[CFD] Relaxation in CFD (Part 3) - Pseudo Transients
Рет қаралды 7,740
Күн бұрынAn introduction to relaxation and how it can be used to help improve convergence in CFD. Time stamps:
0:00 Introduction
0:52 Example Solution
5:36 Recap
9:10 Pseudo-Transient Derivative
14:23 Implicit Euler Approach
16:00 Matrix Form
18:18 Comparison
20:51 Variable Relaxation
23:52 Optimum Relaxation
26:45 Next Lecture
28:07 Summary
29:41 Outro
#relaxation #fluidmechanics101 #cfd
============================================
Some useful references
============================================
1. Ferziger & Peric, 'Computational Methods for Fluid Dynamics', 3rd Edition, Springer-Verlag, Berlin, Heidelberg, New York, 2002.
www.amazon.co.uk/Computationa... [Page 118, 149]
2. S. Patankar, 'Numerical Heat Transfer and Fluid Flow', 1st Edition, McGraw-Hill, 1980. [Page 67]
2. ANSYS Fluent 12.0 User manual (18.4.4 Steady-State Iterative Algorithm)
www.afs.enea.it/project/neptu...
3. OpenFOAM: User Guide v2012. Under relaxation
www.openfoam.com/documentatio...
============================================
Want to learn more?
============================================
Grab a copy of my CFD Fundamentals Course (for beginners):
www.udemy.com/course/computat...
Learn how to write your own CFD code in MATLAB and python (for intermediates):
dr-aidan-wimshurst-s-school.t...
Learn how I draw my figures and diagrams in Inkscape (for everyone):
dr-aidan-wimshurst-s-school.t...
============================================
Did you like the video?
============================================
Download the lecture slides from my website:
www.fluidmechanics101.com/pag...
Buy me a coffee to say thanks:
www.buymeacoffee.com/NKEZrhvg0
Support the channel on Patreon (and get useful extras for your CFD studies):
/ fluidmechanics101
============================================
Donations:
============================================
1) PayPal
www.paypal.me/fluidmechanics101
==================================
Disclaimer
==================================
The methods, algorithms, equations, formulae, diagrams and explanations in this talk are for educational and demonstrative purposes only. They should never be used to analyse, design, accredit or validate real scientific / engineering / mathematical structures and flow systems. For such applications, appropriate trained, qualified and accredited (SQEP) engineers / scientists should be consulted along with the appropriate documentation, procedures and engineering standards. Furthermore, the information contained within this talk has not been verified, peer reviewed or checked in any way and is likely to contain several errors. It is therefore not appropriate to use this talk itself (or any of the algorithms, equations, formulae, diagrams and explanations contained within this talk) as an academic or technical reference. The reader should consult the original references and follow the verification and validation processes adopted by your company / institution when carrying out engineering calculations and analyses. Fluid Mechanics 101 and Dr. Aidan Wimshurst are not accountable or liable in any form for the use or misuse of the information contained in this talk beyond the specific educational and demonstrative purposes for which it was intended.
@winstonong9593 2 жыл бұрын
Truly a gem of a channel - Dr. Aidan has a knack for demystifying seemingly complex CFD concepts!
@jessbuildstech 2 жыл бұрын
First! Thank you Dr Aidan for another great video, I have bookmarked this in my browser thank you
@cmai7032 3 ай бұрын
Thank you so much for helping me understand CFD!
@thomaslee6498 2 жыл бұрын
O Captain! My Captain! I am #301. I've been waiting for this relaxation (part 3) for long enough! Thank you Sir!
@minjoong6687 Жыл бұрын
Thank God and you for uploading this lecture series! You made me a REAL CFD ENGINEER from button clicking technician.
@forestshaner8994 Ай бұрын
This is a fantastic lecture series. Well done!
@patricksouzalima384 2 жыл бұрын
Amazing video! A Brazilian student thanks
@sergniko 2 жыл бұрын
Of course we are looking forward to the next talk! This talk make CFD much more clearer.
@xxl8802 Жыл бұрын
Politicians use ideologies to separate people, while countless brave man like you dissolve their conspiracies by simply giving universal knowledge to the world.
@mo_lan1323 10 ай бұрын
man, this is so great. I thought i underestand CFD. thanks alot ❤
@stefanomonti6581 Жыл бұрын
I am facing these pseudo-transient solver in fluent and your videos are actually helping me out to understand what these values represent. I think I will follow all your videos cause they are so much interesting and enjoyable. Good work!
@carlosmoreno4682 2 жыл бұрын
These lectures are worth sooo much... Thank you for sharing your knowledge as well as the references, it's nice to finally have someone shedding light to the great unknowns of the CFD. Regards from Spain 😄
@schrummy14 2 жыл бұрын
Thanks for the video. Looking forward to the next video's.
@apocalypt0723 2 жыл бұрын
Thank you so much for yet another amazing video.
@nrc9275 2 жыл бұрын
Excellent Dr.👌 wow speechless !
@atakan716 11 ай бұрын
Another eye-opening lecture, thank you so much!
@jk463 2 жыл бұрын
Amazing once again! It is rather confusing to think that having different time steps for each cell is physically able, and quite amazing that still steady solvers with relaxation show appropriate answers for certain(obviously steady) cases. Truly this series is mind-blowing, providing much more fundamental and wider understanding of CFD simulation solvers. Waiting for part 4 already. Thanks a lot for the video.
@fluidmechanics101 2 жыл бұрын
Yea it is a bit confusing! But it does all start to fit together eventually when you start looking at pseudo transient solvers
@pyreaurum676 2 жыл бұрын
These videos are invaluable
@hlanewala Жыл бұрын
I use Fluent for CFD and have almost spent a year trying to figure out why the hell Fluent manual always talks about under relaxation and pseudo-transient together. Today that confusion is gone. Aidan, your series of videos on relaxation factors and pseudo transient have had such a huge impact on my understanding of these settings in CFD.
@fluidmechanics101 Жыл бұрын
That's exactly what I was hoping for! Confusion -> gone
@sujaysawant868 2 жыл бұрын
Amazing content, yet again! lookin' forward for more.... God bless you mate!
@doo_jkk8901 2 жыл бұрын
Best explanations! Thank you!
@MrLote83 2 жыл бұрын
Its amazing video once again, you are doing really a great job.
@user-ic7gn3lz4k 2 жыл бұрын
Dear Aidan, could you please create a video about types of meshes in CFD - structured, unstructured etc. And thank you for your amazing work!
@fluidmechanics101 2 жыл бұрын
Yea I am thinking of making a series on meshing later this year. I want it to be thorough and cover everything, so it could take some time but I am sure it will be worth it
@bahadord3199 2 жыл бұрын
Awesome man with another amazing lecture. Thanks for sharing your knowledge. I am wondering if you have a video about " how to calculate errors and numerical simulation uncertainty". I am looking forward to watch another new series of your lecture.
@fluidmechanics101 2 жыл бұрын
You could check out Hrvoje Jasak's thesis? (You should be able to find it with a Google search)
@bahadord3199 2 жыл бұрын
@@fluidmechanics101 Thanks. i will check it.
@jankracik6915 Жыл бұрын
Hello Aidan, thank you for the amazing videos you have created. However, I have a little remark on this one. I am inclined to believe that on the slide with "Transient Term - Gases" the text after the first bullet point should be corrected to: "For a gaseous system, heat input will also increase the gas VOLUME." as you are using cp (not cv) and pressure should be constant. It is also implied from the figure (not a fixed boundary of the cube). Am I right or not? Thank you for your response!
@fluidmechanics101 Жыл бұрын
Ah yes, well noticed. I brushed over this point, as I really just wanted to focus on solids in this talk
@rahatullah4232 3 ай бұрын
WO......W !!
@tomukurian2649 2 жыл бұрын
Thank you for the nice explanation! Btw, which tool do you use to make those consistent schematic diagrams in your slides ?
@fluidmechanics101 2 жыл бұрын
I made them all in inkscape. I have also made a course showing all the details of how I make figures in inkscape if you are interested: www.udemy.com/course/inkscape-for-scientists-and-engineers/
@lawyinghui7899 3 ай бұрын
Dear Sir, according to my understanding, the case shown in your video is actually transient flow. But since we are using pseudo steady state to simulate the transient flow, we will still get the flow changes when iterations are going more (as shown in 02:17). I tried my case (pseudo-steady state for transient flow) in the same way, but from my observations, I realized that when I set the time scale factor (pseudo-time step factor increment) to a very small value (large value will diverge), the flow pattern will not change. However, if I adjust to larger time scale factor (larger pseudo-time step), the flow pattern will start to change again. Therefore, in this situation, the reason for observing constant flow seems to be due to extreme small time step, the changes is not obvious, rather than a "true" steady flow for the case of transient flow. I want to know, in this case, how can we make sure that the flow pattern has not changing at all, so as to show that the current flow is stable and no longer changes with time? or in simple word, what to dermine when we can stop iterate the transient follow using pseudo steady state (pseudo transient)?
@sukranochani5764 2 жыл бұрын
👍👍👍👍😘😘
@ahmadhijazi8297 2 жыл бұрын
Hello sir, I have a question about the flow courant number that is found in solution controls in Ansys fluent. The default value is 200 for pressure based solver. Is it wrong to specify a value of 1 for the flow courant number since I know that cfl should be below 1 or should I keep it as the default value. Thank you in advance
@fluidmechanics101 2 жыл бұрын
This depends on the type of simulation. Are you running a steady-state? If so, you can probably have a CFL number greater than 1 as you only care about the final solution and not how you got there. Are you running multiphase transient or something complex? Then you will need to keep CFL less than 1
@ahmadhijazi8297 2 жыл бұрын
@@fluidmechanics101 thank you for your reply; really appreciate your work. I am running a 2D transient simulation for a vertical axis wind turbine. I am using the coupled solver and in the solution controls panel, there is a flow courant number which is 200 by default. That is why I got confused, since I am using a small time step in the calculation panel, so I thought if I have to keep it as 200 or I should set it as 1 here as well. Thank you again sir.
@fluidmechanics101 2 жыл бұрын
Ahh ok, turbomachinery! You might need to try some different time steps and see if your result is sensitive to the time step. You can try a large time step to get over the initial transient and then once you reach periodic equilibrium, try switching to a lower time step, run a few revolutions and repeat with different time steps. Normally the result will be stable with CFL> 1 (I went up to 80 with a horizontal axis wind turbine) but the accuracy might be affected, so worth a bit of trial and error to see what happens
@ahmadhijazi8297 2 жыл бұрын
Yes I will try that, but my question is about the flow courant number that is found the relaxation factors in the same panel in Ansys fluent (solution controls panel). Changing it is affecting the solution, I tried so far two simulations, one with the default value(200) and the other one i set a value of 1. I got different results which confused me since both solutions are converging to different results. Thank you for your help.
@fluidmechanics101 2 жыл бұрын
Different results :) then you probably want to go with the lower value. You should also probably try lower values (say 0.5 and 0.2) to see if the results are still sensitive to time step
@anandpatel1143 2 жыл бұрын
What happens if I run a physically transient model using a steady-state solver ?
@fluidmechanics101 2 жыл бұрын
Your steady state solver will just predict a series of 'snapshots' of the solution, which aren't necessarily linked together with an accurate time history. You can think of this as randomly jumping through time to different points. The accuracy of these 'snapshots' depends on how unsteady the real solution is
@anandpatel1143 2 жыл бұрын
@@fluidmechanics101 I am running a Steady State simulation in ANSYS Fluent. When I solved it using COUPLED solver and than also with SIMPLE solver I am getting different results. What could be the possible reasons ?
@fluidmechanics101 2 жыл бұрын
Hmmm there are many different reasons. If your simulation does have some transient flow features, then the mass error associated with these features will result in a different 'looking' flow field. How are you judging convergence? If you look at global mass and energy balances you will probably find that the coupled solution does a better job than SIMPLE. There may also be a region of the flow field that is causing the solver difficulties. Have you tried switching to node based gradients and improving your mesh (i. Lower skewness and better resolution in areas of high gradients?) Lots of things to think about 😊
@hungnguyenthanh8833 6 ай бұрын
Thank God and you for uploading this lecture series! You made me a REAL CFD ENGINEER from button clicking technician.
Fluid Mechanics 101
Рет қаралды 24 М.
Fluid Mechanics 101
Рет қаралды 19 М.
QAZSPORT TV / ҚАЗСПОРТ TV
Рет қаралды 661 М.
Fluid Mechanics 101
Рет қаралды 5 М.
Fluid Mechanics 101
Рет қаралды 5 М.