This appendix briefly describes methods and structures used by other modellers.:
The last series of F1A's I built used a near ribless "D" box together with a molded spar. The spar was laid up using carbon fiber unidirectional pre-preg. I wrapped two layers of 150 gsm pre-preg, about .0065 thick, at +/- 45 degrees over a male plug that I had machined from aluminum. On top of this I applied equal number of layers of unidirectional pre-preg spar caps, you must have an equal layup on both the top and bottom or it will warp, to form the top and bottom spar. I would then put this layup into an aluminum compression mold and cure the pre-preg at 250 Deg. F. After removing the spar from the molds all I had to do was some light sanding to even out the "stepped" layers of carbon fiber. The +/- 45's took care of the torsional strength and the unidirectional top and bottom spars took care of the bending loads. The carbon fiber/Kevlar "D" box described next was mainly there to form a leading edge. Total hands on labor for a molded spar was 1 to 1 1/2 hours.
For the "D" box I made molds for the main panel and each tip using basically the same methods you describe except I use balsa with a basswood leading edge. I also covered the mold with 3/4 oz fiberglass cloth and gave it two or three coats of thinned epoxy to create a very smooth surface. I did this so I could use a spray release agent directly on the mold. My "D" box layup was one layer of 3K carbon fiber cloth on the outside and one layer 1.8 oz/ydsq Kevlar cloth on the inside. The Kevlar was added to support the carbon in case the model glided into something and the "D" box was damaged. Since the Kevlar will almost never fail, the epoxy fails in a Kevlar layup, it holds the carbon fiber pieces together making any required repairs much easier. This layup was .014" thick and weighed .26 g/insq, sorry about the bastard units but that is the way think when it comes to models and weights.
Root ribs were added, 1/4" balsa, at each end of the "D" box so I would have something to glue to at the root and dihedral joints. I didn't lay the "D" box over the spar as I was unwilling to give up the extra spar height. Instead I milled a notch, using the Dremel routing table and a stone wheel, .2" back and .014" deep on the bottom and .02" deep on the top, so I could sand some "round" into the top of the spar, from the molded spar leading edge top and bottom. The "D" box was then epoxied to the spar. The trickiest part of the whole assembly was trimming the "D" box to the correct width so it all came out to the correct width. The inside height of the spar was .281 to accommodate the .281 aluminum tubing I use for the .25 steel wing rod. The tubing was not installed until both wing halves were built. This allowed me align both halves perfectly before gluing in the tubing.
It took me a total of approximately 3 to 4 hours of hands on labor to build a main panel "D" box/spar assembly including all of the layup times. Actually the tip took longer because I had to build up tip spars in the old fashion way. I was very happy with the way this worked and plan on incorporating many of these methods into my F1B's.
Phil Barnes uses the following methods to handle 35 gsm Kevlar when he's vacuum-bagging foam-core wings for Discus RC Hand-Launch Gliders and to make handlable pieces of unidirectional carbon for repairs. These look like suitable departure points for applying light cloths, such as 25 gsm glass or uni-directional carbon to D-box shells with minimal distortion of the constituent fibres or to producing uni-carbon patches for field repairs.
Phil writes:
I have a lot of experience with vacuum bagging of solid foam core wings. I've done a number of layups using the light (aproximately 35gsm) Kevlar that you talk about. There is a technique for handling that material which is common in the RC hand launch community which involves using a "paper transfer" technique. You might find it useful. I have copied a writeup from a group posting and reprinted it below in case you are interested. The writeup mentions the use of 3M77[1]. This is very handy stuff for lots of things.
Roll out the Kevlar on a SMOOTH bench. Brush it out flat with a drafting brush and carefully align all of the weave so it is perfectly laid out flat on the bench. Now get some large sheets of newsprint from the craft / art store. These are sold in pads and are something like 22x28". Lay out a sheet of the paper on a separate bench or the floor and spray it with an incredibly light mist of 3M77[1]. I am talking about such a light mist that you would swear that nothing was on the paper. I hold the can about 3 or 4 feet above the paper and make one very fast pass spraying parallel to the bench and letting the mist settle over the paper. Now pick up the paper and stick it to the Kevlar. Start with a corner and slowly lower it in place. Repeat with other pieces of paper, butting them up to the first pieces until you have enough Kevlar covered with the paper. Trace out the Kevlar patterns that you need with a pencil on the paper. Remember to make lefts and rights as these will not be reversible as a plain piece of fabric would be. Cut the shapes out with Kevlar Scissors. It will cut just like paper, as if the Kevlar were not there. You can handle the cut pieces just like paper, the Kevlar is stuck to the paper and goes along for the ride. To do the layup you roll epoxy on the mylar, drop the paper backed Kevlar pieces in place, paper side up of course. Sometimes if the 3M77 was done just right the paper will unstick from the Kevlar as soon as it gets wet with epoxy. Usually you need to pull it off though. To do this you need to carefully lift a corner with a knife point , get a hold of the paper and yank it off with one quick motion. Be sure that all of the Kevlar is satuarated with epoxy first. This is what will hold it in place on the mylar. The Kevlar will not stay put if you try to slowly pull the paper off but will if you yank it off quickly. I would recommend practicing with little scraps of Kevlar on little pieces of mylar until you become a believer as to how little 3M77 it takes. Too much will make removal of the paper too difficult.
The above method also works if you just cut individual paper patterns and then stick these individual patterns to the Kevlar. That is what you would do if you only needed one or two small pieces rather than a a larger number of pieces.
Here are a couple links to groups I am active in in case you wanted to research more details of our style of vacuum bagging;
"Composite fabric" is my own term for a thicker material made by sticking two different fabrics together to produce a material with better dry handling properties ot that will produce a superior structure than if each fabric was applied separately.
A good trick for handling uni-directional carbon is to stick a piece of 25gsm glass to it with 3M77[1]. I only use this composite for making tail booms or repairs so the extra weight is OK.
Here is how you do this; spread the glass out flat on a smooth bench. Spray a piece of poster paper with a light spray of 3M77 and then stick this to the glass. Cut around the glass with scissors and then spray the glass side of this with a somewhat heavier spray of 3M77. Stick this to the carbon fabric. You can then peel the assembled "composite fabric" off the poster paper backing. This makes the carbon handle just like a piece of paper. You can cut it to any shape without it falling apart or fraying. You can save weight if the composite patch is stuck on with the glass on the outside. The glass layer can then be sanded off after the epoxy has cured.
I use the "composite fabric" method for lots of things apart from making unidirectional carbon easy to handle: for instance for the narrow strips of fabric that go on the wing core leading edge before bagging. In this case the fabrics are usually Kevlar and glass or maybe Kevlar and light, balanced carbon. I also sometimes use a Kevlar and glass "composite fabric" for the aft portions of wing skins..
Here is a link to a post with a better description of the method. I normally use it for bias applications so that is why the write-up talks about parallelogram shaped pieces.
| [1] | I've never seen 3M77 outside the USA. Its the strongest
sprayable contact adhesive that 3M sell, with instant grab and no
chance whatever of repositioning a part once it has been put in
place. You'd have to check with your local graphic arts supplier
for the strongest equivalent available elsewhere. "Spraymount",
which is readily available, is a lot weaker and allows easy
repositioning. It may work in this case, but a heavier layer will
be needed. Experimentation is needed. - Martin Gregorie |
Bill learnt his techniques from a video made by Phil Barnes. The video is being sold by Bill Haymaker, who shot and edited it. If you want a copy, you can find details at http://www.paonline.com/hayman/video.htm
The starting point is a hot wire cut foam core ( I opted to have the foam cores cut by a professional) along with its attendant "beds" (the outside parts - top and bottom that are like female molds.)
Patterns for the carbon cloth are cut from 12 mil Mylar sheet. They are waxed.
If one wants to include a top coat of paint then the top Mylar is spray painted. Use epoxy for a fuel-proof finish is needed. Otherwise any paint that is compatible with epoxy can be used.
The cloth is sprayed with 3-M 77 to hold the fibers in place. The 3.5-Oz carbon cloth is cut to approximate size and placed on the Mylar patterns. RC wings use a veil of 1 Oz glass, which I eliminate because of weight considerations.
The pre-warmed epoxy is applied to the cloth which is then rolled out and de-bulked using rollers... like a carpet seamer. This must be done with care as the roller can pick up a strand of carbon and strip it out of the cloth. From Phil's video one can see the reflective appearance of the cloth changed as it is compressed with the yarn pressed down to fill up the normal voids between the weave. Of course excess epoxy is sopped up.
The cloth is then trimmed around the edges to fit the Mylar outline.
After contact-cementing the ply ribs to the ends of the foam core I then add a strip of uni-web carbon material sprayed with contact cement then fitted to the nose. Is this unidirectional carbon on 25gsm glass as described by Phil? This will underlay any gaps left by the main cloth material not correctly conforming to the nose.
The foam core is placed between the Mylar sheets like ham in a sandwich. The leading and trailing edges are taped together.
Phil uses a two-bag system. The inner "bag" is made from 4 mil plastic drop cloth material having the bleeder cloth on its outside... it is merely folded around the Mylar sandwich. The whole works is then placed inside the usual Nylon bag for decompression with a vacuum pump ( I have a really great one I got from Bob Mattes).
The two beds are placed on the OUTSIDE of the package to conform the sandwich in matching the external shape of the airfoil. Weights are used to hold the top down tightly.
I use an electric blanket for the cure... set for about 90 degrees... too much heat during the early part of the cure is to be avoided.
I'm using the low viscosity Pro-Set Epoxy with the middle cure hardener... that provides the lowest viscosity of the three choices.
When de-bagged, after curing, the trailing edge and leading edge are then smoothed out. Of course the leading edge is the tricky part and must be handled with care. Because of the bulk of the cloth I'm not sure how accurate the nose will turn out on my very thin section. This is the reason for adding the strip of uni-web carbon and what makes this technique possible. (Out-of -date systems don't attempt to form the nose in the molding operation so the leading edge must be shaped out of wood then added after the fact.).
Putting in spars and such is a different matter. In my case since my tubular spar is full depth I'm opting to cut the hole for it after the panels are formed. It was suggested that I use urethane glue rather than epoxy, as it will expand to fill any voids. The expansion does worry me in that it might lever the carbon off the foam core; so perhaps it should be cured with the panel under compression.