balsa core, endgrain or not?

[sigurd]

yeah it was, but it went from 40usd/m2 to 12usd (woven) or 3.75 ukp (mat) with some searching...

So with that in mind I am interested in assessing the suitability of the materials. I am quite hung up on weight but also robustness. If I can prevent the spruce core from ever soaking, I think the veneer can easily be patched when damaged.

Discountcomposite has a lot of different stuff, spectra fabric for instance... Wonder how it compares to kevlar for this application. It is stronger in tension per weight than kev, but stretches less I think.

From this to that: Is it true that the inside of the cored skin will be in tension only, so that any compression strength there will be useless?

In another thread some people recommended corecell for a whitewater kayak. I have checked the price, it is not completely insane, almost a possibility for this project. A550 would make the boat quite a bit lighter, if it is not very resin hungry. Granted, I'd then have to add some transverse fibers, no big deal I think. But, will the panel be as tough as with the spruce core? It seems unlikely to me, since its density is 1/4 of spruce.

[robherc]

The strength of the end-result panel for sandwich core depends more on the stiffness, thickness, and shear-strength of your core material. With that in mind, the spruce will have higher shear-strength, but the corecell is pretty comparable everywhere else, I think. If your main goal is lightweight, then go corecell...if it's cheap, or extreme durability, then spruce may win.
As far as compression or tension; the CORE material needs high compression strength, but the important part for BOTH skins is the TENSION strength.fiberglass/kevlar/carbon-fiber do nothing to increase the compression strength of epoxy/polyester, but they do a LOT to increase the tensile strength. Since tensile strength is all that matters most in the skins, they're great.
If you're happy with the pricing, I think I'd go with the Kevlar woven cloth, because it allows you to have a higher ratio of fiber to resin. That way you're adding less weight in epoxy/polyester to your finished boat (ESPECIALLY weight per strength).

[sigurd]

It is the first time I have heard kevlar gives a better fiber to resin ratio. It is funny that you mentioned it now and I just five seconds ago wrote a question about that in the ganja reinforcement thread.

I am going to have to ponder what you said about compression. When you look at an I beam in bending, one of the flanges (skins) are compressed, the other tensioned.

EDIT: If resin ratio is tied to fiber density, and skins only need tension, spectra should be spectacular skin material (Don't play with puns!)?

re. aramid, it seems some people are concerned it will soak water. Any views on that?

I found a great price for 9mm A550, could be split with hotwire to 4.5mm?

[robherc]

I didn't mean Kevlar gives better fiber to resin ratio...I meant that WOVEN fabric gives better fiber to resin than MAT...because of the way the strands lie...the more uniform & flat the strands lie, the less space there is to be taken up by resin (thus unidirectional "fabric" can be the best for fiber to resin ratio).

In an "I" beam, the flange under compression always fails first...BUT that eventuality is slowed by the height of the center, connecting portion. This happens because steel has a much higher tensile strength than compression strength. If you were to replace the central portion of the beam with a material that had a higher COMPRESSION strength, then you could significantly increase the strength of the "I" beam, because now you'd be taking more advantage of the TENSILE strength of the flanges (though only one at a time, mind you).


....I hope I made sense there...the ganja might be clouding my judgment.......

[sigurd]

Here is a drawing of one of the things that is befuddling me.
It tries to illustrate how the outside skin can be tensioned or compressed, depending on a lot of factors.

Shear strength for sitka spruce parallell to grain is 2.6 MPa, A550 is 1.1.
However, I haven't found the shear values in the *transverse* direction for spruce, which I imagine are the important values as a shear web for the 0.6mm ply skin.

Along the grain the bending stiffness and strength of the spruce would be the baseline for a calculation of the transverse fibers that would have to be added with foam.

It appears that spectra unsurprisingly has some issues with bonding as well as the previously mentioned lack of compression strength (apparently lower than resin alone!)

[robherc]

If you ignore the red "compressed skin" line in those drawings, all of the rest will make sense to you.
If you think about it realistically, if your core is stiff & has good compressions strength, then all of the energy that is applied that WOULD compress the near-skin, the core transports the energy directly to the outer skin as TENSION force. This way, the tensile strength of each skin can be used to make up for the poor compression strength of the other. Basically, in plain english, you're using the core as a lever to allow one skin to see a strain as a "tension" load so the other doesn't need compression strength.

[robherc]

In the second picture, the core was too soft, and/or the impacted skin wasn't bonded well enough/didn't have enough tension strength, or the impact was simply more energy in a smaller area than the panel was designed for (ex. if you shoot a .50 Barrett at a rowboat, it'll penetrate because there was no reason to design for that load). (all three explanations work, just depends on how you want to look at it)

[sigurd]

robherc, I think you are wrong, that the outside skin experience compression. Look here if in doubt.

http://www.oneoceankayaks.com/Sandcore.htm

[robherc]

I wasn't trying to tell you that the skin never experiences compression, what I was trying to illustrate to you is that a core with good compression and shear strength can work with the tensile strength of the opposite skin to "make up for" the lack of compression strength on the compressed skin. If the goal were to make skins with the highest possible COMPRESSION strength, we would NEVER use reinforcing fibers in the skins, as they're almost exclusively useful for their TENSILE strength, at the expense of compression strength.
You could build a laminate with high-compression strength and low tensile strength skins if you want...just use wood instead of fiber-reinforced laminates. Wood has more compression strength than it has tensile strength, so this would work; I think the reason it's not used is that the tensile strength of fiber-reinforced laminates is FAR SUPERIOR to the compression strength of the wood (at least on a strength to weight measurement).
By using cored construction, we are able to take advantage of whichever strength (tensile or compression) of our skins is the best, and have one skin's strength "cover for" the other skin's weakness when stress is applied to the panel.

...If you wanted to build an "ideal panel" for laboratory conditions where all stresses would come from the same direction (i.e. inside skin, or outside skin), you could achieve some pretty amazing results by using a wood veneer for the compression-side skin, and a fiber-reinforced laminate for the tension-side skin. The problem with this panel would be that it would have comparatively almost NO strength it the stresses were reversed as the fiber laminate has VERY little compression strength, and the wood has little tension strength...so the panel would fail VERY early that way.

[sailor2]

Originally Posted by robherc

Wood has more compression strength than it has tensile strength,

The problem with this panel would be that it would have comparatively almost NO strength it the stresses were reversed as the fiber laminate has VERY little compression strength, and the wood has little tension strength...so the panel would fail VERY early that way.

In dry laboratory conditions most spruce samples have ultimate tension strength more than twise of their compression strength along grain. But the strength for extended time periods is substantially lower than instantionous strength for all wood both in compression & tension. Any higher moisture content decreases all strength properties of wood, and stiffness more so making it easier to bend without breaking it.

For the laminate the same applyes, they are stronger in tension but the strength in compression is still 10...20 times stronger than resin alone. If it were not so, composite masts would not contain max fiber to resin ratios for max compression strength. The important part when long & thin sample is in compression is to make sure the laminate is not going to fail by buckling mode long before max compression strength is accheaved. This requires correct sequence of orientations of fibres as well as excellent bond of fibres to resin around it.

Shear failure in wood results almost always grain being slided with respect to other grain. Never grain breaking in shear. As a result wood like spruce has 2 different shear values depending orientation of shearload. When the wood is still growing in the forest, the tangential direction is 60%...100% stronger than radial direction regarding shear load. This means laminated veneer lumber has same difference in shear properties. Plase the grain running in x-direction and apply load vertically along y-direction and the beam is strongest when glue joining the plys has thinnest dimension (<0.05mm) in z-orientation.
In tension the tangential direction can be upto 10 times stronger than radial. along the grain is of course the strongest.
In case of plywood there is always some plys where this weakest radial direction is aligned with one of the edges. One of those plys will always break first when whole plywood is under enough shear loading to cause shear breakage. Tension or comression breakage can of course be the limiting one instead of shear.

[sigurd]

If I read you correctly, an axe shaft should have the stripes from the season rings going parallell to the axe head (true, from experience), but if you laminate the shaft, the stripes from the glue line should be 90' to the axe head. Is this what you ment?