The word "Accurate" needs to be explained. Explicit integration schemes are accurate in time. Implicit schemes converge to the correct steady state solution but their time evolution is not accurate. Forward Euler is an example of an explicit scheme. Backwards Euler is an example of an implicit scheme The video also needs to explain when Algebraic multigrid methods can be used in Openfoam
I completely agree - CFD is a very deep and intricate subject. My goal with this video was simply to give some high-level "rules of thumb" for one particular aspect of OpenFOAM - I didn't want to get too bogged down in the details e.g. Explicit schemes have the same order of accuracy as implicit schemes, the only difference is that implicit schemes don't have a CFL limit but require a matrix solve. But what does this inaccuracy manifest as? Numerical diffusion - negative diffusion in the case of high CFL explicit schemes. And if the implicit scheme doesn't have a CFL limit by itself, but the solver still crashes for large CFL, where does the limit come from? From the CFL limit introduced by pressure-velocity coupling scheme etc. etc. The rabbit hole goes very deep - all very interesting and useful stuff, but can be confusing / offputting for someone new to OpenFOAM. Besides, OpenFOAM doesn't have any explicit schemes, so I wouldn't have an example to show. I had planned on making an fvSolution video which would have certainly included a discussion of Algebraic Multigrid, but I've now written my own CFD code so haven't had the time (or motivation) to do this. The target audience is those people who have the time / motivation for a short 5 minute video, but not much more. I highly recommend reading a CFD textbook (e.g. Versteeg) to anyone who has more time, even if you're proficient with the software.
@@cfdforeveryone-nikl9742 I read and coded CFD solvers a long time ago from Hirsch (Contributed a little bit to rhoCentralFoam) and i would like to use CFD again. So i found your video useful and thought provoking as many old terms got refreshed in my memory
This may have been the best video on OpenFOAM in regards to quality of presentation and it's content. Thanks for the amazing work!
I’m out of words to describe how precisely accurate and concise you pass on information. Thanks a lot mate
Indeed!
such a legendary video. it's a pity this channel isn't active anymore.
video on fvSolutions would have been epic
Great video! Would love to see similar ones for other concepts in OpenFOAM.
I suggest you to include this video and the github link in your fvSchemes tutorial file
Thanks for making this! I'm going to go check your other open foam videos now
Great video! Hope you are still planning to make more videos like this!
Congratulations! Keep on going. fvSolutions next?! Thank you for such a clear contribution.
cheers!!! could you make one on fvsolutions? it would be dope!
Concise and to-the-point, thanks!
Very good explanation, thank you 🙂
unbelievably good video..
Great info. Thank you!
Amazing！！Your video is very useful for me！！！
Very helpful !! Thanks Bro
thank you !!!
Hi just wanted to ask if you have had any success using LocalEuler ddt scheme to achieve local time stepping? I'd appreciate the help thanks
This goat just dropped four videos putting in evidence the rest of shitty openfoam tutorials and then disappeared
nice work I appreciate.
For gradschemes, could you tell me what does div(rhoPhi, U), div(phi, Alpha) and div (phi, (k | omega)) mean?
phi is the flux which is basically the mass flux for compressible solvers: rho*u (dot) face_normal_vector
thank you sir
The word "Accurate" needs to be explained. Explicit integration schemes are accurate in time. Implicit schemes converge to the correct steady state solution but their time evolution is not accurate.
Forward Euler is an example of an explicit scheme.
Backwards Euler is an example of an implicit scheme
The video also needs to explain when Algebraic multigrid methods can be used in Openfoam
I completely agree - CFD is a very deep and intricate subject. My goal with this video was simply to give some high-level "rules of thumb" for one particular aspect of OpenFOAM - I didn't want to get too bogged down in the details e.g. Explicit schemes have the same order of accuracy as implicit schemes, the only difference is that implicit schemes don't have a CFL limit but require a matrix solve. But what does this inaccuracy manifest as? Numerical diffusion - negative diffusion in the case of high CFL explicit schemes. And if the implicit scheme doesn't have a CFL limit by itself, but the solver still crashes for large CFL, where does the limit come from? From the CFL limit introduced by pressure-velocity coupling scheme etc. etc. The rabbit hole goes very deep - all very interesting and useful stuff, but can be confusing / offputting for someone new to OpenFOAM. Besides, OpenFOAM doesn't have any explicit schemes, so I wouldn't have an example to show.
I had planned on making an fvSolution video which would have certainly included a discussion of Algebraic Multigrid, but I've now written my own CFD code so haven't had the time (or motivation) to do this.
The target audience is those people who have the time / motivation for a short 5 minute video, but not much more. I highly recommend reading a CFD textbook (e.g. Versteeg) to anyone who has more time, even if you're proficient with the software.
@@cfdforeveryone-nikl9742 I read and coded CFD solvers a long time ago from Hirsch (Contributed a little bit to rhoCentralFoam) and i would like to use CFD again. So i found your video useful and thought provoking as many old terms got refreshed in my memory