- Форумы
- Гидродинамика и теплообмен
- ANSYS CFX
- Help me!!! Ошибка при расчёте в CFX
26 октября, 2015 — 13:26
#1
Доброго времени суток уважаемые форумчане, помогите пожалуйста решить проблему начанающему пользователю этой не простой программы Ansys.)))
При расчёте на 68 интерации выдаёт ошибку. Расчётный описывающий файл прикладываю ниже (CFX_002). Также прикладываю саму работу в rar документе.
UTER LOOP ITERATION = 68 CPU SECONDS = 7.384E+02
+———————————————————————+
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| Floating point exception: Overflow |
| |
| |
| |
| |
| |
+———————————————————————+
+———————————————————————+
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| Stopped in routine FPX: C_FPX_HANDLER |
| |
| |
| |
| |
| |
+———————————————————————+
+———————————————————————+
| An error has occurred in cfx5solve: |
| |
| The ANSYS CFX solver exited with return code 1. No results file |
| has been created. |
+———————————————————————+
End of solution stage.
+———————————————————————+
| The following user files have been saved in the directory |
| F:/Bondarenko/blad_pending_tasks/dp0_CFX_Solution/CFX_002: |
| |
| mon |
+———————————————————————+
+———————————————————————+
| Warning! |
| |
| After waiting for 60 seconds, 1 solver manager process(es) appear |
| not to have noticed that this run has ended. You may get errors |
| removing some files if they are still open in the solver manager. |
+———————————————————————+
This run of the ANSYS CFX Solver has finished.
Решатель CFX-Solver вылетает с ошибкой
+———————————————————————+
| *** INSUFFICIENT CATALOGUE SIZE *** |
| |
| ACTION REQUIRED : Increase the file catalogue size. |
| |
| If the situation persists please contact the CFX Customer Helpline |
| giving the following details:- |
| Current catalogue size : 101712 |
+———————————————————————+
Details of error:-
—————-
Error detected by routine MAKLNK
COLDNM = /FLOW/ASSEMBLE/TIME-0/SOLUTION/ZN1/ CNEWNM = /FLOW/GETVAR/SOLN_DPA
CRESLT = FCAT
Current Directory : /FLOW/SOLUTION/TSTEP1/CLOOP1/ZN1/VERTICES
+———————————————————————+
| An error has occurred in cfx5solve: |
| |
| The ANSYS CFX solver exited with return code 1. No results file |
| has been created. |
+———————————————————————+
В Solvere увеличил распределение памяти, вылетает также, с тем же
Current catalogue size : 101712
WinXP, 2,5ГБ
Помогите! Плиз!
— CFX
(https://www.cfd-online.com/Forums/cfx/)
| kola77 | June 29, 2011 16:22 |
The ANSYS CFX solver exited with return code 1
Hello everybody,
I am trying to conduct a simulation of a simple shell-tube heat exchanger, but as soon as I start the run the solver crashes with an error with the following explanation:
Quote:
If the isolated regions do not have the pressure level set either by the boundary conditions or using a reference pressure equation, you may encounter severe robustness problems.
This situation may have arisen because a domain interface was not properly defined during problem setup. Please carefully check the setup.
The solver will stop now and write a results file. The isolated regions can be visualised in CFX Post by making plots of the variable «Isolated Volumes».
If you are sure that the pressure level is set in each isolated fluid region then you can force the solver to turn off this check by setting the expert parameter «check isolated regions = f».
+———————————————————————+ | An error has occurred in cfx5solve: | | | | The ANSYS CFX solver exited with return code 1. | +———————————————————————+
End of solution stage.
I know this usually has something to do with problems of the geometry/mesh, but I couldn’t find any noticeable glitches in mine. There were also no unconnected regions, the the interfaces seem to be correct. Nevertheless, what I have tried is the following:
- I have carefully and several times reviewed my setup
- I have redesigned my model from scratch
- I have refined the mesh
- I have tried to simulate a single pipe inside of a shell
All of the above ended with the same error, yet when I start an old file of another heat exchanger I have simulated a while ago it works fine. I have carefully examined my setup of that one and I can’t fine any differences in the setup. If anybody has ANY idea what might be causing this I would be the happiest man on earth, since I am slowly getting desperate.
Is it possible that my installation of ANSYS is damaged in some way?
Thank you in advance for any potential help!
| ghorrocks | June 29, 2011 18:07 |
The error message tells you exactly what the problem is. If you have two or more fluid regions which are not connected then you need to set a pressure value in each of them. So the easiest way is to make sure each fluid region has a pressure inlet or outlet.
| kola77 | July 1, 2011 11:51 |
That’s the problem. There are no regions which wouldn’t have a boundary condition set. All of them are set to inlet/outlet, wall or interface. The reference pressure is set in each of them. I have tried several configurations, but still the solver keeps crashing.
Would you please be kind enough and take a couple of minutes to check my setup? I’m sure it’s some newbie mistake that can be easily recognized. If so, it’s here: http://www.megaupload.com/?d=N33WZTOW
Check duplicate faces
Hi,
As Glenn said, its definitely problem of isolated fluid regions. Its because of unconnected regions. Sometimes duplicate/multiple faces remain unattended. Hence one or more fluid regions remain unconnected/ isolated. Please check the faces joining the regions. If faces are not connected, you could see the nonconformal mesh or multiple faces at same location.
You may also carry out one solution with the expert parameter «check isolated regions = f». This would enable you to understand the problem region.
Hope this helps.
Regards,
-Sanyo
| ghorrocks | July 2, 2011 06:58 |
Quote:
All of them are set to inlet/outlet, wall or interface.
That is not what I said. A velocity inlet and a mass flow outlet has not defined the pressure. One of the inlet and outlets in each region needs to be a pressure boundary to set the pressure.
| kola77 | July 2, 2011 09:11 |
All of the outlets had a pressure defined.
Anyway, I have managed to get it running, after changing the geometry a bit so i only had one outlet/inlet that splits into 13 later and not several regions defined as either one of them. After doing this it worked without a problem. Thank you for your help guys, much appreciated! Cheers!
| mohw2002 | January 1, 2013 15:48 |
HI FOR EVERY ONE
PLEASE I WORK ON MY GEOMETRY PIPE WITH ONE PERFORATION AND I MESH IT BY HEXA NOT TETRA BECAUSE I WANT TO CALCULATE WALL SHEAR STRESS BECAUSE HEXA GIVE US SMOOTH LINE WHEN WE ARE DRAWING THE BEHAVIOR OF WALL SHEAR AND PRESSURE DROP. WHEN I SENT THE MESH TO CFX IT CANNOT DID THE RESOLVING BECAUSE THE ERROR IS COMING AS BELOW HOW CAN I TREAT THIS PROBLEM. THE SIDE PLANE I PUT IT AFTER ONCE THE ERROR APPEAR. tHE SIDE PLANE IT IS VERTICAL AXIS FOR THE SIDE OF THE PIPE BECAUSE I DRAW HALF PIPE.
ERROR #002100013 has occurred in subroutine Chk_Splane. |
| Message: |
| The symmetry boundary condition requires that the boundary patch |
| mesh faces form a plane or axis. However, face set 6 in the |
| symmetry boundary patch |
| |
| SIDE PLANE |
| |
| is not in a strict plane, which means that at least one of its |
| faces is not parallel to the others. To make the solver run |
| you can do one of the following: |
| |
| (1) Make sure that this symmetry boundary patch is in a plane or |
| axis by checking and regenerating the mesh. |
| (2) If the symmetry boundary patch is an axis rather than a |
| plane, change the tolerance of the degeneracy check by |
| increasing the value of the Solver Expert Parameter |
| ‘degeneracy check tolerance’ (the default value is 1.e-4). |
| (3) Increase the value of the Solver Expert Parameter |
| ‘vector parallel tolerance’ (the default value is 1 deg.). |
| Note that the accuracy of the symmetry condition may decrease |
| as the tolerance is increased. This is because the tolerance |
| is the number of degrees that a mesh face normal is allowed |
| to deviate from the average normal for the entire face set.
| Sanyo | January 2, 2013 07:44 |
Dear mohw2002,
It is a problem with symmetry boundary condition. Instead using symmetry, try using wall type BC with free slip.
Regards,
-Sanyo
| mohw2002 | January 2, 2013 16:11 |
Quote:
Originally Posted by Sanyo
(Post 399735)
Dear mohw2002,
It is a problem with symmetry boundary condition. Instead using symmetry, try using wall type BC with free slip.
Regards,
-Sanyo
thanks for reply
I solved this problem because I made a mistake at ICEM CFD, I chose axis after did the mesh I was trying to solve it but without any success. I returned to ICEM CFD and delete the mesh and rearrange the parts.
Thanks
I have a diffucelet when I mesh by hexa do you have simple procedure or good tutorials?
Regards
| susilo | March 7, 2014 01:30 |
Error cfx solver
If the isolated regions do not have the pressure level set either by the boundary conditions or using a reference pressure equation, you may encounter severe robustness problems.
This situation may have arisen because a domain interface was not properly defined during problem setup. Please carefully check the setup.
The solver will stop now and write a results file. The isolated regions can be visualised in CFX Post by making plots of the variable «Isolated Volumes».
If you are sure that the pressure level is set in each isolated fluid region then you can force the solver to turn off this check by setting the expert parameter «check isolated regions = f».
+———————————————————————+ | An error has occurred in cfx5solve: | | | | The ANSYS CFX solver exited with return code 1. | +———————————————————————+
End of solution stage.
| shelkeprithvi | September 10, 2014 01:11 |
exited with return code 1
3 Attachment(s)
I am also having same problem with my geometry. In pre-CFX when I am giving symmetry as boundary condition at the mid section of the pipe those red arrows are surrounding only three sides of the green rectangle shown in the pictures attached. Can anyone tell me why I am getting this type of error.
| ghorrocks | September 10, 2014 06:08 |
You mean you are getting the isolated regions/pressure level is not set error? Isn’t it obvious? Then you need to set a pressure level.
| Sanyo | September 10, 2014 09:20 |
Be Specific. . . Please!
Prithvi,
Be specific please. It is not clear whether you have an issue of arrows showing in 3 directions only or your solver exited with returned code 1.
If you are asking of arrows, it might be just a representation & actually solver has considered symmetry correctly. Have your solve crashed? If yes try to check physics setup. Give more details. :confused:
Regards,
Sanyo
| umar afzal | January 10, 2017 23:47 |
Quote:
Originally Posted by ghorrocks
(Post 314129)
The error message tells you exactly what the problem is. If you have two or more fluid regions which are not connected then you need to set a pressure value in each of them. So the easiest way is to make sure each fluid region has a pressure inlet or outlet.
hi
please tell me what does it mean?
The ANSYS CFX solver exited with return code 1. No results file has been
created.
when add more nodes and i want use slover it couldnt run just show me abut line . is there have any problem in my BC ? or factor effect ??please
| ghorrocks | January 10, 2017 23:52 |
Please read the error message carefully. The error code does not tell you much, there is no list of definitions for error codes that I am aware of (so I do not know why ANSYS put it in there….). All the information is in the error text.
Your simulation needs the pressure set somewhere. This could be by a boundary condition, an initial condition or some other function. But it needs to be set somehow. Either you did not set any pressure condition or the condition you set is not valid.
| umar afzal | January 10, 2017 23:58 |
problem in cfx solver
hi everyone i hope all of you fine and happy with good health
i have problem during the cfx solver using the problem is that
The ANSYS CFX solver exited with return code 1. No results file has been
created.
when i want to run the solver give me above wearing and i spend much time to solve it by ,y self but still cant any one tell me please how to solve this kind for problem ???is there have any problem in BC ??or nodes
| ghorrocks | January 11, 2017 00:00 |
Did you read my previous post? I have already answered your question.
| umar afzal | January 11, 2017 02:28 |
Quote:
Originally Posted by ghorrocks
(Post 632784)
Did you read my previous post? I have already answered your question.
yeas i read it thank you so much …
| umar afzal | January 11, 2017 02:31 |
Quote:
Originally Posted by ghorrocks
(Post 632781)
Please read the error message carefully. The error code does not tell you much, there is no list of definitions for error codes that I am aware of (so I do not know why ANSYS put it in there….). All the information is in the error text.
Your simulation needs the pressure set somewhere. This could be by a boundary condition, an initial condition or some other function. But it needs to be set somehow. Either you did not set any pressure condition or the condition you set is not valid.
i set pressure on the outlet surface and temp used for inlet i and heat flux on the all heated wall .. is there have any other place to need pressure please i mean with out outlet…..
| ghorrocks | January 11, 2017 16:36 |
Please post an image of what you are modelling and your output file.
|
|
|
#1 |
|
New Member Anonymous Join Date: May 2016 Posts: 4 Rep Power: 9
|
Hi ERROR #002100048 has occurred in subroutine SU_BNEXT. | +———————————————————————+ End of solution stage. +———————————————————————+ This run of the ANSYS CFX Solver has finished. |
|
|
|
|
|
|
#2 |
|
Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,476 Rep Power: 140
|
The error message is very clear, tells you exactly what the problem is and how to fix it. So my suggestion would be to read the error message and do what it says. |
|
|
|
|
|
|
|
#3 |
|
New Member Anonymous Join Date: May 2016 Posts: 4 Rep Power: 9
|
I just need help interpreting what the message means. Yes, I did read the message but still don’t know how to do it. |
|
|
|
|
|
|
|
#4 |
|
Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,476 Rep Power: 140
|
You have a domain where all the nodes lie on the boundary — so no internal nodes. You can only do this under a limited set of circumstances. If you are intending doing a 3D simulation the fix is to remesh so that you have internal nodes in the domain. If you are intending doing another type of simulation then please explain what you are trying to do. |
|
|
|
|
|
|
|
#5 |
|
New Member Anonymous Join Date: May 2016 Posts: 4 Rep Power: 9
|
So basically so far I have drawn a rectangle, then i extruded it. I applied a fine mesh (i tried medium and coarse before same errors), then I created an inlet boundary with velocity of 0.00333 m/s, then I created an outlet with 0 Pa pressure. I also modified the density and viscosity. |
|
|
|
|
|
|
|
#6 |
|
Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,476 Rep Power: 140
|
nodes = mesh. So you need to fix the mesh. If you can’t find the problem post an image of your mesh on the forum. |
|
|
|
|
|
|
|
#7 |
|
New Member Anonymous Join Date: May 2016 Posts: 4 Rep Power: 9
|
||
|
|
|
|
|
|
#8 |
|
Senior Member Maxim Join Date: Aug 2015 Location: Germany Posts: 415 Rep Power: 12
|
This doesn’t seem like a fine mesh to me… Quote: A common cause for this error is a mesh which is only one element thick That’s exactly your problem. You need to refine your mesh so that you have at least some cells that are not directly part of the wall. |
|
|
|
|
22.04.2009
Возможно многие владельцы карт ATI с определённого времени сталкивались с подобными проблемами при отрисовке контуров (Contour) в Ansys CFX-Post:
Проблема связана с новыми картами и драйверами (в рамках примерно 2х лет), но старые драйвера, где проблемы нет, невозможно адекватно использовать с новыми картами.
Чем же она вызвана и можно ли это вылечить?
Чтобы понять суть этой проблемы я воспользовался инструментарием для отладки приложений OpenGL, а именно GLIntercept 0.5. Данный инструмент представляет собой библиотеку-транслятор для вызова функций из настоящей библиотеки OpenGL32.dll. В результате по логу можно понять как работает то или иное приложение, естественно для тех кто знает OpenGL. Получив такой лог для достаточно небольшой геометрии, я решил перенести проблемный код в своё приложение, где я мог бы методом исключения найти проблемные места. Сказано — сделано, мое собственное приложение продемонстрировало те же самые проблемы (http://malefice.fatal.ru/files/GLCFXTest.rar):
В первую очередь была проверена гипотеза с дисплей-листами. Оказалось, что тот же самый код даёт другой результат при непосредственном выводе геометрии (без дисплей-листов):
Примечание:
Дисплей-листы представляет самый простой и эффективный способ вывода геометрии, при компиляции дисплей-листа (в момент создания) данные помещаются в видеопамять (во всяком случае у ATI, у NVidia встречаются некоторые неоднозначности). При необходимости нарисовать содержимое дисплей-листа он вызывается одной командой. Другие стандартные средства OpenGL (за исключением расширений) не помещают данные в видеопамять. Это даёт практически на порядок большую скорость вывода, особенно для больших моделей. Большинство продуктов ANSYS пользуются этим механизмом вывода (за исключением Fluent).
Дальше был проанализирован OpenGL-код. Оказалось, что проблемы заключаются в том, что внутри дисплей-листов (при создании) текущий цвет не устанавливается. И в некоторые моменты получается неопределённым, из-за чего и выплывают треугольники неправильной окраски. Все неправильные треугольники и квада (четырёхугольные примитивы) появляются после разрыва следующего вида:
glBegin(GL_TRIANGLES);
glColor4f(0.000000f,0.131579f,1.000000f,1.000000f);
glNormal3f(0.000000f,1.000000f,0.000000f);
glVertex3f(0.305661f,0.000000f,1.412402f);
…
glNormal3f(0.000000f,1.000000f,0.000000f);
glVertex3f(0.500620f,0.000000f,-0.180434f);
glEnd();
glBegin(GL_TRIANGLES);
glNormal3f(0.000000f,1.000000f,0.000000f);
glVertex3f(0.429916f,0.000000f,-0.129703f);
…
glNormal3f(0.000000f,1.000000f,0.000000f);
glVertex3f(-7.540471f,0.000000f,2.155247f);
glEnd();
Формально в подобных местах разрывать последовательность задания примитивов не обязательно т.к. количество вершин между glBegin()-glEnd() никак не ограничено спецификацией OpenGL. Кроме того, видно, что после последующего glBegin() нет задания цвета (glColor) и все треугольники (квады) задаваемые на этом участе (до следующего вызова glColor) имеют неправильный цвет.
Если удалить такое разделение (glEnd()-glBegin()) на стыке, то всё рисуется правильно.
После анализа ситуации видно, что с одной стороны проблема кроется в драйверах ATI и, частично, в стиле программирования.
Осознание проблемы наметило некоторые пути её решения. Решено было создать библиотеку-корректор действующую по аналогу с GLInterceptor, которая будет служить прослойкой между CFX-Post и системной OpenGL32.dll. Логика решения проста: отслеживать задание цвета внутри дисплей-листов, сохранять текущее значение и при вызове glBegin добавлять недостающий вызов glColor. В качестве основы был взят другой инструмент для отладки OpenGL — GLTrace, из которого было выкинуто всё относящееся к отладочной информации. В результате была получена библиотека http://malefice.fatal.ru/files/opengl32.dll которую можно расположить в каталоге проблемного приложения (моего теста или CFX-Post — ..\CFX\bin\winnt). Ошибки исчезли.
К сожалению, такой способ срабатывал только на 32х-разрядных ОС.
Но, имея на руках данные экспериментов, пользователю под ником Dick удалось вырвать секретную переменную VIEWER_CACHE_COLORS, которая позволяет исправить проблему в версии 12. Надо создать переменную с таким именем и приравнять её нулю.
Действительно, эта переменная решает все проблемы в CFD-Post (т.е. CFX-Post V12).
Анализ перехвата команд OpenGL показывает, что в данном случае просто для каждой вершины задаётся цвет.
Какие неудобства это может вызвать?
Так как геометрия в OpenGL задаётся вершинами то количество данных пропорционально количеству вершин. При задании 4х компонент цвета, 3х компонент нормали и 3х координат, получается 10 чисел одинарной точности на вершину. Для 1 миллиона треугольников это будет: (4 + 3 + 3)*4* 3*1000000/1024^2 ~ 114.4 Мб. Таким образом 4 миллиона треугольников практически полностью займут видеопамять объёмом 512 Мб.
Правда что делать с версией 11?
Вооружившись названием переменной VIEWER_CACHE_COLORS и открыв на просмотр PostGui_ogl.exe, рядышком я обнаружил две других переменных с интригующими названиями VIEWER_FACE_LIST и VIEWER_LINE_LIST. Найденные переменные были сразу же опробованы.
VIEWER_FACE_LIST, значение не важно (лишь бы была, я пробовал 0 и 1), заставляет CFD-Post отрисовывать всё с помощью полигональных примитивов (GL_POLYGON). Каждый полигон задаётся индивидуально и драйвер в этом случае правильно отрисовывает контура. Размер дисплей-листа существенно не вырастает.
VIEWER_LINE_LIST, значение не важно (0,1), заставляет CFD-Post отрисовывать линии индивидуальным заданием отрезка, т.е. вызов glBegin(GL_LINES)-glEnd() для каждой пары вершин. С точки зрения производительности это ничего не меняет т.к. в любом случае glBegin(GL_LINES) означает, что вершины трактуются парами. Размер дисплей-листа такой же. Но линии тоже становятся правильного цвета, в противном случае могли быть линии белого цвета.
Но, что самое важное, версия 11 тоже понимает две последние переменные!
Размер дисплей листа для 1 миллиона треугольников: (3 + 3)*4*3*1000000/1024^2 ~ 68.7 Мб.
Таким образом, на мой взгляд, для решения проблем с CFX-Post (CFD-Post) достаточно добавить две переменные: VIEWER_FACE_LIST и VIEWER_LINE_LIST.
VIEWER_CACHE_COLORS использовать не стоит, если только не будет каких-нибудь проблем с новыми версиями драйверов. Всё проверялось для Catalyst 9.4, но и более ранние версии (например 8.x) должны работать так же.
Как добавить переменные (названия для англоязычной версии XP)?
Открываем свойства My Computer (Properties, по правой кнопке). Щёлкаем на вкладке Advanced (Дополнительно?). Жмём кнопку Environment Variables. Далее имеют место две группы — User Variables и System Variables. В зависимости от того хотите ли вы, чтобы переменная была у данного пользователя или у всех пользователей выбираем место.
Пусть мы хотим установить системную переменную (для всех пользователей).
Жмём кнопку New в группе SystemVariables (нижняя). В диалоге в качестве имени переменной (Variable Name) указываем VIEWER_FACE_LIST, в качестве значения 0 или 1 (без разницы). Аналогично для VIEWER_LINE_LIST. Дело сделано.
Кирилл «deLuther» Сергеев
e-mail:
Этот e-mail адрес защищен от спам-ботов, для его просмотра у Вас должен быть включен Javascript