A composites approach to backpack building-no sew backpack

The term specific strength is one that measures the overall tensile strength of a material, divided by the density of the material in question.  As an extreme, upper end example, highly oriented UHMWPE fibers, such as sold under the brand names of Dyneema and Spectra, have a ridiculously high specific strength because they are both low density and have a very high tensile/breaking strength.

Carbon fibers have a fairly high specific strength and so doesn’t kevlar and S grade fiberglass fibers–this latter, some people seemed to be surprised by, pretty much rivals carbon fiber’s specific strength. Carbon fiber beats it out in Young’s Modulus aka stiffness strength though, which makes CF better for applications that require high stiffness (and very high tensile strength) at lowest weight. Such as say an airplane wing, a bike frame, or the like.

What many people don’t know though, or might not think about is that regular cheap E grade fiberglass fibers vs nylon is a very interesting comparison.  The specific strength of E-fiberglass fibers is some 18 times higher than nylons.  That means, mathematically and theoretically at least, you could have a 1 oz/yd2 fiberglass cloth that has a similar tensile strength as an 18 oz/yd2 nylon fabric.

I don’t think it actually quite works out to that in the real world, as that specific strength figure is often derived from the maximum tested limit of highest quality possible E-glass fibers. In reality, quality is going to widely vary.

But nonetheless, and no matter how you cut or measure it, E-glass fibers are much stronger tensile wise than nylon, even when factoring they have a higher density (roughly 2.2X that of nylon). Even if it’s only 10X or 5X in more realistic, real world terms, it’s still pretty impressive.

However, there are some good reasons why woven fiberglass cloth is not used to make things like backpacks or the like. The main reason is that fiberglass cloth is most often used in combination with a rigid matrix like epoxy or polyester resins. These help fiberglass to more efficiently use it’s relatively high specific strength and ok modulus strength to good effect. The combination creates a fairly tough and strong material with relatively low weight.

Also, fiberglass while less brittle than one might think, it still has a limit to how much it can be flexed and bended, especially at the same points. Then it also doesn’t handle abrasion super well.

It’s not like you can take some fiberglass cloth and just sew up a backpack out of it.  It would come apart. The woven fibers easily separate.  It needs to be set in some kind of binding/bonding matrix, and preferably adhered to some sort of thicker core. Also, when fibers come out of the weave, they can get in and irritate the skin, as fiberglass batt insulation is known for.

But these issues can be over come, and it can be used to make a backpack. The material is just too cheap, available, and too strong for it’s weight not to try. This is where a systems approach comes in… Tbc in next post.

Designer is interested in making a relatively lightweight backpack that is very comfortable and that can handle very high loads if needed.  Since designer’s shared sewing machine had been unexpectedly absconded (ok, not really, but taken to another location), he had some extra time to consider this project and looking around at materials already on hand, he wondered–could he counter balance the weaknesses of E-glass fibers to take advantage of their strengths?  And so some imaginative conceptual pondering and then some testing commenced.

Designer has 2mm thick EVA foam on hand and which is not particularly useful for too much (accidental purchase). “Hmm, why not try bonding some fiberglass cloth to each side of this rather thin and lightweight foam?”

Ideally, the bonding material would well adhere the cloth to the substrate core, and hold the glass fibers together within a matrix, while still allowing the combination of foam+fiberglass to flex and bend some. First experiment involved using a water-PU based bonding agent and some .75 oz/yd2 E-fiberglass. This worked well to bond the fiberglass to the foam and to itself.  It however, didn’t seem to increase the tensile and sheer strengths of the composite at all.  This is perhaps not too surprising since we know from WP nylon and polyester fabrics that use of PU as the WP part, this can actually slightly weaken the base fabrics (or at least is weaker than silicone treated ones).

The next experiment will be done with thinned silicone that also is mixed with some crystalline nanocellulose concentrated and isolated from a base stock of already highly crystalline nanocellulose content material (approximately 70% CC content). We know that silicone, unlike PU, actually slightly increases the tensile and shear strength of WP nylon and polyester fabrics. And research shows that CNC can increase tensile, sheer, and modulus strengths (at least within epoxy).  If it adheres well to the EVA foam, it may be the ideal bonding matrix.

In the next post, the big picture of the project will be outlined, including also some trees within the forest.

First general concept.  Take a large, long sheet of 2mm thick EVA foam cut to size, fold in over in half in an envelope, folded/crease part being the bottom. Unfold again. To that EVA foam will be bonded two layers, on each side, of .75 oz/yd2 E-fiberglass cloth and one outer, lightweight nylon (40D) layer on the outside part of the pack. That is your basic semi flexible container.

But then it gets more complicated. To be able to carry very heavy loads, you would need high stiffness in some parts of the pack.  Particularly on the part of the pack that is against the back, and that the shoulder straps and hip belt connects to, so as to distribute load well.  You will also need a very strong bond at these areas and between the two folded together “seams” in lieu of sewing.

So before using any fiberglass and silicone or PU, map out that area of shoulder straps and hip belt.  For this, designer will use G-Flex epoxy in combination with some carbon fiber, to make a square or rectangular border of higher stiffness on the outer foam sheet that will be against the back. Then, the two fiberglass cloths (and 1 outer layer of 40D woven nylon) will be placed over that and also epoxy bonded on at that square/rectangular border.

Later on, making sure as little contact as possible happens between the epoxied areas, then bond the rest of the material using silicone or PU, but leaving the very edges where the folded together seam part will go together.  That part (the connecting seam) will also be epoxy bonded for maximum strength, and increasing rigidity some there too. No carbon fiber though.

There is some more detail to it than this, but this is the general gist.  The hip-belt and shoulder straps, will be epoxy bonded to the carbon fiber+epoxy+fiberglass+40D nylon+foam, stiffer areas, and excess “fabric composite” will be allowed to make sure plenty of surface area is bonded.

As mentioned, afterwards, the entire outer pack will also be covered in a combination of 2, .75 oz/yd2 fiberglass layers and 1 outer layer of 40D nylon (inside of the pack, probably doesn’t need the 40D nylon imo).   The 40D nylon is for specific color, and for abrasion protection for the fiberglass underneath and to protect the skin from any possible stray glass fibers. Either way, it all will be held within some sort of rubbery matrix so there is little chance of glass fibers getting to the skin anyways, but better safe than sorry.

The main difficulty will be in keeping the two different bonding materials separate,as I think that epoxy can interfere with the curing of some or most silicone’s.  However, there is one type of silicone that I know of, that uses a similar curing agent as most epoxies, so that might be the best bet. It’s usually the dark colored, automotive gasket stuff I think. (?)

I might just stick to that, in case there is any cross contamination. But if you get the proper viscosity down for each, it should be relatively easy to keep these mostly separate except for some very tiny overlapping.

There will also be some thicker, squishy EVA foam pieces bonded strategically to the “spinal” area of the back in between the shoulder straps and hip belt area. For comfort/cushion and as spacers for cooling air flow.

Hypothetically, the combination of all the above materials, should create a bomber, high load carrying, very durable, extremely, and permanently so, waterproof, and very comfortable backpack at fairly low weight. It will be equivalent to about a 7 to 9 oz/yd2 or so weight single fabric equivalent, but so, so, so much stronger and more durable . You may want to put some UV gloss finish over the exposed epoxy bonded areas, to maximize it’s longevity, as UV is one of epoxy’s biggest weaknesses over time.

“You may be right, I may be crazy…  but it just may be a lunatic you’re looking for.”

Some additional notes and observations:

I did some rip and pull tests on the .75 oz/yd2 bonded to 2mm EVA foam with PU-water based, rubbery bonding matrix. I wasn’t impressed by the strength of the composite, yeah, it was improved over straight foam for sure, but not as much as I expected nor wanted. There is already some important influencing variables that I have observed that may be involved in this less than ideal performance.

1. As already noted, PU doesn’t increase the strengths of woven fiber cloths, but may actually slightly weaken same, or at the very least, doesn’t slightly strengthen like silicone does.
2. I only did one layer of .75 oz/yd2 on each side.  When making epoxy fixed composites, almost always, multiple layers of fiberglass cloth are used and often alternating directions, which is known to dramatically increase the overall strength and durability of the end composites.
3. It appears to be very low quality fiberglass cloth. It’s an inexpensive, Hobbico brand  I noted the weave was rather loose and it had runs in it.
4. I read that water moisture can decrease the tensile strength of fiberglass. The PU bonding agent is water based, which makes for lower VOC (and is much more pleasant to work with than epoxy, thinned silicone, polyester resin, etc), but it’s possible that it wasn’t fully dried when I tested same. I should have lightly oven dried it for awhile first.  Naphtha thinned silicone will not have this issue when used.

I’m tempted to invest in a heavier and better quality fiberglass cloth such as this one, to use as the first layer on the outside (and only 1 layer on the inside of the pack):
http://shop.fiberglasssupply.com/Glass_Fabrics-Style_1080_1_4_oz_X_38in.html

Note, it is 1.4 oz/yd2 and has a warp breaking strength of 120lb/in and a weft breaking strength of 90lb/in (both of which should increase slightly when within a fully cured silicone matrix). That is pretty impressive.  And note, if you buy at least 4 yards, it only costs 5.45 per 38″ wide X 1 yd length. And if you buy at least 16, it only costs 4.64 per yd, which is quite reasonable.  38″ wide is about how wide I want it anyways (since the main part of the pack is just a simple, folded in half envelope, it needs to be wider than normal so that it can expand out some when filling up).

I’m really hoping that the silicone adheres well to the foam/fiberglass composite.  The next test will still be with the .75 oz/yd2, not so high quality cloth, but two layers each side and within a silicone matrix. Will be bonded today, but will let it cure at least 24 or so hrs before trying to rip, and pull, it apart. (Actually, I will do 1 sample with two layers of fiberlgass each side, and then 1 sample of 2 layers of fiberglass each side and 1 layer of 40D nylon on one side).

If this concept ends up working well (in an actual full prototype), it really could open up a lot of possibilities for ultra strong backpack designs, and especially be helpful for people that don’t have sewing machines, but would like to make their own packs. When I eventually make a full prototype (its queued #3 or 4 project), I think it will deserve a full video of testing, including loading it up with extreme weight. Hunters may especially appreciate such designs if the concept works well.  

Oops – somehow posted in the wrong thread…