Author Topic: Drag coefficient fundamentals  (Read 26169 times)

0 Members and 2 Guests are viewing this topic.

Offline jacksoni

  • Hero Member
  • *****
  • Posts: 1507
Re: Drag coefficient fundamentals
« Reply #30 on: August 27, 2017, 12:10:56 PM »
Ron, you've asked a good question - What does it cost to spend a session in a wind tunnel?  How many bucks do you need?

I dunno.  Back when the A2 tunnel (guy) was active on this Forum I seem to remember a cost in the $500/half day area.  That's only a hazy memory, though, and fresh information would be welcome.  I have no idea what Layne and Tom are charging at the Darko Tunnel - only that a lot of racers (and others) are using it - but they've got open times available.
I think is (still) $495/hour with 2 hour minimum. In October spent 6+hours and made 24 runs there. Runs are pretty quick. Changes between not so much. Need to be prepared with lots of tape, foam, strips, aluminum, cardboard etc etc and have a good idea what you want to do before arriving to save time.
Jack Iliff
 G/BGS-250.235 1987
 G/GC- 193.550 2021
  G/FAlt- 193.934 2021 (196.033 best)
 G/GMS-182.144 2019

Offline Bratfink

  • Full Member
  • ***
  • Posts: 81
Re: Drag coefficient fundamentals
« Reply #31 on: August 28, 2017, 10:32:31 AM »
Thanks for opening up the debate. I was getting worried there that I was ostracizing myself by raising the concern.

A2/ Darko (basically the same tunnel design) are certainly an affordable tunnel for trying out ideas. And they both have some really smart guys working for them.

The racing industry in the US has moved towards using Windshear which is significantly more expensive. But does correct a number of the inherent issues with a small closed section fixed ground tunnel. It's not perfect, but it's closer.

I am a little out of touch with F1 these days, but it doesn't surprise me to hear of RedBull's issues. Rules limit full scale testing time, and scale model testing is under increased limitation. Most teams are doubling down on CFD, but mistakes are still made in basic validation between the 3. In an f1 car all it takes is a slight separation point miscalculation and you go from the front of the grid to the back.

CFD is great tool (Woody you can forward the commission check later ;-) ) However, without validation to real or tunnel data you might as well be P*&$ing in the wind. The entire premise of CFD is based around solving 2 unknowns. In order to get one unknown (Lift/Drag) you have to make up the answer to the other. This is refereed to as the turbulence model. And which model you choose wildly changes the answer of the Lift/Drag. Not to mention all the other variables, Steady State vs transient etc. It's a real minefield. On a vehicle program (in my day job) I might do 10-15 transient CFD runs before I am happy the model correlates with my tunnel, not many people can afford 15 runs to throw away for validation. 

Those who say they prefer to test on the track are not as insane as they sound. With some data logging and some other basic principles a lot of data can be gathered about the aerodynamic performance of the car in the real world.

In short, continue to test and use tunnel facilities and CFD. But don't do so assuming the answers are right. always be critical of what you are looking at.

I am not trying to set myself up here, but if people feel they need help to understand numbers or procedures I'll try to help. I don't have all the answers, but I'll do my best.

I just want racers to be safe. Fundamentally these cars are on the edge of stability, and decisions made in a tunnel or CFD without knowing all the facts could push you over that edge. You don't know what you don't know. And the first key to solving a problem is knowing what you don't know.

Offline Bratfink

  • Full Member
  • ***
  • Posts: 81
Re: Drag coefficient fundamentals
« Reply #32 on: August 28, 2017, 02:41:31 PM »
Ron, you've asked a good question - What does it cost to spend a session in a wind tunnel?  How many bucks do you need?

I dunno.  Back when the A2 tunnel (guy) was active on this Forum I seem to remember a cost in the $500/half day area.  That's only a hazy memory, though, and fresh information would be welcome.  I have no idea what Layne and Tom are charging at the Darko Tunnel - only that a lot of racers (and others) are using it - but they've got open times available.
I think is (still) $495/hour with 2 hour minimum. In October spent 6+hours and made 24 runs there. Runs are pretty quick. Changes between not so much. Need to be prepared with lots of tape, foam, strips, aluminum, cardboard etc etc and have a good idea what you want to do before arriving to save time.

On the subject of brokered test time:
  • When Ford DTF (Sverdup) was still available at least one team from the flats went there. Unfortunately it's not available now.
  • AAWT (Chrysler) has been used recently by a factory supported team (I'll let you guess which one). But essentially is not available to racers.
  • GMAL is a) private and b) down for a refit.

So what is available?
  • Windshear is really too expensive for any but the best funded teams to run at, which begs the question, why don't the best funded teams run there?
  • ARC is more reasonable in tunnel time cost but since it's scale model only there is the added cost of building a model.
  • Lockheed is available, anyone asked if they can get in there?
  • NRC in Canada is available, is rolling road and has a huge test section, sounds perfect, right?

I am deliberately not releasing any prices here but needless to say they are all well above $500/hour.

What could be done for these high cost facilities is that a bunch of guys running similar class cars could all club together buy a shift and all share the data. Extend that idea and for a scale model facility, a reconfigurable model could be developed to fit a number of different classes, the cost shared among all the teams interested. Then combined tests could be had for those on a a budget and those wanting private testing could buy their own shifts. I had approached this idea previously but the facility in question didn't want to commit to the cost of a model without a firm commitment of sales.   

But the option is there and it can be revisited if there turns out to be the demand.

Offline Stainless1

  • Administrator
  • Hero Member
  • ***
  • Posts: 8939
  • Robert W. P. "Stainless" Steele
Re: Drag coefficient fundamentals
« Reply #33 on: August 28, 2017, 10:22:11 PM »
BF... ever been to Bonneville?  We do not have cars like nascar or F1, everything identical so info can be shared.... every vehicle running in the same class is different... even roadsters....
There may be some well funded teams, but rarely factory teams run, and when they do they often lose to backyard guys.... ask GM
Stainless
Red Hat 228.039, 2001, 65ci, Bockscar Lakester #1000 with a little N2O

Offline Bratfink

  • Full Member
  • ***
  • Posts: 81
Re: Drag coefficient fundamentals
« Reply #34 on: August 29, 2017, 09:02:00 AM »
BF... ever been to Bonneville?  We do not have cars like nascar or F1, everything identical so info can be shared.... every vehicle running in the same class is different... even roadsters....
There may be some well funded teams, but rarely factory teams run, and when they do they often lose to backyard guys.... ask GM


I may not be a veteran like many here but I have been most years since 08. I have worked on 4 cars in various different capacities from mechanic to driver to crew chief. No records yet, but working on it.

I disagree with you previous statement, many of the classes are similar enough (particularly in the roadsters) that a modular model system could be developed to replicate a large proportion of the field. They don't have to be identical to be aerodynamically similar. The point is is that that many of the basic tuning tips used by guys on the salt have a certain aerodynamic effect, it'd be quite useful for teams to understand the aero implications of making those changes to a roadster style vehicle. And having worked in Indycar, Nascar and on the periphery of F1 I can tell you the cars are far more dissimilar than you might imagine.

By factory teams I am not really referring to JCB, GM or any of that, I am referring to teams that through network contacts or sponsorship have managed to wrangle test time at a private facility.   

Offline Rex Schimmer

  • Hero Member
  • *****
  • Posts: 2624
  • Only time and money prevent completion!
Re: Drag coefficient fundamentals
« Reply #35 on: August 29, 2017, 04:56:13 PM »
BF,
What are your thought of testing, either wind tunnel of CFD, different parts of a car. My son and I have a lakester which means the wheels and supporting parts are in the wind, how about making say a front wheel assemble, wheel, spindle and axle, and then running that in a wind tunnel, then make changes to things like wheel covers, spindle axle fairings etc. and one of my real interest the distance the wheel if from the body? The wheels and axles of a lakester are the largest contributor aero drag so maybe just looking at them would (could) be effective.

Rex
Rex

Not much matters and the rest doesn't matter at all.

Offline Bratfink

  • Full Member
  • ***
  • Posts: 81
Re: Drag coefficient fundamentals
« Reply #36 on: August 29, 2017, 06:20:59 PM »
BF,
What are your thought of testing, either wind tunnel of CFD, different parts of a car. My son and I have a lakester which means the wheels and supporting parts are in the wind, how about making say a front wheel assemble, wheel, spindle and axle, and then running that in a wind tunnel, then make changes to things like wheel covers, spindle axle fairings etc. and one of my real interest the distance the wheel if from the body? The wheels and axles of a lakester are the largest contributor aero drag so maybe just looking at them would (could) be effective.

Rex

Excellent question!

More than just position of the wheel to body, you also want to consider the effects of the wheel rotating.

Single component testing used to be done in CFD when computing power was minimal and model complexity and compute times were an issue. These days it has fallen out of fashion. First it expanded to half car models and now full car. In the aircraft industry wing section 2d component testing resulted in the NACA profiles we all know and rear wing element assembly testing has been done by Indy teams for many years.

Benefit:
  • You can gauge the individual drag/lift increment of a change from testing an assembly without having to test the whole car.

Cost:
  • Is it worth the investment in creating the test rig to test just that one thing?
  • What you can't get out of that form of testing is the interaction to the rest of the car. And this can be enormously important. For drag reduction small changes in upstream parts can interact with the wake at the back of the vehicle in ways that result in much larger increases (or decreases) in the drag than the independent change would suggest.

The type of test you are suggesting I would set up in CFD using a Design Of Experiments (DOE) method. I would use the whole car model and in my meshing software create a morph of the suspension system allowing me to increase and decrease the offsets. I would run a number of increments of this and use another program to mathematically interpret the sensitivities of the certain variables. That way I don't have to do 100 CFD runs to get the full picture of the results.

I would then pick a few of the best outputs from that analysis to validate in the wind tunnel.

I realize that is way outside the scope of what you are asking. Just trying to give you an impression of how the industry might approach it.

Since wheels and suspension are so important in Lakesters. It might well be worth considering building a half car wind tunnel model to tune just that. And this sort of thing can be done in the back yard. If you have a body, you might well have a body mold which could be re-purposed as a wind tunnel buck. A simple suspension geometry could be fabricated, it doesn't need to hold the loads of the real car, just have the aerodynamic shape. The most complex thing is deciding if you want to rotate the wheels. Universities often have small wind tunnel facilities and no end of students looking for projects to help with grades. Might be worth approaching your local school. Wind is easy to create, just need to figure out how to keep it straight and find a novel method for measuring the loads. The Wright brothers did it, and they started as Bicycle mechanics. I have built my own simple tunnel to show school kids about aerodynamics, I might throw some plans up here for others to play with and expand on.       

Offline Stan Back

  • Hero Member
  • *****
  • Posts: 5875
Re: Drag coefficient fundamentals
« Reply #37 on: August 29, 2017, 06:48:47 PM »
This'll show you how sophisticated I am!

Whataya think about testing models in a captive hydraulic environment by measuring the overall pull of the flow?  (Huh?)
Past (Only) Member of the San Berdoo Roadsters -- "California's Most-Exclusive Roadster Club" -- 19 Years of Bonneville and/or El Mirage Street Roadster Records

Offline Elmo Rodge

  • Hero Member
  • *****
  • Posts: 1653
Re: Drag coefficient fundamentals
« Reply #38 on: August 29, 2017, 06:58:27 PM »
Stan, since I moved to Utah I don't have my swimming pool anymore. I sure thought about that, though.  :cheers:
Wayno

Offline manta22

  • Hero Member
  • *****
  • Posts: 4135
  • What, me worry?
Re: Drag coefficient fundamentals
« Reply #39 on: August 29, 2017, 07:13:59 PM »
This'll show you how sophisticated I am!

Whataya think about testing models in a captive hydraulic environment by measuring the overall pull of the flow?  (Huh?)

Stan;

I thought about that, too. How about letting a model roll down an underwater ramp and timing how long it takes to travel between two marks on the ramp. It won't give you quantitive data but it should show up differences between model shapes. The bigger the model the better.

Regards, Neil  Tucson, AZ
Regards, Neil  Tucson, AZ

Offline jacksoni

  • Hero Member
  • *****
  • Posts: 1507
Re: Drag coefficient fundamentals
« Reply #40 on: August 29, 2017, 08:53:14 PM »
This'll show you how sophisticated I am!

Whataya think about testing models in a captive hydraulic environment by measuring the overall pull of the flow?  (Huh?)
I have seen exactly this done with a model using some fancy apparatus nature of which I forget to make bubbles to show the flow. I don't remember if the guy measured forces or just the path and turbulence. Is on YouTube. I'll see if I can find it. Bratfink likely knows about it and may chime in about it.
Jack Iliff
 G/BGS-250.235 1987
 G/GC- 193.550 2021
  G/FAlt- 193.934 2021 (196.033 best)
 G/GMS-182.144 2019

Offline Jack Gifford

  • Hero Member
  • *****
  • Posts: 1568
Re: Drag coefficient fundamentals
« Reply #41 on: August 30, 2017, 12:34:19 AM »
So that's what Breedlove was doing in the canal! :lol:
M/T Pontiac hemi guru
F/BFL 1-mile Loring record 2020

Offline kiwi belly tank

  • Hero Member
  • *****
  • Posts: 3129
Re: Drag coefficient fundamentals
« Reply #42 on: August 30, 2017, 01:25:01 AM »
From my novice standpoint, that's the way I've always dealt with aerodynamics, using fluid dynamics. It allows me to get a better picture of whats going to happen & after all, air is a fluid.
  Sid. (Hillbilly with a hammer!)

Offline Seldom Seen Slim

  • Nancy and me and the pit bike
  • Administrator
  • Hero Member
  • ***
  • Posts: 13156
  • Nancy -- 201.913 mph record on a production ZX15!
    • Nancy and Jon's personal website.
Re: Drag coefficient fundamentals
« Reply #43 on: August 30, 2017, 07:24:17 AM »
I do remember Scott Guthrie telling me how the Charlie Toy bodywork had been designed using water (water tunnel).
Jon E. Wennerberg
 a/k/a Seldom Seen Slim
 Skandia, Michigan
 (that's way up north)
2 Club member x2
Owner of landracing.com

Offline jacksoni

  • Hero Member
  • *****
  • Posts: 1507
Re: Drag coefficient fundamentals
« Reply #44 on: August 30, 2017, 09:05:25 AM »
This'll show you how sophisticated I am!

Whataya think about testing models in a captive hydraulic environment by measuring the overall pull of the flow?  (Huh?)
I have seen exactly this done with a model using some fancy apparatus nature of which I forget to make bubbles to show the flow. I don't remember if the guy measured forces or just the path and turbulence. Is on YouTube. I'll see if I can find it. Bratfink likely knows about it and may chime in about it.

Here is the youtube link. This is the first of maybe 8 or more videos this fellow made using his water tunnel to illustrate flow around several cars. An accurate model is needed.  https://www.youtube.com/watch?v=quDLzxmJl5I
Jack Iliff
 G/BGS-250.235 1987
 G/GC- 193.550 2021
  G/FAlt- 193.934 2021 (196.033 best)
 G/GMS-182.144 2019