Honestly, you taught me more than what I was able to learn in the last year at university. Thank you so much. Love your talks. I have watched em all atleast twice
Amazing! I am trying to make the talks like a resource that we can all jump back into when we need them. I sometimes forget things as well and have to look back at them 😅
Thank you so much for your well-organized slides and clear teaching! Your teaching style is to ask HOW before conveying a new idea, which makes me follow the main idea better.
awesome explanation for the modified pressure. I was confused about this and I always though we are applying static pressure. But, when I checked fluent guide and openfoam, there they mentioned regarding the modified pressure which I failed to understand. But, this video cleared out my doubts.
Great explanation about modified pressure bc. For the OpenFOAM users this is the p_rgh, which generates thousands of doubts at the CFD Online Forum The screen you appear on sometimes cover part of text on the slide.
At 17:39 M should be M>0.3 (1>M>0.3) as flow to Subsonic Compressible flow. Rest all of your videos are amazing. When I get doubts about working on CFD, I look into your videos!
I wanna take a moment and appreciate your work and the effort you put into creating these. Even if there is only a niche of viewers, in my case, your content has really helped me understand the concepts well. Keep doing more content of FM like the awesome ones that you have done!
This is the only TH-cam channel that I subscribed and clicked the bell icon. You're a great teacher. Love your content. I would really appreciate if you would do videos about following topics: 1- General talk about 'Large Eddy Simulations' 2- Smagorinsky-Lilly turbulence model 3- Vortex visualisation 4- Unsteady RANS I am an university student. My final project is about the efficiency of a wind turbine with leading-edge slat. I am trying to eliminate the tip losses.
These are some great suggestions, thanks Ekin. I am planning on doing the LES type models soon, starting with a general introduction on scale resolving simulations. I still have a bit of research left to do on the subject though before i can make these ones! As you are looking at tip vortices and tip losses for wind turbines, you might find my thesis useful for your studies (Tip Flow Corrections for Horizontal Axis Wind and Tidal Turbine Rotors). Best of luck with your studies!
Clear explanations, easy to understand, excellent work as always. Thank you Aidan! I would really appreciate if you do some videos on the programming theory or programming skills related to OpenFOAM.
Sir, thanking you so much because your videos are really outstanding and it's really valuable. We always need a person like you who teaches us at its best. Once again thank you sir
Wonderful explanation!!! It helps so many people. Hope you will provide more knowledge to many more. You have mentioned Subsonic compressible flow as (M
Hi, thanks for the video. Very useful and instructive. I just think it would have been better to use the term "total pressure" for p0, for the simple reason that "stagnation" mainly refers to a specific case of the total pressure, i.e. a point where the fluid is at rest (like at the leading edge). It's true, at the beginning of the computation the code assumes u=0 (at inlet), but at the next iteration this value is updated. Anyway, great videos, well explained!
Very nice explanation. Many thanks. If possible, I would appreciate some theory about periodic flows. I'm dealing with such case now (without HT), and would like to better understand the theory, especially for turbulent flows.
Your CFD lectures are very useful, especially for the learners. My question about the inlet pressure condition is that how do you specify the value of the stagnation pressure at the inlet which also is unknown if you do not know the inlet velocity?
You will have to do a quick hand calculation. If you take the velocity (that you know), calculate Mach number, then use the stagnation pressure equation to calculate the stagnation pressure from the static pressure (the static pressure will probably be atmospheric) 👍 this is also a useful way to check your CFD before you start running the case
You have to know something about your inlet condition ☺️ what information do you have? Mach number, static pressure, stagnation pressure, altitude, velocity, temperature, total temperature, mass flow rate?
@@fluidmechanics101 Got the idea!! Thank you very much for these videos!! I am new to the channel and I am enjoying every single episode. Keep it up! I would love to see more applications though 😊
Thank you for this great video! I have a question. At 9:10, let us say we use static pressure at the inlet but we still need the velocity to be specified as the boundary condition at the inlet. So, how velocity is calculated at the inlet for this case. You have very well explained that this case will diverge but to solve the equation and to get a diverged solution, we need to specify the velocity boundary condition at the inlet. Can you please explain? Just curious to know!
In this case, the velocity will initially take the value of your initial condition (the initialisation in Fluent or 0/U in OpenFOAM). The velocity is then updated from the solution of the momentum equations. Moving next to the pressure equation (see my video on PISO or simple algorithm) the pressure equation will struggle to reach a solution as it can't find a pressure field that is consistent with the velocity field you prescribed. Bit confusing but hopefully that helps!
Hi, I really appreciate everything you post in your channel. You videos are very informative and I've learned a lot from them. I would like to ask Could you please let me know how can I find the inlet boundary conditions and the exit boundary conditions , if I have only those inlet parameters incidence angle, Mach number and Reynolds number. Thank you.
Try working them out by hand. If you know your Reynolds number, Mach number and temperature (presumably at an altitude) then you should be able to work out the velocity, density, static pressure, total pressure, total temperature using a combination of the ideal gas law and isentropic relationships
This is all fascinating and highly instructive. Having tried to actually apply it, I have discovered that it is actually even more complicated than this, and I am ending up not actually using these formulas. Not exactly. The problem, as I am seeing it, is that in the equation for U in the subsonic compressible case, you need the temperature. If this is the stagnation temperature, no problem. Except that I've concluded it isn't. You need the static temperature, or it doesn't make sense. I think. And how do you get it? I can say all this because I've been consulting Blazek, Computational Fluid Dynamics, Principles and Applications. It has a section on boundary conditions, and there proposes a way of doing this calculation that is actually more efficient, I think, in terms of number of cpu operations, lines of code, etc, than you are proposing here.
I understand that it is the stagnation pressure that we specify for a pressure inlet, but what I still don't understand is how do we calculate that value when we don't know the velocity at the inlet? If p0 = p+0.5*rho*U^2, and I know rho and p but not U, how can I specify p0 at all? If I know U then I could calculate p0 but I might as well use a velocity inlet. How to solve this?
What type of problem are you solving? Is this compressible flow, natural convection or internal flow with loss coefficients? Most of the time I always go with velocity inlet + pressure outlet unless it is one of the above flows. What do you have?
Hello. I am mostly wondering about the theoretical aspect. But say we have the problem you show in the video with an open window feeding a fire inside a house. So we have a pressure outlet at atmospheric and a pressure inlet at atmospheric. If we set both to 101325 Pa then we have divergence according to the video. In order to specify stagnation pressure at inlet we must know U to calculate p0... but if we know U then why would we use a pressure inlet over a velocity inlet? I am wondering how we could ever calculate a numerical value of p0 for the BC without knowing U - and if we know U why not just use a velocity inlet? See the catch-22? Thanks!
Yes! The trick is to specify the pressure inlet with a stagnation pressure of 101325 Pa and the pressure outlet with a static pressure of 101325 Pa. The CFD code will initially guess zero velocity at the inlet but as the CFD solution evolves (because we specified the inlet stagnation pressure rathe than static) the velocity increases and the static pressure at the inlet will drop to below 101325 Pa. This is fine and will allow the CFD code to converge. Fluent allows you to do this very easily. A pressure-inlet with a gauge stagnation pressure of 0 and a pressure outlet with a static gauge pressure of 0 👍
@@fluidmechanics101 That's more intuitive. Your explanations are often really helpful for less experienced people like me. I have one more question: for a natural convection case involving a cube, what should the boundary conditions be for the 6 faces of the cube of {air region}? In this case, all the lateral faces should have pressure inlets, with the top face as a pressure outlet and the bottom face as adiabatic. (Please correct me if I'm mistaken.)
Thank you for the video. A small question: for pressure inlets in subsonic cases, the value of static pressure is obtained from the previous iteration. How is it obtained for the 1st iteration? Is it just guessed, and then corrected as part of the simple algorithm?
Clear and awesome explanation! I have a question to ask. For segregated flow solver, what is the difference between correcting the velocity on pressure boundary after solving the pressure correction equation and the method in this video? Are both equivalent? Thanks.
thanks a lot for the video. I am simulating flow through a box with fans inside. I will be using dynamic meshes to rotate the fans and the objective is to see how much flow rate the fans can generate at the outlet. Do you think pressure inlet in this case is an apt boundary condition? thanks again
Yes, pressure inlet and pressure outlet seems fine to me. You can always try with a coarse mesh first to check your boundary conditions and then switch to a fine mesh once you are happy
@10:13, u mentioned P-o as static pressure (in subtitles its showing) U r saying for static pressure it will take value from previous iteration, but what happens for the first iteration ? Is it same as taking values we initialized the whole domain ...?? for hydrostatic pressure case, as u mentioned CFD automatically accounts for hydrostatic pressure variation (rho*g*z) as it takes modified pressure p' for calculation. but if u see the expression p'=p-(rho*g*z), we are actually subtracting this (rho*g*z) value from static pressure. doesn't it mean we are neglecting the hydrostatic pressure variation ....?? correct me if my understanding is wrong. As usual ur interesting videos are great help for CFD beginners like me. Thank you
Thank you so much for all your great content. I found your video while looking into pressure inlets as I'm working on a case that I would like to have your opinion on if you may. I'm working on a centrifugal fan (blower) case where I'm trying to calculate the mass flowrate at its outlet coming out of the volute; we do not know any velocity, the only given we have is the rpm of the impeller. The blower is in an ambient room. I would like the steady-state case to simulate the suction effect of the rpm and to find the flowrate at the outlet. However, I do not have a value for total or stagnation pressure at the inlet; would it be physically correct in the model to add it as zero gauge pressure? since that the blower is in an ambient room. And would it be okay for the model to have both inlets and outlets as zero gauge pressure; one that is total at the inlet and static at the outlet.
Yep zero gauge for total pressure at the inlet and zero gauge static pressure at the outlet should be fine. The mass of rate will then be the output of the CFD simulation for the particular rpm that you choose. An easy way to check these types of confusing boundary conditions is to make a really coarse mesh first (say 5000 cells) which runs really quickly. That way you can check what the pressure and mass flow rates are doing. When you have the boundary conditions you want, just rerun the case with your full mesh
@@fluidmechanics101 Hello Sir, one question from side: if we are giving 0 gauge total pressure, isnt we are giving static head + velocity head as 0 instead of static head as 0 (as the domain is in atmosphere). i am trying to find this answer but no luck till now
@@himanshushrivastava7062 Yes it did solve my problem. And to answer the question you're asking above: Yes, this would mean that you are making the static + velocity = zero. But these boundary conditions are not constant ones; the RPM of the rotating part would then change this boundary condition anyways as the case starts to converge the velocity component in the total pressure would also change.
sir, I am learning ansys fluent recently and i have almost same problem that you showed at 6:50. There is one inlet(window), one outlet(chimney) and one heat source(fireplace) in the house. In this situation how can i set inlet & outlet boundary conditions? I have already tried out pressure total inlet = 0 and pressure static outlet = 0. But as expected, the inlet 'static' pressure was computed less than 0 (-) and backflow happened. it is logically right but doesn't make sense. How can i accurately simulate reality in this situation?
You may need to 'encourage' the flow to go in the right direction by applying a momentum source to the domain. You can turn on this momentum source to push the flow in the right direction initially and then turn it off as your simulation converges. However, depending on the geometry of your case there is no guarantee that you won't get any backflow at all. If the heat source is close to the window for example, you may get some reverse flow back out of the window. If your heat source is closer to the chimney you are more likely to get all the flow going in the 'correct' direction with no backflow. This is why traditionally we have chimneys installed directly above fireplaces in houses ☺️
@@fluidmechanics101 Thank you for replying! I see what you mean about backflow. Then b.c. pressure inlet(total) = 0 and pressure outlet(static) = 0 is correct? Is it natural that static pressure contour should decrease with increasing altitude? Because inlet total pressure is 0 inlet static pressure should be less than 0, and If the inlet static pressure is less than outlet static pressure, Isn't it impossible to achieve the static pressure distribution I mentioned? (decreasing with increasing height) The figure is almost same as your example.
Yes, this bit can be quite tricky. You need to check if your CFD code is solving for static pressure or static pressure without the hydrostatic component (I think OpenFOAM calls this p_rgh). Your intuitions are correct, that the hydrostatic pressure increases as you go down but it can be confusing when you are looking at static pressure changes due to losses and the hydrostatic pressure. I would recommend making a very simple example case (flow in a vertical box for example) and check what is going on using some simple hand calculations. Then you can go back to your real case when you know what you are looking at
@@fluidmechanics101 Thank you so much. By testing a simple vertical box model i could understand what the static pressure contour in ansys fluent actually means. You taught me more than my professor did. ☺️
hi, thank you for all the hard work! i have a question if you could help me figure it out.....that would be awesome! the problem i have is that if there is a room and on one of the room wall we have exhaust fans. now i want to study the ventilation of the room. one way is to actualy model the fan and then use sliding mesh method to make the air flow out of the room but that is very hard and time consuming instead what i want to do is that create outlet face on the domain wall and use some suitable boundary condition to mimic the behavior of an exhaust fan (ie move air from inside to outside) now, i want to figure out what boundary conditions should i use.... thank you!
Just use a 'velocity inlet' but specify the velocity components (U, V and W) as pointing out of the domain. This will act as a velocity outlet and pull the flow out of the domain, mimicing the effect of the ventilation fan. For your other boundaries, use pressure inlets. 👍
@@fluidmechanics101 Thank you very much! so, my model would be like this, use velocity inlet with outward vector direction to mimic exhaust fans and i have a door that is used for air inlet....just an opening (un restricted, un forced and open flow) should i use pressure inlet for this? Thank you for your help!
Yep perfect. You can have the open doors as a pressure inlet or opening 👍 pressure inlet should work fine but if it is being difficult to converge (if you get a lot of reversed flow) you can always switch to an opening
Your video are really very helpful. I have one question related to the boundary conditions only, can we use total pressure inlet and mass flow outlet type boundary conditions in case of unknown outlet static pressure? Also one request if possible then please can you make a one video on the micro gas turbine engine combustion chamber combustion analysis using Fluent or CFX. Just what kind of inlet and outlet boundary conditions we can apply and flamlet set-up.
You could try the mass flow outlet and total pressure inlet. I suspect it will probably be ok, as the static pressure difference between the inlet and outlet can still develop across the domain. Sadly I don't have much experience with combustion models, so I can't offer anything yet but I am looking into it!
@Fluid Mechanics 101 Thank you sir for your response. As you said the static pressure difference between the inlet and outlet can still develop across the domain but I'm getting constant pressure across the domain ( static equal to static). I have set the inlet total pressure which is comming from the compressor and the mass flow outlet as totat mass flow of air and fuel. To check the pressure loss inside the combustion chamber. But from the result, I have not found any difference in pressure. So, I confused where exactly I'm making mistakes wheather boundary conditions setup or something mistake in my geometry. Now, with your response I'm bit confidence about my boundary conditions setup So, I will check combustion chamber geometry.
It would also be worth a bit of detailed post processing. Like have a look at the forces on walls, heat losses, mass flow rates and see if you can work out what is going on. I am sure you will probably find a small mistake
This presentation is very helpful thank you. I have a question about BC for a forced convection study of a self ventilated electric motor cooling. Is it possible to only use "pressure-outlet" BC for the 6 faces of the enclosure surrounding my motor or do I have to set at least 1 face of the enclosure with a "pressure-inlet" BC ? In both case I want to fix the temperature at theses boundaries, will the result be the same ? Thank you
These simulations are always tricky to get the right boundary conditions. My recommendation would be to make a very coarse mesh and try out some different boundary conditions (so you can do this quickly). Once you have it working and you are happy, then run the same boundary conditions on your fine mesh
Thank you so much! Excellent explanation! I have a question about pressure inlet BC. In the scenario of bouyancy-driven flow showed in your video, the modified pressure is not constant across the inlet, because the streamlines will contract when it is drawn into the room from the ambient. In addition, there will also be pressure loss at the inlet due to that contraction. So i was wondering if there is any way of applying more accurate boundary condition at the inlet.
Have you tried extending the geometry further upstream, so that the contraction is included as part of the CFD geometry? Often people will add large boxes around the inlet of a geometry so that you can capture the inlet losses more accurately.
Yes, you will need to load a profile for each of your inlet variables and then apply the profiles at your inlet boundary condition. Maybe have a look for a tutorial which shows parabolic flow at the inlet? You can then work out how to do a shear profile from this
This is a subtle point. Assuming that you have an inlet and outlet with equal area, the static pressure difference has to balance the total force applied to the domain exactly! You would have to ensure that your calculated pressure difference is 100% correct. In general this is not possible to do and the CFD solver will diverge if you try and run the simulation as it can't balance the total force with the pressure difference you have applied.
For supersonic compressible flow, doesn't specifying the static and stagnation pressure on the inlet equate to specifying the flow velocity? If so, why not just specify the flow velocity?
Often with compressible flow, engineers are more interested in Mach number than velocity (i.e flight Mach number). Static pressure comes from altitude, and we get stagnation pressure from an isentropic flow equation. So it is just convenience really
while applying stagnation BC in the inlet, how do i know the exact value of it, do we need to assume some sort of a velocity at the inlet, but that assumption will not be accurate, and my BC is therefore not accurate.
Yep, if you are doing aerodynamics you can calculate it from the Mach number. For general systems the stagnation pressure is equal to the static pressure far away from the inlet where the flow is stationary (reservoir pressure)
Thanks for the video! I have a question. What if I am trying to simulate a transonic flow past a wing at an angle of attack. Sure I can specify the Gauge total pressure at the inlet and a static pressure at the outlet since it is a subsonic case. But then the calculation of the velocity from the total pressure gives a velocity that is the magnitude and not the individual components. I have set a pressure far field boundary condition with the components of the velocity though. So will this ensure that the velocity components at the leading edge of my wing correspond to the desired angle of attack by using information from the far field or will it not?
Exactly what mathematical process shows the influence that the outlet static pressure has over the inlet static pressure when the flow is subsonic? I can't find anything online for it.
Thank you for the video. When I tried pressure-driven flow, for which I imposed pressures (static pressure + dynamic pressure) at the inlet and oulet, the solution became divergent. To solve this problem, should I impose velocity condition? When I impose zero velocity at inlet, flow became convergent. Or should I impose only static pressure at outlet, not stagnation pressure (static pressure + dynamic pressure)? A result I want is that flow which is purely driven by pressure, not velocity.
Your divergence could be from a number of different things (mesh, timestep, boundary conditions, source terms etc). Have you done all your checking and you are sure it is due to the boundary condition? It is a good idea to check that you can get a stable solution with the simpler case (velocity inlet and pressure outlet) and then switch to pressure inlet and pressure outlet if that converges
Could you please elaborate how the outlet pressure p propagates upstream, from what I understand -- the atmosphreric pressure (static pressure=0) applied at the outlet would propogate back as ripple? Further, is this also the pressure which is used in case of calculating velocity at the inlet when we give stagnation pressure as inlet condition?
Really helpful video as always! I have a question that is somewhat related to this topic. Why is it more appropriate to apply stagnation pressure to fluid entering the domain (either through an inlet or backflow through an outlet) but to apply static pressure to fluid leaving a domain (either through an outlet or backflow through at an inlet)? I am trying to understand why recirculating flow is treated differently at an inlet or outlet boundary compared to the rest of the flow. A get the impression that the reasoning goes beyond numerical stability.
Thank you again for your high quality video. I have a question about one of your slide. when you said that "The computed Velocity field (U) has to balance the pressure gradient exactly" and that "any numerical errors will lead to divergence", I am not sure to clearly understand this mechanism of divergence and how it happens. Could you provide some details about it or at least refered me to some litterature references that could clarify this point? Thanks !
Sadly I don't have any references for this. The way to think about it is with a flow in a box with one inlet and one outlet with equal area. Taking a control volume around the box, the static pressure gradient has to balance the sum of the forces on the box. As we are specifying the static pressure at the inlet and the outlet, we are specifying the static pressure gradient. This has to balance the sum of the forces on the box (skin friction and pressure drag). But we don't know the sum of these forces yet, and so our guess for the static pressure gradient is likely to be wrong! If it is wrong then the code will never be able to converge to a solution. I realise this is complicated but I think a bit of control volume analysis might help you out 😃
Dear Aidan Could you please give a lecture or introduce a reference about propeller CFD analysis? There is some tutorials about it in TH-cam but i guess most of them are wrong at BCs, zone interactions, mesh and other part of setup.
Ah ok then. I would go with sliding mesh approach. CFX tends to be easier to setup for turbo machinery, so i would use this. Have you made a good mesh? What is the y+ on the propeller surfaces?
Thats great. Have a go at running the CFD simulation now and see if you can get a result. Then go back to your mesh and refine it to bring y+ down 👍 In CFX you can just use ‘reload mesh files’ so you wont need to setup the case again. You can just read in the new mesh and run it again 😄
one question: as you shown in picture of house with chimney, we dont know the inlet velocity. so how do we calculate stagnant pressure to give as input (we get stagnant pressure by adding static and velocity head). as I am understanding, we have inlet and out static pressure i.e. 0 Pa as the domain is open to atmosphere.
Heeyy, good explanation!! I wanna ask u sth, if the pressure inlet was set as zero, but in operating condition set as 101325. how does the simulation works? TIA
That should be fine. The CFD code will calculate the pressure relative to 101325pa. So you will see pressures around 101325Pa (the solution) but internally the CFD will subtract 101325 and calculate the pressures around 0Pa to get the extra decimal places
The Navier-Stokes equation uses the total pressure term, but in OpenFOAM boundary conditions, we specify the static pressure. Can you correct my misconception about this?
It was extremely helpful. As shown in the video it's easy to do this in ANSYS Fluent, just set pressure-inlet and pressure-outlet for inlet and outlet . I have a question about how to implement these boundary conditions for incompressible flow in OpenFOAM, I know for p i can set inlet with totalPressure and outlet with fixedValue, but how to set the boundary conditions of U, are pressureInletVelocity or pressureInletOutletVelocity for inlet and zeroGradient or inletOutlet for outlet correct ? and I'm also confused about the 'value' key word in 'pressureInletVelocity' and 'pressureInletOutletVelocity' boundary condition in OpenFOAM, does this mean the CFD code initial guess value for U of the inlet?
Yes you are correct. OpenFOAM lets you input two values, one is the boundary value and the other is the initial condition for that boundary. I think pressureInletVelocity or pressureInletOutletVelocity would work (although I am a bit rusty with OpemFOAM) so you should probably check the documentation. An easy way to check if your boundary conditions are working is to make a really simple case (a few hundreds cells in the mesh) and check the boundary conditions one by one. Then use the working boundary conditions on your real case
@@fluidmechanics101 Thanks！I'm a beginner of CFD，your courses on Udemy are excellent which give me a lot of help. Hopefully you'll have videos on periodic boundary condition in the future，there seems to be another 'modified pressure' too.
Hi. Can you guide me on how to use periodic boundary conditions in two directions simultaneously in Fluent? One set of PBs is having an associated pressure drop while the other doesn't.
Define both sets as periodic boundary conditions in fluent as normal. You then need to apply ‘periodic conditions’ to the pair of boundaries that you want the pressure drop. Have a look for periodic conditions in the fluent manual and you should find your answer 👍
@@fluidmechanics101 Hi. Thanks for the prompt response. I already tried to do it, however, Fluent considers the pressure drop condition common for both the PBs. I do not know how to apply pressure drop separately to a single PB condition. I couldn't find the answer in the Fluent manual.
Can you please comment on overdefinition of boundary conditions? In an incompressible pipe flow problem, is it bad to define pressure inlet and mass flow rate outlet BCs? If I define a velocity inlet and a pressure outlet, I am getting the expected mass flow rate, although the pressure at the inlet has changed. But if I define a mass flow rate outlet and a velocity or pressure inlet, the pressure at the outlet is not as expected. I am breaking my head over this...
If you have a mass flow rate outlet, you are setting the flow rate, so you want your other boundaries to be pressure inlets. Otherwise if you have a velocity inlet and a mass flow rate outlet your boundary conditions are overly defined. Once you have solved the calculation you can read off the pressure drop as the difference between the inlet and outlet pressure. Remember: your pressure drop may not be as you expect because you have additional losses in the system. This is why you are doing the CFD, the CFD calculates these losses for you 👍
@@fluidmechanics101 Thank you for the quick reply. What if I define a velocity inlet and a pressure outlet? Is that badly defined? I want my results to match the expected results because I am verifying my model with CFD simulation data from literature.
@@fluidmechanics101 If I define pressure inlet and mass flow rate outlet boundary conditions, I am able to match the results in the literature up to a certain velocity (
Depending on your geometry you may have some flow bypassing your geometry. It is difficult to say without looking at the geometry in detail. Also, yes velocity inlet and pressure outlet is also fine. Just use whichever is easiest and gives better convergence
Good Evening, In the same thing i am using a udf, for sinusoidal input but it is not working correctly. Could you help me please. Ur vedios have much more knowledge than the textbooks of cfd i have read till now. Thank you !
at 10:08 you say 'static pressure' twice, I guess that's a slip of the tongue, correctly p0 is the _stagnation_ pressure and p is the static pressure (as it is written in the first point of the slide)
Thank you for the video! is there any chance that you can make a video about setting backflow condition at the outlet? I tried to find one for OpenFoAM but I cant understand which boundary conditions to use for p, p_rgh, u, T, alphat, k, epsilon, nut. I can't be sure which BC to use as I don't understand the theory behind it. Great work anyway!
Hi Harish, this will probably be one of the boundary conditions in your model (either the inlet or the outlet). As I don't know what model you are solving I can't really say any further!
@@fluidmechanics101 Sir, i am solving partial arc bearing under laminar case & I know the supply pressure of the lubricant. So, in this case, can I use pressure inlet boundary conditions. But I don't know how to use supply in that case
I do a lot of reading and research to find the information (textbooks, source code, internet searches, looking at the forums) and then try and derive things by hand for myself. I then spend a while putting all the information together so that everyone can understand it. So there isnt really a source, i do a lot of the work myself 😅
Hi ! I really like those videos, really good explanations. Sometimes I wonder why fluent documentation is not made like that. I have a question concerning pressure inlet BC : in "thermal" I can specify a "total temperature". If I understand Fluent vocabulary well, it means that I can specify a stagnation temperature. I do not really understand what is the point of specifying one and how it can change the velocity calculation since only static temperature is involved in velocity calculation ( in equation 13). So I wonder : Does that total temperature have any role in changing the velocity being calculated after specifying Po and P ? Thanks !
When simulating a flow bench test (incompressible), where you don't know the flow rate of the system and want to find it, is it correct to apply a negative static pressure at the outlet and a 0 Total pressure at the inlet?
Yep, you could do it that way. For an incompressible flow, all that matters is the pressure difference between the inlet and outlet. As long as you have the correct pressure difference, you will get the flow rate you are looking for 👍
@@fluidmechanics101 But how do we guess the first stagnation pressure ? I have a diameter of 100mm fan rotating at 1000 rpm. I will try to calculate its one wing's downwash angle and velocity and add that velocity as a dynamic pressure to static pressure. So that will be my guessed stagnation pressure. Am I in the right way ?
Have you had a look at OpenFOAM? You can do a lot of CFD coding with the OpenFOAM framework. I would recommend having a look on their website and giving the tutorials a go 😊
Hello. I have studied fluid mechanics most of my career but it is been since months now that I am approaching problems with numerical methods. I mean I know the whereabouts of most the common methods (FEM, FVM, FDM, Spectral Methods) but I have been mainly a user of packages such as FLUENT or COMSOL. I would like to ask you, what is your take on Lattice Boltzmann Method for solving incompressible flows. I have read encouraging and discouraging opinions about its use. What would be your take regarding LBM for solving N-S? Thank you very much for your answer. Cheers.
![[CFD] How Fine should my CFD mesh be?](http://i.ytimg.com/vi/60fDz2cVdy8/mqdefault.jpg)
![[CFD] How Fine should my CFD mesh be?](/img/tr.png)
![[CFD] Hexcore Meshes for CFD](http://i.ytimg.com/vi/59xfM0ayFE0/mqdefault.jpg)
Honestly, you taught me more than what I was able to learn in the last year at university.
Thank you so much. Love your talks. I have watched em all atleast twice
Amazing! I am trying to make the talks like a resource that we can all jump back into when we need them. I sometimes forget things as well and have to look back at them 😅
Thank you so much for your well-organized slides and clear teaching!
Your teaching style is to ask HOW before conveying a new idea, which makes me follow the main idea better.
I am enjoying these episodes much more than Game of Thrones
... especially Season 8
LOL
That is absolutely true :)
People are lucky that these videos are free.
awesome explanation for the modified pressure. I was confused about this and I always though we are applying static pressure. But, when I checked fluent guide and openfoam, there they mentioned regarding the modified pressure which I failed to understand.
But, this video cleared out my doubts.
Great explanation about modified pressure bc. For the OpenFOAM users this is the p_rgh, which generates thousands of doubts at the CFD Online Forum
The screen you appear on sometimes cover part of text on the slide.
Yes! p_rgh is the modified pressure in OpenFOAM 😄
@@fluidmechanics101 Hope you may include OpenFOAM based samples also. Haha. But it really helps us, to navigate more in the foam directories.
At 17:39 M should be M>0.3 (1>M>0.3) as flow to Subsonic Compressible flow. Rest all of your videos are amazing. When I get doubts about working on CFD, I look into your videos!
This series just solved some of my OpenFOAM problems! Wutt! A talking forum.
Your explanation is very clear. I taught many things. Allah bless you. Thank you!!
Your way of explaining things is superb
I wanna take a moment and appreciate your work and the effort you put into creating these. Even if there is only a niche of viewers, in my case, your content has really helped me understand the concepts well. Keep doing more content of FM like the awesome ones that you have done!
This is the only TH-cam channel that I subscribed and clicked the bell icon. You're a great teacher. Love your content.
I would really appreciate if you would do videos about following topics:
1- General talk about 'Large Eddy Simulations'
2- Smagorinsky-Lilly turbulence model
3- Vortex visualisation
4- Unsteady RANS
I am an university student. My final project is about the efficiency of a wind turbine with leading-edge slat. I am trying to eliminate the tip losses.
These are some great suggestions, thanks Ekin. I am planning on doing the LES type models soon, starting with a general introduction on scale resolving simulations. I still have a bit of research left to do on the subject though before i can make these ones! As you are looking at tip vortices and tip losses for wind turbines, you might find my thesis useful for your studies (Tip Flow Corrections for Horizontal Axis Wind and Tidal Turbine Rotors). Best of luck with your studies!
What a beautiful talk! Amazing from start to end!
Excellent talk! Carry on mate! Hope to see a full lecture series on CFD!
Clear explanations, easy to understand, excellent work as always. Thank you Aidan! I would really appreciate if you do some videos on the programming theory or programming skills related to OpenFOAM.
Amazing! Thanks for all your effort doing such videos. Kudos!
You do a very good job with your videos. They explain things in a very understandable way.
Sir, thanking you so much because your videos are really outstanding and it's really valuable.
We always need a person like you who teaches us at its best.
Once again thank you sir
I am eternally gratefull for your guidance
Wonderful explanation!!! It helps so many people. Hope you will provide more knowledge to many more. You have mentioned Subsonic compressible flow as (M
Yep well spotted
Please continue to spread the knowledge !!
Excellent work Prof. Aidan I enjoy your pretty clear explanations. Amazing !
Brilliant as usual
You actually are amazing teacher, thanks a lot !!!!
Just Stumbled upon your videos while learning Fluent. Great Work. Could you please do a video on other common boundary conditions? Thank You!
excellent ,really appreciate your valuable knowledge and helping us improve rapidly. thanks a lot.
Thank you for a video from Belarus!
Thank you found this video very useful for my project.
Amazing!!! Thank you so much for this wonderful explanation. May your tribe increase!! I have liked and subscribed 👍
Thank you so much for this. Really good for understanding the basic concepts.
Thanks so much, Aidan. Very helpful.
Really great talk !
Enjoying your videos. Great Work!! 👏👏💕
This has been super helpful. Thank you!!
Amazing video as always!
If you ever get around to making more videos on compressible flow CFD, a video on non-reflecting BCs would be great!
Great suggestion! I will add it to my list
Crystal clear
Thanks for inspring videos, please keep going
Thank you so much for making these videos, it helped me a lot!
Great work, very interesting talk
Thank you very much! it helps me a lot!
you are a blessing man, thank you very much.
Excellent talk! Thank you for nice explanations.
Thank you for the lecture
a wonderful video, thanks you so much
Amazing job mate
It was very useful, thank you so much
Спасибо за видео !
Hi, thanks for the video. Very useful and instructive.
I just think it would have been better to use the term "total pressure" for p0, for the simple reason that "stagnation" mainly refers to a specific case of the total pressure, i.e. a point where the fluid is at rest (like at the leading edge). It's true, at the beginning of the computation the code assumes u=0 (at inlet), but at the next iteration this value is updated. Anyway, great videos, well explained!
Very nice explanation. Many thanks. If possible, I would appreciate some theory about periodic flows. I'm dealing with such case now (without HT), and would like to better understand the theory, especially for turbulent flows.
Your CFD lectures are very useful, especially for the learners. My question about the inlet pressure condition is that how do you specify the value of the stagnation pressure at the inlet which also is unknown if you do not know the inlet velocity?
You will have to do a quick hand calculation. If you take the velocity (that you know), calculate Mach number, then use the stagnation pressure equation to calculate the stagnation pressure from the static pressure (the static pressure will probably be atmospheric) 👍 this is also a useful way to check your CFD before you start running the case
Yes, but the velocity in unknown. Does that mean that we take an initial guess for the velocity?
You have to know something about your inlet condition ☺️ what information do you have? Mach number, static pressure, stagnation pressure, altitude, velocity, temperature, total temperature, mass flow rate?
@@fluidmechanics101 Got the idea!! Thank you very much for these videos!! I am new to the channel and I am enjoying every single episode. Keep it up! I would love to see more applications though 😊
Thank you for this great video! I have a question. At 9:10, let us say we use static pressure at the inlet but we still need the velocity to be specified as the boundary condition at the inlet. So, how velocity is calculated at the inlet for this case. You have very well explained that this case will diverge but to solve the equation and to get a diverged solution, we need to specify the velocity boundary condition at the inlet. Can you please explain? Just curious to know!
In this case, the velocity will initially take the value of your initial condition (the initialisation in Fluent or 0/U in OpenFOAM). The velocity is then updated from the solution of the momentum equations. Moving next to the pressure equation (see my video on PISO or simple algorithm) the pressure equation will struggle to reach a solution as it can't find a pressure field that is consistent with the velocity field you prescribed. Bit confusing but hopefully that helps!
@@fluidmechanics101 Thank you for the reply and yes, that helps! Thank you once again for your great effort!!
Hi, I really appreciate everything you post in your channel. You videos are very informative and I've learned a lot from them. I would like to ask
Could you please let me know how can I find the inlet boundary conditions and the exit boundary conditions , if I have only those inlet parameters incidence angle, Mach number and Reynolds number. Thank you.
Try working them out by hand. If you know your Reynolds number, Mach number and temperature (presumably at an altitude) then you should be able to work out the velocity, density, static pressure, total pressure, total temperature using a combination of the ideal gas law and isentropic relationships
This is all fascinating and highly instructive. Having tried to actually apply it, I have discovered that it is actually even more complicated than this, and I am ending up not actually using these formulas. Not exactly. The problem, as I am seeing it, is that in the equation for U in the subsonic compressible case, you need the temperature. If this is the stagnation temperature, no problem. Except that I've concluded it isn't. You need the static temperature, or it doesn't make sense. I think. And how do you get it? I can say all this because I've been consulting Blazek, Computational Fluid Dynamics, Principles and Applications. It has a section on boundary conditions, and there proposes a way of doing this calculation that is actually more efficient, I think, in terms of number of cpu operations, lines of code, etc, than you are proposing here.
Great sir
hope there is a video about transient analysis and courant no as it is very important topic
Yep, the Courant number video should be out in the next few weeks 😄 i am just finishing it off
I understand that it is the stagnation pressure that we specify for a pressure inlet, but what I still don't understand is how do we calculate that value when we don't know the velocity at the inlet? If p0 = p+0.5*rho*U^2, and I know rho and p but not U, how can I specify p0 at all? If I know U then I could calculate p0 but I might as well use a velocity inlet. How to solve this?
What type of problem are you solving? Is this compressible flow, natural convection or internal flow with loss coefficients? Most of the time I always go with velocity inlet + pressure outlet unless it is one of the above flows. What do you have?
Hello. I am mostly wondering about the theoretical aspect. But say we have the problem you show in the video with an open window feeding a fire inside a house. So we have a pressure outlet at atmospheric and a pressure inlet at atmospheric. If we set both to 101325 Pa then we have divergence according to the video. In order to specify stagnation pressure at inlet we must know U to calculate p0... but if we know U then why would we use a pressure inlet over a velocity inlet? I am wondering how we could ever calculate a numerical value of p0 for the BC without knowing U - and if we know U why not just use a velocity inlet? See the catch-22? Thanks!
Yes! The trick is to specify the pressure inlet with a stagnation pressure of 101325 Pa and the pressure outlet with a static pressure of 101325 Pa. The CFD code will initially guess zero velocity at the inlet but as the CFD solution evolves (because we specified the inlet stagnation pressure rathe than static) the velocity increases and the static pressure at the inlet will drop to below 101325 Pa. This is fine and will allow the CFD code to converge. Fluent allows you to do this very easily. A pressure-inlet with a gauge stagnation pressure of 0 and a pressure outlet with a static gauge pressure of 0 👍
@@fluidmechanics101 That's more intuitive. Your explanations are often really helpful for less experienced people like me. I have one more question: for a natural convection case involving a cube, what should the boundary conditions be for the 6 faces of the cube of {air region}?
In this case, all the lateral faces should have pressure inlets, with the top face as a pressure outlet and the bottom face as adiabatic. (Please correct me if I'm mistaken.)
Kindly make a video on aeroacoustics as well if possible using fluent. How we are going to set up the case and validate the case
Thank you for the video. A small question: for pressure inlets in subsonic cases, the value of static pressure is obtained from the previous iteration. How is it obtained for the 1st iteration? Is it just guessed, and then corrected as part of the simple algorithm?
Respected sir
Please make video on large eddy simulation your all videos are great and helped me lot in my project you are great
I am writing the LES video in the next few weeks! It should be a good one 😄
Clear and awesome explanation! I have a question to ask.
For segregated flow solver, what is the difference between correcting the velocity on pressure boundary after solving the pressure correction equation and the method in this video? Are both equivalent? Thanks.
thanks a lot for the video. I am simulating flow through a box with fans inside. I will be using dynamic meshes to rotate the fans and the objective is to see how much flow rate the fans can generate at the outlet. Do you think pressure inlet in this case is an apt boundary condition? thanks again
Yes, pressure inlet and pressure outlet seems fine to me. You can always try with a coarse mesh first to check your boundary conditions and then switch to a fine mesh once you are happy
@@fluidmechanics101 thank a lot :)
@10:13, u mentioned P-o as static pressure (in subtitles its showing)
U r saying for static pressure it will take value from previous iteration, but what happens for the first iteration ? Is it same as taking values we initialized the whole domain ...??
for hydrostatic pressure case, as u mentioned CFD automatically accounts for hydrostatic pressure variation (rho*g*z) as it takes modified pressure p' for calculation.
but if u see the expression p'=p-(rho*g*z), we are actually subtracting this (rho*g*z) value from static pressure. doesn't it mean we are neglecting the hydrostatic pressure variation ....??
correct me if my understanding is wrong.
As usual ur interesting videos are great help for CFD beginners like me.
Thank you
Excellent.. thanks
Thank you so much for all your great content. I found your video while looking into pressure inlets as I'm working on a case that I would like to have your opinion on if you may.
I'm working on a centrifugal fan (blower) case where I'm trying to calculate the mass flowrate at its outlet coming out of the volute; we do not know any velocity, the only given we have is the rpm of the impeller.
The blower is in an ambient room. I would like the steady-state case to simulate the suction effect of the rpm and to find the flowrate at the outlet. However, I do not have a value for total or stagnation pressure at the inlet; would it be physically correct in the model to add it as zero gauge pressure? since that the blower is in an ambient room. And would it be okay for the model to have both inlets and outlets as zero gauge pressure; one that is total at the inlet and static at the outlet.
Yep zero gauge for total pressure at the inlet and zero gauge static pressure at the outlet should be fine. The mass of rate will then be the output of the CFD simulation for the particular rpm that you choose.
An easy way to check these types of confusing boundary conditions is to make a really coarse mesh first (say 5000 cells) which runs really quickly. That way you can check what the pressure and mass flow rates are doing. When you have the boundary conditions you want, just rerun the case with your full mesh
@@fluidmechanics101 Thank you so much for replying I much appreciate it! I will definitely use that trick as well, thank you.
@@fluidmechanics101 Hello Sir, one question from side: if we are giving 0 gauge total pressure, isnt we are giving static head + velocity head as 0 instead of static head as 0 (as the domain is in atmosphere).
i am trying to find this answer but no luck till now
@@mostafaseif4409 Hi, I am doing the same thing as you. did it solve your problem?
@@himanshushrivastava7062 Yes it did solve my problem. And to answer the question you're asking above: Yes, this would mean that you are making the static + velocity = zero. But these boundary conditions are not constant ones; the RPM of the rotating part would then change this boundary condition anyways as the case starts to converge the velocity component in the total pressure would also change.
very interesting
sir, I am learning ansys fluent recently and i have almost same problem that you showed at 6:50. There is one inlet(window), one outlet(chimney) and one heat source(fireplace) in the house. In this situation how can i set inlet & outlet boundary conditions? I have already tried out pressure total inlet = 0 and pressure static outlet = 0. But as expected, the inlet 'static' pressure was computed less than 0 (-) and backflow happened. it is logically right but doesn't make sense. How can i accurately simulate reality in this situation?
You may need to 'encourage' the flow to go in the right direction by applying a momentum source to the domain. You can turn on this momentum source to push the flow in the right direction initially and then turn it off as your simulation converges. However, depending on the geometry of your case there is no guarantee that you won't get any backflow at all. If the heat source is close to the window for example, you may get some reverse flow back out of the window. If your heat source is closer to the chimney you are more likely to get all the flow going in the 'correct' direction with no backflow. This is why traditionally we have chimneys installed directly above fireplaces in houses ☺️
@@fluidmechanics101 Thank you for replying! I see what you mean about backflow. Then b.c. pressure inlet(total) = 0 and pressure outlet(static) = 0 is correct? Is it natural that static pressure contour should decrease with increasing altitude? Because inlet total pressure is 0 inlet static pressure should be less than 0, and If the inlet static pressure is less than outlet static pressure, Isn't it impossible to achieve the static pressure distribution I mentioned? (decreasing with increasing height) The figure is almost same as your example.
Yes, this bit can be quite tricky. You need to check if your CFD code is solving for static pressure or static pressure without the hydrostatic component (I think OpenFOAM calls this p_rgh). Your intuitions are correct, that the hydrostatic pressure increases as you go down but it can be confusing when you are looking at static pressure changes due to losses and the hydrostatic pressure. I would recommend making a very simple example case (flow in a vertical box for example) and check what is going on using some simple hand calculations. Then you can go back to your real case when you know what you are looking at
@@fluidmechanics101 Thank you so much. By testing a simple vertical box model i could understand what the static pressure contour in ansys fluent actually means. You taught me more than my professor did. ☺️
hi, thank you for all the hard work!
i have a question if you could help me figure it out.....that would be awesome! the problem i have is that if there is a room and on one of the room wall we have exhaust fans. now i want to study the ventilation of the room. one way is to actualy model the fan and then use sliding mesh method to make the air flow out of the room but that is very hard and time consuming instead what i want to do is that create outlet face on the domain wall and use some suitable boundary condition to mimic the behavior of an exhaust fan (ie move air from inside to outside) now, i want to figure out what boundary conditions should i use....
thank you!
Just use a 'velocity inlet' but specify the velocity components (U, V and W) as pointing out of the domain. This will act as a velocity outlet and pull the flow out of the domain, mimicing the effect of the ventilation fan. For your other boundaries, use pressure inlets. 👍
@@fluidmechanics101 Thank you very much! so, my model would be like this, use velocity inlet with outward vector direction to mimic exhaust fans and i have a door that is used for air inlet....just an opening (un restricted, un forced and open flow) should i use pressure inlet for this? Thank you for your help!
Yep perfect. You can have the open doors as a pressure inlet or opening 👍 pressure inlet should work fine but if it is being difficult to converge (if you get a lot of reversed flow) you can always switch to an opening
Sir, can you upload a video on exhaust fan and fan boundary condition
Your video are really very helpful. I have one question related to the boundary conditions only, can we use total pressure inlet and mass flow outlet type boundary conditions in case of unknown outlet static pressure? Also one request if possible then please can you make a one video on the micro gas turbine engine combustion chamber combustion analysis using Fluent or CFX. Just what kind of inlet and outlet boundary conditions we can apply and flamlet set-up.
You could try the mass flow outlet and total pressure inlet. I suspect it will probably be ok, as the static pressure difference between the inlet and outlet can still develop across the domain.
Sadly I don't have much experience with combustion models, so I can't offer anything yet but I am looking into it!
@Fluid Mechanics 101 Thank you sir for your response.
As you said the static pressure difference between the inlet and outlet can still develop across the domain but I'm getting constant pressure across the domain ( static equal to static).
I have set the inlet total pressure which is comming from the compressor and the mass flow outlet as totat mass flow of air and fuel. To check the pressure loss inside the combustion chamber. But from the result, I have not found any difference in pressure. So, I confused where exactly I'm making mistakes wheather boundary conditions setup or something mistake in my geometry. Now, with your response I'm bit confidence about my boundary conditions setup So, I will check combustion chamber geometry.
It would also be worth a bit of detailed post processing. Like have a look at the forces on walls, heat losses, mass flow rates and see if you can work out what is going on. I am sure you will probably find a small mistake
This presentation is very helpful thank you. I have a question about BC for a forced convection study of a self ventilated electric motor cooling. Is it possible to only use "pressure-outlet" BC for the 6 faces of the enclosure surrounding my motor or do I have to set at least 1 face of the enclosure with a "pressure-inlet" BC ? In both case I want to fix the temperature at theses boundaries, will the result be the same ? Thank you
These simulations are always tricky to get the right boundary conditions. My recommendation would be to make a very coarse mesh and try out some different boundary conditions (so you can do this quickly). Once you have it working and you are happy, then run the same boundary conditions on your fine mesh
Thank you so much! Excellent explanation! I have a question about pressure inlet BC. In the scenario of bouyancy-driven flow showed in your video, the modified pressure is not constant across the inlet, because the streamlines will contract when it is drawn into the room from the ambient. In addition, there will also be pressure loss at the inlet due to that contraction. So i was wondering if there is any way of applying more accurate boundary condition at the inlet.
Have you tried extending the geometry further upstream, so that the contraction is included as part of the CFD geometry? Often people will add large boxes around the inlet of a geometry so that you can capture the inlet losses more accurately.
@@fluidmechanics101 Thanks for your suggestion! I will try it.
Is it possible to apply a shear flow velocity profile at the inlet right away in ANSYS Fluent?
Yes, you will need to load a profile for each of your inlet variables and then apply the profiles at your inlet boundary condition. Maybe have a look for a tutorial which shows parabolic flow at the inlet? You can then work out how to do a shear profile from this
8:24: Why we can't specify the static pressure at inlet and outlet
This is a subtle point. Assuming that you have an inlet and outlet with equal area, the static pressure difference has to balance the total force applied to the domain exactly! You would have to ensure that your calculated pressure difference is 100% correct. In general this is not possible to do and the CFD solver will diverge if you try and run the simulation as it can't balance the total force with the pressure difference you have applied.
For supersonic compressible flow, doesn't specifying the static and stagnation pressure on the inlet equate to specifying the flow velocity? If so, why not just specify the flow velocity?
Often with compressible flow, engineers are more interested in Mach number than velocity (i.e flight Mach number). Static pressure comes from altitude, and we get stagnation pressure from an isentropic flow equation. So it is just convenience really
while applying stagnation BC in the inlet, how do i know the exact value of it, do we need to assume some sort of a velocity at the inlet, but that assumption will not be accurate, and my BC is therefore not accurate.
Yep, if you are doing aerodynamics you can calculate it from the Mach number. For general systems the stagnation pressure is equal to the static pressure far away from the inlet where the flow is stationary (reservoir pressure)
Thanks for the video! I have a question. What if I am trying to simulate a transonic flow past a wing at an angle of attack. Sure I can specify the Gauge total pressure at the inlet and a static pressure at the outlet since it is a subsonic case. But then the calculation of the velocity from the total pressure gives a velocity that is the magnitude and not the individual components. I have set a pressure far field boundary condition with the components of the velocity though. So will this ensure that the velocity components at the leading edge of my wing correspond to the desired angle of attack by using information from the far field or will it not?
Exactly what mathematical process shows the influence that the outlet static pressure has over the inlet static pressure when the flow is subsonic? I can't find anything online for it.
Thank you for the video.
When I tried pressure-driven flow, for which I imposed pressures (static pressure + dynamic pressure) at the inlet and oulet, the solution became divergent.
To solve this problem, should I impose velocity condition? When I impose zero velocity at inlet, flow became convergent.
Or should I impose only static pressure at outlet, not stagnation pressure (static pressure + dynamic pressure)?
A result I want is that flow which is purely driven by pressure, not velocity.
Your divergence could be from a number of different things (mesh, timestep, boundary conditions, source terms etc). Have you done all your checking and you are sure it is due to the boundary condition? It is a good idea to check that you can get a stable solution with the simpler case (velocity inlet and pressure outlet) and then switch to pressure inlet and pressure outlet if that converges
so, in the case of turbulent flow with incompressible flow, the inlet pressure can be also used instead of velocity inlet?
Yep, you can use whichever you want
@@fluidmechanics101 nice! thx a lot.
by the way, your videos are very useful!
Could you please elaborate how the outlet pressure p propagates upstream, from what I understand -- the atmosphreric pressure (static pressure=0) applied at the outlet would propogate back as ripple? Further, is this also the pressure which is used in case of calculating velocity at the inlet when we give stagnation pressure as inlet condition?
Really helpful video as always! I have a question that is somewhat related to this topic. Why is it more appropriate to apply stagnation pressure to fluid entering the domain (either through an inlet or backflow through an outlet) but to apply static pressure to fluid leaving a domain (either through an outlet or backflow through at an inlet)? I am trying to understand why recirculating flow is treated differently at an inlet or outlet boundary compared to the rest of the flow. A get the impression that the reasoning goes beyond numerical stability.
Thank you again for your high quality video. I have a question about one of your slide. when you said that "The computed Velocity field (U) has to balance the pressure gradient exactly" and that "any numerical errors will lead to divergence", I am not sure to clearly understand this mechanism of divergence and how it happens. Could you provide some details about it or at least refered me to some litterature references that could clarify this point? Thanks !
Sadly I don't have any references for this. The way to think about it is with a flow in a box with one inlet and one outlet with equal area. Taking a control volume around the box, the static pressure gradient has to balance the sum of the forces on the box. As we are specifying the static pressure at the inlet and the outlet, we are specifying the static pressure gradient. This has to balance the sum of the forces on the box (skin friction and pressure drag). But we don't know the sum of these forces yet, and so our guess for the static pressure gradient is likely to be wrong! If it is wrong then the code will never be able to converge to a solution. I realise this is complicated but I think a bit of control volume analysis might help you out 😃
@@fluidmechanics101 Thank you for your feedback !
Dear Aidan
Could you please give a lecture or introduce a reference about propeller CFD analysis?
There is some tutorials about it in TH-cam but i guess most of them are wrong at BCs, zone interactions, mesh and other part of setup.
Yep. This is quite a wide topic. What type of analysis are you trying to do? Sliding mesh or moving reference frame? Also ... what CFD code ?
@@fluidmechanics101
I,m working on it on Fluent & CFX.
And looking for best and more accurate type of analysis
Ah ok then. I would go with sliding mesh approach. CFX tends to be easier to setup for turbo machinery, so i would use this. Have you made a good mesh? What is the y+ on the propeller surfaces?
@@fluidmechanics101 just in first steps with low quality, y+ about 300!!
Thats great. Have a go at running the CFD simulation now and see if you can get a result. Then go back to your mesh and refine it to bring y+ down 👍 In CFX you can just use ‘reload mesh files’ so you wont need to setup the case again. You can just read in the new mesh and run it again 😄
one question: as you shown in picture of house with chimney, we dont know the inlet velocity. so how do we calculate stagnant pressure to give as input (we get stagnant pressure by adding static and velocity head). as I am understanding, we have inlet and out static pressure i.e. 0 Pa as the domain is open to atmosphere.
Both are 0Pa
Heeyy, good explanation!!
I wanna ask u sth, if the pressure inlet was set as zero, but in operating condition set as 101325. how does the simulation works? TIA
That should be fine. The CFD code will calculate the pressure relative to 101325pa. So you will see pressures around 101325Pa (the solution) but internally the CFD will subtract 101325 and calculate the pressures around 0Pa to get the extra decimal places
Okay thank you for your explanation
If you don't mind, i want to ask something about FSI. Can i put roughness in fluent ansys and static structural? TIA
The Navier-Stokes equation uses the total pressure term, but in OpenFOAM boundary conditions, we specify the static pressure. Can you correct my misconception about this?
It was extremely helpful. As shown in the video it's easy to do this in ANSYS Fluent, just set pressure-inlet and pressure-outlet for inlet and outlet . I have a question about how to implement these boundary conditions for incompressible flow in OpenFOAM, I know for p i can set inlet with totalPressure and outlet with fixedValue, but how to set the boundary conditions of U, are pressureInletVelocity or pressureInletOutletVelocity for inlet and zeroGradient or inletOutlet for outlet correct ? and I'm also confused about the 'value' key word in 'pressureInletVelocity' and 'pressureInletOutletVelocity' boundary condition in OpenFOAM, does this mean the CFD code initial guess value for U of the inlet?
Yes you are correct. OpenFOAM lets you input two values, one is the boundary value and the other is the initial condition for that boundary. I think pressureInletVelocity or pressureInletOutletVelocity would work (although I am a bit rusty with OpemFOAM) so you should probably check the documentation. An easy way to check if your boundary conditions are working is to make a really simple case (a few hundreds cells in the mesh) and check the boundary conditions one by one. Then use the working boundary conditions on your real case
@@fluidmechanics101 Thanks！I'm a beginner of CFD，your courses on Udemy are excellent which give me a lot of help. Hopefully you'll have videos on periodic boundary condition in the future，there seems to be another 'modified pressure' too.
Hi. Can you guide me on how to use periodic boundary conditions in two directions simultaneously in Fluent? One set of PBs is having an associated pressure drop while the other doesn't.
Define both sets as periodic boundary conditions in fluent as normal. You then need to apply ‘periodic conditions’ to the pair of boundaries that you want the pressure drop. Have a look for periodic conditions in the fluent manual and you should find your answer 👍
@@fluidmechanics101 Hi. Thanks for the prompt response. I already tried to do it, however, Fluent considers the pressure drop condition common for both the PBs. I do not know how to apply pressure drop separately to a single PB condition. I couldn't find the answer in the Fluent manual.
Probably best to drop an email to ansys customer support 👍
Is there a video which talks a bit more about those outlet conditions for supersonic flow?
Not yet 😅
@@fluidmechanics101 trying to do transonic simulation for my undergrad project, I will keep up the search lol. Thanks for the great content tho!
Can you please comment on overdefinition of boundary conditions? In an incompressible pipe flow problem, is it bad to define pressure inlet and mass flow rate outlet BCs? If I define a velocity inlet and a pressure outlet, I am getting the expected mass flow rate, although the pressure at the inlet has changed. But if I define a mass flow rate outlet and a velocity or pressure inlet, the pressure at the outlet is not as expected. I am breaking my head over this...
If you have a mass flow rate outlet, you are setting the flow rate, so you want your other boundaries to be pressure inlets. Otherwise if you have a velocity inlet and a mass flow rate outlet your boundary conditions are overly defined.
Once you have solved the calculation you can read off the pressure drop as the difference between the inlet and outlet pressure. Remember: your pressure drop may not be as you expect because you have additional losses in the system. This is why you are doing the CFD, the CFD calculates these losses for you 👍
@@fluidmechanics101 Thank you for the quick reply. What if I define a velocity inlet and a pressure outlet? Is that badly defined? I want my results to match the expected results because I am verifying my model with CFD simulation data from literature.
@@fluidmechanics101 If I define pressure inlet and mass flow rate outlet boundary conditions, I am able to match the results in the literature up to a certain velocity (
Depending on your geometry you may have some flow bypassing your geometry. It is difficult to say without looking at the geometry in detail. Also, yes velocity inlet and pressure outlet is also fine. Just use whichever is easiest and gives better convergence
Good Evening, In the same thing i am using a udf, for sinusoidal input but it is not working correctly. Could you help me please.
Ur vedios have much more knowledge than the textbooks of cfd i have read till now. Thank you !
at 10:08 you say 'static pressure' twice, I guess that's a slip of the tongue, correctly
p0 is the _stagnation_ pressure and p is the static pressure (as it is written in the first point of the slide)
Yep, I misspoke!
Thank you for the video! is there any chance that you can make a video about setting backflow condition at the outlet? I tried to find one for OpenFoAM but I cant understand which boundary conditions to use for p, p_rgh, u, T, alphat, k, epsilon, nut. I can't be sure which BC to use as I don't understand the theory behind it. Great work anyway!
can we take the instantaneous velocity calculated by CFD as a transient case , so that i can calculate the transient flow rate? Thanks
Great Work man!
Sir, If I know the supply pressure of the lubricant, then where I need to enter this supply pressure in Fluent ?
Hi Harish, this will probably be one of the boundary conditions in your model (either the inlet or the outlet). As I don't know what model you are solving I can't really say any further!
@@fluidmechanics101 Sir, i am solving partial arc bearing under laminar case & I know the supply pressure of the lubricant. So, in this case, can I use pressure inlet boundary conditions. But I don't know how to use supply in that case
Brother , can i know what are the resources you use to get these ?
I do a lot of reading and research to find the information (textbooks, source code, internet searches, looking at the forums) and then try and derive things by hand for myself. I then spend a while putting all the information together so that everyone can understand it. So there isnt really a source, i do a lot of the work myself 😅
Hi !
I really like those videos, really good explanations. Sometimes I wonder why fluent documentation is not made like that.
I have a question concerning pressure inlet BC : in "thermal" I can specify a "total temperature". If I understand Fluent vocabulary well, it means that I can specify a stagnation temperature. I do not really understand what is the point of specifying one and how it can change the velocity calculation since only static temperature is involved in velocity calculation ( in equation 13). So I wonder : Does that total temperature have any role in changing the velocity being calculated after specifying Po and P ?
Thanks !
When simulating a flow bench test (incompressible), where you don't know the flow rate of the system and want to find it, is it correct to apply a negative static pressure at the outlet and a 0 Total pressure at the inlet?
Yep, you could do it that way. For an incompressible flow, all that matters is the pressure difference between the inlet and outlet. As long as you have the correct pressure difference, you will get the flow rate you are looking for 👍
@@fluidmechanics101 But how do we guess the first stagnation pressure ? I have a diameter of 100mm fan rotating at 1000 rpm. I will try to calculate its one wing's downwash angle and velocity and add that velocity as a dynamic pressure to static pressure. So that will be my guessed stagnation pressure. Am I in the right way ?
Yes, that sounds sensible to me. If you have any additional pressure losses upstream you can always add these as well
Sir
I want to do coding in CFD . Can you suggest me some Book / Research Paper ?
Thanks for these incredible videos .
Have you had a look at OpenFOAM? You can do a lot of CFD coding with the OpenFOAM framework. I would recommend having a look on their website and giving the tutorials a go 😊
@@fluidmechanics101 I will sir . Thanks for giving me the direction .
Hello. I have studied fluid mechanics most of my career but it is been since months now that I am approaching problems with numerical methods. I mean I know the whereabouts of most the common methods (FEM, FVM, FDM, Spectral Methods) but I have been mainly a user of packages such as FLUENT or COMSOL. I would like to ask you, what is your take on Lattice Boltzmann Method for solving incompressible flows. I have read encouraging and discouraging opinions about its use. What would be your take regarding LBM for solving N-S? Thank you very much for your answer. Cheers.