Thank you for the summary of the rabbit hole ahead!

All of my relevant amateur experience is from aviation, and wet layup, so
perhaps I am orders of magnitude off from expectations.

Thank you for taking the time to educate me. Please forgive my ignorance.

I'll continue below inline.
 

On April 24, 2018 4:52:15 PM UTC, "Björn bjornmail@gmail.com [harryproa]" <harryproa@yahoogroups.com.au> wrote:
>I have also been tempted by this. Many have. But very few actually
>build a
>boat with extruded polystyrene as core. Maybe it is because if you
>consider
>the total cost of the boat, the cost of the core material in the
>sandwich
>is a very minor part. And the risk to gain ratio is not good, because
>if
>the core fails, the boat is more or less scrap.

Should not the engineering help us know before we begin?

>I think that considering how light Harryproas are, they will be cheap
>enough to build.

Yes, probably nibbling at the edges.

>I have been doing a lot of calculations to compare the price per square
>meter with different core and resin choices actually. Considering how
>much
>of the cost is epoxy, I think that if you want to save money, it might
>be
>cheaper to use a less expensive resin, (and stick to a standard core
>since.
>If the core has half the strength as you say, it also needs to be twice
>as
>thick, i think.

I don't think it works that way. If the skins are expected to take a load that will subject the core to some load (sheer?), the core needs to take that load, or fail when the part is stressed to that point. I don't see how one can sub in a lesser material, and expect it to work. Also i don't think stiffness can sub for sheer.

Here is a source
http://www.hexcel.com/user_area/content_media/raw/Honeycomb_Sandwich_Design_Technology.pdf

>This means that even more resin is consumed in all the
>holes which has to be drilled through the core to aid the infusion
>process.

I do not understand.
If one makes some small number of layers of glass, in a flat infusion, does the resin not flow well across them and well to each top and bottom surface? How does having the foam against the composite, instead of the table/bag change the flow?

This is my guess; the skin core interface is the weak point! The high performance cores need grooves to get more core surface into the adhesion interface and strengthen the skin core bond to reach the loads that require high performance cores. Thus the come pre grooved.

>The cost of that extra resin (epoxy in the case of polystyrene) is not
>insignificant.

That's a good point. But i cannot see how to use less epoxy, assuming Intelligent Infusion.

>But even though I have done all the calculations, I keep coming back to
>the
>temptation of using it, maybe for the first build. You can get the 20mm
>thickness with standard density for 2 USD per square meter. Which is to
>be
>compared with 20-40 dollars or so for the "real stuff".
>
>One drawback seems to be that the surface usually as some kind of oily
>surface which the epoxy will not stick well to. So maybe it is better
>to
>instead of buying 20mm, buy the thicker (100mm) blocks and slice them
>to
>the required thickness with a homemade hot wire cutting machine. But
>then
>you have to make that machine. And then you have to drill all those
>holes.
>And maybe also cut channels for the resin to flow during infusion. And
>so
>on. So it is a lot of work (when you can buy the proper core which has
>the
>channels and holes already made), with little gain, and with an unknown
>risk of making a really poor sandwich construction. That is my
>conclusion.

It seems that only a core material is needed depending on the strength of the skins? Surely where a thin skin is used, one does not need the high performance core?

Do the expected loads over most of the HP require such large skins that the core needs this high sheer strength?

I have read a few (far to many) of Rob's posts on a few (far to many) forums regarding Intelligent infusion. It seems he often says something like "600gsm per side". I am not suggesting that is the expected glass weight everywhere, or even anywhere, but it seems that low thickness of glass does not require a high performance core?

But, I am often wrong.

I imagine foam sheets could be washed to remove the oily coating, but perhaps hotwiring, which is very cheap, ( a few dollars) is just a good idea to expose a virgin surface to the glue. Like a foam version of peelply?

Assuming the engineering lets us do it.

Can a smaller (10m) harryproa get away with lower performing cores in more
places?

>
>/Björn

Thank you for your time Björn