Submind YouTube summaries
Thumbnail for Week 5: Lecture 25: Non dimensionalisation of governing equations

Week 5: Lecture 25: Non dimensionalisation of governing equations

Watch on YouTube

Video summary

The lecture begins by refining the derivation of the energy equation, emphasizing the importance of understanding every step rather than skipping details. The instructor explains that enthalpy ($H$) can be expressed as the product of specific heat capacity and temperature ($C_P T$), which transforms the governing equation into a form involving convection, conduction, pressure work, viscous dissipation, and internal heat generation. In this expanded version, the convective terms represent temperature changes due to fluid motion, while the instantaneous change accounts for temporal variations. The instructor uses an analogy of entering an air-conditioned room to illustrate how convection causes a gradual temperature shift along the flow direction, whereas sudden impacts like hitting an ice ball represent instantaneous thermal changes. The core of the lecture focuses on applying the principle of similarity by non-dimensionalizing both the momentum and energy equations for steady, two-dimensional flows. For the momentum equation, variables such as length, velocity, and pressure are normalized using characteristic scales like plate length or free-stream velocity. After substituting these dimensionless variables and dividing through by appropriate inertial terms, the equation simplifies to reveal that the resulting flow characteristics, including velocity and pressure distributions, are governed primarily by the Reynolds number. This dimensionless group represents the ratio of inertial forces to viscous forces, highlighting its critical role in determining fluid behavior in momentum transfer problems. When the same non-dimensionalization process is applied to the energy equation, a more complex set of dimensionless parameters emerges. The analysis shows that temperature distribution depends not only on the Reynolds number but also on the Prandtl number and the Eckert number. The Prandtl number is identified as the ratio of momentum diffusivity to thermal diffusivity, linking viscous effects to heat conduction capabilities. Meanwhile, the Eckert number represents the ratio of kinetic energy to enthalpy difference (thermal energy), indicating how much kinetic energy is converted into heat due to fluid motion. Consequently, while friction factors in momentum problems rely solely on the Reynolds number, heat transfer phenomena like the Nusselt number are functions of all three parameters: Reynolds, Prandtl, and Eckert numbers. The lecture concludes by summarizing these key takeaways and previewing future topics involving scale analysis to better understand boundary layer thicknesses and friction factors. The instructor stresses that non-dimensionalization is a powerful tool that reduces complex physical equations into forms governed by specific dimensionless groups, allowing for broader applicability across different scales and conditions. By identifying which forces or energy mechanisms dominate through these numbers, engineers can simplify problems and predict fluid behavior without solving the full differential equations for every specific scenario. This foundational understanding sets the stage for deeper investigations into heat transfer correlations in subsequent classes.
Read the full video transcript
[music] [bell] [music] [bell] [music] So we are we were into the interesting derivation of the energy equation. It may sound a little longish but I would implore upon all the students to derive the way I have derived. I have tried not to skip any step so that you are with me all the time. Okay, please do derive it's once in a lifetime at least you should derive once. Okay. So we were here we were we had derived this equation. Now we were just about to see that H in the form of temperature. So to get that I need to write H as equal to CP into T. From thermodynamics we know that. So if I do that, if I do that, H is equal to CP into T and my equation row D H by DT equal to K into D ² T by DX² + D ^ 2 T by D Y 2 + DP by DT + 5 + Q dot prime Now if I substitute H is equal to CP into T and CP is constant. So row cp d t by dt dt by dt is equal to k into do ² t by dx² + d ² t by d y² + dp by dt + 5. Remember this is the viscous dissipation term which had all the velocity gradients. This is this this is what you can see that viscosity and velocity gradient. Viscosity velocity gradient this is stretching this is stretching. This is this may cause either rotation or angular deformation. Okay. So depending on the uh plus or minus plus q dot triple prime. So if you divide through this is the whole energy. This is what we were up to. We can see that now this is the energy equation. Okay. And this if you expand this row cp instead of dt dt let us expand this do t by d t plus u do t by dx plus v do t by dy is equal to k into d ² t by dx² + d ² t by d y² plus dp by dt this is the total derivative of the pressure plus viscous dissipation term plus q dot triple prime. So this is the total derivative that is the temperature is varying with time and this temperature is varying with x and y by virtue of convection. If you have to physically feel this, if I consider this room, when I'm entering into this AC room with a velocity U, I experience a velocity temperature gradient of delt by d x. That is the convective term. But someone while entering someone hits me with an ice ball so that I will experience instantaneously a temperature change that is dot t by dt. So these terms are because of the temperature variations because the by virtue of convection that is u and v because they have the velocity and this is instantaneous temperature change. So this is convection term. This is heat transfer by conduction and this is pressure work and this is viscous dissipation. You can see that it's all because of the velocity gradients and the viscosity viscous dissipation term. Okay. So just recall viscous this is viscous dissipation term. This is viscous dissipation term. Okay. Now and dp by dt is nothing but dp by d t + u d p by d x + v dp by d y. And this is the heat volumetric heat en heat heat generation either through chemical reaction or radiation or electrical heating within my control volume. So now I have conservation of mass momentum energy equation. Now the question is we will apply what is called as principle of similarity and nondimensionalize this equations and see what what are the nondimensional numbers which are going to emerge while we do this equation. So in the conservation of momentum if you just take conservation of momentum this is your conservation of momentum equation you can see that this is the x momentum equation you have the inertia force this is the inertia force this is the pressure force these are the viscous forces and this is the body force and du by dt if you expand you get du by dt and these are the convective terms If you consider, let me consider steady and twodimensional flows. If I consider steady and two-dimensional, the moment I say steady, nothing varies with time. Then d u by dt is zero. And zed term that is d u by zed and w term does not exist because it is a two-dimensional flow. So this term this term will vanish. So let me write the steady flow for steady and twodimensional flow. The x momentum equation I write x momentum equation. So if I write X momentum equation, you have row into [snorts] u d u by del x + v into del u by del y is equal to - p by dx + mu into d² u by dx² + d ² u by d y² 2 + mu by d by dx of mu by 3 into d u by dx + dv by dy plus f of x. So what I do is I non-dimensionalize x with a characteristic length L. Characteristic length. L is characteristic length. For a flow over a flat plate, it can be length of the plate. For internal flows that is pipe, it can be diameter of the pipe. And yar is y by l. Yar xar are non-dimensional and u star that is velocity u non-dimensionalized with u infinity. U infinity is the free stream velocity. That means if you take flow over a flat plate, what is the velocity with which it is flowing on the flat plate and V star equal to V by U infinity and P star is equal to P by row U infinity squared. If you do non-dimensionalization like this, if you substitute this in this above equation, if you substitute all of them here, so what is that you get is row into u star u becomes u* into u infinity. Okay. And do of do xar x is xar into length. So you have L into U is U infinity into U star. U infinity is constant. So U infinity into U infinity U star. Similarly for V you have Var into U infinity that is 1 U infinity and do YAR YAR into L you have L and U do U is do U* into U infinity. So that's what you get. Now here again you have P. So you have do p star that is you have row u infinity squared minus sign is there divided by do xar divided by l plus same way you can do mu into u infinity by l² into d ² u* divided by dx* squ + d² u* divided by d Y*² plus mu into U infinity by L² D by D XAR into 1 by3 into d u* by dx* plus do var by d yar plus fx here I'm neglecting neglecting body force. Neglect body force. There are no body forces. So now divide throughout. Now divide through out by throughout by U infinity squared by L. Divide throughout by U infinity. So left hand side will become row U infinity squar by L U divide through. So if you do that you get u star d u* by dx* [snorts] plus v* d u* by d y* that is on the left hand side and u infinity squar row u infinity squar by l we'll get so you have minus row u infinity squar by l into l by row u infinity squar into dopar by do xar R + mu U infinity by L² into L by sorry L by row U infinity squared into do ^ 2 U* by D XAR squar + D ^ 2 U* divided by D Y* squar plus mu U infinity by L into L by row U infinity² into d by dx* into 1x3 do by d of u* xar plus do var by do yar you get so here in the if this gets canceled out okay here what is that I end up getting I end up getting I end up getting row u mu by row u infinity l and here also mu by row u infinity l. So if I write that so the equation turns out to be u* d u* by dx* plus v* d u* d yar equal to -1 by row dopar by dox* plus what is that mu by row u infinity Infinity L that is nothing but 1 by Reol's number. If you recall Reol's number is row U infinity L by mu that is viscous inertia force divided by viscous force that is the Reol's number. So you get 1 by R E into do² u* divided by do xar squar + do ² u* divided by d y* squar + 1 by r e into d by d xar into mu by 3 into d u* divided divided by dx* plus d vstar divided by d yar. The point is the takeaway from this is that when you do non-dimensionalization of momentum equation the non-dimensional number which emerges is reol's number. So if the anything which comes out of the momentum equation the velocity distribution the pressure distribution all that is decided by reol's number remember that that is what is emerging and reol's number is nothing but row u infinity l by mu so similar exercise let us do it for energy equation so if you go to the energy equation If you go to the energy equation, this is my energy equation. So my energy equation is consisting of temperature. My energy equation is row CP do T by DT that is this is the convection term. This is the conduction term. This is the pressure work. This is the viscous dissipation. This is the volutric heat generation. So what I do here is this is the viscous dissipation term. Okay. And the pressure term if you put that you get dp by dt u dp by dx + v dp by dy + 5 q q dotle prime. For steady flows you don't have d by dt. So let me start off writing the energy equation for steady flows and nondimensionalize that equation. So let me start with steady flows energy equation for steady flows. Okay. So the energy equation would be row cp u do t by dx plus v do t by dy which is the convection term which is left hand side is equal to conduction term k into do ² t by dx² + d ² t by d y² this is the conduction term and the pressure Work is U del P by del X plus V del P by del Y plus 5 + Q dot triple prime. Five is the viscous dissipation term which is essentially the velocity gradients that is phi = 2 mu into del u by del x² + 2 mu into delv by del y² minus of 2x3 mu of d u by d x + del v by del y² ² + mu into d u by d y + del v by del x² now to non-dimensionalize I take xar = x by l y* = y by l and u* = u by u infinity l u infinity I have already defined v by u infinity And P star is P by row U infinity squared. And T star is what is new for us. T minus of T S divided by T infinity minus T S. T S is the surface temperature which is assumed to be constant. Surface temperature of the flat plate. Let us say surface temperature of the plate or a pipe. Okay. T infinity is free stream temperature. Restall we have already defined. So if I substitute this now what is that I get? So first let us do for row CP. Row CP U. If you do you get U infinity by L. That is you have U infinity U u* into U infinity and do YAR this is sorry this is X do XAR into L and now do T star. So what do you get? do tar is t infinity minus s and t ss is constant. So do ts by do x is not existing because ts is constant. So that's what you get. Similarly you get var for v star also you can pull out the same terms do tar by do y* is equal to k into t infinity minus t s you get when you do for do tar okay divided by for x you get x do xar squar by l² plus do ^ 2 tar divided by d y* squar plus plus okay plus you have u for u you have to write u infinity into u star divided by do xar that is l for p star p star do p star into row u infinity squared Similarly for plus V star do P star by do Y star. Now for viscous dissipation term. So let us take up the viscous dissipation term. The viscous dissipation term. The first term let us say you can write I think I will write directly. You can see that this this is the viscous dissipation term. So you have plus mu is common everywhere. Mu mu mu mu. So I pull out that mu. And you have velocity squared that means u infinity squared. And you have x² that means u infinity squar by l² that's there velocity squar x² velocity squared. So you get velocity squared everywhere. So you can write this as five star row u infinity squar by l² l. So now l². So now what you do is you divide throughout. You divide through out with this term. Okay. Let us take up one by one and divide. So if I take up this term that is k into t infinity minus t s divided by l² into row cp u infinity into t infinity minus t s by l. So t row k t t t t t t t t t t t t t t t t t t t t t t t t t t t t t t t t t t t t t t t t infinity minus t S T infinity minus T S. So I have 1 by row CP U infinity L coming here in this term. Similarly, what do I get for row U infinity squared? that is row u infinity² into u infinity by l into l by row cp into t infinity minus t s into u infinity so what is that l gets canceled out 1 u infinity 1 u infinity gets canceled out row and row gets canceled out so you get u infinity squ divid by CP into T infinity minus T S. So for this term you got this and for this term for this term and this term you got this. So let us write those term with those terms this equation. So what I get here is u*ar u* do t star divided by do xar plus v star do tar by d yar is equal to k by row cp l u infinity into do ^ 2 tar divided by do xar squared plus do 2 tar divided by do y* squar plus u infinity squar divided by cp into t infinity minus t s into U do P star by do XAR plus V do P star by do Y star plus U infinity squared divided by CP into T infinity minus T S okay into mu divided by row U infinity L. Okay, that's what you get here. Yeah, this term I didn't do. So you have mu u infinity squar by l² if you divide it by row row cp u infinity into t infinity minus t s by l. So u infinity 1 u infinity gets canceled out. L and L gets canceled out. So you get mu by row cp l t infinity by t s that is what I have written here. U infinity squared that is you get mu infinity mu that is mu u infinity squar divided by l² into l divided by row cp into u row cp into t infinity minus t s into u infinity so that I have written retained that this is what I have put U infinity squared by CP into T infinity S I have put this term and the left term is mu by row U infinity L I have put here into into five star now this if you see this K by this is nothing but E R so that is E pr equal to row U infinity L by mu into pantal number is mu cp by k. So that is you get this as row cp mu mu gets canceled out row cp l u infinity by k that is nothing but this in 1 by r. So now u infinity squared by cp into t infinity minus t s is called yakert number. So what does this represent? It represents the u infinity squared means it is kinetic energy. The kinetic energy with which my fluid particle is moving. and CP T infinity minus T S is the thermal energy. So the echert number represents how much of the kinetic energy is transformed into thermal energy when it is moving. So I can now write this equation as u* deltar by del xar plus v* del tar by del y* is equal to 1 by r e p r into del² tar divided by d xar squared plus d² tar divide divided by del y* squared plus yurt number into this is star u* del p* by del xar plus var del p* by del y* plus this is echert number and this is 1 by e so echert by E 5 star. So from the energy equation the non-dimensional numbers which are emerging are R E P R Y number. So the takeaway from this class is that if you nondimensionalize momentum equation the non-dimensional number which emerged is Reol's number. So the velocity profile the velocity gradients the pressure gradients the pressure profile is dependent on Reol's number. But when you non-dimensionalize energy equation the non-dimensional numbers which emerged are R E P R Yakert number okay where R E is viscous sorry inertia force divided by viscous force viscous force that is row U infinity L by mu. But on the other hand, pantal number is momentum diffusivity divided by thermal diffusivity that is muc by K. And yakert number. Yakert number is U infinity² divided by CP into T infinity minus T S that is kinetic energy divided by thermal energy that means my temperature distribution is not only dependent on reol number it is also dependent on number and Yakerta number so this is the takeaway so we are going In the next class we are going to represent that the friction factor is a function of reol number only. Here nasalt number is a function is a function of not only reol number but also parental number and a correct number. So we will study this in detail. So much for today's lecture. In the next class, let us derive or do what is called scale analysis and try to get much more understanding on boundary layer thicknesses and the friction factor and nusled number. Thank you. [music] >> [music] [bell]