[fazaghifari]

Hello, I'm a newbie in CFD. Could someone explain me how to model a fluid inside a pipe with cracks in the pipe wall? Do I have to create the cracks in geometry and mesh the cracks one by one or there are some techniques that can be used in this case?
Thanks a lot

[ghorrocks]

What are you trying to model? What do the cracks do? What happens in the cracks? What are you hoping to learn by modelling this?

There are many different ways of approaching this depending on what you are trying to model.

[fazaghifari]

Originally Posted by ghorrocks

What are you trying to model? What do the cracks do? What happens in the cracks? What are you hoping to learn by modelling this?

There are many different ways of approaching this depending on what you are trying to model.

I'm trying to model a heat transfer inside between fluid and a two layers wall. The inner layer of the wall is a concrete and the outer layer is a steel. And the inner wall has some cracks. And I want to obtain the temperature distribution at the outer wall.

[ghorrocks]

Can you post an image of what it looks like? Do the cracks contribute to the heat transfer? If so, how?

[fazaghifari]

Originally Posted by ghorrocks

Can you post an image of what it looks like? Do the cracks contribute to the heat transfer? If so, how?

Well, it look like this:wall.PNG

The blue part is the outer wall, The yellow part is the inner wall. The inner wall has some cracks (not modeled in the geometry). And hot fluid is flowing from the front side to the back side.

For the second question I don't know, that's why I want to do the analysis with CFD. As long as i know that the fluid might go into the cracks and contribute to the heat transfer.

Then, back to the question. Do I have to draw the cracks in the geometry? or are there some techniques that can be used to represent the fluid inside the cracks?

Thank you

[ghorrocks]

Do I have to draw the cracks in the geometry?

That is the key question. To answer that question you need to know whether the cracks are likely to significantly affect the results. And to know whether they are significant, you need to know what flow is likely to occur in it, what the temperatures are likely to be, any other physics (multiphase, radiation, multicomponent etc). Hence my previous questions.

or are there some techniques that can be used to represent the fluid inside the cracks?

There are many methods of modelling cracks. There are simple approaches and complex approaches. And if they are not significant you can ignore them completely.

Based on what I guess could be happening in a device like this I would suspect the cracks are not significant and therefore can be ignored. But I still have little idea of what you are doing and why you think the cracks might be important, so that is nothing more than a guess.

[fazaghifari]

Oooohh, thank you very much!
By the way, would you like to explain or just give me some references to read about the approaches for modelling fluid inside a crack?

Sent from my BV6000 using CFD Online Forum mobile app

[ghorrocks]

There are no specific methods for modelling flow in a crack. It is the same as any flow - you establish the important physics and choose a model appropriate for the physics.

Let me help you. Are the cracks blind or do they go through? How big are they? What flow do you expect to get through them?

[fazaghifari]

Well, the cracks are blind. But I don't know the exact dimensions. I only know the depth is same with the thickness of the inner wall.

Sorry for incomplete information. Because my lecturer just challenged me to model a heat transfer from a fluid to the outer wall, with some cracks at the inner wall, and he said that the cracks dimension is up to me. But this is the real problem, I don't know how big the cracks that i should use.

*sorry for the bad english

Sent from my BV6000 using CFD Online Forum mobile app

[ghorrocks]

If the cracks are blind then the air will be stationary. Stationary air is a good insulator. And I presume the cracks are small relative to the total volume - so this means the total heat transfer will be determined by the overall heat conduction through the steel and concrete. The cracks will not significantly contribute to heat transfer.

So just do a model of the steel, concrete and air with no cracks.

Even better - You can reduce this to a 1D heat transfer model and use analytical heat transfer to solve it. Then you can solve the heat transfer by hand with no need for CFD. This is far easier and is likely to be more accurate.

[fazaghifari]

Thank you very much

Sent from my BV6000 using CFD Online Forum mobile app

[fazaghifari]

Originally Posted by ghorrocks

If the cracks are blind then the air will be stationary. Stationary air is a good insulator. And I presume the cracks are small relative to the total volume - so this means the total heat transfer will be determined by the overall heat conduction through the steel and concrete. The cracks will not significantly contribute to heat transfer.

So just do a model of the steel, concrete and air with no cracks.

Even better - You can reduce this to a 1D heat transfer model and use analytical heat transfer to solve it. Then you can solve the heat transfer by hand with no need for CFD. This is far easier and is likely to be more accurate.

I'm sorry. I need to revise my statement. The cracks have significant influence on the heat transfer, because there are some hot spots region on the outer wall