Need info on composite

Discussion in 'Composites' started by brandon81, Mar 2, 2010.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Mar 4, 2010 #21

    orion

    orion

    orion

    Well-Known Member

    Joined:
    Mar 3, 2003
    Messages:
    5,800
    Likes Received:
    135
    Location:
    Western Washington
    Actually my techniques are no different than that of the industry - a very large fraction of aerospace composites, glass or graphite, are made with conventional ovens, not autoclaves. The trick in achieving good properties is not so much a trick as it is an exercise in following proper layup procedures, especially in layups that have a substantial number of plies. Since thick laminates would tend to have difficulties in achieving a proper level of compaction (flatness), it is very important to laminate, bag and vacuum debulk the layup every five layers or so. Failure to do so can result in poor compaction (a laminate that is not as flat as you'd want it to be) or a layup with voids. Both are of course to be avoided.

    Keep in mind however that the autoclave will not necessarily fix this so even then it is important to use the debulk techniques throughout the layup schedule.

    And regarding the higher epoxy fraction, that's more an issue of choice. In the past I've been ordering prepreg with about 34% to 36% resin content but in doing so I've noticed that my surface quality tends to vary a bit. While this is somewhat an issue of the resin content, it seems to be more the issue of my tools (longer story here so I wont bore you with the details - bottom line, tooling cost issue) so to fix that I've gone to a resin content of 38% to about 40%. This is a bit stickier than I like but it gives me a very consistent surface without the need for a sheet molding compound (sheet layer of epoxy without the fiber reinforcement).

    Strength wise this higher resin content is not much of an issue since the laminate properties are calculated for the fiber and material rather than the finished laminate thickness (which doesn't vary all that much anyway).
     
    Norman likes this.
  2. Mar 7, 2010 #22

    durabol

    durabol

    durabol

    Well-Known Member

    Joined:
    Mar 2, 2010
    Messages:
    169
    Likes Received:
    3
    Location:
    Lethbridge, Alberta / Canada
    I disagree with what most people have written about needing to do complex and detailed structural analysis in order to build a safe airplane. A good book is Composite Basics by Andrew Marshall, although it is just an elementary treatise. I think the math is good enough along with copying proven designs (buy some used plans) and it is important to test composite structures since the strength can vary depending on construction skills. Regular checks for delamination and load tests are a good idea as well.

    For an airplane that flies above 120mph control flutter can be an issue so control surfaces should be built as light as possible and mass balanced on either side of the control hinge.

    For optimum strength pre-preg cloth, vacuum bagging, heat curing in an autoclave all result in about twice as much strength as moldless techniques. Also I have read the carbon fibre doesn't give its great strength with moldless techniques.

    Although I like the idea of "rolling my own" if proven plans are available I would definitely go that way. Unfortunately most plans nowadays seem to come as "kits" which are rather expensive and often don't have to construction types one wants.

    Brock
     
  3. Mar 8, 2010 #23

    orion

    orion

    orion

    Well-Known Member

    Joined:
    Mar 3, 2003
    Messages:
    5,800
    Likes Received:
    135
    Location:
    Western Washington
    While the monkey-see-monkey-do and simple math approach to structural configuration does have some merit, it generally does so only as a methodology for checking ones work, rather than be an end all process for design. One of the problems in taking a given structure and adapting it to one's own work is simply that it is rarely known what assumptions (design loads and load factors, operating envelope, materials and processes, quality control, etc.) the initial designer used for said development. It is also critical to know what precise materials were used and of course how.

    Saying that you're going to build a component in a particular manner because someone built something similar is sort of like saying that you're going to build an aluminum structure of the same metal, without knowing the precise makeup of the aluminum materials (alloy) themselves, nor how to properly form them, join them or treat them, nor what loads and conditions were used in the design of the assembly. The new structure may look functional but may not have the structural margins intended. Yes, some things can be copied but usually at some level of risk.

    In composites this level of risk is magnified by at least a magnitude since the variables in design are that much greater. True, if the designer has a lot of experience with the materials, and with the design process itself, it would be possible to design a reasonably functional structure, at a controllable level of risk. But to suggest that without said experience one can simply copy similar work, without having a full understanding of what went into said structure, is simply our industry's form of Russian Roulette. In other words, are you really willing to fly behind "good enough"?

    On the practical side, the poster really makes the point for me. If a knowledgeable person, with the proper data and experience designed said structure, it is quite possible that testing will most likely not be necessary, nor will continuous tests for delamination or other forms of structural deterioration, since the structure will most likely will have been designed right in the first place. Do you really want to fly in an airplane that you have to worry about all the time?

    It's important to remember that composites do not yield so there is little or no warning prior to total failure. It is possible that the eyeballed structure may provide functional service for some time however, at some point the structure may be required to provide safe service at the extreme of the design envelope. Can it? Without the proper analysis there really is no way to know. Yes, you could build it and test it to destruction but is that all there's to it? Hardly. Most of these tests that we hear about load only in static bending, totally neglecting torsion, effect of gust (short period load), drag, etc. Actual flight loads are anything but static. Without the proper design, you will never know what you can expect from your plane, nor whether it will, some time down the road, bite you.
     
  4. Mar 8, 2010 #24

    durabol

    durabol

    durabol

    Well-Known Member

    Joined:
    Mar 2, 2010
    Messages:
    169
    Likes Received:
    3
    Location:
    Lethbridge, Alberta / Canada
    Certainly if one is going to copy something the structure to be copied should be very similar, but if it is very similar then I see no reason for "reinventing the wheel".

    When the **** hits the fan almost all structural engineering and engineering in general is monkey-see-monkey-do. The people who do structural analysis are just copying and applying other work, who developed the methods and procedures in the first place, regardless of how involved the structural engineering is. Also the data for structural engineering comes from testing materials in a real way. The more we know about a material the more accurate we can theoretically analysis it. There is no "magical" factor in stress engineering as some seem to believe.

    As of yet I don't believe we don't have the data on fatigue life for composite materials (even the big boys) and inspection and testing should be routinely done. I know of a study in Australia that determine composite gliders fatigue life at just a few thousand hours, make everything beefier resulted in 10s of thousand of hours. If composite parts are build to high tolerances than life long expectancy can be used. Burt Rutan claims his designs will last a lifetime, similar to some sheet metal airplanes.

    Also one can accurately test for dynamic loads an airframe may encounter with simple "static" tests.

    Brock
     
  5. Mar 8, 2010 #25

    orion

    orion

    orion

    Well-Known Member

    Joined:
    Mar 3, 2003
    Messages:
    5,800
    Likes Received:
    135
    Location:
    Western Washington
    Data for composite fatigue life has been around for over fifteen years. I actually remember seeing this data in the early eighties when i was still at General Dynamics. Since then, with a bit of digging you can find S-n plots for most of the popularly used materials and weaves. Conveniently though, the S-n plot shapes and values are very similar for most classes of material so an exact match is not always necessary.
     
  6. Mar 8, 2010 #26

    autoreply

    autoreply

    autoreply

    Moderator

    Joined:
    Jul 8, 2009
    Messages:
    10,732
    Likes Received:
    2,542
    Location:
    Rotterdam, Netherlands
    As for composite fatigue; as far as I know the Janus (one of the very first aircraft to use carbon structurally, back in 79 or so) recently got it's lifetime upgraded to 18000 hours. That plane was designed with a ultimate/max ratio of 1.75 and no fatigue in mind. If I recall correctly the residual strength (in crashed ones) was only a few percent below design values after over 10000 hours.
    Those planes are used for teaching and thus frequently are loaded to the maximum g-ratings and severe turbulence.
     
  7. Mar 8, 2010 #27

    bmcj

    bmcj

    bmcj

    Well-Known Member HBA Supporter

    Joined:
    Apr 10, 2007
    Messages:
    12,903
    Likes Received:
    4,847
    Location:
    Fresno, California
    Plus, you run into the problem of not necessarily knowing whether the design you are copying has been fully engineered, or is a copy of someone else's structure. It's kind of like making a Xerox copy of a copy... each time you add a level, you lose quality.
     
  8. Mar 8, 2010 #28

    brandon81

    brandon81

    brandon81

    Member

    Joined:
    Sep 2, 2009
    Messages:
    13
    Likes Received:
    0
    Location:
    ohio
    What method would a home builder use I know a lot of people have built their own designs .What method are the pros using is it all on a computer and where is the data coming from .If a person wants to build a airplane what would be the steps to take and what books would they use to gain the knowledge to build their original design out of composite ? Thanks
     
  9. Mar 8, 2010 #29

    autoreply

    autoreply

    autoreply

    Moderator

    Joined:
    Jul 8, 2009
    Messages:
    10,732
    Likes Received:
    2,542
    Location:
    Rotterdam, Netherlands
    Raymer (aircraft design) is a well known (and good) book. Whether that book (and at least half a dozen others) are enough depends on your design and especially the complexity and speed.
     
  10. Mar 8, 2010 #30

    brandon81

    brandon81

    brandon81

    Member

    Joined:
    Sep 2, 2009
    Messages:
    13
    Likes Received:
    0
    Location:
    ohio
    I have that book and many other my question is still in regards to the actually construction of the aircraft like the structures I have some books on composite as stated early but they only give ideas on construction method which is fine but I want to know how many piles to use and where to apply more .
     
  11. Mar 9, 2010 #31

    orion

    orion

    orion

    Well-Known Member

    Joined:
    Mar 3, 2003
    Messages:
    5,800
    Likes Received:
    135
    Location:
    Western Washington
    As far as the analysis goes, homebuilders have several choices. First, assuming they have an engineering background and are willing to learn the proper forms of analysis of composite lamina, they can do the analysis themselves either through classical methods (hand calculations) or if they wish to invest a certain amount of time, they can write their own programs from the classical methods so as to enable them to run multiple simulations in order to achieve a certain amount of optimization. If you read through the posts submitted by Billski (wsimpso1), that's precisely the method he chose and by all accounts, seems to be doing a great job of systematic design and analysis.

    If you don't have the technical background though the road is much harder. It's not so much that it can't be done (it certainly can with a good amount of research), it's just that it takes substantially longer to do since much of the effort is just figuring out the right questions to ask and the right data to search for. To a point the simpler methods of analysis can be used however it then becomes a very good idea to have someone look over your work in order to verify whether it's complete and applicable or whether there are still considerations missing. But if you read my posts, I would certainly urge folks not to depend on the simplified methods alone, even if using other, similar airframes as a guide.

    And of course the third option is to hire someone to do this for you. No, it's not cheap but depending on how much value you place on your time and how much faith you have in your own ability to do all this in a proper fashion, the consultant doing the design work may just end up being cheaper and safer than if you did all the leg work yourself.

    Virtually all folks I know that did their own (who didn't have the background), hired someone to do the specialized design tasks for them. Unfortunately I also know at least two who did it themselves, who are no longer with us because by all accounts they decided they didn't need the help - unfortunately though, they also didn't have the background to do the work right.

    The folks that do this for a living do use computers, utilizing a program capable of finite element analysis. For composite structures the program must be able to model and analyze orthotropic laminates.

    As far as material properties are concerned, that's a bit more difficult. Although there is much data available, it is generally from material producers and unless you're using exactly the materials tested, the data may not be applicable to your project. At that point you either have your own materials tested, or you hire someone who has a sufficient database and experience needed to determine the properties of the material components you wish to use.

    The only real and useful database publicly available is that which was developed as part of NASA's AGATE program. Since that includes prepregs and wet layup materials, it is one of the more complete pieces of information out there. But again, if you use it for your design, it is extremely important to use exactly the materials that were tested. Don't think you can get something different and/or similar but cheaper and get the same values.

    For this you may want to use the search function here and look for other threads that discuss this very topic. We have delved into this topic on quite a few occasions over the years and some of the best texts available are often referenced numerous times.
     
  12. Mar 9, 2010 #32

    durabol

    durabol

    durabol

    Well-Known Member

    Joined:
    Mar 2, 2010
    Messages:
    169
    Likes Received:
    3
    Location:
    Lethbridge, Alberta / Canada
    Here is some very general information for typical composite construction.

    One layer of fibreglass is often used to just protect the surface although some recommend 2 layers. Two to four layers over core structures such as fairings, fuselage and wing skins. Three layers is normal minimum for molded coreless parts. Four layers to join fuselage to bulkhead. For PVC ribs a single layer has been used on each side. For tail surfaces 3 layers on either side of a foam spar has been used.

    In general a part made with 5-6 layers of fibreglass can be made with 2-3 layers of Kevlar.

    I have heard that a Long-EZ takes 90 yards of fibreglass.

    The thickness of a layer of typical cloth and resin is about .01".

    For a 1200lbs airplane the spar thickness should be about 1sq.inch at the wing root(100%), about 50% of this at the 75% span point, 25% at 50%, 10% at 25%.

    Brock
     
  13. Mar 9, 2010 #33

    brandon81

    brandon81

    brandon81

    Member

    Joined:
    Sep 2, 2009
    Messages:
    13
    Likes Received:
    0
    Location:
    ohio
    Thanks Orion that is what I wanted to find out I think I would like to contact you about doing some of this if you are interested .Brock that is some useful information.And Thanks to all so far for their time in answering some of these question
     
  14. Mar 9, 2010 #34

    lr27

    lr27

    lr27

    Well-Known Member

    Joined:
    Nov 3, 2007
    Messages:
    3,216
    Likes Received:
    465
    That is SO oversimplified! How fast does this 1200 lb airplane go? What is the span? How thick is the airfoil and what is the chord and planform? Etc. etc. etc. Are these numbers for a Long EZ?

    Still, what about just getting approximately there, using good design practices that don't promote fatigue, and then doing actual thorough testing on the airframe?

    Perhaps one of the reasons there isn't more trouble with fiberglass is that often it has to be designed for stiffness and that leaves it much stronger than necessary.* Still, that's not going to help if you use microballoons where flox ought to be.

    *I'm not saying there aren't problems, I don't know enough about the record, though I've worked with the stuff in a number of applications, a few of which fly. (Not carrying people.) .
     
  15. Mar 9, 2010 #35

    wsimpso1

    wsimpso1

    wsimpso1

    Super Moderator Staff Member

    Joined:
    Oct 19, 2003
    Messages:
    5,861
    Likes Received:
    3,077
    Location:
    Saline Michigan
    Generic numbers are certainly useful for checking your design, but can be very dangerous to just apply. Yeah, a bunch of 160-200 knot airplanes use 3 UNI on the outside of the plane, and they have good histories. That thickness will keep the average airshow moron from poking a hole with is pencil and prevent starlings strikes from doing the same thing in an airplane that fast, but it might be terrible overkill in a 90 knot airplane. It might be way inadequate for an aerobatic monoplane with a 350 knot dive speed and only three ribs per side. Likewise the spar using those rules might be adequate for one ship with a pretty thick wing. But if someone built a wing with an aspect ratio of 18 and a thickness of 12%, I can tell you that the spars will not be up to the task unless the wing skins are pretty darned beefy too.

    I knew a guy who redesigned a particular well known kit airplane around homebuilding techniques. He did not seem to know that the base airplane had carbon fiber spar caps in the outer panels. I later found out that he had copied them dimensionally, but in glass... Now the airplane did not give up in the main wing spars, but my analysis indicates he had about a 4 g wing with no factor of safety. He also guessed at the tail structures, which would have folded up around a 3 g pull. He never tried for those g's in the flying he did, but he would come tumbling out of the sky in a broken airplane if he had. He did die in that airplane, but that is another story.

    Better to do good conservative design work based upon good aerodynamic calculations and good material data.

    Billski
     
  16. Mar 10, 2010 #36

    durabol

    durabol

    durabol

    Well-Known Member

    Joined:
    Mar 2, 2010
    Messages:
    169
    Likes Received:
    3
    Location:
    Lethbridge, Alberta / Canada
    Sorry for being vague. I should have written that each spar cap should have that area on either side of a shear web on an I-beam spar.

    Brock
     
  17. Oct 22, 2014 #37

    WonderousMountain

    WonderousMountain

    WonderousMountain

    Well-Known Member

    Joined:
    Apr 10, 2010
    Messages:
    1,853
    Likes Received:
    192
    Location:
    Clatsop, Or
    A little bit of a newbie question. Is prepreg available in strips of say 1 inch width.

    I am wanting to use strips of fabric clamped in-between wood. This may avoid chemical exposure that turns me away from tree based construction.

    Wood has been my favorite, but I don't get along well with glue.

    Thanks,

    LuPi
     
  18. Nov 3, 2014 #38

    tunnels

    tunnels

    tunnels

    Banned

    Joined:
    Sep 7, 2014
    Messages:
    559
    Likes Received:
    21
    Location:
    china is my home , place of work but was born in N
    prepregs glass does not have sufficient resin to use clamped between layers of wood !
    wood needs resin penetration into the grain and sufficient thickness to cure properly !! so no over clamping to tight and prime the wood with resin 30 minutes before gluing
     
  19. Nov 15, 2014 #39

    cblink.007

    cblink.007

    cblink.007

    Active Member

    Joined:
    Jul 7, 2014
    Messages:
    27
    Likes Received:
    10
    Location:
    Tennessee, USA
    My friend, I would highly recommend learning first about how to design an aircraft, because from the way you are indicating, you have a very unsure and potentially complex idea set. Even a basic book like "Simplified Aircraft Design for Homebuilders" by Dan Raymer will help get you started on the right path. I did many tough years in undergraduate and graduate studies in aircraft design, and many years in the world of experimental flight test, and I am still learning and hitting occasional roadblocks with my own design (of which I just started to build). Try some sketches, calculate and test the idea, improve it and repeat as necessary until you get an acceptable product. Best of luck to you, and never hesitate to ask for help!
     

Share This Page

arrow_white