From: "Scott Dahlgren" <dahlgren@itsnet.com>
To: "___Lancair list" <lancair.list@olsusa.com>
Subject: fiberglass & carbon fiber
Date: Fri, 25 Sep 1998 10:35:42 -0600
Message-ID: <000f01bde8a2$898b5dc0$275b29c0@scottdah>
Importance: Normal
Mime-Version: 1.0

          <<<<<<<<<<<<<<<<--->>>>>>>>>>>>>>>>
          <<  Lancair Builders' Mail List  >>
          <<<<<<<<<<<<<<<<--->>>>>>>>>>>>>>>> 
>>
John Cooper,

here are the answers to some of your questions:

>"But what about the Young's Modulus of the spruce longerons (E=1,500,000)
>vs. the Young's Modulus of the fiberglass fuselage (E=10,600,000), hmmm?
>This is a much bigger ratio than between glass-epoxy and carbon fiber
>(E=35,000,000). Shouldn't the "stiffer" glass-epoxy fuselage break before
>the spruce longerons in the standard design?"

The answer is that the spruce (or foam, or honeycomb) are the core materials
in composite construction and are designed to take no tension or compression
parallel to the major load, or loads acting along the major axis.  The core
material is deliberately matched to be a lower modulus to allow the outer
material (glass, or carbon) to take all loads along the major axis, while
increasing the beam thickness or stiffness.  The core materials feel only
the shear, tension, and compression loads that act perpendicular to the load
axis, which are very small. That's why you can use foam as a structural
material. It allows composite construction to be both light and stiff.

>"And wouldn't the epoxy (E=2500) joining the two layers fail long before
>either the glass-epoxy or the carbon fiber failed?"

The epoxy, spruce and other core materials never fail in major axis shear
because in comparison to the carbon or glass their modulus is too low to
create any load stress within the core under strain. The stresses along the
top of the longeron along the major (load bearing) axis for .003" of flexing
strain are:

material                        modulus              stress
glass  - 45 deg. lay-up         3.1 MSI             9,300 psi
carbon - uni lay-up            19.0 MSI            60,000 psi
epoxy                              2500                 9 psi

(lay-up configuration governs the lay-up's modulus, which is lower than the
pristine filament values you listed). You can see in this example that the
glass holds very little load. At the transition point (where the carbon
ends) for a 60,000 psi load to continue into the glass (and it must) would
require .050" strain in the glass fibers. The transition is not
instantaneous so there is strain transition that breaks the part, just like
the sheet/patch example in my last post .

When an extra layer of carbon is added atop the glass, it will load along
the major axis along with  fiberglass with equal strain.  The loads the
epoxy are holding are perpendicular to the load axis, not parallel, and will
hold the glass and carbon together (major axis) so that they can break each
other, not delaminate. This is what allows weak core materials to hold two
layers of stiff material together allowing them to create strength and
stiffness along the load/major axis.

>"Yes, the worst that will happen to my plane is that cracks develop where
>the epoxy holding the carbon fiber to the pre-preg fails in shear, in which
>case I am back to the "standard" strength and will have to repaint my
plane."

Hoping for a bad lay-up is not a good bet. Uni-carbon fiber is bonded
directly to fiberglass shear webs (cores) regularly to make spars for
aircraft.  When bonded correctly they do not delaminate in shear when loaded
past their limit. They break.

A common cause of structural failure and weakening, as reported by the EAA,
is due to beefing up, not messing up. I think it is safest to stick to the
plans when it comes to complex structural applications such as monocoque
construction, unless you can do the analysis.   Martin Hollmann's "Composite
Aircraft Design" has an excellent section on modeling strain and subsequent
stress using maximum strain failure theory, broken down into simple matrix
equations. He also has this compiled as a computer program. If you read this
you will see some good examples of what I am talking about.

hope this helps..

Scott Dahlgren