How to bevel Nomex Honeycomb

Discussion in 'Composites' started by PittsS1S, Oct 26, 2019.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Nov 13, 2019 #21

    Scheny

    Scheny

    Scheny

    Well-Known Member

    Joined:
    Feb 26, 2019
    Messages:
    152
    Likes Received:
    76
    Location:
    Vienna, Austria
    Hi Bob. The method I referred to, is using vacuum infusion for the outer carbon layers, then applying a super thin glass (64g/m2) soaked in resin with some excess resin, then honeycomb and vacuum again.

    Afterwards another thin glass layer plus the carbon, but without a soaking layer on the outside. All excess goes to the honey. This method is used for some of the new planes.

    BTW, I have a question for you, but I will PM you.

    BR, Andreas
     
  2. Nov 13, 2019 #22

    RSD

    RSD

    RSD

    Well-Known Member

    Joined:
    May 20, 2019
    Messages:
    210
    Likes Received:
    42
    If you workout how to PM him can you please let me know the secret?:)
     
  3. Nov 13, 2019 #23

    wsimpso1

    wsimpso1

    wsimpso1

    Super Moderator Staff Member

    Joined:
    Oct 19, 2003
    Messages:
    6,004
    Likes Received:
    3,260
    Location:
    Saline Michigan
    I suspect that a nice sharp carbide router bit spun at router speeds would do the job nicely. Marshall's book describes cutters that resemble engine intake valves sharpened to razor edges for sculpting Nomex cores.

    Filling the edges for a 30 degree bevel with dry micro can be heavy (4.5 pounds/gallon is the best I can do with the stuff) and negate gains from the rest of the core. You have to run your own numbers on this stuff to decide what your tradeoffs are.

    I just ran 3 PCF honeycomb, a 30 degree chamfer, and edge fill with dry micro at 4.5 lb/gallon to fill the edges for machining. I compared it to just using 6 pcf PVC foam (Divinycel). I picked square panels, 1/4 PVC foam core is lighter when the total core is less than 19" on a side, 3/8" PVC foam core is lighter when the core is less than 31" on a side, and 1/2" PVC foam core is lighter when core is less than 41" on a side.

    So, if your parts are big, even micro filled edges on honeycomb are lighter than PVC. Smaller parts are lighter in PVC. Me, If I found that I could not machine the cores plain, I would prefer to either just use PVC foam cores for the tapered edges or fill the edges with pour foam.

    Have fun,

    Billski
     
    Last edited: Nov 13, 2019
    BoKu likes this.
  4. Nov 13, 2019 #24

    Bille Floyd

    Bille Floyd

    Bille Floyd

    Well-Known Member

    Joined:
    Sep 26, 2019
    Messages:
    64
    Likes Received:
    28
    You can practice on a few test parts first . Once honeycomb
    is properly adhered, to a 3k- woven carbon or glass, and the
    epoxy has gone off ; that's when You break out the block sander.

    Honeycomb is glued to itself to make each line of cells, and each
    cell holds the integrity for the next row of cell . You can sand your
    bevel by stroking WITH the honeycombs glue lines, and the cells
    will hold together quite well ; sand against those glue
    lines, and the cells will disintegrate. You can sand a bevel
    at most any angle ; and the cells can be tapered to the thickness
    of a piece of paper if needed.

    Try doing above , when one side of the honeycomb is not
    glued to anything ; then sanding a bevel becomes quite
    problematic.

    Don't attempt to adhere Kevlar, directly to the cell walls
    of honeycomb ; it will rip off kinda easy.

    Boku is correct --- don't try to learn what you NEED to know
    about the proper use of honeycomb from off the internet ;
    it ain't gonna go well ! Honeycomb is Awesome stuff when
    used correctly ; need i say more ?

    Bille
     
    wsimpso1 likes this.
  5. Nov 13, 2019 #25

    PittsS1S

    PittsS1S

    PittsS1S

    Well-Known Member

    Joined:
    Nov 22, 2011
    Messages:
    48
    Likes Received:
    3
    Location:
    Paris/France
    First,i will infuse the outer carbon layers ,one spreadtow of 80 grs and one biax + - 45 ° of 200 grs. Second,put some resin one a flat surface and soak one side of the honeycomb in that resin, then install the honeycomb to the outer skin and vacuum .Third ,soak the inner layer,one biax +- 45° of 100 grs , with resin one a flat surface like making some prepreg and vacuum it to the first layer with honeycomb...May be the Andreas idea to use a thin layer of glass is very nice
     
  6. Nov 13, 2019 #26

    BoKu

    BoKu

    BoKu

    Well-Known Member HBA Supporter

    Joined:
    Aug 16, 2013
    Messages:
    2,348
    Likes Received:
    2,982
    Location:
    Western US
    Three vacuum cycles for a single part sounds like a huge amount of work per unit area. Just out of curiosity, what are you building that requires such heroic attention to weight?
     
    wsimpso1 likes this.
  7. Nov 13, 2019 #27

    PittsS1S

    PittsS1S

    PittsS1S

    Well-Known Member

    Joined:
    Nov 22, 2011
    Messages:
    48
    Likes Received:
    3
    Location:
    Paris/France
    Skin for Pitts S1 wings
     
  8. Nov 13, 2019 #28

    BoKu

    BoKu

    BoKu

    Well-Known Member HBA Supporter

    Joined:
    Aug 16, 2013
    Messages:
    2,348
    Likes Received:
    2,982
    Location:
    Western US
    I would definitely just rip those out using Divynicell H60 PVC foam core and be done with them.
     
    wsimpso1 likes this.
  9. Nov 13, 2019 #29

    PittsS1S

    PittsS1S

    PittsS1S

    Well-Known Member

    Joined:
    Nov 22, 2011
    Messages:
    48
    Likes Received:
    3
    Location:
    Paris/France
    Will be a lot easier,but heavier;reason why i want to use honeycomb . 29 kgs/m3 vs 60 kgs/m3. With a 3 mms core ,i will go from 87 grs/m2 to 180 grs/m2
     
  10. Nov 13, 2019 #30

    wsimpso1

    wsimpso1

    wsimpso1

    Super Moderator Staff Member

    Joined:
    Oct 19, 2003
    Messages:
    6,004
    Likes Received:
    3,260
    Location:
    Saline Michigan
    OK, let's do some math. 6 pcf foam 1/4" thick is 0.125 lb/ft^2, while the honeycomb is 3 pcf or about 0.0625 lb/ft^2. If you methods add 0.0625 lb/ft to the honeycomb, you will be ahead to just use foam. How much is 0.0625 lb/ft^2? Well, let's figure out how much area of paper we have. 1/8" cell size means we have about 96 cells the short way across the cells and 96*0.866=83 the long way. The side of the hexogon is about 0.072. Let's be generous and assume all of the next layer of paper is glued to this one - that gives us three full lengths of that side dimension per cell, and we get one side wet per cell. The area of paper that is likely to get wetted when we do this by homebuilders methods are:

    t*n1*n2*L*3 = 0.25*96*(96*0.866)0.072*3 = 431 in^2.

    OK, how much thickness is 0.0625 lb spread over that area. 0.0625 lb/ft^2 of panel is 1.45E-4 lb/in^2 of nomex paper and resin is somewhere around 0.052 lb/in^3. 1.45E-4/0.052 = 0.0028" So, if your edges tapering and bonding added nothing and you wicked a layer of epoxy 0.003" thick onto the Nomex paper, you would be heavier than if you just used the same 1/4" thickness of PVC foam. Let's go metric 6 mm cores, 3 mm cells, and a layer of resin 0.075mm thick on the Nomex, and you are lighter with foam.

    Now to finish the reality check:
    • The methods being described above will add weight around the edges and in the spar troughs, so you can afford even less wetting of the cores to stay on par with foam;
    • Each of the three steps has both process risk (the chance to scrap the part) and some incremental extra weight that you will not have in doing the skin in one shot;
    • I fully expect that if you build a couple skins each way and put them on a scale, your Nomex cored parts will be heavier than the PVC cored ones.

    Pitts wings? I assume they are going to only have ribs at the Butt Line and the tips. Seriously, use Divinycel cores with each of your eight skins, each molded in one shot. Then smile that the job is done in eight sessions with parts about the same weight as the ones that took twenty four sessions. Well, more likely nine sessions instead of thirty or more, because you will have a learning curve with scrap, but a lot worse curve with the honeycomb.

    And yeah, I did all of that math to just end up agreeing with Bob (above).

    Billski
     
    PittsS1S, BoKu and stanislavz like this.
  11. Nov 13, 2019 #31

    BoKu

    BoKu

    BoKu

    Well-Known Member HBA Supporter

    Joined:
    Aug 16, 2013
    Messages:
    2,348
    Likes Received:
    2,982
    Location:
    Western US
    Yeah, good luck with that.

    Heavier? Yes. A lot? That is relative. The sandwich core is a relatively small portion of the overall wing skin mass.

    The process you describe has a lot of steps, and each step has potential to ruin the part.

    Also, for an aerobatic airplane I think you're going to want to use more carbon on the outside. I'd go with 12oz carbon biax on the outside, 6oz biax inside, and 6mm core, with tapes of 12oz biax along the leading edge and trailing edge or aileron cove.

    [Edit add: TL;DR: yeah, what Bilski said.]
     
    petroelb, PittsS1S and wsimpso1 like this.
  12. Nov 13, 2019 #32

    wsimpso1

    wsimpso1

    wsimpso1

    Super Moderator Staff Member

    Joined:
    Oct 19, 2003
    Messages:
    6,004
    Likes Received:
    3,260
    Location:
    Saline Michigan
    I do not believe you will see much weight difference. If your methods are perfect, you will still have to add the film adhesive layer on both sides, the edges and spar troughs will gain weight, your technique will not be really good until you are done with the job, and your parts will save little weight but expend way more effort in the build.

    If you just have to try to prove it to yourself, go ahead and try a couple parts both ways in one mold, and see how much difference there is or isn't. Oh, and please report back to us on that.

    Thread Drift Alert! If you are really after a weight reduction, Pitts Special wing chord is in the zone for you to save weight by doing the wing on a massive core, Rutan Long-Ez style. Generally, if your chord is less than 6 feet, you can save weight with a massive core redesign. Yes, the foam weighs something, but the foam takes the place of: fiber resin inner layers on the cores; ribs; flanges and glue lines for bonding. Until the chord exceeds 6 feet, foam is lighter than hollow. Hardpoints are easier and lighter to make work too. Look hard at hotwiring the cores and then going full Long Ez, you might be surprised at how much more labor and weight you can save. The big drawback to this is the wings will have to be white or yellow or a pastel shade... Thread Drift Alert OFF.

    Billski
     
    stanislavz, PittsS1S and BoKu like this.
  13. Nov 13, 2019 #33

    PittsS1S

    PittsS1S

    PittsS1S

    Well-Known Member

    Joined:
    Nov 22, 2011
    Messages:
    48
    Likes Received:
    3
    Location:
    Paris/France
    Thank you all for the advices
     
  14. Nov 13, 2019 #34

    BJC

    BJC

    BJC

    Well-Known Member HBA Supporter

    Joined:
    Oct 7, 2013
    Messages:
    9,714
    Likes Received:
    6,505
    Location:
    97FL, Florida, USA
    Pitts:

    What are you wanting to accomplish with composite skins? Better aero? Standard S-1S airfoils?

    What internal structure? Composite or standard wood?

    Thanks,


    BJC
     
    BoKu likes this.
  15. Nov 13, 2019 #35

    PittsS1S

    PittsS1S

    PittsS1S

    Well-Known Member

    Joined:
    Nov 22, 2011
    Messages:
    48
    Likes Received:
    3
    Location:
    Paris/France
    I cannot have more than of 4 mms thick for the sandwich . The core will run only between each spars ..19 " from the front axis spars to rear axis spars .The wings chords are from 36,25 " to 37,5 ".Each spars will have 3 mms thick x 30 mms wide carbon pultruded caps . Less ribs ;only butts , at ailerons hinges,near flying and landings wires attach and may be one more if i have to keep the drag wires .All those ribs will be compression ribs build with spruce capstrips and carbon skin.
    Pitts S1 uncovered wings weigh around 50 kgs with and area of 9,2 m2
    My wings have a lot bigger ailerons in chord and span,but those original wings will not have enough torsional strength to handle them ;so the carbon pultruded caps with the biax sandwich skins could be a good answer..may be...with not more than 8 kgs....may be it is not possible.
    I am going to have a look to divinycell and rohacell core....May be ,as advices,it is the way i have to follow
     
  16. Nov 13, 2019 #36

    BoKu

    BoKu

    BoKu

    Well-Known Member HBA Supporter

    Joined:
    Aug 16, 2013
    Messages:
    2,348
    Likes Received:
    2,982
    Location:
    Western US
    I'd suggest caution regarding that assumption. That works fine in gliders largely because the span loading is so low, so no particular slice of wing is strongly loaded. With lower aspect ratio wings, the shear that reacts lift loads from the skin into the spar is greater, so a sandwich skin on its own might not have enough shear strength to do that. I'd think some analysis is called for.

    When my friend Steve Smith started building carbon fiber RV-6 wings in my shop, I was surprised at how many ribs were required.
     
  17. Nov 14, 2019 #37

    RSD

    RSD

    RSD

    Well-Known Member

    Joined:
    May 20, 2019
    Messages:
    210
    Likes Received:
    42
    Pitts - take a look at that youtube video that I posted of the Lantor Soric with different carbon fibre layups - it may give you the answer to what you are seeking.
     
    PittsS1S likes this.
  18. Nov 14, 2019 #38

    wsimpso1

    wsimpso1

    wsimpso1

    Super Moderator Staff Member

    Joined:
    Oct 19, 2003
    Messages:
    6,004
    Likes Received:
    3,260
    Location:
    Saline Michigan
    Was just assuming that because there have been other composite wings built and sold for the Pitts that are "ribless", ribs only at tip and root. I thought maybe this one was doing that too.

    Most of us will end up with a few ribs for things like supporting hinge points and aileron bellcranks and the like. Thicker cores let us omit some ribs that are in there for skin panel strength reasons. How much airfoil distortion we are willing to put up with also comes into finding the lightest combination of skin core thickness and rib count.

    If we are comparing same core thickness with Nomex honeycomb and PVC foam, I imagine the rib count issue will scale the same in both core types. I suspect that there is little to be gained on composite rib count with one or the other core material, particularly as rib weight is largely the attaching flanges, transfer flanges, and glue lines.

    Billski
     
    Last edited: Nov 14, 2019
    stanislavz and BoKu like this.
  19. Nov 14, 2019 #39

    PittsS1S

    PittsS1S

    PittsS1S

    Well-Known Member

    Joined:
    Nov 22, 2011
    Messages:
    48
    Likes Received:
    3
    Location:
    Paris/France
    I will have a total of 7 ribs by wings,all compression ribs from 3/4" to 7/8" thick .The thickness of each spruce spars is 11/16" .
    As Boku said,i am wonderind about the shear from the skin into the spars...and the skin will be very thin,around 5/32" .
    BJC. Standard S1SS airfoils,spruce spars with pultruded carbon caps,Compression ribs with spruce capstrips ans carbon skin.The goal,better tortional strength to handle the bigger ailerons and get a better roll rate..above 360° per seconde espected.
    Billsky .I did start to have a look at the PVC and Rohacell core....Of course,i could to the skin in one step using them and the infusion method ...and there is more area glued from the carbon to the core compared to the Nomex Honeycomb...May be using a thinner core,let say 2 mms and a thicker outer layer of carbon will be better.
    RSD. I had a look to that video ,thank you
     
  20. Nov 14, 2019 #40

    BJC

    BJC

    BJC

    Well-Known Member HBA Supporter

    Joined:
    Oct 7, 2013
    Messages:
    9,714
    Likes Received:
    6,505
    Location:
    97FL, Florida, USA
    I’m not certain of the roll rate of Wolf wings, but they probably are in the 360 degrees per second range. Their internal structure is not very different from standard wings.


    BJC
     

Share This Page



arrow_white