Computational Fluid Dynamics

Need of Computational Fluid Dynamics (CFD) –

Computational Fluid Dynamics or CFD for short, this is a case for we use the computer software like Ansys, Hypermesh etc.to solve fluid dynamics problems. Why do we even need this? Why do we need fluid dynamics at all? 

We live in a world of motion and every movement that you make interact with fluids, walking down the corridor air flows around you. Your body hits the air over your head which then rises to the ceiling. A ship turning in a channel, is driven by the water flowing around the hall and the rider. Every aspect of change in our world gets shaped by fluid dynamics. Even that pervasive presence we need the ability to predict and control those fluid dynamics. CFD is one of the main tools that we use for this. To talk about CFD we first need to put in context of all of the methods available for fluid dynamics analysis. Methods are as follows:

For Free, Demo classes Call: 8149467521
Registration Link: Click Here!

1)Analytical Methods –

Analytical method contains generalized equations and runs for very simple specialized equations, so we don’t see this too often. This only works for really simple cases.

2)Experimental Methods –

Experimental methods are very accurate but it is limited and it is very expensive. We have to build specialized facilities; high construction costs and they have to be very highly calibrated and very carefully designed. So, we can really use them to explore new possibilities, we can only test for situation that we are already known about.

3)Computational Methods –

Computational method can be just as accurate as an experimental method but it can also be cheaper than experimental methods and gives us a lot more flexibility. So, it offers us lot of flexibility than analytical methods and lots of detailed accuracy than experimental methods. It is a median of both.

How does CFD process works?

Fluid dynamics has been around since the dawn of time and mathematicians must been working on this problem from years and decades and centuries. So why have they only now turned to computational methods? Why do we need computer software to solve the fluid dynamic problems? Haven’t the mathematicians solved this already? The basic answer to this is yes, they have been trying to solve this and fluid dynamics are hard. They are very complicated.

The full equations for fluid dynamics are what we call the Navier stokes equations. There is no general analytical solution to this equation that can work for every situation, so we have to turn to computational method to find solutions that work for all certain situations.

Navier Stokes equations –

DqD=-p+u2q+13uq+ρg

So, we are driving for here is some sort of generalized tool to solve Navier stokes equations. This is where the computational method came in. We said we can solve Navier stokes equations if we keep, the geometry simple. Let’s just keep it as a box. We can solve a Navier stokes equation if is just a box.

As long as we know the conditions at boundaries, we can solve that. That is good because now we have a generalized tool for this as long as we keep the geometry simple and this is what we start combining that of the Brute force capability of the computer. It takes that idea of solving Navier stokes equations on a little tiny box.

• It divides your object into millions of tiny little boxes, we call these as cells and we call the whole thing a mesh.
• Each cell interacts with its neighbors.
• The engineer sets known values for the boundaries along the edges of the mesh.
• Then the computer uses the power of a Brute force iteration to bounce up all of the cells along he edges and in between all of the interactions.

There is a lot of Brute force solution to the equations, there is a lot of iterations that goes on, but that’s okay. Computers are really good at Brute force iteration and that is the computational process.

The really good thing about this is that its generalized algorithm that we can apply to any geometry, it works for any situation, the same math just keeps working again and again.

For Free, Demo classes Call: 8149467521
Registration Link: Click Here!

Advantages of CFD –

• Computers are relatively cheap, you don’t have to pay too much for that Brute force power.
• You do get a lot of details out of your CFD simulation which is really good about answering not jus what your values are but why, it helps us explaining why the answer is good or why the answer is bad, it’s very good for making corrections and improving upon your results and It allows you to simulate your final product with real world physics.
• You are not making approximations, you are not building models. You can simulate things as they actually are in the real world. And because of this and because of the fact that CFD is a generalized solver you can really tackle this without any prior knowledge of exact what you expect, so that allows you to do a lot of optimization and exploration of your design.
• You don’t have to invent the solution every single time, you don’t have to reinvent the tool bar. You can just keep using the same tool for different designs.

Drawbacks of CFD –

• CFD has a lack of standardization. If we go to the experimental and fluid dynamics analysis is, there is a lot of standardization and that is very nice for having lot of confidence in comparing different vendors for experimental testing, not so much for CFD.
• We don’t really have a lot of standardization and so it’s really hard to tell one vendor from the other, we don’t come with label saying good vendor or bad vendor. That does lead to lot of misunderstanding and that does lead to lot of misinterpretation and thinking that some vendors are bad.
• CFD is an experimental analysis still and there is no working backwards. You can say what your perfect solution should be and then ask to find out what the exact wind shape is for that. You have to try wind shapes and just keep trying to get your perfect solution.
• Experimental methods can be just as expensive as CFD but CFD may take longer. Experimental method can be done in a day, CFD takes more time than experimental methods.

Author:-

| SevenMentor Pvt Ltd.

© Copyright 2021 | Sevenmentor Pvt Ltd.