Extruded Polystyrene (XPS)

I reckon you'd do this for similar money
http://www.polycore.com.au/
And, you could use polyester resin and thin gelcoat (shudders), smooths finished face off of flat table

or a compromise, balsa ply with a 220gsm DB skin (there was some light db getting around a while ago)

 

I have been shying away from honeycomb because it seems very difficult to edge and round off.

 

I've done some more reading and learned a little about the engineering of sandwich panels... In relation to 400kpa XPS foam, I used the published data for its shear strength and shear modulus etc to work out a maximum woven glass laminate. The shear strength of the core, being lower than PVC foam, can only handle a glass/epoxy laminate thickness of 0.72mm or approx 676gsm cloth both sides. Increasing the skin thickness won't make a stronger panel as the core will fail first.

If using the highest rated XPS I can find, which is 700kpa compressive, the shear strength is also higher and the core should support a glass\epoxy laminate of around 1.27mm skin thickness. If anyone would like to check my math please chime in if I've made a mistake...

 

groper, core thickness needs to be factored into this, what thickness core did you calculate with?

 

From the reference material I have read, the thickness is not relevant... Its the tensile and compressive strength in the skins in proportion to the shear strength of the core... So The thicker the core is, the stronger and stiffer the panel becomes, however it is of zero benefit to increase the skin thickness beyond what is calculated balanced skin strength vs core shear... So you can choose the core thickness based on your strength and stifness requirements. If you need more strength from a thinner panel, then you increase the core strength and it will carry a heavier laminate.
The literature I've been reading is a really good PDF on the hexel website, sorry I don't Have the link as I'm on the iPhone away from home at present...

It also goes into detail about the different loadings, whether it be a panel in bending or a panel with end loading, such as bulkheads and stiffeners etc. for a bulkhead, the considerations include panel buckling, skin stress, skin wrinkling etc .... I've done the calcs on most of this stuff and XPS seems ok as the main influences are compressive modulus and the sandwich panel stiffness... the core shear strength is not part of these equations...
I've done up all the equations in an excel spreadsheet, let me tidy it all up and I'll email it you like the one you did for me

 

Email?? Could you please put them up here?

 

Yeah, we wanna see too!

 

Frequently there are requirements for both bending strength and stiffness, and both need to be checked.

If bending strength is the limiting factor, ie the maximum load before structural failure, then balancing the tensile/compressive strength of the face laminates with the shear strength of the core makes sense. Increasing the strength of the face laminates beyond what the core can take in shear would result in core failure before laminate failure.

However if bending stiffness is the limiting factor, ie the maximum deflection under a give load, then the deflection can be decreased by increasing core thickness without increasing face laminate thickness.

The above assumes the core has a much larger elastic strain limit than the face laminate.

 

All things considered, one can easily select a core material to get the job at hand, done. Whether it be strength stiffness weight or cost...

Google "hexcel honeycomb sandwich design technology" the PDF file is GOLD.

Lastly, the next couple of boats I build will make extensive use of XPS foam, in combination with PVC foam in higher stress areas.

 

This is the PDF file i was referring to earlier...

http://www.hexcel.com/Resources/Dat...eets/Honeycomb_Sandwich_Design_Technology.pdf

the formulas contained within it pertain to sandwich and plate theory... amongst other things, it will allow you work out whether a particular core material is suitable for a desired purpose. You can also use it to determine what density foam is required to give enough shear strength for a given laminate schedule...

I suggest automating the process by way of excel spreadsheet....

If you go ahead and do the calculations, you can see that the med and high density XPS can be used to build some very strong composite panels, particularly if you are prepared to use a reasonably thick core which is certainly realistic considering XPS is 1/10th the price of PVC foam and offers the same longevity and resistance to moisture.

 

There seems to be some confusion here. In a firmly fixed panel loading, thickness is very important to determine the minimum shear strength of the core. If the shear strength (of the core) is inadequate, you can do one of three things;

1. Increase the modulus of the face by;
a. using high modulus fiber such as carbon fiber or
b. increase the modulus by using a process that will increase the glass content of the skin, thereby increasing the modulus.
2. Increase the thickness of the core, reducing the requirement for core shear strength. But in doing so, the skins takes up the load (compressive and tensile) and must increase in area/thickness to compensate for larger load bearing. Strength remains the same. Remember it is moving away from the neutral axis (center of core)
3. Use a core with higher shear properties if you want to maintain the thickness.

 

RX,

in your point 1, i dont think this is correct. heres why...

Using the formula from beam theory, where core stress = F/h*b
F= P/2 for simply supported and fixed both ends, panel loading.
P= applied load
h= panel thickness
b= panel width

You can see that the modulus of the skins is nowhere in this formula, therefore has no effect on the shear stress in the core.

Your other points i agree with entirely...

Ill clarify what i said earlier about the core thickness being irrelevant... What i was referring to is the ability of a particular core MATERIAL and therefore fixed material properties of shear strength and shear modulus etc, being compatible with a particular laminate schedule. So, If you have a laminate schedule of lets say, 1mm of of UD carbon/epoxy both sides, the skins have a fixed amount of tensile and compressive strength contained within this schedule. Whatever panel you design, the tensile and compressive strengths cannot ever be exceeded by the skin stress or skin failure will result - no brainer here. However, the designed panel core material must also be able to cope with shear stresses or the panel still fails, this time by core shear stress exceeding the core shear strength. So the core MATERIAL must be compatible with this laminate or you are wasting your time designing a panel for best weight, strength, and cost. This was the subject of debate earlier in the thread whereby some people were saying that XPS foam was not suitable for structural purposes etc. Well the truth is, that XPS is suitable for structural purposes, it just depends on how much strength you need and the type of laminate applied to it...

Using the formulas i could see that the stress in the core exceeded the strength of 400kpa XPS, when a woven glass/epoxy laminate exceeded 0.72mm, regardless of total panel thickness. If i were to consider a woven carbon laminate instead of the woven glass on the same core material, then the max laminate thickness would be approx 50% thinner... Why? The skin stress does not change for a given load... There is the point where the tensile/compressive stress in the skin = the tensile/compressive strength of the skin, AND the shear stress in the core = shear strength of the core. This is the ideal point engineering a panel as neither core or skin will fail long before the other. The load carrying capacity and stiffness of the panel changes with core thickness, but the ability of the core material to carry a stronger laminate does not. The laminate thickness and type always balances with core material properties (not core thickness) in an ideally designed panel.

If you go back to my example above of the 1mm thickness UD carbon laminate and the material properties associated with it, using the formulas you can see that the core material must have a shear strength of around 2.6Mpa before it will fail around the same time as the laminate will.... so to use this laminate schedule, you should buy a foam with at least 2.6Mpa shear strength or you are wasting your time placing this laminate on it, regardless of the core thickness- so you would need one of the high density PVC foams to carry it, XPS is definately not upto the task in this example...

By way of the same reasoning, you have wasted alot of money buying high density PVC foams for building your boat, if you are only using light laminate schedules because the shear strength of the core far exceeds the stresses it will ever be subjected to... There are other reasons to use a heavier foam besides simple bending loads however...

 

Looking at the http://www.3d-core.com/en/products.html site I notice that they offer E|PIR™ non flammable foam from Polyisocyanurate PIR, but only available in thicknesses 5mm up to 15mm. In his post yesterday, rxcomposite suggests increasing the thickness of the core, reducing the requirement for core shear strength.

Polyisocyanurate foam is considered the best performing house build insulation. Strength is vital. We built a monopitch extension and laid Celotex 140mm thick panels, 2240mm x 1220mm, mechanically fastened directly to the joists set at 400 mm centres. The roof totals 42 square metres and walking & jumping all over it has still shown no sign that the panels are flexing. It's waterproofed now with an ERDM membrane bonded to the Celotex. The undersurface is hidden with plasterboard.

I mention this because purchasing Celotex or other PIR boards from builders' merchants can save considerable money. Kerfing the boards helps drapability when using thick panels. Over specifying PIR bulkheads seems laudable. Skin stress is reduced by the sheer bulk of the core.

http://en.wikipedia.org/wiki/Polyisocyanurate

Consider how dugout canoes are created. That would be wasteful with PIR, but construct the canoe's form with 140mm (5.5inches) thick planks bonded wth epoxy resin and biax cloth and a strong light vessel is the result.

Sized up with PIR frames and 5.5 inch planks, a lightweight Matthew could be built.

http://www.matthew.co.uk/gallery.php

http://www.boatdesign.net/forums/materials/inexpensive-lightweight-core-interior-uses-37233.html

 

Pericles, the main difference you see with house panels VS boat panels is the core thickness... Typical house panels are around 100mm thick whereas boat panels are typically much less.... the reason for this is that boat panels are usually curved, whereas house panels are typically flat... is much harder to bend a very thick sandwich panel as its stiffness is great, so making boats with 100mm thick foam sandwich hulls is difficult... Also the insulation properties of house panels is much more important for energy efficiency and the main reason for using them as opposed to conventional building methods, thus the thickness of insulation required makes a thicker panel.

 

Groper,

Kerfing the PIR panels & lengths by hand saw, on the inside of the bends as per page six of this pdf enables curves to be followed.

http://www.diabgroup.com/americas/u_literature/u_pdf_files/u_bul_pdf/Foam_Core_Marine_TB.pdf

If the frames and bulkheads were fabricated out of PIR and then planked with lengths of kerfed PIR & then laminating with Biax cloth and epoxy resin inside & out, the hull would be stiff & strong. Not all hull shapes are possible with this method, but these offerings from Jordan suggest some ideas

http://jordanboats.co.uk/JB/current_designs.htm

Working with 140mm for the transom, 100mm for the frames & bulkheads & 50mm for planking & decks, it'd be a tough as old boots & very well insulated.

I was able to purchase seconds. http://www.aandainsulationservices.com/