S Glass

Why isn't S glass used more?

In rough terms, S glass is about 20% stronger, 20% stiffer, and 20% more expensive than E (ordinary) glass fiber. In principle, any successful design in E glass could be made with 20% less S glass at slightly lower cost, with a substantial savings in weight (both glass and resin).

As far as I've been able to discover, builders don't use S because designers don't design for it, designers don't because there aren't any scantlings for S, there aren't any scantlings because there aren't enough boats out there to develop experience. Also the selection of mat and roving is limited and prices are high. The suppliers say this is because demand isn't out there.

This seems to be a classic chicken-or-the-egg. I think there's another factor: carbon is sexy. Carbon fiber looks different, so everyone knows you spent a fortune to make your boat lighter, stronger, faster. S looks the same as E, so it isn't impressive; you also run the risk of paying extra for S, but some shyster pawned off E on you.

I wonder if the folks making S glass could add an additive to make it a distinctive color.

Usually, when I mention S glass, people say something about carbon fiber, like "Why not go whole hog?". The short answer is that carbon is expensive and likely to stay fairly pricey for my lifetime. S is as strong as high-modulus CF and nearly as stiff as high-strength CF, at a fraction of the cost.

Some folks then suggest putting a little bit of CF into an E glass design. This can work, in very specific designs, but mixing different modulus materials isn't generally good design: as the structure deforms, the softer material will deform away from the stress, leaving the stiffer material holding the bag. Without very careful deformation design a mixed structure is essentially the high-modulus material, with a bunch of filler.

The raw fiber is readily available; the whole infrastructure of E glass should work for S with minimal tweaking. Why is everyone so excited about carbon, aramid, etc. and completely ignoring a much simpler to adapt and more widely applicable fiber technology?

 

You can buy black 'glass in the various weaves of carbon (3x3 twill, etc.) for not much more than regular fabric of the same construction. Unless you're very familiar with carbon, it looks the same at a fraction of the cost.

Physical spec's for these fabrics are known, though some less available than others. In high end racing applications, you do see better fabrics and laminates, but for the most part, you can afford the additional weight and bulk of the more commonly available stock, so a simple business decision is made. Carbon has now taken on a similar mystique as epoxy, in regard to marketing. The word epoxy in the title or name of a product means you'll sell more, even if it's only a fraction of the epoxy molecule. Kevlar reinforced products have the same advantage, plus you can jack the price up, well beyond what you have in it. The same is true of carbon. You really don't need a carbon tiller at 10 times the cost of a length of aluminum tubing that serves the same role, but damn, it's easy to sell this to some folks, so jump on, the band wagon is rolling baby. I have a carbon fiber/kevlar reinforced house, set in epoxy and painted with epoxy I'd like to sell you. Yeah, it's several times the market value, but look at what you're getting.

 

A thinner laminate will be more flexible, which requires a more complex structure on the stiffeners. Unless you are building a high performance boat, it doesn't make that much of a difference. A 10% savings on the hull weight is probably less than 3% overall savings on a cruiser, most likely less.

 

I go here, pick a sample and I get 82% more $ for S.

http://www.uscomposites.com/cloth.html

> S is ...nearly as stiff as high-strength CF

Do you example figures?

Maybe basalt fits in somewhere.

 

I can't give you a definitive answer, but I do have a few remarks.

1. I don't think that the main reason S-glass is used is that the actual designs can't be done. True, with less experience designers and builders can have less confidence, but considering the enthusiasm that has gone behind introducing some newer materials and considering the fact there is a decent amount of work done with S-glass, I think it is certainly regarded as a viable option from a "do we know how to analyze and design it" standpoint.
   
2. I'm not sure the weight savings are as great as one may hope. If S-glass is 20% stronger/stiffer (nothing's that straightforward, of course), that does not automatically mean that replacing E-glass with S-glass will make a vessel 20% lighter. Here are some examples why.
   
   Efficient designs very often use hull plating which is the minimum thickness allowed to meet durability or skin stability requirements. Using a stronger, lighter material won't necessarily help if the strength wasn't what drove your design. This might let you increase unsupported panel size, but this may or may not actually be helpful.
   
   For a stronger, lighter material to have the same properties, the entire section can't be shrunk unless you want to add more material. This can lead to using more core material (which may detract from weight savings) or more complex geometries (which can be very problematic).
   
3. Adding carbon to an e-glass design, for example in stiffener caps and around penetrations, is not all that problematic. It's true there's a discontinuity, but construction methods allow a good bond. Modern FRP design is full of composite built-up sections, ranging from drastically different layer properties for fabrics to dramatic examples such as sandwich construction. Carbon stiffening in e-glass vessels has been done a lot and I don't think there have been too many problems. (We've been adding steel to concrete for even longer and that seems to be working out well.)
   
4. The risk of a shyster selling you E-glass and calling it S can be mitigated by testing. If you're trying to build a light GRP boat, you need to be testing anyhow. If you don't care that much, you're probably going to cheap out on the material.
   

Just some thoughts.

 

jonr: You're absolutely correct; the suppliers of woven, mat, etc. I've found charge a substantial premium for S. The 20% price difference I referred to was for simple fiber.

The fiber costs more because the high silica content requires higher temperatures to work, which increases costs in several ways.

Once the fiber is made, I don't see any reason the rest of the processing should cost any more than E. So why does US Composites (and all the other suppliers I've looked at) think it should command an 82% premium? The only answers I've come up with are "Volume" and "They figure the traffic will bear it". In the absence of any fundamental cost difference, one could imagine some specialist coming into the market, willing to accept a somewhat lower, but still very profitable, markup and providing the selection and specific knowledge needed by builders.

I'm not an expert on CF, but my understanding is that turbostratic carbon fibers tend to have high tensile strength, whereas heat-treated mesophase-pitch-derived carbon fibers have high Young's modulus.

From http://www.netcomposites.com/guide/carbon-fibrefiber/34, I get a Young's modulus of 230GPa , from http://en.wikipedia.org/wiki/Stiffn...mate_specific_stiffness_for_various_materials, density 2.15 (assuming CF all has the same density), for a specific modulus of 106. With S specific modulus of 36, I guess "nearly as stiff" is a bit of an exaggeration; I plead faulty memory. There is a range of about 2:1 for strength and modulus among various kinds of CF, so I find it confusing to sort out various claims.

Mike Graham:
1. Of course the engineering could be done; I just don't see anyone doing it.

2. I was mainly thinking of cored design, where the overall thickness is determined independently of the laminate thickness. I wasn't trying to claim that the hull would be 20% lighter, just that it would be somewhat lighter.

There certainly are instances where the laminate needs to be X thick, regardless of strength or stiffness; for those, E glass is just fine. My impression is that for modern designs, those places are a small fraction of the total laminate. I also imagine that a skilled design could replace many of those with cored structures and achieve the needed durability and skin stability or consolidate the structure with a load-bearing structure for a net reduction in weight.

3. I've no doubt that adding CF can be done properly, but I'm also sure that doing so properly involves a great deal of "overhead". A structure that is half CF and half E glass is going to be about half as stiff as an all CF structure and about double the weight. A structure made of a single modulus fiber will use the specific modulus more efficiently. Steel "works" in concrete because we don't expect any more tensile strength from concrete than that provided by the steel.

I should think that, if the design considerations do push toward a mixed-modulus construction, that the discontinuity and other difficulties from using a fair amount of S in a mainly E design would be smaller, resulting in a more cost-effective solution than a tiny bit of CF in E.

4. I'm not sure what kind of testing you have in mind. The sorts of testing that spring to mind are destructive. Do the folks building with CF, aramid, and other exotica test their fibers? I had assumed most of them relied on their suppliers to provide them per specification. I was mainly thinking about the end buyer (or their surveyor), who might have concerns about why all the layers of glass are thinnish.

 

Forget your misunderstanding concrete, what do you think you mean by "A structure that is half CF and half E glass is going to be about half as stiff as an all CF structure and about double the weight."?

Are you re-enforcing air?

 

I am not sure your numbers here are quite right, and I'm not clear what you're describing about reinforced concrete (I apologize for bringing up the irrelevant topic). Mixed modulus construction is unavoidable here -- fiber and resin and core are drastically different in stiffness. I realize that insofar as this is a problem the use of carbon exacerbates it, but I don't think it's as much of a problem in the general case as you're making it out to be.

Yes, for high-performance FRP designs (be them e-glass, carbon, or other), destructive tests on coupons or sometimes for elements and even sub-components are performed. The real properties and reliability of an FRP structure depend heavily on things the fabric supplier does not know (your resin, your construction procedure, etc.) For higher-performance classed vessels, this can lead to approval for relaxed allowables, compared to the rules-based allowables which necessarily assume the worst possible construction conditions. For super-high-performance vessels, this can allow design by quite thin margins without major safety risks.

 

Look here http://www.christinedemerchant.com/carbon-kevlar-glass-comparison.html
Read chapter 'Modulus of Elasticity'. The difference in elasticity is the key..
BR Teddy

 

Like I thought.

Although you did mention a good article.

 

Seriously, the article was good.

And contrary to previous claims, it does not suggest you do not mix CF and Glass, or even CF, Glass, and Kevlar (Aramids - there are different ones, and ironically they mentioned using nomax .... )

One of the points made in the article was to intentionally mix all three to lessen the weaknesses inherent to each one, use the strength of another fiber to overcome an inherent weakness.

A very good read.

 

I'd be glad for an explanation of what I misunderstood about concrete. If you have a reference to a concrete structure that has more tensile strength than the steel it contains, I'd love to see it. If you're talking about the fact that concrete is usually used for compressive loads, then I agree, but that has little to do with why S glass isn't used more widely.

If one is to make a structure containing both carbon fiber and E glass, it will, of course, need to have a resin. Since the resin will be needed, regardless of the choice of fiber, one can examine the affect of different fibers by regarding the resin as constant. The fiber's role is primarily to provide tensile strength and stiffness, so one can study the tensile properties of the structure with the known, but nearly constant, approximation that the resin provides no contribution to the tensile properties. An actual structure will need a satisfactory resin and the strength properties depend vitally on satisfactory performance, but the resin is pretty much irrelevant to stiffness.

The carbon fiber is usually used as a reinforcing layer, laminated between layers of E glass, with resin saturating the whole laminate. I call this a parallel structure. When such a structure is stressed, the resin stretches, leaving the load to the fibers, of which the glass stretches, leaving the bulk of the load to the carbon fiber.

The Young's modulus of epoxy is 3.5, E glass 81, and carbon fiber 230, so this is like jello, reinforced with rubber bands and string: when you pull on it, the load is taken by the part that has the least give (string/carbon fiber).

Sometimes, the carbon fiber is in the form of a strut, tying 2 larger E glass members together. I call this serial, to distinguish from the parallel case, above. In this case, you get most of the deformation in the softer material, unless the E glass has a much larger section than the carbon fiber (by a factor of 3). In most cases, this kind of structure requires careful design to bond the 2 materials together without a stress concentration that produces a progressive failure. A lot of designs solve this problem by using a lot of excess material in an region overlapping the joint.

If we ignore the joint overhead, serial mixed-modulus is still inefficient. If we compare 1 meter of E glass to half a meter of carbon fiber in series with half a meter of E glass (with the same cross-sectional area), the mixed is slightly lighter (0.5*2.15 + 0.5*2.62)/2.62 or 91%, again assuming the resin is the same for both. The series structure is also somewhat stiffer 1 - (0.5/81 + 0.5/230)/(1/81) or 32% less deflection. (This is a tiny bit harsh on carbon fiber, because optimum uses 1/3 the cross-sectional area for the carbon fiber piece. If someone wants to crunch the numbers, you'll get a small incremental advantage in weight for a bit less stiffness.) On the other hand, switching to S glass for the whole thing, one can reduce the cross section to get to the same weight as the series and end up with 1 - 81/(89/(0.91*2.62/2.5) or 13% less deflection (compared to the original E glass). The overhead for bonding and joining is more complicated than I care to research at the moment, but there will, of necessity, be some; whether it eats up all the gains depends on details of a particular application. The point is not that carbon fiber is inferior, but that S glass is A LOT cheaper than carbon fiber and seems like it should be a better choice than E glass in any application where both price and performance matter.

If you don't care about performance, then E glass is clearly the way to go. If you've got an unlimited budget, then carbon fiber is clearly a good choice.
I imagined that, like me, most real-world applications for most people are concerned with both.

 

I can agree with that much more than the previous.

I am happy now.

 

> In the absence of any fundamental cost difference,

I am suspicious that there are some fundamental cost differences. It's not like S-glass is new or that there isn't any competition. At the current cost for fabric, S-glass usually doesn't make sense.

 

Sorry about the double counting, that should have been half the stiffness at same weight OR twice the weight at the same stiffness. Those were both crude approximations (grossly favoring the structure with mixed fibers of different moduli) and switching the "or" to "and" was just a brain fade.

I had assumed that everyone would understand that I was only talking about composite materials. Of course resin is much lower in modulus, but good construction assures that the thickness of resin between adjacent fibers is TINY and that the bond between resin and fiber is nearly as strong and stiff as the resin itself. I had intended the phrase "mixed modulus construction" to mean various fibers, with different moduli, since the way you are using it seems to mean the same thing as the more common phrase "composite material". Composite materials do, indeed combine many of the best properties of the fill (resin) and the reinforcement (fiber), provided the bond between the 2 is strong and stable. CF and glass do not bond directly, so one needs to look at the problem somewhat more deeply than a simple analogy to a 2-component composite material.

For the case of mixed fibers of different moduli (until someone suggests a more apt term), http://www.christinedemerchant.com/carbon-kevlar-glass-comparison.html does suggest that blends might help when there are 2 or more desirable characteristics that need to be met, although it doesn't present any plan or analysis as to how to achieve that. The case discussed is Kevlar with other fibers to provide stiffness and compressive strength. This doesn't guarantee that the tradeoff between properties is efficient, eg. one might have to give up 90% of the difference in strength to achieve 10% of the difference in stiffness. In any case, the design task is much more complex than with the simple 2-component composite material.

As far as testing, I guess I was vaguely aware that builders test new techniques and unusual structures, but I doubt that production builders test every batch of material that comes in to make sure they aren't getting black glass labelled as "Carbon Fiber". What I've read and heard from designers and engineers is generally that they use ABS layup schedules for most applications and that the schedules are planned for E glass. For cost-is-no-object organizations, I'm sure you're right that they do a lot of testing, but I doubt it goes so far as to build and destroy every structure in the hull.

I'm sticking to my main point: S provides substantial advantages over E, is a lot cheaper than carbon and all the other exotic fibers, and simpler to design than any mixed construction using fibers of different moduli, but I'll stop beating that particular dead horse.