How do I into CFD, /sci/? I've majored in math and computer science and I'm starting a Ph.D. In the fall. I may very well be working with some CFD apps so I want to get a head start.
Also I have access to some testbed supercomputers I can play with, so I thought it'd be interesting to try implementing some stuff.
Where do I start?
Learn fluid mechanics, play with the free baby version of adina until you understand how it works.
>>8085247
Lmfao is that cow aerodynamics or something? Hahahaha I can't stop laughing
>born too late to explore the earth
>born too early to explore the universe
>born just in time to witness the research of cow aerodynamics
![elephant[1].png](https://i.imgur.com/EAEHm5f.png "elephant[1].png")
>>8085247
>>8085282
>>8085273
http://escholarship.org/uc/item/03k0g43p
>bovine aerodynamics in terms of fuzzy lagrangian dynamics
I couldn't make this shit up if I wanted to.
pic semi related
>>8085247
These are all PDEs right?
>>8085267
Is there a good textbook for fluid mechanics you would recommend?
>>8085316
Physics is one big PDE.
>>8085304
I...why?
Why the fuck is anyone researching this? Is it to help coastal farmers breed more aerodynamic cows to prevent them from falling of cliffs?
How the fuck did this funding get approved when that computing time could be used to research something actually usefull.
>>8085510
Computing power is cheap. Actually it is to breed more tornado resistant cattle for the midwest. Cattle losses due to tornados can be rather large.
>>8085526
Can't stop laughing.
>>8085247
Solarwind CMHDfag here
Assuming you know PDEs. Familiarize yourself with differential/integral forms of conservation equations, i.e. mass, momentum, and energy. Understand their physical meanings. Pending what you are simulating, you may also want some understanding of the thermodynamics involved, particularly in the case of high speed flows.
If you are writing your own code, then books wise, I recommend taking a look at Computational Fluid Dynamics by Hoffman and Chiang. Fairly standard introductory grad text utilizing finite difference method. FD is probably the most basic way to simulate PDEs and this text does a good job explaining. John Anderson (basically the Led Zeppelin of Aerospace Engineering) also has an excellent book from what I've read.
>>8085526
I actually laughed out loud
>>8085526
>it is to breed more tornado resistant cattle
lmao wtf
>>8085661
Not him, but you don't really need PDEs in the way mathematicians/undergrads study them, the books the other Anon posted are fine, you can also study some numerical analysis and build on more advanced (applied) math being researched from that foundation.
>>8085652
We don't know what hoverbike you are talking about and you should read the article because it is fucking great. It appears a physicist got forced into writing for a student newspaper as part of the class, giving a succinct and pithy explanation of lagrangian dynamics.
There is no reason to fund research on hoverbikes.
>>8085684
Someone made a thread about hoverbikes, loads of people said it was a stupid concept, I supported OP that it's a good idea and I'm working on one myself. There was a 200 reply shitstorm. I'm not going to get into it again but compared to the bullshit I've seen scientists get funding for (nanoputians http://pubs.acs.org/doi/abs/10.1021/jo0349227 ) a hoverbike is a good use of funds in comparison
>>8085684
>read the article because it is fucking great.
He's funny, but that author doesn't know jack shit about modelling.
>>8085705
>> making fun of basic research
Get out
>>8085675
Something like this then? Seems pretty widely used
>>8085397
I just use the one that was recommended back in undergrad, which is particularly terrible. Fluid mechanics can be sort of odd, so you might wanna get several textbooks.
Also, fair warning, adina is a terminally german cfd package. It will work and work very well, but putting in your models is a messy process.
>>8085316
Yeah. Typically they use the kappa-epsilon model as an approximation.
>>8085397
fluid dynamic drag
hoerner
>>8085510
>Why the fuck is anyone researching this?
Real answer: Because you need to research something (anything) to get a MSc
>>8086282
I have a copy of that, the numerical are too sparse it's too old for what you are looking for.
>>8085652
>literal bull
>>8085247
>when the grass hit
learn how to use Ansys CFX or Fluent and Solidworks Flow Simulation
![heatminer-vector-field-heatmap[1].png](https://i.imgur.com/1A7kiTV.png "heatminer-vector-field-heatmap[1].png")
>>8085247
what is the technical name for these kind of graphics that tell the direction and intensity of fluid simulations via vectorel arrows ?
>>8086886
arrow plot.
matlab function is quiver
>>8086282
If anyone's interested in cheap books, I believe dover publishes a version of this now.
>>8086845
He can learn Adina for free, though, without the need to pirate anything (wink, wink). Solidworks flow sim is annoying anyway.
>>8085705
>a 200 reply shitstorm
lumping together hundreds of posts shitting on your idea for a large variety of reasons into a single, retarded buzzword does nothing but make you look like even more of a fucking tool than you already are
>a hoverbike is a good use of funds in comparison
we can all be glad that you're in no position to dictate the allocation of research funding ressources
if you're so starved for attention just go ahead and start yet another hoverbike thread, which we all know will generate the exact same response pattern as the previous dozen
but stop shitting up this thread with your simple-minded rant about who gets his research funded and who doesn't
it's annoying and deters from the actual discussion
