by - trl.co.uk · for a normal structural grade concrete with high yield deformed bar...
TRANSCRIPT
TRANSPORT and ROAD RESEARCH LABORATORY
Department of the Environment Department of Transport
TRRL LABORATORY REPORT 947
CRACK CONTROL IN CONCRETE BEAMS BY SURFACE REINFORCEMENT WiTH GLASS FIBRES
by
K D Raithby
Any views expressed in this Report are not necessarily those of the Department of the Environment or of the Department of Transport
Bridge Design Division Structures Department
Transport and Road Research Laboratory Crowthorne, Berkshire
1980 ISSN 0305-1293
Abstract
1.
2.
3.
CONTENTS
Introduction
Mechanism of crack suppression by glass fibre reinforcement
Tests on reinforced concrete beams with grc surface reinforcement
3.1 Description of test beams
3.2 Mix proportions and nominal strength properties
3.3 Method of test
4. Results of tests
4.1 Ultimate failure loads
4.2 Deflections
4.3 Strains and crack development
5. Discussion of results
5.1 Deflections
5.2 Ultimate strength
5.3 Crack development
5.3.1 Cracks in concrete
5.3.2 Cracks in grc
6. Effectiveness of grc in controlling cracks in the concrete
7. Conclusions
8. Acknowledgements
9. References
Page
1
1
2
3
3
4
4
4
4
4
5
5
5
6
6
6
7
8
8
9
9
© CROWN COPYRIGHT 1980 Extracts from the text may be reproduced, except for
commercial purPoses , provided the source is acknowledged
Ownership of the Transport Research Laboratory was transferred from the Department of Transport to a subsidiary of the Transport Research Foundation on I st April 1996.
This report has been reproduced by permission of the Controller of HMSO. Extracts from the text may be reproduced, except for commercial purposes, provided the source is acknowledged.
CRACK CONTROL IN CONCRETE BEAMS BY SURFACE REINFORCEMENT WITH GLASS FIBRES
ABSTRACT
Tests were made on composite reinforced concrete beams incorporating a surface layer of glass reinforced cement (grc) to investigate the effect of local surface reinforcement on serviceability behaviour. The grc was in the form of 6 mm thick shallow channel sections moulded into the tension surface of the test beams, which were 3.5m long and were tested in four-point bending on a simply supported span of 3m. On beams which had been kept dry, the grc surfacing had a significant effect in reducing the size of cracks in the concrete at the design working load but had no effect on deflections; no visible cracks appeared on the surface of the grc until the steel had yielded, at a load well above the design ultimate. Some beams were subjected to artificial ageing before test, by immersing them in hot water to represent natural exposure of the grc for periods of up to 30 years. On these beams embrittlement of the grc surfacing rendered it virtually ineffective in controlling cracks in the concrete. The observed behaviour compares well with predictions made from a knowledge of the relevant material properties. Considerable improvements in performance could be expected from newer develop- ments in fibre cement composite technology and by using deeper sections for the surface reinforcement. A logical development, which is likely to have economic advantages, is to use suitable sections of rigid fibre rein- forced cement sheet as permanent formwork for reinforced concrete members.
1. INTRODUCTION
Fibre reinforced materials are sometimes used as permanent fofmwork for placing in situ concrete on
bridge decks. Glass reinforced cement in the form of stiffened panels has been used for this purpose on
several major bridges, for example the Kenfig viaduct on the M4 in South Wales and the Marron bridge
on the A66 in Cumbria. Advantages of such materials are that they are easily handled, ate resistant to
impact damage, are not affected by corrosion and bond well to the concrete. Although not usually taken
into account in the design, the surface skin provided by such permanent formwork can have a beneficial
effect on the resistance of the concrete to cracking.
The concept of using a thin layer of fibre reinforced cement sheet acting compositely with conventional
reinforced concrete to improve its serviceability performance has been developed by N J Dave at the
University of Salford 1,2. Tests on a variety of reinforced and pre-stressed beams and slabs have demonstrated
the efficacy of such materials as asbestos cement in reducing deflections and crack widths under working
loads when the tension zone of the concrete incorporates a surface layer of fibre reinforced cement sheet
(frc). Health hazards associated with the use of asbestos have stimulated the development of frc sheet
materials based on other types of fibre, such as polypropylene and glass. The most readily available alternative
to asbestos cement in sheet form is glass reinforced cement (grc) which is made from chopped strands of
alkali-resistant glass fibre in a matrix of cement paste or cement mortar.
Although grc has properties which are suitable in the short term, with prolonged exposure to moisture
it becomes brittle and therefore is likely to be less effective in controlling the onset and development of
cracking in the adjacent concrete. This Report describes a series of tests in which composite beams
incorporating grc surfacing were subjected to accelerated ageing to determine what changes in overall
structural performance might be expected due to prolonged exposure to the atmosphere.
2. MECHANISM OF CRACK SUPPRESSION BY GLASS FIBRE REINFORCEMENT
It is well known that concrete is weak in tension and develops visible cracks when the tensile strain teaches
about 0.01 per cent (100 microstrain). In a conventional reinforced concrete member subjected to tensile
or flexural loading a well defined pattern of cracking develops in which the crack spacing on the surface
depends mainly on the effective concrete cover 3. Once the crack pattern is established, further increase of
load results in a widening of the individual cracks by an amount roughly proportional to the strain in the
steel reinforcement at the cracked section. For a normal structural grade concrete with high yield deformed
bar reinforcement, the crack spacing might be of the order of 100 mm and the crack width would probably
reach about 0.2 mm at the onset of yielding of the steel.
Recent design codes 4,5 restrict permissible crack widths at serviceability limit states, the permissible
width depending on the severity of the environment. Although there is now some doubt about the validity
of the concept of allowable crack widths insofar as liability to corrosion of internal reinforcement is concerned 6,
any move towards reducing the susceptibility to cracking of concrete in tension is likely to have beneficial results.
By a proper choice of materials, the use of surface reinforcement intimately bonded to the concrete
offers the prospect of providing a surface layer which will accommodate large strains without fracture. I f
there is no slip between the surfacing and the underlying concrete it should then be possible theoretically to
replace isolated wide cracks in the concrete by a larger number of finer cracks for a given level of effective
strain. The tests reported here were carried out to study the effectiveness of glass reinforced cement in
achieving this objective, taking account of changes in its properties due to prolonged exposure to moisture.
If grc is subjected to wetting and drying, the strain at failure reduces with time until eventually it falls
to that of the unreinforced matrix 7. Results of tests at the Building Research Establishment 8 have shown
the following changes in direct tension properties of test specimens subjected to natural weathering.
UTS - N/mm 2
Matrix cracking stress (BOP) - N/mm 2
Young's modulus - GN/mm 2
20 years 28 days 1 year 5 years i0 years (estimated)
14--17
9 - 1 0
20--25
11-14
9 - 1 0
20--25
7 - 8
7 - 8
25-32
7--8*
6--8*
27--30*
5 - 7
5 - 7
25-32
* Difficulty in testing because of brittle behaviour.
2
It has been found empirically that similar changes can be produced to an accelerated timescale by
immersing the grc in hot water. The following approximate conditions apply:
One month in water at 60°C is equivalent to 10 years natural weathering,
Three months in water at 60°C is equivalent to 30 years natural weathering.
Changes in strain capacity would be expected to have a significant influence on the effectiveness of surface
reinforcement and this was one of the main points of the tests reported here.
3. TESTS ON REINFORCED CONCRETE BEAMS WITH GRC SURFACE REINFORCEMENT
3.1 Description of test beams
Four-point bending tests were carried out on two types of reinforced concrete beam as shown in
Figure 1. Both types were 3.5m long and over the central portion were reinforced on the tension side only
by two 12 mm diameter deformed bars ('Torbar'). Shear reinforcement was included towards each end of
the beam.
Type R was a conventional reinforced concrete beam with an effective depth to the reinforcement of
214 mm and with an effective cover* of 36 mm. Type G had a thin channel section of glass reinforced
cement cast into the tension surface. In this case, although the effective depth to the internal reinforcement
remained the same, the reinforcing bars were in direct contact with the grc, resulting in a slight reduction in
overall depth. Some preliminary tests had been made on an intermediate type of beam which also had grc
surface reinforcement but had the same overall depth as the conventional reinforced concrete beam. These
tests, however, were inconclusive because the relative levels of the internal reinforcement and the top of the
grc channel tended to induce premature failure. The results of these preliminary tests are not reported here,
although some further mention of them is made in Section 5.1.
The grc contained 5 per cent by weight of alkali-resistant glass fibres and was supplied by the Building
Research Establishment in the form of shallow channel sections, 3.5m long, 50 mm deep and 120 mm wide.
They were made from fiat sheet produced by the spray suction method, cut into strips while wet and draped
over wooden formers before being left to harden. The finished thickness of the grc sheet was nominally
6 mm. After delivery to TRRL the grc channels were kep~ for about nine months in an air conditioned
laboratory (20°C and 65 per cent RH) before use. Each channel section was then put in the bot tom of a
wooden mould and the steel reinforcing cage placed in position on the grc channel. Concrete was placed in
the mould in the usual way, using pokers for compaction. The beams were cast in an unheated building,
demoulded after one day, and then covered with wet sacking and polyethylene sheeting and left for three
months before being tested. After the three months curing period, some of the grc reinforced beams were
subjected to artificial ageing by immersing them for periods of up to three months in water at 60°C. This
was done in a thermostatically controlled tank at Pilkington Bros. Research Laboratories. Samples of the
grc channel sections were also subjected to artificial ageing at the same time. After soaking in hot water
the beams were allowed to dry in the laboratory atmosphere for about 2 weeks before being tested.
* ie distance from centre of reinforcing bar to surface of beam.
3.2 Mix proportions and nominal strength properties
The concrete was made from Thames Valley flint gravel aggregate with a maximum size of 19 mm.
The mix proportions and nominal and achieved strengths are given in Table 1. The method of preparing
and mixing the concrete was broadly as described in reference 9. This involved oven-drying all aggregates
before batching to ensure that the design water/cement ratio was achieved on all test beams, with minimum
variability of concrete strengths throughout. That this was achieved is shown by the extremely low values
of coefficient of variation obtained on 28 day cube and flexural strengths, 2.1 per cent and 4.0 per cent
respectively. Table 1 also includes design and measured strengths for the deformed bar longitudinal rein-
forcement and mean strength values for the grc surfacing.
3.3 Method of test
Each test beam was simply supported on a span of 3m and load was applied to the upper surface at
the middle third points through a single compression jack acting through a steel load-spreading beam to give
four-point bending, as shown in Figure 1.
Vertical deflections were measured at the mid-span position.
Longitudinal strains were measured at a typical cross-section in the constant moment region to
determine strain distributions at various stages of loading, using C and CA type 'portal' gauges with 100 mm
gauge length. Demec points were attached to a vertical face of the beam at 50 mm intervals at the level of
the internal reinforcement and just above the top of the grc channel to enable crack opening displacements
to be measured at these two levels. Load was applied in increments under deflection control until the
beam failed by excessive yielding of the reinforcement.
4. RESULTS OF TESTS
4.1 Ultimate failure loads
Loads at failure for each type of specimen are given in Table 2, for comparison with design loads and
calculated failure loads. Design loads are based on normal reinforced concrete theory, assuming that both
concrete and grc are cracked below the neutral axis and that tensile forces are resisted by the internal steel
reinforcement only. The design ultimate load is calculated from the nominal characteristic strengths of the
concrete and steel (Table 1). The design working load is derived from the design ultimate by arbitrarily
assuming an overall safety factor of 1.7.
The calculated failure loads are based on mean test values for the strength of the concrete and steel
(Table 1) and on typical stress-strain curves and tensile strengths for the grc derived from tests at BRE
(Figure 6). Unaged grc has an effective tensile strength at strains well beyond the yield strain of the steel
reinforcement and is assumed to carry its share of the flexural loading. The failure strain of artificially
aged grc is very much lower and in this condition the grc does not contribute to the flexural strength of
the beam.
4.2 Deflections
Typical load-deflection curves for each of the four types of test beam are given in Figure 2 and the
central deflection of each beam at the design working load is given in Table 2. There is not much difference
between them, showing that the presence of the grc channel had little overall effect on the flexural
stiffness of the beam. 4
4.3 Strains and crack development
Longitudinal strains measured with the portal gauges showed a generally linear distribution over the
depth of each test beam, thus confirming that there was a good bond between the concrete and the grc
right up to failure and that for subsequent analysis of the results it was reasonable to assume full composite
action, with plane sections remaining plane and with a gradual shift of the neutral axis towards the
compression surface. Typical examples of the strain distribution are shown in Figure 3. The non-linearity
in Figure 3(a) is probably due to the influence of a crack developing in the concrete near the lowermost
gauge position.
Crack width measurements derived from Demec readings are illustrated in Figure 4 a t two levels
(a) at the level of the internal reinforcement and (b) just above the level of the top of the grc channel.
These measurements are plotted as apparent crack widths, since the total displacement over the gauge
length included the effect of some strain in visually uncracked concrete and grc as well as the displacement
due to the opening of existing cracks. For ordinary reinforced concrete, errors due to neglecting concrete
strains are negligible. In the case of the beam without grc and the beams with grc which had undergone
artificial ageing, discrete cracks at the level of the reinforcement became visible at a load of 15-18 kN.
With the beams containing unaged grc, however, no cracks were visible in the grc until the steel had yielded,
even though the measured strains were well in excess of the probable cracking strain of the cement mortar
matrix. This point is discussed further in Section 5.3.2. At the higher level, cracks were observed in the
concrete at loads between 15 and 18 kN for all four types of specimen but the crack opening under a given
load was rather less with the unaged grc specimens. Typical crack patterns at the design ultimate load are
illustrated in Figure 5. The development of cracking for each individual test beam is summarised in Table 3.
5. DISCUSSION OF RESULTS
5.1 Deflections
A comparison of the central deflections (Figure 2 and Table 2) shows very little difference between
beams with and without grc surfacing, whether artificially aged or not. This observation apparently conflicts
with the results of comparable tests at Salford University on reinforced concrete beams with asbestos
cement surfacing 2, which indicated that the inclusion of surface reinforcement should result in reduced
deflections up to and beyond the design working load.
The apparent differences in behaviour between the TRRL and Salford tests can be explained by
differences in the stress-strain relationships for grc and asbestos cement. A typical stress-strain curve for
grc is reproduced in Figure 6. This shows a steep initial slope, corresponding to the uncracked sheet,
a well-defined 'bend over point' and a secondary slope which is at least an order of magnitude less than
the initial slope. It is unlikely, therefore, that the grc will have any appreciable influence on the overall
stiffness of the beam once the effective strain in the concrete has exceeded the bend over point at about
300 microstrain. This occurs well below the design working load. With asbestos cement, however, the
slope of the stress-strain curve is quite different 10. The initial slope is maintained up to much higher strains,
so that at the design working load the asbestos cement surfacing would still be expected to have an influence
on the stiffness of the beam.
A further minor difference between the TRRL and Salford tests is that the concrete had rather
different compressive strengths (mean values 57 and 45 N/mm 2 respectively). This would tend to make
5
any contribution of the surfacing to the overall stiffness of the beams less in the case of the TRRL tests,
with the higher strength concrete. (In some preliminary tests at TRRL which involved grc reinforcement
of similar beams with a lower grade of concrete - mean cube strength 46 N/mm 2 - there was some
indication of slightly reduced deflections on the beams containing grc.)
5.2 Ultimate strength
The failure loads agreed closely with those predicted from normal reinforced concrete theory (Table 2).
There was little difference between beams with and without grc, although the beams with unaged grc were
about 10 per cent stronger until the grc fractured. Final failure always occurred by yielding of the steel
reinforcement, which meant that once yield developed, the fracture strain of the grc was exceeded even
for the beamswhich had not been artificially aged. In the latter case this usually resulted in a single
major crack right through the grc channel. With the artificially aged beams several major fractures of
the grc had occurred by this time (see Section 5.3.2 below). Once the grc channel had fractured, each
beam behaved as a normal reinforced concrete beam approaching collapse.
5.3 Crack development
See Table 3 and Figure 4.
5.3.1 Cracks in concrete. For the beam without surface reinforcement the development of flexural
cracks followed the usual sequence, in which a fairly regular pattern of cracks developed in the constant
moment span, with an average spacing of about 80 mm by the time the design ultimate load was reached.
This compares with a value of 83 mm from the empirical formula s m = 2.3 Ce, derived from large scale
beam tests at BRE 11. Here s m = average crack spacing, c e = effective cover as def'med by the distance
from the centre of the bar to the surface.
For the beams with grc surface reinforcement cracks became visible in the concrete at the top of the
upstanding leg of the grc channel section before the design working load was reached. The average crack
spacing was much the same as for the beam without grc, whether the beams had been artificially aged
or not. It was not possible to say how far cracking of the concrete extended downwards behind the grc
surfacing. In all cases cracks extended upwards with increase in load, with a corresponding shift in the
position of the neutral axis at the cracked section. Before artificial ageing, the grc had some effect in
reducing the width of cracks in the concrete, the reduction being about 33 per cent at the design working
load and 26 per cent at the design ultimate load (Figure 4(b)). After artificial ageing (one month or three
months) there was little or no effect.
If it is assumed that there is some local debonding between the concrete and the grc in the vicinity
of wide cracks in the concrete'it may be inferred that such cracks would extend down at least to the level
of the reinforcement. The apparent crack widths of Figure 4(a) may then be taken to approximate to
actual crack widths in concrete hidden by the grc. At the design working load the unaged grc has the effect
of reducing the maximum apparent crack widths at the level of the reinforcement by about 30 per cent
from 0.095 mm to 0.065 mm. This compares with a design value for reinforced concrete of 0.06 mm for
short term loading using the formula 2.3a_ r e 1 given in BS 54005, or 0.075 mm for average crack width
(0.15 mm maximum) given by 2.3c e e I 1 ~. In these two formulae e I is the tensile strain in the reinforcing
steel and acr is the perpendicular distance from the point considered to the surface of the nearest longitudinal
bar. After artificial ageing the grc itself developed visible cracks similar in magnitude and spacing to those
which developed in the reinforced concrete without surface reinforcement.
6
5.3.2 Cracks in gre. For the beams which had not undergone artificial ageing, no cracks were visible
in the grc channel until the steel had yielded at an applied load of about 45 kN, nearly 30 per cent above
the design ultimate load. However some quite large strains had been measured by this time, indicating
that cracking of the cement matrix must have occurred (Figures 6 and 7).
Theoretical predictions of matrix cracking strain for 2-dimensional random dispersions of chopped
glass fibre in a cement paste 12 indicate strains of the order of 0.02 per cent (200 micro strain) for a
composite containing 5 per cent by weight of alkali-resistant fibres, slightly higher than the failure strain
of a cement paste without fibres. For unaged grc the theory of Aveston, Cooper and Kelly 13 predicts
that cracking should be fully developed at strains of about 0.13 per cent, with an average crack spacing of
about 1 -2 mm. Such values are consistent with the typical shape of stress-strain curves obtained in direct
tension tests carried out by BRE on samples cut from grc boards produced by the spray suction technique
(Figure 6). At higher strains, the load is carried entirely by the fibres and final fracture occurs when the
overall strain reaches about 1 per cent. Figure 6 also indicates how the failure strain reduces after exposure
to natural weathering.
These observations may be related to typical strain measurements taken on the test beams.
Figure 7 shows apparent strain, as measured across Demec points in the vicinity of a crack in the concrete,
plotted against applied load. This suggests that before artificial ageing micro-cracking of the grc matrix
probably started at about the same time as discrete cracks began to form in the concrete and that multiple
cracking of the grc then developed, becoming complete at about the design working load. At this stage
the crack spacing in the grc is likely to have been of the order of 2 mm. Further increase in applied load
would then result in opening up of these multiple fine cracks until eventual failure by fibre pull out at
quite high strains. An estimate of the probable crack width at the onset of yielding in the steel reinforcement
can be obtained from Figure 7. Assuming yield at 45 kN applied load, the integrated crack opening
displacements over the 50 mm gauge length ('apparent crack width') were 0.17 mm. With a crack spacing
of about 2 mm this means that the average crack width was only about 0.007 mm, well below the threshold
of visible cracking.
After artificial ageing, cracking of the grc surface was observed at loads below the design working
load and was consistent with fracturing of the grc at strains of less than 0.02 per cent,both for one month and
for three months ageing periods. The resulting cracks had similar width and spacing to those observed on
the beam without surface reinforcement. Direct tension tests were attempted on some samples sawn from
aged and unaged grc sections in order to get some idea of the effect of the artificial ageing on the basic
properties of the grc. Although the testing machine available was not entirely suitable, the results do give
an indication of the extreme brittleness of the grc after soaking in hot water. Based on the mean of about
12 tests in each case, the results were as follows:-
Unaged
UTS - N/mm 2 10.2
Indicated strain at failure - per cent 3.1
1 month at 60°C 3 months at 60°C
6.0 3.7
0.019 0.011
7
J
6. EFFECTIVENESS OF GRC IN CONTROLLING CRACKS IN THE CONCRETE
It may be concluded that initially the grc was effective in controlling cracks in the concrete in the vicinity of
the internal steel reinforcement. At the design ultimate load, crack widths were reduced to values not much
greater than those corresponding to the design working load of beams without surface reinforcement. With
deeper sections of grc surfacing, extending upward to the neutral axis, for example, the surface reinforcement
is likely to be even more effective 2. After periods of artificial ageing ranging from one to three months
immersion in hot water at 60°C (equivalent to 10 to 30 years natural exposure), the effectiveness was
• practically eliminated, however, as the grc became extremely brittle and the strain to failure dropped below
that expected of the unreinforced matrix.
It is arguable that the most important period when cracks need to be controlled is during the early
life of a structure when creep and shrinkage rates are at a maximum and the concrete has not yet reached
its fully developed strength. From this point of view surface reinforcement in the form of grc permanent
shuttering could be quite effective for up to 5 years or even more. It would obviously be preferable,
however, to maintain its effectiveness for a longer period. Recent developments in alkali-resistant glass
fibres 14 should lead to improved performance here; the strain at failure of the new Cem-Fil 2 fibre after
30 years exposure is claimed to be nearly four times that of the original Cem-Fil. Other promising
possibilities include new developments in polyolefin fibres 15.
As far as protecting the internal steel from corrosion is concerned the presence of a cement-rich dense
fibrous surface layer is likely to increase the effectiveness of the concrete cover to the internal reinforcement
and provide some protection against local impact damage. Results of exposure tests on reinforced concrete
at BRE 16 have shown that the susceptibility to corrosion of internal steel is very much reduced by an
increase in cement content and a decrease in permeability, both of which lead to a marked reduction in the
rate of carbonation of the concrete surrounding the steel. Although it remains to be proved conclusively
by long-term exposure tests it is reasonable to suppose that placement of reinforcing bars directly on the
cement-rich fibre reinforced layer, when it is to be used as permanent formwork, will still give adequate
protection for the steel against corrosion. Provided that gross cracking does not develop such protection
may be maintained even when the grc has become embrittled by exposure to moisture. If this is so there
are potential economic benefits to be gained, resulting from a considerable simplification of steel fixing
for the reinforcing cage in practical applications such as deck Slabs for concrete bridges. There is also a
prospect of reducing the amount of reinforcing steel needed for crack control in such structures.
7. CONCLUSIONS
Tests on reinforced concrete beams having a thin layer of glass reinforced cement as surface reinforcement
on the tension side have confirmed that the presence of such a layer can have a significant effect on the
cracking behaviour under load.
Within the first few years of casting the concrete the effectiveness of the grc is such as to reduce the
width of cracks in the concrete and to eliminate yisible surface cracking in the zone covered by the grc,
fight up to the ultimate failure load. After prolonged exposure to natural weathering, the grc becomes
less effective because of a marked reduction in fracture strain. This results in multiple fractures similar
to the cracking which develops in tension zones of ordinary reinforced concrete beams.
8
Whilst there would be benefits during the first few years of service from using grc as permanent
formwork in practical structural applications, it would obviously be much better to use fibre reinforced
materials which retained their extensibility over a longer time scale. There are good prospects of achieving
this by using some of the new developments in fibre reinforced cement composites. Among these are grc
based on improved alkali-resistant glass fibres, recently introduced, and fibre reinforced sheets based on
fibrillated polypropylene film.
Improved long-term effectiveness of surface reinforcement could lead to a reduction in the amount
of reinforcing steel needed for crack control in reinforced concrete structures, with consequent economic
advantages.
8. ACKNOWLEDGEMENTS
The work described formed part of the research programme of the Bridge Design Division (Division Head:
Dr G P Tilly) of the Structures Department of TRRL. All the testing was done by Mr J W Grainger who
also carried out a preliminary analysis of the results; his help in preparing this Report is gratefully .
acknowledged. Thanks are due to Mr J F Ryder of the Building Research Establishment for the supply of
grc channel section and to Messrs D A Warrior and B A Proctor of Pilkington Bros. for the use of their
facilities for the artificial ageing tests.
9. REFERENCES
. DAVE, N J. Fibre reinforced cement (frc) composite concrete construction - a new approach.
RILEM Symposium on Fibre Reinforced Cement and Concrete. London, 1975. Fol 2, p 615.
. DAVE, N J and J D PENNINGTON. Fibre reinforced cement composite concrete construction
employing asbestos cement as surface reinforcement. Conference on fibre reinforced materials:
design and engineering applications. Institution of Civil Engineers. London, 1977. p109.
. BROMS, B B. Crack widthand crack spacing in reinforced concrete members. Journal o f the
American Concrete Institute, 1965 (62), No. 10, p 1237.
4. BRITISH STANDARDS INSTITUTION. The structural use of concrete CP110, 1972.
. BRITISH STANDARDS INSTITUTION. Steel, concrete and composite bridges. Part 4 Code of
Practice for design of concrete bridges. BS 5400, 1978.
. BEEBY, A W. Corrosion of reinforcing steel in concrete and its relation to cracking. The Structural
Engineer, March, 1978, Vol. 56a, No. 3, p 77.
7. BUILDING RESEARCH STATION. A study of the properties of Cem-Fil OPC composites.
Building Research Establishment Current Paper CP38/76. 1976.
. BUILDING RESEARCH STATION. Properties of grc: 10-year results. Building Research Establishment
Information Paper IP36/79. November, 1979.
.
10.
11.
12.
13.
14.
15.
16.
FRANKLIN, R E and T M J KING. Relations between compressive and indirect tensile strengths of
concrete. Department o f the Environment, RRL Report LR 412. Crowthorne, 1971 (Road Research Laboratory).
ALLEN, H G. Tensile properties of seven asbestos cements. Composites. June, 1971, p 98.
ILLSTON, J M and R F STEVENS. Long-term cracking in reinforced concrete beams. Proceedings o f the Institution o f Civil Engineers, Vol. 53. December, 1972, p 445.
ALI, M A, A J MAJUMDAR and B SINGH. Properties of glass fibre cement - the effect of fibre
length and content. Journal o f Materials Science, Vol. 10, 1975, p 1732.
AVESTON, J, G A COOPER and A KELLY. Single and multiple fractures. NPL Conference on the
properties of fibre composites. IPC Science and Technology Pres& 1971, p 15.
FIBREGLASS LTD. Cem-FIL 2 alkali resistant glass fibre. Cem-FIL Product Leaflet. October, 1979.
HANNANT, D J and J J ZONSVELD. Polyol~fin fibrous networks in cement matrices for low cost
sheeting. Royal Society discussion meeting. New fibres and their composites. May, 1978.
GRIMER, F J. The durability of steel embedded in lightweight concrete. Building Research Station Current Papers. Engineering Series 49, 1967.
10
TABLE 1
Material properties
Concrete
Reinforcing steel
Grc
Mix proportions by weight:-
19-10 mm aggregate 2.02
10 - 5 mm aggregate 0.96
< 5 mm aggregate 1.83
cement 1.00
water/cement ratio 0.46
Slump - mm
28 day cube strength - N/mm 2
28 day flexural strength - N/mm 2
12 mm dia Torbar yield stress - N/mm 2
Sand/cement mortar with
25% by weight of sand and
5% by weight of chopped Cem-FIL
AR glass fibre, sprayed and dewatered
Tensile strength at 28 days - N/mm 2
after 5 yrs weathering
after 10 yrs weathering
Strain at failure at 28 days
after 5 yrs weathering
after 10 yrs weathering
Characteristic design value
60-180
30
460
Mean test value
96
57
6
489
16 t
7t 7t
0.85% t
O.os t
t BRE data
11
TABLE 2
Overall behaviour of test beams
Beam No.
1
2
3
4
5
6
7
Type (see Figure 1)
R (no grc)
G O
G O
G 1
G 1
G 3
G 3
Design working load
kN
20.5
20.5
20.5
Design ultimate load
kN
35
35
35
20.5 35
20.5 35
20.5 35
20.5 35
Load at failure kN
Calculated Test
45 45.0
49.5 48.7
49.5 49.1
45 37.8 t
45 37.6 t
45 43.5
45 43.7
Central deflection at design
working load mm
5.0
4.7
4.7
5.2
6.0
4 . 5
5.0
t Test terminated before complete yielding of steel reinforcement to avoid local shear failure at supports.
12
oO
.J oo
~ I~1 ' o I ~.~
t ~
0
,-,,,i
o oo N 2 ~o o~
o . ~ = x~ ~ i I ~ ,'~ , ~
~ ~ ~_~, ,
o ' ~ ~ , o . . . . . ~ o o
. I ~° i
.~ ~ ~ I
o
N ~
+ - . ~ -
° I oo ~ oo
o 0 o o , 0
I ~ - ~ ~
z
I
"F.
L ) . . . . . .
6 Z
0
0
r ~
0
, .o
o
- t - -
z
II II
13
129 Torbar
2, I I
30 ~ - - ~ 3 0
Type R
1T F P4
io_o / / / / ' ~ ( /
6mm thick GRC
Go= No artificial ageing G 1 = Immersed in water at 60°C for 1 month G3= Immersed in water at 60°C for 3 months
Type G (a) BEAM SECTIONS
1000 I
1000 -=' = 1000
Appliedjir load I
l
I1~ II//lllllllllllll'~lll~'i'j*'~'q~Al~t I ' I ~ Crack width rl-' " - ~
Strain gauged measurements (DEMEC) section
Fig. 1
(b) METHOD OF LOADING AND INSTRUMENTATION FOR TYPE G BEAMS
All dimensions in mm
DETAILS OF TEST BEAMS AND LOADING ARRANGEMENT
50
40
30
"O
O
._~ O. O.
20
10
• W
° " ~ l i , N ©
No GRC GRC unaged GRC aged 1 month at 60°(3 GRC aged 3 months at 60°(3
ol m m m 0 5 10 15
Calculated failure loads:-
With unaged GRC
With aged GRC or no GRC
Design ultimate load
Design working load
20
Central deflection (mm)
Fig. 2 TYPICAL LOAD-DEFLECTION CURVES
A
E E
v
C.}
~ L
o
E o
N - -
¢ . ) ¢--
E3
=
O A
Strain (per cent)
--0.10 0 0.10 0.20 O
50
100
150
200
250
(a) TYPE R -- NO GRC
: A t onset of cracking in concrete
O A t design working load
"- A t design ult imate load 1 A
E E
v
== Q _
o
E o
N - -
C -
Strain (per cent)
- 0 .10 0 0
5 0 -
100 -
150 -
200
226
25O
0.10 !
0.20 I
; [
\ :'
(b) TYPE G o - UNAGED
A
E E
v
¢. ) ~ 3
t
Q . o
E o
4 - -
G~ { . ) t "
° ~
E3
Strain (per cent)
- 0 . 1 0 0 0.10 0.20
Gauge positions 50 as for (b)
100 -
150 -
200 -
226
(c) TYPE G1 - AFTER 1 MONTH IN HOT WATER A T 60°C
- 0 0
-~ 50
E
.~= 100 Q .
9 E 2 "- 150
t -
a
200
226
Strain (per cent)
10 0 0.10 0.20 ' '
Gauge positions as for (b)
w
(d) TYPE G 3 - AFTER 3 MONTHS IN HOT WATER A T 60°C
Fig. 3 T Y P I C A L M E A S U R E D S T R A I N D I S T R I B U T I O N S
.~°
0 - - ~ " 0
= - o ~ v ._o
c- t-
O • " ~
0 0
c, t- O ._o ~
E E ~"
n" ~ n -
'- E ," E b • ~-- 0 .~-- 0 ~
~ E ~ E -~
5 ~ 7 ~ of- "~ "- ~ '~" c'~ ~
0 -~ "~ z N ~
~ z
.~g
" I , ~.~. I ~ . ~ . . . ~ " 0 0 0 0 0 0
(N>I) peol Pa!lddV
x DEMEC gauge points I
500
~( X X /X ~ X X X /
! , /x X ( X X X X
Type R - No GRC Beam
X X X
X X X
15 x:x xr.x xlx, xlx:x (/////// l / lX//x /S///~
Type G - Before artificial ageing
Loading ), ~ )oint
Type G - After
x~x ~~ ,~/x/x /kA//b~A/Jl,
month in hot water at 60°C ._J.
j xJ xSx xlX x Type G - After 3,months in hot water at 60°C
..i=.d Fig. 5 TYPICAL PATTERNS OF VISIBLE SURFACE CRACKING AT
D-ESIGN ULTIMATE LOAD
15
E E z
==
10
Ultimate tensile strain after natural weathering:-
2 years
6 months
5 years L !
1 year
28 days
Theoretical strains in unaged GRC (Ref. 12) F
Matrix cracking
t ~ I I
Development
o~ mult i~e~ 0.2 crackin~l I
1000 2000
Crack opening
0.4 0.6 I I I I I
3000 4000 5000 6000 7000
Strain (microstrain)
0.8 (per cent) I J
8000
Fig. 6 STRESS-STRAIN CURVES FOR GRC IN TENSION AFTER NATURAL WEATHERING (From Ref. 7)
± I Beam with unaged GRC
~ Beam with fully aged GRC (3 months at 60°C)
A
z
o -o
Q. <
40
30
20
10
/ /
/ /
Design ultimate load
1st Crack seen in aged GRC
Approx. fracture strain of I ful ly aged GRC
Approx. cracking strain of GRC
- - Design working load
Development of multiple cracking in GRC
1 ,~t Crack opening in GRC (unaged)
Approx. fracture strain of unaged GRC
o l l i I , , I I I 0 0.2 0.4 0.6 0.8 1.0
Apparent strain (per cent)
Fig. 7 MEASURED STRAINS IN GRC AT LEVEL OF REINFORCEMENT
(1112) Dd0536380 1,400 7/80 HPLtdSo ' ton G1915 PRINTED IN ENGLAND
ABSTRACT
Crack control in concrete beams by surface reinforcement with glass fibres: K D RAITHBY: Department of the Environment Department of Transport, TRRL Laboratory Report 947: Crowthorne, 1980 (Transport and Road Research Laboratory). Tests were made on com- posite reinforced concrete beams incorporating a surface layer of glass reinforced cement (grc) to investigate the effect of local surface reinforcement on serviceability behaviour. The grc was in the form of 6 mm thick shallow channel sections moulded into the tension surface of the test beams, which were 3.5m long and were tested in four-point bending on a simply supported span of 3m. On beams which had been kept dry, the grc surfacing had a significant effect in reducing the size of cracks in the concrete at the design working load but had no effect on deflections; no visible cracks appeared on the surface of the grc until the steel had yielded, at a load well above the design ultimate. Some beams were subjected to artificial ageing before test, by immersing them in hot water to represent natural exposure of the grc for periods of up to 30 years. On these beams embrittlement of the grc surfacing rendered it virtually ineffective in controlling cracks in the concrete. The observed behaviour compares well with predictions made from a knowledge of the relevant material properties. Considerable improvements in performance could be expected from newer developments in fibre cement composite technology and by using deeper sections for the surface reinforcement. A logical development, which is likely to have economic advantages, is to use suitable sections of rigid fibre reinforced cement sheet as permanent formwork for reinforced concrete members.
ISSN 0305-1293
ABSTRACT
Crack control in concrete beams by surface reinforcement with glass fibres: K D RAITHBY: Department of the Environment Department of Transport, TRRL Laboratory Report 947: Crowthorne, 1980 (Transport and Road Research Laboratory). Tests were made on com- posite reinforced concrete beams incorporating a surface layer of glass reinforced cement (grc) to investigate the effect of local surface reinforcement on serviceability behaviour. The grc was in the form of 6 mm thick shallow channel sections moulded into the tension surface of the test beams, which were 3.5m long and were tested in four-point bending on a simply supported span of 3m. On beams which had been kept dry, the grc surfacing had a significant effect in reducing the size of cracks in the concrete at the design working load but had no effect on deflections; no visible cracks appeared on the surface of the grc until the steel had yielded, at a load well above the design ultimate. Some beams were subjected to artificial ageing before test, by immersing them in hot water to represent natural exposure of the grc for periods of up to 30 years. On these beams embrittlement of the grc surfacing rendered it virtually ineffective in controlling cracks in the concrete. The observed behaviour compares well with predictions made from a knowledge of the relevant material properties. Considerable improvements in performance could be expected from newer developments in fibre cement composite technology and by using deeper sections for the surface reinforcement. A logical development, which is likely to have economic advantages, is to use suitable sections of rigid fibre reinforced cement sheet as permanent formwork for reinforced concrete members.
ISSN 0305-1293