 |
 |
 |
|
|
|
[Sponsors] | ||||
August 2, 2020, 14:06
|
Entry 'U' not found in dictionary ""
|
#1 |
|
Member
Join Date: Jun 2020
Posts: 37
Rep Power: 5  |
Hello all,
I have recently started to learn OF. I have made a geometry using topoSet and the conditions using changeDictionary. The solver is porousSimpleFoam. when I run Allrun, the following error appears: Create time Create mesh for time = 0 SIMPLE: no convergence criteria found. Calculations will run for 10 steps. Reading field p Reading field U Reading/calculating face flux field phi Selecting incompressible transport model Newtonian Selecting turbulence model type laminar Selecting laminar stress model Stokes No MRF models present No finite volume options present Creating porosity model list from porosityProperties Porosity region porosity1: selecting model: DarcyForchheimer creating porous zone: InternalSpace origin: (0 0 0) e1: (1 0 0) e2: (0 1 0) local bounds: (0.0003007 0.000259944 0.00025) Using pressure explicit porosity Starting time loop Time = 0.01 --> FOAM FATAL IO ERROR: Entry 'U' not found in dictionary "" From function const Foam::dictionary& Foam::dictionary::subDict(const Foam::word&, Foam::keyType:  ption) constin file db/dictionary/dictionary.C at line 532. FOAM exiting Any help would be appreciable Regards |
|
|
|
|
|
August 5, 2020, 18:15
|
|
#2 |
|
New Member
Anup Singh
Join Date: Mar 2020
Posts: 22
Rep Power: 6 |
It seems like you are missing entries for U in your dictionary files ...
Supposedly in one of the source or include dictionary files like transport properties.. |
|
|
|
|
|
|
August 6, 2020, 03:07
|
|
#3 |
|
Member
Join Date: Jun 2020
Posts: 37
Rep Power: 5 |
Thank you for your reply.
I think the the transport dictionary does not include any information about U. In my problem it is as below: transportModel Newtonian; nu 1e-6; rho 1000; I don't know which dictionaries i should search for missed U? The BC of each region is put in changeDictionaryDict, and the P and U dictionaries exist in 0 folder. Regards |
|
|
|
|
|
|
August 6, 2020, 04:36
|
|
#4 |
|
New Member
Anup Singh
Join Date: Mar 2020
Posts: 22
Rep Power: 6 |
BC are defined in the 0 folder for the considered variable (in your case P and U) or you can modify them in solver as per your coding convenience. I have no idea how are you defining your BC elsewhere . ...
It seems like you are using some OF based source includes in your solution which some times require some feedback from the current data for adjustment of sources per iteration.. Try to look carefully if your are missing something there... |
|
|
|
|
|
|
August 7, 2020, 03:15
|
|
#5 |
|
Member
Join Date: Jun 2020
Posts: 37
Rep Power: 5 |
The error still exists. Any help might be useful.
Thanks |
|
|
|
|
|
|
August 7, 2020, 04:30
|
|
#6 |
|
New Member
Anup Singh
Join Date: Mar 2020
Posts: 22
Rep Power: 6 |
If its possible for you to upload your case along with solver, then I can try to take a look.
|
|
|
|
|
|
|
August 8, 2020, 10:54
|
|
#7 |
|
Member
Join Date: Jun 2020
Posts: 37
Rep Power: 5 |
Dear Anup Singh
Thank you very much for your response. My case consists of 4 regions including 3 partial cylinder full of fluid and a half cubic porous media in which the fluid is penetrated from 2 of the cylinders and removed to the third one. for the internal media i set the following BC in changeDictionaryDict Code:
dictionaryReplacement
{
U
{
internalField uniform (0 0 0);
boundaryField
{
".*"
{
type zeroGradient;
}
InternalSpace_to_PA1
{
type porousBafflePressure;
patchType cyclic;
jump uniform -1000;
D 2e20;
I 0.9e10;
length 250e-6;
value uniform 0;
}
InternalSpace_to_PA2
{
type porousBafflePressure;
patchType cyclic;
jump uniform -1000;
D 2e20;
I 0.9e10;
length 250e-6;
value uniform 0;
}
InternalSpace_to_PV
{
type porousBafflePressure;
patchType cyclic;
jump uniform 1000;
D 2e20;
I 0.9e10;
length 250e-6;
value uniform 0;
}
}
}
epsilon
{
internalField uniform 0.01;
boundaryField
{
".*"
{
type epsilonWallFunction;
value uniform 0.01;
}
}
}
k
{
internalField uniform 0.1;
boundaryField
{
".*"
{
type kqRWallFunction;
value uniform 0.1;
}
}
}
p_rgh
{
internalField fixedValue;
value 10000;
boundaryField
{
".*"
{
type fixedFluxPressure;
}
InternalSpace_to_PA1
{
type fixedGradient;
gradient unifoem -1000;
}
InternalSpace_to_PA2
{
type fixedGradient;
gradient uniform -1000;
}
InternalSpace_to_PV
{
type fixedGradient;
gradient uniform 1000;
}
}
}
p
{
internalField uniform 10000;
boundaryField
{
".*"
{
type calculated;
value uniform 0;
}
}
}
}
and for the partial cylinders the changeDictionaryDict is as below except that for the output conduit the values are opposed: Code:
dictionaryReplacement
{
U
{
internalField uniform (0 0 18e-6);
boundaryField
{
".*"
{
type fixedValue;
value $internalValue;
}
sides
{
type zeroGradient;
}
PA1_to_InternalSpace
{
type porousBafflePressure;
patchType cyclic;
jump uniform 1000;
D 2e20;
I 0.9e10;
length 250e-6;
value uniform 0;
}
}
}
epsilon
{
internalField uniform 0.01;
boundaryField
{
".*"
{
type epsilonWallFunction;
value uniform 0.01;
}
}
}
k
{
internalField uniform 0.1;
boundaryField
{
".*"
{
type kqRWallFunction;
value uniform 0.1;
}
}
}
p_rgh
{
internalField calculated;
value uniform 0;
boundaryField
{
".*"
{
type fixedFluxPressure;
}
PA1_to_InternalSpace
{
type fixedGradient;
gradient unifoem 1000;
}
}
}
p
{
internalField calculated;
value uniform 0;
boundaryField
{
".*"
{
type calculated;
value uniform 0;
}
}
}
}
Code:
solvers
{
p
{
solver GAMG;
tolerance 1e-03;
relTol 0.1;
smoother GaussSeidel;
nCellsInCoarsestLevel 20;
}
"(U|k|epsilon)"
{
solver smoothSolver;
smoother symGaussSeidel;
nSweeps 2;
tolerance 1e-03;
relTol 0.1;
}
}
SIMPLE
{
nUCorrectors 4;
nNonOrthogonalCorrectors 4;
convergenceCriterion 1.0e-5;
pRefCell 0;
pRefValue 0;
residualControl
{
p 1e-4;
U 1e-4;
}
}
relaxationFactors
{
fields
{
p 0.3;
}
equations
{
U 0.3;
k 0.7;
epsilon 0.7;
}
}
Last edited by efsolat; August 10, 2020 at 15:56. |
|
|
|
|
|
|
August 10, 2020, 12:23
|
|
#8 |
|
Super Moderator
Tobias Holzmann
Join Date: Oct 2010
Location: Tussenhausen
Posts: 2,708
Blog Entries: 6
Rep Power: 51   |
Please !!! use code tags for formatting your post. Please correct your above post and add the code tags.
If you are using 3 regions you need to have the following structure: Code:
0 |- fluidRegion1 | |-> U | |-> p | |-> ... |- fluidRegion2 | |-> U | |-> p | |-> ... |- solidRegion1 | |-> T | |-> p |- solidRegion2 | |-> T | |-> p ...
__________________
Keep foaming, Tobias Holzmann |
|
|
|
|
|
|
August 10, 2020, 16:05
|
|
#9 |
|
Member
Join Date: Jun 2020
Posts: 37
Rep Power: 5 |
Thank you for reply. I have edited the previous post.
it seems that the changeDictionaryDict is not recognized by the solver and i don't know whether it is required or not to add something to the controlDict or somewhere else for the changeDictionaryDict to be recognized. Regards |
|
|
|
|
|
|
August 11, 2020, 04:05
|
|
#10 |
|
Member
Join Date: Jun 2020
Posts: 37
Rep Power: 5 |
Hello all
Now i can take a step farther and say that i almost found out where the problem originates from but don't know how to resolve it! As i described a few posts earlier, my case includes 4 zones, 3 fluid conduits and a porous media. For all zones, the Navier-Stokes or+Darcy Eqs are to be solved. therefore i concluded that just 1 region exists in my case and i didn't used regionProperties. when running Code:
changeDictionary -region <zone> Code:
FOAM Warning :
From function int main(int, char**)
in file changeDictionary.C at line 709
Requested field to change dictionaryReplacement does not exist in "/home/efsol/Desktop/case/0/<zonenam>"
I tried Code:
changeDictionary <zone> Any help/suggestion is kindly appreciated Regards |
|
|
|
|
|
|
August 17, 2020, 11:19
|
|
#11 |
|
Senior Member
Yann
Join Date: Apr 2012
Location: France
Posts: 1,064
Rep Power: 26 |
Hello efsolat!
Which solver are you using? Regards, Yann |
|
|
|
|
|
August 17, 2020, 11:57
|
|
#12 |
|
Super Moderator
Tobias Holzmann
Join Date: Oct 2010
Location: Tussenhausen
Posts: 2,708
Blog Entries: 6
Rep Power: 51 |
As it is written, you need the files in
Code:
0/yourZone/ in which you should have your boundary fields such as, T, U, p, k, omega ... Code:
0/yourZone/U 0/yourZone/T . . .
__________________
Keep foaming, Tobias Holzmann |
|
|
|
|
|
|
August 23, 2020, 01:40
|
|
#14 | |
|
Member
Join Date: Jun 2020
Posts: 37
Rep Power: 5 |
Quote:
Hi Tobi Thanks for the reply. I had made the files you mentioned in all 0/zones and i found and resolved the changeDictionary problem. now, changeDictionary works properly and replaces the BCs in the related files of 0 folder. But the error 'Entry 'U' not found in dictionary ""' remains unchanged!! Regards |
||
|
|
|
||
|
August 24, 2020, 04:04
|
|
#15 |
|
Senior Member
Yann
Join Date: Apr 2012
Location: France
Posts: 1,064
Rep Power: 26 |
Hi Efsolat,
I think there is a confusion about what is a region in OpenFOAM. The example given by Tobias with the "0/zone/" structure is only valid for multiregion solvers (like the chtMultiRegion(Simple)Foam solvers). But AFAIK, porousSimpleFoam is not a multiregion solver and your variable files should be located in 0/ directory. You can set your BC directly in the files located in 0/, or you can use the changeDictionary utility to modify these files. This is a trick used in some tutorials, but it is not mandatory and it's perfectly fine to manually edit the files in 0/. Could you post a case here so we can have a look at what might cause your problem? (ideally a light version of your case just to reproduce the error) Regards, Yann |
|
|
|
|
|
|
October 9, 2020, 14:56
|
|
#16 | |
|
Member
Join Date: Jun 2020
Posts: 37
Rep Power: 5 |
Quote:
Thank you very much for your reply. As you said my case is a multi zone model in which one zone is porous and the others are fluid filled. I do not know whether it could be considered muiti zone or not. I had defined the region properties as below: Code:
regions
(
fluid (zone1 zone2 zone3 zone4)
solid ()
);
The other thing is that i firstly defined a block and then divided it into four zones using topoSet utility. Each zone has different boundary conditions i.g the top face is divided into four and each partition has its own boundary condition which i put them in the 0/zone/U,.... I don't know how to set the BCs of zones if i have to define all in 0/U,.... I used the recommendation of this thread (Two porous zone buoyantSimpleFoam) and defined the BCs of the initial block in the files located in 0/. The BCs of each zone is defined in 0/zone/. I am not sure about it. I have edited my case as a single region and now it runs. This is the log result up to the first time step but it goes to the end of time loop: Code:
Create time
Create mesh for time = 0
SIMPLE: convergence criteria
field p tolerance 1e-08
field U tolerance 1e-08
Reading field p
Reading field U
Reading/calculating face flux field phi
Selecting incompressible transport model Newtonian
Selecting turbulence model type laminar
Selecting laminar stress model Stokes
No MRF models present
No finite volume options present
Creating porosity model list from porosityProperties
Porosity region porosity1:
selecting model: DarcyForchheimer
creating porous zone: InternalSpace
origin: (0 0 0) e1: (1 0 0) e2: (0 1 0)
local bounds: (0.0003007 0.000259944 0.00025)
Using pressure implicit porosity
Starting time loop
Time = 0.1
GAMG: Solving for p, Initial residual = 7.43343e-07, Final residual = 1.80714e-07, No Iterations 1000
GAMG: Solving for p, Initial residual = 2.9989e-07, Final residual = 9.36922e-08, No Iterations 414
GAMG: Solving for p, Initial residual = 2.03912e-07, Final residual = 9.76252e-08, No Iterations 153
GAMG: Solving for p, Initial residual = 1.90802e-07, Final residual = 9.39956e-08, No Iterations 8
GAMG: Solving for p, Initial residual = 1.67922e-07, Final residual = 8.89795e-08, No Iterations 114
GAMG: Solving for p, Initial residual = 1.67383e-07, Final residual = 9.12886e-08, No Iterations 133
GAMG: Solving for p, Initial residual = 1.63465e-07, Final residual = 8.78786e-08, No Iterations 9
GAMG: Solving for p, Initial residual = 1.70195e-07, Final residual = 9.91503e-08, No Iterations 2
GAMG: Solving for p, Initial residual = 1.63968e-07, Final residual = 9.8325e-08, No Iterations 216
time step continuity errors : sum local = 4.1674e-17, global = 3.07302e-18, cumulative = 3.07302e-18
ExecutionTime = 694.54 s ClockTime = 698 s
Any help is greatly appreciated. Kindly Redards Efsolat |
||
|
|
|
||
|
October 10, 2020, 07:06
|
|
#17 |
|
Senior Member
Yann
Join Date: Apr 2012
Location: France
Posts: 1,064
Rep Power: 26 |
Hello Efsolat,
porousSimpleFoam is not a multi region solver. the regionProperties file is meant to be used with the chtMultiRegion solvers and it is useless with porousSimpleFoam. To model porous zone, we use cellZone, which is basically a group of cells in your domain where you want to apply a source term to simulate a porous media using a model such as Darcy-Forchheimer. You don't need to define boundary conditions at the frontier of your zone because there is no boundary here. It's just a set of cells, the only difference is the source term which is applied to it. You just have to create as many cellZone as you need (using snappyHexMesh or toposet for instance) and to define the porous model you want to apply on each zone (in constant/porosityProperties if you are using porousSimpleFoam, or in fvOptions for other solvers) That's it! Cheers, Yann |
|
|
|
|
|
|
October 10, 2020, 08:35
|
|
#18 | |
|
Member
Join Date: Jun 2020
Posts: 37
Rep Power: 5 |
Quote:
Dear Yann Thank you very much for your time and reply. Although cellZones are not necessarily regions but in the case i am to model it seems that different zones should be separated regions because experimentally there are different boundary conditions in the interface of zones. My simplified case is like this: Consider a uniformly fluid filled cylinder with permeable wall that is connected to a porous medium and there is a fixed value pressure gradient in their interface. At the other extreme of the porous medium there is an other cylinder with a different fixed value pressure gradient. Moreover, in the first level i defined a uniform velocity in the the input of the cylinders but different for each one, but at last they should be time dependent. Therefore i separated the zones using splitMeshRegions -cellZones -overwrite I don't know is it true to define zones, if not how can i define interfaces pressure gradients and input velocity only for cylinder zones? Best Regards |
||
|
|
|
||
|
October 11, 2020, 03:45
|
|
#19 |
|
Senior Member
Yann
Join Date: Apr 2012
Location: France
Posts: 1,064
Rep Power: 26 |
Dear Efsolat,
I'm not sure to understand what you are trying to achieve. Could you post a drawing describing your domain and the boundary conditions you want to apply to it? What I know for sure is that porousSimpleFoam is not designed to work with multiple regions. You won't be able to use it on meshes splitted with splitMeshRegions. Best regards, Yann |
|
|
|
|
|
|
October 12, 2020, 06:36
|
|
#20 | |
|
Member
Join Date: Jun 2020
Posts: 37
Rep Power: 5 |
Quote:
Dear yann Thanks for your reply Here i posted an image of my case geometry. The regions shown in red and blue and yellow are fluid filled and the green volume is porous. Best Regard |
||
|
|
|
||
 |
|
|
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| error in fireFoam, when running the case wallFireSpread2D | zhoubiao1088 | OpenFOAM Running, Solving & CFD | 9 | February 1, 2018 18:45 |
| [Other] Mesh Importing Problem | cuteapathy | ANSYS Meshing & Geometry | 2 | June 24, 2017 05:29 |
| 8x icoFoam speed up with Cufflink CUDA solver library | kmooney | OpenFOAM Running, Solving & CFD | 42 | November 6, 2012 11:37 |
| missing vtf3.h BPatch.h papi.h | linch | OpenFOAM Installation | 41 | July 24, 2012 14:45 |
| OpenFOAM 1.7.1 installation problem on OpenSUSE 11.3 | flakid | OpenFOAM Installation | 16 | December 28, 2010 08:48 |
3Likes
Linear Mode
