Tailoring Extruded HPFRCC to be Nailable

(ProQuest: ... denotes formulae omitted.)

INTRODUCTION

Residential housing is a large market, accounting for 15% of the U.S. gross domestic product.1 A variety of woodbased, metallic, polymeric, and cement-based materials are currently used in residential construction. Durability, economic, and environmental concerns, however, exist with these materials. Durability problems include fire resistance, resistance to water intrusion (resulting in freezing-andthawing damage, mold, and wood rot), vulnerability to termite infestation, radiation, and rust.2-6 Furthermore, the use of wood-based materials in residential constructions contributes to deforestation and can be economically disadvantageous due to the drastic increase in the cost of wood-based materials.7 Therefore, new building materials that are durable, cost-effective, and environmentally friendly are needed.

Cement-based composites are excellent materials to be used in residential applications. These composites can be designed to be strong, ductile, durable, cost-effective, environmentallyfriendly, and design-versatile. Currently, there are primarily only two types of cement-based materials that are used in the housing market. The first are backerboard materials, which are effective in interior applications. The second are Hatschekproduced composites that are used in external siding applications. Research has indicated that these materials may be vulnerable to freezing-and-thawing attack.3,8 In this research, an alternate cement-based material that could be used in both exterior and interior residential applications is investigated-extruded high-performance fiber-reinforced cementitious composites (HPFRCC).

Extrusion is an advanced processing technique that can be used to produce HPFRCC. These materials exhibit a strain hardening response under tensile loading, with a tensile strength and strain capacity that is superior to conventional fiber-reinforced materials. Researchers have demonstrated that this high performance can be achieved in a variety of ways, including using micromechanical modeling,9 tailored fiber geometries,10 and advanced processing techniques.11-15

In extrusion, a stiff cementitious dough is forced through a die of desired cross section using either an auger or a ram. During this process, the material is subjected to high compressive and shear forces, which densify the matrix, improve the fiber-matrix bond, and align fibers in the direction of extrusion.13,16-19 As a result, the mechanical performance and durability of the composites are superior to similar cast materials.14,17

This work aims at developing extruded HPFRCC for use in residential applications. Specifically, the constructibility of the composites is examined by investigating nailing performance. With current extruded HPFRCC formulations, nailing is difficult, typically resulting in excessive cracking. It is anticipated that, for this technology to be accepted by industry, the material must be able to be nailed in the same manner as conventional materials. Therefore, the nailing performance of extruded HPFRCC is evaluated experimentally using a technique previously developed.20 Next, fracture mechanics-based and cavity expansion-based models are used to determine the material parameters that govern nailing. Using the results from the experimental and modeling work, nailable composites are produced. These nailable composites contain lightweight fillers and are reinforced with either glass or polypropylene (PP) fibers.

RESEARCH SIGNIFICANCE

Extruded composites offer a number of advantages over the materials currently used in residential construction, including environmentally-friendly compositions, design versatility, and superior mechanical performance. The constructibility of these materials, however, is still not desirable, with composites being difficult to nail. In this work, the material parameters that govern nailability are systematically investigated and the properties that are needed to achieve a good nailing performance are determined. Using these results, composites can be tailored for residential applications.

MATERIALS

In this research, two types of composites are studied: laboratory-produced extruded composites and commercially available materials. The intent is to compare the nailing performance of the extruded composites to materials that are currently used in residential applications and are, therefore, known to be nailable.

Extruded composites

Details about production of the extruded composites are given elsewhere and are only summarized herein.21,22 The matrix consisted of 30% Class F fly ash, 11% silica fume, 13% cement, 39% water, 1% high-range water-reducing admixture, 1% hydroxypropyl methylcellulose, and 5% fibers (except in the case of polyvinyl alcohol (PVA) where only 2% fibers were used because higher fiber dosages caused clumping) by volume. Unreinforced material and singlyreinforced cellulose, glass, PP, and PVA composites were produced. The properties of the fibers are given in Table 1.

Commercially available materials

Four commercially available materials were examined and compared with the lab-produced extruded materials: two cement-based and two engineered wood. One of the cementbased materials was a tile backerboard that consists of a light cementitious core sandwiched between glass fiber mesh. The other was a Hatschek-produced, cellulose fiber-reinforced cement board for siding and backerboard applications that contains approximately 30% cellulose fibers by volume and is laminated. Oriented strand board (OSB) and plywood were the two engineered-wood products used, and are the two most commonly used residential building materials.1 OSB is composed of wood strands that are bonded together with a resin. Plywood is a material composed of multiple layers, or plies, that are adhered to each other using a resin. The plywood used was a three-ply plywood with an American Plywood Association rating of B-C, Group 1, Exterior 221, PSI-95 underlayment. Specimens were selected such that the thicknesses were as close as possible to the 8 mm (0.31 in.) thickness of the lab-produced extruded composites, with a nominal thickness of 8, 8, 6.35, and 8.7 mm (0.31, 0.31, 0.25, and 0.34 in.) for tile backerboard, the cellulose fiber-reinforced cement board, OSB, and plywood, respectively.

Comparison of material properties

Table 2 compares the density and the mechanical performance of the extruded and commercial materials. Details about these measurements can be found in Reference 21. The densities of the extruded composites are similar, regardless of type of fiber reinforcement. Compared with the commercial materials, the densities of the extruded materials are slightly higher (10 to 20%) than the cement-based materials and are approximately three times greater than the wood-based materials. The elastic modulus of the extruded composites is higher than the moduli of the commercial materials. The flexural strength and toughness of the extruded composites cover a broad range, and these values fall within the range of the performance of the commercial materials.

EXPERIMENTAL EVALUATION OF NAILING PERFORMANCE

Evaluation

Nailing performance was evaluated using a methodology that was developed by the authors and can be found elsewhere. 21,22 Three criteria were defined to describe nailability: 1) ease of nailing; 2) resistance to cracking; and 3) nailholding ability. A material with good nailability should be easy to nail, have a high resistance to cracking, and be able to hold the nail after it has penetrated the material. Because the nail-holding ability is dependent on the substrate used, it is not examined in this work.

Ease of nailing is defined as the load required to drive a nail through the specimen and is measured using a dynamic nail test that was developed by the authors. A Charpy impact setup was modified for the nailing test.20 Test specimens were 50.4 x 177.6 x 8 mm (1.98 x 4.63 x 0.31 in.). Previous research shows that the failure modes of composites nailed using this dynamic test are similar to the failure modes experienced by materials nailed using a pneumatic nail gun, which is the method most commonly used to fasten materials in residential construction, indicating that this test method is representative of true field nailing.21-23 The lower the nailing load, the higher the ease of nailing.

Resistance to cracking was evaluated after the nailing test. The flexural strength of the specimens that had been nailed was evaluated and compared with specimens that had not been nailed. The residual flexural load was calculated as

... (1)

The greater the residual flexural load, the higher the resistance to cracking.

Results

Nailing parameters-The ease of nailing and resistance to cracking of the composites are presented in Fig. 1 and 2, respectively. When considering all of the commercially available materials, the nailing loads of the extruded composites fall within the range of the commercial materials. If only the cement-based commercial materials are considered, however, the dynamic nailing loads of the extruded materials are approximately 50% higher. For the extruded composites, only the PP and glass composites have a residual flexural load after dynamic nailing (the plain, cellulose, and PVA composites break), and these residual loads are similar to the loads of the commercial materials.

Failure modes-The failure modes of the composites are observed after the nailing test by looking at the nail exit face. Figures 3 and 4 show the failure modes for the extruded and commercial composites, respectively. The plain matrix simply breaks, suggesting a flexural-induced failure. The cellulose and PVA composites break but also show signs of scabbing, indicating a combination of flexural and punching shear failures. The PP composite exhibits both radial cracking and scabbing, suggesting penetration and punching shear types of failure. Finally, the glass composite only shows a scabbing type of failure. On the other hand, the commercially available composites fail primarily due to penetration.

Discussion-Based on a penetration type of loading, three different failure mechanisms can occur, as is demonstrated in Fig. 5. The first mode is pure penetration, in which the nail simply penetrates through the material by creating a cavity that has a diameter approximately equal to the diameter of the nail. A penetration type of failure creates tensile radial forces that can cause radial cracking if the forces exceed the tensile strength of the material. A bending type of failure can also occur, causing the material to break. Third, a punching shear failure can be induced, which would be evidenced by scabbing on the nail exit face. For a composite to have good nailability, pure penetration failure is preferred because this type of failure should induce the least amount of damage.

A variety of different failure modes were observed for the extruded and commercial materials: 1) penetration; 2) scabbing; 3) radial cracking; and 4) complete material separation (breaking). These failure modes are ordered based on preferred failure mode, with the penetration mode being the best (as with the commercial composites), followed by a scabbing type of failure, then radial cracking and, finally, the broken composites. This order is made based on minimizing damage. According to this classification, the glass and PP composites have the most desirable failure modes of the extruded composites.

MODELING NAILING PERFORMANCE

Nailing is an extremely complex process that can include a variety of different failure mechanisms. In addition, the fracture properties of HPFRCC, such as the extruded materials studied herein, are also complicated, with a number of toughening mechanisms that can occur during failure. To help explain the observed nailing performance of the extruded composites, a number of different existing models were employed. Limitations exist with the application of these models; however, they do still help to provide insight into the material properties that govern nailing.

The contribution of both the fiber reinforcement as well as the matrix was considered. Fibers are anticipated to primarily influence the residual load, whereas the matrix should affect the nailing load. The role of the fiber reinforcement in the nailing performance was examined using fracture mechanics. In addition, a cavity expansion analysis was used to understand the influence of the matrix properties on the nailing process. Using the results from these models, a better understanding of the material parameters that influence nailability is attained.

Fracture mechanics modeling-role of fiber reinforcement

Fibers influence the development of cracks. With conventional fiber-reinforced materials, fibers generally improve the post-peak toughness by bridging and delaying the coalescence of macrocracks. When high-performance is achieved, however, a homogeneous distribution of microcracking occurs, leading to a composite that undergoes strain hardening, with an improved tensile strength and strain capacity. As multiple cracking occurs, the cracks do not widen because the fibers suppress the extension of matrix cracking by crack bridging, interfacial debonding, and frictional slide.24

The extruded materials studied herein show strain-hardening types of responses when subjected to flexural loading. It is hypothesized that the fracture toughness of these materials should play an important role in nailability. As the nail is driven, or penetrated, into the material, cracking will occur. If the material has sufficient fracture toughness, the damage due to this cracking will be minimized, thus giving higher residual flexural loads. To evaluate the effects of the fiber reinforcement on nailing performance, the toughening behavior of the extruded materials was investigated using existing fracture mechanics modeling techniques. R-curves and brittleness indexes were analyzed to evaluate the role of the fiber reinforcement on the peak load and a micromechanical model was used to evaluate the influence of the fibers on strain hardening.

R-curves and brittleness indexes-Fracture resistance curves, or R-curves, were used to give an indication of the toughening behavior of the extruded composites. R-curves were obtained using the approach proposed by Ouyang et al.,25 in which experimentally-determined fracture parameters are used to generate R-curves by solving a series of differential equations.25,26 Fracture parameters were determined according to the Jenq-Shah two-parameter model, which describes fracture using two geometry-independent parameters, the critical stress intensity factor K^sub IC^ and the critical crack tip opening displacement CTOD^sub c^.27 This model assumes that the material is homogeneous and that failure occurs due to the propagation of a single crack. Therefore, the analysis is only valid until the peak load, at which point microcracks coalesce to form a macrocrack.

The Jenq-Shah fracture parameters, K^sub IC^ and CTOD^sub c^, were also used as an indication of the size of the fracture process zone. Jenq and Shah27 proposed a brittleness index Q given by

... (2)

where E is the elastic modulus. The larger the value of Q, the more ductile a composite is. Typical ranges for Q are 12.5 to 50 mm (0.5 to 1.97 in.) for cement paste, 50 to 150 mm (1.97 to 5.91 in.) for mortar, and 150 to 350 mm (5.91 to 13.78 in.) for concrete.28

Double-edge-notched tension specimens were loaded and unloaded in crack-mouth opening-displacement control to obtain K^sub IC^ and CTOD^sub c^. The fracture parameters are given in Table 3. Due to the brittle nature of the plain composites (making it hard to mechanically secure the specimens before testing) and the quick yielding of the PP composites (causing it to be difficult to obtain the unloading compliance at the peak load), the fracture parameters of these composites could not be determined. More details about the analysis are given in Reference 22.

Figure 6 presents the R-curves for the cellulose, glass, and PVA fiber-reinforced composites. Inspection of the R-curves indicates that the fracture toughness of the glass composite is much higher than either the PVA or cellulose fiber-reinforced materials. The brittleness index Q, given in Table 3, shows a similar trend, with the glass having the highest brittleness index, followed by PVA and cellulose. Both the R-curves and the brittleness index suggest that the glass composites have a higher fracture toughness than the cellulose or PVA fiber-reinforced materials.

It is important to note the limitations in the modeling approach used previously. The behavior of the highperformance composites is being evaluated only until the peak load, assuming that the material is homogeneous and that failure is due to the propagation of a single crack. It is known that the failure mechanisms in high-performance materials are quite complex, however, with a variety of toughening mechanisms, including fiber crack bridging, interfacial bonding, and frictional slide. Nevertheless, the models do give an indication of the relative toughening behavior of the cellulose, glass, and PVA fiber-reinforced composites.

Micromechanical modeling-Researchers have shown that the most fundamental property for fiber-reinforced composites is fiber bridging across a matrix crack.29-31 By examining this property, it is possible to determine what conditions are necessary for strain hardening to occur. Takashima et al.32 used micromechanical modeling to derive the following relationship, which describes the critical volume fraction V^sub critical^ needed in aligned fiber-reinforced cementitious composites for strain hardening32

... (3)

with

... (4)

... (5)

where d^sub f^ is the diameter of the fiber; L^sub f^ is the length of the fiber; τ is the interfacial shear stress; J^sub tip^ is the crack tip toughness; E^sub f^ and E^sub m^ are the moduli of the fiber and matrix, respectively; V^sub f^ and V^sub m^ are the volume fractions of the fiber and matrix, respectively; and K^sub m^ is the matrix fracture toughness. This model assumes that the fiber-matrix interface can be characterized by the following: fibers do not break, the interfacial friction between the fiber and the matrix is constant, and fibers completely debond and pull out of the matrix.

Using Eq. (3), the critical volume fraction of fiber reinforcement for strain hardening was calculated for the extruded composites. The fiber properties are given in Table 1. The matrix fracture toughness K^sub m^ was assumed to be 26 MPa*mm^sup 1/2^ (0.75 ksi*in.^sup 1/2^), which was calculated by Akkaya et al.20 for a similar mixture design. The value of E^sub m^ was approximated as 14.85 GPa (2146.6 ksi) corresponding the modulus determined from the flexural testing of the plain composites (Table 2). Values for the interfacial shear strength were obtained from the literature and are given in Table 4. No value could be found for the interfacial properties of cellulose fiber. Therefore, the critical volume fraction was not obtained for the cellulose-fiber reinforced composites.

Table 4 presents the critical volume fraction calculated for the glass, PP, and PVA composites. A comparison of V^sub critical^ and V^sub used^ (corresponding to the V^sub f^ actually used in the extruded composite) indicates that the glass and PP composites have enough fiber reinforcement to undergo strain hardening with multiple cracking, whereas the PVA fiber-reinforced composites do not. This result suggests that the fracture toughness of the glass and PP fiber-reinforced materials is greater than that of the PVA composites.

Equation (3) suggests that by reducing the crack tip toughness J^sub tip^, a lower critical volume fraction of fibers is needed to achieve multiple cracking. Thus, a reduction in the modulus or fracture toughness of the matrix should improve the strain hardening response of the extruded materials as discussed in the Discussion that follows.

It is important to note that the τ values reported in the literature depend on the properties of the matrix and the fiber surface used by the authors and are, therefore, not identical to the properties of the materials used in this work. Additionally, the model makes assumptions about the fiber-matrix debonding properties, which might not be valid for all of the fibers investigated.

Cavity expansion modeling-role of matrix

The process of driving a nail through a cementitious material is similar to that of a projectile penetrating a concrete target. One of the more commonly-used analytical models to describe this phenomenon is the cavity expansion model, in which penetration is assumed to cause ductile hole-enlargement based on either cylindrical or spherical expansion approximations. The hole is enlarged by the radial motion of the material, expanding to the diameter of the projectile.

Luk and Forrestal36 used cavity expansion models to derive analytical equations for the penetration of projectiles into concrete targets. These models included constitutive relationships that required that the target material be characterized by triaxial testing. Evaluating the triaxial behavior of cementitious materials, however, is difficult because most triaxial equipment does not have the required capacity for testing these materials. Therefore, in subsequent work, Forrestal et al.37 developed a dimensionally consistent empirical equation that contains the functional form of previously published analytical models, but describes the concrete by only its compressive strength.

By assuming that the impact of the projectile is normal to the target material and that the projectile does not deform, the following relationship was derived for the axial force on the projectile F34

F = πa^sup 2^ (τ^sub 0^A + NBρV^sup 2^) (6)

where the projectile is described by the radius a, the nose shape factor N, and the impact velocity V. The target concrete has a density ρ, and the constants τ^sub 0^A and B, which are material parameters determined based on the triaxial response. The value of B was found to have a narrow range and mostly depended on the compressibility of the target.38-40 Based on this previous research, the authors approximated B to be equal to 1. The term τ^sub 0^A depends on the shear strength of the concrete material and is set equal to Sf^sub c^', so that Eq. (6) becomes

F = πa^sup 2^(Sf^sub c^' + NρV^sup 2^) (7)

where S is a dimensionless empirical constant. This relationship only applies when the penetration depth is greater than 4a. When the penetration depth is less than 4a, surface-cratering can dominate the response. Experimental data was obtained for a variety of data sets and data fitting was then used to determine the value of S, which was found to be equal to 82.6f^sub c^'^sup -0.544^.

A few limitations exist with using the cavity expansion approximation given by Eq. (7). First, the model assumes that flexural failure is not dominant. Second, the empirical relationship was derived using velocities ranging from 250 to 800 m/second (820 to 2625 ft/second), which are higher than the nailing speed used in this work, although not higher than the velocities of the pneumatic nail guns used in the field. Furthermore, the empirical equation was developed for concrete targets, not fiber-reinforced cementitious composites. It is assumed that Eq. (7) can be used to evaluate the nailing load of the extruded composites, provided that flexural failure mechanisms are not dominant. This relationship suggests a material will have a higher ease of nailing if the density or compressive strength is reduced and will be discussed in the following Discussion.

Discussion

The fracture mechanics models used suggest that the fracture toughness of the glass and PP composites is greater than the cellulose and PVA composites. These composites demonstrated the best nailability, with preferred failure modes and residual loads that were comparable to the commercially available materials. This finding suggests that the fracture toughening properties of the extruded materials play a significant role in the nailing performance, with the composites with a greater fracture toughness having a better nailability. The failure modes of the glass and PP extruded composites, however, are still not penetration-dominated like the commercial materials. To improve the nailing performance, a penetration type of failure is desired.

The cavity expansion relation suggests that by reducing the density and/or compressive strength of the extruded materials, the resistance to penetration should be reduced. As the resistance to penetration is lowered, it is anticipated that a better nailability will be attained, with composites exhibiting penetration-dominated failures. Furthermore, according to the fracture mechanics-based equation describing the critical volume fraction for strain hardening (Eq. (3)), a reduction in the crack tip toughness of the matrix (from either a lower matrix modulus or fracture toughness) will reduce the amount of fibers needed for strain hardening. Therefore, composites were extruded with lightweight fillers to try to improve the contribution of the matrix to the nailing process. It is anticipated that by using lightweight fillers in the extruded composites, the density, compressive strength, and elastic modulus can be reduced.

Because the glass and PP fiber-reinforced composites demonstrated the greatest nailability potential, both based on the experimental results as well as the modeling, these two types of reinforcement were used with the lightweight fillers.

EXTRUDED COMPOSITES WITH LIGHTWEIGHT FILLERS

Material production

Glass and PP composites were extruded with four different types of lightweight fillers: two sizes of expanded polystyrene beads, perlite, and porous glass spheres. The matrix design was the same as was used previously. Mixing was conducted in the same manner as was used for the extruded composites without lightweight fillers.20 The lightweight materials were added after the extrudate came together to form a cohesive dough and mixing continued until a cohesive dough was again achieved. Lightweight fillers were incorporated at 15, 30, and 45% of the total volume. Either glass or PP fiber reinforcement was used.

The effects of the lightweight filler on the processing parameters, physical properties, and mechanical performance of the extruded composites were evaluated. Expanded polystyrene beads (EPS) showed the most promising results.22 The beads had a density of 0.035 g/cm^sup 3^ (0.001 lb/in.^sup 3^) and had diameters that ranged from 1.4 to 2.2 mm (0.051 to 0.072 in.) (EPS 1) and from 0.8 to 1.4 (0.029 to 0.051 in.) (EPS 2). The EPS beads did undergo some damage due to extrusion, but the damage appeared to be reasonable (measured by estimating theoretical densities). The material properties of the glass and PP fiber-reinforced specimens containing the EPS are given in Tables 5 and 6, respectively. Generally, as the amount of lightweight filler increased, the flexural strength, elastic modulus, and density decreased.

Nailing performance

The nailing performance of the extruded composites containing lightweight fillers was evaluated using the modified Charpy impact setup. Figure 7 presents the failure modes for the glass fiber-reinforced materials. Similar trends were observed for the PP fiber-reinforced composites. Generally, as the amount of lightweight filler increases, the failure mode improved.

Figure 8 summarizes the types of failure observed for the different composites. Based on the failure modes observed, five of the composites demonstrated good nailability (penetration-dominated failures): for the glass-reinforced, EPS1 30% and EPS2 45%, and for the PP-reinforced, EPS1 30%, EPS1 45%, and EPS2 45%.

Generally, as the amount of lightweight filler increased, the ease of nailing improved and the resistance to cracking improved slightly or remained the same.22 Figures 9 and 10 present the nailing load and residual flexural load, respectively, for the commercial composites and for the lightweight composites that were defined as nailable (penetration dominated failures). The nailing load of the nailable extruded composites with EPS falls within the range of the commercial materials. Furthermore, if only the cementbased commercial materials are considered, all of the extruded composites containing EPS have similar nailing loads, except for the glass composite with 30% EPS1. When considering the residual flexural load, it is seen that, except for the glass fiber-reinforced composite with 45% EPS2, the residual load for the extruded composites falls within the range of the residual load of the commercial materials.

Discussion

The nailability results show that glass or PP fiber-reinforced composites containing lightweight fillers have good nailability, demonstrating that by properly tailoring the fiber reinforcement and the matrix properties, nailable composites can be achieved.

The fracture mechanics modeling and experimental work had previously indicated that the glass and PP fibers were effective at controlling cracking and the inclusion of lightweight material did not adversely affect their contribution to the nailing performance. According to Eq. (3), the critical volume fraction decreases as the crack tip toughness (J^sub tip^ [asymptotically =] K^sub m^^sup 2^/E^sub m^) decreases. Due to the brittle nature of the extruded composites without fibers, it is not possible to evaluate the critical stress intensity of the matrix. It is reasonable to assume, however, that as the density of the matrix is reduced, the critical stress intensity factor will decrease. Additionally, the elastic modulus decreased with the increasing addition of EPS.22 Therefore, Eq. (3) suggests that the addition of EPS will reduce the amount of fibers needed to achieve multiple cracking and, therefore, the toughening behavior improves.

Figure 11 presents the resistance to penetration (Eq. (7)) plotted versus the dynamic nailing load for the extruded composites. Both the composites with and without EPS are plotted. Only composites that remain intact after nailing are shown (that is, the plain, cellulose, and PVA are not included) because the cavity expansion relationship assumes that flexural failure is not dominant. A good relationship is observed between the resistance to penetration and the nailing load for the materials, which demonstrates the benefit of decreasing the density and elastic modulus of a material to enhance the ease of nailing.

CRITERION FOR NAILABLE COMPOSITES

By adding EPS to the glass and PP composites, nailable composites were achieved. The results from this work can be used to define the nailability parameters needed for good nailability as well as the material parameters that govern nailing performance. For a composite to be nailable, it should demonstrate: 1) high ease of nailing (low nailing load); 2) low extent of damage (high residual flexural load); and 3) penetration-dominated failure.

Furthermore, the following materials properties will improve the nailing performance of the extruded composites: 1) high fracture toughness; 2) low density and compressive strength; and 3) low modulus of elasticity.

It is important to note that the focus of this work is to determine the material parameters required to achieve a nailable composite, not to propose a specific formulation for use in residential housing applications. EPS were used as a lightweight filler in the fiber-reinforced, cement-based extruded composites to achieve a high fracture toughness, low density and compressive strength, and a low elastic modulus. For EPS to be used in residential applications, the constructibility (handling and cutting), durability, and cost of these composites would need to be systematically evaluated. This research is outside of the scope of the current work.

CONCLUSION

The objective of this research was to determine the material properties that govern nailing. Nailable extruded fiberreinforced composites were produced by tailoring both the fiber reinforcement as well as the matrix properties to achieve a material with a high resistance to cracking and high ease of nailing. Glass and PP fibers were successfully used. These fibers demonstrated the greatest potential when nailed without lightweight filler, which existing fracture mechanics models suggest could be due to the high fracture toughness of these materials. EPS were successfully incorporated to decrease the density, elastic modulus and compressive strength of the extruded composites, which, according to the cavity expansion mechanics, decreases the resistance of the material to penetration. Five different extruded composites were produced that exhibited good nailability (based on penetration-based failure modes). The nailing performance of these composites, evaluated by nailing load and the residual flexural load, were comparable to the performance of the commercially available materials. Based on these results, criterion for evaluating nailing performance and for the material properties necessary for good nailability were defined.

ACKNOWLEDGMENTS

This work was funded by NSF PATH Grant No. CMS-0122045. Cellulose, glass, PP, and PVA fibers were provided by Weyerhaeuser, Saint Gobain, Forta Corporation, and Kuraray Corporation, respectively. Silica fume was supplied by W.R. Grace.