GRP/FRP shell

[glassice]

If you use less than 3 layers it will be in the trash in a few years UV gets through any paint and water get in and will pop I done this for a few years

[kennyrayandersen]

The cowling isn’t the blades, and these days the blades are designed for around 8000 hours without having to do anything to them, so maybe there has been some progress ï

[glassice]

[kennyrayandersen]

We are obviously building something very different. The target weight for my tear including suspension is in the 250 Lb range and will be special built to be towed behind a car that only has 21 HP and weighs 1100 Lb. For 400 pounds there isn't much reason to use composites IMHO since I could build a robust tear out of conventional materials with that weight. Frankly, whether it lasts 50 or a 100 years is irrelevant since I've only got another 48 years till I hit 100 and after that I could give a hoot. A well built and sealed tear will last a lifetime.

The real potential of composites is that is if engineered and executed well there is a potential for weight-saving. Sure, fiberglass doesn't corrode, but even a well executed conventional tear will hold up fine and is cheaper, easier and faster to build. If I wasn't so hot on saving weight I wouldn't even be considering it.

Before I build I will do a series of tests (strength, stability, peel and impact, and bearing). This building-block approach is standard for all material development programs and is the key to making a robust structure. If I need more plies than I'm thinking, that should show up during testing. If something breaks, I'll fix it.

[glassice]

no foam you will need some honycone strips maybe ever 10 " 4.5 to 7 in wide and the walls can not be flat

[kennyrayandersen]

no foam you will need some honycone strips maybe ever 10 " 4.5 to 7 in wide and the walls can not be flat

I tried to stay with you until now, but I have no idea what you are talking about. I'm using foam everywhere instead of honeycomb. Honeycomb is nice but it is very expensive. I can get 2 Lb/cu ft. core at Home Depot for cheap -- around $10 bucks for a 4X8x1" thick sheet. Some surface prep will be required, but it’s not that different strength-wise than the honeycomb at a fraction of the cost (OK, it’s half the strength, but a 1/10 the cost).

The walls will be a sandwich of glass, foam, and glass 1 inch think (standard foam thickness) whereas the floor will be the same configuration but with 2 inch thick foam. There will be a 1X2 frame around the perimeter of the floor as well as a 2X2 member that will coincide with the rear axle location imbedded into the sandwich; so, it’s not merely foam, but the bulk of it is a foam sandwich. This is a real common building practice in aerospace with whole wings built this way in some kit planes. It’s a little different (but not even new) in the case of a teardrop, so I can’t even claim to be the first to do that! I’m just hoping to integrate several of the ideas that I’ve seen people do here and come up with a hybrid of sorts. There is no reason that the side walls can’t be flat – that’s actually the easy part; the curved top is more difficult, but that’s doable as well, so no worries there.

I will test to get the strength allowables (core shear, facesheet wrinkling, crimping, strength etc. – incidentally the facesheet dimpling is only a failure mode with honeycomb) and then I will give myself some big factors of safety. Since it’s an over-the-road application I’ll use a g-loading of 3, but with a total weight of 250 Lb, and 60 of that suspension (single torsion axles), wheels and tires, and hubs, the actual loading of the frame and body shell is not particularly high.

If it doesn’t work, it doesn’t work, but I‘m pretty confident that if I go step by step and verify by test what I am doing then it will be OK.

peace out

[glassice]

[kennyrayandersen]

The good news is that you can predict before you build where the extra plies need to be located. I do finite element analysis and composite structure is one of my specialty areas. I’ve been doing structural analysis for 25 years for companies like Boeing, Bell, Vought, Lockheed Martin as well as working with Eurocopter, Airbus some Korean firms. I’m not the guy in the factory who assembles it, and I’m absolutely positive I could learn something from nearly anybody in the factory, but determining the allowables (material development) and analyzing the structure to determine the amount of material and number of plies [as well as ply orientation] is what I do.

Following general design guidelines and practices will eliminate the bulk of bad details. Also, it should be noted that most structure generally always fails at local load introduction points or local stress concentration areas caused by some geometric feature such as a hole or cutout and rarely in an open field area. That’s why one doesn’t need too much material (strength) for the bulk of the part. In the cowling I mentioned earlier for instance, most of it had just one ply one the inside of the cowling with 2 plies on the outside; but, local areas were reinforced with both high-density core and up to 6 or 8 plies. Additionally where there were loads normal to the surface (several hundred pounds worth) back-up plates were used to prevent fastener pull-through.

So, I’m not disagreeing with you exactly, just suggesting that I’m confident I know where to put the extra plies once I have done a bit of testing to correlate the model results with the allowables.

[Vindi_andy]

OK following the meeting with my fibreglass lady it looks like the foamcore fibreglass Tear will be goiong ahead in the UK Still dependant on fitting a towbar but I am confident this can be done.

She has some Insulated wagons to build at work soon and these will be exactly this sort of construction. She will be doing some stregnth testing of the foamcore construction for me. Nothing complicated like Kennyray would be doing, she is goin to take a large offcut make up the sandwich clamp it to the bench so part of it hang over and sit on it. OK i know its not exactly High tech but it will give us an idea.

had a very long chat with her and it looks promising. OK it wont be the cheapest but it will be light we will be using woven cloth rather than CSM so the prep work for painting will be minimised and strength maximised. As kennyray said it wont be the first made like this but it will be Unique and it will be MY WORK

Kennyray looks like you could be becoming my new bestfriend in america as I will probably want to pick your brains regarding the computer modelling stuff you were on about that I recognised the terms but dont have a clue how to do

[kennyrayandersen]

OK following the meeting with my fibreglass lady it looks like the foamcore fibreglass Tear will be goiong ahead in the UK Still dependant on fitting a towbar but I am confident this can be done.

She has some Insulated wagons to build at work soon and these will be exactly this sort of construction. She will be doing some stregnth testing of the foamcore construction for me. Nothing complicated like Kennyray would be doing, she is goin to take a large offcut make up the sandwich clamp it to the bench so part of it hang over and sit on it. OK i know its not exactly High tech but it will give us an idea.

had a very long chat with her and it looks promising. OK it wont be the cheapest but it will be light we will be using woven cloth rather than CSM so the prep work for painting will be minimised and strength maximised. As kennyray said it wont be the first made like this but it will be Unique and it will be MY WORK

Kennyray looks like you could be becoming my new bestfriend in america as I will probably want to pick your brains regarding the computer modelling stuff you were on about that I recognised the terms but dont have a clue how to do

Now if you could just calibrate her arse...

Hey just use a fish scale or a bathroom scale and a jack -- whatever you got. Generally you can do a 3 point or 4-point bending test -- it's not complicated and I can send you some information (I just have to dig it up). Never waste an opportunity to get data! It may not be lab quality, but it certainly could get you in the ballpark as they say.

We can probably work something out for the modeling, but generally what I discovered is the the load on a teardrop aren't very high, so for the most part you don't need too much. Of course, you still have to worry just a bit about the local attachments where load is introduced into the teardrop body.

[Vindi_andy]

Kenny my friend opoved your comment about clabirating her arse.

If you could get me the stuff you mentioned that would be great.

There will be pics when we start probably even a whole build diary along the lines of the one for the towcar, which is called "Tango" in case you didnt know.

Car build diary here for thoase that may be interested.

http://sites.google.com/site/lawsvindicatorfamily/

[kennyrayandersen]

I'm in Korea, so home for the day -- the info is at work, so I'll send in the morning (assuming I can remember -- there is a reason I'm not a doctor!).

Now, if I might talk about sandwiches and failure modes and some info you might need.

The home depot core is 2 lb/cu ft. This is on the light side for structure -- the lightest I've used is 1.8 Lb core, but for a tear, and even some home-built airplanes it's enough.

The main thing is that you don't want to load the panel up with any concentrated loads, though distributed loads (like a body) are OK. Using light-weight core and thin facesheets (one or two plies) mostly narrows down the failure modes so let's talk about those. Mostly the failure is going to be facesheet stability due to compression. You can get compressive stress by applying a direct compression load (in-plane), or by applying bending. In the bending case one of the skins goes into compression and the other in tension. If I have a beam and I walk out on it, the lower side goes into tension, and the upper side will see compression. you probably don't have any situation on the tear where tension will be the primary mode of failure.

So, in you 'test' you can make some one inch wide strips that are 12 inches long (if I'm remembering right). You lay it flat so that one of the skins is on the bottom and then support it at either end. the load is then applied to the middle. You measure the force it takes to break the specimen. It will break on the upper surface. You can make some specimens with 1 ply and some with 2 plies, I doubt for the bulk of the build that you will need more plies that this except for local areas of high load introduction.

There is a similar test for the core where the specimen is shorter and you drive the critical failure to be in the core (to verify the core shear properties).

The challenge with sandwich panels is how to close them out. What I plan on doing is to frame-out the core with 1X2 inch wood -- that way it gives something to glue and screw to since the foam doesn't have anything but shear capability (nearly).

Practically you are going to also want to do the knee test. I think you will likely need two plies for this. Make a 6 inch square of sandwich and then simply kneel on it with one knee. Does it crush? It probably won't, but actually this test is a little harsh anyway since your floor will be covered with a mattress which will help spread the load out and make it less concentrated.

If you are making a galley, you will probably want to put 3 or 4 layers on the bottom if you plan on dragging a cooler in and out that would be the same for any high wear or traffic area.

What plan on doing is burying 1X1 or 2X1 fir into the foam (joined by epoxy) wherever something need to be joined (wall or shelf etc.) then skinned over with the glass. The joints (side to center or shelf to side) will then be glued and screwed or I might opt to put biscuits in as well (as they can react normal loads as a shear through the adhesive.

There is one other way to make attachments and that is to use an insert (plug). They make commercial ones that are potted (epoxy with some filler) into the core. You could make a poor-man's insert buy burying a plug of plywood. The plug can then have a bolt or screw attached at that location.

you can check resistance to impact by dropping a steel ball from 20 ft up (there is actually a calibrated test for this as well). It may be that over the bottom half of that front you add another ply to keep the rocks from poking holes in the sandwich.

That's most of the main points -- I believe your friend will confirm most of what I'm saying.

[Vindi_andy]

Kenny you are the man.

I apprecate you giving me this info. My friend also suggested wood stringers between the foam. This is going to be primarily for going to kitcar shows so is going to be one night etc so getting a little mad here but is it possible to make the internal shelves in glass over foam same as the walls. There wont be anything heavy going on them maybe the kids portable DVD player.

I was looking at the more dense PU foams as for the roof line there will need to be some shaping and one of the websites suggest using the denser foam for this.

Regarding the impact resistance yes an extra layer or two over the front or some satin finsh stainless ( i know this goes against the lightweight principle but is matches in with the car then.)

[glassice]

but is it possible to make the internal shelves in glass over foam .That how we make ours make it a real stiff shell best way to tie the side wall together

[kennyrayandersen]

the problem is the denser foam is expensive (unless you have a good soure for it) whereas the 2 Lb stuff can be got at the local builders supply (i.e. cheap).

For the roof I think there are two options that would be reasonably easy to do. One is to lay you inner ply and foam together, but leave off the outer facesheet. Score the foam side to side about 1 inch apart 75% through leaving about a 1/4 inch which will let it bend very easy. Using the tear, or a mold piece in the tear as a guide, just lay it in there and it would follow the contour right around. Then you could either fill the gaps with filer made mostly of microballs (can order on-line), or the second option is you could glue down directly a 1/8 inch think plywood to the foam (unfortunately you probably still have to put one more ply on top of that followed by paint to seal and UV protect). Of course you would also fillet and tape the inner and outer corners which would complete the 'structural' connection.

Personally, I think I will try filling the gaps and wet laying the outer ply since it's a little lighter. I still have see what it's like, but 1 ply should actually do it for a very small trailer. for a larger one 2 plies is probably better for the outside. The inside probably only needs 2 plies on the floor, assuming that passes the kneeling test.