Linear & quadratic resistance coefficient - Fishing net as membrane/porous media

Burken · November 13, 2017, 08:17

Hello!

I am doing a school project together with two other course participants, where we are supposed to simulate the water flow in a channel to analyse the behavior (deformation, drag, reaction forces etc.) of a net in the channel. To simulate this problem, we consider the net as a membrane/porous medium.

We use three domains, where the middle domain is the membrane/porous medium, but to simulate the behavior of the membrane, we need to know either the permeability and loss coefficient or the linear and quadratic resistance coefficient.

For alternative 1, the permeability should be determined by:

$Q = - \frac{\kappa A}{\mu}\frac{dp}{dx},$
but since we haven't done any experiments yet, a pressure difference in the channel isn't possible to determine.

We are therefore considering the use of the linear and quadratic resistance coefficients instead in alternative 2, and from Ansys help, we have found the expressions:

$C_{R1} = \frac{\mu}{K_{perm}}$ , (Linear resistance coefficient)
and
$C_{R2}=K_{loss}\frac{\rho}{2}$ (Quadratic resistance coefficient),

where $K_{perm}$ is the permeability and $K_{loss}$ the quadratic loss coefficient.

So once again we have the problem with determining the permeability without having any experimental data to evaluate, so the problem we need to find the answer to is: Is it possible to determine the permeability and/or the linear and quadratic resistance coefficients without experimental data? We were supposed to evaluate the results from the CFD analysis with experimental data, but the water pump is broken in the lab so we have started with the CFD in advance.

If we have to find the permeability via experiment, do you set arbitrary values for the coefficients, run the simulations and take the pressure difference to calculate a new value for the constants? In that case, is it possible to take the average pressure at the whole surface before and after the membrane, instead of looking at pressure differences in just points? If this isn't the supposed way to do, please give me some leadings.

Regards, Burken

JuPa · November 13, 2017, 10:14

Hi Burken,

Interesting project!

The obvious question to ask is the pressure difference or the coefficients available in the literature? Has someone already done the experiment or CFD and published the data?

Could you get the answer by simplifying? E.g.
1. Simulate a very small pipe with the net in the middle (physically draw the net, include it in your geometry, and grid it up and run the CFD)
2. Set the Re number in the pipe to be the same order of magnitude as the Re in the channel
3. Get the pressure loss or coefficients across the net
4. Use the results from point 3 to do a full channel simulation with the net as the porous media

Burken · November 14, 2017, 07:12

Quote:

Originally Posted by JuPa

Hi Burken,

Interesting project!

The obvious question to ask is the pressure difference or the coefficients available in the literature? Has someone already done the experiment or CFD and published the data?

Could you get the answer by simplifying? E.g.
1. Simulate a very small pipe with the net in the middle (physically draw the net, include it in your geometry, and grid it up and run the CFD)
2. Set the Re number in the pipe to be the same order of magnitude as the Re in the channel
3. Get the pressure loss or coefficients across the net
4. Use the results from point 3 to do a full channel simulation with the net as the porous media

Thank you for your input JuPa!

We are trying to design a net now instead to simulate our problem. We decided that it would be easier to just ignore the idea of using membrane and just go with a drawn net instead.

sk11 · March 13, 2018, 05:29

I found this interesting! Since you are going to simulate the flow over the net by drawing them, I hope you have some results now.
But to simulate bigger nets for example trawl nets, it is very difficult and computationally expensive when you draw a net. Do you have mind to share some literature if you have, on considering net as a porous membrane?
Thank you.

Burken · March 13, 2018, 09:56

Quote:

Originally Posted by sk11

I found this interesting! Since you are going to simulate the flow over the net by drawing them, I hope you have some results now.
But to simulate bigger nets for example trawl nets, it is very difficult and computationally expensive when you draw a net. Do you have mind to share some literature if you have, on considering net as a porous membrane?
Thank you.

Hi Sk11!

I did create the geometry for the net by drawing it, but as you already know it is computational heavy. The computers at my university was way to slow to mesh a 30*30 cm^2 net so that idea was scrapped.

Instead we created a porous medium where the permeability and porousity was determined by JuPa's method by simulating a single netmask. The flow of the medium had to be converted for the smaller channel (flow to velocity with respect to the smaller cross sectional area). Unfortunately my school does not keep files on the computers older than a certain date, so I don't have any result files anymore. Also, we focused on the behavior of the net (deformation and forces acting upon it) and not so much on the behavior of the flow through the net. The reason was that the PIV pump to visualize the flow at my university was broken so we couldn't compare experimental flow from a real channel with CFD simulations.

If you plan to do the same kind of simulations, i recomend you to do the same step (JuPa's method) with simulations of a single netmask in order to find the permeability for your porous domain.

The best book I know of the topic is "Handbook of hydraulic resistance", written by Idelchik.

https://www.amazon.com/Handbook-Hydr.../dp/156700251X

I don't recall the chapter, but the author has a chapter with a lot of necessary information about net, like permeability. I think he has formulas for calculating variables like permeability, but he calls it something else. Like friction resistance coefficient or something like that.

I hope I have answered your question to your satisfaction. sorry for bad english.

/Burken

sk11 · March 13, 2018, 10:26

Quote:

Originally Posted by Burken

Hi Sk11!

I did create the geometry for the net by drawing it, but as you already know it is computational heavy. The computers at my university was way to slow to mesh a 30*30 cm^2 net so that idea was scrapped.

Instead we created a porous medium where the permeability and porousity was determined by JuPa's method by simulating a single netmask. The flow of the medium had to be converted for the smaller channel (flow to velocity with respect to the smaller cross sectional area). Unfortunately my school does not keep files on the computers older than a certain date, so I don't have any result files anymore. Also, we focused on the behavior of the net (deformation and forces acting upon it) and not so much on the behavior of the flow through the net. The reason was that the PIV pump to visualize the flow at my university was broken so we couldn't compare experimental flow from a real channel with CFD simulations.

If you plan to do the same kind of simulations, i recomend you to do the same step (JuPa's method) with simulations of a single netmask in order to find the permeability for your porous domain.

The best book I know of the topic is "Handbook of hydraulic resistance", written by Idelchik.

https://www.amazon.com/Handbook-Hydr.../dp/156700251X

I don't recall the chapter, but the author has a chapter with a lot of necessary information about net, like permeability. I think he has formulas for calculating variables like permeability, but he calls it something else. Like friction resistance coefficient or something like that.

I hope I have answered your question to your satisfaction. sorry for bad english.

/Burken

Hey Burken,
Thank you for your fast response and the reference. I am more into the flow through and over the trawling net even though the final motive is to accurately predict the behavior of the net. I was aiming to use openFoam for the cfd calculations but I don't know upto what extent the flow through porous media are developed in Openfoam. I will have a look at the book you have referred.
Thanks again
Sarath

Burken · March 15, 2018, 07:51

Quote:

Originally Posted by sk11

Hey Burken,
Thank you for your fast response and the reference. I am more into the flow through and over the trawling net even though the final motive is to accurately predict the behavior of the net. I was aiming to use openFoam for the cfd calculations but I don't know upto what extent the flow through porous media are developed in Openfoam. I will have a look at the book you have referred.
Thanks again
Sarath

We tried to simulate the behavior of the net as a result of the forces acting upon it from the flow. It didn't work though. We were able to analyze the nets behavior in Ansys Mechanic (deformation) using a load function we calculated analytically with a mixture of catenary and FEM, but in the CFD simulations we couldn't capture the nets behavior.

I hope you can acchieve it yourself. I would really like to know how to simulate it if you manage to do it!
/Burken

November 13, 2017, 08:17	Linear & quadratic resistance coefficient - Fishing net as membrane/porous media	#1
Burken New Member Johan Björk Join Date: Nov 2017 Posts: 9 Rep Power: 9	Hello! I am doing a school project together with two other course participants, where we are supposed to simulate the water flow in a channel to analyse the behavior (deformation, drag, reaction forces etc.) of a net in the channel. To simulate this problem, we consider the net as a membrane/porous medium. We use three domains, where the middle domain is the membrane/porous medium, but to simulate the behavior of the membrane, we need to know either the permeability and loss coefficient or the linear and quadratic resistance coefficient. For alternative 1, the permeability should be determined by: $Q = - \frac{\kappa A}{\mu}\frac{dp}{dx},$ but since we haven't done any experiments yet, a pressure difference in the channel isn't possible to determine. We are therefore considering the use of the linear and quadratic resistance coefficients instead in alternative 2, and from Ansys help, we have found the expressions: $C_{R1} = \frac{\mu}{K_{perm}}$ , (Linear resistance coefficient) and $C_{R2}=K_{loss}\frac{\rho}{2}$ (Quadratic resistance coefficient), where $K_{perm}$ is the permeability and $K_{loss}$ the quadratic loss coefficient. So once again we have the problem with determining the permeability without having any experimental data to evaluate, so the problem we need to find the answer to is: Is it possible to determine the permeability and/or the linear and quadratic resistance coefficients without experimental data? We were supposed to evaluate the results from the CFD analysis with experimental data, but the water pump is broken in the lab so we have started with the CFD in advance. If we have to find the permeability via experiment, do you set arbitrary values for the coefficients, run the simulations and take the pressure difference to calculate a new value for the constants? In that case, is it possible to take the average pressure at the whole surface before and after the membrane, instead of looking at pressure differences in just points? If this isn't the supposed way to do, please give me some leadings. Regards, Burken

March 13, 2018, 05:29	Interesting!	#4
sk11 New Member Sarath Join Date: Mar 2017 Location: Spain Posts: 22 Rep Power: 9	I found this interesting! Since you are going to simulate the flow over the net by drawing them, I hope you have some results now. But to simulate bigger nets for example trawl nets, it is very difficult and computationally expensive when you draw a net. Do you have mind to share some literature if you have, on considering net as a porous membrane? Thank you.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Quadratic resistance Coefficient	Ramana	CFX	4	November 26, 2012 17:25
Porous Media Inertial and Resistance	giogio	FLUENT	3	June 21, 2011 19:09
Viscous resistance coefficient in physical v-model	AL	FLUENT	0	December 15, 2008 15:52
Quadratic Resistance Coefficient	Ramana	CFX	0	January 2, 2006 04:43
Flow resistance of porous media	Desheng Kong	Main CFD Forum	3	December 19, 2000 12:07

November 13, 2017, 10:14		#2
JuPa Senior Member Mr CFD Join Date: Jun 2012 Location: Britain Posts: 361 Rep Power: 15	Hi Burken, Interesting project! The obvious question to ask is the pressure difference or the coefficients available in the literature? Has someone already done the experiment or CFD and published the data? Could you get the answer by simplifying? E.g. 1. Simulate a very small pipe with the net in the middle (physically draw the net, include it in your geometry, and grid it up and run the CFD) 2. Set the Re number in the pipe to be the same order of magnitude as the Re in the channel 3. Get the pressure loss or coefficients across the net 4. Use the results from point 3 to do a full channel simulation with the net as the porous media