whether you will provide the software for the CFD course for our college during the two week programs? Yeah, I understand your question. In this particular course, we are not going to use any of the software such as say Fluent or CFX that you may be referring to. What we are planning to do is we are going to generate our own programs in an open source software called either Sylab or something like that. And we will be providing you all these programs beforehand. And since it is an open source software, this software can be loaded in the computers in the remote centers. And the participants can run our programs during the lab session directly on this open source software. So, there is no need to go for any commercial software such as Fluent or CFX. We are not really interested right now in this first level course to teach you how to use the CFD software that is generated commercially. What we are interested is the basic theory behind all this and some example problems which will help you understand how to use the commercial software later. So, I do not think you need to bother about this part much at all. Over to you. Sir, one more question. Participants were interested to see your reference books which one of your slide is showing that I think probably in the last slide. Yeah, we will put that up. Our participants want to see the reference books please. Yes. Can you show me that? Yes, just a second please. Yeah, the slide with the reference books is now projected. Over to you.