Hydrofoil exercise to validate CFD analysis

Discussion in 'Hydrodynamics and Aerodynamics' started by quequen, Oct 1, 2014.

  1. quequen
    Joined: Jul 2009
    Posts: 370
    Likes: 15, Points: 28, Legacy Rep: 199
    Location: argentina

    quequen Senior Member

    This is intended for newbies in CFD (like me). Attached is a 3D file of an hydrofoil, the idea is to calculate and compare results from different sources, given some predefined conditions. Anyone learning CFD could compare his own results with other's ones. I thought on this initial conditions:

    Immersion: 95% of height
    Velocities: 4m/s, 8m/s
    Angle of Attack: 0º, 4º
    Fluid: Fresh Water (999kg/m3)
    No free surface effects
    No lateral inclination

    This is an "L" hydrofoil intended for a little dinghy (something like this). Airfoil section is Tom Speer's H105.
    Measures are in milimeters.
    I'll post my own outcomes soon.

    -
     

    Attached Files:

  2. tspeer
    Joined: Feb 2002
    Posts: 2,319
    Likes: 303, Points: 83, Legacy Rep: 1673
    Location: Port Gamble, Washington, USA

    tspeer Senior Member

    If you want to validate your ability to do a CFD model, I suggest starting with a configuration for which experimental data exist. A good example would be Kuhn, John C. and Scragg, Carl A., "Analysis of Lift and Drag on a Surface Piercing Foil", Eleventh Chesapeake Sailing Yacht Symposium, January 1993.

    If you can reproduce the results for a simple straight foil, then you can move on to L foils.
     
  3. quequen
    Joined: Jul 2009
    Posts: 370
    Likes: 15, Points: 28, Legacy Rep: 199
    Location: argentina

    quequen Senior Member

    Thanks Mr. Speer, that was my first attempt, in fact I searched for T foils (mainly from Moths) but couldn't find any CFD analysis, so decided to make my own move. I will try to get that paper, off course!

    Anyway, this are my first results for 0º AoA:
    AoA: 0º
    Velocity: 4m/s
    Lift: 52N
    Drag: 7.5N

    Velocity: 8m/s
    Lift: 300N
    Drag: 35N

    I'm using SolidWorks FlowSimulation, the only CFD tool I have access to, it's an external analysis, automatic grid is set to his maximum (8). Convergence is left automatic. I expected more differences between flow trajectories at different speeds.
    Some images attached

    -
     

    Attached Files:

  4. daiquiri
    Joined: May 2004
    Posts: 5,371
    Likes: 258, Points: 93, Legacy Rep: 3380
    Location: Italy (Garda Lake) and Croatia (Istria)

    daiquiri Engineering and Design

    I have used the Cosmos FlowSimulation (now SW Floworks) some 8-9 years ago, and though I don't remember all the details of it, I do recall that it wasn't giving much control and choices to the user. Only the k-epsilon turbulence model was available. The mesh type was fixed and the resolution was roughly controlled with a slider, with no further info. That's how it was back then, and I guess the product has evolved over the years. The way it was made, it was a scarcely useful software except for a rough check that the flow will go in the right direction inside a 3-way valve or similar.

    Now, if the mesh is what is visible in this pic: Grid.jpg , then it is absolutely insufficient for any serious analysis. It is too rough, and hence misses all the important things which are happening close to the foil surface.

    You need to run some simulations of known cases first, as Tom Speer said, in order to compare the reliability of the numerical outputs from your CFD. Right now, without prior benchmark testing, I wouldn't feel too confident about the numbers you got.

    Cheers
     
  5. quequen
    Joined: Jul 2009
    Posts: 370
    Likes: 15, Points: 28, Legacy Rep: 199
    Location: argentina

    quequen Senior Member

    Daiquiri, thanks for the advice. FlowSimulation is pretty the same old tool you describe, computational grid can be refined manually but I didn't reach that point yet. My hope is someone else run this analysis using stronger tools like Aqua or OpenFoam in order to see where I'm standing. It is difficult to find the 3D file in wich each analysis is based on, that's why I published my own file.
    I'll appreciate references to other documents similar to the one indicated by Tom.

    This are my outcomes for the AoA=4º condition:

    Velocity: 4 m/s
    Lift: 198 N
    Drag: 6.8 N

    Velocity: 8 m/s
    Lift: 1076 N
    Drag: 34 N

    and some images attached:
    (Note that I'm calling Lift to the total vertical force perp to water plane, and drag to the total horizontal force)

    -
     

    Attached Files:

  6. daiquiri
    Joined: May 2004
    Posts: 5,371
    Likes: 258, Points: 93, Legacy Rep: 3380
    Location: Italy (Garda Lake) and Croatia (Istria)

    daiquiri Engineering and Design

    Just one note to the above numbers. Frankly, I don't trust calculations which yield a Lift/Drag ratio of 30 for surface-piercing foils without seeing it confirmed by experimental data. I am not saying that it is not possible, just that one should be cautious before trusting it.
    But Mr. Speer is certainly in a better position to evaluate your numbers and compare them with real-life data.

    Cheers
     
  7. Joakim
    Joined: Apr 2004
    Posts: 892
    Likes: 53, Points: 28, Legacy Rep: 422
    Location: Finland

    Joakim Senior Member

    Have you already done a simple 2D CFD simulation of an airfoil? That would be a better starting point, if you are a newbie in CFD. There are plenty of measured data at different AoA and Re. It is much easier to experiment the effect of grid and turbulence models in 2D.
     
  8. baeckmo
    Joined: Jun 2009
    Posts: 1,654
    Likes: 670, Points: 113, Legacy Rep: 1165
    Location: Sweden

    baeckmo Hydrodynamics

    Very valid Point; pure calculation ("number twisting") can be regarded one-dimensional and the logical next step is to get familiar with two-dimensional analysis.

    A simple 2D cfd, like "EasyCFD" or similar, will get you started with the grid manipulation techniques and also the shortcomings of the process (which are considerable....).
     
  9. daiquiri
    Joined: May 2004
    Posts: 5,371
    Likes: 258, Points: 93, Legacy Rep: 3380
    Location: Italy (Garda Lake) and Croatia (Istria)

    daiquiri Engineering and Design

    I agree. EasyCFD is a very well-made, easy to use and flexible low-cost tool, although with limited choice of turbulence models. But more than sufficient for simple 2D fluid-dynamics analysis. A simple 3D tool which can be used for quick analysis of cases like this one is XFLR5, which is open-source by the way.
    Cheers
     
  10. Joakim
    Joined: Apr 2004
    Posts: 892
    Likes: 53, Points: 28, Legacy Rep: 422
    Location: Finland

    Joakim Senior Member

    XLFR5 is a potential flow program with boundary layer correction. You can use it as a verification tool for CFD, but you won't learn anything about grids, turbulence models etc. from it. For 2D foils in "normal" situations XLFR5 is much more accurate than CFD unless you really know how do get most accurate CFD results.

    I think it is best to use the same software you are going to use for 3D, but just use it for 2D while learning. Then you can use the same models and grid densities in the real problem.
     
  11. Mikko Brummer
    Joined: May 2006
    Posts: 574
    Likes: 83, Points: 28, Legacy Rep: 258
    Location: Finland

    Mikko Brummer Senior Member

    Quequen,

    Daiquiri is right, your mesh resolution is much too coarse, as shown by the screenshots. You should refine the initial mesh closer to the foil, and also use a local initial mesh, with more pronounced refinement.

    The Solidworks Flowsim is a version of FlowEFD by Mentor Graphics. It uses wall functions to model the boundary layer close to the model, so you won't need nearly as dense a bodymesh as with many traditional codes. It's good enough for preliminary work, especially for blunt bodies (like vehicles), but not so good for lifting bodies when lots of separation is present. Easy to use, fast and the cartesian mesh is no hinder when it is well resolved.

    Your foil is surface piercing, and Flowsim does not do free surface, so you cannot expect very good results even if you had something to validate against. What boundary condition are you using on the surface now - symmetry? I agree with Tom that you should validate with some experimental results, not against another CFD result. But a surface piercing foil may not be the best choice for validation.

    Another somewhat similar product is Dr. Hanley's Stallion, more oriented towards lifting bodies (airplanes really). Dr. Hanley shows good validation cases, and the code is relatively inexpensive, too.
     
  12. Doug Halsey
    Joined: Feb 2007
    Posts: 640
    Likes: 212, Points: 53, Legacy Rep: 160
    Location: California, USA

    Doug Halsey Senior Member

    A good friend of mine is considering purchasing this program for a project he is working on, so we would appreciate any feedback about it.
     
  13. quequen
    Joined: Jul 2009
    Posts: 370
    Likes: 15, Points: 28, Legacy Rep: 199
    Location: argentina

    quequen Senior Member

    Joakim, FS can do 2D cuts from a 3D body and compute it as a full 2D analysis of infinite span, anyway I don't think a 2D analysis can make big contribution to this special case where the foil has an arbitrary shape far from a standard planar shape. My sandard approach to wings of conventional planar form is using xflr5 and/or Javafoil to determine Cl/Cd, and maybe a lifting line spreadsheet. Unfortunately, xflr5 can't create 3D bodies of arbitrary shape and has no import capabilities.

    Mikko, I use to think that mirroring is a good assumption for standard (almost vertical) daggerboards and rudders under a relatively big almost-planar hull. I'm not shure it's a good idea for this case. I know free surface effects should be strong in this "L" foil, but not as strong as a standard daggerboard with big lateral lift. Hence I'm trying to minimize them setting the foil vertical and well submerged, with 0º AoA on the horizontal plane. By the way, $1900 is far away from my budget...
    Have you considered making some runs with some of your powerful CFDs ? I can export the file on the format you need.

    I've finally got the paper Mr. Speer suggested. Seems to be a rectangular planar form of a conventional daggerboard/rudder, under a flat hull and at 25º of heel. That's a case where free-surface effects should be terrific, hence I can't model it with FS.

    Here are my outcomes with a partially refined cartesian grid all around the horizontal branch of the L foil. Daiquiri's doubts on the accuracy of this analysis are even clearer now as L/D is even higher.

    Velocity: 8 m/s
    AoA: 4º
    Lift: 1498 N
    Drag: 36.4 N
    _
     

    Attached Files:

  14. Joakim
    Joined: Apr 2004
    Posts: 892
    Likes: 53, Points: 28, Legacy Rep: 422
    Location: Finland

    Joakim Senior Member

    The 2D simulation is not supposed to give results for your case. It is meant for learning how to model foils in CFD. When you have learned how to do it accurately for a simple 2D case, you have a chance of getting good results for the real case in a 3D simulation.
     

  15. Mikko Brummer
    Joined: May 2006
    Posts: 574
    Likes: 83, Points: 28, Legacy Rep: 258
    Location: Finland

    Mikko Brummer Senior Member

    Your resolution on the horisontal wing is sufficient now, but you should use more on the vertical part too. Looking at you convergence curves, your second one is labeled "GG friction" - I assume your drag number is not just friction, but you have included pressure & induced drag as well?
     
Loading...
Similar Threads
  1. Maarten88
    Replies:
    5
    Views:
    999
  2. S V
    Replies:
    0
    Views:
    1,876
  3. Eytan Levi
    Replies:
    0
    Views:
    1,688
  4. container
    Replies:
    7
    Views:
    4,086
  5. vejas
    Replies:
    34
    Views:
    11,946
  6. revintage
    Replies:
    8
    Views:
    2,604
  7. revintage
    Replies:
    14
    Views:
    17,341
  8. rallyhybrid
    Replies:
    39
    Views:
    6,409
  9. optiwings
    Replies:
    1
    Views:
    2,482
  10. Tommifin
    Replies:
    5
    Views:
    3,102
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.