ACI MATERIALS JOURNAL TECHNICAL PAPER

Title no. 100-M15

Bending and Punching Shear Strength of

Fiber-Reinforced Glass Concrete Slabs
by Bin Mu and Christian Meyer

An experimental study was carried out on fiber-reinforced glass the advantage of simplicity and economy of production,
aggregate concrete slabs under a central patch load. The slab performance specifications of thin-sheet products are more
specimens were reinforced either with randomly distributed short readily satisfied with continuous fiber mesh, especially if large
fibers or with continuous fiber mesh with equal fiber volume ratios.
The influences of fiber type, form and volume ratio on the two-way fiber volume ratios are called for.7,8,9
bending behavior, and punching shear capacity of the glass A considerable body of literature exists on the punching
concrete slab were investigated. shear behavior of reinforced concrete slabs. When modeling
Test results revealed that fiber mesh is decidedly more effective in such behavior for small-scale laboratory experiments, it is
bending than randomly distributed fibers; however, randomly important to accurately reproduce the boundary conditions
distributed fibers are somewhat more effective in punching shear.
The shape and location of the critical punching shear perimeter is
that exist in real structures. Slabs are typically supported
independent of fiber type, form, and volume ratio. But crushed continuously along their edges and restrained by adjacent
glass aggregate has some influence on both strength and failure panels to various degrees. Such restraints typically affect the
mode of the slabs. in-plane deformations of a representative slab panel subjected
to a concentrated transverse load by forcing membrane or
Keywords: concrete; fiber-reinforced concrete; glass fiber-reinforced arch action. This can completely alter the load-carrying
concrete; punching shear.
characteristics and failure mode10-12 by greatly enhancing
the punching shear capacity of such restrained slab panels.
INTRODUCTION
Thin-sheet concrete products have attracted considerable The one-way bending behavior of fiber-reinforced concrete
attention in recent years.1,2 There are numerous potential members with glass aggregate has been reported earlier.13 This
applications that were previously difficult or impossible to previous study showed that fiber mesh is clearly more effective
realize with conventional concrete materials. The intro- than randomly distributed short fibers as reinforcement. The
duction of nonferrous reinforcement, either in the form of proper design of concrete slabs, however, whether used for
randomly distributed short fibers or continuous fiber mesh, thin-sheet products or more conventional structural slabs,
greatly reduces the cover requirements, thereby facilitating depends on a thorough understanding of the two-way bending
significant reductions of minimum thicknesses for such thin and punching shear behavior.
sheets. Their mechanical behavior differs considerably from This study reports on the experimental study of slab elements
that of conventional reinforced concrete panels, and so do reinforced either with randomly distributed short fibers or
the manufacturing processes. Thin-sheet products are often
continuous fiber mesh. The effect of the glass aggregate is
manufactured using extrusion and pultrusion processes.
assessed by also testing samples produced with regular river
If thin-sheet concrete panels are subjected to loads,
conventional structural theory needs to be applied to assure sand as aggregate. Three types of materials were studied for
that such loads can safely be resisted. Given their small reinforcement: alkali-resistant glass (AR-glass), polyvinyl
thicknesses, punching shear performance may become acetate (PVA), and polypropylene. In the punching shear
important, especially if the reinforcement is dimensioned for test, only AR-glass fibers were studied.
flexural strength and the panels are subjected to large
concentrated forces either by design or by accident. RESEARCH SIGNIFICANCE
Specific architectural surface treatments can open up an Fiber mesh reinforced thin-sheet concrete products with
entire new category of applications such as claddings, veneers, crushed glass aggregate lend themselves to numerous novel
and face panels that previously were mostly the domain of applications. If used for relatively large panels, their two-
natural stone. The use of crushed glass particles as aggre- way bending and punching shear behavior needs to be
gate also lends itself to a multitude of architectural treatments known. A characterization of the behavior will facilitate the
including polished surfaces or exposed aggregate finishes.3,4
use of such panels for applications that until now were
The problem of alkali-silica reaction (ASR) needs to be
considered, but technology exists to control the potentially primarily the domain of natural stone. The research reported
damaging effects.5,6 herein provides needed insights for their safe use.
It was the objective of the study presented herein to evaluate
the bending and punching shear strength of fiber-reinforced ACI Materials Journal, V. 100, No. 2, March-April 2003.
MS No. 02-131 received April 8, 2002, and reviewed under Institute publication
concrete slabs with glass aggregate. The reinforcement policies. Copyright © 2003, American Concrete Institute. All rights reserved, including
consisted of either randomly distributed short fibers or the making of copies unless permission is obtained from the copyright proprietors.
Pertinent discussion will be published in the January-February 2004 ACI Materials
continuous woven fiber mesh. Whereas short fibers have Journal if received by October 1, 2003.

ACI Materials Journal/March-April 2003 127

 obtain the desired workability. The compressive strength of the
Bin Mu is a postdoctoral research fellow, Center for Advanced Cement-Based
Materials, Northwestern University, Evanston, Ill. He received his BS from Hohai mixture tested on 2 in. cubes after 28 days was 97.7 MPa.
University, Nanjin, People’s Republic of China; his MS from Nanjing University of
Aeronautics and Astronautics, Nanjing; and his PhD from the Hong Kong University
of Science and Technology. His research interests include fracture mechanics of Two-way bending test
concrete and fiber-reinforced concrete, durability of high-performance concrete, The test specimens were square plates of 152.4 mm length
extrusion technology, and the finite element method. and 19 mm thickness, loaded at their center with a round steel
Christian Meyer, FACI, is a professor of civil engineering and engineering mechanics,
pressure head of 12.7 mm diameter to simulate a concentrated
Columbia University, New York, N.Y. He received his PhD from the University of load. The clear test span was 101.6 mm (Fig. 1). The load
California at Berkeley, Calif. He is a member of ACI Committees 446, Fracture was applied by a 50 kN universal testing machine under
Mechanics; 544, Fiber Reinforced Concrete; and 555, Concrete with Recycled
Materials. He is also a member of Joint ACI-ASCE Committee 447, Finite Element displacement control at a rate of 1.0 mm/min. A linear
Analysis of Reinforced Concrete Structures. His research interests include concrete variable differential transducer (LVDT) was mounted to
structures and materials, structural analysis, and design. measure indirectly the center deflections such that extraneous
influences from the supports and loading fixtures were
MATERIALS AND TEST PROGRAM eliminated (Fig. 1). Three data channels, representing applied
Materials load, displacement of load cell, and center displacement of
A single concrete mixture design was used throughout the specimen, were connected to a PC and recorded automatically.
test program. The water:binder ratio was 0.35. Crushed Two sets of specimens were prepared, one reinforced with
post-consumer glass was used as the aggregate, with maximum short random fibers, distributed over half of the slab on the
particle size No. 16. Strictly speaking, the material should tension side, and one with fiber mesh. Three types of fibers
therefore be referred to as mortar instead of concrete. The were studied: AR-glass fibers 12.7 mm long and with tensile
cement:aggregate ratio was 1:2. Fifteen percent of the Type III strengths of 1800 MPa, PVA fibers 6 mm long and with
cement was replaced by metakaolin to suppress the potentially tensile strengths of 1400 MPa, and polypropylene fibers
harmful effects of ASR.5 Suitable admixtures were used to 12.7 mm long and with tensile strengths of 620 MPa. Two fiber
volumes were considered for each fiber type, designated as
Table 1—Test program for two-way bending Vf and 2Vf , corresponding to one and two layers of the fabric
specimens meshes. For polypropylene fibers, Vf = 0.67%, while for
Volume of short Ultimate load,
glass fibers, Vf = 0.25%, and for PVA fibers, Vf = 0.44%. The
No. of
Slab Type of fiber mesh layers fibers Vf , % kN mesh-reinforced specimens contained one or two layers of
S-C — — — 2.41
mesh, positioned on the tension side of the slab with 2 mm
concrete cover. Both the AR-glass and the PVA mesh had a
S-G1 AR-glass — 0.25 2.75
5 x 5 mm grid, while the grid of the polypropylene mesh was
S-G2 AR-glass — 0.50 3.44
4.5 x 4.5 mm. The former two were woven and the third one
S-G3 AR-glass 1 — 4.13 knitted (Fig. 2). All specimens were demolded one day after
S-G4 AR-glass 2 — 5.14 casting and placed in a moisture room for two months before
S-A1 PVA — 0.44 2.91 being tested. Each batch consisted of three samples. Table 1
S-A2 PVA — 0.88 3.86 contains an overview of the test program.
S-A3 PVA 1 — 4.38
S-A4 PVA 2 — 5.46 Punching shear test
S-P1 Polypropylene — 0.67 2.87 To eliminate warping of the slabs, circular specimens with
S-P2 Polypropylene — 1.34 3.72 a diameter of 127 mm and a thickness of 19 mm were cast.
S-P3 Polypropylene 1 — 4.08
S-P4 Polypropylene 2 — 4.98
Note: AR-glass = alkali-resistant glass, and PVA = polyvinyl acetate.

Table 2—Test program for punching

shear specimens
No. of Position
Type of mesh of fiber Short fiber Ultimate
Slab aggregate layers mesh volume, % load, kN Coefficient k
Fig. 1—Test setup for two-way bending of glass concrete slab.
RS-C1 River
sand — — — 3.45 —

RS-C2 Crushed — — — 3.88 1.62

glass

RS-G1 Crushed
glass — — 0.25 4.39 1.69

RS-G2 Crushed
glass 1 Top — 4.05 1.70
Crushed
RS-G3 glass 1 Bottom — 4.09 1.67

RS-G4 Crushed
Top and
2 — 4.17 1.73
glass bottom
River
A-C1 sand — — — 2.84 —

Crushed
A-C2 — — — 3.35 —
glass Fig. 2—Structures of fabric mesh.

128 ACI Materials Journal/March-April 2003

The specimens were supported on a simple ring with a diam- demolded one day after casting and placed in a moisture
eter of 101.6 mm. The test setup, testing machine, and load- room for seven days before being tested. Two control batches
ing rate were similar as in the two-way bending tests and as were cast, one with plain glass concrete and the other with
shown in Fig. 1. plain normal concrete, using river sand as aggregate. To
The specimens were reinforced with either short random achieve the arch action, a plastic tube with a diameter of
AR-glass fibers distributed throughout the slabs or AR-glass 127 mm and thickness of 6.35 mm was used to confine the
fiber mesh. The mesh-reinforced specimens contained one or specimens. The test program is summarized in Table 2.
two layers of mesh (Vf = 0.25% or 2Vf = 0.50%), positioned
either on the slab’s tension side, compression side, or both, TWO-WAY BENDING TEST RESULTS
with 2 mm concrete cover in each case. All specimens were The load-deflection curves for all 13 test specimens are
shown in Fig. 3. Each of the three plots contains five curves:
one for the control specimen without fiber (S-C), two for the
specimens reinforced with randomly distributed short fibers
(Vf and 2Vf), and two for the specimens reinforced with one
or two layers of fiber mesh (Vf and 2Vf ). Figure 3(a) to (c)
show the responses of specimens reinforced with AR-glass,
PVA, and polypropylene fibers, respectively. The ultimate
strengths are summarized in Table 1.
Regardless of what type of fiber is used, continuous fiber
mesh is seen to be clearly more effective than randomly
distributed fibers. A similar observation was made with
one-way beam bending tests.13 The reasons are as follows.
First, the fiber mesh is placed in the optimum location and
effectively bridges cracks in any direction. Second, the bond
between the fiber mesh and matrix is better, benefiting from
the yarn curvature in the warp direction. Third, the effective
bridging length of the fiber mesh is greater. The difference
(a) between the slab and beam cases is that in two-way bending,
yarns in both the warp (or weft) direction and fill direction are
stressed. This leads to an interlocking effect at the intersection of
orthogonal yarns, which greatly improves the anchorage of
the fiber mesh. The larger the bending moment, the more
effective the bridging yarns become. This may explain why
the ultimate strengths of mesh-reinforced specimens are on
average around 45% higher than those of short fiber-reinforced
specimens in the two-way bending test but only 12% higher
for the one-way bending cases13 (refer to Fig. 4). Furthermore,
Table 2 shows that in two-way bending, randomly distributed
short fibers (Cases S-G1, S-A1, S-P1) increase the ultimate
strength of the control case (S-C) by only 18% on average,
whereas in the one-way bending case the corresponding
improvement was with 10% even less.13 Thus, the effectiveness
of fibers is considerably higher in two-way than in one-
way bending. This implies that the use of beam theory to
(b)

(c)
Fig. 3—Load-deflection curves for slabs in two-way bending: Fig. 4—Ultimate strength ratios of specimens reinforced with
(a) AR-glass fiber; (b) PVA fibers; and (c) polypropylene fibers. fiber mesh and short fibers.

ACI Materials Journal/March-April 2003 129

design fiber-reinforced two-way slabs would lead to the highest ultimate punching shear load and ductility. The
overly conservative results. fiber mesh reinforced specimens failed at slightly lower
In the test, a sudden change in the slope of the load-deflection load, but whether the mesh was positioned at the top or
curves corresponds to a major crack transverse across the bottom face of the slab, or both, seems to have had relatively
specimen’s center at the bottom face. As the applied load little effect (RS-G2, G3, and G4). The reason is that a concrete
was further increased, more cracks appeared in the center slab cracks when the diagonal tension or combined action of
region, and the cracks continued to propagate primarily in a shear and direct stress exceeds the tensile strength of the
radial direction. Typical cracking patterns at failure are concrete. In this case, fiber mesh is less effective because it
shown in Fig. 5. The cracking patterns of Samples S-C, S-G1, has been placed in areas of maximum flexural stress, not
and S-G2 are very similar, while those of fiber mesh rein- maximum diagonal tension. Short fibers, on the other hand,
forced slabs (S-G3 and S-G4) are quite different. The better are uniformly distributed and randomly oriented such that
bridging effect of mesh-reinforced specimens causes more some of them effectively bridge the diagonal tension cracks,
cracks and increases the ultimate load capacity. Because the thereby increasing the slab’s shear strength. By comparing
two orthogonal sets of fiber yarns in a mesh have different the two control specimens, it is seen that Sample RS-C2 with
wavy yarn structures and different geometries, their bridging crushed glass as aggregate has a higher punching shear
effects are also different, as shown in Fig. 5 (Specimens S-G3 strength than the Sample RS-C1 with river sand as aggregate.
and S-G4), where the warp direction is horizontal and fill This may be due to the glass aggregates’ irregular shapes
direction vertical. After cracking, the horizontal warp yarns and sharp angles, which can increase shear transfer
assume a larger share of the applied load because they are across cracks.
stronger. Thus, more cracks appear in this direction. Figure 7 and 8 present the crack patterns of confined
specimens after punching shear failure. The crack patterns at
PUNCHING SHEAR TEST RESULTS the top face were all nearly identical, with a small circle of a
The punching shear load-displacement curves are shown diameter, which is almost the same as that of the loading
in Fig. 6, and the ultimate punching shear loads are listed in plate. On the bottom face, the crack patterns of the last five
Table 2. According to these results, Sample RS-G1, which specimens are very similar (refer to Fig. 8). Flexure cracks
was reinforced with short randomly distributed fibers, had do not appear to be extensive. Maximum principal tension
occurs near the mid-plane, initiating web-shear cracks that
propagate toward the top and bottom faces of the slab and
creating the typical conical failure surface.
When comparing the bottom face cracks of Sample RS-C1
with that of RS-C2 (Fig. 8), a marked difference is noted. In
RS-C1, near-vertical flexural cracks formed at the bottom
face when the applied moment exceeded the cracking
moment. With increasing load, these cracks propagated
toward the middle surface of the slab to form flexure-shear
cracks, followed by the appearance of transverse cracks that
then led to punching shear failure. This observation implies
that glass aggregate gives a higher flexural strength than
ordinary river sand, possibly because the irregular shapes
and sharp angle increase the bond between aggregate and
matrix. Also, the punching shear strength of the glass
aggregate concrete slab is higher. According to the load-
deflection curves of Fig. 5, specimens made with glass
Fig. 5—Typical cracking patterns of slabs after two-way aggregate are not more brittle than those made with river sand.
bending. (For description of specimen IDs, refer to Table 1.) To further investigate the relationship between flexural
and punching shear strengths of slabs and the effect of
confinement, two additional control specimens were prepared
and tested, A-C1 with river sand and A-C2 with glass aggre-
gate. The mixture proportions were identical to those of
Specimens RS-C1 and RS-C2, but the boundary was left
unconfined. Both specimens failed in typical flexure. The
flexural cracks initiated at the center of the bottom face and
propagated towards the circular edge. The ultimate strengths
are given in Table 2 and found to be lower than those of the
specimens with constrained edges. As is known, the edge
confinement facilitates compressive membrane action,
which increases the shear friction across cracks and, there-
fore, the punching shear strength. In fact, this boundary
restraint also increases the flexural strength, and this
increase is proportional to the slab thickness.
Whether the edge is constrained or not, specimens with
glass aggregate have higher punching shear capacity than
those with river sand, as shown in Table 2. By comparing the
Fig. 6—Load-deflection curves for slabs in punching shear. strength of Specimen RS-C1 with that of RS-C2 and

130 ACI Materials Journal/March-April 2003

Specimen A-C1 with A-C2, it is seen that the strength Two-way bending and punching shear behavior of slabs
difference is larger for the unconfined case (18%) than the with and without restraints are two different phenomena.
one with confinement (12%). Similarly, the confinement The introduction of fiber reinforcement, either short and
increases the strength of the sample with sand by 21% and of randomly distributed or fiber mesh, makes them more difficult
that with glass by 16%. This implies that the specimen with to analyze. Both types of reinforcement are gaining acceptance
river sand tends to expand in the radial direction more than in practice. The attractiveness of short fibers is due to their
that with glass aggregate or the flexural cracks of the specimen simplicity and economy of concrete production, especially for
with river sand are wider than those of the specimen with glass lower fiber volume ratios. The use of high-performance poly-
aggregate. This means that glass aggregate provided a better meric fiber mesh has gained increased attention in structural
bridging effect in the specimen than the normal river sand engineering applications since the mid-1980s. Advantages
due to its irregular and sharp edges. So, the specimen with of such fabric reinforcements include high strength, low unit
river sand is more likely to be influenced by the boundary weight, and ease of coiling and handling. The fabric meshes
restraint. Whereas the confinement changed the failure are especially suitable for automated fabrication processes
mode of the glass concrete slab from a flexural to a punching for thin sheet products (such as pultrusion or extrusion), as
shear failure, in the case of the specimen with river sand, the well as for repair and strengthening of existing structures.
two failure modes were coupled. This implies that the flexural Such thin sheet reinforcements can provide a complete
resistance of the glass concrete slab is more easily improved integrated armature system to enhance the bending and shear
by the arch action than the ordinary concrete slab. From a resistance of beams and slabs.
design point of view, this result suggests that the restrained The study reported herein systematically compared the
glass concrete slab is more likely to fail in punching shear. In reinforcing effects of short randomly distributed fibers and
either case, proper reinforcement has to be provided. fiber meshes on both the two-way bending and punching
To calculate the shear stress associated with punching shear behavior of glass concrete slabs. It permits the following
shear failure, the failure load is usually divided by the conclusions to be drawn:
slab depth h and the average circumference of the failure
surface l. The average circumference of the failure surface l is
determined on the assumption that the effective section is located
a distance kh from the face of the loaded area and has a
geometrically similar shape. For a round patch load with
diameter d, the effective length l can be expressed as14

l = π ( d + kh ) (1)

The failure surface is typically assumed to have a 45-degree

slope for which k = 1. The value of k can be determined by
measuring the top and bottom diameters of the failure surface.
Using the average of six measurements, the value of k was
estimated for each slab using the following expression

D–d
k = ------------- (2)
2h
Fig. 7—Typical cracking patterns of slabs after punching shear
(top face). (For description of specimen IDs, refer to Table 2.)
where D is the average diameter of the bottom face of the
failure cone. The values thus obtained are listed in Table 2.
They vary from 1.62 to 1.73, with a mean value of 1.68,
which corresponds to a failure surface inclination angle
of about 30 degrees. There are no significant differences
between the plain, short fiber-reinforced, and fiber mesh-
reinforced glass concrete specimens. This k value agrees
with others reported in the literature,12,15 where values of
about 1.5 were given for steel fiber-reinforced concrete slabs
with either confined or free boundary.

CONCLUSIONS
The use of crushed waste glass as aggregate for concrete is
a relatively novel concept. The reservations about such use,
which stem from a concern about long-term ASR problems,
have been addressed by extensive research efforts reported
elsewhere. Solutions to overcome ASR-related problems are
available; therefore, it is now possible to address other issues
related to glass concrete products. The question of flexural and Fig. 8—Typical cracking patterns of slabs after punching
punching shear of fiber-reinforced glass concrete slabs has been shear (bottom face). (For description of specimen IDs, refer
the topic of this study. to Table 2.)

ACI Materials Journal/March-April 2003 131

 1. For two-way bending, fiber mesh is very efficient as 2. Daniel, J. I., and Shah, S. P., eds., Thin-Section Fiber Reinforced
reinforcement. Even at the very low fiber volume ratio of Concrete and Ferrocement, SP-124, American Concrete Institute,
Farmington Hills, Mich., 1990, 441 pp.
0.25%, strain-hardening response can be achieved. For the
3. Jin, W.; Meyer, C.; and Baxter, S., “Glascrete—Concrete with Glass
same fiber volume ratio, the fiber mesh is clearly superior Aggregate,” ACI Materials Journal, V. 97, No. 2, Mar.-Apr. 2000, pp. 208-213.
over randomly distributed short fibers because of better 4. Meyer, C.; Shimanovich, S.; and Vilkner, G., “Precast Concrete Wall
interfacial bond between the matrix and the yarns and the Panels with Glass Concrete,” New York State Energy Research and
locking phenomenon at the yarn intersections. This superior Development Authority, Albany, N.Y., 2003, 144 pp.
reinforcing effect is even more pronounced in two-way 5. Jin, W., “Alkali-Silica Reaction in Concrete with Glass Aggregate: A
bending than was observed in one-way bending; Chemo-Physicomechanical Approach,” PhD thesis, Columbia University, 1998.
2. Short randomly distributed fibers improve the flexural 6. Meyer, C., and Baxter, S., “Use of Recycled Glass for Concrete
Masonry,” New York State Energy Research and Development Authority,
strength and ductility of glass concrete slabs but do not Albany, N.Y., Report No. 97-15, 1997, 144 pp.
change their failure modes. The yield lines are very similar 7. Swamy, R. N., and Hussin, M. W., “Woven Polypropylene Fabrics—
to those of unreinforced glass concrete slabs. For the same An Alternative to Asbestos for Thin Sheet Application,” Fiber-Reinforced
fiber volume ratio, fiber mesh not only greatly increases the Cement and Concrete, Recent Development, R. N. Swamy and B. Barr,
strength and ductility of glass concrete slabs but also forces eds., Elsevier Applied Science, 1989, pp. 90-100.
multiple cracking and, consequently, strain hardening; 8. Peled, A.; Bentur, A.; and Yankelevsky, D., “Flexural Performance of
3. All restrained circular concrete slabs failed in punching Cementitious Composites Reinforced with Woven Fabrics,” Journal of
Materials in Civil Engineering, V. 11, No. 4, Nov. 1999, pp. 325-330.
shear. Glass concrete slabs reinforced with short randomly
9. Reinhardt, H. W., “Integral Formwork Panels Made of GFRC,” High-
distributed fibers had somewhat higher shear strengths Performance Fiber-Reinforced Concrete Thin Sheet Products, SP-190,
than those reinforced with fiber mesh, whether this was A. Peled, S. P. Shah, and N. Banthia, eds., American Concrete Institute,
positioned on the slab tension side, compression side, or Farmington Hills, Mich., 2000, pp. 77-96.
both. In specimens with boundary restraints, the beneficial 10. Rankin, G. I. B.; Niblock, R. A.; Skates, A. S.; and Long, A. E.,
effect of arch action on the flexural strength is more pronounced “Compressive Membrane Action Strength Enhancement in Uniformly
Loaded, Laterally Restrained Slabs,” The Structural Engineer, V. 69,
in the case of glass concrete than with ordinary concrete; and No. 16, 1999, pp. 287-295.
4. Fiber form and volume ratio, fiber mesh position, and 11. Fang, I. K.; Lee, J. H.; and Chen, C. R., “Behavior of Partially
restrained boundary condition do not significantly influence Restrained Slabs under Concentrated Load,” ACI Structural Journal, V. 91,
the shape of the punching shear failure cone. No. 2, Mar.-Apr. 1994, pp. 133-138.
12. Mansur, M. A.; Ahmad, I.; and Paramasivam, P., “Punching Shear
ACKNOWLEDGMENTS Behavior of Restrained Ferrocement Slabs,” ACI Structural Journal, V. 97,
The senior author was supported by Echo Environmental, Inc., New York No. 5, Sept.-Oct. 2000, pp. 765-773.
and Columbia University. This support is gratefully acknowledged. The 13. Mu, B., and Meyer, C., “Flexural Behavior of Fiber Mesh-Reinforced
authors also thank Kuraray Co. Ltd., Nippon Electric Glass American, Inc., Concrete with Glass Aggregate,” ACI Materials Journal, V. 99, No. 5,
and Synthetic Industries, Inc., for providing the PVA, glass and polypropylene Sept.-Oct. 2002, pp. 425-434.
fibers, and fiber mesh. 14. ACI Committee 318, “Building Code Requirements for Structural
Concrete (ACI 318-95) and Commentary (318R-95),” American Concrete
REFERENCES Institute, Farmington Hills, Mich., 1995, 369 pp.
1. Peled, A.; Shah, S. P.; and Banthia, N., eds., High-Performance Fiber- 15. Fang, I. K.; Lee J. H.; and Chen, C. R., “Behavior of Partially
Reinforced Concrete Thin Sheet Products, SP-190, American Concrete Restrained Slabs under Concentrated Load,” ACI Structural Journal, V. 91,
Institute, Farmington Hills, Mich., 2000, 214 pp. No. 2, Mar.-Apr. 1994, pp. 133-139.

132 ACI Materials Journal/March-April 2003