CFD - Computational Fluid Dynamics, Motorsport, Formula 1

Here are our CFD links and discussions about aerodynamics, suspension, driver safety and tyres. Please stick to F1 on this forum.
CFDruss
CFDruss
0
Joined: 08 Sep 2003, 18:47
Location: Tamworth (nr Birmingham) UK

Post

Monstrobolaxa

Your Processor speed is fine, and graphics card should render at a fairly decent speed, unless the geometry is highly complex, even my 128mb card (Asus beast!!! mmmm) strugles to render quickly in fluent, and thats just simple grey scale rendering :(
The memory, well it will do, for basic models, but the better you get with catia the more you computer will lag, and the memory is definatly pretty useless with Fluent, unless a course mesh is used. The memory is bottlenecking you CPU. Not sure how many spare slots you have on your laptop, but if you have one more consider popping in a nother 512mb, will help, or if you can afford it, pop in a 1Mb.

Russ

CFDruss
CFDruss
0
Joined: 08 Sep 2003, 18:47
Location: Tamworth (nr Birmingham) UK

Post

By the way guys, if any of you have problems with running solution due to memory or any thing, then feel free to send me the files and I will be happy to run for you. Or if you are struggling to mesh then I can send you a Journal file and you can go through it and try and see what is the best route, and even edit that journal file when you get confident.

Russ

By the way, hey paul, good to hear from you again

F1 CFD
F1 CFD
0

Post

Dear all ,
I feel so exciting to see so many pictures about F1 CFD technicals,
and I am a CFD modelor too in the field of building Service and Fire,
Could someone provide me some information about the F1 modelor's
certification, my email is
timhe2004@hotmail.com
some other issues about CFD is also welcome.

Monstrobolaxa
Monstrobolaxa
1
Joined: 28 Dec 2002, 23:36
Location: Covilhã, Portugal (and sometimes in Évora)

Post

Only now I saw your post Russ...well I'll be learning/working with CATIA in August...I'm going to take my laptop and the tutorials on holidays with me....so first comes CATIA....and when I know a little about it and 3d modelling and putting parts together....I'll start with CFD...about the memory I can expand it! Actually after Christmas I'm planning to expand the memory, hard drive and the processor! (when I bought the computer I made sure that I could upgrade almost everything on it.....

crystalclear
crystalclear
0

Getting started in CFD

Post

okay, there is mention on this site that CFD is hard and you've got to be prepared to handle the mathematics and computing, and just being an F1 petrolhead is not enough. Then, the site goes on to imply that there is a shortage of CFD engineers and that F1 is almost crying out for people that are naturally gifted in CFD for F1. And that all makes sense too: whether you believe Descates (that running into a crowd twice as fast causes you to hit twice as many people twice as hard), or believe conventional physics (that you hit the same number of people with four times the energy - lots magically getting out of your way), the result is the same. Aerodynamics rules in F1.

So would one of these F1 CFD companies provide software free for me to learn, in the knowledge that they'd make money if I put it to serious professional use later?

I'm just wondering how one could get started at this. That's the crux of my question.

Basically, somebody should be able to post a bog standard 'out of date' formula one part on the web somewhere. Then anybody that thinks they are any good at this CFD stuff could post their modified part and anybody could CFD-it to check the results. Then, if Joe Bloggs the F1Technical Amateur comes up with a part that gives numbers comparable to the 'state of the art' F1 parts, Joe Bloggs has proved his worth, and an F1 company should at least consider him as a potential asset to them.

So who are the major companies in the CFD stuff? I get the feeling that this site is sponsored or run by a CFD company, as the site has a CFD bias relative to other F1 technology. Any chance of a play-copy of the software to see if I can hack it, or whether the computing and mathematics is beyond me?

I have a decent understanding of Raytracing and gearboxes. I made a video explaining how modern electronically controlled 6-speed automatics work by animating the ray traced pictures, with splines on the clutches, helical cuts on the teeth of planetary gears, semi-transparent hydraulic oil, cutaways to allow viewing inside, moving camera and zooming effects, etc.

I've got a basic idea about aerodynamics too, as you would if you'd had some flying lessons and bought some cheap kites that initially wouldn't fly too well: designed for different windstrengths, less blustery conditions, etc. I was particularly proud when I read about a NASA scientist's theory 'the wing should adapt to the wind and not the wind adapt to the wing' and then tried to reproduce a kite with no solid parts. Friends and work colleegs said it would never fly because it had no solid parts: just string and ripstop nylon, and it frustrated me as I experimented with different adjustments, but when it finally went up, it was glued in the sky for half an hour.

Anyway, back to my question.

Is it possible to get started with a VERY low budget?

CFDruss
CFDruss
0
Joined: 08 Sep 2003, 18:47
Location: Tamworth (nr Birmingham) UK

Post

Hello crystalclear, welcome to the site.

CFD is not a hard tool to use, infact, you can pick programs like Fluent up quite easily for analysis of simple models, and then compare your results to known data. However, if you want a highly accurate solution, then a in-depth knowledge of Fluid Mechanics and the Code is essential. You need to know what’s happening behind the screen to interpret the results correctly, you need to understand the solution may be failing if it is not converging. I quote by John D Anderson Jnr, "If you do not physically understand the meaning and significance of each of the equations associated with CFD - Indeed, of each term in these equations - then how can you even hope to properly interpret the CFD results obtained by numerically solving these equations??"

I am not aware the F1 in crying out for aerodynamicists/CFD users, if that’s the case I wish they would respond to my applications, but as always, unless you have stacks of experience behind you then companies wont even sniff at you. That’s why im having such a tough time getting in to aerodynamics/CFD, if one chance popped up I would prove my worth and work damn hard for that company, but they just don’t give you that chance.......catch 22: you need experience to get a job, but you need a job to get experience!!!!

The site is not sponsored by a CFD site, If only it/I was, again, that might given many budding engineering a chance.

The main contenders in CFD codes are Star-CD, Ansys CFX, and Fluent, however, as im aware, it is difficult to obtain any limited lisence software from them, and if they do they tend to only like the software to be installed in a industrial surrounding, such as on a computer at a university you may be attending.
However, there are a number of low level Codes available for download free of charge, but these a limited to basic use, just as wing analysis.

As for the old models, people wishing to carry out CFD on a old F1 part have to rely on the hope that a CAD engineer out they will offer to model a part as close to scale and detail as possible. As original models from the F1 companies are like gold dust.


Hope this help.
Would be nice if some company owner out there read this and gave a few of use budding engineers a chance, after all, its the ones that never give up, even when they cant get a job in the field they wish, yet carry on self learning for the shear interest in the subject and not just as a job that pays the bills, are the people that I feel will contribute to a company far more.

Russell Harrison
CFDRuss
Russell Harrison
Forced Convection Design Engineer, Comair Rotron Europe Ltd
CFD is based around assumptions; the accuracy of the solution depends not only on the knowledge of the mathematics behind the software but the assumptions the user makes!!!

SKRAT
SKRAT
0
Joined: 30 Aug 2003, 01:34

Post

Best/Cheapest way I can think to start (and stay legit) is to be a student at a university. Most engineering (mech/aero) depts should have at least a few licenses of one or more of the top CFD codes for their use. If there's an intro class, take it, or just go through the progression of tutorials to get the gist of what's going on. After that, it's just hammering away at tougher and tougher problems (and validating the results of course) and then picking up the infinite number of subtlies and tricks.

CFD shops, like Advantage CFD, seem to use separate (and specialized)packages for grid, solve, and plot/analysis. If anyone knows what their guys/gals went through to get their current jobs, it would be interesting to learn.

BTW Russ - nice reference to JD Anderson (Anderson's Fundamentals of Aerodynamics text seems to be the standard aero-intro book, at least here in the States). I applaud your tenacity in getting a CFD-motorsport job. Alas, I am now at the transition in my professional career to taking on managerial/project lead tasks (it's where the bucks are) and doing less and less CFD. Good luck.

Paul C

Aerodramatics(UK)
Aerodramatics(UK)
0
Joined: 10 Oct 2004, 16:39

Post

CFDruss wrote:
I am not aware the F1 in crying out for aerodynamicists/CFD users,
.... unless you have stacks of experience behind you then companies wont even sniff at you. That’s why im having such a tough time getting in to aerodynamics/CFD
Russ don't despair + don't give up.... from what I can see both Keith & you have performed some credit-worthy "generic" CAD/CFD modelling, with a sensible design flow condition.

That's a great start... I reckon, one way into F1 teams is to become a competent authority in the efficient application of particular commercial CFD code, as you say, like FLUENT and STAR-CD (and gaining fans - CFX). So get this experience by taking a less "prestigous" job in the Aerospace/Automotive industry for 2-5 years.

By efficient, I mean a demonstrated ability to grid complex geometries (e.g. front, rear assemblies and full/half car) and to perform the pre-/post-processing in a highly automated fashion.

Iteration is king....(it allows some degree of practical optimisation in the design process - since it's tough to take out the creative human element in innovative design) - teams are v. competitive and know this very well!Being able to slot into and perform one's job in fast design cycles is a key/prerequisite employee attribute I'd venture.

Also a decent understanding of configuration integration aerodynamics would go a long way into making you a race team's kinda person.... those entry level junior scarydynamicist positions as advertised in Autosport are well within your grasp...

... you've already taken the first steps.... pursue your ambition & you will suceed!

...but don't burn out in your first year in the job...! Save something for the long haul... ! :lol:

P.S.:

Btw, gaining demonstrable competence in some third-party pre-pro s/w like ICEM-CFD/GRIDGEN in addition to GAMBIT/TGRID and post-pro s/w like Fieldview/Ensight could add a string to your bow - though if you can use GAMBIT/FLUENT these'll be a doodle to pick up...!

Also consider getting some basic surfacing CAD skills and maybe familiarity with optimisation s/w would score points....

...old world panel methods/coupled b.l.s have a role to play here - though
have all the usual flow physics limitations that drives one to RANS and detached eddy sim. :wink:

However, be warned track testing, tunnel & race engineering are still king, with CFD a junior partner...

... like I'm telling you nothing you don't know already... doh! :roll:


PPS:

Expect your girlfriend/wife to ditch you (or play away) in the first 6 months
when you get sucked into the F1 work ethic!!! :shock:

PPPS: If they offer to renew the initial basic, low pay (you're out the door
when we say and banned from working again with no pay) contract make sure you work (insist) on getting some better(cut-down Tech. Dir. gardening leave style) terms, pay-rises, points bonuses and promos...) -> then again you might just have to take what you're given :) -> also stay friends with colleagues and the boss (or at least his daughter...! :wink:)

PPPPS: An aside:- I saw your comments on Hypersonics & Anderson in and about this topic.... ironic as Anderson seems to have made his name in Hypersonics (especially w/t testing heat transfer measurement techniques) + might well not object to being labelled a dyed-in-the-wool experimentalist in his time (ok cfd was in its infancy)... so in this respect I found his CFD book an ironic chuckle, but he's got a great teaching style and is open minded about CFD, wouldn't you agree?

User avatar
KeithYoung
24
Joined: 02 Jul 2003, 20:21
Location: USA

Post

Unfortunately due to time zone differences and changes in work shifts Russ and I havent been able to work together very much recently. I am sure he is looking for a designer who has a compatable schedule. I still have time during lunch to model things up, and I would like to find someone who is very capable in CFD and the theories behind it, someone with the same schedule. Email me at keith.a.young@comcast.net if you are interested in working with me. I dont want to send my work out to everyone, and I can only work with someone I can trust and who I think is as good at CFD as I am in CAD. These projects benefit both sides of the coin, I cant do CFD as well as I need done, and chances are if your good enough at CFD, you dont have time to be as good as me in CAD, its exactly why Russ and my teamwork worked so well. If you are interested just email me for further discussion, and I ask you please have MSN instant messenger working for further communications. Not only is this great for the two partners involved, but from what I gather everyone else likes reading about the findings :wink: .

Cheers

nazo
nazo
0

Post

this site is amazing....
i have a problem, i am designing a rear wing for a formula student car and I will go in the wind tunnel in 2 weeks time. I have fluent but i dont know how to use it. can u import different file typesmaybe cad, or catia? i need to put endplates but i think i cannot do that with fluent!

hope somebody will reply... thanks.

naz.

User avatar
KeithYoung
24
Joined: 02 Jul 2003, 20:21
Location: USA

Post

I usually use step files. There is a list in there which says what can be imported, parasolid is one, but I havent ever gotten it to work.

Yaquiya
Yaquiya
0

Post

Hi everyone;) I'm a spanish girl and i'm interested in doing a research project about F1 aerodynamics. I've seen your topic and it seems very very interesting 8) I wonder if there is any Pc software which allows me to design a F1 car and study the air's influence in each design :oops: If there is one, please tell me where can i find and (if it is not too much :oops: ), how to use it (because I supose i wouldn't be easy to use :cry: )
I've seen CFDruss' articles and i think they are quite interesting, so if you could explain me more about them, i'll be very pleased :lol:
Thank you a lot

User avatar
Yaquiya
0
Joined: 27 Feb 2005, 15:25
Location: Barcelona, Spain

Post

I'm here again, now to tell i've registred myself here :lol: , so if you want to send me any PM...
I hope you help me with my RP 8) :lol:
Image

Guest
Guest
0

Interesting Thesis

Post


Guest
Guest
0

Post

Fluent is supposedly the leader in F1 CFD. You'll need some modeler tho to build the 3D models, like Solidworks (my friends say it's really easy to use), Unigraphix (according to an old professor, who takes classes by Oakley sunglasses, is the best one for assembly), or Pro/E (Ferrari). Of course, you need a license to operate them, and if u do get one, plan to spend thousands on software and some more thousands on a computer to build/analyze models. Basically you import the CAD drawing into Fluent, make some meshes based on equations and theory, and you can see how fluid works on ur shapes

For FEA, you'll want something like CATIA

West

BTW this is just a very low-level understanding of these things; don't bother nagging me on it.