Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
1 336742
TITLE
Composites of Stretch Broken Aligned
Fibers of Carbon Reinforced Resin
Backqround of the Invention
This invention relates to a process for stretch
breaking carbon and glass filaments and using the stretch
broken slivers therefrom to form a composite of either a
matrix reinforced with carbon fibers or a matrix
reinforced with glass fibers.
Composite sheets of either continuous filament
carbon fiber reinforced resin or continuous filament
glass fiber reinforced resin have been made. One
technique is to prepare a warp of filaments as by winding
on a frame, impregnating them with resins and hot
pressing to form a thin flat sheet which is cut from the
frame. Several such sheets are then cross lapped and
again hot pressed to form the final reinforced composite
product. Such products have high strength and stiffness.
Problems occur when attempts are made to
produce deep drawn three dimensional articles by hot
pressing continuous carbon or glass filament containing
resin sheets. The articles in many instances exhibit
uneven areas and wrinkles. The use of staple carbon or
glass fibers as reinforcement substantially overcomes the
above-stated problems but at a great sacrifice to
strength and stiffness.
In a similar situation involving P-aramid
fibers, a solution to the aforementioned problem was the
use of certain stretch broken P-aramid fibers as
disclosed by Fish and Lauterbach in U.S. Patent No.
4,552,805. However, because carbon and glass filaments
exhibit little or not cohesive capability when processed
according to known stretch-breaking processes, slivers of
carbon or glass fibers have not been able to be formed by
these known processes.
1 336742
The present invention permits forming cohesive
~livers of stretch broken filaments of carbon and glass
for use in forming a composite carbon or glas6 fiber
reinforced re~in useful for deep drawing purposes with
little sacrifice of ~trength and stiffness.
Summary of the Invention
A cohesive filiver of stretch broken gla6s or
carbon fibers having a high degree of axial alignment
and a coating of a finish compri~ing a vi~cous lubricant
and an anti-static ingredient. Composite~ of a matrix
resin reinforced with such ~livers and ~haped ~tructures
formed therefrom are al~o encompassed.
Brief Description of the Drawing
Fig. 1 is a schematic illustration of a
preferred embodiment apparatus for use with a continuous
proce~s in the practice of the pre~ent invention.
Fig. 2 i5 a schematic illustration of apparatus
for applying finish to a carbon or gla~ filament yarn.
Fig. 3 i5 a schematic illustration of apparatus
for stretch-breaking a cohesive carbon or glass yarn.
Detailed Description of the Preferred Embodiment
Referring to Fig. 1, the preferred embodiment
generally includes a creel 10 for yarn eupply packages
12, a plurality of yarn tensioning bar~ generally
designated 14, a finish applicator 16 comprised of a
rotatable finish roll 18 emer~ed in a pan 20 filled with
a liquid finish 22 a pair of grooved roller guides 24,26
are located between the finish applicator 16 and a Turbo
Stapler 28 ~manufactured by the Turbo Machine Co.,
Lansdale, Pa.). The Turbo-Stapler include~ a pair of
driven nip rolls 30,32 which firmly grip the tow band 34
that has been consolidated from the individual yarns in
guide 29. The nip rolls 30,32 feed tow band 34 at a
constant rate to a pair of front rolls 36,38 which also
grip the tow band 34 and withdraw it from breaker bar~
39 and feed it as a sliver to a condensing guide 40 from
1 336742
which the sliver is fed to a windup (not shown) for
packaging.
In operation, glass or carbon yarn 13 from
individual packages 12 is fed from creel 10 over finish
roll 18 where it is coated with finish 22. The yarns
are con~olidated in guide 29, tensioned between rolls
30,32 and front rolls 36,38, then randomly broken by
6harply deflecting them laterally by the breaker bar6
39. The coating of finish on the yarn in the sliver i6
sufficient to enable the sliver to be pulled through
guide 40 to the windup without disassociation of the
fiber6 in the sliver.
While the continuous process illustrated in
Fig. 1 is preferred, the application of finish to
continuous filament carbon or glas6 fibers and the
~tretch-breaking of the coated filaments can be carried
out in two steps; i.e., separate finish application and
stretch-breaking processes, according to Fig~. 2 and 3
and as described ~ubsequently in Example 1. More
particularly, in Fig. 2, glass or carbon yarn 13 from
package 12 i5 fed over yarn tensioning bars 14' over
finish roll 18 where it is coated with fini~h 22 and
wound onto a bobbin 12' and allowed to dry. The yarn
from bobbins 12' is then stretch-broken by breaker bars
39 (Fig. 3) in the turbo-6tapler as describcd above in
connection with Fig. 1.
The finish used in this invention i6 a material
that causes an interfilament viscou6 drag ~ufficiently
high to permit the handling required to make a
composite, such as winding and unwinding from a package.
More particularly, the finish used for the carbon fiber
application is a m~xture of a one part of a ~uitable
anti~tat and two part~ of a non-tacky viscou~ lubricant
of a consistency to impart to the chopped sliver
adequate cohesiveness (minimum of .01 gram6 per denier)
without tackiness or without compromising the
1 33674~
fiber-matrix adhesion in the final composite. The
antistat portion of the mixture could be reduced or even
eliminated if the reinforcing fiber is electrically
conductive (e.g., carbon fibers).
A suitable viscous lubricant is polyethylene
glycol (400 mol wt) monolaurate and a lauric amide while
a suitable antistat is mixed mono and di-phosphate esters
of C8-C12 fatty alcohols neutralized with diethanol
amine.
Preferably, the percent finish on fiber is in
the range of from about 0.3% to about 0.5% by weight.
Formable planar and shaped non-planar
composites are contemplated by the present invention.
For the formable composites, that is, those composites
that can be formed into shaped non-planar three-
dimensional structures at elevated temperatures (where
necessary), matrix resins of the thermoplastic variety or
of the not fully cured thermoset type may be employed.
In the latter case the thermosettable resin is cured
after the composite has been shaped. Suitable
thermoplastic resins include polyesters (including
copolyesters), e.g., polyethylene terephthalate, Kodar~
PETG copolyester 6763 (Eastman Kodak); polyamides, e.g.,
nylon 6,6; polyolefins, e.g., polypropylene; also
included are the high temperature resins such as an
amorphous polyamide copolymer based upon bis (para-
aminocyclohexyl) methane, a semi-crystalline polyamide
homopolymer also based on bis(para-aminocyclohexyl)
methane, and polyetheretherketone. Thermosetting resins
that are useful include phenolic resins, epoxy resins and
vinyl ester resins.
The ratio of reinforcement to matrix can vary,
but preferably is between 40% to 75% by volume. The
average fiber lengths also may vary but preferably range
1 336742
from about 1/2 to about 6 inches in lenqth with a random
overlap distribution. About B5 percent of the fiber6
are aligned within +10 degrees, preferably +5 degrees of
the axial direction.
The composite may be made by a variety of
procedures. Thus, a 6tretch broken sliver may be wound
on a frame covered with a film of thermoplastic resin to
form a warp. The warp of stretch-brokcn sliver,
however, can be made by any technique known to those
skilled in the art, e.g., by creeling or beaming. A
preform is obtained when another film of thermopla6tic
resin is placed over the warp to form a sandwich which
is heated in a vacuum bag and then removed from the
frame. Several of such preforms may be stacked while
offset to provide multi-directionality and then the
~tack may be heated under pressure to form a compo6ite
rtructure.
Other techniques for applying matrix polymer
include sprinkling of powdered resin on the sliver warp
followed by heating to melt the resin, flowing liquid
resin over the 61iver warp, intermingl~ng thermopla~tic
fiber with the sliver warp and then heating to melt the
thermoplastic fiber thereby forming the matrix resin,
calendering the warp between layers of matrix film, etc.
Test Procedures
Composite Tensile
The composite tensile te6ts followed the
general procedure described in ASTM Te~t D 3039-76
entitled "Standard Test Method for Tensile Properties of
Fiber--Resin Composites. n
Short Beam Shear
The 6hort beam shear tests followed the general
procedure described in ASTM Method D 2344-76 entitled,
"Standard Test Method for Apparent Interlaminar Shear
Strength of Parallel Fiber Composites by Short Beam
1 336742
Method" with the following exception, the loading nose
was 1/16 inch radius instead of 1/8 i~ch.
Sliver Cohesion
The yarn to be tested for sliver cohesion was
S placed in the clamps of an ~nstron*ten6ile testing
machine set to a gauge length of 17 inches, a cros6head
speed of 10 inches per minute and a chart speed of 12
inches per minute. The crosshead was started to apply
tension to the sample and the maximum force in grams
indicated on the chart was recorded and divided by the
sliver denier to give the sliver cohesion.
Finish on Yarn
Finish on yarn i6 determined in a method
wherein weighed ~pecimens are extracted gravimetrically
with prescribed solvent(6) at room temperature, the
solvent containing dissolved fini6h and any other
materials which may wash off the specimens, is
transferred to a preweighed container and evaporated.
The extractable residue is weighed. Percentage
extractables based on extractable-free specimen weight
is calculated. Aerothane (l,l,l-trichloroethane) is
u~ed as the solvent for all finish material6 except
glycerine and methanol is used as the solvent for that
material.
High Temperature Tensile Drawing
The sample to be tested was placed in the
clamps of an Instron tensile testing machine get to a
particular gauge length and a crosshead speed depending
on the sample. A thermocouple was attached the surface
of the sample midway between the clamps and an 8 inch
long electrically heated cylindrical oven was placed
around the sample leaving a one inch space between the
bottom of the oven and the lower clamp. The open ends
of the oven were plugged with insulation material to
prevent convective heat loss and heating of the clamps.
* denotes trade mark
1 336742
The oven was turned on and the sample heated to reduce
its viscosity to permit drawing (temperature determined
by the viscosity, time, temperature data of the matrix
material. Samples made with thermosetting matrix resins
must be tested in their uncured state.). The sample was
held at this temperature for 15 minutes to ~nsure
thermal equilibrium. The crosshead was then started and
allowed to run until the heated section of the sample
was drawn 50%. The oven was removed and the sample
inspected to determine whether it had broken.
Fiber Orientation
A photomicrograph of the surface of the
composite (enlarged 240X) was prepared. The angle
between each fiber axis and the axial direction of the
composite was measured with a protractor on the
photomicrograph and tabulated. The percentage of fibers
with an angle within + 5 degrees of the axial direction
was reported.
Example 1
Four bobbins of 2000 denier continuous
filament carbon fiber (3K AS-4 from Hercules ~nc.) were
prepared for stretch-breaking by applying a finish
composed of two parts of a lubricant ~polyethylene
glycol monolaurate and a lauric amide) and one part of
an antistat (mixed mono and diphosphate esters of C8-C12
fatty alcohols neutralized with diethanol amine). The
finish was applied by running the continuous filament
carbon fiber, one bobbin at a time, at 75 yards/minute
over a finish roll which was wet with a 4% aqueous
emulsion of the lubricant-antistat mixture (Fig 2). The
four bobbins were allowed to stand overnight to
evaporate the water. Finish level after drying was
0.33%.
The four bobbins of carbon fiber were
stretch-broken on a Turbo-stapler (Turbo Machine Co.,
B 1 336742
Lansdale, PA) as ~hown in Fig 3. The 6urface 6peed of
the roll~ ~30,32) wa~ 3S.4 yards/minute and the turface
6peed of the front roll~ (36,38) wa~ 110 yards/minute.
The tip speed of the breaker bars (39) was 71
yards/minute. The resulting sliver wa~ 2422 denier and
had a cohesion value of 0.18 grams/denier which was
6ufficient to allow winding without twi6t on a
cylindrical paper tube u~ing a Leesona*type 959 winder.
The average fiber length of fifty measurement6 of this
61iver was 3.2 inches (shortest 0.7 inch, longest 5.6
inches).
A warp was prepared from this sliver by winding
it from the paper tube, 25 ends to the-inch on a 16 inch
square metal plate. A 2.0 mil thick film of
lS thermoplastic re6in (an amorphous polyamide copolymer
based on bis~para-aminocyclohexyl) methane) was placed
on the frame before winding the 61iver and another was
added after winding was complete. The entire sandwich
was vacuum bagged at 280-C for lS minute~ after which
time it was cut from the plate. Thi6 product, called a
preform was now a well-impregnated, relatively stiff
matrix/6tretch-broken 61iver 6andwich, in which all the
61ivers were aligned in one direction.
Twelve of the~e preform6 were 6tacked on top of
2S one another with all the fibers in the same direction.
Thi6 6tack wa6 heated ~n a mold at 305C at 500 pounds
per square inch for 3S minute6 to make a
well-consolidated plate 93 mil6 thick and fiber volume
fraction of SS%. Short beam shear test6 conducted on
O.S inch wide strip6 cut from this plate gave a value of
13,700 pounds per ~quare inch. It wa~ concluded that
the presence of the finish did not adver~ely affect the
adhesion of the fiber to the matrix polymer.
A 6econd plate was made from ten of the6e
3S preforms by stackinq them so that the direction of the
* denotes trade ~ark
1 336742
stretch-broken fibers were offset by 45 degrees in a
clockwise direction in successive layers. The bottom
plane of the fifth layer was considered a reflecting
plane and the next five layers were stacked 60 that the
warp directions of the stretch-broken sliver were mirror
images of the top five layers with respect to that
plane. This sandwich was molded as above to make a well
consolidated plate with a fiber volume fraction of 55%.
This plate was heated to 322C and molded into a
hemisphere with a radius of 3 inches. The plate
conformed very well to the shape of the mold and it was
concluded that the product was deep drawable without
wrinkles.
Example 2
A sliver of stretch-broken glass fiber was
prepared by the method in Example 1 except that 6700
denier continuous filament glass fiber was u6ed (T-30
P353B from Owens-Corning Fiberglass) and the finish was
applied by spraying a 10% aqueous emulsion on the fiber.
The emulsion was pumped to the spray nozzle at 5 cc. per
minute and the air pres6ure used was 3 psi. The yarn
was pulled past the spray head at 55 yards per minute by
a pair of nip rolls and wound on a cylindrical paper
tube. After drying, the finish level was 0.35%.
Stretch-broken sliver was prepared from two finish
treated continuous filament bobbins and had a cohesion
of 0.09 grams per denier which was adequate for winding
as in Example 1. Further, the finish controlled static
generation in the stretch-breaking process to an
acceptable level. The average fiber length of fifty
measurements of this sliver was ~.4 inches (6hortest 1.0
inch, longest 10.2 inches).
A unidirectional plate was made from this
sliver and PETG film (Kodar PETG copolyester 6763,
Eastman Kodak) by the method of Example 1 cxcept that
the sliver spacing was 26 ends per inch, the film
1 336742
thickness was 3.0 mils and 8 layers of preform were used
to 55% fiber volume fraction. Short beam shear tests on
0.5 inch wide strips cut from this plate gave a result
of 5,400 pounds per square inch. It wa6 concluded that
the presence of the finish did not affect the adhesion
of the fiber to the matrix polymer.
Example 3
A sample of carbon fiber sliver was prepared
using the stretch-breaking process of Example 1 except
that finish was not pre-applied to the continuous fiber
and two bobbins were used instead of four. The two ends
of carbon fiber were contacted by a felt pad saturated
with glycerine which was placed between thc tension
guide and the infeed roll. Glycerine level on the
sliver was 0.5%. The average fiber length of fifty
measurements of this sliver was 3.2 inches (shortest 0.6
inch, longest 7.9 inch). Cohesion was measured as a
function of time vs. the sliver from Example 1 with the
following results.
Cohesion, grams per denier
Days Glycerine Example 1
1 .58 .15
9 .79 .24
16 .02 .25
22 .02 .25
.02 .21
Example 4
Glycerine treated sliver from Example 3 was
made into a warp, preforms and a unidirectional plate by
the method of Example 1. The end count was 12 per inch,
the film was 3.0 mil thick PETG (Rodar PETG copolyester
6763 from Eastman Rodak) and 6 preforms were stacked to
make the plate which was 40% fiber volume fraction. The
plate was cut into 0.5 inch strip6, provided with
aluminum tabs and subjected to tensile te6ts at 8 inch
guage length with the following results:
11 1 336742
Tensile strength, psi. 127,400
Modulus, psi. 11,600,000
It was concluded that the product had very high strength
and modulus. The uniformity of orientation of the
fibers on the surface of thi~ plate were measured from a
photomicrograph and it was found that 85% of the fibers
were within + 5 degrees of the axial direction.
Example 5
Continuous filament 2000 denier carbon fiber
was made into a warp, preforms and a unidirectional
plate. The end count was 12 per inch, the film was 3.0
mil thick PETG (Kodar PETG copolye~ter 6763 from
Eastman Rodak) and 16 preforms were stacked to make the
plate which was 40% fiber volume fraction. The plate
was cut into 0.5 inch strips, provided with aluminum
tab~ and subjected to tensile tests at 8 inch guage
length with the following results:
Tensile strength, psi. 139,800
Modulus, psi. 11,600,000
It was concluded that the product of Example 4 exhibited
the strength and stiffness expected of continuous
filament carbon fiber. The product of Example 4,
although made of stretch-broken di6continuou~ staple
fiber, came within 90% of the ~trength and 6tiffness of
the continuous filament product. This excellent
performance i~ believed due to the high degree of order
of the ~tretch-broken fibers.
Example 6
Stretch broken glass sliver was prepared by the
method of Example 2 except that finish was not
pre-applied to the continuous fiber. Instead, the fiber
being 6upplied to the Turbo-stapler was sprayed
periodically with Jif-Job*antistatic spray (Schafco,
~ancaster, PA). The roll and breaker bar speeds were
one-half the values in Example 2. The average fiber
* denotes trade mark
1~ 1 33674~
length of fifty measurements of this sliver was 3.1
inches (shortest 1.0 inch, longest 5.8 $nch). This
sliver was made into a warp, preforms and a
unidirectional plate by the method of Example 1. The
end count was 21 per inch, the film was 3.0 mil thick
PETG (Rodar PETG copolyester 6763 from Eastman Rodak)
and 5 preforms were ~tacked to make the plate which was
40% fiber volume fraction. The plate was cut into 0.5
inch strips, provided with aluminum tabs and subjected
to tensile tests at 8 inch guage length with the
following results:
Tensile strength, psi. 67,200
Modulus, psi. 4,950,000
It was concluded that the product had very high strength
and modulus.
Example 7
Continuous filament 6700 denier glass fiber was
made into a warp, preforms and a unidirectional plate.
The end count was 13 per inch, the film was 3.0 mils
thick PETG (Rodar0 PETG copolyester 6763 from Eastman
Rodak) and 5 preforms were stacked to make the plate
which was 40% fiber volume fraction. The plate was cut
into 0.5 inch strips, provided with aluminum tabs and
subjected to tensile tests at 8 inch guage length with
the following results:
Tensile strength, psi. 67,900
Modulus, p8i. 5,460,000
It was concluded that the product of Example 6 exhibited
the ~trength and stiffness expected of continuous
filament glass fiber. The product of Example 6,
although made of discontinuous staple fiber, came within
90% of the strength and stiffness of the continuou~
filament product.
1 336742
13
Example 8
A preform of stretch broken carbon fiber 61iver
in an epoxy resin ~Hercules 3501-6) was made by the
following procedure:
1) The frozen resin was thawed at room
tcmperature, then heated to 180F for 15 minutes.
2) A film of resin was cast onto release paper
then chilled to 40F to stop the polymerlzation reaction
and the exposed surface was covered`with polyester film
for protection.
3) The paper-resin-film ~andwich was wound on
a 7-foot diameter drum and the polyester film removed.
4) 2300 denier graphite sliver made by the
process of Example 1 was wound on the exposed resin at 9
ends per inch for a total width of 10.5 inches. The
average fiber length of fifty measurements of thi~
~liver was 3.2 inches (shortest 0.7 inch, longest 5.6
inches).
5) The polyester film was removed from a
second paper-resin-film sandwich and wound over the
graphite layer on the drum to make a
paper-resin-graphite-resin-paper sandwich.
6) The sandwich was unwound from the drum and
vacuum bagged flat at 140F for 10 minutes to force the
resin into the graphite layer, then frozen for later
use. The thickness of the resin-graphite part of this
sandwich was 7 mils.
A unidirectional compo6ite strip made by
~tacking together ten layers of 3/4-inch wide and
14-inch long strips (fiber direction parallel to the
14-inch dimension) of the graphite-resin preform was
vacuum bagged for two minutes. One inch on either end
of the strip was partially cured by heating it to 120C
for two hours while keeping the middle 12 inches of the
strip cold with dry ice. At a guage length of 11 inches
1 336742
14
and a cros6head speed of 5 inches per minute, a high
temperature tensile drawing test was conducted at 124C
on the 14 inch long by 0.75 inch wide 6trip which 6howed
the composite could be drawn 50% without breaking,
predicting a high degree of formability.
A composite plate was made from 10 layers of
the 6andwich from step 6 above by removing the release
paper, cutting the graphite-resin preform into sheet6
and stacking them so that the direction of the
stretch-broken fibers were offset by 45 degrees in a
clockwise direction in successive layers. The bottom
plane of the fifth layer was considered a reflecting
plane and the next five layers were stacked 60 that the
warp directions of the stretch-broken sliver were mirror
images of the top five layers with respect to that
plane. This sandwich was vacuum-bagged at ambient
temperature for 2 minutes to stick the layers together.
This plate was molded into a hemispherc with a radius of
3 inches and cured in the mold at 175C for 2 hours.
The plate conformed very well to the shape of the mold
and it was concluded that the product was formable.
Example 9
Four bobbins of 2000 denier continuous filament
carbon fiber (3K AS-4~from Hercules ~nc.) were
stretch-broken on a Turbo-stapler (Turbo Machine Co.,
Lansdale, PA) set up as shown in Fig 1. A 10% aqueous
solution of the finish described in Example 1 was
applied with a wetted roll. The surface speed of the
intermediate rolls was 17.7 yards/minute and the surface
speed of the front rolls was 55 yards/minute. The tip
speed of the breaker bars was 35.5 yards/minute. The
resulting sliver was 2250 denier. The average fiber
length of fifty measurements of this sliver was 3.3
inches (shortest O.B inch, longest 5.5 inches).
1 336742
A warp was prepared from this sliver by winding
it, 27 ends to the inch ~n a 18 inch square metal plate.
A 3.0 mil thick film of thermoplastic resin (PETG
copolyester) was placed on the frame before
winding the sliver and another was added after winding
was complete. The entire 6andwich was vacuum bagged at
220C for 15 minutes after which time it was cut from
the frame. This product, called a preform was now a
well-impregnated, relatively stiff matrix/stretch-broken
sliver sandwich, in which all the sliver6 were aligned
in one direction.
Seven of these preforms were stacked on top of
one another with all the fibers in the 6ame direction.
This stack was heated in a mold at 200C at 400 pounds
per square inch for 30 minutes to make a
well-consolidated plate 82 mils thick and fiber volume
fraction of 50%. High temperature tensile drawing tests
at a guage length of 10 inches and cros6head speed of 10
inches per minute conducted at 262C on 12 inch long by
0.75 inch wide strips cut from this plate with the fiber
direction parallel to the 12 inch dimension 6howed the
composite could be drawn 50% without breaking,
predicting a high degree of formability.
Example 10
Two bobbins of 6700 denier continuous filament
glass fiber ~T-30 P353B from Owens-Corning Fiberglass)
were stretch-broken on a Turbo-stapler (Turbo Machine
Co~, Lansdale, PA) set up as 6hown in Fig 1. A 10%
aqueous 601ution of the finish described in Example
1 was applied with a wetted roll. The surface ~peed of
the intermediate rolls was 17.7 yards/minute and the
surface speed of the front rolls was 55 yards/minute.
The tip speed of the breaker bars was 35.5 yards/minute.
The resulting sliver was 4100 denier. The average fiber
length of fifty measurements of this sliver was 3.4
inches (shortest 0.9 inch, longest 8.7 inches).
1 336742
16
A warp was prepared from this 61iver by winding
it, 22 ends to the inch pn a 18 inch square metal plate.
A 3.0 mil thick film of thermoplastic resin (PETG
copolyester) was placed on the frame before winding the
sliver and another was added after winding was complete.
The entire ~andwich was vacuum bagged at 220C for 15
minutes after which time it was cut from the frame.
This product, called a preform was now a
well-impregnated, relatively stiff matrix/stretch-broken
61iver ~andwich, in which all the sliver6 were aligned
in one direction.
Seven of these preforms were stacked on top of
one another with all the fibers in the same direction.
This stack was heated in a mold at 200C at 400 pounds
per 6quare inch for 30 minutes to make a
well-consolidated plate ~2 mils thick and fiber volume
fraction of 50%. High temperature tensile drawing test6
at a guage length of 10 inches and cros6head ~peed of 10
inches per minute conducted at 262C on 12 inch long by
0.75 inch wide strips cut from this plate with the fiber
direction parallel to the 12 inch dimension showed the
composite could be drawn 50% without breaking,
predicting a high degree of formability.
Example 11
Sliver from Example 10 was re-broken to reduce
the fiber length by passing it through two set6 of
elastomer coated nip rolls with a separation of 2.50
inches between the nips. The surface 6peed of the
second set of rolls was 10 yards per minute and the
~urface 6peed of the first set of roll~ was 7.1 yards
per minute giving a draft of 1.4. Denier of this
re-broken sliver was 5371 and the average fiber length
of fifty measurements of this sliver was 1.57 inches
(shorte6t 0.5 inch, longest 3.6 inches~.
1 336742
17
A ~warp~ was prepared from this sliver by
winding it , 17 ends to the inch on a 18 inch square
metal plate. A 3.0 mil thick film of thermoplastic
resin (PETG copolyester) was placed on the frame before
winding the sliver and another wa~ added after winding
was complete. The entire sandwich was vacuum bagged at
220C for 15 minutes after which time it was cut from
the frame. This product, called a preform was now a
well-impregnated, relatively stiff matrix/6tretch-broken
sliver sandwich, in which all the slivers were aligned
in one direction.
Seven of these preforms were stacked on top of
one another with all the fibers in the ~ame direction.
This stack was heated in a mold at 200C at 400 pounds
per square inch for 30 minutes to make a
well-consolidated plate 80 mils thick and fiber volume
fraction of 50%. High temperature tensile drawing tests
at a guage length of 10 inches and cro~shead speed of 10
inches per minute conducted at 262C on 12 inch long by
0.75 inch wide strips cut from this plate with the fiber
direction parallel to the 12 inch dimension showed the
composite could be drawn 50% without breaking,
predicting a high degree of formability.
Example 12
Sliver from Example 9 was re-broken to reduce
the fiber length by passing it through two set6 of
elastomer coated nip rolls with a separation of 2.50
inches between the nips. ~he surface speed of the
second set of rolls was 10 yards per minute and the
surface speed of the first set of roll6 was 7.1 yards
per minute giving a draft of 1.4. Denier of this
re-broken sliver was 4623 and the average fiber length
of fifty measurements of this sliver was 1.33 inches
~6hortest 0.6 inch, longest 3.1 inches).
1 336742
18
A warp was prepared from thi6 61iver by winding
it, 13 ends to the inch on an 18 inch 6quare metal
plate. A 3.0 mil thick film of thermoplastic resin
(PETG copolye6ter) was placed on the frame before
winding the 61iver and another was added after winding
was complete. The entire sandwich wa6 vacuum bagged at
220C for 15 minutes after which time it wa6 cut from
the frame. This product, called a preform was now a
well-impregnated, relatively stiff matrix/~tretch-broken
sliver sandwich, in which all the sliver6 were aligned
in one direction.
Seven of these preforms were 6tacked on top of
one another with all the fiber6 in the same direction.
This 6tack was heated in a mold at 200C at 400 pounds
per 6quare inch for 30 minutes to make a
well-consolidated plate 80 mils thick and fiber volume
fraction of 50%. High temperature ten6ile drawing
test6, at a guage length of 10 inches and a cro66head
speed of 10 inches per minute, conducted, at 262C, on
12 inch long by 0.75 inch wide strip6 cut from thi6
plate with the fiber direction parallel to the 12 inch
dimension 6howed the composite could be drawn 50%
without breaking, predicting a high degree of
formability.
Example 13
A pre-laminate was prepared from gla6s fiber
from Example 2 by a continuou6 proce66 as follows:
46 ends of sliver were fed from a creel into a 6 inch
wide warp which was 6andwiched between two 1.0 mil PET
poly(ethylene terephthalate) films to give a
pre-laminate of 55% fiber volume fraction. A release
film of Rapton* polyimide was placed on each 6ide of
thi6 sandwich to prevent 6ticking of molten PET to hot
6urface6. This sandwich was then passed at 10 feet per
minute through the nip of two steel roll6 heated to
278C to tack the assembly together.
* denotes trade mark
1 336742
19
A composite plate was made from this
pre-laminate by removing the release film, trimming the
excess PET from the edges and placing strips of
pre-laminate in layers in a 16 inch 6quare mold. Each
5 layer was made up of side-by side strip6 of pre-laminate
to reach the required 16 inch width.
A plate was made from 10 layer6 of pre-laminate
by arranging them 80 that the direction of the
stretch-broken fiber~ were offset by 45 degrees in a
10 clockwise direction in successive layers. The
bottom plane of the fifth layer wa6 considered a
reflecting plane and the next five layers were 6tacked
~o that the warp directions of the 6tretch-broken sliver
were mirror images of the top five layers with respect
15 to that plane. This 6andwich was molded as ln Ex~mple 2
to make a well-consolidated composite plate with a fiber
volume fraction of 55%. Thi6 plate was heated to 280C
and molded into a hemisphere with a radiu6 of 3 inches.
The plate conformed very well to the ~hape of the mold
20 and it was concluded that the product wa6 formable.
Example 14
A plate was made from 10 layers of pre-forms
made by the method of Example 11 by arranging them in a
16 inch square mold 60 that the direction of the
25 ~tretch-broken fibers were offset by 45 degree6 $n a
clockwi6e direction in 6uccessive layer6. The bottom
plane of the fifth layer was considered a reflecting
plane and the next five layers were 6tacked so that the
warp directions of the 6tretch-broken 61iver were mirror
30 images of the top five layers with respect to that
plane. This 6andwich was molded as in Example 2 to make
a well-consolidated composite plate with a fiber volume
fraction of 55%. This plate was heated to 280C and
molded into a hemisphere with a radius of 3 inches. The
35 plate conformed very well to the shape of the mold and
it was concluded that the product was formable.
1 336742
Example 15
Continuous filament 2000 denier carbon fiber
was made into a warp, preforms and a unidirectional
plate by the method of Example 1. The cnd count was 25
per inch, the film was 2.0 mil thick film of
thermoplastic resin (an amorphous polyamide copolymer
based on bis(para-aminocyclohexl) methane). Seven
preforms were stacked to make the plate which was 55
mils thick and 55% fiber volume fraction. The plate was
cut into 0.5 inch strips, provided with aluminum tabs
and subjected to tensile tests at 8 inch gauge length
with the following results:
Tensile strength, psi. 243,200
Modulus, psi. 18,200,000
It was concluded that the product had very high strength
and modulus.
Example 16
A warp was prepared from sliver from example 9
by winding it, 21 ends to the inch on a 18 $nch square
metal plate. A 2.0 mil thick film of thermoplastic
resin (an amorphous polyamide copolymer based on
bis(para-aminocyclohexl) methane) was placed on the
frame before winding the sliver and another was added
after winding was complete. The entire ~andwich was
vacuum bagged at 280C for 20 minutes after which time
$t was cut from the frame. This product, called a
preform was now a well-impregnated, relatively stiff
matrix/stretch-broken sliver sandwich, in which all the
slivers were aligned in one direction.
Seven of these preforms were stacked on top of
one another with all the fibers in the same direction.
This stack was heated in a mold at 305C at 600 pounds
per 6quare inch for 40 minutes to make a
well-consolidated plate 5B mils thick and fiber volume
fraction of 55%. One half inch strips cut from this
1 336742
21
plate were subjected to tensile tests at 8 inch gauge
length with the following result6:
Tensile strength, psi 246,000
Modulus, psi 18,800,000
The uniformity of orientation of the fibers on the
surface of this plate were measured from a
photomicrograph and it was found that 92% of the fibers
were within + 5 degrees of the axial direction. The
product of this example, although made of discontinuous
staple fiber, was equivalent to the strength and modulus
of continuous filament fiber (Example 15).
Example 17
Continuous filament 6700 denier glass fiber was
made into a warp, preforms and a unidirectional plate by
the method of Example 1. The end count was 15.5 per
inch, the film was 3.0 mil thick PET (poly(ethylene
terephthalate)) and 5 preforms were stacked to make the
plate which was 55% fiber volume fraction. The plate
was cut into 0.5 inch strips, provided with aluminum
tabs and subjected to tensile tests at 8 inch gauge
length with the following results:
Tensile strength, psi. 156,000
Modulus, psi. 7,300,000
It was concluded that the product of Example 17
exhibited the strength and stiffnes6 expected of
continuous filament glass fiber.
Example 18
A unidirectional plate was made from
pre-laminate from Example 13 by stacking 5 layers in a
mold with all slivers in the same direction and heating
in a press as in the reference example to give a final
thickness of 103 mils. One-half inch strips cut from
this plate were subjected to tensile tests at 8 inch
gauge length with the following results:
Tensile Strength, psi 86,800
Modulus, psi 5,900,000
1 336742
0 22
It was concluded that strength and modulus of
the product of this example, although not as high as
those from continuous filament glass (Example 17) were
far superior to those of randomly oriented glass
composites of equivalent fiber volume fraction reported
in the literature (ref. B.D. Agawarl, L.J. Broutman,
"Analysis and Performance of Fiber Composites" p. 92)
which are:
Tensile Strength, psi 23,000
Modulus, psi 2,400,000
This is a divisional application based on
Canadian Serial No. 616,305 filed February 5, 1992, which
was a division of Canadian Serial No. 554,032 filed
December 10, 1987.
