Read sampe_paper.doc text version

Novel Processing Technique for Construction of High Strength Composite Joints

Henry K. Obermeyer Obermeyer Hydro, Inc. 303 West County Rd. 74, Wellington, CO 80549 Eric N. Gilbert GCC Services, LLC 504 S. Grant Ave., Fort Collins, CO 80521

ABSTRACT

Established methods for design and construction of steel reinforced concrete structures universally utilize continuous reinforcement through all tensile load carrying joints. This stands in sharp contrast to current practice in the field of fiber reinforced composites, which relies almost exclusively on adhesively bonded joints or discrete fasteners. These bonding techniques entirely lack fiber continuity between primary fiber reinforced components and result in inherently low joint efficiency. Techniques for providing controlled zones of non-resin- infused "dry fiber", for incorporation into fully reinforced secondary operation joints, have been investigated. Load paths, applications and techniques for both prepreg structures and resin transfer molded structures are outlined.

KEYWORDS: Adhesive bonding, Joint design, Joining

1.

INTRODUCTION

Efficient manufacturing practice dictates that complex fiber reinforced composite structures be produced in several stages and incorporate multiple prefabricated components. The efficiency and reliability of the joints between stages of construction is critical to the acceptance of fiber reinforced structures and to the success and growth of the composites industry. Simple prefabricated composite components are less expensive than complex components because of lower tooling costs, less onerous equipment sizes and costs, lower scrap risk and more numerous manufacturing method options. Most adhesively bonded joints in fiber reinforced composites lack any continuity of reinforcement, with the notable exception of Z-Pin technology. To civil engineers, who use concrete to carry shear and compression loads and steel to carry all tensile loads, especially those across joints, the lack of tensile member continuity through joints in fiber reinforced composites is troubling. Accordingly, a method has been identified where fiber reinforced composites might benefit from continuity of reinforcement through adhesively bonded joints or between portions of a structure. This novel joining technique also lends itself to situations where resin infusion must occur in stages, or where multiple resin systems with discrete boundaries might be desirable within a single structure.

Figure 1. Continuity of tensile members through joints according to civil engineering practice. Continuity of fiber reinforcement across joints is desirable because the fiber reinforcement is far stronger than the polymer matrix and is stronger yet compared to adhesively bonded joints within a polymer matrix. Polymer joints lacking fiber reinforcement can be expected to lack the inherent fatigue resistance of fiber reinforced structures. Additionally, adhesively bonded joints between polymer structures are subject to environmental degradation. Cured resin systems of composite components leave relatively few sites for chemical bonding to occur when the article is later joined in a secondary adhesive bonding step. These limitations of conventional adhesively bonded joints in fiber reinforced composites have been well documented [1,2,3,4]. The limitations of conventional adhesively bonded joints in fiber reinforced composites are well known [1,2,3,4]. US Patent 5264059 to Jacaruso et al. discloses partial embedment of reinforcing fabric in thermoplastic materials for subsequent connection of thermoset composite structures, but without fiber continuity through the completed thermoset/thermoplastic/thermoset joints suggested therein. The various processes for increasing surface energy and availability of potential bond sites are labor intensive, expensive, of dubious reliability, and are subject to reversal by brief environmental exposure. Figure 1 demonstrates a joint structure consisting of rebar dowels protruding from preceding stages of concrete. In this work, novel composite joining techniques have been identified that leave non- infused exposed fiber at component surfaces to be subsequently joined. As a result of this technique the resin boundary at the dry fiber interface is precisely controlled. The type of system which was selected for experimentation was based on the use of thermal gradients to precisely control the infusion of resin blocking substance. In the selection of candidate resin blocking substances, preference was given to environmentally benign materials and required solvents. Other manufacturing techniques were considered for the creation of the controlled dry fiber area such as the establishment of cold zones and the use of microwave, UV, or electron beam curing of resin dams. However, these methods were dropped for a variety of reasons including ease of application during manufacturing and limitations of part geometry.

2. EXPERIMENTAL

The creation of a dry fiber area in a cured composite designed for high strength adhesive bonding is accomplished in a two step process. The fiber rich area designated for bonding is first infused with a thermoplastic that will not react with the matrix during cure. After cure, the thermoplastic is then washed off the part leaving the fiber exposed for bonding. The braided glass fiber cord shown in figure 2 was initially infused with paraffin wax to ascertain the feasibility of using a thermal gradient to control the matrix blocking substance. Samples of partially paraffin infused fiberglass cord were then infused with a standard room temperature curing epoxy. After the epoxy was cured the samples were heated and the paraffin was removed by evaporation with the aid of a vacuum. Some waxy residue remained and was removed with a commercially available and non-toxic wax removing agent, "Fit® ". Optical micrographs, shown in figure 2, confirmed that a distinct and sharp resin boundary had been achieved. Paraffin could only be used for low temp curing systems because its low melting range reduces maximum allowable cure temperatures. As a result alternative resin blocking substances that would not prevent the use of elevated temperature cures were sought. Paraffin may still prove to be a useful choice for the blockage of low temperature cure phenolic resin systems, where water solubility may prevent the use of PVA based materials [5].

a) Figure 2:

b) Paraffin blocked/epoxy infused glass fiber cord

Various aqueous solutions including gelatin, pectin, and a novel PVA (polyvinyl alcohol) compound were next selected for experimentation. The advantage of the aqueous solutions over paraffin was that the effective melting points of these substances could be reduced by increasing water content. The first experiment involving a water soluble thermoplastic for fiber isolation was simply performed by applying five minute epoxy to the knitted nylon liner of a commercially available PVA glove used for handling non-polar solvents. The nylon fabric wetted readily with the epoxy. After the epoxy cured, hot water was used to place the PVA compound in solution. This left a distinct resin boundary with non-resin infused fibers protruding.

The PVA compound found in the gloves mentioned above was extracted for analysis. FTIR analyses of the proprietary PVA glove compound indicated the presence of approximately 4% glycerin (a plasticizer for PVA) and approximately 4% paraffinic hydrocarbons. Additionally, the aforementioned compound incorporated a bright orange pigment which proved to be extremely useful for the purpose of verifying complete coverage of the surface with PVA, and verification of complete removal by solution of the PVA compound after the resin was cured. In order to test various aqueous mixtures of fiber-isolation compound s in conjunction with structural fabrics, a length of 3 inch diameter copper pipe was used as a heat sink over which the fabrics were tightly wrapped. The solutions were then applied to various fabrics with a paint brush. The fabrics selected for experimentation were a 190 gram per square meter square woven carbon fabric, a 1127 gram per square meter square woven glass fabric and a Kevlar® fabric. The Kevlar® fabric was abandoned after PVA application due to non- uniform penetration. The gelatin treated fabric was brittle and porous after drying and was abandoned. The pectin proved brittle and porous after drying and discolored during a subsequent 134 deg. C autoclave cure. As a result, further work with this fiber isolating compound was abandoned. The PVA solution used was approximately 20% solids by weight and was applied at a temperature of approximately 66 deg. C. The copper pipe mandrel was maintained at a temperature of approximately 10 deg. C, which was above the ambient dew point of approx. 4 deg. C. After application of the PVA, the coated fabric was allowed to cool to room temperature and partially dry, after which it was oven dried at 66 deg. C. for 24 hours. Dry weight of the carbon fiber plus PVA was 347 g/sq. meter. Dry weight of the glass fiber plus PVA was 1330 g/sq.meter.

Figure 3: Carbon fiber (from left to right) untreated, PVA infused surface, opposite surface.

Figure 4: Glass fiber (from left to right) untreated, PVA infused surface, opposite Surface.

The above PVA blocked fabric samples were incorporated into several laminates including use as a surface layer of a carbon pre-preg fabric laminate with an autoclave cure at 134 deg. C. and as the surface layer of a wet lay up epoxy-carbon laminate with a 66 deg. C oven cure. Figure 5 provides a general illustration of the process sequence.

Figure 5:

Process sequence and relative temperatures.

3. RESULTS AND DISCUSSION

The laminates included an ordinary peal ply on one surface and a layer of PVA partial-thicknessinfused carbon fiber cloth at the opposite surface for the purpose of subsequent bonding. In all cases the PVA and resin formed continuous relatively planar boundaries which allowed removal of the PVA blocking compound with hot water (approx. 80 deg. C) after laminate curing. After removal of the PVA compound, the laminate specimens were oven dried at 66 deg. C for 6 hours. Details of the exposed non-resin infused carbon fibers are shown in Figures 3 and 4.

6a)

6b) Figure 6: Scanning electron micrographs of exposed fiber surface of cured laminate.

4. POTENTIAL APPLICATIONS

The PVA blocked fabric (HYPERBONDTM) appears to be very versatile and useful for the establishment of surfaces featuring exposed non-resin infused fiber. Both room temperature cured and autoclave cured structures may be designed to include exposed non-resin infused fiber. This process appears to be applicable to a wide range of fabrication methods including pre-preg fabrication, wet lay-up, and a variety or resin transfer based production processes. In addition to the provision of discrete joint areas, the same materials and methods may be used to provide extensive areas of exposed fiber. Such areas of exposed fiber might be used for a variety of purposes such as a flexible joint. Additional potential uses include the application of refractory coatings, fire retardant coatings, elastomeric coatings, paint systems, and the temporary attachment of objects by hook and loop means similar to Velcro®. New classes of discrete fasteners and Z axis reinforcement may also be enabled by the use of resin blocking systems. Examples include Z axis staples, nails, rivets, and pins which might be installed in a rigid configuration comprised of a cured resin infused zone and a PVA stiffened zone. After curing of the assembly, the PVA could be dissolved, leaving exposed flexible fibers which could be advantageously folded against the surface of the structure where they could be secured with adhesive. Figures 7 through 11 provide a sketch of both potential methods for fabricating this new material and a variety of applications.

Figure 7: Proposed continuous process for treatment of fabric with controlled infusion of resin blocking compound.

Figure 8: Arrangement for controlled edge-wise infusion of a resin blocking compound.

Figure 9: Multiple lap tensile joint between laminates of fabrics with resin-blocked edges

Figures 10a through 10e illustrate how a strip of HYPERBONDTM prepared with the fiber isolating PVA on one side, and a high performance matrix on the other could be applied to the production of a composite part

Figure 10a) Application of HYPERBONDTM strip to surface of uncured composite structure.

Figure 10b) Co-curing of HYPERBONDTM strip to surface of composite structure.

Figure 10c: Removal of PVA resin blocking compound with warm water.

Figure 10d: Application of adhesive and assembly with similarly prepared mating part.

Figure 10e: Curing of the joint. 11a)

11b)

Figure 11: Use of one strip of HYPERBONDTM material on each of two opposing components to be bonded with a stepped lap joint. Each strip of HYPERBONDTM material would have been co-cured with its corresponding laminate

5. CONCLUSIONS

The feasibility of providing dry fibers that protrude from a cured composite part for use in later stage bonding operations has been demonstrated. A thermal gradient was used to control the infusion of a fiber isolating substance into structural reinforcing fibers. This substance preserved the fibers while a matrix was cured around the remaining fibers. The preserved fibers could subsequently be used in fabricated joints after removal of the fiber- isolating substance. Plasticized PVA compounds appear to be particularly promising for blocking non-polar resin systems due to their environmentally benign characteristics and lack of toxicity. The flexibility of plasticized PVA infused fabric allows spooling, handling and application to curved surfaces requiring subsequent bonding operations. The fabrics tested were selected for the purpose of initial evaluation of the temperature gradient controlled infusion process. Other fabric configurations such as needle-felted materials would be expected to have greater effectiveness as a Z axis joint reinforcement and will be the subject of ongoing investigations.

6. REFERENCES

1. J. M. Koyler, et al, Intl. SAMPE Tech. Conf. Series, 45, 365 (2000). 2. D. M. Gleich, et al, Intl. SAMPE Tech. Conf. Series, 45, 818 (2000). 3. R. H. Bossi, R. L. Nereberg ,Intl. SAMPE Tech. Conf. Series, 45, 1787 (2000). 4.Heselhurst R. B., Joining Composite Structures, Tutorial notes SAMPE 2001 5. A. Mekjian, Intl. SAMPE Tech. Conf. Series, 45, 1205 (2000).

ACKNOWLDEGMENTS

We wish to thank Hemant Thakkar and Malcolm Wilborn of Akron Rubber Development Laboratory for their technical support. We also wish to thank Gary Zito and Prof. Hans-Joachim Kleebe of the Colorado School of Mines for the scanning electron microscopy work.

BIOGRAPHIES

Mr. Henry Obermeyer is president of Obermeyer Hydro, Inc. He holds a number of patents in the fields of hydropower, water control, and composite connections. He received his B.S. in Mineral Engineering in 1976 and his M.S. in Metallurgical Engineering in 1978 from the Colorado School of Mines and is a Registered Professional Engineer in the State of Colorado. Dr. Eric Gilbert is President of GCC Services in Fort Collins, Colorado. He received his PhD in chemical engineering from the University of Washington in 2002. His work there focused on the application of composite materials to a variety of sporting good applications.

Information

sampe_paper.doc

11 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

847080


You might also be interested in

BETA
No Job Name
Microsoft Word - master_pmc_vol_1.doc
Microsoft Word - ASC Agenda Website Posting rev3.docx
ASTM Complete Index
asacRevFO.xls