Submind YouTube summaries
Thumbnail for Week 7: Lecture 31: Internal Flow: Introduction

Week 7: Lecture 31: Internal Flow: Introduction

Watch on YouTube

Video summary

This lecture introduces the concept of internal flow, which refers to fluid motion within confined channels such as circular pipes or non-circular cross-sections, marking a shift from the previously discussed external flows like those over flat plates or around cylinders. The fundamental distinction lies in how viscosity affects the fluid; in external flow, the boundary layer grows until it merges far downstream, leaving a large inviscid core, whereas in internal flow, the boundary layers developing along the entire circumference eventually merge at a specific distance from the inlet. Once these layers merge, the entire cross-section becomes dominated by viscous effects, and the region is termed "fully developed," where the velocity profile no longer changes with distance along the pipe, meaning the velocity distribution depends only on the radial direction and remains constant downstream. In the context of heat transfer within these confined flows, the analysis relies heavily on two primary reference temperatures: the wall temperature and the bulk mean temperature, often called mixing cup temperature. Unlike external flow which utilizes a free-stream temperature at infinity, internal flow defines the fluid temperature as an average value obtained by hypothetically mixing all the fluid in a cross-section. This bulk mean temperature is crucial because heat transfer occurs due to the temperature difference between the solid wall and this averaged fluid temperature. The lecture emphasizes that convective heat transfer cannot be separated from fluid mechanics; therefore, understanding the velocity profile developed in fluid dynamics is essential for analyzing thermal behavior, as the shape of the flow dictates how heat is transported and distributed within the pipe. The practical application of these concepts extends far beyond simple textbook examples to complex engineering scenarios like nuclear reactor rod bundles. Even though a fuel rod bundle consists of multiple rods generating heat in a complicated geometry, engineers analyze it by simplifying the problem into equivalent circular pipe flows using subchannel analysis. In such real-world applications, boundary conditions are often approximated as either constant wall heat flux or constant wall temperature to make the problem solvable analytically and numerically. For instance, in nuclear reactors, the power distribution along a fuel rod is not perfectly uniform but can be modeled as a series of segments with different constant heat flux values, simulated using electrical heating in experiments. This approach demonstrates that while the conditions might seem idealized, they are highly relevant for solving practical engineering problems where controlling both heat input and fluid temperature simultaneously is impossible. The lecture concludes by outlining the roadmap for future modules, which will delve deeper into analyzing these two specific boundary conditions: constant wall heat flux and constant wall temperature. The primary goal is to understand the development of thermal profiles within the pipe, analogous to the hydrodynamic fully developed condition already discussed. By examining how the temperature profile evolves from the inlet until it reaches a state where it no longer changes with axial distance, students will gain the tools necessary to predict heat transfer rates in various internal flow systems. This foundational knowledge is critical for designing efficient thermal systems, ensuring that engineers can accurately model and optimize performance in applications ranging from simple piping networks to high-stakes nuclear energy generation.
Read the full video transcript
Yeah, welcome everyone. Uh this module and the subsequent three or four modules, we will uh look at internal flow. That is flow through pipes or uh flow through uh non-circular cross sections. Uh till now we had looked at uh external flow, flow around a cylinder, flow over a flat plate, flow around a sphere, etc. So, just as in fluid mechanics, we had uh boundary layer formation over a flat plate. And then we also looked at how pipe flows are going to be characterized. So, the same logic is being followed in heat transfer. So, having done external flow, now we are going to look at internal flow. So, as you all know already from fluid mechanics, internal flow is essentially where we are talking of flow through a confined channel. So, by confined channel, I mean something where you have a Let us take a circular pipe. It could be a rectangular channel or any such channel. You're going to have the boundary for flow all around the circumference. Physical boundary. So, this imposes a very important restriction or constraint on the flow because as we know in external flow, if if this is my external flat plate flow situation, this is my coordinate system, we have what we call as a boundary layer which is being formed. And this boundary layer develops as I move further in the X direction, the thickness of the boundary layer goes on increasing. And what this means is the viscous region goes on becoming more and more prominent. So, the inviscid part, which is the external part, which is the outside of the boundary layer, is there, but the effect of viscosity gets propagated to a greater and greater distance away from the solid surface. This is what happens in external flow. Okay, this is because we have uh no boundary on the top. Okay, suppose we had a flat plate and flow between two parallel plates. Okay, this is one one plate and some plate is there very far off, whatever. The boundary layer will develop like this. And from the top also a boundary layer would develop, and somewhere here the boundary layers may even merge. So, this is what is going to happen in case of flow through a circular pipe or any such closed uh uh geometry, where the boundary layer, which is formed all around the circumference, would end up merging at some location some distance away from the inlet. That depends on various conditions. Uh we'll see what they are, but the important point to notice, if you take a look at this flow between two parallel plates. So, this is parallel plate one. This is one and two. We will have the viscous region here. This is also the viscous region. This is the inviscid part. And as you proceed along the flow direction, the inviscid part goes on decreasing, and eventually, when the boundary layers merge, the entire flow region becomes viscous. So, viscous effect is there throughout the flow. Okay, unlike having been able to segregate this into two parts, whereas this is the inviscid part, which was decreasing with respect to distance. So, inviscid part decreases with X. Viscous region increases. How much does it go to? It will go to the entire space that is going to be there. So, that is what happens. The flow becomes entirely viscous. So, here also in flow through a circular pipe, we do circular pipe as our example because most real-life situation, engineering applications would have pipe flow as one of the simplest geometries of interest. And I'll tell you something. Even when you're talking of very, very fancy uh height, high level of stuff like flow through a nuclear reactor rod bundle. So, a rod bundle essentially consists of a set of fuel rods. Uh you could have 5 by 5, 7 by 7, 9 by 9 such This is a fuel rod bundle. I'm just drawing a representative here. Okay, so if I'm looking at this kind of a setup, this is a nuclear fuel rod bundle. This is fuel rod. And this entire arrangement is called a bundle. This is generating heat. These are generating heat. This is generating heat, etc. So, if I look at a region which is here. Region A. If I look at another region, which is a region B. If I look at another region, which is a region C. And region D. These are all empty spaces through which flow is going to happen. Flow is happening through into the plane of the board or outside, whichever way, and this rod is supplying heat to it. Okay, so I'm just shading them to indicate these are solids and heat generating surfaces. Um So, this what happens? The fluid as it passes through gets hot. Okay, so this length of this heater rod bundle is about 3 3.6 m. Diameter of the fuel pin is about 12 mm, and pitch is probably 15 or 20 mm. So, whatever it is, the area available is a very small narrow area through which coolant is flowing, and the heat generated by the fuel rods is taken away for steam generation, etc. The point I'm trying to make is even such a complicated geometry, okay, is translated is analyzed as if it is a pipe flow. So, even this geometry, they will analyze as if it is a pipe flow problem. So, this is going to be analyzed as if this is a flow through a pipe, and this heat The heat that is coming in from these four rods is as if it is being supplied through a pipe across the wall. So, this is the Q double prime from the four surrounding fuel rods. So, each region A is surrounded by rod one, rod two, rod three, rod four. Each of this is supplying some amount of heat to this fuel of fluid inside here, and this heat is taken and said that it is as if there is a flow through a circular pipe of some diameter. This diameter DH I will try to find out what it is, but same mass flow rate, same heat flux as is done through the uh fuel rods channel. This is called a subchannel analysis. So, point I'm trying to make is even when you have a very complicated geometry, uh very very real-life application, flow through pipes, circular pipe is what is used for analysis. Therefore, whatever we are doing now is not something which is only confined to textbooks or is it it's a it's a approximation, we don't know to do to do anything better, therefore we are doing it. It is something which is very very practical and therefore it's being done. Okay, so I will just draw the pipe uh wall a little better. This is the pipe wall. And as the flow comes in, so we typically generate uh show the flow like this. Uh Now, if you look at external flow, and if you look at internal flow, velocity is U infinity. In internal flow, there is nothing like U infinity because infinity is where you have free stream, where nothing is affected by the presence of the solid surface. Here, you please uh think of it like this. Suppose you are you are walking through, you know, holding hands, like you know, your friends are walking and going through holding hands, and you have to cross a small gate, and the constraint is that all of you need to cross at the same time, pass through the gate at the same time, you will come closer to each other and try to squeeze through. That's what happens to the fluid also. Outside, when it is coming in here, it is all free free, but once it it sees that there is a pipe wall which I have to take care of, the the fluid will have to adjust itself to this particular contour, to this shape. And therefore, what happens to the fluid? The The boundary layer that we have studied in fluid mechanics will start generating, grow, grow, grow, grow, and it will grow around the circumference, and you will see that the boundary layer will merge at this point. Whatever is coming from here would come like this and merge at this location. So, imagine this is like a plastic sheet which is just being pushed and it's allowed to merge here. This is obviously symmetric. It doesn't look symmetric the way I have drawn because of the tablet, but this is not a straight line. This is more of a curved geometry like this. I don't want to Yeah, this is better. It's something like this. And you'll see that the boundary layer merges at a certain distance, at a certain location. Okay? So, this is your viscous part. This is your viscous part. And here, this is the inviscid core or the inviscid central part. Inviscid central part. And as you see, the inviscid region goes on decreasing, and finally, in this After this location, after this, fully viscous flow is there. Okay? Entirely viscous. Okay? So, this part, of course, we know it's called as a fully developed region. Fully developed region. We know this from fluid mechanics. And we know in internal flow, we all we have a quantity called as average or bulk mean velocity. Which takes over the role of what U infinity would take care of. So, I will draw the velocity profile, etc. So, here you had U infinity, here you had bulk mean velocity. Now, in heat transfer, life is a little bit different. So, in external flow, remember, we had this as T wall, and we had T infinity. T wall and T infinity, and we had a local temperature T of X, Y. So, we had T wall, T infinity, T of X, Y. We had this. Now, here, of course, U of X, Y would be there. Here also U of X, Y would be there. Okay. Now, from heat transfer point of view, T wall would remain, but instead of T infinity, there is no scope for this T infinity, we have what we call as bulk mean temperature. Mixing cup temperature, bulk mean temperature, or average temperature. I'll tell you what this is. And of course, we have local temperature. Bulk mean temperature. When we say water flows through a pipe at 2 m/s and 30° C, the 30° C which I'm specifying is the bulk mean temperature. You take this water in a insulated container, put a thermometer, and measure the temperature. That is called as the bulk mean. That's why it's given the name mixing cup temperature. So, whatever is the fluid, take it into a cup, mix it well, stir it well, don't allow any heat to pass out through, put in a put in a thermocouple and measure the temperature. It is an average value which you're going to see. It's called as the mixing cup temperature. Okay. So, unlike external flow, where we had U infinity, here you have U bulk mean, U bulk mean, or U bar, whatever, or whatever you want to call. I'll just write what whatever you see, U bar or V bar or U average, anything. These are all meaning the same. T infinity doesn't have any value, but we have T bulk mean. T wall necessary condition for heat transfer to occur. So, I have to have a temperature difference between the surrounding wall and the T fluid. This has to be there, otherwise there is no uh heat transfer possible. T X Y is in this one, we would give it as R {comma} X or R {comma} Z. In internal flow, the coordinate system will be such that this is my X or Z direction. This is your radial direction. Okay, as we had in pipe flow. I will use Z. Okay, simply because X can refer to thermodynamic quality in uh uh phase change applications. But, we can use X also because entire course is on single phase heat transfer. So, T T wall will also be there. And of course, T local X {comma} Y will be there here also T local X {comma} Y would be there. Okay. Now, heat transfer, convective heat transfer, we all know is coupled with in coupled with the fluid mechanics part. You cannot separate the fluid mechanics from the heat transfer. So, whatever we did for pipe flow, everything has to be taken here in the heat transfer part. That means, whatever I I understood as velocity profile, velocity distribution, etc., those things will continue to be there. I'm just going to revise this very quickly because this has been part of your curriculum already in fluid mechanics. So, if this is the boundary layer which is formed, and let me put in my R and X here. Okay, this is fluid coming in at velocity U infinity. Okay, so if I plot the velocity distribution at some location here, I would see and then I would have a constant value and then a varying value like this. If I plot the velocity distribution at some other location, some other location here, second region, I would have constant here again in this part and doing this. So, this part, which I'm shading, is where velocity is constant. Because that is a part of the inviscid region, where the necessary thing to slow down the velocity, which is the effect of viscosity, has not propagated. This is the viscous region, entire part, and the shaded part is where is we call as the inviscid part. Inviscid. And this after the boundary layers merge, we call this part as fully developed flow. And what do we know about fully developed flow? The velocity distribution anywhere is exactly the same at any other location. So, U Z, which is a function of R, but not a function of Z. So, dU by dZ is equal to zero. Velocity is invariant along the flow direction. In fully developed region. This is extremely important. dU by dZ or dU by dX, whatever you want to call. I'm somehow used to using Z but it's same as X. Okay? This condition we all know from fluid mechanics. That means if I take a picture or velocity plot at this location A and do this do the measurements of velocity at some other downstream location B and overlap the profile at A to all the profile at B, they will be exactly the same. That means velocity is only a function of the radial direction and independent of the actual direction. This is the actual direction. Actual direction. X is the actual direction. Independent of the actual direction, which means dU by dX is equal to zero. Please remember this condition. Okay? So, UZ is only a function of R. In the part in this part which we call as the developing region, that is still here. Velocity, as you can see, there is the viscous part velocity is changing, the inviscid part where the velocity is constant, and because m. is constant by continuity equation and at steady state, the integration of the velocity profile The integral of the velocity profile that is whatever you see row UDA is constant. That means if I do the velocity distribution at location A1, A2, A3, etc. All the integrals should add up to m. Okay, I mean should be equal to m. That means if I have a slow moving portion here the center velocity would come out to be something such that it satisfies the conservation of mass which is given by this U infinity business. Now as I move downstream the viscous region is increasing between section A1 and section A2 the viscous part has increased. That is what do I mean by that? This region is thicker here compared to here. Correct? So you have a slower moving fluid extending to a greater uh distance away from the wall. That means mass conservation has to dictate that the central core which is inviscid should move faster so that this constraint is satisfied. That means U center line is not constant. in developing region. But in fully developed region U center line is constant because the profile itself doesn't change. That means whatever is the U center line or U max U center line is same as U max that is going to increase. So if I talk of U center line it is going to increase this way and settle to a value. Okay, some value. So this is a function of X. Okay, I'm not saying it's it will not start at zero. This need not be a zero zero, okay? The idea is use centerline will go on increasing until until the flow becomes fully developed, and after that it will become constant. Life is easy in fluid mechanics, okay? But now if you go to heat transfer, the same situation, the same situation that I have, for heat to flow, for heat to flow, we need T wall, we need T fluid. And this one is proportional. Correct? Now, what is this? T fluid, what is the temperature of the fluid that I'm talking about? So, let us look at it in a little better way. I'll just draw a bigger diagram so that we can draw many things, write many things here. So, this is the centerline of the pipe. Okay? And this is my R, and this is my Z. Fluid is I have In heat transfer, I can have two conditions. One is constant wall heat flux. That is Q double prime is equal to constant. Second one is constant wall temperature. Which means T wall is equal to constant. What is so special about these, and why is this even being mentioned? Well, when I have to do analysis, first of all, I can control or I can I I control the wall heat flux and allow the fluid temperature to vary or I can control the wall temperature and allow the heat flux to vary. For me to control both is impossible. I cannot control the amount of heat I put in and also the temperature of the fluid. I can control only one. If I put in so much amount of heat, so much is the variation in temperature so on and so forth. So, this is something which I have to be I have to live with. Otherwise, if I say this is how my wall temperature is going to behave, the heat transfer will only behave in this manner or it is controlled it is constrained by the fact that the wall equal to constant. Why not and why are these two cases important? Because in real life, in most engineering application, one or the other condition is what is you can approximate a real life problem to. That means I might have something which is not a function of I might have a situation where uh the wall temperature may not be constant neither is the wall heat flux constant. But, you with with certain approximations, you can at least have piecewise constant values for uh either the wall heat flux or the wall temperature. And why do we need these? Because these two cases are the cases where both analytical as well as numerical solutions are possible. Experiments can be done. Experimental data is available. Because how do you how do you give constant wall heat flux? So, if I want to supply heat to a pipe carrying a fluid, m. Fluid comes in at 30°C, I want it to go out at 60°C, I have to supply heat. So, I would have a you know, resistance heating, which is going to be like a wire which is wrapped around the circumference. Okay, so Q is equal to V * I. Okay, so I will give that much amount of power so that in a given length L, this will come out to be 60°. So, I have control over it. So, Q is known. Q double prime is Q divided by pi d l, which is known. This is heat flux is equal to constant case. Okay, so when I do experiments also, I will have to incorporate one of these conditions. So, either this or this condition will have to be incorporated or will have to be created in the experiment. Numerical also, this is what we'll have to do. And more importantly, just because it is convenient, I'm not doing it because it is convenient. I'm doing it because many real-life engineering applications, either this or this. I'll just give you one very simple, quick example. Again, it comes from the reactor nuclear reactor business only. This is typically the power shape of a nuclear fuel rod. So, this is a single rod, if I'm going to say. This is the rod, fuel rod. And this is what I call as a power profile. That means this is a sign kind of profile. This is typically Z direction. Q double prime is Q double prime not sin pi Z by L. This is my L. Okay, length of the rod I will And this is my Z direction. So, typically it is this way. Now, uh when you conduct experiment, when you do experiments for these kind of things, you don't have a fission nuclear uranium dioxide or any such material here. You do these experiments using electrical heating. So, you simulate this using electrical heating. By that, I mean I have this this region split into Just show this here. This would be the And then, I mean Okay? So, this is for this length L, I will have small regions where there is electrical heating of different levels, so that it mimics this power profile. How is that possible? So, from zero to say L1, L2, L3, L4. So, I would I would have a certain resistance of heating here, so which will give me so much amount of heat flux. Next portion, I want to have higher heat flux, so I would have a higher resistance. Next small distance, still higher. How small can this be? Ideally, it should be like this. Okay? But, practicality also is there. So, we cannot have such a you know, small step step. So, what is done is over a small region, this is done. And this power profile is essentially mimicking this. Midpoints of these, if I take and join, it's essentially a what you have seen here. So, experiments are conducted where in this region, you have constant wall heat flux. In this region, you have another heat flux value, which is different from Q1 double prime, but it's still a constant. Third region, some other constant heat flux value greater than Q double prime two, but still it's a constant. So, even in a nuclear reactor assembly situation, to conduct experiments of this type, you have electrical heating, but nothing changes in in real life abruptly. So, you you you will have step changes which will be modeled as or which is which is seen to mimic this kind of a profile. And therefore, here, when you do the analysis for this part, it is constant wall heat flux situation. In this part, it's also a different constant wall heat flux. In this part, another constant wall heat flux. This is how the analysis of heat transfer in a nuclear fuel rod bundle also progresses. So, uh what I try what I'm trying to say is even though these seem to be very textbook-like conditions, these are very very important, very practical, very relevant to real-life engineering. In the next module and further modules, we will look at how to analyze both these cases. So, first we will look at how the temperature profile is there for a flow through a pipe situation, try to understand the concept of thermally fully developed just as we had velocity fully developed condition. This one was for velocity, hydrodynamics. We'll try to look at what is there for heat transfer similarly, and then proceed with the analysis. Thank you. >> Hey.