Epoxy over XPS method

[ondarvr]

RhinoKore made foam filled honeycomb, it was a good product,
With no market.

 

[Eric ruttan]

I found this. It has an example of how I was imaging. The foam filled example seems closest to the idea.
It would be neat if rxcomposite could step through this example with his foam cored example.

Neat that the throw away former foam may actually help performance.

 

[rxcomposite]

I found this. It has an example of how I was imaging. The foam filled example seems closest to the idea.
It would be neat if rxcomposite could step through this example with his foam cored example.

Neat that the throw away former foam may actually help performance.

You are imagining it correctly. It has been done a long time ago. There is a research center for composite in California that designed this corrugated beam for bridges. No foam, just corrugations and a flat plate for floor. If I remember it correctly, the final product is a woven/stitched preform and the resin was injected.

The corrugated panel has been a practice in ships for bulkheads. They just added a flat plate for the cars to roll on. My initial calculations showed that hollow tophat stiffeners arranged closely together showed that it is too strong. The problem is deflection, not vertical load/shear. With spans of 3 to 4 meter for support, typical of bridges, span deflects too much. It will take some research or a lot of imagination to come up with an arrangement.

I will post my core sizing spreadsheet later. Just checking if there are errors cropping up.

 

[Dejay]

Cool! I guess unsurprisingly there is quite a bit of research on these topics. For trusses you can also find some interesting papers. The last one concludes that such sandwich trusses are comparable to honeycombs. It's not surprising that these can be very stiff and lightweight, but more importantly for boat building they can also have high specific compressive strength.

I haven't found anything yet about using cheap foam as a former for vacuum infusion though.

---
Ultralight Cellular Materials

Sun, Y., & Gao, L. (2013). Mechanical behavior of all-composite pyramidal truss cores sandwich panels. Mechanics of Materials, 65, 56–65. doi:10.1016/j.mechmat.2013.06.003
https://sci-hub.tw/10.1016/j.mechmat.2013.06.003

Finnegan, K., Kooistra, G., Wadley, H. N. G., & Deshpande, V. S. (2007). The compressive response of carbon fiber composite pyramidal truss sandwich cores. International Journal of Materials Research, 98(12), 1264–1272. doi:10.3139/146.101594
https://www.researchgate.net/publication/250196695_The_compressive_response_of_carbon_fiber_composite_pyramidal_truss_sandwich_cores

Kim, H., Cho, B. H., Hur, H.-K., & Kang, K.-J. (2015). A composite sandwich panel integrally woven with truss core. Materials & Design (1980-2015), 65, 231–242. doi:10.1016/j.matdes.2014.08.064
https://sci-hub.tw/10.1016/j.matdes.2014.08.064

 

[rxcomposite]

Cool! I guess unsurprisingly there is quite a bit of research on these topics. For trusses you can also find some interesting papers. The last one concludes that such sandwich trusses are comparable to honeycombs. It's not surprising that these can be very stiff and lightweight, but more importantly for boat building they can also have high specific compressive strength.

Waffle or pyramidal type core have been around at the same time as honeycomb. They just did not become popular. Waffle type may offer some increase in shear properties but loses some in vertical load bearing like the honeycomb. If you analyze the load bearing micro structure, the elements in pyramids are slanted while the honeycomb is vertical. In order to be "efficient" it must have vertical and diagonal members like a bridge structure. Note that the honeycomb is in the lower group of cores when it comes to shear properties. Cores are shear driven and in computing cored laminate properties, only the shear is computed in the calculations for the web part. The glue line/bondline or the web to flange region is sometimes considered but only if the design is so short that it becomes critical. The shear is greatest at the neutral axis and vanishes to almost next to nothing at the glue line.

<bbcode tag should be fixed now>

 

[rxcomposite]

something wrong with the web. I quoted Dejay and answered but turned out as above. No amount of editing corrected it.

 

[Dejay]

something wrong with the web. I quoted Dejay and answered but turned out as above. No amount of editing corrected it.

Probably the editor got confused by the formatted text I copy pasted, splitting some formatting tags. My bad

Waffle or pyramidal type core have been around at the same time as honeycomb. They just did not become popular. Waffle type may offer some increase in shear properties but loses some in vertical load bearing like the honeycomb. If you analyze the load bearing micro structure, the elements in pyramids are slanted while the honeycomb is vertical. In order to be "efficient" it must have vertical and diagonal members like a bridge structure. Note that the honeycomb is in the lower group of cores when it comes to shear properties. Cores are shear driven and in computing cored laminate properties, only the shear is computed in the calculations for the web part. The glue line/bondline or the web to flange region is sometimes considered but only if the design is so short that it becomes critical. The shear is greatest at the neutral axis and vanishes to almost next to nothing at the glue line.

So if I understand you correctly the cores are shear driven for boat applications, then waffle or pyramidial type core would actually be better?

But what's interesting about this isn't the top end of possibilities but being theoretically able to manufacture these yourself cheaply and locally with a CNC and use vacuum infusion. It "just" has to be strong enough, lightweight, cheap and easy to use.

Varying the orientation or density of the stitched struts in the truss core could also be an advantage then. You could even manufacture each stitched core panel with optimized patterns and density of struts for example near junctions or point loads. A FEA software could output a list of optimized sheet sizes and stitch patterns for your CNC to make.

And apparently the straightness of the woven struts is important for compression strength and bonding to the surface layer is as well.

But I've just started to read and trying to comprehend the last paper and should probably just finish reading first, but this is kind of exiting

 

[rxcomposite]

So if I understand you correctly the cores are shear driven for boat applications, then waffle or pyramidial type core would actually be better?

For uniformly loaded such as boat bottoms and sides, yes. But in decks or floors, it is generally point load, live loads, or concentrated loads. You need both vertical and horizontal components.

There is always a vertical and horizontal shear and it shifts from vertical to almost 45 degree at certain load models. Sometimes they cross, +45 and -45 degree. This is very evident in concrete girders. You can easily see how the crack propagates. Design is always a compromise, that is one that gives you the least trouble.

 

[rxcomposite]

Attached is the spreadsheet for core sizing. This Lloyd's derived formula and can be validated with an LR software. I have posted this in the forum a long time ago. The program contains provisions for adjustment in safety factor and derating factor in accordance with the rules used. This is not meant for full strength analysis, just core sizing.

The first part of the program tells you the required shear strength and thickness of the core. It reads from the database. It also includes the required thickness of the laminate face. As a rule, CSM is not used because of its low modulus. 11,000 is the minimum.

The second part lets you choose a different thickness of core, its corresponding shear strength and adjustment in face thickness. As you go for thicker core, it requires less shear and the stress on the laminate face is reduced.

From the example, I went to extreme and used the lowest shear available (XPS foam). It showed that for a given load (typical load for a small boat or dinghy) and panel size, I would need to use 150 mm thick foam. The skin thickness is also reduced to impractical values. A single layer of 7781 cloth is about 10 mils thick. The calculation showed I need only 1 mil thick. In practical use, aircraft will tolerate a single skin thickness provided you don't step on it. In boat use, ISO rules that 2 mm bottom outside skin is the barest minimum and LR requires 4 mm minimum.

 

[Dejay]

Thank you very much RXComposite! This is very helpful.

I've looked a bit around and finally found some values for sheer strength of Styrofoam XPS foam. Ok it was right here on the forum It seems if you get to densities of 40-45kg/m³ you get shear strength of 0.4 to 0.5 MPa. So comparable to lower density PET or PMI foam.

XPS Styrofoam RTM-X
Density 40 kg/m³, Sheer strength 0.4 MPa

XPS Styrofoam HD 300F-X
Density 45 kg/m³, Sheer strength 0.5 MPa

I've attached my spreadsheet with some collected material properties for foams, no guarantees that I typed off the values correctly though.

 

[rxcomposite]

Thank you Dejay. I will update my spreadsheet. It has been quite some time since I used it.

 

[Prettypicturegirl]

Hi,

interesting and long going thread. Read most of it, but not all.
I just thought of two possible ways to improve the shearstrength. Kind of a low cost DIY version of the stitching mentioned earlier. Not counting worktime cost.
Essentially we are tying the skins together with fibers to help.
Of course sand the panel with about 36 grit before starting this. Than dedust.

1) Threading a bunch of roving strands through the panels.
Punch holes through the (say 20mm) XPS. Then thread a number of roving strands through each hole. Alternatively punch the holes with an oversized needle and pull the roving through. About 50-100mm intervalls between holes.
Then, either leave the roving overhanging by 100mm at each side. Alternatively thread back in the next hole running a continous strand through the panel this way.
Maybe afix the ends with a dab of instand glue (cyanoacrylate) not sure if that eats up foam though. Alternatively a tind dab of hot glue could work, but it has to be tiny or it eats the foam away.
Alternatively one could slit the panels about 15-20mm wide and use UD tape. Personally I'd favour the roving as I feel round holes create less of a breaking spot.

2) Strip plank the surface with glass in between the strips connecting the skins.
Cut yourself a bunch of XPS strips of say 50-100mm width.
Lay down the inner skin laminate for say the first 5 strips. Use a slow hardener.
Put a tape of glass down which overlaps the seam by about 30mm. Width of the tape 30mm+XPS thickness+30mm. Wet out the glass tape as well. Lay down your first foam strip. Fold glass up. Lay down second foam strip and proceed as before. Proceed until panel is layed up & glass over the top.

In both cases there is one issue which may be difficult. Wicking. How do you avoid that the glass in the roving holes or the glass tapes are not getting "wet" in the area where they exit the topside of the panel? If the are full of resin, they become stiff and with it the top skin gets hard to laminate. One reason why the threading method of continous strands mentioned under 1) might be easier, as the strands are pulled flat already.
When laminating the top skin we should perhaps apply some unthickened resin at the fiber "exit" holes though to ensure that they are wetted inside the holes.

Thoughts? Comments? Could that help?

 

[Prettypicturegirl]

Thats quite interesting. I thought PET fam would fair more akin to PVC foam, but that seems not the case.

Thank you very much RXComposite! This is very helpful.

I've looked a bit around and finally found some values for sheer strength of Styrofoam XPS foam. Ok it was right here on the forum It seems if you get to densities of 40-45kg/m³ you get shear strength of 0.4 to 0.5 MPa. So comparable to lower density PET or PMI foam.

XPS Styrofoam RTM-X
Density 40 kg/m³, Sheer strength 0.4 MPa

XPS Styrofoam HD 300F-X
Density 45 kg/m³, Sheer strength 0.5 MPa

I've attached my spreadsheet with some collected material properties for foams, no guarantees that I typed off the values correctly though.

 

[fallguy]

All of the 'solutions' people propose modify the shear and compressive (sometimes) properties of foams designed for alternative uses than marine foams. Or they modify the foam enough that it only becomes a former for a web of intricate fitments.

This is largely an exercise in reverse engineering a product whose primary goal was insulation into a 'cheaper' alternative to costlier marine foams designed at the outset with better properties. No, ta da, right?

Well, there is...none of the aforementioned 'solutions' accounts for the costs of said. And there is the ta da.

By the time you invest large sums of capital into machines and plants to modify the cheap core and add the labor of people running the stuff through a stitching machine; the stuff is no longer a cheap alternative, but a fantasy. And. And. Still not as good.

Now, there is some good in all the banter. Some of us gain a deeper understanding. But it is important to not keep attempting to solve the problem; that is dysfunction.

Those are my thoughts and comments. And I have used foamular 150 for amas. But its purpose is largely as a former and flotation. It is not designed for much hull stress. There are no G forces!

 

[Prettypicturegirl]

Thanks Fallguy.

Well, I agree on a commercial level its a different ball game in terms of machinery, warranties and so on. I agree in this case absolutely go for high quality marine foams (mainly non-cross linked closed cell PVC).

My thoughts above were relating to this more on the level of a kind of cheaper solution for a low budget DIY home builder.
I am not proposing to use this on a complete boat hull, but possibly parts of it. Primarely as former. Maybe as part of a bow modification, while keeping the existing bow underneath.

You said you used for it for amas. Those are experiencing considerable forces when settling in waves too, no?
Did you strip plank them out of the stuff? Did you glass both sides? How big is the boat? How thick a core did you use?

All of the 'solutions' people propose modify the shear and compressive (sometimes) properties of foams designed for alternative uses than marine foams. Or they modify the foam enough that it only becomes a former for a web of intricate fitments.

This is largely an exercise in reverse engineering a product whose primary goal was insulation into a 'cheaper' alternative to costlier marine foams designed at the outset with better properties. No, ta da, right?

Well, there is...none of the aforementioned 'solutions' accounts for the costs of said. And there is the ta da.

By the time you invest large sums of capital into machines and plants to modify the cheap core and add the labor of people running the stuff through a stitching machine; the stuff is no longer a cheap alternative, but a fantasy. And. And. Still not as good.

Now, there is some good in all the banter. Some of us gain a deeper understanding. But it is important to not keep attempting to solve the problem; that is dysfunction.

Those are my thoughts and comments. And I have used foamular 150 for amas. But its purpose is largely as a former and flotation. It is not designed for much hull stress. There are no G forces!