How accurate is our "cheap" CFD ?

Here are our CFD links and discussions about aerodynamics, suspension, driver safety and tyres. Please stick to F1 on this forum.
Uwe
Uwe
1
Joined: 18 Dec 2020, 13:40

Re: How accurate is our "cheap" CFD ?

Post

Hoffman900 wrote:
19 Dec 2020, 22:19


From a Boeing friend, I don't think they would put much stock into low end CFDs on desktop computers, except for some very course measurements (maybe). That said, airplane aerodynamicists have to work with a greater factor of safety and confidence than even F1 designers.
I think F1 aerodynamics is far more complex than in aviation.
F1 has lots of aerodynamics sick complex parts that aircraft not need..

Jolle
132
Joined: 29 Jan 2014, 22:58
Location: Dordrecht

Re: How accurate is our "cheap" CFD ?

Post

Zynerji wrote:
19 Dec 2020, 21:17
LM10 wrote:
19 Dec 2020, 21:10
Talking about CFD at home, a bit offtopic: How is the FIA managing the ban on CFD work for the 2022 car for instance? What if engineers simply start their home computer and do some CFD?
I'm absolutely sure there is some of this going on, and if F1/FIA doesn't know, it is ridiculous. Especially with blockchain tech. A team could have given each engineer a 64core Threadripper with 512GB RAM, multiple GPU desktop to take home, while having a distributed blockchain CFD cruncher running in the background. Get 300-400 of these rigs up and running, and you could be doing some serious crunching in a 100% invisible, un-auditable way...
A lot of these rules are under a gentleman's agreement and good sportsmanship. There are many ways around these rules if they wanted to. It's in the interest of all investors into the sport that this play field is levelled to some degree, else spending (and financial results) would go haywire again. As for CFD work, I assume everything is done by remote sessions on the network back in base. Not just to monitor the rules, but also for safety.

n_anirudh
28
Joined: 25 Jul 2008, 02:43

Re: How accurate is our "cheap" CFD ?

Post

Personally, I think accuracy is the "wrong" question to be asking as many previous posters have said. For instance, researchers may look at wakes of tractor-trailers and come up with a particular flow topology when they run LES; but may end up with a completely different flow topology whilst using a different turbulence model/numerical scheme or a method. This may happen in F1 with RANS predicting one type of flow state, while another may exist in the wind tunnel.

In my experience, researchers of automotive hatchbacks traditionally use the Ahmed body as a benchmark - but a standard body actually lies at the cusp of two different flow states (wrt to its width, back slant angle etc); and would be a bad starting point for someone to use the experimental data to tune CFD models.

These are small issues compared to what F1 teams usually look at - performance modelling and perhaps 0D or 1D models for initial assessment

Hoffman900
163
Joined: 13 Oct 2019, 03:02

Re: How accurate is our "cheap" CFD ?

Post

Uwe wrote:
20 Dec 2020, 00:26
Hoffman900 wrote:
19 Dec 2020, 22:19


From a Boeing friend, I don't think they would put much stock into low end CFDs on desktop computers, except for some very course measurements (maybe). That said, airplane aerodynamicists have to work with a greater factor of safety and confidence than even F1 designers.
I think F1 aerodynamics is far more complex than in aviation.
F1 has lots of aerodynamics sick complex parts that aircraft not need..
Not at all. Ever look at a wing and all the control elements? Also, planes have a much further range of speed (MUCH higher than Formula 1), pitch, temperature changes (air density), crosswinds / downwinds / updrafts / downdrafts, balance, etc. then add in the factor of safety aspect of it all.

The smartest aerodynamics folks are aerospace for a reason (and get paid accordingly) and why they have MUCH more computing power than any F1 teams. This isn't a slight on the intelligence in F1 either, incredibly smart people there too, but there are just less brains (it's a smaller industry). Don't confuse little add on vortex generators and the like with complexity. Sure, it's not easy, they are two different disciplines and people know that, but aircraft (to do it well and reliably) are hard. I really don't know if I would be confident enough to fly in a plane designed by Ferrari's aero department :shock:

Remember, even aircraft aerodynamicists understand and use vortex generators. They have been playing with them well before Formula 1 did. There is a famous crash of the XB-70 in 1966 where a F-104 got sucked into the wing tip wake vortex and crashed into it.

Uwe
Uwe
1
Joined: 18 Dec 2020, 13:40

Re: How accurate is our "cheap" CFD ?

Post

Hoffman900 wrote:
20 Dec 2020, 02:52
Uwe wrote:
20 Dec 2020, 00:26
Hoffman900 wrote:
19 Dec 2020, 22:19


From a Boeing friend, I don't think they would put much stock into low end CFDs on desktop computers, except for some very course measurements (maybe). That said, airplane aerodynamicists have to work with a greater factor of safety and confidence than even F1 designers.
I think F1 aerodynamics is far more complex than in aviation.
F1 has lots of aerodynamics sick complex parts that aircraft not need..
Not at all. Ever look at a wing and all the control elements? Also, planes have a much further range of speed (MUCH higher than Formula 1), pitch, temperature changes (air density), crosswinds / downwinds / updrafts / downdrafts, balance, etc. then add in the factor of safety aspect of it all.

The smartest aerodynamics folks are aerospace for a reason (and get paid accordingly) and why they have MUCH more computing power than any F1 teams. This isn't a slight on the intelligence in F1 either, incredibly smart people there too, but there are just less brains (it's a smaller industry). Don't confuse little add on vortex generators and the like with complexity. Sure, it's not easy, they are two different disciplines and people know that, but aircraft (to do it well and reliably) are hard. I really don't know if I would be confident enough to fly in a plane designed by Ferrari's aero department :shock:

Remember, even aircraft aerodynamicists understand and use vortex generators. They have been playing with them well before Formula 1 did. There is a famous crash of the XB-70 in 1966 where a F-104 got sucked into the wing tip wake vortex and crashed into it.
Greater speed dont make calculation more complex,execept we are going into supersonic speeds.
Aircraft dont have and it will never has spliters,gurney flaps,tyres,suspensions,mirrors,sidepod,shark fin,diffuser,side duct,DRS,side skirt,canards,spoilers,moving ground,etc etc etc etc etc

Puting all this thousendes thing all together make F1 extremly more complex to solve in one CFD software..
Airflow around F1 car is far more complex/turbulent then aorund any aircraft,separated flow and turbulence is almost impossible to solve with any CFD with great accuracy..

User avatar
jjn9128
769
Joined: 02 May 2017, 23:53

Re: How accurate is our "cheap" CFD ?

Post

Uwe wrote:
20 Dec 2020, 10:10
Greater speed dont make calculation more complex,execept we are going into supersonic speeds.
Aircraft dont have and it will never has spliters,gurney flaps,tyres,suspensions,mirrors,sidepod,shark fin,diffuser,side duct,DRS,side skirt,canards,spoilers,moving ground,etc etc etc etc etc

Puting all this thousendes thing all together make F1 extremly more complex to solve in one CFD software..
Airflow around F1 car is far more complex/turbulent then aorund any aircraft,separated flow and turbulence is almost impossible to solve with any CFD with great accuracy..
Errrrrr Gurney flaps were called Nolders before Dan Gurney first used one. Most F1 tech comes from aerospace, only thing I can think of going the other way is moving ground plane in a wind tunnel.

Aerospace and motorsport aerodynamicists are the same people - some go from one industry to the other.
#aerogandalf
"There is one big friend. It is downforce. And once you have this it’s a big mate and it’s helping a lot." Robert Kubica

Uwe
Uwe
1
Joined: 18 Dec 2020, 13:40

Re: How accurate is our "cheap" CFD ?

Post

jjn9128 wrote:
20 Dec 2020, 10:15
Uwe wrote:
20 Dec 2020, 10:10
Greater speed dont make calculation more complex,execept we are going into supersonic speeds.
Aircraft dont have and it will never has spliters,gurney flaps,tyres,suspensions,mirrors,sidepod,shark fin,diffuser,side duct,DRS,side skirt,canards,spoilers,moving ground,etc etc etc etc etc

Puting all this thousendes thing all together make F1 extremly more complex to solve in one CFD software..
Airflow around F1 car is far more complex/turbulent then aorund any aircraft,separated flow and turbulence is almost impossible to solve with any CFD with great accuracy..
Errrrrr Gurney flaps were called Nolders before Dan Gurney first used one. Most F1 tech comes from aerospace, only thing I can think of going the other way is moving ground plane in a wind tunnel.

Aerospace and motorsport aerodynamicists are the same people - some go from one industry to the other.
I just want to say that airflow around F1 car is far more complex turbulent then around aircraft...so it must be harder to solve for CFD,much harder..

Uwe
Uwe
1
Joined: 18 Dec 2020, 13:40

Re: How accurate is our "cheap" CFD ?

Post

jjn9128 wrote:
19 Dec 2020, 22:37
Heck anyone can buy time on an HPC relatively cheaply nowadays, don't even have to simulate it on your home PC. Really depends how accurate you want to go, the sort of mesh you can run now on a home PC in a day is within 1% of real world data. Or you can run on a Chinese HPC with 80 billion cells and get within 0.5% :lol:
.
Last edited by Uwe on 21 Dec 2020, 15:56, edited 1 time in total.

Uwe
Uwe
1
Joined: 18 Dec 2020, 13:40

Re: How accurate is our "cheap" CFD ?

Post

Hoffman900 wrote:
20 Dec 2020, 02:52
Uwe wrote:
20 Dec 2020, 00:26
Hoffman900 wrote:
19 Dec 2020, 22:19


From a Boeing friend, I don't think they would put much stock into low end CFDs on desktop computers, except for some very course measurements (maybe). That said, airplane aerodynamicists have to work with a greater factor of safety and confidence than even F1 designers.
I think F1 aerodynamics is far more complex than in aviation.
F1 has lots of aerodynamics sick complex parts that aircraft not need..
Not at all. Ever look at a wing and all the control elements? Also, planes have a much further range of speed (MUCH higher than Formula 1), pitch, temperature changes (air density), crosswinds / downwinds / updrafts / downdrafts, balance, etc. then add in the factor of safety aspect of it all.

The smartest aerodynamics folks are aerospace for a reason (and get paid accordingly) and why they have MUCH more computing power than any F1 teams. This isn't a slight on the intelligence in F1 either, incredibly smart people there too, but there are just less brains (it's a smaller industry). Don't confuse little add on vortex generators and the like with complexity. Sure, it's not easy, they are two different disciplines and people know that, but aircraft (to do it well and reliably) are hard. I really don't know if I would be confident enough to fly in a plane designed by Ferrari's aero department :shock:

Remember, even aircraft aerodynamicists understand and use vortex generators. They have been playing with them well before Formula 1 did. There is a famous crash of the XB-70 in 1966 where a F-104 got sucked into the wing tip wake vortex and crashed into it.
I think F1 rear wing has all aerodynamics in it, that airliners need to fly...
F1 much more complex to solve

Just_a_fan
591
Joined: 31 Jan 2010, 20:37

Re: How accurate is our "cheap" CFD ?

Post

Uwe wrote:
20 Dec 2020, 10:31
jjn9128 wrote:
20 Dec 2020, 10:15
Uwe wrote:
20 Dec 2020, 10:10
Greater speed dont make calculation more complex,execept we are going into supersonic speeds.
Aircraft dont have and it will never has spliters,gurney flaps,tyres,suspensions,mirrors,sidepod,shark fin,diffuser,side duct,DRS,side skirt,canards,spoilers,moving ground,etc etc etc etc etc

Puting all this thousendes thing all together make F1 extremly more complex to solve in one CFD software..
Airflow around F1 car is far more complex/turbulent then aorund any aircraft,separated flow and turbulence is almost impossible to solve with any CFD with great accuracy..
Errrrrr Gurney flaps were called Nolders before Dan Gurney first used one. Most F1 tech comes from aerospace, only thing I can think of going the other way is moving ground plane in a wind tunnel.

Aerospace and motorsport aerodynamicists are the same people - some go from one industry to the other.
I just want to say that airflow around F1 car is far more complex turbulent then around aircraft...so it must be harder to solve for CFD,much harder..
An airliner at cruise, yes. That same airliner on approach with landing gear down, flaps and slats extended, the effect of changing configurations, add in a cross wind too, is a different thing again.
If you are more fortunate than others, build a larger table not a taller fence.

Uwe
Uwe
1
Joined: 18 Dec 2020, 13:40

Re: How accurate is our "cheap" CFD ?

Post

Just_a_fan wrote:
20 Dec 2020, 13:19
Uwe wrote:
20 Dec 2020, 10:31
jjn9128 wrote:
20 Dec 2020, 10:15


Errrrrr Gurney flaps were called Nolders before Dan Gurney first used one. Most F1 tech comes from aerospace, only thing I can think of going the other way is moving ground plane in a wind tunnel.

Aerospace and motorsport aerodynamicists are the same people - some go from one industry to the other.
I just want to say that airflow around F1 car is far more complex turbulent then around aircraft...so it must be harder to solve for CFD,much harder..
An airliner at cruise, yes. That same airliner on approach with landing gear down, flaps and slats extended, the effect of changing configurations, add in a cross wind too, is a different thing again.
All aircrafts are design for cruise flight.
They dont care about wheels/gear aerodanyamics like F1 do, F1 wheels are critical part how airflow will moving around car and has great influenece of car performance.

F1 even must include rotating wheels in calculations becuase stationary wheels give completly wrong results

for example airliners dont include rotating fan at turbo fan engine when simulate plane flight...

User avatar
jjn9128
769
Joined: 02 May 2017, 23:53

Re: How accurate is our "cheap" CFD ?

Post

Hoffman900 wrote:
19 Dec 2020, 23:09
That seems optimistic. 1% is for what kind of design? A glider or a F1 car? How do you know what 1% is if you have nothing to go off of?

From an engineering side, just throwing a general value of 1% out there is kind of reckless without qualifying it.
You're right I was exaggerating for effect, one should never engage in hyperbole on internet forums. I have seen a paper comparing 35million cells to 2.3billion cells to experimental data, for automotive so large separated regions, and the difference is not all that great a few percent at most.
#aerogandalf
"There is one big friend. It is downforce. And once you have this it’s a big mate and it’s helping a lot." Robert Kubica

User avatar
jjn9128
769
Joined: 02 May 2017, 23:53

Re: How accurate is our "cheap" CFD ?

Post

Uwe wrote:
20 Dec 2020, 13:11
I think F1 rear wing has all aerodynamics in it, that airliners need to fly...
F1 much more complex to solve
Aviation isn't subject to some arbitrary resource requirement though, a lot of F1 (in the recent past at least) is RANS because they can churn though quite fast before going into LES/DES for more detailed work before signing off parts.

Dont' see the need for all this willy waving, both are complex problems.
#aerogandalf
"There is one big friend. It is downforce. And once you have this it’s a big mate and it’s helping a lot." Robert Kubica

Hoffman900
163
Joined: 13 Oct 2019, 03:02

Re: How accurate is our "cheap" CFD ?

Post

Uwe wrote:
20 Dec 2020, 13:36
Just_a_fan wrote:
20 Dec 2020, 13:19
Uwe wrote:
20 Dec 2020, 10:31


I just want to say that airflow around F1 car is far more complex turbulent then around aircraft...so it must be harder to solve for CFD,much harder..
An airliner at cruise, yes. That same airliner on approach with landing gear down, flaps and slats extended, the effect of changing configurations, add in a cross wind too, is a different thing again.
All aircrafts are design for cruise flight.
They dont care about wheels/gear aerodanyamics like F1 do, F1 wheels are critical part how airflow will moving around car and has great influenece of car performance.

F1 even must include rotating wheels in calculations becuase stationary wheels give completly wrong results

for example airliners dont include rotating fan at turbo fan engine when simulate plane flight...
I think you’re grossly underestimating things. Both are complex problems though and there is respect from both, for both disciplines within the aerodynamics community.

Hoffman900
163
Joined: 13 Oct 2019, 03:02

Re: How accurate is our "cheap" CFD ?

Post

jjn9128 wrote:
20 Dec 2020, 13:57
Hoffman900 wrote:
19 Dec 2020, 23:09
That seems optimistic. 1% is for what kind of design? A glider or a F1 car? How do you know what 1% is if you have nothing to go off of?

From an engineering side, just throwing a general value of 1% out there is kind of reckless without qualifying it.
You're right I was exaggerating for effect, one should never engage in hyperbole on internet forums. I have seen a paper comparing 35million cells to 2.3billion cells to experimental data, for automotive so large separated regions, and the difference is not all that great a few percent at most.
Right. I think the computing power is accessible to get pretty close from home (by renting time on clusters (I believe Fluent has this available)), but it’s the calibration of the model that is well beyond what anyone can do at home.

Even CFD of the a cylinder head / intake tract is going to require building an instrumented engine at some point, which depending on the performance envelope, can get pretty pricy (tens of thousands of dollars). You still need to establish the in cylinder pressure curve, which will effect your cfd model as well as ideally measuring port pressure for confirmation. You can use a 1d pressure dynamics sim to do help with development, but at some point, you need to calibrate that too.