Тёмный
No video :(

[CFD] Conjugate Gradient for CFD (Part 1): Background and Steepest Descent 

Fluid Mechanics 101
Подписаться 54 тыс.
Просмотров 9 тыс.
50% 1

An introduction to the conjugate gradient method and other gradient descent based methods (steepest descent method) for CFD. This is Part 1 of a 2 Part lecture series.
Time stamps
0:00 Introduction
2:53 Gauss-Seidel Recap
7:20 Quadratic Form
14:48 Gradient of f
19:37 Finding the minimum
24:00 General algorithm
25:53 Steepest descent method
29:50 Search direction (pk)
32:25 Search distance (alphak)
39:55 Steepest descent summary
41:33 Conjugate gradient motivation
43:02 Summary
44:33 Outro
References
J. Shewchuk 'An Introduction to the Conjugate Gradient Method Without the Agonizing Pain', Edition 1 1/4, 1994.
www.cs.cmu.edu/~quake-papers/...
============================================
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.

Опубликовано:

 

12 авг 2024

Поделиться:

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист
Посмотреть позже
Комментарии : 33   
@bartosh3959
@bartosh3959 8 месяцев назад
Could not be more clear, thanks!
@matthias1991
@matthias1991 4 месяца назад
I was deterred from watching the video for the longest time because of the name of the channel, because I'm not studying Fluid Mechanics but honestly this is the best explanation of CG I found on youtube! You're really breaking it down into bits that are easy to understand and explain the relevant concepts from basics and built up on that. It only required attention but I finally understand it now. Thank you so much! :-)
@sammartens1090
@sammartens1090 3 месяца назад
I really, really like your approach of explaining things. Like when you go through concrete examples with numbers instead of variables or explain physical meaning also circling back to the basics when needed, it makes it possible to understand this even for people like myself, who has very little experience if CFD solvers. Thank you very much.
@k0185123
@k0185123 11 месяцев назад
what a great introduction to steepest descent method! Looking forward to your next lecture on CG and perhaps GMRES :D
@bsreekanthreddy2622
@bsreekanthreddy2622 4 месяца назад
The best explanation, anyone could ever find
@aminfathi9116
@aminfathi9116 8 месяцев назад
Best lecture I found on the Internet. Thank you so much. Now, I'm forwarding to the next lecture.
@felipelopez5656
@felipelopez5656 8 месяцев назад
Great explanation to begin with gradient methods! Thank you for this video 😀
@jessbuildstech
@jessbuildstech 11 месяцев назад
A fantastic video mini series, I'm excited for the Part 2!
@giahuypham1857
@giahuypham1857 5 месяцев назад
thank you for the clear instructions. You are more than a god to me
@isaacpaten1495
@isaacpaten1495 6 месяцев назад
Great lecture, thanks so much. You took time to explain lots of things which other videos just stated.
@user-rw1py2wq8w
@user-rw1py2wq8w 11 месяцев назад
Great effort, thanks
@heloisacristina8688
@heloisacristina8688 11 месяцев назад
Woooow! Very nice class! Please send the part 2!
@raghavamorusupalli7557
@raghavamorusupalli7557 9 дней назад
Wonderful demonstration! Thanks for sharing. The initial discussion did not interpret what f is
@AC-tn4it
@AC-tn4it 5 месяцев назад
Wow, this was such an amazing explanation
@bushraw66
@bushraw66 4 месяца назад
I wish you had solved some examples, your explanation is great but hands on practice would really drill the method into my brain, specially that there is very few examples on youtube about those methods
@weigleec
@weigleec 11 месяцев назад
Very nice class! Steepest gradient is a lot like gradient descent (used in neural networks).
@opencast1819
@opencast1819 11 месяцев назад
Thank you Aidan, a great lecture! Currently I have encountered in my multi-phase magneto-hydrodynamics simulation using OpenFOAM convergence problems for the Poisson equation for the electric potential due to the presence of a low conductive phase. Algebraic multi grid (GAMG in OpenFOAM) was fast, but it stuck at low convergence level. Conjugated gradient (PCG in OpenFOAM) managed to converge (I need 1e-18 absolute tolerance at least), but it takes ~800 iterations. The great solution appeared to be a combo of PCG solver with GAMG preconditioner with Fast-DIC smoother. It allows to converge in 16 (only!) iterations. Would be nice to have some lecture on such hybrid approaches. Thank you in advance!
@falanger6659
@falanger6659 6 месяцев назад
Great video, helped me a lot. Big thanks to you!
@fzigunov
@fzigunov 11 месяцев назад
Really nice. I just implemented this algorithm (CG) for a problem I was working on, so it's really cool to see your take on it! Very detailed and beginner-friendly! Edit: why does it sometimes diverge? Maybe you can discuss that in the next lecture?
@wile9763
@wile9763 11 месяцев назад
CG methods are a member of a class of linear solvers called Krylov subspace methods. The iteration error of these methods can be shown to depend on the condition number of the matrix which is in turn determined by the eigenspectrum of the matrix. So, in short, if a linear system is sufficiently poorly conditioned, Krylov methods will diverge. Preconditioners are used to remedy this problem and represent a large area of interest in the literature.
@fluidmechanics101
@fluidmechanics101 11 месяцев назад
Exactly! Great response. I didn't want to get into condition number and pre-conditioners here, as I wanted to focus on actually understanding the method first. Hopefully I have succeeded!
@fzigunov
@fzigunov 11 месяцев назад
@@wile9763 Thanks for the easy-to-understand response! This is also nice to know so I can improve my solver 🙂
@opencast1819
@opencast1819 11 месяцев назад
@@wile9763 great answer! You can find some related experience in my comment above. Cheers!
@aderstaarey3438
@aderstaarey3438 9 месяцев назад
Great work
@naimaabdallah3490
@naimaabdallah3490 4 месяца назад
So nice 🎉
@freevoice6611
@freevoice6611 8 месяцев назад
These methods also solve the storage issue of sparse Matrix.. but how? Can someone please explain?
@mohmmedabdghany5696
@mohmmedabdghany5696 11 месяцев назад
Nice lecture, can you took about multiphase, vaporization and condensation ?
@user-yr8ze2zb7j
@user-yr8ze2zb7j 8 месяцев назад
Hello, thank you very much for your video. However, in your video 34:20, you mentioned that all residuals point towards the global minimum. I would like to know, is this a special case or an inevitable situation? Because the residual equals the gradient,but the gradient does not necessarily point towards the global minimum.
@murat9268
@murat9268 11 месяцев назад
There is better alternative to gauss seidel method. LUSGS method, gauss seidel method with approximate LU factorization.
@brandon7939
@brandon7939 8 месяцев назад
I find that u look like Sheldon
@fluidmechanics101
@fluidmechanics101 8 месяцев назад
😂😂
@caleb4485
@caleb4485 10 месяцев назад
Promo_SM
Далее
[CFD] Gauss-Seidel Method in CFD
28:31
Просмотров 10 тыс.
[CFD] H/A (HbyA) in OpenFOAM - Part 1
38:34
Просмотров 2,8 тыс.
Ranking EVERY Weapon in Minecraft
21:13
Просмотров 142 тыс.
Ray Tracing Harmonic Functions
9:19
Просмотров 1,4 тыс.
BRILLIANT WIN!!!!!!!
25:27
Просмотров 251 тыс.