Unlock all 39 lessons of Computational Combustion Using Python and CanteraFull course ₹40,000 · paid
What Computational Combustion Is and How It Links to CFD
A free lesson from Skill-Lync's Computational Combustion Using Python and Cantera course, on Grow by Skill-Lync · Last updated 16 Nov 2026 · Session: “Week 1 - Introduction to computational combustion”
What will you learn in this free Computational Combustion Using Python and Cantera lesson?
The instructor outlines the 12-week course, its Python, NumPy, matplotlib and Cantera tools, and how assignments work. The lecture then defines computational combustion, shows a Stanford hypersonic nozzle simulation, and explains how the species transport source term connects the flow solver to chemistry, including the GRI methane mechanism and validation against experiments.
- Definition of computational combustion
- Navier-Stokes plus species transport equations
- Source term linking CFD and chemistry
- Detailed kinetics models and experimental validation
What comes after lesson 1?
The next free lesson is lesson 2, “Installing Python and Sublime Text on Windows” (4 min). Later in Skill-Lync's Computational Combustion Using Python and Cantera course (paid): lesson 4, Week 1 - How to install cantera; lesson 5, Week 2 - Basic concepts; lesson 6, Week 2 - Getting started with Python. The full course has 39 lessons (9 h 21 min of video); see the course overview or its price and syllabus.
Lesson transcript
Machine transcript of the lesson audio (lightly cleaned). Technical terms may contain errors; the video is the reference.
Hey guys, I hope you're all doing good. Welcome to the first lecture on Computational Combustion using Python and Cantera. My name is Sarang and I'm going to be the primary instructor for this course. All right, so in today's class, I'll be first talking about what you will be learning in the coming 12 weeks. All right, I will be talking about the software programs that we will be using for this course and also tell a little bit about the type of problems that you will be solving. Now, once we have done that, I will be talking about progress tracking.
Once you have watched a few lectures, you will be given assignments. These assignments help us gauge your performance and understanding in this particular course. These assignments will be hosted in our leaderboard interface and you will be routinely checking it out and you will be solving the problems that are posted over there. So, here's a list of topics that I will be teaching you throughout this course. I have color coded them in terms of difficulties. Topics that are colored in green are a bit easy to understand. Topics that are colored in yellow, you know, they require a little bit of work on your side.
You will be doing some assignments to make sure that you understand these topics correctly. And topics on the right hand side are advanced topics or basically, you know, that's the core of computational combustion. So, that will take a little bit of time for you to understand. And towards the end, you will be working on two projects. As far as the topics are concerned, in the first week, you know, we are talking about introductory level thermodynamic concepts like low rating value, the NASA thermodynamic database, some simple combustion terminologies.
And, you know, pretty much the introduction to Python part is in the first part of this course. Again, no programming expertise is required. We will be teaching you from the basics. So, don't worry if you have not done any programming before. In the second level, you know, that is what I would like to call as an intermediary level. We talk about intermediary thermodynamic concepts and also the math part gets a little bit tough and also and so does the computational part. Since we will be dealing with nonlinear systems with constraints, we will be working on solving those type of problems.
We will dwell deep into combustion models and this is where we will be learning the first combustion model which is called as the equilibrium combustion model. And then the advanced part of this course starts where we will first develop some compact mathematical schemes which will help us consistently represent a chemical mechanism. This is an essential part in converting chemical mechanisms into computer programs. OK. I know that at this point it doesn't make sense, but this compact notation is what helps us tell the computer, hey, this is so we are going to represent the entire set of equations.
When I say equations, you know, it's chemical equations. All right. So that part is super important and you will learn that this is actually a system of ordinary differential equations. So we will talk about how to solve ODE systems. We will look at what a stiff ODE is and also discuss methods that can be applied to solve such a system. Then we take up advanced combustion concepts like auto ignition, negative temperature coefficient. During this point, you know, we would be using Cantera, which is the chemical kinetics toolkit that we will be using.
All right. And we will be performing flame speed and ignition delay sensitivity analysis. And your project will be related to that topic. And once we talk about plug flow reactor and some advanced combustion concepts, really control this course. So to follow along in this course, you don't have to read any textbooks. OK, it's not required. But if you feel like you need to follow along in a textbook so that, you know, you can increase your knowledge in combustion, then a very good book to start with is introduction to combustion.
OK, by Stephen turns. Now that we have looked at in detail what you will be learning, I just wanted to summarize it in case the first slide was a bit more complex for you. At the end of the day, you are learning the fundamentals of fuel and combustion. You will be brushing upon intermediary level thermodynamics. You will definitely be learning computer programming using Python, which is a very good skill set to have. And we will be familiarizing ourselves with intermediary and advanced topics in computational combustion by performing chemical kinetics calculation using Python and Cantera.
Now, it is important to understand that having knowledge in these topics is very useful. A lot of companies in India and outside India that perform combustion simulations require engineers with exactly the same skill set. So you're not learning something that is outdated. You're learning something that is relevant. And depending on what you want to do after your graduation, you can use the projects that you do in this course to your advantage. Now, if you want to discuss carrier options after this course, I would be happy to engage in that discussion with you.
And as I mentioned before, we'll be using Python for this coursework. You will be using NumPy module to perform numerical analysis and the matplotlib module to create charts. Cantera is the chemical kinetics toolkit that I talked about. And as I mentioned previously, we provide challenges periodically to gauge your performance and understanding of this coursework. These challenges are hosted in our learning management system. You can take a look at it and anytime you have a question with the challenge or one of the course videos, you can get in touch with us through email, phone or our official forums.
So whenever you have a question regarding a challenge or a particular video, you can get in touch with us via phone, email or through our official forums. All right. So that being said, let us jump into the actual lecture. So today we are going to primarily understand the topic computational combustion on why it is very important. And I will explain to you the relationship between computational fluid dynamics and computational combustion. We will be doing this by looking at some mathematical equations. All right. So what is computational combustion?
So to get started, let me just play this simulation here. So what you're looking at here on the left hand side is hypersonic nozzle. OK. And on the right hand side, you're actually looking at the flow that's coming out of a hypersonic nozzle. So as you might know, hypersonic nozzles are extremely high Mach numbers. In this particular case, the Mach number is around 16, I believe. So you're looking at temperature. Right. So you're looking at the odd exhaust gas products and simulations like this help us understand.
A lot about the system, right? How air and fuel are mixed, how pollutants are formed. So here we are looking at a simulation that was done by the Center for Turbulence Research at Stanford University. It's a very well known group for computation fluid dynamics and also computational combustion. So this is what we call as computational combustion. Simulating the combustion process by solving equations in a computer is what we want to do. And there are so many users of it. In this case, it's a hypersonic nozzle.
Right. So testing this out is going to require a lot of costly equipment. On the other hand, if you're doing simulations, you can do something like this using powerful computers, which are available a lot. This particular simulation was run in a supercomputer. I believe it used around 100,000 processors. So which means it's a parallel simulation and the simulation got completed in four days. OK. So this is a very interesting simulation. If you're interested, you can go to CTS website to read more about these type of simulations.
All right. So to summarize, computational combustion refers to simulating combustion using a computer. That is a very simple way of defining it. So you need to remember that a successful simulation can predict the thermodynamic states and species concentration before and after the combustion process. Now, this might sound a bit complex, but actually it's not. If you think about it, let us say I'm simulating an internal combustion engine. Now, at the end of the combustion process, I'm interested in calculating the emissions. Right. If you think about it, emissions are nothing but products of a particular reaction.
So if you're simulating combustion accurately, you are simulating those chemical reactions accurately. And that is why we are able to calculate those emission products accurately. So in other words, information from a combustion simulation can be used to calculate engineering results. The next thing that I wanted to talk about is the relationship between computational fluid dynamics and computational combustion. So let me go back to this simulation here. So like I said, we are simulating the flow that occurs at the exit of a hypersonic nozzle. Right. So there is a flow component and there is a combustion component.
How are these two things related to each other? OK. And now if I go back, these are the equations that are being solved in that type of simulation. We first solve the Navier-Stokes equations. Now, if you are not familiar with these equations, the Navier-Stokes equations are what govern fluid flow. So when you take fundamental laws such as mass conservation, energy conservation, momentum conservation, and you convert it into mathematical equations in a differential form, you get the Navier-Stokes equations. So what we are looking at here is what you call as a non-dimensional form of the Navier-Stokes equation.
And here we are looking at continuity equation, the momentum equation and the energy equations. These equations can predict how a flow is going to take place, what's the temperature in the flow and so on. All right. On the other hand, if you are talking about a simulation where there is also chemical reactions taking place, then we will be. Poisoning the drainage front of the query. All being an additional equation. This equation is called as the species transport equations. Now the species transport equation is very similar to your other governing equations. So for example here you can see that you have a time derivative term and then we have spatial derivative terms.
So here XI basically refers to X axis when I is equal to 1 and when I is equal to 2 it refers to the Y axis and when I is equal to 3 it refers to the Z axis. It's just a compact way of writing the equations. And if I go to the previous slide you can see that you have spatial derivatives that is derivatives with respect to X, Y and Z. Very similar to what we saw in the species transport equation. The main difference is the term. All right. But you can see that essentially the same terms occur and this is what is called as a species transport equations.
The species transport equation basically is saying that hey this species is a conserved quantity. Now in this particular equation you see that I have a term called S. This term is what you call as a source term and this is what links computational fluid dynamics and computational combustion. All right. So to explain this a little bit more let me go to the whiteboard. All right. So imagine that we have a burner and then I'm going to say that at the inlet of this burner we have air and fuel coming in.
We are also going to say that the inflow temperature is 1200 Kelvin. Now if we have to simulate the flow through the burner we need to first create a computational mesh. All right. So once we have created a computational mesh. Our objective is to solve the governing equations in each and every cell. So for example if we take this particular cell and then let us write down the equations that we will be solving in that particular computational cell. All right. So we will be solving the continuum equation the momentum equation the energy equation the species transport equations plus additional transport equations which can be your turbulence equations.
But the idea is when you solve the continuity momentum and energy we are just trying to predict how the fluid flow is going to move around how it's going to get compressed and how it might get heated up. Right. This is all the information that these three equations can predict. Now remember that your fuel contains a particular species it can be a hydrocarbon right. So when your fuel and air react it's going to form products which can be carbon dioxide carbon monoxide and water vapor correct.
Now what's going to happen is the species transport equation has a source term. Now the source term is going to be responsible for consuming the fuel and it will be responsible for creating products. So in other words in this particular cell you might first start with a particular fuel composition. So for example let us assume that in this particular cell we have 50% fuel 40% air and 10% nitrogen. Once the species transport equations are solved the chemistry calculations are performed and the chemistry solver can then determine that okay when fuel and air are going to react they are going to form a set of products.
So after the combustion simulation we might end up with the following species concentration and this is what the combustion solver is doing. So you can see that the computational fluid dynamics solver and the computational combustion solver are speaking to each other through the source term in your species transport equation. And this is an important equation and this is an important equation that you need to understand alright. So moving on the other interesting thing that we need to tackle here is how mechanisms are fed into computer programs.
And to demonstrate this I have a snapshot of what we call as the GRI mechanism. So the GRI mechanism is the list of reactions that represent the combustion process of methane okay. So here you can see that I have a keyboard called reactions and I have a set of reactions. You can see that there's a bunch of numbers that are getting printed on the right hand side and also below the equations. For now just ignore them we will be learning what those terms mean individually. But this equations represent how combustion process is going to take place at least they instruct the computer how the combustion process is going to take place.
Now if you remember in the starting of the lecture I talked about something called as a compact notation. The compact notation helps us take this information and feed it into a computer and get the results out of it alright. So if you're curious about it don't worry we will be learning about those topics very soon. But the general idea is while feeding mechanisms to a computer we use this particular format. Okay this format is also called as Chemkin 2 format. It's a popular format that has been used for a long time now.
So many types of software programs like ANSYS Fluent, StarCCM and ConvertCFD can read this particular reaction table and perform simulations with it. So that being said I wanted to give you a 50,000 feet view of computational combustion. You know it all starts with what we call as ab initio calculations. There is something called as a reaction rate which determines how fast or slow a reaction is taking place. Now to calculate the reaction rate we have to measure couple of important parameters which cannot be solved for very easily.
To calculate those numbers we perform quantum calculations and figure out accurate reaction rates. So once we have accurate reaction rates the second step is to determine species thermodynamic properties. Why do we need that? Well it turns out when you use these species in governing equations you need to have access to their specific heats, the enthalpies and entropies. Okay. Now if you do not calculate those properties accurately your combustion simulation is going to be wrong. Both of these concepts are extremely hard. People have dedicated their entire lifetimes in calculating accurate reaction rates and develop processes that helps calculate thermodynamic properties accurately.
So once you have these two information you will have something called as a reaction path. What does that mean? Well reaction path represents how a particular reaction is going to lead to another reaction, right? What is the flow of reactions, right? So when we talk about computational combustion or when you have studied combustion in your undergrad, you would have simply studied that air and fuel react with each other and they form water vapor carbon dioxide and nitrogen. This is called as a global reaction. In reality chemical reactions do not take place like that.
You have intermediary reactions that takes place to produce intermediary species. You can have ten thousands of reactions and ten thousands of species. All right. So this is what is called as a reaction path. You need to determine which reaction leads to which reaction under which condition. You know, it's not an easy task. It's really hard. Once we figure that out, we need to talk about something called as reaction rate rules. But for now, let's just skip over the topic. Let us just assume that it has something to do with how those reactions are calculated.
Right. Once we have this, then we have something called as a detailed chemical kinetics model. We have put in so much hard work to develop these detailed chemical kinetics models, but we cannot use them right away. Why is that? Well, we can't be sure that these models are accurate. So in order to do that, we need to perform validation work. And that is why people spend a lot of time in experimental combustion, where they work on rapid compression machines, shock tube, internal combustion engines or some other type of combustion device.
And they measure quantities. They measure a lot of data, like what's the temperature profile in the combustion chamber? What's the species profile in the combustion chamber? Right. They measure all this data. And now your detailed chemical kinetics computational model is compared against experimental data. Now you need to get reasonable results for a wide range of condition. And only then is a particular chemical kinetics model acceptable. For example, in the context of IC engines, you can think of applying this detailed chemical kinetics model to predict emissions from combustion simulation and comparing that results with experiments that are run on the same engine.
Now, this is the entire life cycle when it comes to computational combustion. And the interesting thing to note here is there are specific companies or research organizations that are working on each of these topics individually. All right. So typically your automotive companies or engineering companies are working down the line. Right. Where they are using the chemical kinetics model that has already been developed by some established research lab. And they are applying that to predict the emissions of their upcoming engine. Okay. Whereas scientists and people who are involved in research, people who are involved in universities, they focus on the top end of the tree.
Right. Where more fundamental research needs to be done and a lot of calculations have to be done very accurately. Right. So this is the entire picture. And I would like to thank Lawrence Livermore National Lab for this particular slide. If you're interested in learning more about computational combustion, they have some excellent articles over there. You can read it in your free time. All right. With that, I would like to conclude this particular lecture. In the next lecture, we will get started by discussing some fundamental chemistry concepts that you need to be aware of in order to do computational combustion.
Please get in touch with me if you have any questions. Thank you. Bye.