CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > OpenFOAM > OpenFOAM Programming & Development

Drag coefficient implementation

Register Blogs Members List Search Today's Posts Mark Forums Read

Like Tree1Likes
  • 1 Post By geth03

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   November 30, 2020, 03:44
Default Drag coefficient implementation
  #1
New Member
 
Sumit Peh
Join Date: Oct 2018
Location: Beijing
Posts: 20
Rep Power: 7
Jamessmp23 is on a distinguished road
Hi Foamers

I would like to apply equations of other Drag coefficient in 4 conditions of Renold's number.
[Re<1 // 1<Re<100 // 100<Re<1000 // Re>1000]

SchillerNaumann has 2 conditions of Re which are Re<1000 and Re>1000
Here is the SchillerNaumann code on ~dragModels
I supposed neg(Re - 1000) is this Re<1000 condition and pos0(Re - 1000) is the Re>1000.

Code:
{
volScalarField Re(max(Ur*phase1_.d()/phase2_.nu(), scalar(1.0e-3)));
volScalarField Cds
(
neg(Re - 1000)*(24.0*(1.0 + 0.15*pow(Re, 0.687))/Re)
+ pos0(Re - 1000)*0.44
);

return 0.75*Cds*phase2_.rho()*Ur/phase1_.d();
}
Is there a way to write code for more than 2 conditions ?

As my imagination, I guess maybe my case would be
Code:
{
volScalarField Re(max(Ur*phase1_.d()/phase2_.nu(), scalar(1.0e-3)));
volScalarField Cds
(
neg(Re - 1)*(Equation1)
+ neg1(Re - 100)*(Equation2)
+ neg100(Re - 1000)*(Equation3)
+ pos0(Re - 1000)*(Equation4)
);

return 0.75*Cds*phase2_.rho()*Ur/phase1_.d();
}
Please confirm if i'm right and please correct if i'm wrong.

Thanks in advanced

James
Jamessmp23 is offline   Reply With Quote

Old   November 30, 2020, 05:04
Default
  #2
Senior Member
 
Join Date: Dec 2019
Location: Cologne, Germany
Posts: 355
Rep Power: 8
geth03 is on a distinguished road
if the compiler does not complain about your code, you still need to verify it with easy to check simulation, maybe a quadratic pipe with one cell depth and height and 30 cells long. set different inlet velocities for both phases and use the calculator in paraview to calculate your drag coeff., also you can output the drag coeff. with openfoam, so you can cross check.
geth03 is offline   Reply With Quote

Old   December 7, 2020, 23:02
Default
  #3
New Member
 
Sumit Peh
Join Date: Oct 2018
Location: Beijing
Posts: 20
Rep Power: 7
Jamessmp23 is on a distinguished road
Quote:
Originally Posted by geth03 View Post
if the compiler does not complain about your code, you still need to verify it with easy to check simulation, maybe a quadratic pipe with one cell depth and height and 30 cells long. set different inlet velocities for both phases and use the calculator in paraview to calculate your drag coeff., also you can output the drag coeff. with openfoam, so you can cross check.
Bad news, the compiler did complain about neg1, neg100.
So i tried this instead. Everything go well when compile but not sure its correct or not.

+ (1 - Re -100)
+ (100 - Re - 1000)
Jamessmp23 is offline   Reply With Quote

Old   December 9, 2020, 04:52
Default
  #4
Senior Member
 
Join Date: Dec 2019
Location: Cologne, Germany
Posts: 355
Rep Power: 8
geth03 is on a distinguished road
ok, let us first clarify what neg() and pos() mean:

neg(scalar s) checks if the value of s i negative, if yes, it will return
the value 1. so if neg(Re-1000) is negativ, than you will get a 1 back.

the same goes for pos(s) or pos0(s). pos(s): is s larger than zero,
pos0(s): is s larger than zero or 0. if yes, return 1.

which means, if you have neg(s) and pos0(s) in your code, it will only return
for one case 1, otherwise 0.

if you have many of that, as you did, it can and will be the case, that you will return more 1 than expected.

let me give an example:
if Re = 10 and you have

+ neg(1 - Re -100)
+ neg(100 - Re - 1000)

like you did, you will have

+1
+1

as return values, because in both cases, the if-check will give "true".
so, in that case, you will overpredict the drag coefficient.

as a workaround, i suggest you loop over all your cells and depending
on the Re, you can check in which value area Re is for each cell and save
Cds accordingly. afterwards you can insert that one in the place where this short formula would normally be.

i hope things are clear now.
geth03 is offline   Reply With Quote

Old   December 10, 2020, 04:50
Default
  #5
New Member
 
Sumit Peh
Join Date: Oct 2018
Location: Beijing
Posts: 20
Rep Power: 7
Jamessmp23 is on a distinguished road
Quote:
Originally Posted by geth03 View Post
ok, let us first clarify what neg() and pos() mean:

neg(scalar s) checks if the value of s i negative, if yes, it will return
the value 1. so if neg(Re-1000) is negativ, than you will get a 1 back.

the same goes for pos(s) or pos0(s). pos(s): is s larger than zero,
pos0(s): is s larger than zero or 0. if yes, return 1.

which means, if you have neg(s) and pos0(s) in your code, it will only return
for one case 1, otherwise 0.

if you have many of that, as you did, it can and will be the case, that you will return more 1 than expected.

let me give an example:
if Re = 10 and you have

+ neg(1 - Re -100)
+ neg(100 - Re - 1000)

like you did, you will have

+1
+1

as return values, because in both cases, the if-check will give "true".
so, in that case, you will overpredict the drag coefficient.

as a workaround, i suggest you loop over all your cells and depending
on the Re, you can check in which value area Re is for each cell and save
Cds accordingly. afterwards you can insert that one in the place where this short formula would normally be.

i hope things are clear now.
Hi geth03,

Thanks for your reply, you made me clear about selecting case for different Reynold's number.

I also found a Lain.C in /drags, it use the code like this so I changed it.

Code:
#include "Lain.H"
#include "phasePair.H"
#include "addToRunTimeSelectionTable.H"

// * * * * * * * * * * * * * * Static Data Members * * * * * * * * * * * * * //

namespace Foam
{
namespace dragModels
{
    defineTypeNameAndDebug(Lain, 0);
    addToRunTimeSelectionTable(dragModel, Lain, dictionary);
}
}


// * * * * * * * * * * * * * * * * Constructors  * * * * * * * * * * * * * * //

Foam::dragModels::Lain::Lain
(
    const dictionary& dict,
    const phasePair& pair,
    const bool registerObject
)
:
    dragModel(dict, pair, registerObject)
{}


// * * * * * * * * * * * * * * * * Destructor  * * * * * * * * * * * * * * * //

Foam::dragModels::Lain::~Lain()
{}


// * * * * * * * * * * * * * * * Member Functions  * * * * * * * * * * * * * //

Foam::tmp<Foam::volScalarField> Foam::dragModels::Lain::CdRe() const
{
    volScalarField Re(pair_.Re());

    return
        neg(Re - 1.5)*16.0
      + pos0(Re - 1.5)*neg(Re - 80)*14.9*pow(Re, 0.22)
      + pos0(Re - 80)*neg(Re - 1500)*48*(1 - 2.21/sqrt(max(Re, SMALL)))
      + pos0(Re - 1500)*2.61*Re;
}
As your example, why Re = 10 get +1 result ? It isn't in the range of neg(100 - Re - 1000).

According to my understanding, we can write code by this way :
pos0(Re - 1)*neg(Re - 100)
for a condition of 1 < Re < 100
Jamessmp23 is offline   Reply With Quote

Old   December 10, 2020, 06:28
Default
  #6
Senior Member
 
Join Date: Dec 2019
Location: Cologne, Germany
Posts: 355
Rep Power: 8
geth03 is on a distinguished road
if Re = 10, then neg(100 - Re - 1000) = neg(-910).
neg(-910) will check if -910 is a negative value, and return +1
if true, which it is, in that case.

as in the Lain.C - case, you can use pos0(s) neg(s1) in combination
so if your Re falls between two transition values, both return
+1 and +1, otherwise +1 and 0, which makes +1 x 0 = 0, and
so other Re-ranges are neglected.

i didn't think about combining those pos() and neg(), that is way smarter
then looping over all cells.
Jamessmp23 likes this.
geth03 is offline   Reply With Quote

Reply

Tags
cfd, code, drag, drag coeffcients, openfoam

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Drag Force Ratio for Flat Plate Rob Wilk Main CFD Forum 40 May 10, 2020 05:47
problem with saving drag coefficient colopolo FLUENT 5 April 12, 2013 11:59
Calculation of Drag Coefficient, Help Please teek22 CFX 1 April 26, 2012 19:41


All times are GMT -4. The time now is 08:22.