Science topic

Computational Fluid Dynamics - Science topic

Numerical methods to solve and analyze problems that involve fluid flows.
Questions related to Computational Fluid Dynamics
Question
4 answers
I am doing research in CFD and currently using RK Gill. Can you suggest a better method?
Relevant answer
Answer
CFD is the simulation of fluids engineering systems using modeling and numerical methods.
CFD is a set of PARTIAL differential equations - PDE’s, describing conservation laws for transport of mass, momentum and energy.
CFD is the solution of such fluid dynamic problems by means of computer based simulation
Components of a numerical simulation are :
• Mathematical model
• Discretisation method
• Coordinate system
• Computational mesh
• Approximation
• Solution method
• Convergence criteria
CFD code consist mainly of three modules:
• Pre-processor- problem formulation and mesh construction
• Solver- solution of discretisation of governing equations
• Post-processor- analysis and display of results
The continuous Initial Boundary Value Problems - IBVPs are discretized into algebraic equations using following numerical methods :
• Discretization methods
• Solvers
• Numerical parameters
• Grid generation and transportation
• High performance computation and post processing
Discretization methods are as follows:
• Finite difference method
• Finite volume method
• Finite element method
Types of CFD Codes;
1. Commercial codes:
a. FLUENT
b. STARCD
c. CFDRC
d. CFX/AEA
2. Research codes:
a. CFDSHIP-IOWA
3. Public domain codes:
a. PHI3D
b. HYDRO
4. Grid Generation and Visualisation codes :
a. GRIDGEN
b. GAMBIT
c. TECPLOT
d. FIELDVIEW
Question
9 answers
As we know, the Nusselt number for an isothermal flat plate is written as Nu=h*x/k where h is the convective coefficient, x is length and k is the thermal conductivity of the fluid. In addition: h=q/(Tw-T_inf) and q=k*dT/dy at y=0.
Now, if the thermal conductivity is a function of temperature, then, the thermal conductivity of free stream (k_inf) is different from thermal conductivity at the surface (k_w). For q and Nusselt which k is adequate? I believe that for q we should use k_w and for Nusselt we should use k_inf. Because q is calculated at the plate, but Nusselt is calculated regarding to flow. However, in some papers (not all of them) i see that they used k_inf for both of the Nusselt and q.
What is the best definition for Nusselt number and why?
Relevant answer
Answer
Hi mohammad,
don't have experience in that specific case but as far as I know a local Nusselt number can always be calculated by an energy balance on the surface in this case you only need k_w. But obviously dt/dy|_w changes and so does Nu_x.
I have used a similar approach in the following paper: International Journal of Heat and Mass Transfer. 01/2013; 59:219 - 229. for calculating the local Nusselt numbers around a cylinder. Of course k was a constant in that case but this doesn't change the formulation.
For a simulation it would be quit easy to calculate dt/dy and correspondingly the local values of the Nusselt number then you can integrate this for a mean Nusselt number. Of course it is possible to use an average temperature to advise a mean Nusselt number but then the question is how to calculate the film temperature (1/2 1/3 and other suggestions).
Anyway for engineering purposes I guess using a film temp and an average k would be just fine.
Question
2 answers
The equivalent sand grain height approach is widely used in modern Computational Fluid Dynamics (CFD) calculations on rough surfaces. I am conducting a few simulations against available experimental data on a rough aerofoil. The experiment reported the rough surfaces were sandpaper P80 and grit 36 without mention the equivalent one. Surface roughness research is not a field I'm familiar with, because I concentrate on roughness modeling in aerodynamics. So I tend to ask if any one who knows any data or correlations to sandpapers. Don't mentioned the method in Boundary Layer Theory by Schlichting, because the experiment didn't provide me sufficient data to use that. Dirling's correlation is also not viable because I can't do the roughness geometrical measurement.
Question
29 answers
I am working on optimising the design of the static guide vanes found encasing a vertical wind turbine however I am finding it difficult picking out aerodynamic theories that would be relevant to this area of study. Any suggestions would be welcome.
Relevant answer
Answer
We have done a lot at our department in the Otto von Guericke University Magdeburg to optimize the Savonius turbine. Send me an email, I can send you a doctor thesis.
Question
4 answers
What is the right way to change the moisture content of the grain from wb% to mol/m3 to put the initial and equilibrium moisture contents in COMSOL?
Relevant answer
Answer
Thank you very much Kamil it really help
Why are pressure and specific enthalpy commonly the choice of state variables when modeling two-phase thermal-hydraulics?
Question
15 answers
Why not use specific internal energy or entropy instead of specific enthalpy?
Relevant answer
Answer
Pressure is the only parameter controls the saturation condition (via saturation temperature) and consequently the transfer between the two phases. specific evaporation enthalpy controls the amount of heat being transferred between the two phases during evaporation or condensation process (at the saturation line during thermodynamic non-equilibrium). A. Hainoun
Question
3 answers
Any basic book to understand the code of a Lattice Boltzmann Solver for CFD? I basically want to use XFlow.
Relevant answer
Answer
There is the book "The Lattice Boltzmann Equation: For Fluid Dynamics and Beyond" by Sauro Succi, which is a standard reference. If you just want to download a summary from the web try http://www.ndsu.edu/fileadmin/physics.ndsu.edu/Wagner/LBbook.pdf
Question
4 answers
The Euler equation is used with vorticity confinement methods.
Relevant answer
Answer
Do you mean the Euler's equation of inviscid flows? viscosity is assumed zero for an inviscid flow and Re is not defined
Question
4 answers
I have to simulate a feed nozzle design for FCC (Fluidized Catalytic Cracking) column in a petroleum refinery. It is a multiphase problem. For this, I am asked to choose the most appropriate simulation package b/n COMSOL and FLUENT?
Can anyone help me in this regard by sharing your views on which is best for solving such cases?
Relevant answer
Answer
Hi Sairamu Madala. In my opinion, you may model the kinetic and behavior of the fluid flow using FLUENT through the user-defined functions. Multiphase problem can be solved by volume of fluid (VOF) model, FLUENT has this function. However, if you want to solve mass balances for more than 50 species, there is a limit for FLUENT. May be you have to make sure what kind of model that need for you simulation, and then compare both simulation packages. I think the attachment and the link may useful for you.
Question
49 answers
Considering a fully developed flow in an annular pipe. The velocity is only a function of radius. What a bout the pressure? I know the pressure is a function of pipe length. Is the pressure also a function of radius?
Relevant answer
Answer
pressure drop is the driving force to push fluid to flow through a pipe, that means as far as there is no pressure difference between two points along the pipe length there is no flow. Generally pressure drop takes place even radially, but the value is too smaller than that in length. therefore we just consider the pressure drop in pipe length.
when you have a fully developed flow it doesn't mean that there is no pressure drop, it just means there is no velocity difference along the axis where we have fully developed flow.
Question
10 answers
I know how to bilinear interpolation in a unit square grid but I am struggling to find bilinear interpolation involving radial and direction(angle)
Relevant answer
Answer
Uma,
Your interpolation equations appear to be correct on the unit grid, but to put them in polar coordinates, you're probably going to want to ditch the special unit grid case and use the generic case.
Doing so, you get the equation:
p(r,t) = ( (r1-r)*[p0*(t1-t)+p3*(t-t0)]+(r-r0)*[p1*(t1-t)+p2*(t-t0)] ) / ((r1-r0)*(t1-t0))
where
r: radius
t: theta(angle)
p: f(r,t)
p0: p(r0,t0)
p1: p(r1,t0)
p2: p(r1,t1)
p3: p(r0,t1)
You can derive these equations using your attached proof, without assuming a unit grid.
PLEASE NOTE
- Your interpolation is now linear in POLAR SPACE, not Cartesian (real distances). I am not sure what problem you are trying to solve, but it sounds like this is maybe what you want.
Best,
Troy
Question
3 answers
I need some suggestions on CFD Analysis of microturbine ( speed above 1,00,000rpm)..
i am planning to use ANSYS-CFX as i need to proceed with FSI of blades after getting forces from CFD..
Details- Curved blade rotor, 8 blades, axial flow, clearance btw wall 1mm.
I am planning to use sliding mesh to the rotating interfaces since unsteady conditions may exist.
My Query- How to mesh fluid and solid region, whether rotating parts alone meshed separately or stationary volume also need to be meshed along with it..
Does it requires to define the boundary btw blades and hubdisk since both are rotating..but separate..
I am using Sliding mesh for first time,, looking for some tips
Relevant answer
Answer
Hi there.
I have been working on Microturbine Generation system modelling and simulation using MATLAB. I'm getting speed about 2,73,000 rpm when connected to permanent magnet synchronous generator. As i'm from electrical field, i couldn't able to understand the basics of those control blocks (temp control, radiation shield, exhaust temp, fuel valve actuator...). Could you please help me out?
And i saw one thesis for your CFD analysis...I can forward that copy if you're interested.
Question
12 answers
I have been working on rotor and propeller blade modifications that shift L/D and also Cd at high and low AoA respectively. The work is focused on rotorcraft at present, with rotor diameters from 25' to 65', symmetric and asymmetric blades. Are there any testing programs involving HAWT blades with diameters from 60' through to 180'? The boundary layer modification that is being employed indicates that some rotor models such as Camrad II are not suitable to a CFD analysis of the dynamics involved.
Relevant answer
Answer
Arturo, COMSOL becomes constrained at higher Re values to being RANS rather than a DNS. This results in limitation of the validity of the outcome. In principle it would otherwise permit the evaluation of the development of a vortex filament on a surface subject to radial acceleration. Thanks for the thought.
Question
30 answers
Give your personal opinion for others to read.
Relevant answer
Answer
I did not elaborate in my earlier comment because I thought most people are aware of h index and impact factor. But it seems it is necessary to say a few things.
By and large, impact factor relates to quantification of the quality of a Journal whereas h factor to that of an author. A journal has higher impact factor if it has more readership and thus popular. Now we know all over the world the number of scientific community is much more in biological and medical sciences. In physics it is not that big. In chemistry still there is large community. Therefore the impact factors f journals in physics remain low as compared to those in medical and biology fields. Now a days there is a tendency to have broad based generals like "biomedical and materials" having high impact factors because the audience they cater to is very large. Anyway this has become a technique to increase impact factor o a journal. The classical journals in physics who are well rated like Physical Review do not get their impact factors increased very much above 3. Still they have their place in the society of physicists.
As regards h index, it has not much to do with impact factor of a journal directly. It only is an indicator how many of your papers are cited how many times. It does not depend on the quality of who cites them or which journal cites them. In big groups this h index tends to rise enormously because all collaborators refer invariably the group citations and continues endlessly.
Of course there is a problem for a relatively new entrant to establish and find ppl who can cite his/her work. It seems one has also to be smart in addition to good worker.
Question
7 answers
I wish to make a simple program that solves tridiagonal matrix . now TDM can be solved using Thomas' algorithm, which is very serial in nature. I have come across cyclic reduction , even-odd reduction method in one of nvidia's papers. Can some one please explain me this algorithm, and help me with the pseudo code?
Relevant answer
Answer
Indeed, the cusparse library from nvidia already have a tridiagonal solver, so you dont't have to implement it from scratch, just use the one provided in cusparse.
Question
2 answers
I am working now on the grain drying with CFD simulation using Comsol and I need to start with one kernel simulation for moisture transfer during drying. This is the first time for me to use the Comsol. So, How can I start?
Relevant answer
Answer
We can coupling CFD and diffusion models in COMSOL to simulate the heat and mass transfer during drying.
Question
15 answers
I have been trying to simulate droplet falling, impact and spreading. I have used COMSOL and Fluent. I got some results but they are not satisfactory. Is there anyone who worked on similar things? I need to :
1) define an initial velocity of the droplet
2) apply dynamic contact angle during spreading
For COMSOL, the droplet don't deform that much after the impact.and for fluent I having trouble defining a free / open boundary. I have been using pressure inlet for free boundary. Here is some simulation result.
1. VOF+Level Set - with surface tension - with wall adhession (http://www.youtube.com/watch?v=P4VkHX9UYeE)
2. VOF+Level Set - with Surface Tension - no wall Adhession (http://www.youtube.com/watch?v=4UVQ-AM-vkY)
3. VOF (no Level Set) - without surface tension - no wall adhession (http://www.youtube.com/watch?v=rtVbNEHD6zE)
Can anyone help me regarding regarding my queries?
1. How can I assign initial velocity for the droplet?
2. How can I get more realistic deformation?
3. How can I apply wetted / non-wetted surface condition?
4. How do I implement dynamic contact angle?
Can anyone tell me whether Smooth Particle Hydrodynamics (SPH) is a better option for this problem?
Relevant answer
Answer
I think those droplets are big enought that you should be able to use the Stoke's settling velocity as a decent estimate, though less than about twenty microns might need a better estimation. Ther teminal velocity of the 100 micron and smaller droplets is surprisingly slow. Down to about 10 microns they fall so slowly that I suspect the surface tension is going to overwhelm much of your other deformation simulations.
A good first step might be to plot your settling velocity for your range of particle diameters. (They're droplet shaped, but you can use the aerodynamic diameter, which is basically the average diameter.) I suspect that you're going to see that you'll need a few different simulations for the impact and spreading.
What is the relationship between the lift coefficients of upper and lower transition points for the surface of airfoils at increasing AOA up to 10?
Question
35 answers
In the upper surface of airfoil, transition occurs in 30% chord and moves towards the leading edge and pressure side 40% chord from the leading edge(moves towards trailing edge), the lift value increases up to 10 of AOA. What is the relationship of upper and lower transition points with lift coefficient?
Relevant answer
Answer
Basically the transition point is within the boundary layer. Before stalling takes place, the detached flow from the upper surface again becomes attached before becoming completely turbulent in nature. Mainly boundary layer phenomena needs to be known. Also because of reattachment of flow before final detachment from surface, the pilots get a warning message about imminent stall. Thus.
Question
7 answers
I have meshed my design model but i am not able to export it to fluent (CFD). Everytime it displays an error in fluent. Please guide me.
Relevant answer
Answer
If "mesh exported successfully" message is displayed at bottom of Gambit transcript window, then only one possible mistake would result in displaying an error in Fluent. Either you are reading a 2D mesh in 3D fluent session or 3D mesh in 2D Fluent session.
Question
8 answers
I want to solve a CFD problem with a mesh generated by my own code in OpenFOAM. The question is: in what format is better to export the mesh for OpenFOAM?
Relevant answer
Answer
Thank you Bruce, the problem I have is that writing the point file is not difficult, quite intuitive actually, but the others, since they require the knowledge of the relationship between an element and the other, are a bit difficult. Anyway, thanks a lot!
Cheers
Question
4 answers
Total drag is sum of viscous and pressure drag . code Cdp and Cdf in C++??
Relevant answer
Answer
Hi Krunal,
You just need to integrate the total stress tensor on the surface of your immersed object. Have a look at: international Journal of Multiphase Flow. 04/2012; 40:38-55,
There are some suggestion for the interpolation and integration of the stress tensor on the surface in the context of fixed mesh methods if using a body conformal mesh the problem would be much simpler. Of course there is no code in the paper but should be pretty straightforward to implement the method in any Languages.
Not sure what you mean by FDM (finite difference method maybe?) but I think the important thing is the type of mesh you are using and you need some interpolation and integration to calculate the surface forces.
good luck
Sina
Question
3 answers
I'm not very clear about the no-through condition and no-slip condition. When I write programs to do the impulsively started cylinder simulation, I found that after convect the particles, a few enter the boundary. I want to know whether the no-through condition should take into account this time and how to do it.
Relevant answer
Answer
Yes, it is. I used the discrete vortex particles to simulate the flow, and the kinematic boundary condition is called no-through flow boundary condition. I want to know how to ensure this when use particles.
Question
2 answers
How can we raise the limits of integration for a triangular element in cylindrical coordinates in FEA? How are the limits in the radial and tangential direction?
Relevant answer
Answer
I don't really understand your question. If you could phrase it better and/or in a different way...
Question
16 answers
I am looking for software for wind simulation for Mac OS. I use Ansys but I heard that there is no version for Mac. Is there any other software that you are satisfied with using? Is there a free version for students?
Relevant answer
Answer
By the way... virtual machine consume lot of resources.
Question
11 answers
Is it possible to diagnose a disease from mechanical properties of blood ?
Relevant answer
Answer
The mechanical properties (Viscosity) of blood are very much influenced by the patient’s health. In fact, various diseases change the blood rheology such as diabetes, myocardial infarction, hypertension and rheumatic diseases. The increase in the blood viscosity might be due to an increase in blood’s cholesterol levels. The cholesterol levels are not constant throughout the day, as after a big meal the levels rise and after sleep the levels lower. Since an increase in cholesterol levels mean an increase in blood viscosity, there is an inherent uncertainty in blood viscosity for a patient on a daily basis.
Question
6 answers
I am a Student of M Tech (CFD) and working on Nasal Airflow simulation. I want to prepare a 3D geometric reconstruction of Nasal cavity. The computer tomographic (CT) images are used to obtain 3D model, but I dont know how ? Can anybody help me by explaining how these images are processed to get 3D model and is there any software used in medical technology to convert into 3D model ?
Relevant answer
Answer
Hi,
I think I have an answer for you! 3D animation program Blender can help you to prepare the nasal cavity, the air and the simulation of it. It is a free software which is worth thousands of dollars.
Check the links in the following order:
Good luck!
Question
2 answers
I calculated the flow in arteries and I would like to show that for some time steps the wall shear stress has the opposite direction, for example in the curves, branches, etc. For the straight pipes it is easy but how to do it for the more complicated geometries?
Relevant answer
Answer
I was trying to import the project to Matlab but as it takes 4 GB it did not manage. What I need is only to indicate direction of the flow. As it is done for the arrows. I could export data to the txt file but I have 500 time steps then it will take some time. I need to do it somehow in automatic way.
Question
9 answers
I'm a CATIA V5- beginner. I have designed with CATIA V5 a 3D re-entry capsule. The surface of the capsule is not straight but curved similar to a spherical surface. The Purpose is to integrate and connect this model with a grid generator and then perform as generic CFD model simulations.
Of course for numerical simulations I have to define a "far-field" for this model for example for a spherical model. Regarding this matter I have two questions with please for your support:
1) How can I define using CATIA v5 more extra-circular surfaces on this sphere surface for a side jet, so that the circular surfaces exactly to be placed on the curved surface of the sphere?
2 How can I define then to this spherical - model a solid room for CFD simulations as a "far-field"? Then the generic model (capsule) overlaps or intersects with far field as a solid volume then is a problem or not?
Relevant answer
Answer
For 1) : if you want simulate this file, you must insert new part body in part design of CATIA, after then you can define new surfaces.
Question
1 answer
My question is related to different softwares/tools that we use to predict the temperature rise in the Electrical distribution enclosures such as switchgears, motor control centers, etc
Apart from various ANSYS modules like Themal/CFX/Fluent, etc. do we have any other popular packages which we can use to predict temp rise ? If yes, how do we compare the solvers performance with ANSYS ?
Relevant answer
Answer
PHOENICS(see applications over Internet)
Question
2 answers
I looked up through the literature and I ended up with a work done by Olsen H. and here is the citation "Olsen, H. (1970). Unpublished sloshing experiments at the Technical University of Delft. Technical report, Delft, The Netherlands." However, I was not able to get this reference.
Relevant answer
Answer
Here are few references:
[43] G. G. Stokes. On the theory of oscillatory waves. Transactions of the Cambridge
Philosophical Society, 8:441–455, 1847.
[44] G. G. Stokes. Mathematical and Physical Papers. Cambridge University Press,
pages 197–229, 1880.
[45] I. A. Svendsen. Introduction to nearshore hydrodynamics. World Science Publishing
Ltd, 2005.
[46] J. M. Vanden-Broeck and M. C. Shen. A note on solitary and cnoidal waves
with surface tension. Journal of Applied Mathematics and Physics, 34:112–117,
1983.
[47] G. R. Whitham. Nonlinear dispersion of water waves. J. Fluid Mech., 27:399–
412, 1967.
[48] A. S. Peters and J. J. Stoker. Solitary waves in liquids having non constant
density. Communications on Pure and Applied Mathematics, 8:115–164, 1960.
Question
5 answers
Because Fluent is a commercial code, it is not always good to use the default turbulence model inside for research purposes. To define an omega equation based turbulence model, in Fluent, we can choose the k omega model, define our own user defined scalar (UDS) transport equations, and then solve the newly defined transport equations instead of the default k omega model. The k and omega values are also necessary to be copied back to the default k and omega field, because the momentum and energy equation needs the value of k and turbulent viscosity (computing the unknown Reynolds stress) as well.
So, my first question is, is there anyone has done similar things before?
I managed to write an UDF to define all necessary UDS, boundary conditions, and initialization, but I find the k and omega jump to negative values in some region of the flow field. Linearization of the source term, limiters for k and omega may help, but I don't exactly know how to do it. So, my second question is, how to deal with the solution divergence when using UDS k and omega equations.
My ultimate goal is to define all k omega based turbulence models, such as SST, and γ-Reθ SST.
Relevant answer
Answer
Hello, I have written an UDF code for k-w SST model firstly. (I enclosed my UDF here)
For compiling, I have done these steps:
1-go to define/user-defined/functions/compile and add the UDF and then bottom "Build" and "Load"
2-go to define/user-defined/scalars and put 2 UDS
3-go to model/viscous and hook Turbulent viscosity
4-go to "Fluid" and click on "Source terms" and hook two source terms for two UDS
4-go to Boundary conditions and select "plate" as wall and then select "fixed value" for both UDS, put zero for UDS1 (as turbulent kinitic energy) and select wall UDF for UDS2 (as specific dissipation rate)
5-go to pressure outlet boundary condition and click on "UDS" tap, then put the value of UDS1 and 2. I do the same procedure for velocity inlet boundary, too
6-go to solution control/equations, and select User Scalar 0" and "User Scalar 1". Moreover, I uncheck "Turbulence"
7-start calculation
But, this UDF does not work very well. It would be appreciated if you help me to solve the problem. I just want to get a simulation with this UDF and compare with Fluent one.
Question
8 answers
I am researching biomimetics of fish and sea creatures, especially their motion. Any paper, documentary, or piece of information would be useful for me.
I am trying to understand the idea behind motions of fish and sea creatures, regarding efficiency and hydrodynamics.
Relevant answer
Answer
Hi,
maybe this webpages are helpful:
EPFL Lausanne in Switzerland
Max Planck Institute Tübingen Germany
Cheers Stephan
Question
6 answers
I want to model the propagation of a flame front using MATLAB. What model can be used? Any references from where I can get information?
Relevant answer
Answer
Try this book: "J. Warnatz, U. Maas, R.W.Dibble. Combustion. Physical and Chemical Fundamentals, Modeling and Simulation, Experiments, Pollutant Formation. ISBN-10 3-540-25992-9. Springer-Verlag Berlin Heidelberg 2006." I have used it for modeling the gas kinetics for gasification.
Question
11 answers
I think mesh-based methods can be completely replaced by mesh-free methods, but maybe mesh-based methods such as FEMs have some useful properties that mesh-free methods don't have them.
Relevant answer
Answer
In general, for similar problems and at the same level of demanded accuracy, FEM is faster than mesh-free methods. The reason is the necessity of seeking the neighbouring nodes at every step, it's extremely time-consuming job.
Nowadays, mesh-free methods (again, in general) should be used only in two cases:
1. When mesh deformation is a significant problem. In such cases mesh-free is faster, or just only available method. For FEM deformation of elements can be overcame by dynamic remeshing, but it's time-consuming too and not always works correctly - than one can use mesh-free method.
2. When many crack surfaces are developed in the material - like for example in crushing a ceramics. For FEM, alternatively one can use XFEM method, but it's slow.
So answering the main question - in the near future, mesh-based methods rather will not be fully replaced by mesh-free methods - they are too slow.
Comparing Wind Tunnel Tests with Fluent Results - Dimensional Modifications
Question
115 answers
I have taken wind tunnel experimentation for 20 cm chord and 25 cm length airfoil, and want to compare the results with 1 m length and 20 cm chord of airfoil in FLUENT. Lift and drag are inversely proportional to the lift and drag coefficient. Is this right or do I have to model for 25mm length of airfoil?
Relevant answer
Answer
Under the assumption that you are interested in the two-dimensional airfoil characteristics, you only have to make sure that you are comparing at the same Mach and Reynolds number. This is resulting from non-dimensional analysis of the basic equations. At least in the computation, as long as you can guarantuee an inifinite wing span by using symmtery consitions at the sides, the computed result should not depend on the span at all. At least this is a test for proper two-dimensional simulations using three-dimensional CFD. But, since your model in the tunnel has a finite aspect ratio, and you didn't mention the cross section of your tunnel, you will observe differences when you simply compare at the same angle of attack. As long as your cross secttion is not too small for the model (see the comment on blockage above), that's why wind tunnel corrections should have been applied. Usually your tunnel should be able to provide you with informations about the corrections applied. Depending on the investigated range of AoAs your finite aspect ratio will introduce a spanwise loading, making it impossible to compare measurements and computations at the same angle of attack - in this case it is better to vary the AoA to match the pressure distributions to assess the AoA-shift between wind tunnel and CFD (free air). Of course you can try to model the complete experiment of the airfoil in the tunnel, but you should be aware that this is not an as easy task as it sounds. At least you have to make sure that the boundary layer at the wind tunnel walls is developped in the same way, meaning that you first should simulate the empty tunnel and compare boundary layer profiles at all side walls along the tunnel axis with appropriate measurements. If you don't do this the uncertainty may be reduced but the risk of introducing new unknown deviations is of a high probability.
Question
9 answers
I am working on an analysis in which oil flows through a tube and causes a turbine to spin. I would like to add friction between the spinning turbine and the shaft as it is spinning around.
Currently, according to CFD, the turbine is spinning at over 7 million RPM at average flow which seems completely impossible. Any insight on this matter would be greatly appreciated.
Relevant answer
Answer
Hello Isabelle,
Since I am here on campus, you should feel free to drop by, room 007A Hayden. Something is indeed wrong with your simulation, and it is not just a matter of friction: a turbine even without skin friction or form drag will still have a fixed angular velocity relative to the flow velocity, just like a screw turning through wood... Jim Papadopoulos
Question
13 answers
I have a reasonably complex experimental electro-mechanical hydraulic system (using water medium) design I would like to have simulated (to include pressure and flowrate versus time response, valve operation - steady state and transient) before I spend a lot on hardware to build it. If anyone has any connections/suggestions, please let me know.
ICEM CFD help needed
Question
21 answers
I modeled wind turbine blade in solid works and meshed with ICEM CFD. After naming the parts, while meshing: "The application has run out of memory and will terminate after doing an emergency save". My system has configuration of 3GB RAM and 320 GB HD. Can anyone help me with this problem?
Relevant answer
Answer
This is memory problem for your computer.please try another system otherwise increase the element sizes that means reduce the total number of elements.
Question
2 answers
I am doing analysis of centrifugal pump in turbine mode. I meshed model in turbine mode. Then exported in Gambit for boundary conditions. Then read mesh file in fluent and done 3 D analysis. After doing complete analysis I closed fluent. When I start fluent again and read cas and data file it shows error of grid interface. It shows like this:
Building...
mesh
Error: %%create-sliding-interface: wta(sb2-id)
Error Object: sb1-id
materials,
interface,
domains,
mixture
zones,
default-interior:014
default-interior:001
default-interior
inlet
outlet
outlet_pipe_wall
casing_walls
wall
interface_1
interface_2
interface_3
interface_4
Skipping zone wall-18 (not referenced by grid).
Skipping zone wall-19 (not referenced by grid).
wall-21
wall-22
casing
impeller
outlet_pipe
interior-17
interior-20
Error: CDR: invalid argument [1]: wrong type [not a pair]
Error Object: #f
Relevant answer
Answer
you are doing something wrong when your are defining boundary conditions in the gambit might be your are not able to give proper interface between two interior faces or didn't define flow field very well that the reason this error come so i would like to suggest you to do meshing and give boundary conditions at meshing packages like gridgen, icem  or you can use fluent mesh as well to resolve this issue.
Question
8 answers
To better understand the discretization and also to narrow the mesh independent test.
Relevant answer
Answer
If other equation than turbulence one is to be solved, you must consider of course the associated Peclet number which compare convection transfer Vs Diffusion transfer. This will allow you to estime the transfer phenomenon boundary layer thickness. The rule to obey is "you need more than one mesh inside the real boundary layer". So your mesh dimension must be chosen according with this rule.
Question
6 answers
of temperature, the model doesn't seem to handle the heat produced by Joule dissipation. It is going further
Relevant answer
Answer
Many things could have been wrong. It could the code problem in COMSOL, the time scale problem, mesh issue, etc. You could split the problem to get rid of the CFD part first.
Question
2 answers
I have been asked to represent a rectangular shape as a mesh. I am doing a numerical work which deals with flow inside tanks in case the tank subjects to external excitation. The dimension of the tank is considered to have 104 points in x-direction and 60 in y-direction. Does anyone know how can I handle this task?
Relevant answer
Answer
Thanks Ravi, Although it seemed a trivial task, but the accumulation of the research work made this a little bit annoying matter. Thanks anyways for your valuable answer.
Question
8 answers
Velocity profile is a factor, that you could find the velocity in any depth of fluid by changing the variable.
Relevant answer
Answer
Basically, the profile of velocity not only shows you the magnitude of velocity but also shows you the characteristics of the flow like direction, change due to shape of the domain or increase-decrease in the velocity magnitude with respect to the geometry and so on. So in general, it helps you to understand how the fluid behave while it is transported through the domain.
Question
3 answers
I would like to verify my FLUENT model for the dam-break process on a positive slope (upward, like a beach). But my reservoir is in the horizontal plain. Can you suggest some literature for this purpose.
Secondly, can dam-break be modeled with laminar flow conditions (instead of turbulent k-epsilon model)?
Relevant answer
Answer
What is the flow properties.
Question
3 answers
I am learning to develop algorithm for flooding and drying. And am using the 2D shallow water model.
Relevant answer
Answer
Note that the 2D-shallow water model excludes viscous effects; thus it inherently cannot describe the dynamics of dewetting that would lead to dry spots. Further, the shallow water model cannot describe the flow of thin liquid films near a boundary, the dynamics of which are dominated by viscous effects. I think the approach you will have to take is one based on a singular perturbation analysis, in which the outer solution is described by the shallow water wave model and the inner solution ( when the layer thickness is sufficiently thin) described by a lubrication-like flow dominated by viscosity and surface tension. There are many examples in the literature that discuss deweting of thin films using a lubrication model.
But your challenge will be to define the appropriate scalings such that you have the correct limits as the layer becomes vanishingly small in thickness, and have the correct matching conditions.
Perhaps others on the forum know if this has been done before- I am not familiar with the literature of shallow water waves....
Question
3 answers
When I compile the code (written by Stephane Popinet), I am faced with a code warning like "max residual...". A screenshot is attached.
Relevant answer
Answer
Hello,
Gerris uses a prediction-correction method to solve the NS equations. The correction phase requires to solve a Poisson equation to obtain the pressure and a divergence-free velocity field. I think that the problem comes from this part of the computation.
This kind of problem is not supposed to happen if you use the default Gerris cases without modifying them. If you obtain this error with the standard version of the script file, you should contact directly Stéphane Popinet (he is on RG).
If you have modified the script file, please tell us what you have changed and I'll try to help you.
Question
26 answers
In CFD for which applications do we take the conservative governing equations and for which applications do we use the non-conservative governing equation?
Relevant answer
Answer
We drive the governing equations by considering a finite control volume. This control volume may be fixed in space with the fluid moving through it or the control volume may be moving with the fluid in a sense that same fluid particles are always remain inside the control volume. If the first case is taken then the governing equations will be in conservation form else these will be in nonconservation form.
See the book "Computational Fluid Dynamics - The Basics and Applications" by Jhon D. Anderson
Question
11 answers
I made a model on CATiA V5 and I am facing difficulty in doing the analysis part. I really do not know how to do the analysis. I haven't used Gambit or fluent before.
Relevant answer
Answer
You can export your file with *.stp after then open ANSYS Workbench and select "Fluid Flow (Fluent)". Now import file in "Geometry". Then you can create mesh with the mesh section. Then you can select boundary condition in "setup".
Question
1 answer
I am using the SSME geometry to perform a CFD analysis for the flow characteristics, thrust and plume. I am using an Inviscid density based solver with 20 MPa and 3500 K for my inlet. I am having difficulties in defining the pressure and temperature at the exit. Also should I be using a different type of solver (i.e. Pressure based SA Modle, etc.)?
Relevant answer
Answer
Sorry, may I know the SSME geometry. Usually we are doing cfd analysis for control volume. if we define inlet boundary conditions, then cfd gives output values for control volume.
How can I simulate multi bubble cavitation in comsol 4.2?
Question
10 answers
I have a system with multi bubble cavitation, and I want simulate the bubble dynamics include the Bjerknes force (first and second).
Relevant answer
Answer
Hi Vicente, I can not give you a direct answer to your question, however, there has been s recent paper in sono-chemistry by D. F. Rivas (i.e. Enhancing cavitation with micromachined surfaces). Hopefully that would help! With kind regards, Nikolay
Question
1 answer
Hi,
I have a set of PIV vector data recorded at 5000 Hz (!), representing a flow field around a bluff body. Downstream of the body, the flow seems to be turbulent. I want to find out, if the flow has an impact on small droplets. Therefore i want to calculate the Stokes number St=(Droplet relaxation time)/(Kolmogorov time scale). It is no problem to calculate the droplet relaxation time, but for the Kolmogorov time I need the turbulent energy dissipation (epsilon). Epsilon is difficult to calculate from the PIV data, since spatial and temporal resolution are limited. Today a collegue told me someting about an analogy between the Kolmogorov time scale and a Reynolds time scale (?). He said something about Reynolds time being about six times larger than Kolmogorov time, but I was not able to find his equation in literature.
Has anybody a helpful comment on this topic? Is there a way to approximate the Kolmogorov time scale?
Greetings,
Greg
Relevant answer
Answer
We typically use other statistical functions to get the estimate of dissipation, e.g. structure functions, Taylor surrogate, etc. If you're interested in more details or cooperation, welcome off-board to http://www.eng.tau.ac.il/efdl
Question
1 answer
I mean there is any simulatio program or softwear can do it simulatio for bod human or for any part of body?
Relevant answer
Answer
You are not very specific as to what exactly you want to simulate- physical motion (kinematics), chemical processes, ergonomics, or any number of other approaches. Perhaps this paper (http://www.cescg.org/CESCG-2000/RFilkorn/) can provide you a starting point.
Question
4 answers
What are the boundary conditions that should be imposed in order to obtain the simplified momentum, energy equations for shock waves? Is it possible?
Relevant answer
Answer
What is your problem exactly?
Question
8 answers
Software- fluent 12.0
Relevant answer
Answer
:) ok I ll check. Thank You Shantanu.
Question
8 answers
Are there any published results for experimental data of measurement of skin friction drag of a triangular plate at zero angle of attack?
Also, can we use theories developed for solving boundary layer over a flat plate (like blasius equation) to estimate the skin friction drag over a triangular shaped plate at zero angle of attack?
Thank you in advance!
Relevant answer
Answer
You can use Blasius theory if the flow is laminar, but those for the fully turbulent flow von Kármán’s empirical formula has to used.
You can check the calculation by Squire and Winterbottom, "Note on further profile drag calculations.
RAE Rep. BA 1634, 1940.".
Problem in Multiphase Modelling
Question
7 answers
I am working on simulation of solar still (water purification with help of solar energy) in FLUENT by using multi-phase modelling..Need some guidance.i didn't found the rate of mass transfer from vapor to liquid..i am using Mixture model,3 phases (air,liq,vapor), k-epsilon turbulence model. Thanks Vaibhav Rai Khare
Relevant answer
Answer
Goto Define>Phases>Interactions in the Mass section set Number of Mass Transfer Mechanism to 1 and define a constant rate for it. Good Luck
Anybody know about Wind tunnel modification project details? What are the data types I should know for that?
Question
5 answers
Wind tunnel modification and analysis in which software and fabrication is best.
Relevant answer
Answer
Hello friend can u send wind energy handbook, or wind energy bible........
Question
7 answers
Thanks
Vikas Sharma
Relevant answer
Answer
I think it could be possible using the Lattice Boltzman approach ?
Question
3 answers
Sir I want to validate this paper. PDf attached.
I have a Simple Flow Situation If you can give a minute , It will Be Solved and my life will be a lot easier.
I have a 3d duct, with a small obstruction/wing to create disturbance.
, Walls are at 400 k constant temperature.
and air at 0.8 m/s at 300k.
the duct is 21 x 14 X 160 mm.
well they are two parallel plates that are 14 mm apart. and the side are open , which are at 21 mm, i have defined symmetry.
It is similar to tutorial "Forced Convection over a plate""http://confluence.cornell.edu/display/SIMULATION/FLUENT+-+Forced+Convection+over+a+Flat+Plate", except, that , that my problem is in 3 d .
Most of friends have replied me like this " As far as finding nusselt number, that is the only way of doing it. ie. report surface integral and manipulate them in excel and then plot. "
i use h*x/k along a line at the bottom plate in its mid, but that gives the local axial variation i think.
but i need average, without using exel if possible.
Anyone intrested in seeing the *.cas, *.dat, *.lay files can mail me, i will send the files so that one can check properly.
Regards Saksham
+919467704656
any help on Tecplot or Fluent is welcome.
Relevant answer
Answer
I suppose you want to solve the Eq. 4 of the paper. Once you've simulated the flow you already have T_w1, T_w2 and T_b values. You can solve the denomiator using any numerical integration algorithm. For the numerator, first you find the heat transfered q_1 and q_2, trough Fourrier's Law. Use some numerical differentiation technique to find dT/dx and then integrate it. I would use MatLab to solve this.
Sory if I wrote something useless, I didn't read the paper neither understand your real question. :)
Question
3 answers
Hi... I am modeling open sea air turbulence (Re ~ 6.0e6) using SST (Menter) K-Omega RANS based turbulence model. What should be turbulent intensity (T.I.) at inlet? 1%, 5% or 10%? Also, does anyone have experience on influence of drag prediction for different T.I.? What's the physics correlating T.I. and Drag?
Relevant answer
Answer
@ Roza: I guess this formula is true for internal flow (e.g. pipelines, inlet duct). Let me correct if I am wrong.
Question
4 answers
I'm trying to simulate the response of an instrument to an input signal using MATLAB. To be specific its a constant temperature hot wire anemometer and I'm interested to simulate its dynamic response. For the input signal I'm planning to take in data from FLUENT(simulation of flow around a cylinder). For this I need the data of a particular point in the flow field. I wanted to know the procedure to import the data of a particular point in the flow field
Relevant answer
Answer
This is probably way too late to answer this, but what the heck.
This probably involves a lot of cas and dat files, because it is my understanding that it is a transient simulation. I suggest that you go with the following steps:
1.Export the data into a programme which can read the fluent files ( I would prefer Tecplot)
2. Use the probe command to obtain the pressure, temperature or any flow variable at any particular point.
3. This can saved as a dat file, which can later be read by MATLAB
I think that the major problem would be in importing numerous dat and cas files to Tecplot for each time step: I suggest that you use a Macro in Tecplot, which can be written in Python.
Regards
Question
2 answers
For example, there is a strip placed in center of a room, and is a high frequency noise source. Does the code allow to simulate the effect of this noise on the surrounding flow?
Thanks in advance!
Relevant answer
Answer
Some suggestions to explore:
COMSOL has an acoustics module
Code_Saturne has compressible flows, two-phase flows (Euler-Lagrange approach with two-way coupling), extensions to specific applications (e.g. for atmospheric environment).
Question
44 answers
I am proficient enough in programming using C and MATLAB but when I open any CFD text book, all i find is calculus. I want to slowly learn writing my own code but the books don't give any idea how to proceed with those hifi mathematical schemes.
Relevant answer
Answer
Hi Surya,
A strong or atleast functional understanding of calculus and numerical techniques is required to truly understand the code behind CFD and program a solver. Familiarity with C, C++, Python and MATLAB will definitely help in this direction. The programming codes are written to implement mathematical schemes.
Text Books : You can refer to 'Versteeg and Malalasekara' for a good introduction and specific guidelines/ an example to programming a pressure based algorithm. I also recommend using John Anderson's - Introduction to CFD, which is a great book for the theory with a gradual progression in the complexity of the mathematics involved. Another book that should help relate to the coding is 'Applied CFD - Rainald Lohner'.
A simply worded introduction to CFD , focused on concepts and much less mathematics (and a free ebook) > http://bookboon.com/en/textbooks/energy-environment/computational-fluid-dynamics
To understand Fluid Mechanics on a fast(er) track, try the videos and notes at > http://web.mit.edu/hml/ncfmf.html
Some very nice sources for coding examples : http://storify.com/ShreyasRagavan/software-and-code-related
Once you get along, you might want to plunge into open source CFD applications like OpenFOAM and SU2.
I'd recommend playing around with SU2 before going for openFOAM.
Have a look at the resources I've collated > http://cfdrevolutions.weebly.com/resources.html
The 'open access' and 'mathematics' sections have resources that can help with streamlining your understanding of the required calculus. The links I've posted above are spread across different sections and are available in the website.
Question
2 answers
I collected raw data from the field and intend to generate a function from the data, how do I do that?
Relevant answer
Answer
You might give Eureqa a try. http://creativemachines.cornell.edu/eureqa
Question
5 answers
I am interested in solutions of Nonlinear hyperbolic conservation laws with jump conditions (propagation of shock wave).
Relevant answer
Answer
me to ....but i need help to proceed...
Question
1 answer
now days i carried out simulation of flow pst circulare ,square ,triangle cylinders by using fluent
Relevant answer
Answer
Please see Karabelas,S.J., et al ,Applied Mathematical modelling,36(1),pp379-398,2012.It is high Re but you just avoid the turbulence model, if you prefer.
Question
11 answers
Can anyone help me in Computational modeling of Wind Turbine's Wake? I am just not sure to how to start this project in 2D? assuming an actuator disc doesn't always work... any new ideas?
Relevant answer
Answer
Wind Turbine wake modelling is not a simple problem and simply selecting one CFD package over another is not the answer.
Modelling the flow over an aerofoil section requires a fine and rather sophisticated mesh to capture the boundary layer separation. This means elements in the 1mm range on the blade surface.
Problem is an wind turbine is massive which means a lot of elements. The second issue is the wake zone, what exactly are you trying to measure or simulate? If it's related to the performance of the blade section then you really don't care what happens downstream because this has no influence on the blade past 1 rotation (turbulence from the proceeding blade due to the solidity factor).
If however you are trying to model the low frequency noise that is driving everyone mad then you need to accurately capture the pressure pulsations.
That is an extremely difficult problem to solve because all of the current CFD commercial packages are highly diffusive due to the dominate face centered solvers.
Question
4 answers
is the Comsol most appropriate cfd software for academicians or lecturers ?
Relevant answer
Answer
"Most appropriate" is a difficult criterion to to evaluate. Before pursuing a major investment such as required by COMSOL, one should at leat evaluate the capabilities of OpenSource alternatives, such as those packaged with CAELinux, which includes OpenFOAM, Code_Saturne, Elmer, Gerris, etc. One major advantage of the OpenSource solution is the extensive user communities available, especially within academia, for support and development.
Question
7 answers
Is there anyone that conducted research on supersonic caviyt flow with FLUENT or other CFD codes?
or, is there anyone interested in this topic?
Relevant answer
Answer
In CFD, is it accurate to simulate half model? for saving the grids or computing resources.
Question
1 answer
I need help in sector , trimmimg and mapping method of combustion analysis of Diesel engine
Relevant answer
Answer
Dear Dr. Sandeep Joseph,
It is better that you ask the your question on STAR-CD online discussion forum in CFD online.
I think you can get more help rather than here.
Regards,
Souichi Murata
Question
4 answers
Hello,
I would like to know if somebody is dealing with this kind of simulation. I did not find any studies concerning this particular subject in CFD. My main issue is to know the forces acting in the human body when it impacts the water. Since I did not find anything more similar, I'm just trying to make a correlation with the slamming studies for boats, but too many assumptions must be made, and I'm afraid I'm not being too precise...
Can anyone help me by telling me some reference or ideas?
Thank you very much
Relevant answer
Answer
It seems to me that this is not an easily simplified reaction, because one is dealing with the interaction between two malleable bodies. How much of the dissipated energy is absorbed in the deformation of the human body on impact, and how much is dissipated in the displacement of the water as the body enters? Of course, the deformation of the human body is theoretically recoverable (assuming the velocity of contact was within a reasonable limit!), while it would seem that the energy displacing the water is pretty much lost (at least until the body leaves the water). Perhaps begin by thinking in terms of a rubber bladder of relatively stiff density sitting on top of a body of water- if you get something reasonable from this, the4n increase the initial velocity of the bladder to see if the model continues to provide reasonable predictions for an appropriate range of impact velocities. Ultimately, however, you are going to have to mount a bunch of force sensors on a diver and actually measure the forces you are seeking in order to validate any model you build...
Question
5 answers
Anyone have experience with ELMER? I have recently begun trying to use it for high pressure gas dissipations.
Relevant answer
Answer
I have limited experience with an older version of Elmer for CFD, and more recent experience using Elmer for some electrostatics analysis. My personal preference for CFD is for Code_Saturne, but that is a personal preference, not a technical assessment- I am more comfortable with the Code_Saturne GUI for setting up the problem, atthough Elmer is just as good (or, possibly better, with some additional capabilities for post analysis- I haven't fully explored all of the reatures). I also prefer Salome for modeling/meshing- again, more of a personal preference than a technical evaluation. The newest version of OpenFOAM has some features I am really intrigued by, but it requires upgrading my system to Ubuntu 10.04, which I haven't gotten around to yet. As a note, all of these packages are bundled with CAELinux, which has just released a new package built on Ubuntu 10.04- probably what I will upgrade to when I finally get around to it. I have also noted that the on-line forums for Sallome and Code_Saturne appear more active and more helpful than the other packages- again, that may be due to the fact that I have focused on my preferred applications.
What version of Elmer are you working with? What issues would you like to discuss? If you are just starting, I would be willing to have a look at your problem setup and give you some pointers in getting to the solution you are after. Of course, I might have to do a bit of studying to get up to speed on high pressure gas dissipation...
Evaluation Journals in the field of CFD, fluid mechanics
Question
5 answers
I would like to have some Journals in the field of CFD, fluid mechanics evaluated like this: Journal Name Publisher Review Time Revision time Publ. Time Total Time The CFD J. example publ. 2 months 3 months 1 month 6 months
Relevant answer
Answer
Unfortunately I had to go to a different library to check this. I didn't find any of the information I was looking for. I only saw information which are for interesting for librarians (e.g. cost, published since) but no information for authors like Revision time. Maybe you can post an example for the "Journal of Fluid Mechanics" and/or point me where I have to look specifically?!
Question
10 answers
I am working on a turbulent flow, I have read about enhanced wall treatment which is used to capture viscous sublayer. At very high Reynold's number (~10^7 or ~10^8) does this needed to be captured as this will be very small and may have less effect?
Relevant answer
Answer
Hi, to definitely answer your question you have to reveal some more detailes on your flow case - e.g. does any separation occur?
If you are not sure on the effect of sublayer (or whether your case is sensitive to Re), test it with two different meshes and compare results.
General guidelines are:
1) To use EWT you have to keep y+ < 1
2) for SWT (standard) you MUST keep it between 30< y+ < 200 (300)
hope it helps for starters...
I highly recommend you visiting CFD online forum - there are a lot of discussions on this topic already.
-Patryk
Question
9 answers
I am trying air injection to a normal flow of water over a flat plate to see the reduction in drag, what model must be suitable, VOF, Mixture, or Eulerian? Speed of water over plate is about 14-20 m/s. and length of plate is about .80m which makes Re~ 10^7.
Relevant answer
Answer
In the problem due to gravity buoyancy will come in picture which will cause the bubbles to rise and flow over torpedo will be unpredictable, hence analytical approach will not work, if buoyancy would not present then this approach will work.
Question
1 answer
1. When the liquid content between grains is small, liquid bridges are formed at the contact points of grains and create the meniscus models.
But I can't imagine where the liquid bridges would be generated when the liquid would flow into the grains. Which grains will be included? That is to say,
How is the connection between the grains generated? Is there a theoretical basis to explain this phenomenon?
2. When the liquid content between grains is small: how to calculate the liquid bridge force accurately?
When the liquid content between grains is more or almost saturated, how to calculate the suction due to the pressure difference between the air and the liquid?
Some references to explain this phenomenon would be very helpful.
3. We know that the wet granular materials can express complex phenomena, such as the repose angle can be larger than 90 degree.
But I wonder that whether cohesion is enough to express this phenomenon? Or should we take the lubrication of solid-solid friction caused by liquid into consideration?
Relevant answer
Answer
Hi Sheng,
I think you have some questions the answers for which are still on-going research. however in-regard to your second and third question I would suggest you to read the following article especially the first 3-4 sections of the paper. You can get a better idea about liquid bridge formation, forces, cohesion, lubrication(little bit) and references to article on angle of repose.
Question
1 answer
If we want to simulate long period case, we must use time scale to reduce our simulation time. But, I still dont understand the effect of NS Equation. How is the effects?
Relevant answer
Answer
Scaling through non dimensionalization makes your max time as 1. When we non dimensionalize NS equation, we non dimensionalize other parameters like velocity, viscosity , pressure as well. But as we scale, our time step decreases, hence there is no effect on final solution for NS in simulation.
Hope this helps.
Question
17 answers
I am trying to simulate BL transition on a flat plate at 50.1 m/s and 0.18% Tu. The plate is 1.8 mts long. Although, I detect transition on the plate it is way too early and is not correlating with the paper whose simulation I am trying to repeat. This is extremely strange because I have followed their instruction and have set the mesh parameters right with a y+ of 0.1. (I have mirrored the authors mesh). The only difference is they are using the SST gamma-theta model in STAR CCM and I am doing it in FLUENT.
Does anyone have any suggestion on how to improve the simulation? I am attaching the reference paper. Please see figure 3 in the attached paper and Case 1-TSK.
Relevant answer
Answer
Are you certain that the turbulence is "created" at your inlet boundary condition?
For some turbulence models, such as the SST, you have to provide values of turbulence properties at the boundaries. For your turbulence closure, you should provide to the code values of the "k" and "w", and maybe other, properties depending on which solver you are using.
I hope it was useful,
Good luck!
Question
1 answer
I am using k-epsilon model , is there a fluent tutorial available
Relevant answer
Answer
Did you try looking at the static mixer tutorial in fluent?
Try googling it. It should be easy to find.
Hope it helps.
Best wishes...
Question
17 answers
I need to generate a mesh over the surface of a 3d object consisting of a number of intersecting cylinders. This needs to be done repeatedly, within the loop of an optimization. I'm hoping there is a simple and free tool out there that could handle the task, so I wouldn't have to make my own. (Panelizing the cylinders is easy, but making an algorithm to deal with panelizing the intersections would be too complicated to code myself.) Are there any such software tools out there? Ideally something that could be called from within matlab or a c++ code?
Relevant answer
Answer
There are a couple of free tools available, for example gmsh (take a search machine to find). However, I suppose you additionally need something with open sources to add your routines for automatic processing. This is not always the case with otherwise free tools. A good start to search for a suitable instrument is:
Question
11 answers
Idea of learning a analysis software.. Which among the four is best with respect to application, popularity, useage and user friendly? I learnt Ansys structural and thermal analysis but yet to learn Fluid Analysis... Hyperworks is mainly for meshing but anyway I include it too... Please suggest me with better options..
Relevant answer
Answer
First learn the concept with any one of that , than you can choose . Because diferent people woring with their requirment.
Thanking you
Best regards
SUJAN
Question
4 answers
Micropolar fluid
Relevant answer
Answer
In Boundary layer theory we can simulate that type of every model by FDM Keller box method using MATLAB
I want to take results of pressure distribution on wind turbine blades, I modeled by solid works.
Question
6 answers
In my college Ansys 12 version is available. I need materials for meshing and analysis (fluent). Kindly suggest and send the tutorials related to my analysis.
Relevant answer
Answer
Meshing depends on the total volume of the project you are going to analyze. If the volume has to be meshed to more than a particular level of elements and nodes, ansys won't support it. You have to save your file in iges format and you have to use hypermesh for meshing it. Then you have to impost the *.msh file into the software. tell me whether you have fluent- Gambit- HummingBird application or Ansys Multiphysics- Fluent intergrated version? I have tutorials for the former which varys a little from the latter. Latter is easier than the former. I'm not sure whether I have the latter. If so I will let you know by 2morrow afternoon. I think the capacity is more too. I will check it out too. But for effective fluid analysis, Ansys is not a better choice. Go for Nastran Femap.
Question
7 answers
I have modeled the arc chamber.
Relevant answer
Answer
that's where you have got your model.... now what kind of simulation you want to do??
export file as iges format to icem cfd and then create mesh in it. Later you can export it to fluent for simulation
Question
6 answers
I am solving lid driven cavity using fvm and am having issues regarding scaling the residuals.
Relevant answer
Answer
it depends on the problem. for unsteady terms like du/dt you must put all the d/dt<0.001, and for pressure equation you can employ ((Pnew- Pold)/Pnew)<1e-8. if you want to get most accurate results, the following conditions are suggested
1. d/dt<0.001
2. ((Pnew- Pold)/Pnew)<1e-8
3. ((Unew- Uold)/Unew)<1e-8
Question
14 answers
I want solve by shooing method but I am not getting it. We can also use F''(-infinty) =0;
and F''(+infinity) =0; can anyone help?
Relevant answer
Answer
As I am sure you know, this is the Blasius equation. Below is a shooting method that I once wrote in matlab, for the typical case where you shoot from y=0 to infinity in order to get the ZPG boundary layer profile. It will be easy to adapt to your case.
Shooting from minus infinity to plus infinity sounds like a bad idea for stability issues - I suggest you shoot from y=0 in both directions, and you adapt the (unspecified but needed) BC for F'(0) iteratively.
To try this routine, pick a y0 not too far from 0.33:
======================
function [T,Ynew]=shootingBlasius(y0)
% by Lutz Lesshafft, 2011
% solves the ODE y''' + y*y'' = 0 with BC y(0)=0, y'(0)=0, y'(inf)=1.
% input parameter y0 is initial guess for y''(0).
% shooting algorithm:
[T,Y0] = ode45(@diffeqn,[0 20],[0 0 y0]);
y1=y0+0.01;
[T,Ynew] = ode45(@diffeqn,[0 20],[0 0 y1]);
while (abs(Ynew(end,2)-1) > 1e-15)
ynew=interp1([Y0(end,2) Ynew(end,2)],[y0 y1],1,'linear','extrap');
y0=y1; Y0=Ynew; y1=ynew;
[T,Ynew] = ode45(@diffeqn,[0 20],[0 0 ynew]);
disp(ynew);
disp([' error: ',num2str(abs(Ynew(end,2)-1))]);
end
end
function dy=diffeqn(t,y)
dy = zeros(3,1);
dy(1)=y(2);
dy(2)=y(3);
dy(3)=-0.5*y(1)*y(3);
end
Question
9 answers
Does anyone know how to time-average NO mass fraction in FLUENT?
After running combustion transient simulation, I enabled NOx model and turned off all equations except NO. Then, I continued the iteration.
However, NO mass fraction, concentration and rate of production being calculated appeared in the data sampling statistics. So, I could not get their time-averaged values.
Please kindly give me some suggestions.
Relevant answer
Answer
I suggest to write an UDF where you store and time-average your results in a UDM. If it is in cell-domain it would only be a couple of programming lines, where you loop through that specific domain, and store the results along something like this: NO(new)=NO(old)+NO(current)*dt. Using such an approach of course need some basic UDF-skills - however the UDF-manual will introduce you, and then CFD-online community or other place may help you debugging
Question
6 answers
I am writing a cfd code for solving N-S equation on a structured collocated grid (lid- driven cavity), my code is slow because of the above reason.
Relevant answer
Answer
I assume you're simulating a case which has been shown to be steady on other similar solvers?
Is the output from your solver also steady per real time step - i.e. you're not hitting some oscillatory limit cycle in the solution?
Another useful check is residual in each pseudotime superimposed for each real time step. When you restart the solution for the next real time step- is the residual going lower at the beginning of the new pseudo time iterations (perhaps not initially but it should after a few real Dt).
I would agree with Carlos to check what you've set as your termination criteria in your pseudo time loop. And also check from the plot of pseudo time residual that there's no convergence issues - i.e. it converges easily to your termination criteria.
Question
3 answers
I am injecting air over plate over which water is flowing(see domain attached), since vf = Qa/(Qa+Qw), Qw and vf are known through this Qa is calculated, and since air inlet dimension is known so velocity at inlet is calculated, but vf should i give at inlet in simulation in fluent, is it 1 or less than 1 or what value.
Relevant answer
Answer
Dear Mr. Aman,
From my experince, I think you should set the intensity of the air at the inlet of the water to 0.1air + 0.9 water (or 0.2air + 0.8 water). For the inlet of the air is 1.0 air + 0.0 water.
How can I model a 2D aerofoil with flaps?
Question
11 answers
Supercritical aerofoil with trailing edge flaps
Relevant answer
Answer
Pat Hanley's multi foil does flaps quite nicely with a quick potential flow or generated grid RANS/Euler analysis. Does cascades quite nicely.
Question
7 answers
Is it possible to simulate fire using Fluent in CFD?
Relevant answer
Answer
It was once done in COMSOL. Please read the abstract :
"Analysis of burning candles is extremely complex; combustion produces a highly non-linear temperature profile through the flame in which local temperatures may exceed 1400 °C. Heat transfer includes radiation, conduction and convection components and the low melting point of the candle wax leads to a phase change that allows mass transport via capillary flow prior to combustion in the flame. COMSOL Multiphysics has been combined with generalized strategies to analyze heat transfer and fluid flow during candle burning. Radiation was included through a radiating surface that was non-locally coupled to the radiating gas volume. The dynamic behavior of the plume was incorporated by artificial diffusion to give a time-averaged approximation. Heat transfer within the liquid wax used an anisotropic thermal conductivity to account for convection in the horizontal direction. The analyses predict the velocity flow field for a half-burned three-wick candle and demonstrate spreading of the flow away from the flame. Predicted temperature distributions in the wax and candle container compare favorably with experimental measurements"
Question
8 answers
Materials that can be useful.
Relevant answer
Answer
I thought my email was visible in my profile page. Is it not? In any case, it is carlos.lange@ualberta.ca.
Question
3 answers
I have to simulate a THREE PHASE SEPARATOR in which water+air+chemical enters into the chamber and water is to be taped out from one (01) OUTLET
and Air from second (02) OUTLET and chemical remains in the chamber...
***The problem here is that MIXTURE is set on all boundary conditions as well as on OUTLET. I am unable to select any one (01) fluid at OUTLET using FLUENT.
Can FLUENT Simulate this problem?
Relevant answer
Answer
Have you find any answer to this problem? I appreciate if you can share it with me
How can you determine a global Damkohler number in the case of chemical non-equilibrium flow with many reaction steps?
Question
5 answers
Is it possible to determine a single chemical characteristic time for a non-equilibrium flow?
Relevant answer
Answer
In my case there are some chemical processes in series and others in parallel, so if I understand correctly, it's possible to compute the overall chemical time as follows - series: tchem = t1 + ... + tns - parallel: tchem = max(t1,...,tnp). Computing eigenvalues of source term Jacobian is more useful to identify slow, fast, and quasi-equilibrium groupings, and to choose the best time integration strategy.
Question
3 answers
Can anyone provide an explanation?
Relevant answer
Answer
Thank u sir.
Question
7 answers
I'm working on elliptic geometry that is made to rotate at a distance from the central axis of the duct. I need to determine the local and average Nusselt number.
The values being generated from FLUENT are too low
Relevant answer
Answer
You should calculate reference Temperature. The reference temperature has a substantial effect on the Nusselt number. You can use mass average surface integral for calculating Tref.
Question
7 answers
Most of the papers I have read say that in the absence of pooling, the boundary condition at ground-level is given by a flux, something like q = K(dh/dz + 1). But I have not found how to specify what q is. It seems to me that it should be related to the humidity, and also the current level of saturation at the soil surface. Is there a common consensus on an elegant way to describe this boundary condition?
Relevant answer
Answer
Dirichlet condition (specified pressure head) can be assigned by using thermodynamic relation - please see Equation (12) in my paper at http://www.angelfire.com/nh/cpkumar/publication/IEI_1996_Evaporation.pdf
Question
3 answers
What is the main reason for this prompt to came when a simulation is being done on FLUENT
Relevant answer
Answer
Before anything, check your mesh, refine it 
Question
1 answer
how exactly is volume fraction defined in FLUENT,
or
if volume fraction of a =Qa/(Qa+Qb), then how can this be related to FLUENT
Relevant answer
Answer
You can use constant volume fraction in the multiphase panel, so that you can the value appears in the bouandry layer conditions panel, but if it is a fonction , you can introduce it by creating an independent file as fluent define it in its help manual and when running the calculation you refer to the file by profile in define panel.
Question
2 answers
Is there some handbook in which I can find density dependence on the temperature for liquid alkali metals, especially for lithium, sodium, potassium, and cesium at coexistance line?
Relevant answer
Answer
Take a look at this paper. It may help you: