Hexacor compared with other composites?

[wjones]

Aluminum Core Panels

There is a company in the US that manufactures aluminum honeycomb core panels that are IMO, DNV, USCG and SOLAS class-C certified.

www.ayrescom.com

 

[juiceclark]

WJ,
I had no idea. I've already began figuring what laminates we'll need on their cores for our bulkheads. They owe you a discount! TC

 

[Edmundo Souto]

PP honey combs are cheaper but foams are easer to use specially when you are infusing a whole hull at once.

In a infusion, you can use vinylester poliester or epoxi resin, but vaccuun bagging you can only use epoxi.

 

[TeddyDiver]

Does someone have an idea where/how to get some 30m2 in Europe. I sended an info request to Hexacor in China but they just want to mingle..
It's not a big deal but..

 

[Analyst]

Details about PP core and general core design

I just wanted to clear up some of the fog on polypropylene honeycomb core materials and what I have found through testing and parametric study for use on high production fishing boats and yachts.

:!: GENERAL CORE DESIGN:

(Not directly concerning polypropylene, but core in general) The proper method of design for sandwich structures is to include the effects of transverse shear deformation, or the sliding of the face sheets due to the shearing component of the core during bending. If displacement is of concern, a low value for shear stiffness (as found in PP and many foam cores) will cause larger deflections in reality, usually much larger than isotropic (uniform properties) equations will give you. Some references would be books by Jack R. Vinson or Chandrashekhara or any modern composite plate and shell text. Hexcell also covers this for (oversimplified, but ballpark) semi-composite analysis in their honeycomb design reference which I believe is still free. Diab also discusses it in their online manual.

:!: Generally the ideal sandwich structure has:

* Thin face sheets (you would be surprised at how thin if you use the correct materials and core height) -- (The major weight and cost savings is dramatic for the proper use of sandwich design). (not necessarily core material dependent, the design philosophy is more important) I have also seen engineers use the same solid laminate and just throw in a core somewhere in the middle, and expect a weight or cost reduction. Strange, huh?)
* A core with extremely low density (PP has at around 5 pcf)
* Low cost (PP is the lowest I've found to date)
* High stiffness (PP is low compared to many foams and balsa, but how much do you need? Comparison studies will tell you what is best if the proper equations are used). It is important to note here that by nature of materials under dynamic loading, high stiffness materials become brittle and softer ones become stiff, so a study should be performed for the specific application to determine the proper stiffness and strength requirement of the core.
* High shear strength (PP is also low in this category, but the question is emphasized as to "how much do you need to not fail?" which is answered through rigorous engineering analysis.)
* High strain elongation to prevent catastrophic failure (PP has a very high strain elongation), formability (Honeycomb is inherently the best at this, but the scrim of PP makes it difficult, but can be scored)
* No resin absorption. (PP absorbs very little)
* Little to no print-through at splice joints (PP was excellent compared to balsa for the same resin content)

:!: POLYPROPYLENE HONEYCOMB:

Therefore PP Honeycomb contains most of the important requirements for ideal core, with the exception of the two major ones: strength and stiffness. However, if proper design (designing to what you actually need instead of some arbitrary standard and using the correct equations to carry out the design) is used, the result is exceptional.

Polypropylene is next to impermeable to water. It is also a recyclable thermoplastic with a higher Tg than many thermoset materials and will melt instead of burn eliminating a major energy source for boat fires.

CAUTIONS WITH THERMOPLASTICS AND PP HONEYCOMB:

Thermoplastics are awesome. When I was in aerospace, only the best applications got thermoplastics becaues of their processing expense. They give you the so needed ductility and resilient that thermosets lack and are recyclable, heat formable and cheap. However for combining PP core with thermosets such as Polyester, Vinylester and Epoxy, care must be taken when bonding to thermosets in regions where the scrim has been removed because PP is a thermoplastic. It is effectively a release agent and will not allow chemical bonds to most materials. Ingenuity must be present when bonding at tapered regions, where the scrim is removed. There are methods to do so, such as thermofusing the scrim onto the exposed angled surface with back a vibrating tool.

The key to PP core, is that the increase in height to maintain deflection allowables or reduce facesheet stresses is usually at an insignificant cost and weight increase and a major cost savings and improvement in durability, water resistance and surface appeal foam and balsa. In other words I can achieve the same deflections and stresses as the stiffer materials while also achieving a cost and weight savings by increasing the core height to match the flexural properties.

* Note that I have not discussed plywood. Why? Plywood is technically not a core material. It is closer to a stack of fiberglass than any core material on the market. Although extremely heavy and disastrous in wet and humid environments, it is a very - very efficient material from a specific strength and stiffness standpoint. Its in-plane stiffness jumps the laminate flexural properties through the roof and its shear stiffness properties put even the best balsa to shame, while maintaining in the range of one-third to two-thirds the weight of fiberglass. So by comparison with other core it is too heavy to be used as an efficient core material and by comparison with fiberglass too environmentally unstable to be exposed. As a result, I don't endorse plywood as a good material for structural use in wet environments for dynamically loaded "lightweight" products such as boats.

THERMAL STABILITY?

An issue that I am concerned with and looking into is thermal degradation effects. It has been observed that while ambient air temperatures do not reach 180 degrees F under UV exposure, insulated surfaces such as FG can, (ever burn you feet on hot sand, or on the pavement or on the boat floor?). The mechanical properties as a function of T could play an important factor in the continued use of PP on UV exposed regions. I have been told that it can be autoclave cure with prepregs, which tells me that it should be mechanically sound at these temperatures, but I wonder if anyone has any information or reference about this topic?

I could say more but as long as this thing already is would just be rude.

Thanks

 

[Edmundo Souto]

Did you noticed DNV DO NOT allow any PP honeycomb on decks and sides...??? Check DNV approval sheet for PP core....

 

[kengrome]

What is DNV?

 

[Edmundo Souto]

It´s independent international company of classification,certification and consulting for materials, construtions ....etc
Used as reference speacially in boatbuilding and others...


DNV - Services for Managing Risk

www.dnv.com/industry/maritime/

 

[Analyst]

Edmundo Souto:

Yes, I have heard (through a PP HC core supplier) that DNV has created a restriction on PP materials on decks and sides.

To that, I ask you, what is the motivation of a certification and standards organization in restricting a material type for any reason other than environmental factors, human health factors, and/or processing hazards?
PP has none of the problems mentioned above.

Unless the material has one of the three problems stated above, there is no intelligent means of restricting a material type.

Otherwise, the motivation for the restriction could only be 1. political (ties with companies that make competitive materials intended for the same application), 2. empirical (equations in standards and testing methods are insufficient in defining the behavior of the material (common problem)) 3. Prejudice (individuals creating the standard have a personal preference in what kind of materials that they want to see for that application and apply a God-like rule over the design for structures requiring their certification).

As stated before, material selection must be done on an application basis. Good certification organizations and self-checking companies apply specifications on the mechanical or physical properties of the material (not the material type itself) on an application based method (hull/liner/door/spar/wing skin/etc.). To go even further, especially with composite structures, laminate and/or sandwich properties are a means of specification, ignoring completely (within the spec) the properties of the plies/or core making up the laminate, although they must be known to create the laminate/sandwich properties.

(i.e. for a hull bottom panel of a ship of X length and Y beam with a panel aspect ratio of a/b constrained by BC1, BC2, ... BCn, at a max speed of ZZ knots, in sea conditions ABC or D which creates a panel pressure zone of C psi or N/mm^2 at that station location, the cross section of the structure/panel must have an overall flexural stiffness of XXXX to maintain a deflection of W.WWW inches or mm and critical buckling load higher than XXXX lb/in or N/mm and must have a stress/strain allowables of YYYYY psi or N/mm^2 or must meet a factor of safety of EEE.)

Note that in the above example, the material type would never enter the calculation.

Also additional constraints that may be more material dependent could be peel stress/strain of a bond, or energy release rate of an cut-out, thermal/moisture expansion rate, thermal/moisture degradation rate, impact energy absorption, surface hardness for debree, degree of chemical resistance, etc. These are all still defined by their performance in the specific application, not by the name of the material.

Material type just does not come into the picture. The requirement is only that the mechanical properties of the materials within the structure meet these requirements. Hence, Foam versus PP versus Balsa are be examined on an equal playing field.

If the motivation for the restriction is test based, what may be likely is that the DNV performed tests on panels with identical core thickness and did not account for the specific properties. You cannot expect a core material of the same thickness but with a lower transverse shear modulus (Gc or Gxz, Gyz or G13, G23 or GL, Gw) to achieve the same deflection as a material with a higher transverse shear modulus. This still cannot be grounds for material-type restriction. It should be designer’s preference. (If the designer wants to save money and eliminate weight and wood by replacing the 0.5 inch Balsa with 0.75 inch PP Honeycomb, it should be allowed to do so. If a designer wants to use 10 inches of HDPE on the sides and bottom of his/her boat and can prove that structural integrity and stability is achieved then this should be his/her choice to do so. It may not be a good idea, but if it works, it works. A certification company cannot limit the ingenuity of the marketplace to materials that they see fit.)

It is also likely (and common) that the wrong test for Honeycomb sandwich panels was performed. C393-62 is acceptable, whereas the common 1 inch wide strip 3-point bending method is inaccurate for HC because of cell size.

Nevertheless, material selection MUST BE APPLICATION BASED NOT MATERIAL TYPE BASED.

Sorry about the length. I will try and keep it short next time.

 

[kengrome]

Hi Analyst,

There's no need to keep it short the next time, I appreciate everything you've posted and more. Please continue to post as much new and additional information as you wish -- the more detailed the better!



Here's where I was coming from when I started this thread ...

I suspect that 9mm thick Hexacor is not thick enough to replace 9mm plywood as the core in a composite sandwich if I want a sheer strength that's high enough for a particular application -- a small, extreme shallow water planing powerboat. I was thinking of using 25mm Hexacor instead, not only for light weight but also to eliminate wood as the core material so buyers won't be thinking of 'rot problems'.

Unfortunately I do not have the ability to do the calculations to determine whether or not 25mm Hexacor will be strong enough to be a good replacement for the 9mm plywood core, and this is why I've been trying to learn more from people who actually have more experience using this material than I do.

 

[Analyst]

Kengrome:

You are correct in assuming that you will need a thicker PP Honeycomb core to achieve the same stiffness and strength properties as plywood.

To assist you in your analysis, I wanted to give you some of the basic simplified equations to use for comparison.

What you will want to know and compare for complete structural analysis are the differences between each construction for:

1. The maximum deflection of the panel.
2. The maximum stresses of the core and the facesheets.
3. The critical buckling load of the panel(s) and facesheets.
4. The natural vibration frequency of the panel.

FOR STARTERS AND TIME CONSTRAINTS, THIS REPLY WILL ONLY BE CONCERNED WITH MAXIMUM DEFLECTION OF A QUASI-ISOTROPIC SANDWICH BEAM WITH TRANSVERSE SHEAR DEFORMATION.

*Note that for simplicity we will assume a linearly static analysis which means that dynamic (time dependent) loads and properties (which are generally not available) are ignored and non-linear terms, such as deformed shape dependent deflection (deflections based on the deformed shape instead of initially undeflected shape) are ignored. For a very detailed analysis, these assumptions should be validated or these terms must be included, but this is seldom the case, if ever, in the marine industry.

*Further, to keep from overloading you with plate and shell theory, which would require more time and room than is available in this format (entire textbooks are devoted to this single subject and many only consider the most simple of cases because of complexity), we will restrict our view to short beam analysis (considering the shorter dimension of the panel as the beam length).

Because I mentioned it before, I will use the equations from the ASTM Standard C393-62 because they include core material and the effect of transverse shear deformation. This method was created using derivations from the principle of minimum potential energy using Reissner's Method.

The maximum deflection can be determined by the following equations, which can then be correlated with an experiment to determine the true values of the properties:

3-Point Bending:

wmax = {(P*L^3)/(48*D11)}+{(P*L)/[4*Gc*((tf+hc)^2/hc))*b]} (1)

4-Point Bending:

wmax = {(11*P*L^3)/(768*D11)}+{(P*L)/[8*Gc*((tf+hc)^2/hc))*b)]} (2)

Where:

* D11 = Flexural stiffness coefficient defined by:

In General:

Dij = (1/3)*(Qbar)ij*(z(k)-z(k-1)) which for isotropic or quasi-isotropic facesheets and core is:

D11 ~ (1/3)*{(2*E11)/(1-v12^2)*[0.75*hc^2*tf+1.5*hc*tf^2+tf^3]+(Ec/(1-vc^2))*(0.25*hc^3)}

for simplified analysis of a sandwich beam with isotropic or quasi-isotropic facesheets.

Where:

-E11 = effective elastic modulus of the facesheets (assuming isotropic), also referred to in data sheets as Ex or Exx. Could also be Ef for Efiber not to be confused with Ef for Eflexural.
-Ec = inplane elastic modulus of the core material, usually neglected because of its small magnitude relative to E11.
- v12, vc are Poisson's Ratios for the facesheet material and core, respectively. v12 is apprx. 0.25-0.35 for woven fiberglass. but be careful because I have seen it at 0.08-0.11. Unidirectional tapes generally have different Poisson Ratios, one being lower and the other higher. Poisson's for most ductile metals is around 0.3. Ceramics see very low Poisson's 0.05-0.1 and rubber is very high 0.4-0.5, with 0.5 being the maximum theoretical value for any material.
-hc = Core thickness
-tf = single facesheet thickness (will be better defined later)

* L = Length of the beam (shorter span length of the panel)

* P = Load in force (lbs or Newtons) (Note that for the purposes of test correlation we are using force instead of a distributed pressure, however we can also go that route later.

* hc = Core height

*tf = facesheet thickness of one side (note that a simplification and a limitation is implied here. 1. Isotropy is being assumed by stating tf as the facesheet thickness. Generally the thickness of each ply must be accompanied by the stiffness of the ply for accurate analysis. However, if your facesheet ply stiffnesses are not too far from each other, the assumption is reasonable. 2. Bending-Stretching coupling effects are being neglected. In lamination theory, or general elasticity theory, a system of equations ( also referred to as the matrix for people familiar with composite analysis) is formed that is based on the fact that eccentricity of a laminate stack causes deformations about the midplane to loads not applied in the direction of deformation. i.e. for a bending load stretching and shearing will occur, or for a stretching load bending and twisting will occur. This is a major problem between manufacturing and design in terms of warpage and residual stresses and can be alleviated by eliminating the matrix. To do this one must make the facesheets of the stack symmetric and balanced about the geometric midplane. What this means is that the you must have the same plies and ply orientations at every incremental distance from the midplane. If you were to cut out a square section of the laminate, a symmetric and balanced design would allow you to flip it over and see the exact same layup in the same order on each facesheet.

With all of that being said (such a seemingly assumption that has a major implication), coupling effects are less important for sandwich laminate with large flexural stiffness terms, however a proper analysis will include them. Once again, this is out of the scope of the format so we will just neglect this for now.

* b = beam width (in the case of this test, the beam width should be taken as: b = 2*(hc + 2*tf). This will eliminate any problems pertaining to core cell size. This is one of the reasons why this test is accurate and the others are not.

* Gc = Transverse (out-of-plane) shear modulus of the core (also referred to as Gxz, Gyz or G13, G23 or GL, Gw in technical data sheets). Beware that this is not the in-plane shear modulus Gxy or G12 unless the material is isotropic.


Some notes on equations (1) and (2):

1. It should be seen that the last term on the right hand side is the deflection due transverse shear deformation and is dominated by the core shear stiffness Gc and core height hc in the denominator. The first term on the right hand side represents the deflection due to flexural stiffness with respect to face sheet modulus and distance from the midplane, dominated by D11. Many equations neglect core shear stiffness in error by the use of only terms of this type.

2. Because Gc is in the denominator, an increase in Gc will reduce the overall deflection and a decrease in Gc will increase the overall deflection. The same can be said for the hc term. The same can be said as well for the D11 term, in which hc has the largest effect.

3. Therefore, if you have a lower Gc and you want to match deflections for the same load you can do 1 of 3 things or a combination of any or all of the 3 possible choices:

1. You can increase the facesheet stiffness E11 to increase D11 (worst possible choice because you are now working on orders of X^3 instead of just X and you will need alot of facesheet stiffness to make of for the loss in Gc.) This also requires a change in material type (i.e. from fiberglass to carbon fiber) and you still may not meet the same deflection constraint because the impact is small.

2. Increase the facesheet thickness tf to simultaneously increase flexural stiffness D11 and shear stiffness. (also a bad choice because now you are using your heaviest and most expensive material to increase bending inertia and adding extra material, weight, processing steps and cost that you did not need before for in-plane (membrane stresses and facesheet buckling).

3. Increase your core height hc. (Best possible choice because you are again simultaneously improving your flexural stiffness D11 and shear stiffness, using your least expensive, lightest weight and easiest to apply material.)

WARNING: There is a limitation to the benefit of increasing core height in that most core material properties decline with increasing core height. It is usually unnoticeable until an critical value of hc is reached. Core manufacturers always (or at least should always) state the range of core height that the values are valid for and if there is a noticeable drop in value at a specific height, then new properties should be provided for that height range. Hexcel does an excellent job of this (best I've ever seen) and in making sure that all the data is available before releasing a product (there's nothing worse than analyzing structures with sparsely fill out data sheets).

NOTE ON PLYWOOD:

As stated before, plywood is more fundamentally a light stack of semi-stiff plies, and more than a core material because of its high in-plane stiffness values. Its density and weight compared to other core materials is an equal motivation for reclassification. It is far too heavy to be considered side-by-side with core and too light for side-by-side comparison with fiberglass. It is interesting that much of the theory and development of modern laminated composite materials and structures comes from the invention and validation of plywood at the Forest Products Laboratory, and now I am having trouble fitting it into the picture.

At any rate, for plywood, the D11 term is greatly affected by Ec which is non-negligible, whereas for Balsa, Foam, PP etc. Ec is negligible compared to the facesheet materials. While a lot of weight is added this also adds a considerable contribution to bending stiffness. To add Gc of plywood is rather large as well contributing to shear stiffness. Check APA website or Forest Product Laboratory for technical manuals and data sheets on plywood.
However, you end up with a heavy, relatively expensive, water absorbent and degradation susceptible core material.

It would be wise to cross reference equations (1) and (2). I am getting it right now out of my memory, because the actual standard is in a reference manual at my office.

 

[naturewaterboy]

Very interesting discussion! I have a simple question: Why not use balsa for decks and everything but hulls? As long as water doesn't get to it (easy to prevent) it has hi strength, low weight and is not affected by temperature. So why use anything else? I'm not being a smartass, just trying to understand more....

 

[masalai]

Naturewaterboy, "horses for courses", and because some fools do not know the difference, then leglislation happens in its very heavy handed way making a total mess of everything. PP honeycombe panels are far lighter and cheaper than balsa core - if done right excellent for "light point impact load roofing" on cabin tops etc and cupboard/shelving on boats but has a disadvantage in some applications which can be partially addresses by adding appropriate cloth/resin skins.... It is easy to "heat mold" and then add more "glass" to hold it in place, is a good thermal & sound insulator. Likewise balsa core is a bit soft for bottoms BUT, when laminated with appropriate thin layer of plywood and glassed over is OK for lightweight boat bottoms.... ONE MUST KNOW the appropriate materials to achieve the desired result and do good engineering testing and analysis before using anything then quality control in construction and manufacture is the final critical phase in building .....

To put it all very briefly... You still gotta do your engineering - - Don't just take my word for it....

 

[naturewaterboy]

I see! :idea: (said the blind man to his deaf wife as he picked up his hammer and saw)...

 

[masalai]

Pity you are not over here, I would like a steel drum (no musical talent whatsoever, but I like them & would live in hope that I would find people who could and would play it )