Direction of stitched glass fiber laminate?

Steve W-

I feel I still owe you some explanation so I researched further for a written text. Here is one from Filament Winding Composite Structures by S mayers, et al.

For the coupling effect, what I can deduce from the text is "when a 0 degree laminate is subjected to off axis load there is shear distortion". This means the ply starts to distort and the fibers move, tugging along at the next layer. This happens very quickly between 10 to 20 degrees in the Coefficient of Mutual Inductance chart (a ply engineering constant chart). If the adjacent ply is close to or near the off axis load, it takes over.

 

Thanks for the feedback. rxcomposite has a lot of good points and I am still investigating this subject myself.

This is a semidisplacement power catamaran LOA 14M / WL 13M @ 8000-10000 kg. Speed should reach 40+ knots at max

Material is Sandwich of E-glass, Iso Polyester Resin and PVC Core

Attached is some of the information I have found on this topic though it is from 1997, refereed from this paper on page 7/8: http://www.shipstructure.org/pdf/403.pdf

 

Hi,

40+ knots is pretty quick!

Going to be quite a machine once it's built..

Quite surprising(to me) on the table that the 0 degree triax outdid the quadraxial, I would have thought on a 3:1 aspect panel that the quad would have had shorter spans across the :1 distance on a higher percentage of fiber..

Jeff

 

I am not sure about the performance. Calculations on a semi-displacement with a new hull design is kind of a blank paper. Not much theory on this subject that I can find atm. - Still searching though.

I have doen some analysis in Orca but it does not handle multihulls well. Neither semi displacement hulls so I am not sure about the report results.

And it does seem strange that the triax outperform the quad... Is this due to the coupling effect..?

Since the resin is the weaker point in the structure, de-lamination would occur here first so this is an important konstant as well

It does not say anything about buckling which would be interesting. I have to read and calculate some more.

I just like to dig into this field to know why and not only be happy with approximations.

 

You might want to look over Gropers thread http://www.boatdesign.net/forums/bo...displacement-powercat-build-thread-38588.html He has designed & built a 25knot + 10meter power cat, well chronicled in the thread.
As for the graph, I can believe the 90 degree triax but seems a strange result on the quad vs the 0 degree triax... it's either correct(& I'm surprised but not a material scientist) or poorly labelled, I'd love some pro design opinion on this though.

Jeff

 

If it is a 0/+-45 triax, it will outperform the quad of 0/90/+-45, assuming both have the same weight. 90 and +-45 only have 10% of the modulus and strength of the 0 degree. What puzzles me is that a 90/+-45 Triax will outperform a 0/+-45 because the 0 has maximum strength while the +-45 and 90's has only 10% of the strength. Must be a drawing error.

 

Your 10% "strength" number is surprising..

My assumption is that the examples are all the same overall weight but the plies each an equal division of that...
so on the 3:1 panel tested via hydromat method that-
-the 90 triax has 33% fiber spanning 1 and 66% spanning approx 1.4
-the 0 triax has 33% fiber spanning 3 and 66% spanning approx 1.4
-the quad has 25% spanning 3, 25% spanning 1 and 50% spanning approx 1.4.
-the 0/90 has 50% spanning 1 & 50% spans 3
the 45/45 100% spans approx 1.4

Just surprised that the quad lands so low and that the 0/90 didn't rate much either on the aspect of panel.

It's a shame that the test method and panels weren't described in some greater detail.... apparently the hydromat is pretty valid over small test coupons in testing especially when 45/45 fiber runs out over coupon edges.

Jeff

 

Not surprising at all. !0% is a general rule. It can get as low as 9% up to 15% depending on the formulation method (simplified and complex) and the type of fiber used.

Generally, pure UD without any stitches has the highest strength/modulus but degrades easily as the direction of load is rotated. Some do reach its lowest even before the 45 degree is reached.

Woven fabrics has lower strength/modulus but is more forgiving at transverse loading. See attached sheet for comparison. I am preparing a new sheet based on published data of 45 degree fabric. Sadly, not much is available.

Delete spreadsheet if not reading correctly. see second post.

 

Maybe I'm oversimplifying the scenario-
That the hydromat is testing against the pressure in the bladder and measuring deflection at panel center- could all plies(as each uni) in each fabric geometry be loaded at their best but at different lengths considering it's a panel(in this case 3:1 ratio represented to table) with support at all edges, that the pressure supplied be roughly at 90 degrees to the panel roughly approximating slamming/landing loads to hull or underwing as an example?

Jeff

 

Waikikin-

I am not very familiar with hydromax testing but I know as they claim is better than 3 point or 4 point load test as it closely approximates a uniformly loaded beam with fixed ends.

In any case, they are measuring the deflection which is a measure of how rigid the panel is. What doesn't show is how the individual layers will react at point of failure.

In calculating stresses, the outermost fibers are in compression/tension and surely, the ones with the lowest strength will fail. With only 10% strength remaining, it will fail. I have fed this into my ply by ply analysis spreadsheet and it shows the 45 and the 90's are failing.

Quad and triax are meant to be facings in sandwich laminate and must be symmetrical on opposite faces. For a single ply laminate, a single triax or quad will fail. It must be laid up in "face down", "face up" system. Meaning 0/45/-45, -45/45/0 or 0,45/-45/90, 90/-45/45/0.

By the way, is the graph in my spreadsheet showing correctly? I cut and pasted it from another spreadsheet and it showed the graph is reading from another spreadsheet. No problem in my computer as it is reading internally. In the web, it will not access the files in my computer.

 

Here is the corrected spreadsheet. It is now stand alone.

 

Hi RX,
Your graph & table showed up but heaps more info on the first... not that I'm spending too much on interpreting the info.

I may be entering some zone of semantics but would think/expect that the 90 degree fiber would break first then the 45s, but because due to it's orientation is spanning the shortest distance so proportionate to it's length is taking/resisting the greater part of the load and deflecting proportionately more over a shorter length, breaking earlier not due to being weaker but due to the greater deflection per length & subsequent mode of failure/buckling/whatever. Compare the scenario of the quad fabric without considering stacking sequence to a panel aspect of 3:1 thru some other aspect ratios say 2:1 and 1:1 to an equal octagon hydromat fixture, presumably the octagon would show little preference to the manufactured orientation of the fiber.
Would be interesting to see some hydomat testing to destruction!

Regards from Jeff

 

You are correct in your assumption. The short side (width) of the panel has a lower bending moment than the length. With the beam analysis deflection formula, the short side and the long side can be analyzed. It is the stacking sequence that is important.

 

Thanks for that RX. I am of the opinion that most of the loads on sailboats in particular are fairly low, so most of what holds true in lab testing does not really come into play. I have in front of me the laminate schedule for a 45ft fast cruising cat from a well known actual naval architect that is simply gelcoat, 3/4oz skin coat,0/90 w/mat,0/90 w/mat, 3/4 oz mat, 3/4" balsa, 3/4 oz mat, 0/90 w/mat,0/90 w/mat, VE resin. No +/- 45 at all. Ive seen others and in fact have built with all double bias and they all seem to work just fine. There are of course areas that carry higher loads but they are dealt with individually but overall the loads are not high so it comes down to designing a schedule that meets your weight goals, which is not that easy because when you are happy with the weight and strength the skins are still pretty thin with todays low profile fabrics. My sons 24ft sailboat which I designed and built in the eighties is only 1 ply of db120 each side of a 1/2" h80 core with ve and has proven to be more robust than it needs to be.

 

Boats are typical.
The actual loads are not generally the worst thing that sizes the boat.
Deflection of the hull can be more critical which can mean you have to up size the structure to the point where there is a lot of excess relative to the load.
As you said local loads can be taken care of with local reinforcement.
Damage or impact can be really dangerous in a thin skinned, designed to the last ounce boat. If we are talking keel boat or even just a centerboard, the impact from grounding is really hard to accurately predict but not an unusual occurrence.

All that can significantly increase the hull strength desire, and may be taken care of by experience.

None of this includes fatique or repetative cycling of loads, which composites are good at but not immune to failure.

There is a reason good NA/ engineers earn their money.