Plywood fuselage design methods

Discussion in 'Wood Construction' started by mcrae0104, Nov 21, 2017.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Nov 21, 2017 #1

    mcrae0104

    mcrae0104

    mcrae0104

    Armchair Mafia Conspirator HBA Supporter

    Joined:
    Oct 27, 2009
    Messages:
    2,961
    Likes Received:
    1,957
    Location:
    BDU, BJC
    In Stress Without Tears, Tom Rhodes discusses the design of a cone-shaped plywood fuselage tail section (chapters 25-27). I understand his method of analyzing this "beam" for bending, shear, and torsion; however, there seems to be little to no guidance out there in either this simplified book or in Bruhn on:

    • Design of the stiffeners (or formers). Just how much moment do they need to be designed for? If we ignore aero loads normal to the fuselage then the only thing to cause bending in the formers would be buckling. Is there a method to determine the forces introduced to the formers by skin buckling? Can you point me to a source I can read? Bruhn does present a method for turning these frames into an equivalent determinate structure. It's a little over my head but I think I can slog through it--but I need to determine what loads to apply to the frames first.
    • Cockpit opening. The tail cone is relatively straightforward as a cantilevered beam. Now where I have a cockpit opening, it seems simplest to treat the two side walls of the fueselage as beams with the longerons acting as caps. How does the curved shear web of this beam affect it? (Imagine a venetian blind used as a beam.)
    • Firewall to canopy. Like the tail section, this is hollow beam. However, because of the concentrated loads at the motor mount points, longerons will be used to distribute the load into the almost-round beam. Also, it would be nice to have some openings in the side of the fuselage here for a forward baggage compartment--precisely where the structure doesn't want them. For these two reasons, I am thinking this section would most easily be analyzed as either a) a beam, similar to the fuselage sides under the canopy, but with openings, or b) as a truss. How would you approach it?
    • Longerons and I of a hollow beam. Am I correct to include the longerons when calculating I of the fuselage? (Rhodes had no longerons in his example and I did not see a similar example in Bruhn.)
    Based on what I am learning, I strongly suspect that most stressed-skin plywood aircraft (particularly the former frames) are not fully analyzed, but taken partway and given either a TLAR check or the "well-it-worked-for-that-other-design" test. I am also beginning to think I may need to build a full-scale test article to gain full confidence in the design (although I don't really want to build two airplanes!). Sure does make a steel truss look easy from a designer's perspective.

    Thanks for any guidance you can offer on how to analyze this animal.
     
  2. Nov 21, 2017 #2

    TFF

    TFF

    TFF

    Well-Known Member

    Joined:
    Apr 28, 2010
    Messages:
    11,545
    Likes Received:
    3,230
    Location:
    Memphis, TN
    Although engineered, any airplane designed Golden Age or before will not have the time spent on it like today. The Falco is the only newer plane and still it was an early post war plane. When computers became the first test pilots, it changed how it was done. What planes are you trying to use as examples?
     
  3. Nov 21, 2017 #3

    mcrae0104

    mcrae0104

    mcrae0104

    Armchair Mafia Conspirator HBA Supporter

    Joined:
    Oct 27, 2009
    Messages:
    2,961
    Likes Received:
    1,957
    Location:
    BDU, BJC
    The Falco would serve nicely as an example of the question regarding fuselage bending in the area of the cockpit opening. See how the fuselage sides form a beam with a curved web? One of my questions is how to analyze that beam.

    Capture.JPG
     
  4. Nov 22, 2017 #4

    Mad MAC

    Mad MAC

    Mad MAC

    Well-Known Member

    Joined:
    Dec 9, 2004
    Messages:
    412
    Likes Received:
    167
    Location:
    Hamilton New Zealand
    Bruhn which one, I understand there is more analysis of wooden structure in "Analysis and Design of Airplane Structures"

    Have you had a dig through "Wood Handbook, Wood as an Engineering Material" might be of some value.
    https://www.fpl.fs.fed.us/products/publications/specific_pub.php?posting_id=18102&header_id=p

    There is a NACA war time report on the state of the art in stress analysis of wooden aircraft design in the UK which pretty much concludes that there was nothing to learn from the UK at that point.
     
  5. Nov 22, 2017 #5

    mcrae0104

    mcrae0104

    mcrae0104

    Armchair Mafia Conspirator HBA Supporter

    Joined:
    Oct 27, 2009
    Messages:
    2,961
    Likes Received:
    1,957
    Location:
    BDU, BJC
    I have the 1949 edition of Bruhn. It took a while, but as soon as I found the earlier edition at a reasonable price I snapped it up. I'll take a look at that FPL handbook as well--thank you.

    I think I may have found a start in ANC-18, which suggests analyzing a fuselage using the same methods as a closed-section wing cell. ANC-18 isn't aimed at the novice but I think I can make sense of it in combination with Bruhn.

    The engineering training I have had was limited to statics & strength of materials and beam theory with application to my particular field of work, which is unrelated. This was a good start, but many of the things we need to analyze on an aircraft structure are a different ball game. Part of my difficulty is that I don't know what I don't know and therefore it's hard to know the right questions to ask. I'll keep working through it and post what I learn.
     
    BoKu and Topaz like this.
  6. Nov 22, 2017 #6

    proppastie

    proppastie

    proppastie

    Well-Known Member

    Joined:
    Feb 19, 2012
    Messages:
    3,710
    Likes Received:
    1,077
    Location:
    NJ
    As I understand it, the gust loading on the tail and the inertial loading of the weights, (pilot, fuel, engine, structure) are what apply. According to "Glider Criteria" my tail loading is 12 PSF for my speed. Interestingly the pilot moment in my bird is almost the same at the wing attach fittings for 6 Gs
     
  7. Nov 23, 2017 #7

    BBerson

    BBerson

    BBerson

    Well-Known Member HBA Supporter

    Joined:
    Dec 16, 2007
    Messages:
    11,861
    Likes Received:
    2,254
    Location:
    Port Townsend WA
    The cockpit opening should be designed with extra material for crash loads, flight loads are not that much.
    Load testing the fuselage for flight loads only shouldn't hurt it.
     
    Last edited: Nov 23, 2017
    BoKu likes this.
  8. Nov 23, 2017 #8

    mcrae0104

    mcrae0104

    mcrae0104

    Armchair Mafia Conspirator HBA Supporter

    Joined:
    Oct 27, 2009
    Messages:
    2,961
    Likes Received:
    1,957
    Location:
    BDU, BJC
    Thanks; my question about loading was about the ring frames in particular (which are just serving as stiffeners to the skin). Perhaps I could have been more specific. As I understand it, they're not carrying any load* until the skin buckles or transitions to a semi-tension field. The question, then, is how to determine how stiff they need to in order to resist skin buckling?

    *Actually, I think they will be carrying some (negligible) aerodynamic loads normal to the skin.
     
  9. Nov 23, 2017 #9

    proppastie

    proppastie

    proppastie

    Well-Known Member

    Joined:
    Feb 19, 2012
    Messages:
    3,710
    Likes Received:
    1,077
    Location:
    NJ
    Not sure about that...the stronger stiffener, is carrying much of the load, the skin gives a larger effective leg thickness and length to that stiffener element. (and I really am unsure about that answer) Bruhn C7 (1973 edition) I use Gerrard Method to calculate Stiffener strength. Curved panel buckling does carry lots of load though. I do not have stessed skin except for my leading edges so I am not the best source. I do have lots of angle stiffeners. Bruhn shows lots of different ways, and lets you decide the best way, which does not help the novice as much as we would like.
     
  10. Nov 26, 2017 #10

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,786
    Likes Received:
    3,011
    Location:
    Saline Michigan
    mcrae104,

    Not analyzed? I have not looked much at wooden fuselages - not many out there in my class airplane, but I will argue that Dr Frati and George Pereira did pay a bunch of attention in these areas.

    First let's look at steel tube fuselages. You typically see one of several things: The cage is on both sides of the opening. All of the high wing birds where there is substantial 3-D structure above and below the door openings. Open cockpit birds either have doors that only go down a foot or so, or have only the top opening, again with substantial 3-D to the structures through the area of the opening.

    The tandems usually have a smaller opening on one side and a bigger one on the other (window left, door right on Cubs, etc) that also allow for less of a "Cut" in the structure that has to be made up for. High wings have it easy as there is a roof and floor that can be built with tubes applied in 3-D. Lessons in wood include substantial frames supporting panels forming the openings and box structure for the floor.

    In wooden fuselages, this means more structural fuss. The Falco has no doors, but there are substantial ring bulkheads and top longerons where there is no top to the fuselage. This allows the fuselage to be a channel through that part and remain stable so that the walls do not buckle. Model the fuselage beam just behind the opening, aft and forward edges of the opening, and just forward of the opening. Similar shear, bending moment, and torsion either side of a discontinuity, but different structures and transition between them. Model the top longeron in buckling with L as the length between only substantial frames. You can put a deep console between seats in side-by, hang a bit of a center console off a ceiling in the same. Tandems can have doors on one side and window on the other so the structure is much more of a channel on its side.

    Now none of this means you can not have an open cockpit or sliding canopy with full depth doors, just that you will pay a weight penalty and make your build a little more complicated. Of all the places to analyze, the cockpit with its openings is the place where you have to get creative to either do analysis or build and test and rebuild. Or, as I like to point out, copy a successful design (monkey-see, monkey-do) and then test. The thought of the cockpit structure collapsing around you and thus trapping you in the cockpit of a doomed bird is just too scary. You gotta make sure it will be sturdy, both in flight and in a forced landing.

    I did talk about ways of analyzing this in the other thread you started on wood. Or you could spend a bunch of money on build-test-iterate.

    Billski
     
    mcrae0104 likes this.
  11. Nov 26, 2017 #11

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,786
    Likes Received:
    3,011
    Location:
    Saline Michigan
    I addressed how to analyze the structure in my misplaced response on your other wood fuselage thread. http://www.homebuiltairplanes.com/forums/showthread.php?t=28251&page=2

    Stiffeners - When you analyze a panel and find that it will buckle, you can thicken the panel, shorten the span or add to its EI in the critical direction. A stiffener is adding a bunch of EI to the panel against the buckling. Your math model of the panel used to be just the local EI of the panel. Add in a stiffener, and you have its local EI in the same orientation plus you get to calculate the neutral axis of the assembly and sum up the EAx^2 of the parts to represent the assembly, then re-run the Euler calculation. Iterate until happy;

    Cockpit opening - talked about that in another post above. The tube becomes a channel, and your frames serve to keep the channel walls straight. Dr Frati's frames near and through the cockpit resemble the flying buttresses of the old cathedrals, don't they? If you put in full depth doors and the top is open, the floor has to take on depth and frames, maybe box structures;

    Firewall to canopy - substantial longerons, substantial frames at the front of the opening, maybe more longerons through this section, perhaps beefier walls from firewall through the cockpit either through doublers or box structures.

    Longerons - covered them in the other thread post too. Very definitely include them. Allocate bending and twisting moments per the EI fractions and calculate.

    Most of this is analyzable (if conservatively) with the methods I talked about earlier. The ones that are not are just how beefy to make the frames to prevent buckling the top longeron. If you do enough analysis of the rest, and you are guessing at the frames, you may need some test here. I strongly recommend monkey-see, monkey-do on frame and top longeron design to increase the probability that the test will only need to be done once.


    Billski
     
    mcrae0104 likes this.
  12. Nov 27, 2017 #12

    DaveD

    DaveD

    DaveD

    Well-Known Member

    Joined:
    May 28, 2010
    Messages:
    320
    Likes Received:
    152
    Location:
    Perth, Western Australia
    It depends on your definition of "negligible" but I think your ring frames will see some loading prior to skin buckling. If the fuselage is acting as a beam, then bending loads will attempt to distort the cross section. For example landing gear loads attempting to bend the fuselage in the vertical plane will cause loads which try and bring the top and bottom surfaces of the fuselage closer together and force the sides further apart (think of the deformation when you bend a pipe). This will translate to loading on the ring frames which I don't believe is insignificant.
     
    wsimpso1 likes this.
  13. Nov 27, 2017 #13

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,786
    Likes Received:
    3,011
    Location:
    Saline Michigan
    Your half-ring frames will carry plenty of load before buckling of the skins. They will be supporting the longerons and skin, breaking them into smaller pieces. They will also be resisting the longeron deflections away from centerline when the tail is pulling up and towards each other when the tail is pulling down. I would taper the half-rings to be deeper at the bottom and thinner at the top to get much more stiffness and strength per unit weight.

    The frames (perpendicular to the long axis of the airplane) serve to carry some of the "inflation" load due to pressure difference between inside and outside of the airplane, and also serve to break up the skin and longerons into shorter elements under compression from fuselage bending. Critical load under compression (buckling onset) goes with the inverse of the square of the length of the element and the degree of constraint. If the skins and longerons are continuous across the frames, they are already well constrained per Euler.

    The half-frames should be beefy indeed to support the skin and top longeron. How beefy? I do not know, but I would look at the Falco and GP4 for inspiration. Definitely tie the fuselage sides together solidly at the seat back bulkhead and at the instrument panel to make the region supported by only a couple half-rings shorter. I would do a test working up to 2 times Max bending moment in both positive and negative g. Simple fixture with only the cockpit area like you intend to fly, the rest can be simple stuff.

    One other comment on this bird. You can probably do yourself some favors by using vertical fuselage walls through the wing, and either straight walls or slightly expanding walls through the wing too. The very sexy tapering elliptical fuselage shape through the wing that the Lancairs and the Falco (and P-40 and Spitfire) usually interfere with the wing and crowd you where you can use that space for half-frames that flare as you go toward the floor... You want to see what a low drag fuselage actually looks like, look at the P-51.

    Billski
     
    Last edited: Nov 27, 2017
  14. Dec 6, 2017 #14

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,786
    Likes Received:
    3,011
    Location:
    Saline Michigan
    What I keep hoping will come out is someone experienced in the art of designing wooden fuselages will come forward and tell us some rules of thumb that are always safe or criteria than otherwise simplify the design process. I was never hoping to terminate the discussion with my full-on analytical approach that may well be over-the-top in areas...

    Billski
     
  15. Dec 6, 2017 #15

    mcrae0104

    mcrae0104

    mcrae0104

    Armchair Mafia Conspirator HBA Supporter

    Joined:
    Oct 27, 2009
    Messages:
    2,961
    Likes Received:
    1,957
    Location:
    BDU, BJC
    Haha... that wasn't over-the-top. For my part, I'm working out some configuration stuff that will inform the fuselage structural design. Expect more chatter in a month or so from me (if I actually get to take my scheduled vacation time around Christmas).
     
    wsimpso1 likes this.
  16. Dec 8, 2017 #16

    mcrae0104

    mcrae0104

    mcrae0104

    Armchair Mafia Conspirator HBA Supporter

    Joined:
    Oct 27, 2009
    Messages:
    2,961
    Likes Received:
    1,957
    Location:
    BDU, BJC
    [video=youtube_share;5hOqjd2tgJw]https://youtu.be/5hOqjd2tgJw[/video]
     
  17. Dec 8, 2017 #17

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,786
    Likes Received:
    3,011
    Location:
    Saline Michigan
    Nice review. He selected his problem so it was easy to reduce the math and then easy to follow. It also had the simplification that the skin contributed nothing to bending and then he did not do torsion, which a vertical tail imposes. Last simplification is that all of the parts in his analysis have the same E. Yeah, if you are just getting all of this stuff in your head for the first time, you should go through this to train your brain, because what comes ahead is built from here..

    You could do the same thing, but then you will really overbuild the longerons in exchange for making the math easier... The plywood skins may contribute more than the longerons, and are definitely in range, so you would be wise to count them all...

    You, my friend are choosing to work in wood, which means you are working in Composite Theory due to the fact that E and G the long way in the longerons is usually substantially different from E and G in the plywood, and both will have significant contributions. You can not work in stress space, but must instead work in strain space. For each of the pieces of the fuselage at any section, you will find EA, EAx, EAy, the centroid x (neutral axis for yaw bending) by Sum(EAx)/Sum(EA), centroid y (neutral axis for pitch bending) by sum(EAy)/sum(EA), Local EIxx, Eiyy, GJ, re-dimension positions of each element relative to the centroids, then global EIxx, EIyy, GJ, then sum EI's and GJ to get the whole section stiffness. Now you can apportion bending in pitch and in yaw and shear per the global fractions of each or by doing the whole [ABBD] matrix and finally go through the solution for global and local strains. I find the [ABBD] to be easier to keep things straight if more involved to write out initially. Either way, discipline is needed and I will look at your work in Excel if you want.

    Billski
     
    mcrae0104 likes this.
  18. Dec 8, 2017 #18

    Pops

    Pops

    Pops

    Well-Known Member

    Joined:
    Jan 1, 2013
    Messages:
    7,023
    Likes Received:
    5,920
    Location:
    USA.
    You mention the construction of the Falco and the GP4. Also much can be learned by the all wood construction of the Falconar F-12, stressed for +9 and - 6 .
     
  19. Dec 8, 2017 #19

    proppastie

    proppastie

    proppastie

    Well-Known Member

    Joined:
    Feb 19, 2012
    Messages:
    3,710
    Likes Received:
    1,077
    Location:
    NJ
    As I do this I often wonder exactly what that means in the promotional flyers.... Is that limit, ultimate, sizing of material, limit + FOS, Ultimate + FOS. That they do not publish their numbers it is sort of hard to know......Certified not much more forthcoming, That gear on the PA 140 or the Mooney airframe/wing, is I imagine a lot higher stressed than limit + FOS.
     
    mcrae0104 likes this.
  20. Feb 18, 2018 #20

    larr

    larr

    larr

    Well-Known Member

    Joined:
    Apr 25, 2012
    Messages:
    157
    Likes Received:
    44
    Location:
    markham, ontario, canada
    Hmmm ... time for an 'Old School" moment:
    ply box girder fuselage p1.jpg ply box girder fuselage p2.jpg symbols p1.jpg symbols p2.jpg symbols p3.jpg fuselage construction p1.jpg fuselage construction p2.jpg fuselage construction p3.jpg fuselage construction p4.jpg fuselage construction p5.jpg fuselage construction p6.jpg fuselage construction p7.jpg
    This is the mid 1930's view, but it may be applicable here.

    The DH Moth Minor might be a good starting point for a practical example.
     
    mcrae0104, Marc W and BJC like this.

Share This Page

arrow_white