Crude simulation for free standing mast - sail - luff curve ?

Discussion in 'Hydrodynamics and Aerodynamics' started by jmf11, Aug 6, 2021.

  1. jmf11
    Joined: Oct 2019
    Posts: 33
    Likes: 1, Points: 8
    Location: Aix en provence - france

    jmf11 Junior Member

    I'm a software guy with little knowledge of basic mechanical law. So I try to tinker with tools I don't fully understand. But it is still a way to learn some basics (and I like that process). Yes, my best bet is that the issue is at the top of the second section, at a joint or not far. Land Yacht seem to be be a mix of quite basic technology + good sails + super low drag => high speed. Masts are sections of Aluminum slotted one in the other, with "recipies" for local reinforcements (tubes in tubes) to achieve a curve (less engineered than my A-Cat CArbon mast). Top of secon section is 35x2 mm tube slotted in 40x2 mm tube, so 1mm gap, filled with tape wrapped round the smaller tube.

    I don't know how to model the stress and strain from the fit between pipes. So as you point out, I don't catch it. I dodel as is the joint was solid, which is quite optimistic.

    I just modeled 100N at the tip of the mast, base clamped. This is how the measure and compare Finn Masts. I could try to have a more realistic load with force in direction of the leech ov may sail.

    What do you mean by "It's only good the sail unloaded -maybe we can relate to leach tension" ? and by "pre-stressed curve needs to be accounted for ?"

    Thanks for all those perspectives Skyak,

    JMF
     
  2. Skyak
    Joined: Jul 2012
    Posts: 1,462
    Likes: 145, Points: 63, Legacy Rep: 152
    Location: United States

    Skyak Senior Member

    See if this helps intuition.
    You push your mast into the sail sleeve and set the outhaul (and vang?). The mast is bent into a curve. All the force vectors are coplanar (in the plane of the sail) and all the peak tension stresses are up the leading edge of the mast. Call these leach tension stresses
    Now think of the wind hitting a solid in the shape of the mast and sail. It would be a distributed load orthogonal to that original plane of the sail and the peak tension stresses would be on the windward side of the mast.
    As long as these stresses are orthogonal they are on different parts of the mast and don't add, but when there is insufficient leach tension for the wind, in reality, the top of the sail twists off and in the middle of the luff the wind induced stresses have come around and added to the leach tension stresses. I think this is the source of our kink, likely at a joint.

    Rather than go nuts trying to model this complex system in use I think we should just see if we can get a good leach tension first.
     
  3. AJB
    Joined: Jul 2021
    Posts: 56
    Likes: 8, Points: 8
    Location: 31 42S 152 04 E

    AJB Junior Member

    Keep it up men...

    nnn Skyak, the wind load distribution is neither planar, nor constant, or linear !

    Nice, simple problem !

    Ready, fire, aim IMO!
     
  4. jmf11
    Joined: Oct 2019
    Posts: 33
    Likes: 1, Points: 8
    Location: Aix en provence - france

    jmf11 Junior Member

    All feedbacks are helpful and the explanation to trigger my intuition also. As explained by Skyak, my mast simulation is too simplistic at the joints and do not capture the "kink" I experience.

    As advised by both, next step is to do a measurement of my mast in the real life, without the sail. Mast clamped in the Land Yacht mast step, cable from top of the mast to the boom attachment to simulate the leach tension, sheet in as in operation, and:
    - measure real mast bending (will help compare with simulation),
    - see if a kink appears at the top of second element, at or near the joint.

    Part of the stuff is at home, part at the sailing club. So no tests before next week.

    By the way, from mechanical perspective, are there advices about how should be designed / sized a "good joint" between 2 tubes (Ex OD 40mm and OD 35mm) => adequate overlap between the 2 tubes, reinforcements length and location ? I know that in structural wood assembly, scarf joints can be considered in the 1:10 to 1:20 range. Anything similar for tubes ? (but it is a bit off topic)

    JMF
     
  5. Skyak
    Joined: Jul 2012
    Posts: 1,462
    Likes: 145, Points: 63, Legacy Rep: 152
    Location: United States

    Skyak Senior Member

    I know! I simplified just to get to the reason why there is a stress kink located where it is.
    Two more oversimplified cases that will illuminate what is going on are a simple beam bending and a C section bending. The mast (leach tension) is a nonlinear spring transitioning from beam to C, or should be. If it takes too much bending to get appropriate leach tension you have a C to c bending. You don't have enough control of geometry or material properties to make C to c bending work in aluminum tube.

    If the mast can't get up to sufficient leach tension a new sail cut can't fix the problem. Have a look at the rig on the Clark's UFO foiling cat. We might be able to salvage the mast and maybe the sail using that concept.
     
  6. jmf11
    Joined: Oct 2019
    Posts: 33
    Likes: 1, Points: 8
    Location: Aix en provence - france

    jmf11 Junior Member

    Ughhh, don't have the background to understand that. I'm just at the level of wondering if adding a reinforcement tube in the faulty area could address the issue, ou increasing the overlap in that area. Also if I could better simulate the coupling of the tubes under FreeCAD FEM workbench, in order to try different configurations on the computer instead of sourcing, cutting, fixing suitable alu tubes...

    (I can share the mast tubes breakdown if could be usefull ; almost all land yachts have alu mast, so it is possible to have this work ; I also consider as alternative the option to used a windsurf mast)

    JMF
     
  7. Skyak
    Joined: Jul 2012
    Posts: 1,462
    Likes: 145, Points: 63, Legacy Rep: 152
    Location: United States

    Skyak Senior Member

    We are on the same page.
    -get some real deflection measurements -the mast under a simple load, and the leach tension of the current rig
    -get the current deflection of the joints to evaluate the value of 'fixing' them
    -do calculations for a new cut -luff tension and stress to look for improvement.

    I will check out freecad FEM capabilities to see if I can model actual wind load.

    BTW, to put this more in your expertise, this can be developed using camera surface data and machine learning. Any ML capabilities or interest?
    Do check out the UFO rig. They addressed the same type of problem and they know more than we will ever learn about small sail rigs.
    I am also curios why you are so partial to high reflex rigs. The impressive land and iceboats I have seen favor stiff wing masts with excellent twist and traveler controls. As you can see, reflex is a crazy complex trial and error mess. The wing mast is one shot certainty.
     
  8. jmf11
    Joined: Oct 2019
    Posts: 33
    Likes: 1, Points: 8
    Location: Aix en provence - france

    jmf11 Junior Member

    When you say "-get the current deflection of the joints to evaluate the value of 'fixing' them", does it looks like measuring the angles between the 2 tubes at joint level? Or something else?

    Unfortunately, no ML skills here. Interest for sure, but not enough spare time as hobby.

    I had a look at the UFO rig, with its inverted spreaders. Not so common layout. Could be the same principle for the Moth mast attachment (https://cst-marine.com/moth/)? However the main benefit of this small land yacht is its transportability, and the fact that it is set-up in 15 min. So I would be happy to keep things simple, without too many fittings.

    Why I'm partial about high reflex rigs: could be different explanations. This land yacht has such a rig. As a windfoiler, I often use those high efficiency windsurf sails. I have some interest in RC gliders, and tried to understand the loads on the wings, the role of the spar in the wing, that looks like the mast of the sail... Aeroelasticity concepts. On the A-Cat, the lateral-longitudinal flex are doing a great job keeping sailing easy by managing gusts automatically. Great sail makers did wonders in sail designs on well defined and consistent masts (Ashby, Landenberger, Glaser, Fiberfoam...)

    In fact two things collided:
    - I did not liked the shape of my mast + sail. Had difficulties to analyse the situation. Manufacturer decided not to address the issue. I wanted to understand and started to digg,
    - I had read some nice academic articles about CFD and simulating rigs (Class C, Finn...). Pictures helped understood what was hapening, and the design of some sails. This was Wow ! They have all those tools to assess the shape, aerodynamics, interaction between sails... But this is a different league from what I can do.

    JMF
     
  9. Skyak
    Joined: Jul 2012
    Posts: 1,462
    Likes: 145, Points: 63, Legacy Rep: 152
    Location: United States

    Skyak Senior Member

    What I had in mind for measuring the mast deflection was to pull the mast into the deflection of your sail luff from a known position somewhere on the boom, then take a laser pointer on the base of the mast and measure (distance dot to mast) at least every joint and the point midway between joints. It might be useful to do this again at a tension you want to achieve.

    I know there are lots of reflex rigs around and they are slick when they are right, but I know nobody that has done one right the first try and the people that have done so many they can make a good first guess are all tight lipped. BTW all the good reflex designs are just at the tip from what I can see.
     
  10. jmf11
    Joined: Oct 2019
    Posts: 33
    Likes: 1, Points: 8
    Location: Aix en provence - france

    jmf11 Junior Member

    Agree, and my mast sail is a good example of not so OK design. I had hoped that new tools could help "crack the code". I believe that Phantom (SAILS — Phantom Windsurfing https://www.phantom-windsurfing.com/sails) is using CFD and FSI for their windsurf sails and have some suceess (they removed their CFD pictures). But still not mainstream...

    So... I down grade my ambition to:
    - real objects measurements and trial and error experiments to solve the mast kink,
    - apply the pragmatic approach for my recut sails to bae luff curve on mast measured bend
    - ... and understand the mechanics and stress at the joint of 2 tubes, basic beam models

    Skyak, I browsed the net for details, and did not found a lot. Some elements for telescopic cranes, but those rely on square sections with sliding parts that bear all the efforts. So different from my case. Could help me identify the concepts I need to understand the sizing of the joint (which stress to assess at which place...)?

    Are there some basic facts / assumptions to simplify the sizing (is the small tube always the culprit, always at the same location?

    I attach my mast "design" and some other references I could find. And the picture of mast over sail where we see thha there is a joint near the expected kink.

    Not any more about aerodynamics... maybe should move to another subforum ;-)

    JMF
     

    Attached Files:

  11. Skyak
    Joined: Jul 2012
    Posts: 1,462
    Likes: 145, Points: 63, Legacy Rep: 152
    Location: United States

    Skyak Senior Member

    The condition of your rig is clearly too much deflection and the greater the deflection the greater the stress/leach tension. So it would be best if your joints had zero deflection (it they didn't rotate we could pre-stress counter-bend).
    If you wanted to do the actual stress strain of the joints you would have to calculate hertz contact stresses -FEA is not simple because the contact is a function of strain -nonlinear.
    That is a mess of math you don't want to do so here are some napkin sized bounds
    Deflection from fit Af=arctan((D-d)/overlap)
    when stressed the outer tube will oval to inner tube diam. As=arctan(((D*pi-d*pi)/2+(D-d))/overlap)
    The contribution to deflection would be =A (in radians)* the length of mast above that joint.
    Of course all the above assume you stay in the linear elastic range of the aluminum.

    I don't know what you filled gaps with, but if it isn't nearly as hard as aluminum it is giving up deflection and stiffness.

    If I understand the mast schematics chart, you have tubes inside tubes -I don't think these contribute much to stiffness other than eliminating buckling mode.

    Edit -for small angles arctan=angle in radians
     
    Last edited: Aug 13, 2021
  12. gonzo
    Joined: Aug 2002
    Posts: 16,790
    Likes: 1,714, Points: 123, Legacy Rep: 2031
    Location: Milwaukee, WI

    gonzo Senior Member

    How do you find a higher force/tension on the leech since it loosens when the mast bends? As the mast bends the distance between the clew and head is reduced.
     
  13. Skyak
    Joined: Jul 2012
    Posts: 1,462
    Likes: 145, Points: 63, Legacy Rep: 152
    Location: United States

    Skyak Senior Member

    I am not certain we will. If the current sail cut is fairly straight and if much of the deflection due to joint fit, maybe we can stiffen the mast and tighten the cut of the sail to get leach tension up. I am pretty sure that is the OPs original idea. Maybe stiffen by replacing a section or two with stiffer better fitting tubes. I would likely do rigging like the UFO -I have an idea for how to make the sail flatten in a gust -OP wants to keep it simple.

    About moving this from aero to some other thread -this thread is where the people who know look. If we get some data maybe one will step in and tell us how far this is from right.
     
  14. jmf11
    Joined: Oct 2019
    Posts: 33
    Likes: 1, Points: 8
    Location: Aix en provence - france

    jmf11 Junior Member

    I crunched the numbers according to the formulas above. Attached at the bottom. Deflexion from stress would account for twice the one of the fit issues. For a total of 7.5° along the complete mast. I yellow is highlighted the "problematic joint", which only gives 1.7°. All sections are of similar length (1150mm), except S4 (550mm).

    Tighter fit is achieved between the tubes with few turns of adhesive tape. Not sure it changes a lot under stress, but feels good in the hands.

    Yes, you understand right. There are tubes in tubes for "stiffening" (or buckling resistance) or for joining.
     

    Attached Files:


  15. jmf11
    Joined: Oct 2019
    Posts: 33
    Likes: 1, Points: 8
    Location: Aix en provence - france

    jmf11 Junior Member

    Oh, and to complement, my intention is:
    - either get my alu mast have an even bending under load, by adding stiffners at joints, if this is identified as the issue to solve (from FEM analysis, I got convinced that the overall sequence of tubes makes sense and should work "good enough" for me), or move to a windsuf carbon mast (stiffer, more consistent),
    - recut a windsurf sail to achieve a sail smaller than the one I have (have 4.7 m2, needs 3m2 and 2m2), and need to have a luff curve to draw,
    - learn in the process as much as I can on different topics.
     
Loading...
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.