# periodic, unsteady, non-constant mass flow rate

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

September 30, 2011, 16:58
#22
Senior Member

Amir
Join Date: May 2009
Location: Montreal, QC
Posts: 739
Blog Entries: 1
Rep Power: 16
Quote:
 Originally Posted by m2montazari hi, consider the inlet thread of cells. we have zero gradient of velocity on it and constant pressure in its plane. so tell me what is the difference between this thread of cells and a duct of the same length having fully developed flow in both inside it? both have zero normal gradient velocity and constant pressure in plane normal to axis. so fixedvalue for pressure and zerogradient for velocity at inlets leads to fully developed flow for internal flows. having zero normal gradient of velocity cannot lead to pressure bc, but the only thing that can lead to pressure bc is constant pressure without any variations in plane. so with considering these variations, we cant use pressure bc and so there would be no doubt about two results with one problem. yours,
Ok, so we cannot use pressure BC to generate fully developed flow. (I checked that)
__________________
Amir

Last edited by Amir; October 1, 2011 at 04:40.

September 30, 2011, 17:10
#23
Senior Member

Amir
Join Date: May 2009
Location: Montreal, QC
Posts: 739
Blog Entries: 1
Rep Power: 16
Quote:
 Originally Posted by m2montazari not a good solution! if you want to have a fully developed flow inside a duct or pipe and you want to model short length, forget about periodic BCs and simply set inlet and outlet as pressure BCs.
__________________
Amir

 October 1, 2011, 04:10 #24 Senior Member   Mohammad Join Date: Feb 2010 Location: Shiraz, Iran Posts: 109 Rep Power: 9 hi, which idea do you mean that I changed? I told that for a fully developed flow you can use pressure BC and I persist on it. as I said the relation between first thread of inlet and a duct in physics, setting pressure-pressure BC results fully developed flow at inlet. think about that again and you will accept my idea. the main reason that I recommend pressure-pressure BC for simple fully developed flow, is you can use udf in it and the other reason is in this example, the periodic BC would result a false velocity and pressure profile after too many iterations(also it may result good profiles in some few iterations but after too many iterations it would result false profiles) because the errors that must exit the domain from outlet, would enter the domain from inlet again and it causes a main error. I persist that pressure-pressure can be used in fully developed flow in a duct. and I persist that presure-pressure cannot be used for flow in a duct containing boxes. but periodic could be use. yours,

October 1, 2011, 04:23
#25
Senior Member

Amir
Join Date: May 2009
Location: Montreal, QC
Posts: 739
Blog Entries: 1
Rep Power: 16
Quote:
 Originally Posted by m2montazari I persist that pressure-pressure can be used in fully developed flow in a duct. and I persist that presure-pressure cannot be used for flow in a duct containing boxes. but periodic could be use.
you can easily check it in FLUENT!!

Bests,
__________________
Amir

 October 1, 2011, 04:49 #26 Senior Member     Amir Join Date: May 2009 Location: Montreal, QC Posts: 739 Blog Entries: 1 Rep Power: 16 Dear Mohammad, we can analytically show that pressure BC is not appropriate for fully developed flow because of pressure inlet BC issue. As you know, there shouldn't be any in-plane pressure variation @ inlet and outlet boundaries in fully developed flow. We don't have any problem with pressure outlet which needs static pressure but @ inlet if we set constant total pressure, static pressure varies in-plane which violate fully developed flow! If you insist on using pressure BC, you'll need an extra UDF which can set a variable total pressure as a function of velocity in order to achieve constant static pressure. Bests, __________________ Amir

 October 1, 2011, 07:59 #27 New Member   mohsen Join Date: Feb 2011 Posts: 13 Rep Power: 8 لطفا با هم دوست باشید بچه ها!

 October 1, 2011, 09:06 #28 Senior Member   Mohammad Join Date: Feb 2010 Location: Shiraz, Iran Posts: 109 Rep Power: 9 hi, to Amir: I didnt say you can use "pressure inlet". I said you can use pressure boundary condition for inlet. so perhaps using pressure outlet for both inlet and outlet results good. but if pressure outlet for inlet boundary results false, you may use udf to get the velocity and by adding dynamic pressure to static pressure, using "pressure inlet" would be a good idea. BTW I'm glad to see that you've changed your idea and accepted the correctness of pressure-pressure boundary for fully developed flows. yours, and to momech: بچه وقتی دو تا مهندس با هم بحث میکنن یک بچه نمیپره وسط!!

October 1, 2011, 11:00
#29
Senior Member

Amir
Join Date: May 2009
Location: Montreal, QC
Posts: 739
Blog Entries: 1
Rep Power: 16
Quote:
 Originally Posted by m2montazari hi, to Amir: I didnt say you can use "pressure inlet". I said you can use pressure boundary condition for inlet. so perhaps using pressure outlet for both inlet and outlet results good. but if pressure outlet for inlet boundary results false, you may use udf to get the velocity and by adding dynamic pressure to static pressure, using "pressure inlet" would be a good idea. BTW I'm glad to see that you've changed your idea and accepted the correctness of pressure-pressure boundary for fully developed flows. yours,

We have a general discussion on appropriateness of two kind of boundary conditions which is available in FLUENT; if we use OpenFOAM, I'll completely agree with you that pressure-pressure BC is proper for fully developed flow because we can set static pressure, but in FLUENT it's not; if you want to write a UDF to use your favorable pressure BC, I can also write another UDF which can generate fully developed flow without restrictions of pressure or periodic BCs! (piece of propaganda for myself )
In contrast of you, I don't used to insist on my wrong suggestion but here, your suggestion doesn't make sense. So in FLUENT the best choice would be PERIODIC BC and not PRESSURE one.
It's better to refer to manual(7.15):
"The second type of periodic BC allows a pressure drop to occur across translationally periodic boundaries, enabling you to model fully-developed periodic flow"
Now it's completely clear and other friends who visit this thread can simply realize who is wrong.

Quote:
 Originally Posted by m2montazari and to momech: بچه وقتی دو تا مهندس با هم بحث میکنن یک بچه نمیپره وسط!!
محمد نظرت چیه این حرکت محسن رو به عنوان یک پیام بازرگانی در نظر بگیریم؟ اینجوری خیلی هم شخصیتش خرد نمیشه
__________________
Amir

Last edited by Amir; October 1, 2011 at 16:03.

 October 2, 2011, 01:19 #31 New Member   mohsen Join Date: Feb 2011 Posts: 13 Rep Power: 8 محمد خون خودت کثیف نکن! ایی فک کرده اینجا مثل فروم اپن فم هست که هرچی غلط غلوط هم بگه هیچکی نیس جوابش بده، امیر تا حالا شرایط مرزی فلوینت رو دیدی یا فقط شنیدی!؟ ا :d

October 2, 2011, 05:53
#32
Senior Member

Amir
Join Date: May 2009
Location: Montreal, QC
Posts: 739
Blog Entries: 1
Rep Power: 16
Quote:
 Originally Posted by m2montazari hi, OK, lets have a conclusion. first of all a problem was defined as a pipe and a periodic flow in it. and you recommend writing a journal and solve the problem for many many flowrates using periodic BC and write each case and data. so having huge space filled with cases and datas just for a very simple problem. then I recommend to change periodic boundary to pressure-pressure boundary(I mentioned "pressure inlet" at nowhere, I just said pressure. even I warn about total pressure in pressure inlet!). then you disagree my recommend because you thought pressure BC doesnt mean something like zeroGradient velocity in OpenFOAM. now you tell that the only problem is that the "pressure inlet" gets total pressure not static pressure(what I have told in first post of my own. it seems you reach this very late!!! ). Ok it seems a victory for me and it is a good victory for start!! then you asked about why fluent gives an option about periodic BC and you thought that you can ignore pressure variation of wake in plane of outlet! now it seems that you agree with me about the sentence I told:"if you ignore these pressure gradient, you've accepted a simple fully developed flow and ignored the box wake effect on boundaries." this can be the second victory!! then you told about total pressure and static pressure and when I told that you can use a simple udf, strangely you said "boundary conditions which is available in FLUENT"!!! I must remind you that udf is one of fluent tools, not one of openfoam tools, not one of cfx tools, and even not one of jet-audio tools!! udf is available in fluent and anyone can use it easily! the first problem was a periodic flow in a pipe. if you dont want to use any udf, how can you solve the problem? with journal?!! then I would say that journal is not a thing that is available in fluent!!! (your third defeat?? ) I think this is enough for me and ac2011 can choose between my way with a simple udf and and your way with lots of gigabytes of hard disk and a very very very hard post processing.
I'm very sorry because it seems that you got angry of this discussion and you don't want to apologize; just try, it's not very difficult.
And in first post, you just wanted to warn about difference between total and static pressure value and you didn't aware of in-plane pressure variation!!! (If you were aware of this, you had mentioned about a UDF)
Finally,I tried to have a general comparison between these 2 BCs in steady state flow which your suggestion couldn't handle such simple case. But I should say that in unsteady flow, if we use journal file, it wouldn't need any huge space, we can save data in any time step we want in journal file!! it seems that you don't know journal file well either.
UDF is an auxiliary program but journal is not a program it's used just for automation! So they are very different for amateurs.
Final suggestion: I can continue this discussion if you think you need it and you can learn more, but if you struggle more, apologizing seems more difficult, I though that this discussion is scientific so I didn't count your defeats! because it would need more than one post!
And about post processing which you said it's very hard!!!! It's very simple; if you have a question about that, you can ask.

Quote:
 Originally Posted by momech محمد خون خودت کثیف نکن! ایی فک کرده اینجا مثل فروم اپن فم هست که هرچی غلط غلوط هم بگه هیچکی نیس جوابش بده، امیر تا حالا شرایط مرزی فلوینت رو دیدی یا فقط شنیدی!؟ ا :d
محسن! من فقط داشتم سعی می کدم که تو رو از یک کارگر افغانی که دسته بیلش شکسته به یک پیام بازرگانی ارتقاء بدم ولی می بینم که لیاقتشو نداشتی. به هر حال اینجا ما کاری واست نمی تونیم انجام بدیم باید بری ابزار فروشی!
__________________
Amir

Last edited by Amir; October 2, 2011 at 06:14.

October 2, 2011, 07:27
#33
New Member

mohsen
Join Date: Feb 2011
Posts: 13
Rep Power: 8
Quote:
سی ای بچو که از صب تا حالا داشته زور میزده که جواب انگلیسی بده؛ بترس از خشم محمد، (به اندازه موهای ریشش که ازموهای کله ی منو تو بیشتره ران گرفته!) میاد یه لقمت میکنه ها از ما گفتن بود بعد گریه نکنی.

 October 2, 2011, 07:29 #34 New Member   mohsen Join Date: Feb 2011 Posts: 13 Rep Power: 8 @ محمد کیف کردی؟ تریپ مشایی ازت حمایت کردما برو حالش ببر!!:d

 October 2, 2011, 07:57 #35 New Member   mohsen Join Date: Feb 2011 Posts: 13 Rep Power: 8 Dear friends i do not know why you guys make it too complicated, this problem has a very simple solution , you just need to be familiar with AVL Fire v8.3 all the best, Mohsen

October 2, 2011, 08:47
#36
Senior Member

Join Date: Feb 2010
Location: Shiraz, Iran
Posts: 109
Rep Power: 9
Quote:
 Originally Posted by momech Dear friends i do not know why you guys make it too complicated, this problem has a very simple solution , you just need to be familiar with AVL Fire v8.3 all the best, Mohsen
hi,
I like this post!!!
and about Amir, it seems that he doesnt want to accept his mistakes. however I even used analytical equations to show him his mistake, but he is grouching about this. so I think its better to finish this discussion and let the thread starter to choose between udf and journal!!
and about Amir's last post, he said some jokes about journal and some other things. it seems he never learns correct solution methods!!
after all I should say that the first problem in this thread is one of "viscous flow theory" coarse project and I and my friends have done this. I did it by openfoam and other students with fluent.
and the last thing I'm so curious about the easiness of post processing of Amir method. can you tell me how you can plot a velocity vs. time in some points in pipe by your method? it is one of two main results of this problem in analytic solution. if you tell me more about that, then I WOULD LEARN IT AND I WOULD ALSO THANK YOU! I know you have learned journal writing and you like using it anywhere you can, even if it is not correct!

امیر برو عامو برای چی چی مردم رو میذاری سر کار؟ میخوای با جورنال بشینه ران بگیره؟
اگه به نصایح من گوش نمیدی خوب حداقل به نصایح محسن گوش بده. بره با فایر ران بگیره بهتره تا با جورنال که!!!

yours,

October 2, 2011, 09:31
#37
Senior Member

Amir
Join Date: May 2009
Location: Montreal, QC
Posts: 739
Blog Entries: 1
Rep Power: 16
First of all, I'll try to answer to the post which may have some valuable information, so I discard momech posts.
Quote:
 Originally Posted by m2montazari hi, and about Amir, it seems that he doesnt want to accept his mistakes. however I even used analytical equations to show him his mistake, but he is grouching about this. so I think its better to finish this discussion and let the thread starter to choose between udf and journal!!
Good idea, it seems that you need more time to understand my suggestion because you don't have enough information about journal file and its capabilities.
Quote:
 Originally Posted by m2montazari and the last thing I'm so curious about the easiness of post processing of Amir method. can you tell me how you can plot a velocity vs. time in some points in pipe by your method? it is one of two main results of this problem in analytic solution. if you tell me more about that, then I WOULD LEARN IT AND I WOULD ALSO THANK YOU!
It's better; just ask what ever you don't know before talking about!
What you want doesn't need any special thing, you can simply use solve->monitor->surface and prepare your files, during run of journal file, these files are also generated!!! Just think that a guy run a case time step by time step! post processing and other thing is not different. (Journal just automate what we want to do)
But I also want to answer in another way, it's very easy and I've done several times even for moving points in dynamic mesh. First create desired points before using journal file. As you know, when we create virtual geometries (here point) you can find its ID in (surface->manage) you'll need that for post processing; then in report panel you can write any variable you want (for point you can use vertex minimum or maximum doesn't have any difference) you have to save results in a file and in next time steps you can simply add data to that file.
Quote:
 Originally Posted by m2montazari I know you have learned journal writing and you like using it anywhere you can, even if it is not correct!
I prefer not to discuss about capabilities of a feature before I know about it. If you've used journal file for different cases; you would agree with me.
Quote:
 Originally Posted by m2montazari امیر برو عامو برای چی چی مردم رو میذاری سر کار؟ میخوای با جورنال بشینه ران بگیره؟ اگه به نصایح من گوش نمیدی خوب حداقل به نصایح محسن گوش بده. بره با فایر ران بگیره بهتره تا با جورنال که!!!
محمد جان این اصلا چیز سختی نیست. از تو تعجب می کنم که فکر می کنی استفاده ازش مشکله! در ضمن اینجا فروم فلونته و افرادی که میان اینجا می خواهند که مشکلشون رو با این نرم افزاری که احیانا کلی واسش هزینه کردن حل کنن!
__________________
Amir

 October 2, 2011, 10:10 #38 Senior Member   Mohammad Join Date: Feb 2010 Location: Shiraz, Iran Posts: 109 Rep Power: 9 hi, you said sth about journal capabilities and then use a simple monitor! I think its a kind of cheat! and its time to end..... آخه امیر این چه لجیه؟ برای چی با جورنال اخه؟ وقتی با فشار میشه، وقتی حتی اگه نخوای از فشار استفاده کنی میشه از massflowinlet استفاده بشه، استفاده از جورنال چه توجیهی داره؟ تازه من اینو چک نکردم ببینم با جورنال میشه یا نه؟ اخه باید طرف تابع بنویسه. تو جورنال تابع مثلا سینوسی بنویسه؟ کلا راه منطقی ای نیست..

October 2, 2011, 10:23
#39
Senior Member

Amir
Join Date: May 2009
Location: Montreal, QC
Posts: 739
Blog Entries: 1
Rep Power: 16
Quote:
 Originally Posted by m2montazari hi, you said sth about journal capabilities and then use a simple monitor! I think its a kind of cheat! and its time to end..... آخه امیر این چه لجیه؟ برای چی با جورنال اخه؟ وقتی با فشار میشه، وقتی حتی اگه نخوای از فشار استفاده کنی میشه از massflowinlet استفاده بشه، استفاده از جورنال چه توجیهی داره؟ تازه من اینو چک نکردم ببینم با جورنال میشه یا نه؟ اخه باید طرف تابع بنویسه. تو جورنال تابع مثلا سینوسی بنویسه؟ کلا راه منطقی ای نیست..
What cheat, look, journal file is not a special thing!!
In another words, think you want to solve this case one time step (1TS) then you change mass flow rate and run another 1TS and so on. We just want to automate this procedure. During this solution, every post-processing file which defined are also generated (I checked that).
Be more unprejudiced; we both use UDF for years and also have experience of using journals, on my opinion, writing a journal is much easier even for me who is familiar with UDF!!!
Additionally, we don't need to write something like sin(t) in journal! You can simply write a simple code (at most 7-8 lines) and generate a journal file with any resolution. I can explain more if it's necessary.

Bests,
__________________
Amir

 October 2, 2011, 10:36 #40 Senior Member     Amir Join Date: May 2009 Location: Montreal, QC Posts: 739 Blog Entries: 1 Rep Power: 16 Dear mohammad, Regardless of these discussions which I think that can help other friends, I prefer to recommend easiest procedures because not all of the users are familiar with C programming and all other UDF issues and I'm sure you conform this. I didn't change my idea because it's completely true and it doesn't violate analytic results (periodic with journal), but if we accept that you were aware of in-plane pressure variation issue @ inlet and using UDF to handle that, you have to announce @ first because other users of this forum may be less experienced than you. So I want you to be more careful in replying posts. Bests, __________________ Amir

 Thread Tools Display Modes Linear Mode

 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 OffTrackbacks are On Pingbacks are On Refbacks are On Forum Rules

 Similar Threads Thread Thread Starter Forum Replies Last Post a.lynchy Fluent UDF and Scheme Programming 22 September 26, 2017 12:50 wildli FLUENT 2 July 8, 2015 11:38 student87 CFX 4 January 2, 2010 05:45 maka OpenFOAM Running, Solving & CFD 3 September 21, 2005 05:18 Min Zhu Main CFD Forum 1 September 29, 1998 15:33

All times are GMT -4. The time now is 03:07.