Complete ANSYS Fluent Tutorial: NACA 2412 Airfoil Simulation and Analysis

hey everyone and welcome to this tutorial for a NACA 2412 airfoil in ansys fluent we're going to go through

the full steps of creating the geometry the mesh on setting up the simulation in ansys fluent I'm going to do this

simulation in answer student so that anyone watching this video can conduct the simulation at home by simply

downloading answer student and conducting the simulation or by doing it on the full version also

in this tutorial we're going to calculate the drag coefficient the lift coefficient on the airfoil we're going

to also plot the velocity Contours the pressure Contours and the velocity vectors of the flow field

we'll also then change the angle of attack and show how that affects how we model the simulation in answers fluent

and we'll also show how varying the angle of attack will change the flow fields surrounding the airfoil so go

ahead and load up answers on your computer and let's get started with this tutorial so the first thing we want to

do here is drag an unsus fluid analysis system into the project schematic like so

now the first step is geometry so we need to go ahead and get the knockout 20012 airfoil curve and import that into

design modeler so how we do that is simply to look up on Google um airfoil

plotter naca2412 so here's our airflow

um we'll leave all these digits the same this basically is identifying that this is the naca2412 curve shape that we want

and a number of points here we're going to change this to 200 because we get the choice between 200 uh and 20 points so

we're obviously going to pick 200 as it's going to be just a little bit more accurate and also we're going to close

the trailing edge here because we don't want to Hull at the trailing edge here and in our curve as we're going to get

errors in our simulation then I'm going to go ahead and click plot here

and send to airflow plotter and one last change I'm going to make here is to change the chord to be 1000 millimeters

which is one meter as this is the standard size for airfoils and it'll make calculating the Reynolds number and

velocity just a little bit easier and so once that's done I'm going to click plot again

and my earphones ready to go and I'm going to click csv's file of coordinates so this will give us an Excel file that

will edit a small bit before we can import into design modeler so I'll download that file and then open it

okay now once I downloaded this file there's a few edits we need to make um to the file so we can actually import

this into design model so the first thing I want to do is get rid of all this stuff at the top here as there's no

need for it so I'm going to click this top row hold hold Ctrl and come all the way down sorry hold shift

all the way down to this row and I just want to delete all them rows there I'm going to insert two rows here one and

two on the left I'm going to change this left very left column here to be hashtag group

I'll explain this shortly and this next one would be hashtag points #x hashtag y

and hashtag Z so hashtag group this is essentially um the I believe it's the type of Point

um all I know is that we're gonna put in here the 0.1 for all of the coordinates here and that seems to work when we

implement it into ansys um so that's what we put in here for group number and then for Point number

uh we're gonna put in one two three four and all the way down until 200. so how I do that how I speed up this process here

is to do equals click the row above and then click plus one and then I'm going to drag this formula here so this little

green square here I'm going to drag that down all the way down to my 200 and first point

all the way down to here I'll remove this in a minute we don't need that camera line

and lastly then the Z coordinate and even though we are doing a 2d simulation uh we will just put in here Z

equals zero and as answers expects it in the file that we're going to upload so put in zero here drag that down all the

way to the end now we don't need any of this stuff here the camera line I think there's some

more stuff at the bottom comb line and chord line we don't need that delete that and one last change you

want to do here so you can see at the end here basically the point start to repeat itself and it goes back to 100

which is the same as the top coordinate because it's trying to basically indicate that it wants to close the

trailing Edge and but how we actually do that in ansys um design model is we remove this point

here like so I change this to one zero and then it will close the trailing edge of

the airfoil so we have a fully closed curve and that's everything we need to do here and to upload this file into

design modeler so I'm going to save this in my downloads as a very bad practice but nonetheless just for video purposes

so I'm going to call it naca2412 and I'm just going to call it a video because I have this file done before

uh this needs to be a text file in tab delimited format just like so

and hit save yes that's okay and our file is now ready to go to be

imported into design model all right now that we've created the airflow curve in Excel to import into

design model we're ready to go ahead and create the full geometry so to do that I'm going to right click on Geometry

hit properties I'm just going to change the analysis type to 2D as we're just going to do the

2D airfoil today then right click geometry and load up design modeler

once design model is opened up I'm going to look directly at the x-load plane and go to concept 3D curve and here

we're going to import the curve we just created in Excel so I first need to change my coordinates to millimeters

that's what we had our coordinate file in I'm going to go to downloads and I'm

going to click the file that we created I'm going to import that and hit generate and now the R4 file is

created here so I'm just going to click on Zoom to fit just to take a proper look at it and all looks okay here

um so now what we want to do is we want to create a surface from this curve and then later on what we'll do is we'll

remove the airfoil surface from our fluid domain so to do that what I need to do is go to

concept go to services from edges click the curve I hit apply and hit generate and

what you'll see now is we now have a surface representing our airfoil okay so now I'm ready to go ahead and start

making the fluid domain around the airfoil and for this case I'm going to do a c-shape domain you can do an O

shape domain too where you just make the domain of fluid a circle around it but I'm going to make a c-shape domain as it

converges a little bit better so a c-check domain is going to have a c in front of the Leading Edge and then a

square behind the Leading Edge so to create that domain I'm going to go to sketching create

I'm going to create a semi-circle then in front of the Leading Edge so Arc by Center

I'm going to click here at the origin but I want to click When I See This P here as that's letting me know that I'm

directly locked on to the origin whereas the C just lets me know that I'm coincident to the x-axis so once I see

the P I click and now I want to create the Arc of the circle to be in line with the y-axis so I click once I see this C

here and the same again down here click once I see the C to let me know that I'm fully in line with my Axis

okay let's add some Dimensions here so I'm gonna make this domain about 15 quarts and

surrounding the airfoil as I found from previous studies that this is sufficient and give accurate results so I'd like it

to be 15 chords from here till here from one end of the semicircle to the other so that means the radius of this circle

would be seven and a half chords which would be seven and a half meters because you remember we made the airflow to be

one meter in chord length so I'm going to make this seven and a half meters as the radius of the circle

now I'm going to create the rest of the domain which is just a square part behind it so I'm going to click here we

want to see the P to be locked onto that point create a horizontal line indicated by that h

uh Dimension this line to be 15 meters which is 15 chords rearward of the Leading Edge

now I'm just going to fully close the domain and it's important here that I click

once I see this C here just so I don't have to add in an extra Dimension and that's my domain fully closed

now there are a few more sketches that I'm going to add in here and to The Domain and that will help me with my

meshing later on so there are two extra lines that I'm going to add in here and and this will you'll see later on I can

add sizings to the to these lines later on and in the matching software and it makes the meshing far more structured

and far easier to Define so I'd like a vertical line here um at the at the sorry at the trailing

edge of the airflow not the Leading Edge so I'm going to create a line just like this and it's not exactly at the

trailing Edge right now so I'm going to Dimension this to be one meter from I'm just going to delete that that's a

vertical Dimension I'm going to create this to be one meter from the origin just like this and once that line is

created I'm going to extend it to the end of the domains by clicking modify it extend

and just that'll extend it to the end of the fluid domain and now I want a vertical line near the

Leading Edge of the airfoil but not exactly so I've found that if I create a vertical line that's about 80

millimeters in the case for this airflow for the naca2412foil and this is the case that I found if you're using a

different airflow it may be different and but for this particular case I've found to create a vertical line at about

80 millimeters from the origin here and gives pretty good results and in terms of a nice mesh so this line here I'm

gonna do 0.08 meters which is 80 millimeters and from the the origin here and then

I'm just going to do the same extend this to the end of the domain okay and they're the lines that I need

now um to properly Define my mesh in a nice structured way so once I've created my

sketch here I'm going to turn this into actual lines so to do that I'm going to go to Concepts lines from sketches and

just select the sketch we just created hit apply and then click generate and you'll see we have actual lines now that

we can do some more um interesting things with and once we have lines we can turn this

into a surface so to do that now I can go to concept and go to services from edges and select

all the outer edges of my domain so be careful of this very small line here and while holding control I select all the

outer edges of the domain and then hit apply and you'll see we get a Surface here

that we're going to use to represent our fluid surface but if you notice if you look in the

tree that answers has created a solid surface and we want a fluid surface this is going to represent our full fluid

domain so I'm going to switch this here to fluid and then hit generate okay now that I have a fluid surface what I

want to do is I want to remove the airfoil surface we created at the beginning from the fluid domain so we

have an airfoil shaped hole in our fluid domain which will represent the solid airfoil so to do that I go to create

Boolean I'm going to do Boolean subtract the target body is the fluid surface

and the tool body is going to be the airfoil surface now to select the air full surface what we need to do is click

the airflow here but you'll see answers actually selects the fluid surface we just created we don't want that we want

to select the airflow in behind that so we need to click this plane here and you'll see it highlights the airfoil

surface then I hit apply and then click generate and you'll see we're now left with a

fluid surface here with an airfoil shaped hole in the domain and this hole is representing our solid airfoil of

course now there's one more thing I need to do before I can bring this to the meshing

software and what I need to do is I need to split the domain into different faces and from the sketches that we created to

help refine our mesh so to do that I'm going to go to tools face split

select this face here and we're going to split the face by the sketches that we created so select the

lines that we created earlier on these lines here and we're also going to select the x-axis here so that we have

six total faces hit generate these six total faces now in our domain

that we can mesh separately now there also is one more thing I need to do and I was getting errors with this

small edge here so I'm just going to combine it with this larger edge here and as it makes it far more easier to

mesh so to repair this Edge what I'm going to do is I'm going to use the edge selection filter hold Ctrl click this

edge here and this edge here go to tools merge

and you'll see we have merge type as edges and then hit apply and click generate and what you'll see is now

instead of having that small line and the the larger line here this is all just the one line now all the way up to

this point here I'm going to do the same thing for this lower half of the domain

I'm going to click on this edge here hold down control and then click this edge here

tools merge and apply and click generate and just to check that that's merged the two lines together when I click this

line here this is my full Edge now which is exactly what I wanted so I'm ready to go here to bring this to answer smashing

so now it's time to properly save the file so properly saving files and answers is

really important if I go to file save as and save my project somewhere I really don't want to just save this to my

desktop or my downloads folder or even into my documents folder because what's going to happen is ansys is going to

save some extra files and alongside this file and if we just save it into a folder all these files

might get lost the files that we need to actually open our project so what we need to do is to create a working

directory or in other words just create a folder specifically for this project so what I'm going to go is what I'm

going to do is I'm going to go to documents create a a new folder here I'm going to

call this Nokia 2412 video and then inside that folder I'm going to

create a the project file and I'm just going to call that the same thing naka2412

video now if I go to my documents and look at the folder I created here you'll see

that where I save my project file which is the file we're working off here there's these extra files in here so

this is what I meant when I said that answers is going to save some extra files and we need these extra files in

order to open our project so this is why it's important to create a proper working directory and a designated

folder for your project otherwise these files are just going to get either lost or it'll be dumped into a folder where

you don't want them for example on your desktop in your downloads and in your documents folder so we don't want that

so it's really important you create this folder for your project okay so once that's done we have our project saved

and we're ready to move on to the meshing step start meshing I'm just going to click

edit and this load up answers mechanical once ansys mechanical is loaded up here I'm just going to look at the XY plane

again and as you can see here we have a line body from the sketches we created I'm

just going to go ahead and suppress that body as we don't need it anymore as we've split our domain into different

faces so there's no need for that now I'm going to add some edge sizings here to the mesh and to do this I used a

there's another tutorial on YouTube from a guy named Anthony T and I'm going to use the sizings that he used in his

video as he managed to get a very structured mesh with these particular Edge sizings and I'd recommend after

this tutorial if you're still struggling to model an airflow to go and look at his tutorial because it's very good and

so yeah we're going to use them Edge sizings in this mesh and then we'll determine afterwards if the mesh is

suitable to this airfoil so I'm going to insert a sizing here select these edges here while holding

down the Ctrl key hit apply change this to number of divisions

250 divisions and then I'm going to add a bias here select the first option and create a

bias factor of 50 000 and that's going to basically bias give us more elements towards the airfoil than the outer edge

of the domain but you'll see here that some of the edges are actually flipped in the wrong direction where we have

bigger elements close to the airfoil and smaller at the edge of the domain so we're going to go ahead and reverse all

the edges that are in the wrong direction and to do that we click on reverse bias hold down control and click

all the edges that are in the wrong direction and then hit apply now we're going to add some more edge

sizings here I'm going to select all three edges here hit apply

the number of divisions set this as 150 set the behavior too hard

and then create a bias again with a bias factor of 300 now and I'm going to reverse these two edges here

also now you'll notice I set the behavior as hard here and what this basically does

is it inflicts on the software a very stringy condition that we want 150 divisions here no more and no less and

that's what we do when we have a very simple shape that we're trying to mesh such as this square but as as before we

used a soft meshing condition as we're meshing multiple different edges and basically when we use the hard condition

we get a very structured mesh so we're meshing very simple structures it's it's okay to use this hard condition and we

get a very structured mesh but when we're using the soft condition and we might get as well as structured mesh but

we're less likely to run into errors with the meshing so in any other cfd simulations that you might be doing if

you're meshing a very simple shape such as a square or a rectangle and we can often use the hard condition and get a

very structured mesh but if it's a structure that's not very simple and then often using the soft condition for

the um element size behavior is often a better

option as then otherwise you'd run into errors using the hard condition nonetheless let's continue on with our

edge sizings here so I'm going to select these four edges here top and bottom of the airfoil and bottom

of the domain hit apply number of Divisions 150 here and set the behavior as hard

and then lastly we're going to select these four edges here at the outer edge front of the airfoil on the bottom of

the domain hold Ctrl and hit apply and we're going to do 100 divisions here for these edges with behavior as hard also

last thing we need to do is to add a face sizing here to let us the software know we want a quadrahedral mesh instead

of triangular mesh and to select all the faces just press Ctrl a on your keyboard hold Ctrl and hit apply and then click

generate and we'll take a look at the mesh so once the mesh has loaded we can

expect inspect it a little bit just have a look at how it looks and to be honest it could be improved a

bit more if I were to do this again I would probably reduce the length um of this point here towards the

Leading Edge of the airfoil so less than the 80 millimeters that I used maybe to 40 as What's Happening Here is I'm

getting kind of a squished mesh here and the the volume is going to change a lot from each cell from this cell to the

next one which is not what we want and these elements are also pretty skewed which is fine because it's far away from

the mesh but yeah there could be some improvements made to this mesh for sure but for purposes purposes of this study

I'm going to leave it alone for now because um

I'm just going to go ahead and get a solution and show you how to run a simulation rather than focusing on

getting the mesh perfect because mesh generation itself really is an art so the one last thing that's really

important that we do need to check and besides for mesh quality is take a look at the Y plus value um at the air for a

while so what the Y plus is is it's a non-dimensional value that lets you know how much of your mesh is capturing the

viscous sub layer so in order to capture this non-linear behavior of the Velocity profile at the wall you need to have a y

plus of one and the reason for that as you can see is this is a y plus value of one here and then as you as we go along

here we get this non-linear behavior all the way so in order to capture this non-linear behavior of the Velocity at

the wall we need to ensure that our first cell height is at what's called a y plus of one and we can use a

calculator in order to figure out what our y plus value should be so in other words how big should the Cell at the

wall be in order to capture this non-linear behavior of the Velocity at the wall and this is going to be

important for this airfoil simulation where we're concerned about maybe the adverse pressure gradient of the wall

and looking at the separation length across the airfoil so it's important to have a y plus of less than one and where

wall effects are important but for other simulations and where wall effects are not important it might not be need to be

captured so nonetheless for this simulation we're going to aim for a y plus of less than one and it will

actually be an output of the simulation because it's a function of the Velocity profile and we don't know the velocity

until we solve the final simulation and but nonetheless we can use calculator to get an estimate of what our y plus is

before we start the simulation we can then go ahead and calculate our solution and then we can get the Y plus as an

output to see if we've kept a y plus of less than one in the majority of the mesh so to calculate the what value

um our cell height needs to be to get a y plus of less than one of course we need to pick a Reynolds number of the

flow and the Reynolds of where I'm picking for this case is compared to this

experimental data um from Abbott and I will leave this also linked in the description so you

can have a look at this experimental data and run some different simulations yourself but here's a Naka 2412 airfoil

with the lift coefficient versus angle of attack and the dry coefficient versus the lift coefficient and there's a list

of Reynolds number here so I just picked an arbitrary Reynolds number of the first one here of 3.1 by 10 to the six

and by using this Reynolds number in our simulation we can then compare later on our lift and dry coefficients and to

this experimental data but nonetheless now that I know my Reynolds number for the flow I can calculate what the

velocity the mainstream velocity of the flow is so by rearranging the formula for the Reynolds number and this is

wrong for a day of viscosity here and I actually use 1.802 e to the minus 5 because this is air at 288 Kelvin

and using these values I got a mainstream velocity of 45.6 meters per second now to calculate what value

um our first cell height needs to be to get a y plus of one I can input these values into a calculator

so I used 1.802 e minus 5 here for the viscosity our reference length is going to be our chord length which is

was one meter as you remember and we want a y plus of one I need to put in the mainstream velocity here

45.6 and it's going to should be much smaller than this so

this is our value in meters so I'm going to convert this to millimeters just so it's a little bit easier to see um in

ounces so meters to millimeters so we need to have 0.008 millimeters and as our first cell height thickness so

I'm gonna go to answers here and let's take a look at the airfoil we want to zoom into the wall and try and

get an estimate of the first layer height so zooming in right down to the band layer what I can

actually do is press this selection here to select a node and if I select this node here and this node here you can see

that the the distance from the wall of the first cell height is equal to 6.3 e to the minus three millimeters and the

Wall height that we need to have a y plus of one is equal to 0.008 which is 8.2 e to the zero e to the minus three

so this value here is less than what we need for a y plus of one so we know that we have a y plus of less than one in the

bench layer and we're going to capture all the boundary layer effects and in our simulation and but of course I

think it it really is important to state that the Y plus is an output of the simulation so actually the Y plus value

that we just we just got is an estimate and the Y plus value that we we need sorry the cell height that we need to

maintain a y plus of one will actually change over the wall so it is important that we check this after the simulation

and ensure that a y plus has remained mainly below one um all around the airfoil wall

one last thing we also need to do is we actually need to create our boundary conditions

um for fluent so what I'm going to do here is I'm going to zoom in on the airfoil

select the edges of the airfoil here press n on my keyboard or actually what I should do is right click

and press create name selection just in case you don't have that shortcut enabled

we're going to name this airfoil and that's going to be our wall condition and we're going to select these outer

edges here all outer edges create a name selection and we're just

going to create this as our Inlet and then finally the edge here is going to be our Outlet

and now we're almost ready to go what we're going to do is we're going to press update and what this is going to

do is it's going to send the mesh to fluent and what that does is it gives us a green tick here on our mesh letting us

know that everything is done it's fully saved it's gone to um fluent so

file close meshing and I'm going to save the project one more time just before we go to the fluent solver

now I'm going to click on setup right click and press edit and we're going to start fluent

I'm just going to enable double precision and I'm going to use four cores here because I know I have four

cores on my PC and that's the limit on the student version as the maximum number of cores I can use and if you

don't have more than four cores or four cores that's fine and it's just going to affect how fast you solve your

simulation so once fluent loads up here we're ready to get going with our simulation

so once everything is loaded up I'm gonna go to General

and we're going to leave pressure based on steady for this simulation let's go to models and I have the sstk Omega

um model selected for the airfoil and this is a great model for predicting airflow separation especially when you

have a y plus of less than one and this is generally recommended for airflow simulations so I'm going to continue

using this model and for this simulation so now I'm going to go to material properties here just double click

materials go to fluid and have a look at the properties that I have for air here so the density is okay but I believe I

had a different viscosity so I used 1.802 as my viscosity so I'll change that here

thank you and hit change create so now let's click on boundary

conditions have a look at the inlet we have a velocity Inlet here which is exactly what we want

and just for good practice I'm going to change the velocity to be in components and we want the x velocity here to be

what velocity we had calculated for our rounds number which was 45.6 so 45.6 for our x velocity and hit apply

let's have a look and make sure that the name selections we created are all okay so the airflow here so there's a wall

condition which is correct and the outlet is a pressure Outlet and the gauge pressure here is zero so

that basically means that the pressure at the outlet is zero pascals above atmospheric pressure and how I know that

is if I look at operating pressure I'm operating at atmospheric pressure and then the gauge pressure would be the

value of both our operating pressure which is zero here which is exactly what I want the outlet is essentially at

atmospheric pressure which is what I want I'm going to leave all the settings alone here

I'm going to go to reference values and I'm going to compute from the inlet and this is important because we're going to

be calculating drag and lift coefficients and we need the values to be the same as the values we use at the

inlet otherwise our drag and lift coefficients will not be calculated correctly

now I'm going to go to solution methods I'm going to leave all the default settings here as second order upwind and

using the coupled and pressure velocity scheme I'm going to go to residuals

and I'm going to reduce this down to one e to the minus 6 because we're running a very simple kind of simulation here so

I'm I'm perfectly fine with um instructing a more strict convergence

condition on the residuals here so then we're going to go to initialization

however initialization is fine and you can use standard if you want all this basically does is

um solving the final volume method is an iterative solution so we need to assign the cells an initial value or an initial

guess and so that the simulation can proceed to continue so it doesn't really matter which initialization scheme we

use it just affects the path to convergence so I'm just going to use hybrid here and

initialize the solution okay once it's initialized we have a few more things we need to set up here so

I'm going to basically set up the drag and lift coefficients and to be an output of the simulation so we can look

at how they're converging so to do that I go to right click on report definitions a new Force report and we'll

do drag we'll change this to drag coefficient

keep it as dry coefficient and we're looking for a drag coefficient on the airfoil force Vector leave that alone

we'll talk about that when we change the angle of attack and we want to print it to console and

repent print a report file to so that's all okay so that's our dry coefficient done and now we just need to create one

for the lift as well so we're going to new Force report lift

click on air foil lift coefficient and we want to put this console 2 and

then click ok double click on Rune calculation and

let's just do one thousand iterations and that should be plenty for the solution to converge so that's

everything we need to Define and we're ready to calculate the solution so once I click on calculate then I

should get a graphic window showing me the lift and the drag coefficients and how they're converging with the number

of iterations and I should also be able to see the residuals and how they're converging

and as I said this is a very simple simulation so there should not be too many problems and with convergence it

tends to be with unsteady simulations or simulations with heat transfer or something like that that we might have

issues with convergence and so I'm not expecting any errors and or any difficulty in convergence of this

simulation and you can see we have the lift and dry coefficients also being reported to the

console and we'll compare these final values to the experimental data from Abbott

so here's our dry coefficient our residuals and our lift coefficient and I will know that the solution is

converged is essentially when this the values just are not changing too much from one point to the next

um and you'll you'll notice that the residuals will keep decreasing all the

way down to the value that we've set them to usually and but the lift and the dry coefficient tend to converge a lot

sooner so for accurate lift and dry coefficients we might only need a couple hundred iterations before the value of

the lift and the drag don't change too much more so you can see that the lift and dry coefficients have not been

changing for a while and but the residuals are still going down so I'm going to leave it for a while

um you can see yep here's our dry coefficient here on the left sorry our lift coefficient left and our drag on

the right so I can go ahead and check this against the experimental data while this is rolling just to see if we're on

the right track so if I go to the experimental data here for a zero degree angle of attack which is what we have

and in our test case for and for a Reynolds number of 3.1 million if I go up along this line here

and have a look at the circular Dot zoom in so zero degree angle of attack it's

quite difficult to look at and go across it's somewhere in between 0.2 and 0.3 for sure probably about 0.25

for the section lift coefficient and what we have here is a lift coefficient of 0.218 so in terms of error and to

experimental data here you're looking at about 0.218 divided by 0.25 which is about 13 error which to be

honest is quite good especially considering the fact that we haven't done a mesh independent study we haven't

done a a domain study so we haven't vary the size of the domain to have a look at how that affects the results and

nonetheless this is cfd this is a steady simulation there's lots of assumptions embedded in here and to compare that to

experimental data and get only a 30 13 error and I think it's very very good and especially for cfd

now let's take a look at the dry coefficient so we have something like 0.009 and if I look at the

data on the right here for a lift coefficient of 0.2 would be about here and I scroll up

the circular dot here for the drag is out about if I go across here is that about the line below is

0.5.6.65 0.0065 for the dry coefficient and you can see there's a little bit more error

here on the dry coefficient so if I go to calculator and do 0.0065 divided by 0.009

let's round that up to seven yeah about a 30 error so that's more like what we will kind of expect from

cfd results and I'm I'm pretty okay with Dem results especially considering we have not

looked into the wall effects too much we haven't put in a friction coefficient or anything so

um I'm quite happy with having a a dry coefficient of about 30 error here and especially given that the lift

coefficient is quite good so these are you you probably could get more accurate values than the values

I've got in here but again that would take uh doing a mesh Independence study doing a domain size study maybe looking

at some some more of the wall effects and improving the mesh quality and taking a look at time dependent effects

and there's lots and lots of different things we could do here to improve our accuracy but with improving accuracy

takes improved effort improved computational costs and so there is trade-offs to trying to get your your

answers as close as you can to experimental data and not only that experimental data and publish such as

this is not the end-all be-all some experiments are different than others or

is tolerances involved in error so we we can't take the values here as set in stone also so you can see that getting

values as close as we can and within the range with reasonable values and getting simulations that follow the same Trend

as the experiments is more than what we're aiming for rather than getting exact numbers from our cfd if that makes

sense so the main value in cfd is we'd expect that as we increase the angle of attack we should get a line similar to

this with a similar slope we should get similar values for our lift and drag and our simulations should follow similar

Trends to our experiments and then we know that we're simulating um our simulations are accurate in the

sense that the representative of the experiments and then the beauty of cfd is then once we know

that our simulations are accurately represent in reality for this case we should expect to change flow conditions

and then still be getting accurate results so I should be able to change the velocity to 44 or 43 down to 40 and

for the most part I should be confident in my results that um I'm representing reality properly so I'm put doing an

experiment for example and it might not always be available to change quantities dot easily or change the geometry that

easily so this is why we use cfd and this is why it's valuable so once the simulation is done let's

take a look at our results so moving down here to the graphics we can create a velocity contour and just

take a look at how the velocity field is behaving so I go to Velocity Contour I'm just going to give this a name

changes to Velocity and velocity magnitude is fine for now and just hit save display and it should

display the velocity um on this symmetry plane here so you can see what we get is we get our

velocity slowing down and reaching a stagnation point at the front of the airfoil and then we can we can see the

effect of the boundary layer being represented and we have our no slip condition after a while so this looks

quite good and then we get a little bit of Separation here at the the not really separation more just the flow and having

a discontinuity let's say at the trailing edge of the airfoil and then that affects the flow all the way back

to the edge of the domain for the most part so this looks quite good and now let's

look at the pressure and this is quite interesting because it I'll be able to explain how an airflow works so let's go

to pressure Contour static pressure is fine and display

so how an airflow actually generates lifts and how airplanes actually fly is what we get is you'll see we have at the

lower end of the airfoil and we have a region of higher pressure and above the airflow we have a region of lower

pressure so what we actually get here when we have low pressure is we have the airfoil airfoils top surface being

pushed upwards uh as it's being sucked by the low pressure whereas below here as you can see the green Contour

represents a positive pressure we should expect to see the force pushing the airflow up below so the way an airflow

works is we get a kind of a soaking Force upwards and a pushing force below and that causes a lift Force now there

is regions of negative pressure here and that's because we have a zero degree angle of attack and what you'll notice

is as we increase uh we'll move now to a different angle of attack and you'll notice you'll get a much more

pronounced effect between the a lower pressure region on top of the airfoil and a high pressure region below that

creates a much bigger lift force on the airfoil um so yeah essentially that's how

airflow works and we'll explore that by changing the angle of attack shortly but one more thing I want to look at before

I move on is have a look at the velocity vectors just going to call this velocity

vector and do it by magnitude again

and further velocity magnitude in this case I had to change the scale here to 10

before I could see anything and click fluid Surface body so the velocity magnitude is scaled by its actual

magnitude so we we have to play around with the vectors A little bit to get the size we want

so to more accurately visualize this flow I will change the vectors A little bit more

let's go in here to Vector options and do fixed length and hit apply and Co dot changes it

Okay so now the vectors are too big so I'll change this scale again back down to one

hit save display still too big so let's try 0.1 yeah it's starting to get a little bit

smaller now so let's maybe change this to 0.05 so depending on what angle of attack and

what airflow you're running sometimes the velocity vectors takes a little bit of playing around with before

um they make any sort of sense and in terms of how they look so I'll actually change

this right the way down to 0.01 and now this is much more clear so you can see the flow

comes in and we get sort of a stagnation region not really recirculation here but just stagnation then the flow will go

over the airflow we get zero flow at the wall of course due to the no slip condition

and then the flow here just continues on so I'm happy enough with these results and now we'll move on to a different

angle of attack so when we change the angle of attack in this simulation what we need to do is

we're going to um enforce components on the flow so for

example um given the angle of attack we're going to change the velocity component such as

the x coordinate to be um the velocity times the COS cosine of the angle and that we're going to be

simulating and then for the Y velocity of the component which was originally zero when we had a zero degree angle of

attack we'll change that to be the velocity magnitude times the sine of the angle and that were that we're

simulating and also what we need to do is we need to change the lift and drag Force vectors because the lift in the

drag is always measured perpendicular and parallel to the relative flow Direction so you'll see the lift is

always 90 degrees to the flow Direction and the drag is always parallel to the flow Direction so whereas before when we

were um we were defining the lift and dry coefficients we left the velocity I we

left the force components the same we're going to actually need to change them now in this simulation when we change

the angle of the flow so if we want to simulate say an 8 degree angle of attack on this airflow what we do is we go to

the velocity section and where we had an x velocity of 45.6 our x velocity is now going to be 45.6 times the cosine of 8

degrees so if I go to on my calculator 45.6 times cost of 8 degrees I get 45.156 as the component of velocity and

the Y velocity upwards is going to be 45.6 times the sine of 8 degrees and that gives me

6.346 and if you were to do calculate the magnitude of the x velocity squared plus the Y velocity squared and this

should still equal to the 45.6 that we originally had for a velocity magnitude so hitting apply now the flow will be

tilted in an eighth grade direction towards the airflow and we also need to change the force

vectors of the drag and lift coefficient to be parallel with the flow so or perpendicular in the case of lift

so if I go to report definitions and go to my dry coefficient as this is the most simple one to do we want the

drag Force Vector to be parallel with the flow and we know we've tilted the flow now 8 degrees so the force Vector

that's parallel with the flow is going to be cosine of 8 degrees in the X Direction and sine of 8 degrees in the y

direction so cosine of 8 degrees I'm just going to type it into my calculator is 0.99

and the sine of 8 degrees is 0.139 and hit OK and what I'm

essentially doing here is just changing the force vectors rather than being one zero along this direction now that

we're tilting the flow we need to tilt the drag Force Vector with the flow and just by simple trigonometry you could

work this out yourself using some simple trigonometry you can work out that this drag Force Vector here and has

components cost 30 sine 30 and the lift is going to be in the X component of the flow it's going to be minus sine of the

angle and in the Y component it's going to be the cosine of the angle okay so minus sine of 8 degrees here in the X

direction for that Force Vector is going to be 0.139 and in the y direction we're going to

have a cosine of 8 degrees and it's going to be 0.99 for the force Vector for the lift

coefficient now if I reinitialize my solution hit okay

we should get a flow at 8 degrees here and our lifts and dry coefficients and should represent the lift and drag for

an 8 degree angle of attack if we have a look at the simulation data here or sorry the experimental data at

an 8 degree angle of attack we'd expect the airfoil to have let's scroll along here so I'm getting about

1.05 by looking at the experimental data here if this line here is 1.1 for Lift coefficient somewhere in between here

will be 1.05 is what we'd be expecting for our lift coefficient

on for our dry coefficient if we go along here and go to a section lift coefficient of 1.05 would be somewhere

about here move our Mouse up go to this circle here this is

representative random number of three million 3.1 million go across here we can calculate our expected drag expected

drag is somewhere about 0.11 0.115 ish somewhere in between point

0.011 and 0.012 is what we're expecting for our dry coefficient so these are expected answers from the simulation

so clicking Rune here we should be able to get our answers and compare.2d experimental data

so having a look at the results here even after about 80 iterations we can see our lift coefficient is almost

exactly what we predicted there of 1.05 which is really good and to be honest that's probably just by chance

um and looking at the dry coefficient we got about 0.0014 so if there's anything I'm

noticing from my simulations here is that my dry coffee my lift coefficient is pretty good but my dry coefficient

um appears to be predicted a bit higher by the um by the simulation so that's one Trend I can notice here by running

different angles of attack so now my solution has converged and let's take a look at the results so I'm

going to take a look at the pressure Contour and you can see here we are getting the

results I described before we get a positive pressure below the airfoil that's causing the pushing force upwards

of the airfoil and we get a low pressure region above the airfoil this is a negative pressure that's causing the

pulling of the airflow which and this is what's creating the lift on specifically what creates a higher lift for a higher

angle of attack it's due to this flow phenomenon here now we'll take a look at the velocity

Contour save display and you see we get similar velocity

profile but the flow is starting to kind of tilt in this direction here and it's essentially as you increase the angle of

attack and once it gets to a certain point the airflow will stall which is where it reaches its maximum lift

coefficient and then any increase of angle of attack above that point causes recirculation of the flow of both the

airfoil and essentially the flow separates and you don't get any increase in lift and it becomes very unstable

now let's take a look at the velocity vectors and it won't even make the vector just a

small even smaller again so we can properly visualize this here so let's make this 0.05

and let's investigate the flow so as you can see we get the flow accelerating around the airfoil here

and our bench layer is being accurately represented by the no slip condition here and

I'm happy enough with the results here so what I will do also is I forgot to show how we check our y plus our final y

plus value for the simulation so I go to plots go to XY plot and create new I'm going to call this y Plus versus

location and if I go to turbulence and wall y Plus

and it's going to do it in the direction of the vector 1 0 which is what we want do it on the fluid Surface body and then

sorry we're going to do this on the airfoil not the fluid Surface body hit save plot and we have to see the Y Plus

at all locations so you can see I'm at the Leading Edge of the airflow we get a y plus of about one just after

1.041 which is just above one so it's it's fine to be honest and then it decreases along the airfoil and this is

where it's really important along the boundary layer here is what's being represented here and it goes down to

about 0.0 .007 so it's it's below one at all

locations uh here for the most part except for the spike here and so I'm very very happy and with the wall y plus

here so that was the uh NACA 2412 airflow simulated in ansys fluent and with the

full geometry tutorial full meshing tutorial and the full results tutorial and we had a look at the effect of

changing the angle of attack um on the airfoil so I hope you enjoyed and leave me a comment below if you'd

like any more simulations for me to do I'm more than happy to take your suggestions into account and do a few I

rather do 2D simulations to be honest because um this simulation here is anyone that

watches the video is able to conduct a simulation um but I'm happy again to do some more

3D simulations the same as before but nonetheless just leave a comment below of um any simulations you'd like to see

me do and I'd be more than happy to do them and thanks for watching