(Michael 1981) Concrete Encased in Fiberglass-Reinforced Plastic
(Michael 1981) Concrete Encased in Fiberglass-Reinforced Plastic
(Michael 1981) Concrete Encased in Fiberglass-Reinforced Plastic
15
and (3)
it carries part of the shear force in the beam through
the two sides. Layers of unidirectional fiberglass can
be added to the bottom face as tension reinforcement
ACI JOURNAL I November-December 1981
(a}
Bidirectional
Unidirectional
reinforcement
(b)
Ribs for mechanical
interlocking
Fig. 4 - Rectangular FRP-encased concrete beams
f-3In.-j
l(76mmll
r-rl2 layers of IOoz./sq yd.
I fiberglass cloth (Type II)
6 in.
(152mm)
L
24 oz./sq. yd balanced
woven roving (2 layers)
13 oz./sq.yd
Unidirectional
Fig. 5 - Cross section of tested FRP-encased concrete
beams
(Fig. 4). The concrete core provides compressive
strength and rigidity and prevents local buckling of the
FRP casing. Adhesion between the concrete and the
FRP is not essential, provided that the FRP box beam
is closed at the two ends and that the unidirectional
layers at the bottom have adequate end anchorage
[Fig. 4(b)]. If for any reason adhesion between the
concrete core and the FRP casing is necessary, me-
chanical interlocking can be introduced by artificial
roughening of the internal surface of the FRP box
beam [e.g., by means of attached ribs, as in Fig. 4(b)].
Five FRP-encased concrete beams were constructed
using zero, five, or ten layers of unidirectional fiber-
glass as reinforcement at the bottom. The concrete mix
used was the one described previously, except that the
water-cement ratio varied from 0.50 to 0.65 to control
concrete strength. The composition of the plastic ma-
trix was the same as in the compression cylinders. The
FRP box beams had the cross section shown in Fig.
5, and were constructed by first hand-laying two
plies of wetted bidirectional fiberglass (24 oz/sq yd
balanced woven roving) against the internal surface of
a wood mold and then adding the layers of unidirec-
tional (if any) at the bottom. After air drying for 24
hr, the three-sided FRP box beam was free-standing
and could be removed from the mold. Concrete was
placed in the box beam afterwards. The top of the
beam was covered with wet burlaps for 7 days. Before
testing in two-point bending at an age of 28 days, the
top and the two sides of the beam were covered with
two layers of wetted 10 oz/sq yd fiberglass cloth.
A description of the five beams tested and a sum-
mary of the test results are presented in Table 2. Beam
E, which can be considered as under-reinforced since
it has no unidirectional fiberglass at the bottom, failed
ACl JOURNAL I November-December 1981
Table 2 - Beam test results
FRP-encased concrete
Layers of
unidirectional f(.
Beam fiberglass psi
A 10 7000
B 10 4000
c
I
5 5000
D 5 4500
E 0 4000
Midspan
deflection
at maximum
load, in.
I
0.9
I
0.75
0.9
M aximum
ment,
kips
mo
in.
200
175
170
130
70
Conventional reinforced concrete
Main
reinforcement Ties
I #6 bar W3 wire at I in. O.C.
2 #3 bars W3 wire at I in. O.C.
2 #3 bars None
I #3 bar W3 wire at 2 in. O.C.
I psi = 6895 Pa.
I in. = 2.54 em.
I in. kip= 113 Nm.
I lb = 0.454 kg.
Midspan
deflection Maximum
[:
at maximum moment,
psi load, in. in. kips
4000 0.4 110
4000 0.4 80
4000 0.15 70
4000 0.3 50
----
Weight
of FRP,
lb
4.5
4.5
3.5
3.5
2.5
Weight
of steel,
lb
--
12
9
3
4.5
------
in a brittle fashion by fracture of the FRP in tension
[Fig. 6(a)]. All other beams were over-reinforced and
experienced a ductile failure mode, which involved
crushing of the concrete at the compression zone.
Bulging of the crushed concrete caused bursting of the
two layers of the fiberglass cloth at the top [Fig. 6(b)].
Removal of the FRP casing after the end of the test
revealed that the concrete inside was severely cracked
in tension and shear and was crushed at the top within
the constant moment region (Fig. 7). However, the
FRP casing was still intact (with the exception of the
damage at the top in the constant moment region) and
elastic, and provided confinement of the concrete. As
a result, unloading after the peak load led to almost
total recovery of the initial shape, and the beam could
even be reloaded to a load almost as high as the peak
load. In other words, over-reinforced FRP-encased
beams have large ductility and retain most of their ca-
pacity after severe overloading.
Typical plots of load versus midspan deflection, con-
crete strain at the top, and fiberglass strain at the bot-
tom (both within the constant moment region of the
beam) are presented in Fig. 8 to 10. Concrete and FRP
strains were measured by 2 in. (50 mm) long electrical
resistance strain gages. These figures clearly show the
capability of over-reinforced beams (Beam B) to return
to their original shape after unloading. Moment-cur-
vature relations for three of the beams are presented
in Fig. 11. Values of ~ d were calculated as the sum of
the measured top and bottom strains. Good prediction
of these plots up to the maximum moment can be ob-
tained by assuming that plane sections of the compos-
ite beam remain plane and by using the FRP properties
in Table 1. Establishment of a reliable prediction pro-
cedure, however, requires additional verification by
test results.
The strength of all five beams tested is greater than
that of typical conventionally reinforced beams of the
same dimensions. Deflections are larger too, because
of the relatively low modulus of the FRP. This can be
443
(a) Under-reinforced beam, with FRP fracture at the bottom
(b) Over-reinforced beam, with concrete crushing and bursting of the
confining FRP
Fig. 6 - Constant moment region of FRP-encased beams after failure
0.25
0.20
-..?
::go.l5
0::
-o
0
0 0.10
_J
0.05
0 0.01
Beam E
0.02
Fig. 7 - Damage of the concrete core in a failed over-
reinforced FRP-encased beam
Fig. 8
beams
Midspan deflection 8/J.
Load-midspan deflection of FRP-encased
444 ACI JOURNAL I November-December 1981
0.25
0.20
-u
....
0.15
"
..0
'
a..
"
0
0.10
0
......J
0.05
2 3
Concrete strain
Fig. 9 - Load versus concrete strain at the top, in the
constant moment region of FRP-encased beams
seen in Table 2, by comparison with the results of tests
on four conventional beams of the same dimensions
as the FRP-encased ones, ranging from over-reinforced
to overly under-reinforced. The same table shows the
total amount of reinforcing steel or FRP in each beam.
Since the cost of FRP used is approximately $1 per lb,
FRP-encased concrete beams are superior to reinforced
concrete beams on a strength/material cost basis.
CONCLUSIONS AND RECOMMENDATIONS
The paper presents a preliminary investigation of the
mechanical behavior of FRP-encased concrete under
short-term compression or bending. Several other me-
chanical aspects must be investigated for further un-
derstanding of the mechanical behavior of this new
composite material. These include the questions of
strength under sustained compression, behavior under
eccentric loading and buckling of columns, and the is-
sues of time-dependent deformations and creep rupture
of beams. Design procedures can be developed only
after a thorough understanding of the behavior is
achieved. Very important questions to be resolved, be-
fore putting the proposed concept into application, are
the problems of fire resistance and durability of the
FRP under long-term exposure to sunlight and atmo-
spheric conditions. These problems can be solved, or
at least mitigated, by applying surface coatings or by
appropriately selecting the composition of the plastic
matrix.
In conclusion, the idea of constructing beams and
columns by encasing concrete in FRP appears very
promising. Provided that satisfactory answers are given
to the problems of fire resistance and durability of the
casing, FRP-encased concrete can provide very eco-
nomic alternatives to conventional reinforced or pre-
stressed concrete because material costs can be re-
duced, reusable forms are not necessary, and
construction costs can be reduced by application of
highly automated prefabrication techniques.
ACI JOU ANAL I November-December 1981
0.25
---Beam E
2 4 6
FRP Strain
Fig. 10 - Load versus FRP strain at the bottom, in
the constant moment region of FRP-encased beams
0.4
(\J
"
..0
'
0.3
0.2
...
c
Q)
E
0
0.1
0 4
Beam B
B
Curvature
12
Fig. 11 - Moment-curvature relations of FRP-encased
beams
FRP-encased circular columns under concentric
compression can attain very high strengths and ductil-
ity due to the confinement of the concrete by the FRP.
If their creep and buckling characteristics are equally
good, such columns may lead to significant savings
over reinforced concrete. Rectangular concrete beams
encased in FRP and reinforced with unidirectional
FRP at the bottom have very good strength and the
remarkable ability to recover their original shape upon
load removal, even after considerable damage has been
done to the encased concrete. Such beams can also lead
to considerable savings, especially if the FRP casing is
optimized.
REFERENCES
I. Furlong, Richard W., "Strength of Steel-Encased Concrete
Beam Columns," Proceedings, ASCE, V. 93, ST5, Oct. 1967, pp.
113-124.
445
2. Knowles, Robert B., and Park, Robert, "Strength of Concrete
Filled Steel Tubular Columns," Proceedings, ASCE, V. 95, STI2,
Dec. 1969, pp. 2565-2587.
3. Bertero, Yitelmo V., and Felippa, C., Discussion of "Ductility
of Concrete" by H. E. H. Roy and Mete A. Sozen, Flexural Me-
chanics of Reinforced Concrete, SP-12, American Concrete Insti-
tute/ American Society of Civil Engineers, Detroit, 1965, pp. 227-
234.
4. Burdette, Edwin G., and Hilsdorf, Hubert K., "Behavior of
Laterally Reinforced Concrete Columns," Proceedings, ASCE, V.
97, ST2, Feb. 1971, pp. 587-601.
5. Roy, H. E. H., and Sozen, Mete A., "Ductility of Concrete,"
Flexural Mechanics of Reinforced Concrete, SP-12, American Con-
crete Institute/ American Society of Civil Engineers, Detroit, 1965,
pp. 213-224.
6. Sargin, M.; Ghosh, S. K.; and Handa, V. K., "Effects of Lat-
eral Reinforcement upon the Strength and Deformation Properties
of Concrete," Magazine of Concrete Research (London), V. 23, No.
75-76, June-Sept. 1971, pp. 99-110.
7. Sheikh, Shamin A., and Uzumeri, S. M., "Strength and Duc-
tility of Tied Concrete Columns," Proceedings, ASCE, V. 106, ST5,
May 1980, pp. 1079-1102.
8. Somes, Norman F., "Compression Tests on Hoop-Reinforced
Concrete," Proceedings, ASCE, V. 96, ST7, July 1970, pp. 1495-
1509.
446
9. Iyengar, K. T. Sundara Raja; Desayi, Prakash; and Reddy, K.
Nagi, "Stress-Strain Characteristics of Concrete Confined in Steel
Binders," Magazine of Concrete Research (London), V. 22, No. 72,
Sept. 1970, pp. 173-184.
10. Martin, C. W., "Spirally Prestressed Concrete Cylinders,"
ACI JouRNAL, Proceedings V. 65, No. 10, Oct. 1968, pp. 837-845.
II. Kurt, Carl E., "Concrete Filled Structural Plastic Columns,"
Proceedings, ASCE, V. 104, STl, Jan. 1978, pp. 55-63.
12. Richart, F. E.; Grandtzaeg, A.; and Brown, R. L., "The Fail-
ure of Plain and Spirally Reinforced Concrete in Compression,"
Bulletin No. 190, Engineering Experiment Station, University of Il-
linois, Urbana, Apr. 1929, 72 pp.
13. Newman, K., and Newman, J. B., "Failure Theories and De-
sign Criteria for Plain Concrete," Proceedings, International Civil
Engineering Materials Conference on Structure, Solid Mechanics and
Engineering Design (Southampton, 1969), Wiley Interscience, New
York, 1972, Part 2, pp. 963-995.
14. lzbal, Mohammed, and Hatcher, David S., "Post-Crushing
Behavior of Bound Concrete Beams," Proceedings, ASCE, V. 103,
ST8, Aug. 1977, pp. 1643-1654.
15. Iyengar, K. T. Sundara Raja; Desayi, Prakash; and Reddy,
K. Nagi, "Flexure of Reinforced Concrete Beams with Confined
Compression Zones," ACl JouRNAL, Proceedings V. 68, No. 9,
Sept. 1971, pp. 719-725.
ACI JOURNAL I November-December 1981