Lesson finished
Up next: Momentum Conservation and Deriving the Navier-Stokes Equations
Unlock all 49 lessons of Turbulence and RANS ModellingFull course ₹40,000 · paid
Course Structure, Averaging and the Continuum Assumption
A free lesson from Skill-Lync's Turbulence and RANS Modelling course, on Grow by Skill-Lync · Last updated 16 Nov 2026 · Session: “Week 1 Session 2”
What will you learn in this free Turbulence and RANS Modelling lesson?
The instructor explains the course format, including weekly sessions, challenges and two projects, one coding a two-dimensional turbulence model. The lesson then shows how time averaging smooths a turbulent velocity signal, covers the continuum assumption and mean free path, and derives the incompressible mass conservation equation.
- Course format and the two projects
- How averaging converts random signals to smooth profiles
- Continuum assumption and mean free path limit
- Deriving mass conservation for a fluid element
What comes after lesson 2?
The next free lesson is lesson 3, “Momentum Conservation and Deriving the Navier-Stokes Equations” (27 min). Later in Skill-Lync's Turbulence and RANS Modelling course (paid): lesson 4, Week 1 Session 4; lesson 5, Week 2 Session 1; lesson 6, Week 2 Session 2. The full course has 49 lessons (19 h 41 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.
Hello, everyone, and welcome back to the second session of week one of the course, Turbulence and its Modeling. In this session, we will focus more on what the course consists of in terms of the structure. We will start diving in into the governing equations of the fluid motion. First of all, let's recap what we did in the last one. So we understood that turbulence is an irregular and apparently random condition of flow. And it is indeed important to understand it, but because there is no mathematical theory to exactly define it, it's a problem.
We, engineering or in physical systems, need to understand the behavior of turbulence and its effects. So it becomes important from the engineering point of view as well. Then we understood that in this course, we will talk in the first half about the physics of turbulence, and then in the second half, more about the turbulence modeling. All right, so in this session, as I said, we go into course details, and we start the mathematical description of the flow, which is the starting point of describing turbulent flows rigorously.
So by the end of this video, you will have a clear idea about the course structure, and you will have a know-how on how to start describing the fluid flows. So this will basically be close to what you have seen in your fluid mechanics course. Right, so regarding the first topic, that is the course details, the course is organized into lectures. So that is, I will be giving you, and then there will be some homework. So the lectures will be organized as four 20 to 30 minute sessions per week.
So I will call them sessions. So each week will have four sessions. Then after every week, or during the week, you will get challenges, which will be related to the lectures. So for example, if we start doing the derivation of Navier-Stokes equations in this session, then I will ask you maybe an alternative way to arrive at a particular expression, an alternative way to describe the same equation in a different form. There will be two main projects. So the project one is that we together develop a two-dimensional code, which is of which the first step is to write the laminar form of the code.
And then we will take a simple turbulence model and code it up in this laminar form of the code. So then we will have a two-dimensional turbulence model. In the second project, we will look at a practical geometry. So something closer to what we see in the engineering world. And we will then compare various turbulence models. So by project two, you will have a good idea, a good clear idea on the range of turbulent models that we have, their disadvantages and advantages. So at that point, then we can compare these different models on practical geometries.
During the course, I will assume some knowledge of fluid mechanics. So if you do not understand some points, I can suggest you some extra reading. Anyway, in the end, there will be a question session where you would be able to ask me questions regarding the lectures or homework or whatever. Right, coming back to the subject matter, you must be remembering this slide from the last session, some of the characteristics of turbulence. And I promised you there that I will explain a bit more the last point, which is that turbulence can be simplified, but the details in the turbulence can be simplified by taking out an average in time and, if permitted, also in space.
So we are trying to simplify the details and we can start understanding it by taking one point in space. So, which is shown by the tail of this arrow. If you plot the velocity on the given location, it will have this random nature that you see in this plot. So here we are plotting the velocity, the y-axis and time on the x-axis. So how the velocity varies in time. So you see that there is a random variation in time. So there is unsteadiness. A property, usually, that turbulence shows is that if you start taking average of these random signals, this average converges to a particular number.
So here, that number is u bar. And this is the property that we are really interested in as engineers. So to explain it a bit more, I will take the example of this 3D flow that you see in the picture. So let's take cut plane through this smoke and see how the averages make the flow simpler. Here, I will try to replicate what we have seen in the picture. So pay close attention to how I draw or how I represent the turbulent structures here. So, well, just the boundary of the picture very roughly.
But now is the important bit. How do I show the interface? Okay. So we need to have small structures as well as large structures. So let's see what we get. Okay. So I made some representation of smaller eddies than the bigger eddies. So this is the interface between the smoke and the air. Okay. Let's say roughly this is the smoke bit that you are looking at. Now, what I want to do is, let's say I take a velocity profile along this line. So along this line, because there are circulations here of different sizes, and if I redraw this line here, and I want to plot the velocity on this axis, and this is the direction y, say, how these velocities will be looking like.
So let's say at an instant t0, so let's say this velocity looks like, okay, so I'm going to redraw all these velocities here for clarity. So let's say this velocity at t0. At the second time instant, the velocity could be something like that. Then at the third time instant, so when we have enough eta, So for example, we have a velocity profile at time t0, t1, t2, until the nth time. So what we can do is really add them all up. So let's say add all the velocities from i0 to n.
And then simply divide by their numbers. So that will be our average velocity profile. So how would that look on your graph is something like a more smoother profile. That represents all of the other, the whole 2D flow would probably look, again, more smooth, especially this interface. Smoother, say like that. And the velocity profile I was showing you would be, again, a more smoother profile. This is what we are after, an averaged velocity profile, or averaged pressure profile, and so on. So coming back to our slide, the question now forms like that.
Okay, I have got the idea of how we go from random looking pictures, which probably have some kind of unsteadiness, to smooth looking pictures, which we are after. But how can we describe? I've this mathematically. So for example, a question could be, for example, at a distance, two kilometers, how wide is this smoke as it moves in the atmosphere? So we have to form this question in more general terms, like how in a generic fluid volume the flow is behaving. So what we are after is how velocity or pressure or density or smoke concentration in this case varies with space.
So for that, we go into the mathematical description of the flow. As we are going into mathematical description, we need some variables. So we have space and time in which the fluid flow is varying. Then we have velocity, pressure, density, which is rho here, and the energy equation. All these variables, so three components of velocity, pressure, density, and energy, would comprise of a general flow description. So there is compressibility involved. If the flow is incompressible, we can get away with just the velocity and pressure. Where these variables are coming from, it originates in the kinetic theory of gases, and I'm not going to go there.
But I presume that we know that these variables are enough to completely describe a given flow condition. In addition, for example, in the smokestack we were looking at, there is concentration. So there can be scalars that are, you know, moving with the flow in simple terms. So for example, in Reynolds' dye experiment, there was dye in our picture. There is smoke, and there can be other such scalars that move with the flow. So these form the variables that are necessary to describe the flow. Then comes the fundamental assumption of continuum.
So most of the fluid flows, especially near the Earth's surface, the flows that we are mostly concerned with, they follow the assumption of continuum. So what it is basically is that we know that fluids on microscopic scale are composed of molecules. But when we want to look at them, we do not want to deal with all the random motions that the molecules are doing. We are mostly concerned with the average properties. So the microscopic properties, such as density, pressure, temperature, bulk, velocity, which can be defined on a small volume, which is bigger than the motion of these small molecules.
So, for example, again, I draw an idealized box, let's say. And in that there are some molecules, so we are really zoomed in into a fluid flow picture. So let's say these molecules, they are interacting with each other. So this could be a typical path of this molecule in motion. What it is doing at bigger level, so if you effectively zoom out, this whole thing, this whole thing, this volume, so it is essentially the same. It's, say, moving with a velocity V, like that, and has a pressure P.
So instead of going into the details of all these molecular motions, we are concerned with the details of this block here. And we call this an infinitesimal fluid element. And as we will go in the mathematics, we will have dimension delta X. So we basically said is that we are not concerned with individual molecular motions, but their averaged properties. So for this assumption to be true, there are two main factors. So the first factor is the average distance of these collisions, which is also called mean free path.
You must remember from the kinetic theory of guesses. So the factor number one is mean free path. And the factor number two is the smallest distance. Turbulent motions that are happening in a turbulent flow. So as we remember, there are big circulations. Then there are smaller ones. Then at some point, there are the smallest ones. So that form the characteristic length scale. So the length scale, or let's say in loose terms, the radii of those smallest motions would be called the characteristic length scale. So what we are assuming is that the characteristic length scale of the fluid motions is much bigger than the mean free path of the molecules.
So in that condition, so what we are saying is that mean free path or the characteristic length scale is much smaller than one. So this is the condition. So this is the condition where continuum assumption is valid. Indeed, there could be cases where this condition becomes invalid and uppermost layers of our atmosphere. So when the spacecrafts, they are reentering into Earth's atmosphere, the usual equations of fluid motions do not apply because there is no continuum. So this condition can also happen if we are dealing with flows that are active at very small scales, so at molecular scales or nanoscales.
Right. So when we will use this continuum properties, I will highlight it to you. So we talked about the basic variables needed and the basic assumption. Now, I think we are ready to go into derivation of the mathematical description of the flow, otherwise known as the Nagier-Stokes equations. So these are basically conservation laws applied to the fluid flow, and we will start with the conservation of mass. Then the conservation of momentum will follow. I will skip the conservation of energy here because in this course, we are mainly focused on incompressible flows.
This is mainly because if you have a grasp on incompressible turbulent flow, it is natural and extend over to the compressible flows as well. Then following that, the conservation laws will not take us to the final form of the equations, but we will need to use a relation between viscous stresses and how these stresses deform the fluid elements. So we will look at this more in details in the coming slides. Okay, so let's start looking at the conservation laws. Very simply put, this conservation deal just means that if you have an arbitrary surface that is, say, S, enclosing a volume V, or you can also make these things differential.
Then what we are basically trying to say is, so if a property, if a quantity enters the surface, so concentration. So smoke concentration, C enters into the surface. C out leaves the surface, and let's say there is a process going on inside the surface where the smoke is being produced. So let's say there is C production. The conservation laws are basically saying is that the rate of change of the smoke concentration is equal to the net change from the boundaries, so C in minus C out, and the contribution from inside the surface.
So what is happening in the volume? So then C production. This is a basic tool that we will use for the conservation of mass and momentum equation. Okay, then we can just extend. This idea to the conservation of mass, but here we will define a bit properly our fluid element. All right, so let me quickly draw a little cube where I will define the coordinate system. So we have x, y, and z-axes. So this is how we define our space. So accordingly, you will have, say, delta x, delta y, and delta z.
So in this fluid element, now you must remember that continuum applies here. So we can say that the properties, they vary smoothly from one fluid element to the neighboring fluid element. And this fluid element is just in an arbitrary position in any fluid flow. So for example, this could be inside that smoke stack that we were looking at. For example, that's the smoke stack. The smoke is moving on that side. And we are just taking a small volume inside that. The rate of change of mass in time, so is equal to the change from the boundaries.
This is the same statement that we have said in the previous slide. Here, there is no production of mass inside a volume. In fact, usually, you would not have any production of mass that is reserved for nuclear reactions, say. But here, we are just going to convert this mass flow rate as density and velocity. And this comes from, again, the basic fluid mechanics course. You should be able to identify this. So then how can we write the whole equation is that, so this is the rate of change.
And now I'm writing the change through the boundaries integrated over the surfaces. So this n is basically the normals of our fluid element. And all this thing sums to zero. So basically, I have taken this side on the left-hand side. So to clarify, I repeat that the rate of change of mass is this term density multiplied by the volume of our fluid element, and here is the contribution from the surface. So the mass entering the surface and mass leaving the surface of the fluid element. Now, this term can be written by divergence theorem as an integral over volume.
And so the whole equation, as the element of the volume goes to infinitesimally small, the whole equation can then simply be written as... Now, we will mostly be dealing with incompressible flows. So that means the change of density is negligible. So this term, we would neglect. And incompressibility means that density is constant. So change of density with any independent variable would be zero. And this will basically mean that this is our mass conservation equation. Let me put it again here more clearly. This is our mass conservation equation.
So I think I will stop here for this session, and then we will move on to the conservation of momentum. More details about that in the next session. Okay. So we have talked a little bit about the conservation law applied to the mass. I would like to summarize this session here. So we talked that the course will give you the mathematical and physical insight of turbulence, but we will also do some hands-on coding and practicality of the experience, a better understanding of what are the turbulence models doing.
Then we understood the concept of averaging in turbulence flows, how it simplifies the problem, and becomes an important tool for engineers. Then we quickly started the mathematical modeling, the general mathematical modeling of flows has the turbulence properties inside. It will become more and more obvious as we move further in the course. Continuum assumption and why is it important? And we have started application of the conservation laws. In the next session, we will continue this. For now, thank you very much.