1. Attachments are working again! Check out this thread for more details and to report any other bugs.

Airflow Analysis by CFD

Discussion in 'Gen 2 Prius Accessories & Modifications' started by 200Volts, Mar 17, 2006.

  1. 200Volts

    200Volts Member

    Joined:
    Nov 19, 2004
    429
    43
    1
    Location:
    Northern CA
    Vehicle:
    2006 Prius
    Model:
    N/A
    OK, don't bust my chops for rough Prius model.
    I had a 2 week software tiral from work to learn Solidworks and Cosmos Floworks, make a 3D model of the Prius and do CFD modeling and outputs ( at night after my full time job).

    What you are looking at took about 3 hours of processing time (per run) on my 2.4 GHz gaming machine to do the CFD calculations(only) on about 20,000 grid cells on and around my 3D model. Some 3D images show clipping due to memory constraints on my PC.

    Calculated pressures are shown in colors (nominal pressure 14.5psi and velocity vectors (75mph oncoming wind) are the arrows(longer arrow means faster flow).

    I think this was pretty close to accurate as shown by:
    1-Flow is curling up around the back-top of the car, similar to streaks left by dew after a short drive.
    2-Windshield flow in front of the driver is about a 45 degree angle, similar to flow seen during rain.
    3-Only subtle pressure increases and decreases are shown over most of the body, indicating mostly laminar flow (good aerodynamics).

    Suprises are:
    a- The rear window-rear spoiler shows a small higher pressure zone (exactly where the BT Tech spoiler goes).
    b- A higher pressure zone is located between the rear wheels.
    c-The flow curling around the two rear sides-rear window area appear to meet exactly as they exit the back of the car (like a perfect vortex generator).
    d-Flow exiting the rear top and and rear bottom seem very symetrical.

    Expectations are:
    i)- The lowest pressure zone on the top of the car is from the driver area back to rear window ( suprise is the ulta low pressure area exactly where the antennae is located-lowest drag).
    ii)air does flow under the front of the car
    iii) air flows horizontally around the front bumper area to the sides.

    I don't pretend to know it all, but I hope this helps our understanding of what's going on.
     
  2. Frank Hudon

    Frank Hudon Senior Member

    Joined:
    Jan 11, 2004
    4,147
    18
    0
    interesting! to say the least, the area of high pressure behind the rear wheel could that be caused by the air exiting thru the wheel opening and the wheel building up in that area? in laminar flow the air flow is tightly on the surface and the air coming out of the wheel opening is unable to break the flow and there by build in that area?
     
  3. V8Cobrakid

    V8Cobrakid Green Handyman

    Joined:
    Jun 6, 2004
    3,790
    152
    0
    Location:
    Park View, Los Angeles, CA. U.S.A
    Vehicle:
    2004 Prius
    Model:
    N/A
    cool... so on the ( i think '05 and '06" they put that extra little black piece ont he front to make more air pass over the front? i know the nose of my car isn't as bold as the newer ones.

    low air between the rear wheels and the back spoiler is. so extending it helps eliminate that low pressure.. does low pressure create lift?

    if it does create lift we are getting it in the back of the car... so the bt spoiler helps bring the back down by giving it more of that ... is it called an air foil?

    what would happen if there was a little "skirt" type of thing installed in front of the rear axle? i wonder what that would do....

    if i'm completely wrong, please correct me.. lol.. it's late and i'm sleep deprived. I'm eager to learn though. :)
     
  4. robinsod

    robinsod New Member

    Joined:
    Jan 10, 2005
    37
    0
    0
    Location:
    Kerrville, TX
    Thanks for sharing your hard work. I learned a lot. Good job.
     
  5. TheForce

    TheForce Stop War! Lets Rave! Make Love!

    Joined:
    May 30, 2005
    3,461
    537
    0
    Location:
    Wheelersburg, Ohio
    Vehicle:
    2006 Prius
    Model:
    N/A
    From the looks of the pictures it explains how the rear bumber gets very dirty when the rest of the car is almost spotless.
     
  6. hdrygas

    hdrygas New Member

    Joined:
    May 22, 2004
    3,650
    6
    0
    Location:
    Olympia Wa
    Vehicle:
    2004 Prius
    Very Nice. Very Nice indeed. Explains why my back up cam is dirty. What about mud flaps. My daughter has access to a quad processor Mac, would that help.
     
  7. 200Volts

    200Volts Member

    Joined:
    Nov 19, 2004
    429
    43
    1
    Location:
    Northern CA
    Vehicle:
    2006 Prius
    Model:
    N/A
    Hdrygas, Thanks for the offer but the 2 week software trial is over.
     
  8. Bill Lumbergh

    Bill Lumbergh USAF Aircraft Maintainer

    Joined:
    Jul 5, 2005
    537
    7
    0
    Location:
    NorCal
    Vehicle:
    2005 Prius
    Model:
    N/A
    Fantastic model! Should you decide to purchase the software, you can do Hummer H2 model very easily. Hint, start with a perfect cube and remove the area over where the hood would be, then suspend it on four slightly smaller cubes to mimic tires. You should be able to run the model after a whole 30 seconds of work, and it should be pretty accurate. :p
     
  9. Sufferin' Prius Envy

    Sufferin' Prius Envy Platinum Member

    Joined:
    Jul 7, 2004
    3,998
    18
    0
    Location:
    USA
    Vehicle:
    Other Non-Hybrid
    The airfoil design is what give a wing, and a Prius, lift. An airfoil is a shape which is curved on one side and flat on the other side. The air which travels past the curve has a longer distance to travel than the air which passes the flat side. When the airfoil is moved through the air as designed, this causes there to be more molecules of air per square inch of surface area on the flat side and, in relation, less molecules per square inch on the larger curved side. That causes a pressure difference. The bottom of the wing, with its higher pressure is pushing the wing up, and the top, with its lower pressure is sucking the wing up . . . together this is lift.

    . . . does low pressure create lift? On a car it depends where the low pressure is located. If it is on top, then it is fighting the effects of gravity and is called lift. If it is on the bottom, the car is being sucked down toward the ground, and is known as “ground-effects†or downforce. This is good for formula-one race cars which need to make sharp turns at high speeds, but not so good for a passenger car because it takes tremendous amounts of power to move a car which is essentially sticking to the ground.
    If the low pressure is behind the car, it is actually sucking the car backwards, and is known as drag.

    The holy grail of car design would be to swap the low pressure at the back with the high pressure at the front. Your car would then be blown and sucked forward, and, if the car were not a perpetual motion machine, energy consumption would be minuscule.

    The BT spoiler helps bring the rear down by “spoiling†the airfoil design at that point, thereby decreasing the lift and exchanging it for more drag. Not a good thing for MPGs, but theoretically better for Prius stability.
     
  10. donee

    donee New Member

    Joined:
    Aug 15, 2005
    2,956
    197
    0
    Location:
    Chicagoland
    Vehicle:
    2010 Prius
    Model:
    III
    Hi Sufferin,

    While that is the common theory of wing lift taught in secondary schools, NASA teaches its incorrect in its extensive K-12 web site (http://www.grc.nasa.gov/WWW/K-12/airplane/foil2.html) .

    They show with models based on results from CFD analysis that the lift is not caused by low pressure, but by the rotating of the direction of air flow off the top of the wing. The resulting mass of the air thrust downward creates a reaction force of the wing upward they say. As evidence they point out that some of the best wings have are just curved plates, with nearly the same topside as downside surface areas. Also, even symetrical air-foils create lift.

    This may help explain while in racing reduction of aerodynmic form drag has not be wildley successful, but increase of downforce has. Air thrust upward goes off into the sky unecumbered creating a great downforce. Air thrust downward hits the road and changes direction suddenly creating drag. Aircraft even with all those wings and control surfaces have have Cd under .02 typically, while the Prius - one of the best cars - is at .26 (http://www.aerodyn.org/Drag/tables.html).

    The socalled Kamm (Prof. Kamm was a german aerodynacist in the 1930's)tail which the Prius and BT Spoiler are based on can be designed to minimize drag or spoil lift, depending on the details of the design. Typicially in road vehicles the goal is to reduce drag. This is done by providing some resistance to the flow over the top of the car thus preventing flow delamination and turbulence. Tubulence creates increased drag. Another common way (but not for cars) that area of turbulence is reduced is the dimples on a golf ball. This phenona is difficult to model due to the mathematically kaotic (Kaos Theory) nature of turbulence. Turbulent flow may not be modeled by 200Volts program. Possibly he knows more about this. That red zone on the rear of the Prius in 200Volts models is probably where turbulent flow is occuring.

    If one could take air from behind the radiator and with low drag guide it into that red area at the rear of the car, that would probably be the next step in reducing drag.
     
  11. Sufferin' Prius Envy

    Sufferin' Prius Envy Platinum Member

    Joined:
    Jul 7, 2004
    3,998
    18
    0
    Location:
    USA
    Vehicle:
    Other Non-Hybrid
    Yes, for simplification purposes I was using the “Longer Path" theory of lift . . . but you must remember, I was responding to V8Cobrakid . . . and was not trying to prove every theory of lift. Hell, I know that even a human head placed at the proper angle of attack (AOA) will produce lift . . . but I wasn't going to get into the theories involved in simultaneously conserving mass, momentum, and energy to describe simple lift.

    “Longer Path" theory of lift is valid . . . but not for all forms of lift. The reason an airfoil described as curved on the top and flat on the bottom is first used to teach aerodynamic lift is due to the fact that this shape will produce lift at a zero degree AOA . . . and, also in this case, it is the difference in pressure across this airfoil which produces the lift.

    Regarding “Longer Path" theory From the NASA link you gave:
    {The upper flow is faster and from Bernoulli's equation the pressure is lower. The difference in pressure across the airfoil produces the lift.} As we have seen in Experiment #1, this part of the theory is correct. In fact, this theory is very appealing because many parts of the theory are correct.
     
  12. clett

    clett New Member

    Joined:
    Jul 15, 2005
    537
    19
    0
    Location:
    Scotland
    Vehicle:
    Other Non-Hybrid
    Man, that big red area where the radiator is looks horrible! The sooner we get rid of that nasty IC engine the better.

    Lack of radiator inlet is one of the major reasons the GM EV1 managed such a cool Cd (0.19) for a production car.
     
  13. BT Tech

    BT Tech New Member

    Joined:
    Apr 28, 2005
    885
    2
    0
    Location:
    South Florida
    Vehicle:
    2005 Prius
    Excellent data Greg.. Thank you for taking the time to post it.


    Thanks!!

    Brian
    BT Tech
    305-652-3115



     
  14. 200Volts

    200Volts Member

    Joined:
    Nov 19, 2004
    429
    43
    1
    Location:
    Northern CA
    Vehicle:
    2006 Prius
    Model:
    N/A
    FYI, I did not cut a hole for radiator intake. The only cut outs are the wheel wells (and not the volume between the wheel wells). I never realized how complex the Prius is shaped. Trying to accurately model it for me was a bear. My biggest hangup was trying to loft complex splines. I couldn't do this hardly at all and therefore I used extrusions and extrusion cuts (plus fillets & chamfers). Here's my simplified approach (borrowed from others on the web and adapted).
    1-Use photoshop to open side, top, front and back views. Proportion all views and save as individual files.
    2-Use Autocad(or another CAD) to trace the outlines of the images from step 1 and save as dxf files.
    3-Use solidworks to import dxf onto drawing planes. (all 4 sides)
    Here is were it gets tricky. I originally wanted an outline of the outside edge of the car and was going to loft from the center to this outline. So instead I did this:
    4-Align all the dxf outlines, with the side view outline running right down the center of your future car.
    5-Extrude the side view way past were the tires are.
    6-Extrude CUT using the front view outline. This will give the bend for the fenders and windows.
    7-Extrude CUT using the top view to shape the front and rear bumpers and sides.
    8-Now try to fillet, or chamfer or plane-fillet the sides and top surfaces. I couldn't get this to work very well either. This rounds the sides to top and bumpers and bottom to sides.
    9-Use new planes to cut the wheel wells.
    10-Mirror the whole thing. Do the mirrors if you're brave. I couldn't get them to mirror.
    11- Draw tires, fillet the edges, duplicate and repeat.
    12-Check the bottom and fill in any areas bewtween the wheel wells if needed.
    Here is your basic shape, the extrude cuts are why the bumpers aren't flat all the way around.
    The better way to make this model would be in Maya, since you can move individual verticies to get a better shape. Then save as IGES and import into solidworks. COSMOS is very well written and relatively simple to use compare to other CFD software.
     
  15. NuShrike

    NuShrike Active Member

    Joined:
    Feb 21, 2005
    1,378
    7
    0
    Vehicle:
    2005 Prius
    Model:
    Five
    Hey great job! It doesn't look like it, but I'm wondering if you modeled those spoiler flaps/spats in front of all 4 wheels and what effect they have on the airflow.

    You've been long time discussing that these flaps could be causing Prius instability at highway speeds hence your front-spoiler mod guide.
     
  16. darelldd

    darelldd Prius is our Gas Guzzler

    Joined:
    Jan 17, 2006
    6,057
    389
    0
    Location:
    Northern CA
    Vehicle:
    2006 Prius
    I'm gonna name my child after you!

    Remove the stupid plate, and just look at the smooth beauty that I used to drive!

    [Broken External Image]:http://www.darelldd.com/ev/images/ev1/ev1_gallery/ev1b.jpg
     
  17. darelldd

    darelldd Prius is our Gas Guzzler

    Joined:
    Jan 17, 2006
    6,057
    389
    0
    Location:
    Northern CA
    Vehicle:
    2006 Prius
    Uh-oh. News to me. Anybody have a pointer?
     
  18. hyo silver

    hyo silver Awaaaaay

    Joined:
    Mar 2, 2005
    15,232
    1,563
    0
    Location:
    off into the sunset
    Vehicle:
    2004 Prius
    Model:
    N/A
    While I admire and respect the effort and intelligence that has gone into the computer modelling, I have to wonder if there isn't an easier way. I recall reading something in an old Scientific American about building your own wind tunnel with a big fan, a few bags, and lots of cardboard. Would a scale model Prius in a scale wind tunnel be of any help?
     
  19. 200Volts

    200Volts Member

    Joined:
    Nov 19, 2004
    429
    43
    1
    Location:
    Northern CA
    Vehicle:
    2006 Prius
    Model:
    N/A
    Yeh, a fan and a model of a Prius. Anyone know where to find a plastic model of a Prius? My initial idea was to build a plastic model and scan it in with a 3D laser scanner.

    Actually the most accurate would be to tape 2 inch strands of yarn all over the front and back of the Prius and video tape what happens during cruise, passing a truck, cross winds, etc.

    And yes, the night after I got my front spoiler mods into 3D (and the BT Tech spoiler) my license expired......dooooh.

    NuShrike, my model used 20,000 grids around the car (including about 6 feet in front and behind the car (plus 3 feet on the sides and top). This is actually kind of coarse, so small details won't necessarily show up (like the air dams). You basically define a volume around the car and then define the liquid stream (air) flowing into the volume. The computer adjusted the grid automatically and I didn't have time to isolate them for a detailed study.