strength.txt

3	Properties of Aggregate      Since at least three-quarters of the volume of concrete is occupied by aggregate, it is  not surprising that its quality is of considerable importance. Not only may the  aggregate limit the strength of concrete, as weak aggregate cannot produce strong  concrete, but the properties of aggregate greatly affect the dura
cific  gravity, hardness, strength, physical and chemical stability, pore structure, colour,  etc. On the other hand, there are some properties possessed by the aggregate but absent  in the parent rock: particle shape and size, surface texture, and absorption. All these  properties may have a considerable influence on the quality of the concrete e
al aggregate.   Tests to date have shown the material capable of producing a concrete  with compressive strengths as high as 50 MPa (7000 psi) at 28 days. There will,  obviously, be problems with variations in the composition of the raw ash, and the long-term  durability characteristics of the material have yet to be determined, although results  
to date look promising. It is not envisaged that this material will be of use in high strength  structural concrete; however, it may be suitable for low strength concrete where currently  high-grade aggregates, of a quality far superior to that really required, are being used.  Previous use of this type of concrete is mentioned on Page 609.      S
ibe, and it is, therefore, convenient to define certain geometrical characteristics  of such bodies.   Roundness measures the relative sharpness or angularity of the edges and corners of a  particle. Roundness is controlled largely by the strength and abrasion resistance of the  parent rock and by the amount of wear to which the particle has been 
d	 texture properties encountered in practice  Some other approaches are reviewed by Ozol.   It seems that the shape and surface texture of aggregate influence considerably the  strength of concrete. The flexural strength is more affected than the compressive  strength, and the effects of shape and texture are particularly significant in the case 
 of high strength concrete. Some data of Kaplan's are reproduced in Table 3.5 but this  gives no more than an indication of the type of influence, as some other factors may  not have been taken into account. The full role of shape and texture of aggregate in the  development of concrete strength is not known, but possibly a rougher texture results
n's experimental results, however, do not confirm  that the surface texture is a factor.    Fig. 3.4. The relation between the angularity number of aggregate and the compacting  factor of concrete made with the given aggregate      Bond of Aggregate    Bond between aggregate and cement paste is an important factor in the strength of  concrete, esp
ecially the flexural strength, the full role of bond being only now  realized. Bond is due, in part, to the interlocking of the aggregate and the paste owing  to the roughness of the surface of the former. A rougher surface, such as that of  crushed particles, results in a better bond; better bond is also usually obtained with softer,  porous, and
actured particles, however, might suggest that the aggregate is too weak. Because it depends on the  paste strength as well as on the properties of aggregate surface, bond strength increases with the  age of concrete; it seems that the ratio of bond strength to the strength of the paste increases with age  Thus, providing it is adequate, the bond 
strength per se may not be a controlling factor in the strength of  concrete. However, in high strength concrete there is probably a tendency for the bond strength  to be lower than the tensile strength of the cement paste so that preferential failure in bond takes place.  The problem of failure of concrete is discussed more fully in Chapter 5.   
	Strength of Aggregate    Clearly, the compressive strength of concrete cannot significantly exceed that of the  major part of the aggregate contained therein, although it is not easy to state what is  the strength of the individual particles. Indeed, it is difficult to test the crushing strength  of the aggregate by itself, and the required inf
ormation has to be obtained usually from  indirect tests: crushing strength of prepared rock samples, crushing value of bulk aggregate,  and performance of aggregate in concrete.   The latter simply means either previous experience with the given aggregate or a trial  use of the aggregate in a concrete mix known to have a certain strength with pre
viously  proven aggregates. If the aggregate under test leads to a lower compressive strength of  concrete, and in particular if numerous individual aggregate particles appear fractured  after the concrete specimen has been crushed, then the strength of the aggregate is  lower than the nominal compressive strength of the concrete mix in which the 
aggregate  was incorporated. Clearly, such aggregate can be used only in a concrete of lower  strength. This is, for instance, the case with laterite, a material widely spread in  Africa, South Asia and South America, which can rarely produce concrete stronger than  10 MPa (1500 psi).   Inadequate strength of aggregate represents a limiting case a
s the properties of  aggregate have some influence on the strength of concrete even when the aggregate by  itself is strong enough not to fracture prematurely. If we compare concretes made with  different aggregates we can observe that the influence of aggregate on the strength of  concrete is qualitatively the same whatever the mix proportions, a
nd is the same  regardless of whether the concrete is tested in compression or in tension. It is  possible that the influence of aggregate on the strength of concrete is due not only to  the mechanical strength of the aggregate but also, to a considerable degree, to its  absorption and bond characteristics.   In general, the strength and elasticit
y of aggregate depend on its composition, texture  and structure. Thus a low strength may be due to the weakness of constituent grains or  the grains may be strong but not well knit or cemented together.   The modulus of elasticity of aggregate is rarely determined; this is, however, not  unimportant as the modulus of elasticity of concrete is gen
erally higher the higher the  modulus of the constituent aggregate, but depends on other factors as well. The modulus  of elasticity of aggregate affects also the magnitude of creep and shrinkage that can be  realized by the concrete.   A good average value of the crushing strength of aggregate is about 200 MPa (30 000 psi)  but many excellent agg
regates range in strength down to 80 MPa (12 000 psi). One of the  highest values recorded is 530 MPa (77 000 psi) for a certain quartzite. Values for other rocks  are in Table 3.6. It should be noted that the required strength of concrete is considerably  higher than the normal range of concrete because the actual stresses at the points of contac
t of  individual particles within the concrete may be far in excess of the nominal compressive  stress applied.    Table 3.6: Compressive Strength of American Rocks Commonly Used as Concrete Aggregates     On the other hand, aggregate of moderate or low strength and modulus elasticity can be  valuable in preserving the durability of concrete. Volu
ists between the strength and modulus of  elasticity of different aggregates. Some granites, for instance, have been found to  have a modulus of elasticity of 45 GPa (6.5 x 10'6 psi), and gabbro and diabase a modulus  of 85.5 GPa (12.4 x 10'6 psi), the strength of all these rocks ranging between 145 and  170 MPa (21000 to 25 000 psi). Values of th
e modulus in excess of 160 GPa (23 x 10'6 psi)  have been encountered.   A test to measure the compressive strength of prepared rock cylinders was prescribed by  BS 812 : 1967. In this test, a 25.4 mm (1 in.) diameter cylinder 25.4 mm (1 in.) high, is  used, and the nominal crushing strength of an oven-dry specimen is determined to the  nearest 0.
 be significant once the rock has been comminuted to the size used in concrete. In  essence, the crushing strength test measures the quality of the parent rock rather than the  quality of the aggregate as used in concrete. For this reason, in 1975, the test was deleted from  BS 812, and tests on prepared specimens are nowadays less used than tests
 on bulk aggregate,  but are nevertheless useful when dealing with a potential new source of crushed aggregate.   Sometimes, the strength of a wet as well as of a dry specimen is determined. The ratio of wet  to dry strengths measures the softening effect, and when this is high, poor durability of the rock  may be suspected.   A test on the crushi
ng properties of bulk aggregate is the so-called crushing value  test of BS 812: Part 3: 1975. There is no explicit relation between this crushing value  and the compressive strength, but the results of the two tests are in agreement (Fig. 3.5).  The crushing value is a useful guide when dealing with aggregates of unknown  performance, particularl
y when lower strength may be suspected, as for instance with  limestone and some granites and basalts.    Fig. 3.5. Relation between the compressive strength of the parent rock and the  crushing value of aggregate obtained from the same rock     The material to be tested should pass a 14.0 mm (1/2 in.) test sieve and be retained on a  10.0 mm (3/8
ssing this sieve to the total weight of the sample is called the aggregate crushing value.   In the United States, where large quantities of artificial lightweight aggregates are used,  attempts were made to develop a strength test for these aggregates, rather similar to the  crushing value test described above, but no test has been standardized. 
	The crushing value test is rather insensitive to the variation in strength of  weaker aggregates, i.e. those with a crushing value of over 25 to 30. This is so  because, having been crushed before the full load of 400 kN (40 tons) has been applied,  these weaker materials become compacted so that the amount of crushing during later  stages of th
e	ten per cent fines value test  is more sensitive and gives a truer picture of differences between more or less weak samples.  For this reason, the test is of use in assessing light weight aggregates but there is no simple  relation between the test result and the upper limit of strength of concrete made with  the given aggregate.      Other Mech
rect correlation  between the crushing value and the performance of aggregate in concrete or the strength  of the concrete is not possible.   One advantage of the impact test is that it can be performed in the field with some  modifications, such as the measurement of quantities by volume rather than by weight,  but the test may not be adequate fo
r	compliance purposes.   In addition to strength and toughness, hardness or resistance to wear is an important  property of concrete used in roads and in floor surfaces subjected to heavy traffic.  Several tests are available, and it is possible to cause wear by abrasion, i.e. by rubbing  of a foreign material against the stone under test, or by a
ttrition of stone particles against  one another.   In the abrasion (Derry) test, a cylindrical specimen, similar to those used in the  crushing strength test, is subjected to wear by quartz (Leighton Buzzard) sand pressed  against the cylinder by a rotating metal disc. The abrasion value is expressed as 20  minus one-third of the loss of weight o
it is quite frequently used in other countries, too, because its results show  good correlation not only with the actual wear of aggregate when used in concrete but  also with the compressive and flexural strengths of  concrete made with the  given aggregate. In this test, aggregate of specified grading is placed in a  cylindrical drum, mounted ho
o. 200 sieve) is produced. No standard  apparatus is available but some development has been made by Meininger.   Table 3.7 gives average values of crushing strength, aggregate crushing value,  abrasion, impact, and attrition for the different rock groups of BS 812 : Part 2 : 1975.  It should be noted that the values for hornfels and schists are b
: Average Test Values for British Rocks of Different Groups     As far as the crushing strength is concerned, basalt is extremely variable, fresh  basalts with little olivine reaching some 400 MPa (60 000 psi), while decomposed basalt  at the other end of the scale may have a strength of no more than 100 MPa (15 000 psi).  Limestone and porphyry s
how much less variation in strength, and in Britain porphyry has  a good general performance - rather better than that of granites, which tend to be  variable.   An indication of the accuracy of the results of the different tests is given by Table 3.8,  listing the number of samples to be tested in order to ensure a 0.9 probability  that the mean 
ctions give an indication of the quality of the aggregate, it is not  possible to predict from the properties of aggregate the potential strength development  of concrete made with the given aggregate, and indeed it is not yet possible to  translate physical properties of aggregate into its concrete-making properties.    Table 3.8: Reproducibility
absorbent.   Although there is no clear-cut relation between the strength of concrete and the water  absorption of aggregate used, the pores at the surface of the particle affect the bond  between the aggregate and the cement paste, and may thus exert some influence on the  strength of concrete.   Normally, it is assumed that at the time of settin
rete. The organic matter present may not be  harmful to concrete or the colour may be due to some iron-bearing minerals. For this  reason, further tests are necessary: concrete cubes are made using the suspected  aggregates and their strength is compared with concrete of the same mix proportion but  made with aggregates of known quality.   In earl
eroxide.   It is interesting to note that in some cases the effects of organic impurities may be only  temporary. In one investigation concrete made with a sand containing organic matter had  a 24-hour strength equal to 53 per cent of the strength of similar concrete made with clean sand.  At 3 days this ratio rose to 82 per cent, then to 92 per c
ent at 7 days, and at 28 days equal  strengths were recorded.    Clay and Other Fine Material    Clay may be present in aggregate in the form of surface coatings which interfere with the  bond between aggregate and the cement paste. Since good bond is essential to ensure a  satisfactory strength and durability of concrete the problem of clay coati
ed.  Sea-dredged coarse aggregate may have a large shell content.  This has no  adverse effect on strength but workability in concrete made with aggregate having a large  shell content is slightly reduced. The shell content of particles larger than 5 mm can be  determined by hand picking, using the method of an amendment of the British Standard  B
density are regarded as unsound and so are soft inclusions  such as clay lumps, wood, and coal, as they lead to pitting and scaling.  If present in large  quantity (over 2 to 5 per cent of the weight of the aggregate) these particles may adversely  affect the strength of concrete and should certainly not be permitted in concrete which is  exposed 
weight of the aggregate have no  adverse effect on the strength of concrete.   The presence of coal and other materials of low density can be determined by flotation in a  liquid of suitable specific gravity, as, for instance, by the method of ASTM Standard C  123 - 69 (reapproved 1975). If the danger of pitting and scaling is not thought importan
t,  and strength of concrete is the main consideration, a trial mix should be made.   Mica should be avoided because in the presence of active chemical agents produced  during the hydration of cement, alteration of mica to other forms may result. Also, free  mica in fine aggregate, even in quantities of a few per cent of the weight of the aggregat
e  affects adversely the water requirement and hence the strength of concrete. It appears that  mica in the form of muscovite is much more harmful than biotite. These facts should be  borne in mind when materials such as china clay sand are considered for use in concrete.   Gypsum and other sulphates must not be present; their existence in many Mi
, such as mine tailings, can contain harmful  substances. For instance, small quantities of lead soluble in limewater (e.g. 0.1 per cent of  PbO by weight of aggregate) greatly delay the set and reduce the early strength of  concrete; the long-term strength is unaffected.      Soundness of Aggregate    This is the name given to the ability of aggr
able. A related problem is that of  combining fine and coarse aggregates so as to produce a desired grading. What, then, are  the properties of a "good" grading curve?   Since the strength of fully compacted concrete with a given water / cement ratio is  independent of the grading of the aggregate, grading is, in the first instance, of importance 
 only in so far as it affects workability. As, however, the development of strength  corresponding to a given water / cement ratio requires full compaction, and this can be  achieved only with a sufficiently workable mix, it is necessary to produce a mix that can be  compacted to a maximum density with a reasonable amount of work.   It should be s
x is undesirable. It has also been assumed  that the greater the amount of solid particles that can be packed into a given volume of  concrete the higher its strength. This maximum density theory has led to the advocacy of  grading curves parabolic in shape, or in part parabolic and then straight (when plotted to a  natural scale), as shown in Fig
.    Fig. 3.13. Fuller's grading curve     Let us now consider the surface area of the aggregate particles. The water / cement ratio of  the mix is generally fixed from strength considerations. At the same time, the amount of  cement paste has to be sufficient to cover the surface of all the particles so that the lower  the surface area of the agg
n be used under given  circumstances and the problem of influence of the  maximum size on strength in general  are discussed on Page 195.   It can be seen that, having chosen the maximum size of aggregate and its grading, we can  express the total surface area of the particles using the specific surface as a parameter,  and it is the total surface
fic surface of the aggregate the water requirement and the compressive strength of  the concrete are the same for very wide limits of aggregate grading. This applies both to  continuously and gap-graded aggregate, and in fact three of the four gradings listed in  Table 3.21, reproduced from Davey's paper, are of the gap type.    Table 3.21: Proper
ties of Concretes Made with Aggregates of the Same Specific Surface     An increase in the specific surface of the aggregate for a constant water / cement ratio has  been found to lead to a lower strength of concrete, as shown for instance in Table 3.22,  reproducing Newman and Teychenne's results. The reasons for this are not quite clear,  but it
 is possible that a reduction in density of the concrete consequent upon an increase in  fineness of the aggregate is instrumental in lowering the strength.    Table 3.22: Specific Surface of Aggregate and Strength of Concrete for a 1:6 Mix with a  Water / Cement Ratio of 0.60     It seems then that the surface area of the aggregate is an importan
 segregation, has some effect on bleeding, and influences the placing and finishing of the  concrete. These factors represent the important characteristics of fresh concrete and affect  also its properties in the hardened state: strength, shrinkage, and durability.   Grading is thus of vital importance in the proportioning of concrete mixes, but i
s	 this is usually corrected at the mixer by a variation in the water content, concrete of  variable strength is obtained.      Practical Gradings    From the brief review in the previous section it can be seen how important it is to use  aggregate with a grading such that a reasonable workability and a minimum segregation  are obtained. The impor
 same workability is to be obtained using aggregates with grading curves Nos. 1 and 4,  the latter would require a considerably higher water content: this would mean a lower strength  if both concretes are to have the same aggregate / cement ratio or, if the same strength is  required, the concrete made with the fine aggregate would have to be con
 aggregate from  10 to 25 per cent results in only a small decrease in the compressive strength of concrete,  typically by 10 per cent.   At the other extreme, the coarse sand of zone 1 produces a harsh mix, and a high sand  content may be necessary for higher workability. This sand is more suitable for rich mixes  or for use in concrete of low wo
ws test results for the case when the same materials  were used but the coarse / fine aggregate ratio was kept constant. The use of finer sand led  to a higher water requirement and consequently a lower strength of concrete.    Table 3.26: Properties of Concretes Made with Aggregates of Fixed Proportions     The requirements of BS 882 : 1973 for t
o be affected. Likewise, Fig.  3.24, showing McIntosh's results, confirms that, using given materials with a fixed  aggregate / cement ratio (but adjusting the sand content), approximately the same  workability and strength are obtained with gap and continuous gradings; Brodda and  Weber reported a slight negative influence of gap grading on stren
gth.    Fig. 3.24. Workability and strength of 1:6 concretes made with gap- and continuously  graded aggregates     Similarly, there is no difference in shrinkage of the concretes made with aggregate of  either type of grading, although it might be expected that a framework of coarse  particles almost touching one another would result in a lower t
nt ratio can be lowered with a consequent increase in  strength.   This behaviour has been verified by tests with aggregates up to 38.1 mm (1 1/2 in.)  maximum size, and is usually assumed to extend to larger sizes as well. Experimental  results show, however, that above the 38.l mm (1 1/2 in.) maximum size the gain in  strength due to the reduced
 water requirement is offset by the detrimental effects of lower  bond area (so that volume changes in the paste cause larger stresses at interface) and of  discontinuities introduced by the very large particles, particularly in rich mixes. Concrete  becomes grossly heterogeneous and the resultant lowering of strength may possibly be  similar to t
nce of the  two  effects depends on richness of the mix as shown in Fig. 3.25. Thus the best maximum size  of aggregate from the standpoint of strength is a function of the richness of the mix.  Specifically, in lean concrete (165 kg of cement per cubic metre (280 lb / yd3)), the use of  150 mm (or 6 in.) aggregate is advantageous. However, in str
uctural concrete of usual  proportions, from the point of view of strength there is no advantage in using aggregate  with a maximum size greater than about 25 or 40 mm (1 or 1 1/2 in.). Moreover, the use of larger  aggregate would require the handling of a separate stockpile and might increase the risk of  segregation. However, a practical decisio
 codes of practice.    Fig. 3.25. Influence of maximum size of aggregate on the 28-day compressive strength of  concretes of different richness      Use of "Plums"    The original idea of the use of aggregate as an inert filler can be extended to the inclusion  of large stones in a normal concrete: thus the apparent yield for a given amount of cem
f the concrete.  14    Properties of Hardened Concrete    The properties of fresh concrete are important only in the first few hours of its history  whereas the properties of hardened concrete assume an importance which is retained for  the remainder of the life of the concrete. The important properties of hardened concrete  are strength, deformat
ion under load, durability, permeability and shrinkage. In general,  strength is considered to be the most important property and the quality of concrete is  often judged by its strength. There are, however, many occasions when other properties  are more important, for example, low permeability and low shrinkage are required for  water-retaining s
tructures. Although in most cases an improvement in strength  results in an improvement of the other properties of concrete there are exceptions. For  example, increasing the cement content of a mix improves strength but results in higher  shrinkage which in extreme cases can adversely affect durability and permeability. One of  the primary object
ts have therefore been developed for the assessment of  concrete quality. The limitations and applications of nondestructive testing together with a  brief description of the techniques is given at the end of this chapter.      14.1 Strength    The strength of concrete is defined as the maximum load (stress) it can carry. As the  strength of concr
ete increases its other properties usually improve and since the tests for  strength, particularly in compression, are relatively simple to perform concrete  compressive strength is commonly used in the construction industry for the purpose of  specification and quality control. Concrete is a comparatively brittle material which is  relatively wea
k in tension.    Compressive Strength    The compressive strength of concrete is taken as the maximum compressive load it can  carry per unit area. Concrete strengths of up to 80 N mm-2 can be achieved by selective  use of the type of cement, mix proportions, method of compaction and curing conditions.  Concrete structures, except for road pavemen
ts, are normally designed on the  basis that concrete is capable of resisting only compression, the tension being carried by  steel reinforcement. In the United Kingdom a 150 mm cube is commonly used for  determining the compressive strength. The standard method described in BS 1881: Part  116 requires that the test specimen should be cured in wat
er at 20 +/- 2 C and crushed, by  loading it at a constant rate of stress increase of between 12 and 24 N mm-2 min-1,  immediately after it has been removed from the curing tank.    Tensile Strength    The tensile strength of concrete is of importance in the design of concrete roads and  runways. For example, its flexural strength or modulus of ru
pture (tensile strength in  bending) is utilised for distributing the concentrated loads over a wider area of road  pavement. Concrete members are also required to withstand tensile stresses resulting from  any restraint to contraction due to drying or temperature variation.   Unlike metals, it is difficult to measure concrete strength in direct t
diametrical plane and the  cylinder splits along the loaded diameter. The magnitude of the induced tensile stress fct at  failure is given by    (formula)    where F is the maximum applied load and l and d are the cylinder length and diameter  respectively.   The flexural strength of concrete is another indirect tensile value which is also  common
ly used (BS 1881: Part 118). In this test a simply supported plain concrete beam is  loaded at its third points, the resulting bending moments inducing compressive and tensile  stresses in the top and bottom of the beam respectively. The beam fails in tension and the  flexural strength (modulus of rupture) fcr is defined by    (formula)    where F
 is the maximum applied load, L the distance between the supports, and b and d  are the beam breadth and depth respectively at the section at which failure occurs.   The tensile strength of concrete is usually taken to be about one-tenth of its  compressive strength. This may vary, however, depending on the method used for  measuring tensile stren
gth and the type of concrete. In general the direct tensile strength  and the split cylinder tensile strength vary from 5 to 13 per cent and the flexural strength  from 11 to 23 per cent of the concrete cube compressive strength. In each case, as the  strength increases the percentage decreases. As a guide, the modulus of rupture may be  taken as 
0.7 / (cube strength) N mm-2 and the direct tensile strength as 0.45 / (cube  strength) N mm- 2 although, where possible, values based on tests using the  actual concrete in question should be obtained.      14.2 Factors influencing Strength    Several factors which affect the strength of concrete are shown in figure 14.1. In this section their  i
nfluence is discussed with particular reference to compressive strength. In general, tensile strength is  affected in a similar manner.    Figure 14.1 Factors affecting strength of concrete    Influence of the Constituent Materials    Cement    The influence of cement on concrete strength, for given mix proportions, is determined by its fineness  
and chemical composition through the processes of hydration (see chapter 12). The gain in concrete  strength as the fineness of its cement particles increases is shown in figure 14.2. The gain in strength  is most marked at early ages and after 28 days the relative gain in strength is much reduced. At some  later age the strength of concrete made 
with fine cements may not be very different from that made  with normal cement (300 m2 kg- 1).   The role of the chemical composition of cement in the development of concrete  strength can best be appreciated by studying table 14.1 and figures 14.3 and 14.4. It is apparent that  cements containing a relatively high percentage of tricalcium silicat
e	(C3S) gain strength much more  rapidly than those rich in dicalcium silicate (C2S), as shown in figure 14.3; however, at later ages the  difference in the corresponding strength values is small. In fact there is a tendency for concretes made  with low-heat cements eventually to develop slightly higher strengths (figure 14.4). This is possibly  d
ue to the formation of a better quality gel structure in the course of hydration.    Figure	14.2 Effect of cement fineness on the development of concrete strength, based on  Bennett and Collings (1969)    TABLE 14.1 Chemical composition of various Portland cements with similar fineness    Figure 14.3 Development of strength of typical concrete mad
e	with different Portland cements  (see table 14.1)    Figure 14.4 Development of strength of typical concrete made with different Portland cements  (see table 14.1)    Water    A concrete mix containing the minimum amount of water required for complete hydration of its  cement, if it could be fully compacted, would develop the maximum attainable 
strength at any given  age. A water - cement ratio of approximately 0.25 (by weight) is required for full hydration of the  cement but with this water content a normal concrete mix would be extremely dry and virtually  impossible to compact.   A partially compacted mix will contain a large percentage of voids and the  concrete strength will drop. 
concrete, as the ratio of  water to cement increases the strength decreases in a manner similar to that illustrated in figure 15.3  (see chapter 15).    Aggregate    When concrete is stressed, failure may originate within the aggregate, the matrix or at the  aggregate - matrix interface; or any combination of these may occur. In general the aggreg
ates are  stronger than the concrete itself and in such cases the aggregate strength has little effect on the strength  of concrete.   The bond (aggregate - matrix interface) is an important factor determining  concrete strength. Bond strength is influenced by the shape of the aggregate, its surface texture and  cleanliness. A smooth rounded aggre
gate will result in a weaker bond between the aggregate and  matrix than an angular or irregular aggregate or an aggregate with a rough surface texture. The  associated loss in strength however may be offset by the smaller water - cement ratio required for the  same workability. Aggregate shape and surface texture affect the tensile strength more 
than the  compressive strength. A fine coating of impurities, such as silt and clay, on the aggregate surface  hinders the development of a good bond. A weathered and decomposed layer on the aggregate can  also result in a poor bond as this layer can readily become detached from the sound aggregate beneath.   The aggregate size also affects the st
rength. For given mix proportions, the  concrete strength decreases as the maximum size of aggregate is increased. On the other hand, for a  given cement content and workability this effect is opposed by a reduction in the water requirement  for the larger aggregate. However, it is probable that beyond a certain size of aggregate there is no  obvi
ous advantage in further increasing the aggregate size except perhaps in some instances when  larger aggregate may be more readily available. The optimum maximum aggregate size varies with  the richness of the mix, being smaller for the richer mixes, and generally lies between 10 and 50 mm.  Concrete of a given strength can be produced with aggreg
ates having a variety of different gradings  provided due care is exercised to ensure that segregation does not occur. The suitability of a grading  to some extent depends on the shape and texture of the aggregate. Aggregates which react with the  alkali content of a cement adversely affect concrete strength. This is rarely a problem in the United
	Kingdom, even though some cases have been reported since 1976 (BRE Digest 258).    Admixtures    As a general rule, admixtures can only affect concrete strength by changing the hydration processes  and the air content of the mix and / or by enabling changes to be made to the mix proportions, most  importantly to the water - cement ratio. Acceler
ating admixtures increase the rate of hydration thereby  providing an increased early strength with little or no change at later ages unless the increased rate of  heat evolution causes internal cracking in which case a lower strength will result. In contrast, with  retarding admixtures the early strength of concrete is reduced owing to the delay 
in setting time.  Provided no air is entrained, the concrete strength will be approximately the same as that of the  control mix within a few days. Air entrainment in concrete will cause a reduction in strength at all  ages and to achieve a required strength the mix cement content has to be increased. However, in  practice the increased yield and 
improvement in workability is also taken into account in the mix  design when using air-entraining admixtures and there is generally no significant change in concrete  strength for the usual range of air content. Most water-reducing admixtures, including  superplasticizers, do not have any significant effect on the hydration of the cement. Thus, w
hen these  admixtures are used to improve workability no significant change in strength should be expected,  providing of course that the air content remains unchanged. On the other hand if the water content of  the mix is reduced while maintaining workability an increase in strength corresponding to the new  water - cement ratio will result.   It
estimate of their effect on concrete strength  in individual circumstances it is necessary to carry out trial mixes.    Influence of the Methods of Preparation    When concrete materials are not adequately mixed into a consistent and homogeneous mass, some  poor quality concrete is inevitably the result. Even when a concrete is adequately mixed ca
on. If full compaction is not achieved the resulting voids produce a  marked reduction in concrete strength.    Influence of Curing    Curing of concrete is a prerequisite for the hydration of the cement content. For a given concrete, the  amount and rate of hydration and furthermore the physical make-up of the hydration products are  dependent on
 the time - moisture - temperature history.   Generally speaking, the longer the period during which concrete is kept in  water, the greater its final strength. It is normally accepted that a concrete made with ordinary  Portland cement and kept in normal curing conditions will develop about 75 per cent of its final  strength in the first 28 days.
 This value varies with the nominal strength of concrete however,  increasing as concrete strength increases. The gain in strength with age up to the time of loading can  now be used only for estimating the static modulus of elasticity of concrete in structures (BS 8110:  Part 2). The age factors for strength prescribed for this purpose are given 
in table 14.2.    TABLE 14.2 Age factors for strength of concrete     The development of concrete strength under various curing conditions is  shown in figure 14.5. It is apparent that concrete left in air achieves the lowest strength values at all  ages owing to the evaporation of the free mixing water from the concrete. The gain in strength  dep
 final apparent strength of concrete. It should also be  noted that moist (or water) curing after an initial period in air results in a resumption of the hydration  process and that concrete strength is further improved with time, although the optimum strength may  not be realised.    Figure 14.5 Effect of curing and condition of concrete when tes
ted on concrete strength,  based on Gilkey (1937)     The temperature at which concrete is cured has an important bearing on the  development of its strength with time. The rate of gain in strength of concrete made with ordinary  Portland cement increases with increase in concrete temperature at early ages (figure 14.6), although  at later ages th
e concrete made and cured at lower temperatures shows a somewhat higher strength.  Figure 14.7 shows how a high temperature during the placing and setting of concrete can adversely  affect the development of its strength from early ages. On the other hand, when the initial temperature  is lower than the subsequent curing temperature, then higher t
emperatures during final curing result  in significantly higher strengths (figure 14.8). A possible explanation for this behaviour is that a rapid  initial hydration appears to form a gel structure (hydration product) of an inferior quality and this  adversely affects concrete strength at later ages. Concretes made with other Portland cements woul
d  respond to temperature in a somewhat similar manner.    Figure 14.6 Comparative compressive strength of concrete cast, sealed and maintained at different  temperatures, based on Price (1951)    Figure 14.7 Comparative compressive strength of sealed concrete specimens maintained at  different temperatures for 2 hours after casting and subsequent
ly cured at 21 C, based on Price (1951)    Figure 14.8 Comparative compressive strength of concrete cast, sealed and maintained at 10 C for  the first 24 hours and subsequently cured at different temperatures, based on Price (1951)     It has been suggested that the strength of concrete can be related to the product  of age and curing temperature,
 commonly known as maturity. However, such relationships are  dependent on a number of factors such as curing temperature history, particularly the temperature at  early ages (see figures 14.7 and 14.8), and are therefore limited in their general applicability for  predicting concrete strength.    Influence of Test Conditions    The conditions und
er which tests to determine concrete strength are carried out can have a  considerable influence on the strength obtained and it is important that these effects are understood if  test results are to be correctly interpreted.    Specimen shape and size    Three basic shapes used for the determination of compressive strength are the cube, cylinder 
and  square prism. Each shape gives different strength results and furthermore for a given shape the  strength also varies with size. Figure 14.9 shows the influence of specimen diameter and it can be  seen that as the size decreases, the apparent strength increases. The measured strength of concrete is  also affected by the height - diameter rati
o (figure 14.10). For height - diameter ratios less than 2  strength begins to increase rapidly owing to the restraint provided by the machine platens. Strength  remains sensibly constant for height - diameter ratios between 2 and 3 and thereafter shows a slight  reduction. The relative influence of slenderness may be modified by several inherent 
characteristics of  concrete such as strength, air-entrainment, strength of aggregate, degree of moist curing and moisture  content at the time of testing.    Figure 14.9 Effect of specimen size on the apparent 28-day concrete compressive strength for  specimens with a height - diameter ratio of 2 and aggregate whose maximum diameter is  one-quart
er of the diameter of the specimen, based on Price (1951)    Figure 14.10 Effect on height - diameter ratio on concrete compressive strength for specimens  moist-cured at room temperature and test wet    BS 1881: Part 116 specifies the use of concrete cubes for determining compressive strength; 150 mm  cubes are widely used in the United Kingdom f
or quality-control purposes. However, cored cylindrical  specimens are used for measuring the compressive strength of concrete in situ and precast members  (BS 1881: Part 120). This standard gives a set of equations for converting the measured strength of a  core into an equivalent in situ cube strength. It should be noted that the estimation of e
quivalent  standard cube strengths from core strengths is no longer considered to be valid, because of large  variations in the relationship between the two strengths arising from variations in site conditions  including percentage of reinforcement, dimensions of members and the methods of compaction and  curing. The effect on the measured strengt
h of the variations in the height - diameter ratio of drilled  cores is taken into account by using the correction factors given in table 14.3. The effect of the shape  of the test specimen is taken into account by multiplying the cylinder (core) strength, for a cylinder  (core) having a height - diameter ratio of 2, by 1.25 to obtain the equivale
nt cube strength. Since the  relationship between height - diameter ratio and strength depends on the type of concrete, the use of  one set of correction factors of the type given in table 14.3 can only be suitable for a limited range of  concrete materials. It should be noted that the correction factors given in table 14.3 are most likely to  fav
our low-strength concretes. The estimation of the equivalent in situ cube strength from  compression tests performed on cored cylindrical specimen can be further influenced by a multitude of  other factors and the reader is advised to consult the Concrete Society Technical Report on concrete  core testing for strength (1976), the British Standard 
Guide on Assessment of Concrete Strength in  Existing Structures (BS 6089), Murphy (1984) and Munday and Dhir (1984).    TABLE 14.3 Correction factor for compression tests on cylinders (cores)    Specimen moisture content and temperature    This should not be confused with the effect of moisture and temperature during curing. The strength  of conc
rete can be influenced by the absence or presence of moisture and by temperature only when  these conditions generate internal stresses which change the magnitude of the external load required  to bring about failure. Since the mode of failure in different strength tests is different, it follows that  the influence of moisture content and temperat
ure on the apparent strength varies.   In the case of compression tests, air-dry concrete has a significantly higher  strength than concrete tested in a saturated condition (figure 14.5). The lower strengths of wet  concrete can be attributed mainly to the development of internal pore pressure as the external load is  applied.   The flexural stren
gth of saturated concrete is greater however than that of  concrete which is only partially dry owing to tensile stresses developed near the surface of the dry  concrete by differential shrinkage. The initial drying is therefore critical with the apparent strength  reaching a minimum value within the first few days; thereafter it begins to increas
e gradually as the  concrete approaches a completely dry state. A thoroughly dry concrete specimen in which tensile  stresses are not being induced has a higher apparent flexural strength than saturated concrete. The  indirect tensile or split cylinder strength is lower for saturated concrete than for thoroughly dried  concrete.   Since the influe
nce of the moisture content on concrete strength varies with  the type of test, standard strength tests (BS 1881: Part 116) should be performed on specimens in a  saturated condition.    Method of loading    The compressive strength of concrete increases as the lateral confining pressure increases (figure  14.11).    Figure 14.11 Effect of lateral
 compression on concrete compressive strength     The apparent strength of concrete is affected by the rate at which it is loaded.  In general, for static loading, the faster the loading rate the higher the indicated strength. However,  the relative effects of the rate of loading vary with the nominal strength, age and extent of moist  curing. Hig
h-strength mature concretes cured in water are most sensitive to loading rate and  particularly so for loading rates greater than 600 N mm-2 min-1. BS 1881: Part 116 requires concrete  in compression tests to be loaded at 12 to 24 N mm-2 min-1 and within this small range of loading  rates variations in the measured strength of concrete will be ins
ignificant. The standard rates of  loading for flexural and split cylinder tests correspond to rates of increase of tensile stress of 1.2 to 6.0  N mm-2 min-1 and 1.2 to 2.4 N mm-2 min-1 respectively (BS 1881: Parts 117 and 118). For the  flexural test the standard requires the use of the lower loading rates for low strength concretes and the  hig
her loading rates for high strength concretes.   When loads on a structure are predominantly cyclic (repeated loading and  unloading) in character, the effects of fatigue should also be considered. This kind of loading produces  a reduction in strength. A reduction in strength of as much as 30 per cent of the normal static strength  value can take
 place, although this depends on the stress - strength ratio, the frequency of loading and  the type of concrete.   Structural concrete is commonly subjected to sustained loads. It is probable that  concrete can withstand higher loads if a constant load is maintained before  loading to failure. Improvement in compressive strength can occur for sus
tained loads up to 85 per  cent of the normal static strength although the actual gain in strength depends on the duration and  magnitude of load, type of concrete and age. The increase in strength is probably due to consolidation  of the concrete under sustained load and the redistribution of stresses within the concrete.      14.3 Deformation   
. The method prescribed in  BS 1881: Part 121 requires repeated loading and unloading before the specimen is loaded for  determination of its secant modulus of elasticity from a stress - strain curve, which then approaches a  straight line for stresses up to one-third of the concrete cylinder strength. The modulus of elasticity for  most concretes
, at 28 days, ranges from 15 to 40 kN mm-2. As a guide, the modulus may be assumed  to be 3.8 / (concrete strength, N mm-2) kN mm-2 for normal weight concrete. For structural design  purposes, the short term elastic modulus values for normal weight concrete given in BS 8110: Part 2  may be used.    Poisson's ratio    When concrete is subjected to 
terial and its resistance to deformation under load is dependent on the  stiffness of its various phases such as aggregate, cement paste and voids, and the interaction between  individual phases. In general the factors which influence the strength of concrete also affect  deformation although the extent of their influence may well vary. The modulu
s of elasticity increases  with strength, although the two properties are not directly related because different factors exert varying  degrees of influence on strength and modulus of elasticity. Although relationships between the two  properties may be derived they are only applicable within the range of variables considered. In short  there is n
o unique relationship between strength and modulus of elasticity.   The influence of stress level and rate of loading on both the axial and lateral  strains is shown in figure 14.14 for a concrete which has been cured in water (20 C +/- 1 C) and  air-dried before loading in uniaxial compression. It should be noted that the slower the rate of loadi
involved. the  provision of members with inadequate strength. In prestressed concrete, allowance must be made for  the loss of tension in the prestressing tendons resulting from the shortening of a member under the  action of creep. Creep strain can also be beneficial in that it can relieve local stress concentrations  which might otherwise lead t
o structural damage. A classic example of this is the reduction of  shrinkage stresses in restrained members.    Factors influencing creep    Both the type of concrete (as described by its ingredients, curing history, strength and age) and the  relative magnitude of the applied stress with respect to concrete strength (stress - strength ratio) aff
ect  the creep strain. For a given concrete the creep strain is almost directly proportional to the  stress - strength ratio for ratios up to about one-third. For the same stress - strength ratio, the creep  strain increases as both cement content and water - cement ratio increase (figure 14.16) and decreases as  the relative humidity and age at l
ituent materials of concrete on creep is somewhat  complex. The different types of cement influence creep because of the associated different rates of  gain in concrete strength. For example, concrete made with rapid-hardening Portland cement shows  less creep than concrete made with ordinary Portland cement and loaded at the same age. Normal  roc
n purposes. It has been noted that,  for a given concrete, creep strain depends on the stress - strength ratio. For practical purposes,  concrete creep strain may be assumed to be directly proportional to the elastic deformation up to a  stress - strength ratio of about two-thirds. On this assumption the ultimate creep strain may be  estimated for
air-entrained concrete  is particularly useful for roads where salts are used for de-icing. BS 8110: Part 1 recommends the use  of entrained air for a concrete of characteristic strength below 50 N mm-2 where it is likely to be  exposed to freezing and thawing actions while wet and its surfaces are subject to the effects of  de-icing salts. The le
 In general, the resistance of concrete to erosion and abrasion increases with  increase in strength. The use of a hard and tough aggregate tends to improve concrete resistance to  wear.    Alkali - Aggregate Reactions    Certain natural aggregates react chemically with the alkalis present in Portland cement. When this  happens these aggregates ex
f concrete and the two are not necessarily related. Absorption  may be defined as the ability of concrete to draw water into its voids. Low permeability is an  important requirement for hydraulic structures and in some cases watertightness of concrete may be  considered to be more significant than strength although, other conditions being equal, c
 the Quality of Concrete From Nondestructive Testing    The quality of concrete is usually taken to mean its strength and durability although other properties  such as resistance to deformation and shrinkage can be significant in determining structural  behaviour. In general most of the properties of concrete improve with increasing strength and f
or this  reason the quality of concrete is often judged by its strength. Nondestructive testing, as the name  implies, requires that the material under test is not damaged during testing.   Direct measurement of the strength of concrete involves destructive stresses and thus cannot  be used for determining the quality of concrete in structures. Fu
rthermore, the compressive strength  test, as described in BS 1881: Part 116 can only indicate the potential strength of the mix. The actual  concrete strength within a concrete unit or structure depends on the conditions of placing, compaction  and curing. Although test samples can be cored from the structure and tested for evaluating the  qualit
	developed (BS 1881: Part 201) but those which have been most widely accepted include vibrational  methods for estimating strength, durability and uniformity and for detecting flaws, and hardness  methods for estimating strength. Although the tests are simple to perform they have certain limitations.  Nevertheless, when applied rationally the tec
e, mix proportions and curing  conditions. However, for a given concrete the variation in dynamic modulus of elasticity can give a  good indication of variations in strength and static modulus of elasticity, and this test is particularly  useful for assessing the progressive change in strength and durability as affected by various factors  such as
 sides is  desirable. As for the dynamic modulus of elasticity, there is no unique relationship between  the pulse velocity and strength as it is influenced by the concrete constituents and curing  conditions. The effect of the coarse aggregate is of particular  significance since its influence on the pulse velocity is more marked than its influen
ce on  strength. Thus the evaluation of the quality of concrete in structures is usually made on a  comparative basis and the technique is frequently employed to detect inferior parts within a  structure (Tomsett, 1980). However, when the mix proportions remain constant and only  one type of coarse aggregate is used then it is possible to determin
e a specific relationship  between strength and pulse velocity for in situ concrete. Since the pulse velocity is affected  by moisture it is important to have moisture conditions in  the test specimens similar to those in the in situ concrete when establishing  strength - pulse-velocity relationships. The pulse is transmitted most effectively by s
surface the rebound reading is  corrected because of the change in the impact energy.    Figure 14.23 Schmidt hammer     Hardness (rebound number) is a relative property and there can be no physical  relationship between it and the other properties of concrete. Empirical relationships  between rebound number and strength have been established and 
in general the higher the  rebound number, the greater the strength. As for the dynamic modulus of elasticity and  ultrasonic pulse velocity, there is no unique relationship between rebound number and  strength. For this reason it is advisable to determine the  strength - hardness relationship for each concrete instead of relying on secondhand val
ues.  When the hammer is used for assessing the strength of in situ concrete the test procedure  and environmental conditions should be similar to those employed during calibration. Since  for a given concrete the rebound number can vary because of differences in hardness  between aggregate and matrix and the possible variation in aggregate minera
logy, it is  necessary that several readings are taken and the average value used. Control over the  preparation of the test surface is important for proper use of the hammer. Provided the  limitations of the method are borne in mind and the hammer is used intelligently, it can be a  useful tool for assessing the strength of concrete in structures
sessment of concrete strength are pull-out, break-off and  penetration tests. The application of these tests, along with  ultrasonic pulse and hardness tests, is dealt with in the British Standard Guide to  Assessment of Concrete Strength in Existing Structures (BS 6089). The pull-out tests  essentially involve determining the force required to pu
ll out a steel rod fixed into the  surface of the concrete (Malhotra, 1975;  Chabowski and Brydon-Smith, 1980). The break-off test measures the flexural strength of  concrete (Johansen, 1979). The penetration test measures the resistance of concrete to  penetration of a probe fixed into the surface (ASTM C 803-75). Introduction    Concrete is a ma
ipes and lightweight  drainage channels. This section deals only with normal concretes in which cement and  water form the binding medium.   In its hardened state concrete is a rock-like material with a high compressive  strength. By virtue of the ease with which fresh concrete in its plastic state may be  moulded into virtually any shape it may b
e used to advantage architecturally or solely for  decorative purposes.  Special surface finishes, for example, exposed aggregate, can also be used to great effect.   Normal concrete has a comparatively low tensile strength and for structural  applications it is normal practice either to incorporate steel bars to resist any tensile forces  (reinfo
amp standards.   The impact strength, as well as the tensile strength, of normal concretes is low  and this can be improved by the introduction of randomly orientated fibres into the  concrete mix. Steel, polypropylene, asbestos and glass fibres have all been used with some  success in precast products, for example, pipes, building panels and pile
s. Steel fibres also  increase the flexural strength, or modulus of rupture, of concrete and this particular type of  fibre-reinforced concrete has been used in ground paving slabs for roads where flexural  and impact strength are both important. Fibre-reinforced  concretes are however essentially special-purpose concretes and for most purposes th
 not been possible, however, to find in the constitution of  cement a complete answer to the problem of durability of concrete: the principle  mechanical properties of hardened concrete, such as strength, shrinkage, permeability,  resistance to weathering, and creep, are affected also by factors other than cement  constitution, although this deter
mines to a large degree the rate of gain of strength. Figure  2.1 shows the rate of development of strength of concretes made with cements of  different types: while the rates vary considerably, there is little difference in the 90-day  strength of cements of all types. The general tendency is for the cements with a low rate of  hardening to have 
a slightly higher ultimate strength. For instance, Fig. 2.1 shows that  Type IV cement has the lowest strength at 28 days but develops the second highest  strength at the age of 5 years. A comparison of Fig. 2.1 and Fig. 2.2 illustrates the fact that  differences between cement types are not readily quantified.    Fig. 2.1. Strength development of
 concretes containing 335 kg of cement per cubic metre  (565 lb / yd3) and made with cements of different types    Fig. 2.2. Strength development of concretes with a water / cement ratio of 0.49 made with  cements of different types     Also, the retrogression of strength of the concrete made with Type II cement is not  characteristic of this type
 of cement. The pattern of low early and high late strength agrees  with the influence of the initial framework of hardened cement on the ultimate  development of strength: the more slowly the framework is established the denser the gel  and the higher the ultimate strength. Nevertheless, significant differences in the important  physical properti
nt in the tropics. The maximum gypsum content is also specified (see  p 18).   Over the years, there have been some changes in the characteristics of ordinary Portland  cement. In particular, modern cements have a higher C3S content and a greater fineness  than 40 years ago. As a consequence, cements have nowadays a 28-day strength perhaps  25 MPa
 higher than in 1925, but the gain in strength between 28 days and 10 years is  unaltered: approximately 20 MPa (3000 psi) for continuously water-cured concrete with a  water / cement ratio of about 0.53. (See Fig. 2.3.)    Fig. 2.3. The rate of gain of strength of cements between 1916 and 1970 measured on  standard concrete cylinders with a water
 / cement ratio of 0.53     The German classification of cements is on the basis of the 28-day strength of 1 : 3  mortars with a water / cement ratio of 0.5: 35, 45, and 55 MPa.      Rapid Hardening Portland Cement     This cement is very similar to ordinary Portland cement, and is also covered by BS 12:  1978. Rapid hardening Portland cement (Typ
e III), as its name implies, develops strength  more rapidly, and should therefore be correctly described as high early strength cement.  The rate of hardening must not be confused with the rate of setting: in fact, the two  cements have similar setting times.   The strength developed by the rapid hardening Portland cement at the age of three days
 is  of the same order as the 7-day strength of ordinary Portland cement with the same  water / cement ratio but the British Standards no longer specify 7-day strength. The  expected rate of hardening is reflected in the minimum strengths specified by BS 12: 1978,  listed in Table 1.10. The increased rate of gain of strength of the rapid hardening
 as for  ordinary Portland cement and need not therefore be repeated.   The use of rapid hardening cement is indicated where a rapid strength development is desired,  e.g. when formwork is to be removed early for re-use, or where sufficient strength for further  construction is wanted as quickly as practicable. Rapid hardening cement is only about
 $4 (f2) per  tonne dearer than ordinary cement, and it is not surprising that rapid hardening cement is used  extensively, accounting for about 10 per cent of all cement manufactured in the United Kingdom. Since,  however, the rapid gain of strength means a high rate of heat development, rapid hardening Portland  cement should not be used in mass
should generally be used within one month of despatch from the cement plant.   Extra rapid hardening cement is suitable for cold weather concreting, or when a very high  early strength is required, but structural use with reinforcement is not permitted by the British Code of  Practice CP 110: 1972 because of the risk of corrosion, and the cement i
s no longer manufactured in the  United Kingdom. The strength of extra rapid hardening cement is about 25 per cent higher than that of  rapid hardening cement at 1 or 2 days and 10 to 20 per cent higher at 7 days. The setting time of extra  rapid hardening cement is short: depending on temperature it can be 5 to 30 minutes so that early  placing i
ide can be classified as an  accelerator; its effects are discussed on p. 101.   Another type of cement with very rapid hardening properties is the so-called ultra high  early strength Portland cement, marketed in Great Britain. This cement contains no admixture and is  therefore suitable for use in reinforced and prestressed concrete; the rapid s
trength development is due  to the very high fineness of the cement: 700 to 900 m2 / kg. Because of this, the gypsum content has to  be higher (4 per cent expressed as SO3) than in cements complying with BS 12: 1978, but in all other  respects the ultra high early strength cement satisfies the requirement of that standard. We may note  that the hi
gh gypsum content has no adverse effect on long-term soundness as the gypsum is used up  in the early reactions of hydration.   The cement is manufactured by separating fines from rapid hardening Portland cement by  a cyclone air elutriator. Because of its high fineness, the ultra high early strength cement has a low bulk  density and deteriorates
 rapidly on exposure. High fineness leads to rapid hydration, and therefore to a  high rate of heat generation at early ages and to a rapid strength development; for instance, the 3-day  strength of rapid hardening Portland cement is reached at 16 hours, and the 7-day strength at 24  hours. There is, however, little gain in strength beyond 28 days
. Typical strengths of 1 : 3 concretes  made with the ultra high early strength cement are given in Table 2.3.    Table 2.3: Typical Values of Strength of a 1 : 3 Concrete made with Ultra High Early Strength  Portland Cement     The cement has been used successfully in a number of structures where early prestressing  or putting in service is of im
portance. Shrinkage and creep are not significantly different from those  obtained with other cements when the mix proportions are the same; in the case of creep, the  comparison has to be made on the basis of the same stress / strength ratio (see p. 401). We should note,  however, that for the same mix proportions, the use of ultra high early str
ength cement results in lower  workability.   The ultra high early strength Portland cement is marketed as Swiftcrete. A somewhat less  fine cement is Speed cement, developed in Belgium. It contains no accelerator and has a specific surface  of 450 to 500 m2 / kg. The standard vibrated mortar cube test gives strengths of about 28 MPa (4000  psi) a
t 1 day, 48 MPa (7000 psi) at 3 days, and 68 MPa (9800 psi) at 28 days. The Speed cement is  suitable for winter concreting or for urgent jobs such as road repair, well-sealing, etc.   In some countries, e.g. Italy and Sweden, extremely high early strength cement is  manufactured by double burning in the kiln.   One more cement of the very high ea
rly strength cement variety should be mentioned.  This is the so-called regulated-set cement, or jet cement, developed in the U.S. The cement consists  essentially of a mixture of Portland cement and calcium fluoro-aluminate (C11A7.CaF2) with an  appropriate retarder (usually citric acid). The setting time of the cement can vary between 1 and 30  
minutes (the strength development being slower the slower the setting) and  is controlled in the manufacture of the cement as the raw materials are interground and burnt  together. Grinding is difficult because of hardness differences.   The early strength development is controlled by the content of calcium fluoro-aluminate:  when this is 5 per ce
nt, about 6 MPa (900 psi) can be achieved at 1 hour; a 50 per cent mixture will  produce 20 MPa (3000 psi) at the same time. These values are based on a mix with a cement content of  330 kg / m3 (560 lb / yd3). The later strength development is similar to that of the parent Portland  cement but at room temperature there is virtually no gain in str
pidly hydrating compounds, C3S and C3A, results in a  slower development of strength of low heat cement as compared with ordinary Portland cement,  but the ultimate strength is unaffected. In any case, to ensure a sufficient rate of gain of strength the  specific surface of the cement must be not less than 320 m2 / kg.   A low heat Portland blast-
arly strength may be a disadvantage, and for this reason a  so-called modified (Type II) cement was developed in the United States. This modified cement  successfully combines a somewhat higher rate of heat development than that of low heat cement with a  rate of gain of strength similar to that of ordinary Portland cement. Modified cement is reco
ate-resisting cement mean that it has  a high silicate content and this gives the cement a high strength but, because C2S represents a high  proportion of the silicates, the early strength is low. The heat developed by sulphate-resisting cement is  not much higher than that of low heat cement. It could therefore be argued that sulphate-resisting  
und granulated slag of the same fineness as cement is added at the mixer as  a partial replacement of Portland cement; Portland blast-furnace cement concrete is thus  manufactured in situ. Like Portland blast-furnace cement concrete, Cemsave concrete has a lower early  strength than when Portland cement only is used, but at later ages at least equ
al strengths are reached  However, with Cemsave the workability is somewhat higher so that some reduction in water / (cement  plus slag) ratio is possible compared with the water / cement ratio of a Portland cement mix with the  same aggregate content. Concrete  with Cemsave exhibits a significantly  smaller temperature rise during hydration. Also
, setting  times and soundness are the same for both  cements. In actual fact, the fineness of Portland blast-furnace cement tends to be higher, but even so the  rate of hardening of Portland blast-furnace cement is somewhat slower during the first 28 days, and  adequate curing is therefore of importance; the strength requirements of BS 146 : 1973
 are therefore  lower than for ordinary Portland cement the requirement for a 28-day strength is 34 MPa  (4900 psi) for  mortar cubes or 22 MPa (3200 psi) for concrete cubes. However, at later ages  there is little difference between the strengths of  Portland blast-furnace and ordinary  Portland cements. Fig. 2.4 shows typical strength - time cur
vers low-heat Portland  blast-furnace cement; this allows a longer final setting time and a lower strength than BS 1370:1979.)  However, in cold weather the low heat of hydration of Portland blast-furnace cement, coupled with a  moderately low rate of strength development, can lead to frost damage.  Because of its fairly high sulphate resistance, 
concrete construction but care must be taken if used in cold weather as the rate  of strength development is considerably reduced at low temperatures. The rate of hardening of  supersulphated cement increases with temperature up to about 50 C (122 F), but at higher  temperatures anomalous behaviour has been encountered. For this reason, steam curi
	Portland cement, so that concrete with a water / cement ratio of less than	0.4 should not be made.  Mixes leaner than about 1 : 6 are not recommended. The decrease in strength with an increase in  the water / cement ratio has been reported to be smaller than in other cements but, since the early  strength development depends on the type of slag 
used in the manufacture of the cement, it is  advisable to determine the actual strength characteristics prior to use. Typical strengths  attainable are given in Table 2.4. It should be noted that for the concrete test BS 4248: 1974  prescribes a water / cement ratio of 0.55 instead of 0.60 used with other cements .      Portland-pozzolana Cements
he  measurement of a pozzolanic activity index. This is established by the determination of strength of  mixtures with a specified replacement of cement by pozzolana. There is also a pozzolanic activity  index with lime, which determines the total activity of a pozzolana. According to BS 4550: Part 2:1970,  pozzolanicity is assessed by comparing t
-lime ashes is sensitive to temperature: specifically, in mass concrete when  a rise in temperature occurs, the products of reaction may not be of high strength. However, the  development of strength is not simply related to temperature, being satisfactory in the  region of 120 to 150 C (250 to 300 F) but not at about 200 C (about 400 F) when the 
products of  reaction are substantially different.   The testing of this type of ash is still being developed but future use is possible.    Use of Pozzolanas    It is not possible to make a generalized statement on the Portland-pozzolana cements because the  rate of strength development depends on the activity of the pozzolanas and on the proport
ion of  Portland cement in the mixture. As a rule, however, Portland-pozzolana cements gain strength  very slowly and require, therefore, curing over a comparatively long period, but their ultimate  strength is approximately the same as that of ordinary Portland cement alone. A typical strength curve  is shown in Fig. 2.5.    Fig. 2.5. Strength de
velopment of concrete made with Portland cement and fly ash     ASTM Standard C 595-79 describes Portland-pozzolana cement as Type IP for general  concrete construction and Type P for use when high strengths at early ages are not required, and  limits the pozzolana content to between 15 and 40 per cent of the weight of the Portland-pozzolana  ceme
as the specific  gravity of pozzolanas is much lower than that cement; for instance, the specific gravity of fly ash  is	1.9 to 2.4, compared with 3.15 for cement. Thus replacement by weight results in a considerably  greater volume of the cementitious material in the mix. With replacement, concrete mixes have a  lower early strength than when Por
tland cement is used, but beyond about three months there is no  loss of strength. With lean mixes, there may even be a long-term gain of strength due to the  replacement (see Fig. 2.6). If equal early strength is required and pozzolana is to be used (e.g.  because of alkali - aggregate reactivity) then addition of pozzolana rather than partial re
reached out, partial replacement of Portland cement by  pozzolana reduces the permeability of concrete: a 7- to 10-fold reduction has been reported.    Fig. 2.6. Effect of partial replacement of  Portland cement by pozzolana on the strength  development of concrete made with 167 kg of "cement" material per cubic metre of  concrete (282 lb / yd3)  
 mixer, but it is essential that the pigments do not affect adversely  the development of strength of the cement or affect air entrainment. For instance, carbon black  reduces the air content of the mix. For this reason, some pigments are marketed in the U.S.A.  with an interground air-entraining agent; it is of course essential to be aware of thi
han usual should be used.   A typical compound composition of white Portland cement is given in Table 2.5 but the C3S and  C2S contents may vary widely. White cement has a slightly lower specific gravity than ordinary  Portland cement, generally between 3.05 and 3.10. The strength of white Portland cement is  usually lower than that of ordinary Po
strength is rather low.   Hydrophobic cement is similar in appearance to ordinary Portland cement but has a characteristic  musty smell. In handling, the cement seems more fluid than other Portland cements.   Hydrophobic cement should not be confused with waterproofed cements, which are claimed to  make a more impermeable concrete than ordinary Po
 air-entraining agent. Masonry cements  make a more plastic mortar than ordinary Portland cement, they also have a greater  water-retaining power and lead to lower shrinkage. The strength of masonry cements is  lower than that of ordinary Portland cement, particularly since a high air content is  introduced, but this low strength is generally an a
 likely to become extensively used in low-strength concrete. Indeed, one can  argue that, for many purposes, the existing high-quality Portland cements are too good,  so that there is intrinsic merit, and not only energy saving, in this development.   In a non-explicit way, some dilutants of Portland cement are allowed in the United  States, as AS
TM Standard C 465-74 allows processing additions, providing they do not  reduce the strength of the cement by more than 5 per cent. ASTM has also set up a  sub-committee to consider a possible standard for blended cements. Finland already allows  15 per cent of filler, and East Germany produces blended cements with 20 per cent of filler. In  Switz
energy expanding cement is quick-setting and rapid-hardening, reaching a strength  of about 7 MPa (1000 psi) in 6 hours, and 50 MPa (7000 psi) in 28 days. The cement has a  high resistance to sulphate attack.   A more recent development is expanding cement known as Type K in the ASTM  classification, developed in California. The ingredients of thi
ions. The important requirement is  that CaO, SO3, and especially Al2O3, become available for  ettringite formation at the right time. Specifically, a major part of it must form  after a certain strength has been attained; otherwise, the expansive force will be  dissipated in the deformation of the still plastic concrete and no stress against the 
ould  otherwise manifest themselves as tensile strains and possibly as cracks. A Standard  laying down the practice for the use of shrinkage-compensating concrete was issued by  the American Concrete Institute in 1977. It may be noted that many properties of  this concrete, such as strength, modulus of elasticity, inherent shrinkage, and  resistan
hough BS 915: 1972 prescribes  the conventional Le Chatelier test.    Hydration    The hydration of CA, which has the highest rate of strength development, results in the formation of CAH10, a small quantity of C2AH8, and of alumina gel (Al2O3.aq). With  time, these hexagonal CAH10 crystals, which are unstable both at normal and at higher  tempera
very high rate of strength  development  About 80 per cent of its ultimate strength is achieved at the age of 24  hours, and even at 6 to 8 hours the concrete is strong enough for the side formwork to  be struck and for the preparation for further concreting to take place. The high rate of  gain of strength is due to its rapid hydration, which in 
 and high-alumina cement clinker as aggregate,  with a water / cement ratio of 0.5, can reach a strength  of about 100 MPa (14 000 psi) in  24 hours, and 120 MPa (18000 psi) in 28 days at cool temperatures. This extremely high  strength development is due to the cementitious character of the aggregate but this  aggregate is, of  course, very expen
ltimate strength of such pastes is quite low.   Because of the rapid setting just described, in normal concrete construction it is  essential to make sure that the two cements do not come in contact with one another.  Thus, placing concrete made with one type of cement against concrete made with the other  must be delayed by at least 24 hours if h
e stress / strength  ratio.    Conversion of High-alumina Cement    The high strength of high-alumina cement concrete referred to on Page 90 (see also Fig.  2.8) is reached when the hydration of CA results in the formation of CAH10 with a small  quantity of C2AH8 and of alumina gel (Al2O3.aq). The hydrate CAH10 is, however,  chemically unstable bo
ng.   The practical interest in conversion lies in the fact that it leads to a loss of  strength of high-alumina cement concrete. The most likely explanation why this is so is  in terms of the densification of the calcium aluminate hydrates: typically, the density  would be 172 g / ml for CAH10 and 2.53 for  C3AH6. Thus, under conditions such that
see Fig. 2.10).    Fig. 2.10. Airflow through concrete: (a) unconverted high-alumina cement concrete;  (b) converted high-alumina cement concrete; (c) Portland cement concrete (temperature 22 to  24 C (72 to 75 F), relative humidity 36 to 41 per cent; pressure difference 10.7 kPa)     As shown on page 271, the strength of hydrated cement paste or 
of concrete is very  strongly affected by its porosity; porosity of 5 per cent can reduce the strength by  more than 30 per cent, and a 50 per cent reduction in strength would be caused by a  porosity of the order of 8 per cent. This order of magnitude of porosity can be induced  by conversion in high-alumina cement concrete.   It follows that, si
nce conversion takes place in concretes and mortars of any mix  proportions, they lose strength when exposed to a higher temperature, and the general  pattern of the strength loss versus time curves is similar in all cases. However, the  degree of loss is a function of the water / cement ratio of the mix, as shown in Fig.  2.11. The mix proportion
s and percentage loss are given in Table 2.8. It is clear that  the loss, either in megapascals or as a fraction of the strength of cold-cured concrete,  is smaller in mixes with low water / cement ratios than in mixes with high water / cement  ratios.  Fig. 2.11. Influence of the water / cement ratio on the strength of high-alumina cement  concre
te cubes curved in water at 18 C and 40 C for 100 days	Table 2.8: Influence of Water / Cement Ratio on Loss of Strength on Conversion.   It may be observed that the shape of the strength versus water / cement ratio curves  for storage at 18 C (Fig. 2.11) is dissimilar from the usual curves for Portland cement  concretes. This is characteristic 
te that the residual strength of mixes with moderate and  high water / cement ratios, say over 0.5, may be so low as to be unacceptable for most  structural purposes.   The relative difference in the loss of strength of high-alumina cement concrete  made with a high water / cement ratio and with a low water / cement  ratio should be noted. In the 
latter case, say at a water / cement ratio of 0.35, even  after full conversion the strength could be deemed to be adequate for all structural  purposes. Two caveats should, however, be expressed. First, under practical conditions  of manufacture of concrete, it is not possible to guarantee that the water / cement ratio will  not be occasionally e
xceeded by 0.05 or even by 0.10; this has been repeatedly demonstrated.  Second, converted high-alumina cement concrete, even if of adequate strength, is more porous and  therefore more liable to chemical attack than before conversion.   Not only the water / cement ratio but also the richness of the mix may affect the loss of  strength on conversi
on. For a given water / cement ratio, the leaner the mix the lower the  porosity of the cement paste and, consequently, the lower the relative loss in strength  with time.   The important point is that, despite some attempts to prove the contrary, all mixes  lose strength on conversion. Test results on the strength of 13 mm (1/2 in.) cubes of neat
	high-alumina cement paste show that even at a water / cement ratio of 0.30, the strength  at 50 C (122 F) is only 34 per cent of the strength at 18 C (64 F); at a water / cement ratio  of 0.50, the fraction is only 20 per cent. These values are much lower than those often  quoted for concrete but it must be stressed that the data refer to neat c
ement pastes.	The value of the minimum strength after conversion was also controversial, but there  are now available long-term data which give a clear and broad picture of the residual  strength which can be expected. Actual values may vary with the particular cement used,  but the following can be used as typical values:    (table).     The lo
ss of strength is often accompanied by a change in the colour of the cement paste  from blackish-grey to brown or yellow-brown. This happens probably because the increase  in porosity on conversion facilitates the oxidation of the ferrous compounds in the  cement paste. This is why conversion and change in colour may occur together, but  concrete 
na Cement    High-alumina cement concrete is one of the foremost refractory materials but it is  important to be clear about its performance over the full temperature range. Between  room temperature and about 500 C, high-alumina cement concrete loses strength to a greater  extent than Portland cement concrete, then up to 800 C the two are compara
ble, but above  about 1000 C high-alumina cement gives excellent performance.   Fig. 2.12 shows the behaviour of high-alumina cement concrete made with four different  aggregates over a temperature range up to 1100 C. The minimum strength varies between 5  and 26 per cent of the original value but, depending on the type of aggregate, above 700  to
 1000 C, there is a gain in strength due to the development of a ceramic bond. This  bond is established by solid reactions between the cement and fine aggregate, and  increases with an increase in temperature and with the progress of the reactions.    Fig. 2.12. Strength of high-alumina cement concretes made with different aggregates as a    func
of concrete while  acceleration primarily to the early strength development, i.e. to hardening (see p. 68),  more rapid setting being generally only coincidental. The classification of the British  Standard BS 5075 : Part 1 : 1974 is substantially similar; the standard lays down the  requirements for the various types of admixtures (Table 2.9).   
 various possible circumstances. This is due to the  marketing of admixtures largely as proprietary products. Acceptance requirements are  laid down by ASTM Standard C 494-79 and by British Standard BS 5075 : Part 1 : 1974.    Calcium Chloride    The addition of calcium chloride to the mix increases the rate of development of  strength, and this a
apid the natural rate of hardening of the cement the earlier  becomes apparent the action of the accelerator. Calcium chloride must not, however, be  used with high-alumina cement. With rapid hardening Portland cement the increase in  strength due to CaCl2 can be as much as 7 MPa (1000 psi) at 1 day while with ordinary  Portland cement this increa
se would be achieved only after 3 to 7 days. By the age of 28  days there is no difference between the strengths of rapid hardening cements with and  without CaCl2, but in the case of ordinary Portland cement the addition of CaCl2 would  still show an improved strength.   Hickey's results for cements of different types are shown in Fig. 2.13. The 
long-term  strength of concrete is believed to be unaffected by CaCl2. Calcium chloride is  generally more effective in increasing the early strength of rich mixes with a low  water / cement ratio than of lean ones.    Fig. 2.13. Influence of CaCl, on the strength of concretes made with different types of  cement (For Type V curve reference is 2.1
,	CaCl2 has been found to raise the resistance of concrete to erosion  and abrasion, and this improvement persists at all ages. When plain concrete is  steam cured, CaCl2 increases the strength of concrete and permits the use of a more  rapid temperature rise during the curing cycle (see p. 327).   The action of sodium chloride is similar to that 
of calcium chloride but is of lower  intensity. The effects of NaCl are also more variable and a depression in the heat of  hydration, with a consequent loss of strength at 7 days and later has been observed.  For this reason the use of NaCl is definitely undesirable. Barium chloride has been  suggested but it acts as an accelerator only under war
pounds.    Retarders    A delay in the setting of the cement paste can be achieved by the addition  to the mix of a retarding admixture. These admixtures generally slow down also the  hardening of the paste although some salts may speed up the setting but inhibit the  development of strength. Retarders do not alter composition or identity of produ
ing are more commonly used; these are described in the next section. Great care is  necessary in using retarders as in incorrect quantities they can totally inhibit the  setting and hardening of concrete. Cases are known of seemingly inexplicable results of  strength tests when sugar bags have been used for the shipment of aggregate samples to  th
of sugar, say 0.2 to 1 per cent of the weight of cement, will  virtually prevent the setting of cement. Such quantities of sugar can therefore be used  as an inexpensive "kill", for instance when a mixer or an agitator has broken down and  cannot be discharged.   When sugar is used as a controlled set retarder, the early strength of concrete is  s
everely reduced but beyond about 7 days there is an increase in strength of several  per cent compared with a non-retarded mix. This is probably due to the fact that  delayed setting produces a denser gel (cf. p. 318).   It is interesting to note that the effectiveness of an admixture depends on the time  when it is added to the mix: a delay of ev
effect of dispersion is to expose a greater surface area of cement to hydration,  which progresses therefore at a higher rate in the early stages. For this reason, there  is an increase in the strength of concrete, compared with a mix of the same water / cement  ratio but without the admixture. A more uniform distribution of the dispersed cement  
throughout the concrete may also contribute to the improved strength. The increase  in strength is particularly noticeable in very young concretes but under certain  conditions persists for a long time.   The influence of admixtures on strength varies considerably with the composition of  cement, the greatest increase in strength occurring when us
exert full hydrostatic pressure.    Fig. 2.17. Relation between German flow table spread and water content of concrete  with and without superplasticizer     The second use of superplasticizers is in the production of concrete of normal  workability but with an extremely high strength owing to a very substantial reduction  in the water / cement ra
tio. Water / cement ratios down to 0.28 have been used with 28-day  strengths of the order of 100 MPa (15 000 psi). The long-term strength is unimpaired, test  results being available up to 13 years. Generally speaking, superplasticizers can reduce  the water content for a given workability by 25 to 35 per cent (compared with half that  value in t
he case of conventional water-reducing admixtures), and increase the 24-hour  strength by 50 to 75 per cent and even greater increase occurs at earlier ages.  Practical mixes with a strength of 30 MPa (4300 psi) at 7 hours have been obtained  (see Fig. 2.18). With steam-curing or high-pressure steam-curing, even higher strengths  are possible.    
Fig. 2.18. The influence of the addition of superplasticizer on the early strength of  concrete made with a cement content of 370 kg / m3 (630 lb / yd3) and cast at room  temperature. Type III cement; all concretes of the same workability     When the strength at later ages is of primary importance, superplasticizers can be used  in concrete with 
ent comes into contact with water, C3S begins to hydrate rapidly, generating a  considerable amount of heat and making a significant contribution to the development of the early  strength, particularly during the first 14 days. In contrast C2S, which hydrates slowly and is mainly  responsible for the development in strength after about 7 days, may
 place very quickly, producing little increase in strength after about 24 hours. Of  the four principal compounds tricalcium aluminate, C3A, is the least stable and cements containing  more than 10 per cent of this compound produce concretes which are particularly susceptible to  sulphate attack. Tetracalcium aluminoferrite, C4AF, is of less impor
tance than the other three  compounds when considering the properties of hardened cement mortars or concrete.   From the foregoing, certain conclusions may be drawn concerning the nature of various  cements. The increased rate of strength development of rapid-hardening Portland cement arises from  its generally high C3S content and also from its i
ncreased fineness which, by increasing the specific  surface of the cement, increases the rate at which hydration can occur. The low rate of strength  development of low-heat Portland cement is due to its relatively high C2S content and low C3A and  C3S contents. An exceptionally low C3A content contributes to the increased resistance to sulphate 
 attack of sulphate-resisting cement. It should be noted that while there can be large differences in the  early strength of concretes made with different Portland cements, their final strengths will generally  be very much the same (see chapter 14).    Fineness    The reaction between the water and cement starts on the surface of the cement parti
cles and in  consequence the greater the surface area of a given volume of cement the greater the hydration. It  follows that for a given composition, a fine cement will develop strength and generate heat more  quickly than a coarse cement. It will, of course, also cost more to manufacture as the clinker must be  more finely ground. Fine cements, 
s the initial set. Further stiffening occurs as the volume of gel increases  and the stage at which this is complete and the final hardening process, responsible for its strength,  commences is known as the final set. The time from the addition of the water to the initial and final  set are known as the setting times (BS 4550: Part 3) and the spec
ol the rapid reaction of C3A with water, this reaction generating a  considerable amount of heat and causing the cement to stiffen within a few minutes after mixing. This  can only be overcome by adding more water and reagitating the mix. The addition of water results in  a reduction in strength. A false set also produces a rapid stiffening of the
ture of cement.    Strength    The strength of hardened cement is generally its most important property. The British Standard  strength requirements for Portland cements, obtained from mortar or concrete tests carried out in  accordance with BS 4550: Part 3, are summarised in table 12.7. It should be understood that cement  paste alone is not used
 for this test because of the unacceptably large variations of strength thus  obtained. Standard aggregates are used for making prescribed mortar or concrete test mixes to  eliminate aggregate effects from the measured strength of the cement.    TABLE 12.7 British Standard requirements for strength of the principal Portland cements    Soundness   
	Portland cements    Ordinary Portland cement has a medium rate of hardening, making it suitable for most concrete work.  It has, however, a low resistance to chemical attacks. Rapid-hardening Portland cement is in many  ways similar to ordinary Portland cement but produces a much higher early strength. The increased  rate of hydration is accompa
ents within the concrete. Its  slow rate of hydration is accompanied by a much slower rate of increase in strength than for ordinary  Portland cement although its final strength is very similar. Its resistance to chemical attack is greater  than that of ordinary Portland cements. Sulphate-resisting Portland cement, except for its high  resistance 
to sulphate attack, has principal properties similar to those of ordinary Portland cement.  Calcium chloride should not be used with this cement as it reduces its resistance to sulphate attack.  Extra-rapid-hardening Portland cement is used when very high early strength is required or for  concreting in cold conditions. Because of its rapid settin
g	and hardening properties the concrete should  be placed and compacted within about 30 minutes of mixing. Since the cement contains approximately  2 per cent calcium chloride dry storage is essential. Its use in reinforced or prestressed concrete is not  recommended (BS 8110: Part 1). Ultra-high early-strength Portland cement, apart from its much
	greater fineness and larger gypsum content, is similar in composition to ordinary Portland cement.  Although the early development in strength is considerably higher than with rapid-hardening cement  there is little increase after 28 days. It is suitable for reinforced and prestressed concrete work. White  and coloured Portland cements are simil
 the concrete  strength. These cements are used for architectural purposes. Hydrophobic Portland cement, owing to  the presence of a water-repellent film around its grain, can be stored under unfavourable conditions of  humidity for a long period of time without any significant deterioration The protective coating is  broken off during mixing and 
ulverized-fuel ash as pozzolana  (BS 6610) with 35 to 50 per cent pfa content. The use of these cements reduces the water demand of a  concrete mix, when compared with that using the corresponding ordinary Portland cement. The rate  of gain in strength and liberation of heat is lower than for ordinary Portland cement and this can be  useful for ma
ss concrete work; the use of pozzolanic cements (BS 6610) is particularly beneficial in  this respect. However, an equal 28 day strength can generally be achieved by increasing the cement  content of a Portland pulverized-fuel ash cement concrete by approximately 10 per cent compared  with that of an ordinary Portland cement mix. Like sulphate res
e high proportion of aluminate, about 40 per cent,  brings about an exceptionally high early strength and consequently it often becomes necessary to keep  concrete, in which this cement is used, continuously wet for at least 24 hours to avoid damage from  the associated heat of hydration.   Structural concrete made with high-alumina cement (common
ly known as HAC) presents a  serious problem however, as it can suffer a substantial reduction in its strength in most normal  environments, owing to the conversion of the hydrated cement to a more porous form. Thus, the use of  high-alumina cement in structural concrete is not permitted in BS 8110.   The conversion of high-alumina cement is parti
cularly sensitive to temperature, the  water - cement ratio and the richness of the mix at a given water - cement ratio, increasing as all these  factors increase. In certain circumstances, the residual strength of concrete can be considerably lower  than its one day strength. Water - cement ratios higher than 0.4 should generally be avoided.   Th
e exceptionally high early strength property of high-alumina cement makes it well suited  for repair work of limited life and for temporary works. It also has a wide application in refractory  concrete and its resistance to chemical attack, particularly sulphate attack, is greater than that of  Portland cements, although in its converted, more por
h, deformation, durability, toughness, hardness, volume change, porosity, relative density and  chemical reactivity.   The strength of an aggregate limits the attainable strength of concrete only when its  compressive strength is less than or of the same order as the design strength of concrete. In practice  the majority of rock aggregates used ar
e usually considerably stronger than concrete. While the  strength of concrete does not normally exceed 80 N mm-2 and is generally between 30 and  50 N mm-2 the strength of the aggregates commonly used is in the range 70 to 350 N mm-2, In  general, igneous rocks are very much stronger than sedimentary and metamorphic rocks. Because of the  irregul
ar size and shape of aggregate particles a direct measurement of their strength properties is not  possible. These are normally assessed from compressive strength tests on cylindrical specimens taken  from the parent rock and from crushing value tests on the bulk aggregate. For weaker materials, that  is, those with crushing values greater than 30
, the crushing value may be unreliable and the load  required to produce 10 per cent fines in the crushing test should be used (BS 812). It should be noted  that the strength test on rock specimens was deleted from BS 812: Part 3 in 1975 and is consequently  less used than the tests on bulk aggregate. The results of these tests for the strength pr
refore  be employed as the test for assessing aggregate strength (Spence et al ., 1974). The AIV test has  advantages over the ACV test of simplicity and cheapness in operation. It does not require the more  elaborate facilities of a testing laboratory, the equipment is portable and the small sample required  makes it a particularly useful test fo
hat of the cement paste this too may adversely  affect the concrete performance.   Aggregate porosity is an important property since it affects the behaviour of both freshly  mixed and hardened concrete through its effect on the strength, water absorption and permeability of  the aggregate. An aggregate with high porosity will tend to produce a le
er content is to be kept constant and the required  workability and strength of concrete maintained. Concrete mix proportions are normally based on the  weight of the aggregates in their saturated surface-dry condition and any change in their moisture  content must be reflected in adjustments to the weights of the aggregates used in the mix. Sever
lag aggregates produce concretes with similar strength to natural  aggregates but with improved fire resistance. Broken-brick aggregates are also very fire resistant, but  should not be used for normal concrete if its soluble sulphate content exceeds 1 per cent.    Lightweight aggregate    Lightweight aggregates find application in a wide variety 
p  water, is acceptable for mixing concrete. The impurities that are likely to have an adverse effect when  present in appreciable quantities include silt, clay, acids, alkalis and other salts, organic matter and  sewage. The use of seawater does not appear to have any adverse effect on the strength and durability  of Portland cement concrete but 
and extent of contamination as  prescribed in BS 3148. The quality of water may also be assessed by comparing the setting time and  soundness of cement pastes made with water of known quality and the water whose quality is suspect.   The use of impure water for washing aggregates can adversely affect strength and durability if  it deposits harmful
results in some reduction in concrete strength. Since improvements in  workability can permit a reduction in the water content the loss in strength can be minimised. The  amount of entrained air in concrete is dependent on the type of admixture and dosage used, as well as  on the cement type, aggregate type and grading, mix proportions, ambient te
rk involving water leakage. Because of their adverse effect  on subsequent strength development these admixtures should not be used where the final concrete  strength is an important consideration. Setting and hardening accelerators increase the rate of both  setting and early strength development. The most common admixture for this purpose is cal
.12) and their  effect on concrete is dependent on dosage, cement type and mix proportions used and the ambient  temperature of concrete. There are many accelerators which can achieve an increase in 1 day strength  of up to 100 per cent over the corresponding normal mix, which is well in excess of the value  specified in BS 5075: Part 1 (see table
 with a corresponding  increase in final concrete strength. The lignin-based retarders result in some air-entrainment and tend  to increase cohesiveness and reduce bleeding although drying shrinkage may be increased . The  hydroxy-carboxylic retarders, however, tend to increase bleeding.   The use of retarders on their own is declining however, wi
an be used in three ways. First, to increase concrete workability for a given  water - cement ratio and nominal strength; this allows easier placing and compaction of concrete.  Second, to increase concrete strength without the addition of further cement owing to the reduced  water requirement of a mix at a given workability; this allows the produ
ction of high strength  concrete. Third, to reduce cement content of a mix at a given workability and strength by reducing its  water content while maintaining the original water - cement ratio; this reduces the cost of a mix. The  reduction in cement content will result in lower maximum temperatures and hence reduce the risk of  shrinkage crackin
nal mix composition and without causing a strength reduction. The self-levelling property of  flowing concrete means that it can be placed with little or no compaction and is therefore particularly  suitable for heavily reinforced and inaccessible sections, or where rapid placement of concrete is  required. To avoid bleeding, segregation and other
aken into account at the design stage of the structure.  An alternative use of superplasticizers is in the production of high strength concrete by reducing the  mix water content, and hence water - cement ratio, while maintaining the original workability. A  reduction in the water content up to 25 per cent is possible.    Bonding admixtures    The
se are organic polymer emulsions used to enhance the bonding properties of concrete,  particularly for patching and remedial work. The bonding admixtures are known also to  increase the abrasion resistance of concrete and its tensile strength, but some reduction in  compressive strength also occurs.    Water-repelling agents    These are the least
inker rather than added during mixing. Pigments used for this purpose are formulated  from both natural and synthetic materials and in the United Kingdom should conform to  BS 1014. Pigments do not normally affect the concrete properties although those based on  carbon may cause some loss of strength at early ages and can also reduce the effective
re  increased workability, decreased bleeding and reduced tendency to segregation. In the  hardened concrete, pfa promotes a denser matrix structure and gives good long-term  strength development. For detailed information on the mechanisms by which pfa affects the  fresh and hardened concrete, the reader is referred to Dhir (1986), which also prov
ides  information concerning the characterisation of pfa and the mix proportioning, engineering  properties and durability of concrete containing pfa. Briefly, a concrete  mix incorporating pfa can be designed for a given workability and strength comparable  with the corresponding mix in which Portland cement alone is used, provided the  variation
ng  to pfa consumption rather than disposal) is of greater national interest.  5 Strength of Concrete      Strength of concrete is commonly considered its most valuable property, although in many  practical cases other characteristics, such as durability and impermeability, may in  fact be more important. Nevertheless, strength usually gives an ov
erall picture of the  quality of concrete because strength is directly related to the structure of the  hardened cement paste.   The mechanical strength of cement gel was discussed on Page 35; below, some empirical  relations concerning the strength of concrete will be considered.      Water / Cement Ratio    In engineering practice, the strength 
of concrete at a given age and cured at  a prescribed temperature is assumed to depend primarily on two factors only: the  water / cement ratio and the degree of compaction. The influence of air voids on strength  was discussed on page 204, and at this stage we shall consider fully-compacted concrete only:  in practice this is taken to mean that t
he hardened concrete contains about 1 per cent of air  voids.   When concrete is fully compacted its strength is taken to be inversely proportional to the  water / cement ratio according to the "law" established by Duff Abrams in 1919. He found  strength to be equal to    (formula)    where w / c represents the water / cement ratio of the mix (ori
ginally taken by  volume), and K1, and K2 are empirical constants. A typical strength versus water / cement  ratio curve is shown in Fig. 5.1.    Fig. 5.1. The relation between strength and water / cement ratio of concrete     Abrams' "law", although established independently, is a special case of  a general rule formulated by Feret in 1896. This 
was in the form    (formula)    where f, is the strength of concrete, c, w, and a are the absolute volumes of cement,  water, and air respectively, and K is a constant.   It may be recalled that the water / cement ratio determines the porosity of the hardened  cement paste at any stage of hydration (see p. 32). Thus the water / cement ratio and th
e  degree of compaction both affect the volume of voids in concrete, and this is why the  volume of air in concrete is included in Feret's expression.   The relation between strength and the total volume of voids is not a unique property of  concrete but is found also in other brittle materials in which water leaves behind pores: for  instance, th
e strength of plaster is also a direct function of its void content (see Fig. 5.2).  Moreover, if the strength of different materials is expressed as a fraction of the strength at  a zero porosity, a wide range of materials conform to the same relation between relative  strength and porosity, as shown in Fig. 5.3 for plaster, steel, iron, alumina 
and zirconia.  This general pattern is of interest in understanding the role of voids in the strength of  concrete. Moreover, the relation of Fig. 5.3 makes it clear why cement compacts (see p.  30), which have a very low porosity, have a very high strength.   Strictly speaking, strength of concrete is probably influenced by the volume of all void
s in  concrete: entrapped air, capillary pores, gel pores, and entrained air if present. An  example of the calculation of the total air content may be of interest and is given in a  footnote.    Fig. 5.2. Strength of plaster as a function of its void content    Fig. 5.3. Influence of porosity on relative strength of various materials     The infl
uence of the volume of pores on strength can be expressed by a power function of  the type    (formula),    where  fc = strength of concrete with porosity p   fc,0 = strength at zero porosity   and n = a coefficient, which need not be constant.   However, the shape of the pores is also a factor. The shape of the solid  particles and their modulus 
of elasticity also influence the stress distribution,  and therefore stress concentration, within the concrete.   The possibility of obtaining very high strengths at extremely low  porosities was demonstrated by the application of high pressure (340 MPa  (50 000 psi)) with a simultaneous high temperature (250 C (480 F)) to  cement paste (see Fig. 
5.4(a)). Porosity of about 1 per cent was obtained.  The degree of hydration ranged between 0.29 and 0.37 but did not seem to be  related to strength. The specimens had a compressive  strength of about 660 MPa (95 000 psi) and tensile splitting strength of 64 MPa (9250  psi). These and other tests show that there is a direct relation between stren
gth and  porosity, although the exact form of this relation has not been established: specifically, it is  not clear whether the logarithm of porosity is linearly related to strength or to its logarithm  (compare Fig. 5.4(a) and (b)).     Fig. 5.4(a). Relation between compressive strength and logarithm of porosity of  cement paste compacts for var
ious treatments of pressure and high temperature    Fig. 5.4(b). Relation between logarithm of compressive strength and logarithm of porosity  of cement paste compacts for various treatments of pressure and high temperature (after  ref. 5.107)    Fig. 5.1 shows that the range of validity of the water / cement ratio rule is limited. At the  lower e
n of  strength, particularly when large size aggregate is used. Thus at later ages, in this type of  mix, a lower water / cement ratio would not lead to a higher strength. This behaviour may  be due to stresses induced by shrinkage, whose restraint by aggregate particles causes  cracking of the cement paste or a loss of the cement - aggregate bond
.   From time to time the water / cement ratio rule has been criticized as not being sufficiently  fundamental (see the following section). Nevertheless, in practice the water / cement ratio  is the largest single factor in the strength of fully compacted concrete. Perhaps the best  statement of the situation is that by Gilkey:   "For a given ceme
nt and acceptable aggregates, the strength that may be  developed by a workable, properly placed mixture of cement, aggregate, and water (under  the same mixing, curing, and testing conditions) is influenced by the:   (a) ratio of cement to mixing water   (b) ratio of cement to aggregate   (c) grading, surface texture, shape, strength, and stiffne
ss of aggregate particles   (d) maximum size of the aggregate"   We can add that factors (b) to (d) are of lesser importance than factor (a)  when usual aggregates up to 40 mm (1 1/2 in.) maximum size are employed They are,  nevertheless, present since, as pointed out by Walker and Bloem, the strength of concrete  results from: (1) the strength of
 the mortar; (2) the bond between the mortar and the  coarse aggregate; and (3) the strength of the coarse aggregate particle, i.e., its ability to  resist the stresses applied to it".   Fig. 5.5. shows that the graph of strength versus water / cement ratio is   approximately in the shape of a hyperbola. This applies to concrete made with any give
n  type of aggregate and at any given age. It is a geometrical  property of a hyperbola  y = k / x that y against 1 / x plots as a straight line. Thus the relation between the strength  and the cement / water ratio is approximately linear in the range of cement / water ratios  between about 1 2 and 2 5. This linear relationship is clearly more con
venient to use than  the water / cement ratio curve, particularly when interpolation is desired. Fig. 5.6 shows  the data of Fig. 5.5 plotted with the cement / water ratio as abscissa. The values used apply  to the given cement only, and in any practical case the actual relation between strength and  cement / water ratio has to be determined.    F
ig. 5.5. Relation between 7-day strength and water / cement ratio for concrete made with  a rapid-hardening Portland cement    Fig. 5.6. A plot of strength against cement / water ratio for the data of Fig. 5.5    It must be admitted that the relations discussed here are not precise, and other  approximations can be made. For instance, it has been 
suggested that as an approximation  the relation between the logarithm of strength and the natural value of the water / cement  ratio can be assumed to be linear (cf. Abrams' expression). As an illustration, Fig. 5.7 gives  the relative strength of mixes with different water / cement ratios, taking the strength at the  water / cement ratio of 0 4 
as unity.   The pattern of strength of high-alumina cement concrete is somewhat different from that  of concrete made with Portland cement, in that strength increases with the cement / water  ratio at a progressively decreasing rate.    Fig. 5.7. Relation between logarithm of strength and water / cement ratio      Gel / Space Ratio     The influen
ce of the water / cement ratio on strength does not truly constitute a law as the  water / cement ratio rule does not include many  qualifications necessary for its validity. In  particular, strength at any water / cement ratio depends on the degree of hydration of  cement  and its chemical and physical properties; the temperature at which hydrati
on takes place;  the air content of the concrete; and also the change in the effective water / cement ratio  and the formation of fissures due to bleeding.   It is more correct, therefore, to relate strength to the concentration of the solid products  of hydration of cement in the space available for these products; in this connection it may  be r
elevant to refer again to Fig. 1.9. Powers has determined the relation between the  strength development and the gel / space ratio. This ratio is defined as the ratio of the  volume of the hydrated cement paste to the sum of the volumes of the hydrated cement  and of the capillary pores.   On page 27 it was shown that cement hydrates to occupy mor
tion of cement that has hydrated.  Then, the volume of gel = 2 06 cvca, and the total space available to the gel is cvca + wo.  Hence, the gel / space ratio is    (formula).     Taking the specific volume of dry cement as 0 319 ml / g, the gel / space ratio becomes    (formula).     The compressive strength of concrete tested by Powers was found t
o be 234 x3 MPa  (34 000 x3 psi), and is independent of the age of the concrete or its mix proportions. The actual  relation between the compressive strength of mortar and the gel / space ratio is shown in Fig. 5.8: it  can be seen that strength is approximately proportional to the cube of the gel / space ratio, and the  figure 234 MPa (34 000 psi
) represents the intrinsic strength of gel for the type of cement and of  specimen used. Numerical values differ little for the usual range of Portland cements  except that a higher C3A content leads to a lower strength at a given gel / space ratio.   If A cm3 of air are present in the set paste, the ratio wo / c in the above  expression would be 
replaced by    (formula) (See Fig. 5.9).    Fig. 5.8. Relation between the compressive strength of mortar and gel / space ratio    Fig. 5.9. Relation between the compressive strength of mortar and gel / space ratio  modified to include entrapped air voids     The resulting formula for strength would be similar to that of Feret but the ratio used h
ere  involves a quantity proportional to the volume of hydrated cement instead of the total  volume of cement and is thus applicable at any age.   The expression relating strength to the gel / space ratio can be written in a number of ways.  It may be convenient to utilize the fact that the volume of non-evaporable water, wn, is  proportional to t
he volume of the gel; and also that the volume of mixing water, wo, is related  to the space available for the gel. The strength, fc, in pounds per square inch, for f, greater than about  2000 psi, when the relation is approximately linear, can then be written in the form -    (formula).    Alternatively, the surface area of gel, Vm, can be used. 
Then    (formula).    Fig. 5.10 shows Powers' actual data for cements with low C3A contents.    Fig. 5.10. Relation between the strength of cement paste and the ratio of surface area of  gel Vm, to the volume of mixing water wo     The above expressions have been found to be valid for many cements but the numerical  coefficients may depend on the 
intrinsic strength of the gel produced by a given cement. In  other words, the strength of the paste depends primarily on the physical structure of the gel but  the effects of the chemical composition of cement cannot be neglected; however, at later ages  these effects become minor only. Another way of recognizing the properties of the gel is to s
ay  that strength depends primarily on porosity but also on the ability of the material to resist crack  propagation, which is a function of bonding. Poor bond between two crystals can be  considered to be a crack.      Effective Water in the Mix    The relations discussed so far in this chapter involve the quantity of water in the mix. This  need
to a saturated and surface-dry condition is considered to be the effective water of the mix.  For this reason, the strength curves of Road Note No. 4 (see Fig. 10.2) are based on the water  in excess of that absorbed by the aggregate. On the other hand, McIntosh and Erntroy's  data refer to the total water added to a dry aggregate. This condition 
rete on its strength has been repeatedly mentioned, and  it should be possible to relate this factor to the actual mechanism of failure. For this  purpose, concrete is considered as a brittle material, even though it exhibits a small amount of  plastic action, as fracture under static loading takes place at a moderately low total strain; a strain 
 of 0 001 to 0 005 at failure has been suggested as the limit of brittle behaviour.    Strength in Tension    The actual (technical) strength of cement paste or of similar brittle materials such as stone is very  much lower than the theoretical strength estimated on the basis of molecular cohesion, and  calculated from the surface energy of a soli
d assumed to be perfectly homogeneous and flawless.  The theoretical strength has been estimated to be as high as 10.5 GPa (1.5 x 10'6 psi).   This discrepancy can be explained by the presence of flaws postulated by Griffith. These  flaws lead to high stress concentrations in the material under load so that a very high stress  is reached in very s
mall volumes of the specimen with a consequent microscopic fracture,  while the average (nominal) stress in the whole specimen is comparatively low. The flaws vary in  size and it is only the few largest ones that cause failure: the strength of a specimen is thus a problem  of statistical probability, and the size of the specimen affects the proba
ble nominal stress at  which failure is observed.   Cement paste is known to contain numerous discontinuities - pores, fissures and voids -  but the exact mechanism through which they affect the strength is not known. The voids  themselves need not act as flaws, but the flaws may be cracks in individual crystals associated with the  voids - or cau
ypothesis postulates microscopic failure at the location of a flaw, and it is usually  assumed that the "volume unit" containing the weakest flaw determines the strength of the  entire specimen. This statement implies that any crack will spread throughout the section of the  specimen subjected to the given stress, or, in other words, an event taki
lexure specimens at the moment of incipient failure are higher  than the strength determined in uniform direct tension: in the latter case the propagation of fracture  is not blocked by the surrounding material. Some actual data on the relation between the  strength in flexure and in direct tension are given in Fig. 8.10.   We can see then that in
 a given specimen different stresses will produce fracture at different  points, but it is not possible physically to test the strength of an individual element without  altering its condition in relation to the rest of the body. If the strength of a specimen is governed  by the weakest element in it, the problem becomes that of the proverbial wea
kest link in a  chain. In statistical terms we have to determine the least value (i.e. the strength of the most  effective flaw) in a sample of size n, where n is the number of flaws in the specimen. The chain  analogy may not be quite correct because in concrete the links may be arranged in parallel as well  as in series but computations on the b
asis of the weakest link assumption yield results of the  correct order. It follows that the strength of a brittle material such as concrete cannot be described  by an average value only: an indication of the variability of strength must be given, as well as  information about the size and shape of the specimens. These factors are discussed in  Ch
o	that fracture can take place. Orowan calculated the maximum tensile stress at the tip of the  flaw of the most dangerous orientation relative to the principal stress axes as a function of  the two principal stresses P and Q. The fracture criteria are represented graphically in Fig. 5.11,  where K is the tensile strength in direct tension. Fractu
 concrete compression test specimens. The nominal  compressive strength in this case is 8K, i.e. eight times the tensile strength determined in a direct  tension test. This figure is in good agreement with the observed values of the ratio of the  compressive to tensile strengths of concrete. There are, however, difficulties in reconciling  certain
ing tensile strain that  determines the strength of concrete under static loading:  this is usually assumed to be between  100 x 10'-6 and 200 x 10'-6 The failure criterion of limiting tensile strain is supported by an  analysis recently advanced by Lowe. It has been found that at the point of  initial cracking the  strain on the tension face of a
between about 0 11 for high strength concrete and 0 21 for weak  mixes (see p. 370), and it is significant that the ratio of the nominal tensile and compressive  strengths for different concretes varies in a similar manner and between approximately the  same limits. There is thus a possibility of a connection between the ratio of nominal strengths
 is probable that ultimate strain is the criterion of failure,  but the level of strain varies with the strength of concrete: the higher the strength the lower the  ultimate strain. Some typical, but by no means general, values are given below:   Under triaxial compression, failure must take place by crushing: the mechanism is, therefore,  quite d
ifferent from that described above. An increase in lateral compression increases the  axial load that can be sustained, as shown for instance in Fig. 5.12. With very high lateral  stresses extremely high strengths have been recorded (Fig. 5.13). It may be noted that if the  development of pore water pressure in concrete is limited by allowing the 
displaced pore water to  escape through the loading platens, then the apparent strength is higher.    Fig. 5.12. Influence of lateral stress on the axial stress at failure of neat cement paste and  of mortar    Fig. 5.13. Influence of high lateral stress on the axial stress at failure of concrete     A lateral tensile stress has a similar influenc
 repeated loading is applied - a condition frequently met with in practice.  Fatigue strength of concrete is considered on page 338.     Microcracking    Investigations have shown that very fine cracks at the interface between  coarse aggregate and cement paste exist in fact even prior to application of the load on  concrete. These cracks remain s
table up to about 30 per cent  or more of the ultimate load and then begin to increase in length, width,  and number. The overall stress under which they develop is sensitive to the  water / cement ratio of the paste. This is the stage of slow crack propagation.  At 70 to 90 per cent of the ultimate strength, cracks open through the mortar  (cemen
etween the beginning of loading and a stress  equal to about 0.85 of the prism strength (100 mm by 100 mm by 300 mm (4 by 4 by 12 in.)  prisms). A further increase in stress resulted in a large increase in the total length of cracks.  At a stress / strength ratio of about 0.95 (determined on prisms), not only interface cracks but also  mortar crac
stress / strength ratio in compression (based on prisms)    Fig. 5.14 also shows the crack development under a cyclic stress alternating between zero and  0.85 of the prism strength. Immediately prior to failure, the cracks became longer and wider.  Likewise, sustained loading at a stress / strength ratio of 0.85 led to an increase in cracking pri
 properties of aggregate affect thus the cracking, as distinct from the ultimate, load in  compression and the flexural strength in the same manner, so that the relation between the  two quantities is independent of the type of aggregate used. Fig. 5.15 shows Jones and  Kaplan's results, each symbol representing a different type of coarse aggregat
e. On the other hand,  the relation between the flexural and compressive strengths depends on the type of coarse  aggregate used (see Fig. 5.16) since (except in high strength concrete) the properties of aggregate,  especially its shape and surface texture, affect the ultimate strength in compression very much less  than the strength in tension or
 the cracking load in compression. In experimental concrete, entirely  smooth coarse aggregate led to a lower compressive strength, typically by 10 per cent, than when  roughened.    Fig. 5.15. Relation between flexural strength and compressive stress at cracking for  concretes made with different coarse aggregates    Fig. 5.16. Relation between c
ompressive strength and indirect - tensile strength for concretes  of constant workability made with various aggregates (water / cement ratio between 0.33 and  0.68, aggregate / cement ratio between 2.8 and 10.1)     The influence of the type of coarse aggregate on the strength of concrete varies in magnitude and  depends on the water / cement rat
io of the mix. For water / cement ratios below 0.4, the use of  crushed aggregate has resulted in strengths up to 38 per cent higher than when gravel is used. The  behaviour at a water / cement ratio of 0.5 is shown in Fig. 5.17. With an increase in the  water / cement ratio, the influence of aggregate falls off, presumably because the strength of
 the  paste itself becomes paramount, and at a water / cement ratio of 0.65 no difference in the  strengths of concretes made with crushed rock and gravel has been observed.    Fig. 5.17. Relation between compressive strength and age for concretes made with  various aggregates (water / cement ratio = 0.5)     The influence of aggregate on flexural
 strength seems to depend also on the moisture  condition of the concrete at the time of test.  The shape and surface texture of coarse aggregate affect also the impact strength of  concrete, the influence being qualitatively the same as on the flexural strength (see p. 695).   Kaplan observed that the flexural strength of concrete is generally lo
wer than the  flexural strength of corresponding mortar. Mortar would thus seem to set the upper limit  to the flexural strength of concrete and the presence of the coarse aggregate generally  reduces this strength. On the other hand, the compressive strength of concrete is higher  than that of mortar, which, according to Kaplan, indicates that th
e mechanical  interlocking of the coarse aggregate contributes to the strength of concrete in  compression. This behaviour has not, however, been confirmed to apply generally, and the  question of the influence of aggregate on strength is considered further in the next  section.    Influence of Richness of the Mix on Strength    The anomalous beha
viour of extremely rich mixes was mentioned on p. 272, but the  aggregate / cement ratio affects the strength of all medium and high-strength concretes,  i.e. those with a strength of about 35 MPa (5000 psi) or more. There is no doubt that the  aggregate / cement ratio is only a secondary factor in the strength of concrete but it has been  found t
hat for a constant water / cement ratio a leaner mix leads to a higher strength (see Fig. 5.18).	Fig. 5.18. Influence of the aggregate / cement ratio on strength of concrete     This behaviour is probably associated with the absorption of water by the aggregate: a  larger amount of aggregate absorbs a greater quantity of water, the effective  w
dverse effect on strength.   Recent studies on the influence of aggregate content on the strength of concrete with a  given quality of cement paste indicate that when the volume of aggregate is increased  from zero to 20, there is a gradual decrease in compressive strength, but between 40 and  80 per cent there is an increase. The pattern of behav
iour is shown in Fig. 5.19. The  reasons for this effect are not clear, but it is the same at all water / cement ratios. The  influence of the volume of aggregate on tensile strength is broadly similar (Fig. 5.20).    Fig. 5.19. Relation between the compressive strength of cylinder (100 mm diameter,  300 mm in length) and volume of aggregate at a 
constant water / cement ratio of 0.50    Fig. 5.20. Relation between direct tensile strength and volume of aggregate at a  constant water / cement ratio of 0.50     The effects are smaller in cubes than in cylinders or prisms. In consequence, the ratio  of cylinder strength to cube strength (cf. p. 543) decreases as the volume of aggregate  increa
 acting. Mather pointed out  that, ideally, it should be possible to express the failure criteria under all possible  stress combinations by a single stress parameter, such as strength in uniaxial tension.  However, such a solution has not yet been found.   To develop equations for concrete strength, Berg considered the stress, ocr, at the  initia
tion of crack propagation, the maximum nominal stress, opr, on a concrete prism,  and the cleavage strength, ocl, of concrete in the direction normal to the applied  compressive stress. The cleavage strength is approximately equal to the strength in  axial tension. He showed that for the general case of a principal stress system o1, o2, and o3  (w
in Fig.	5.21. The criterion ceases to apply, however, when o2  and o3 have values such that the transverse cleavage strength cannot be overcome. The  behaviour of concrete is then no longer brittle but plastic. Some data on the strength  of concrete when o2 = o3 and o1 > o2 are shown in Fig. 5.13.    Fig. 5.21. Berg's equation for strength of conc
g. 5.22  that under a biaxal stress o1 = o3, strength is only 16 per cent higher than in uniaxial compression;  biaxial tensile strength is no different from uniaxial tensile strength. These findings were  confirmed by other workers. Experimental data on interaction are plotted in Fig. 5.23;  these were obtained with brush platen loading and by th
tress as measured by various investigators.  Wet or air-dried concrete (fc = compressive strength )     The level of uniaxial compressive strength virtually does not affect the shape of the  curve or the magnitude of the values given by it; the prism strength range tested  was 19 to 58 MPa (2700 to 8350 psi) and both the water / cement ratio and c
ement content  varied widely. However, in compression - tension and in biaxial tension, the relative  strength at any particular biaxial stress combination decreases as the level of uniaxial  compressive strength increases. This accords with the general observation that the  ratio of uniaxial tensile strength to uniaxial compressive strength decre
ases as the  compressive strength level rises (see p. 301); in these tests, the ratio was 0.11, 0.09,  and 0.08 at a uniaxial compressive strength level of 19, 31, and 58 MPa (2700, 4450, and  8350 psi) respectively.   Generally, triaxial compression increases the strength of weaker or leaner concrete  relatively more than that of stronger or rich
er concrete. For the range of  conventional concretes, Hobbs found that under triaxial compression, the major principal  stress at failure, ol, can be expressed as    (formula),    where o3 = minor principal stress,  and fcyl = cylinder strength,  all values being in MPa.   The limited information on lightweight aggregate concrete suggests that th
e influence  of o3 is not as large as with normal aggregates; therefore the coefficient 4.8 in the  above equation can be reduced to about 3.2.  The combined strength results for concretes in triaxial compression and in biaxial  compression plus tension, may be represented by the equation    (formula),    all values again being in MPa and compress
rength, fcyl.  The generality of this equation should not be overestimated for, as Hobbs points out,  the tensile strength and compressive strength of concrete are not equally affected by  the aggregate type and grading and by the direction of the applied stress relative to  the direction of casting. In each case, the tensile strength is more sens
itive.    Fig.	5.24. Failure stresses in concrete under biaxial stress     Recognizing that the strength of concrete cannot be predicted by considering  limitations on the compressive, tensile, and shearing stresses independently of each  other, Bresler and Pisters proposed a general equation of strength of concrete. The  equation uses criteria in
d shearing octahedral stresses given by    (formula).	Octahedral stresses are so named because they occur on the sides of an octahedral  element formed by planes whose normals make equal angles with the principal stress axes.   To avoid the influence of the properties of the specimen on the compressive strength of  concrete, the octahedral str
esses are rendered dimensionless by dividing them by the  nominal uniaxial compressive strength of the specimen used, fc,. An equation of the type    (formula)    is obtained, and this gives the combination of stresses that will just cause failure. A  typical curve is shown in Fig. 5.25. The effect of I3 has been found to be significant  so that t
er development of an equation of strength is still needed. Such information  will be of value in the design of structures such as shells, plates, pressure vessels  and even in parts of flexural members.    Fig. 5.25. Relation between octahedral shear stress and octahedral normal stress at  failure for 20 MPa (3000 psi) concrete subjected to uniaxi
al compression and to  biaxial compression - tension	Effect of Age on Strength of Concrete    The relation between the water / cement ratio and the strength of concrete applies to  one type of cement and one age only. On the other hand, the  strength versus gel / space  ratio relationship has a more general application because the amount of g
el present in  the cement paste at any time is itself a function of age and type of cement. In other  words, different cements require a different length of time to produce the same  quantity of gel.   The rate of gain of strength of different cements was discussed in Chapter 2, and Figs.  2.1 and 2.2 show typical strength - time curves. The influ
ence of the curing conditions on  the development of strength is considered later in this chapter, but here we are  concerned with the practical problem of strength of concrete tested at different ages.  In the majority of cases, the tests are made at the age of 28 days when the strength of  concrete is considerably lower than its long-term streng
th. In the past, the gain in  strength beyond the age of 28 days was regarded merely as contributing to an increase in  the factor of safety of the structure, but since 1957 the codes of practice for  reinforced and prestressed concrete allow the gain in strength to be taken into account  in the design of structures that will not be subjected to l
oad until a later age except  when no-fines concrete is used; with some lightweight aggregates, verifying tests are  advisable. The values of strength given in the British Code of Practice CP 110 : 1972,  based on the 28-day compressive strength, are given in Table 5.1, but they do not, of  course, apply when accelerators are used.    Table 5.1: B
ritish Code of Practice CP 110 : 1972 Factors for Increase in Compressive  Strength of Concrete with Age (Average Values)     The rate of gain is the strength of concrete is of interest also in connection with  testing. It is often desirable to check the suitability of a mix long before the results  of the 28-day test are available. However, even 
if the curing conditions are carefully  controlled, the prediction of the 28-day strength from that measured at the age of 7  days is difficult, mainly because of the variation in the intrinsic rate of hardening of  commercial cements. Furthermore, mixes with a low water / cement ratio gain strength,  expressed as a percentage of long-term strengt
h, more rapidly than mixes with higher  water / cement ratios (Fig. 5.26). This is because in the former case the cement grains are closer to  one another and a continuous system of gel is established more rapidly. For this reason,  a general extrapolation of, say, the 7-day strength to the 28-day strength is not easy,  even when dealing with one 
cement only.	Fig. 5.26. Relative gain of strength with time of concretes with different water / cement  ratios, made with ordinary Portland cement     When no specific data on the materials used are available, the 28-day strength may be  assumed to be 1.5 times the 7-day strength and, as an alternative to the specified  28-day strength of test 
cubes, the Code of Practice CP 114 (1969) accepts a 7-day strength  equal to not less than 3 of the required 28-day strength. Tests have shown that for  concretes made with British ordinary Portland cement the ratio of the 28-day to 7-day  strengths lies generally between 1.3 and 1.7 but the majority of the results fall above  1.5. The extrapolati
on of 7-day strength according to the Code of Practice is therefore  quite reliable. According to ASTM Standard, American Type 1 cements have a lower ratio  of 28-day to 7-day strength: 57 per cent of them have a ratio of at least 1.32 but only 2  per cent higher than 1.50.   It should be noted that in a hot climate the early strength gain is high
 and the ratio  of the 28-day to 7-day strengths tends to be lower than in cooler weather. This is also  the case with some lightweight aggregate concretes.   In Germany, the relation between the 28-day strength, fc,28 and the 7-day strength,  fc,7 is often taken to lie between    (formula),    fc being expressed in MPa. For strengths expressed in
 psi the two constant terms become  respectively 150 and 850.   Hummel recommends the use of an approximately linear relation between the strength  and the logarithm of age within the range of 3 days to 2 months. Thus if the strength is  determined at 3 and 7 days, it is possible to estimate the 28-day strength by  extrapolation. The validity of t
his approach between about 28 and 250 days is shown in  Fig. 5.27.    Fig. 5.27. Gain of strength of air-stored concrete with time (Austrian Portland cement;  cement and aggregate from one source; water / cement ratio = 0.65; cement content = 260 kg / m3  (440 lb / yd3); curing 7 days in wet sand, then in air at room temperature)     Pioneiro sugg
ested an expression of the type    (formula),   where fc,7 and fc,28 are strengths at 7 and 28-days respectively, and k1 and k2 are coefficients,  different for each cement and curing condition used. The value of k1 ranges from about 0.3 to  0.8, and that of k2 from 3 to 6.  All the expressions mentioned here apply only to concrete made with ordin
ary Portland  cement. Many of the other cements gain strength at different rates, and when they are  used the prediction of strength should be based on experimental results.   As far as the really long-term strength is concerned, American Portland cements made at  the beginning of the century (which had a high C2S content and a low specific surfac
e)  led to an increase in the strength of concrete stored outdoors which was proportional to  the logarithm of age up to 50 years. The 50-year strength was typically 2.4 times the  28-day strength. However, cements made since the 1930s (with a lower C2S content and  a higher specific surface) reach their peak strength between 10 and 25 years and  
thereafter undergo some retrogression of strength. German Portland cements made  in 1941, when used in concrete stored outdoors, led after 30 years to a strength 2.3 times  the 28-day strength. The relative increase in strength was greater at higher  water / cement ratios. By comparison, Portland blast-furnace cement led to a 3.l-fold  increase.  
ins, the higher the  regain of strength, but healing without a loss of strength has been observed at ages up to  three years. The application of pressure across the crack assists in healing.      Relation between Compressive and Tensile Strengths    From the discussion on the strength of compression and tension (both direct and flexure)  test spec
imens it would be expected that the two types of strength are closely  related. This is indeed the case but there is no direct proportionality, the ratio of the  two strengths depending on the general level of strength of the concrete. In other  words, as the compressive strength, fc, increases, the tensile strength, ft, also increases  but at a d
ecreasing rate.  A number of factors affect the relation between the two strengths. The beneficial effect  of crushed coarse aggregate on flexural strength was discussed on page 287, but it seems  that the properties of fine aggregate also influence the ft / fc ratio. The ratio is  furthermore affected by the grading of the aggregate. This is prob
ably due to the different  magnitude of the wall effect in beams and in compression specimens: their surface / volume  ratios are dissimilar so that different quantities of mortar are required for full  compaction.   Age is also a factor in the relation between ft and fc: beyond about one month the  tensile strength increases more slowly than the 
compressive strength so that the ratio  ft / fc decreases with time. This is in agreement with the general tendency of the  ratio to decrease with an increase in fc.   The tensile strength of concrete is more sensitive to inadequate curing than the  compressive strength, possibly because the effects of non-uniform shrinkage of flexure  test beams 
are very serious. Thus air-cured concrete has a lower ft / fc ratio than  concrete cured in water and tested wet.   Air entrainment affects the ft / fc ratio because the presence of air lowers the  compressive strength of concrete more than the tensile strength, particularly in the  case of rich and strong mixes. The influence of incomplete compac
tion is similar to  that of entrained air (see Fig. 5.28).    Fig. 5.28. Relation between compressive and flexural strengths of incompletely  compacted concrete. Compressive strength determined on equivalent cubes;  modulus of rupture determined under third-point loading     Lightweight concrete conforms broadly to the pattern of the relation  bet
ween ft and fc for ordinary concrete. At very low strengths (say, 2 MPa  (300 psi)) the ratio ft / fc can be as high as 0.3, but at higher strengths it  decreases to values comparable with those for ordinary concrete. In the  latter case, the ratio varies generally between 0.16 and 0.07 when the cube  crushing strength is taken as fc, and the modu
lus of rupture under  third-point loading as ft. As shown on page 547, the tensile strength of  concrete depends on the type and method of test, so that the means of  determining ft must be clearly stated.   A number of empirical formulae connecting ft and fc have been suggested, many of them of  the type -    (formula),    where k and n are coeff
rmer value is used by the American Concrete Institute, but  Gardner and Poon found a value nearer the latter, cylinders being used in both  cases.   Comite Europeen du Beton assumes that the mean direct tensile strength is related to  the characteristic compressive strength of cylinders (see p. 656) by the expression    (formula),    the strengths
n view of the numerous factors influencing the ratio of the strengths it is not  surprising that no simple relation is generally applicable. Data obtained at the  laboratories of the Portland Cement Association are given in Table 5.2, and Fig. 5.29  shows the results of Walker and Bloem's tests.    Fig. 5.29. Relation between compressive and flexu
ral strengths of concrete (both air- and  non-air-entrained)    Table 5.2: Relation between Compressive and Tensile Strengths of Concrete      Bond between Concrete and Reinforcement    Since structural concrete is in the vast majority of cases used with steel  reinforcement, the strength of bond between the two materials is of considerable  inter
rgely outside the scope of this book.	In general terms, bond is related to the quality of the concrete, and bond strength is  approximately proportional to the compressive strength up to about 20 MPa (3000 psi).  For higher strengths of concrete, the increase in bond strength becomes progressively  smaller and eventually negligible (see Fig. 5.3
0). This is why most codes of practice  restrict the permissible value of bond in high-strength concrete. For instance, the  British Code of Practice for the Structural Use of Concrete CP110: 1972 gives the values  listed in Table 5.3.    Fig. 5.30. Influence of the strength of concrete on bond determined by pull-out tests    Table 5.3: Maximum Va
lues of Flexural Bond Stress in Concrete according to the  British Code of Practice for the Structural Use of Concrete CP110 : 1972     A rise in temperature reduces the bond strength of concrete: at 200 to 300 C (400 to  570 F) there may be a loss of one-half of the bond strength at room temperature.   Galvanizing and other protective treatment o
f reinforcement may reduce the bond  strength, probably because in treated steel the good bond of a rusty surface is absent.  However, data of the Building Research Establishment show that bond of galvanized  reinforcement is at least as good as that of ordinary steel bars and wires. The  explanation lies in the development, on initial attack of z
te saturated as possible,  until the originally water-filled space in the fresh cement paste has been filled to the  desired extent by the products of hydration of cement. In the case of site concrete,  active curing stops nearly always long before the maximum possible hydration has taken  place. The order of influence of moist curing on strength 
can be gauged from Fig. 5.31,  obtained for concrete with a water / cement ratio of 0.50. The loss of strength due to  inadequate curing is more pronounced in thinner elements but is smaller in lightweight  aggregate concrete. Tensile and compressive strengths are affected in a similar  manner; in both cases, richer mixes are slightly more suscept
ible.    Fig. 5.31. Influence of moist curing on the strength of concrete with a water / cement  ratio of 0.50	The necessity for curing arises from the fact that hydration of cement can take place  only in water-filled capillaries. This is why a loss of water by evaporation from the  capillaries must be prevented. Furthermore, water lost inter
than the water required for  combination. This statement is of considerable importance as it was formerly thought  that, provided a concrete mix contained water in excess of that required for the  chemical reactions with cement, a small loss of water during hardening would not  adversely affect the process of hardening and the gain of strength. It
 that below a  vapour pressure of 0.8 of the saturation pressure the degree of hydration is low, and  negligible below 0.3 of the saturation pressure.    Fig. 5.32. Water taken up by dry cement exposed for six months to different vapour pressures     It must be stressed that for a satisfactory development of strength it is not necessary  for all t
earth, sawdust or straw. Periodically-wetted  hessian or cotton mats may be used, or alternatively an absorbent covering with  access to water may be placed over the concrete. A continuous supply of water is  naturally more efficient than an intermittent one, and Fig. 5.37 compares the strength  development of concrete cylinders whose top surface 
was flooded during the first 24  hours with that of cylinders covered with wet hessian. The difference is greatest at low  water / cement ratios where self-desiccation operates rapidly. The influence of curing  conditions on strength is lower in  the case of air-entrained than non-air-entrained  concrete.    Fig. 5.37. Influence of curing conditio
ns on strength of test cylinders     Another means of curing is to use an impermeable membrane or  waterproof paper. A  membrane, provided it is not punctured or damaged, will effectively prevent evaporation  of water from the concrete but will not allow ingress of water to replenish that lost  by self-desiccation. The  membrane is formed by seali
er rise in the temperature of  the concrete. The effectiveness (as measured by strength) of a white membrane and of  white translucent sheets of polyethylene is the same. Details of the curing  compounds are outside the scope of this book. ASTM Method C 156-74  prescribes tests  for the efficiency of curing compounds and ASTM Standard C 171-69 (re
ents and  exposure conditions in terms of maturity of concrete (see p. 314); however, the  quality of concrete is ignored. The American Concrete Institute Standard 308-71  (reaffirmed in 1978) gives extensive information on curing. Striking times for formwork  are given in a British publication.   High-strength concrete should be cured at an early
f strength is caused by delaying the  curing, as shown for instance by Fig. 5.38 for a 1:3.4 mortar with a water / cement ratio  of 0.70, but in practice early drying may lead to shrinkage and cracking, and delaying  curing is inadvisable.    Fig. 5.38. Relation between total curing time in water and tensile strength of mortar  briquettes      Mat
urity of Concrete    So far we have considered only the time aspect of curing but, as mentioned earlier, the  temperature during curing also controls the rate of progress of the reactions of  hydration and consequently affects the development of strength of concrete. This  influence is shown, for instance, in Fig. 5.39 obtained from tests on speci
mens cast,  sealed and cured at the indicated temperatures. The effect of the temperature at the  time of setting is considered on page 318.    Fig. 5.39. Ratio of strength of concrete cured at different temperatures to the 28-day  strength of concrete cured at 21 C (70 F) (water / cement ratio = 0.50; the specimens  were cast, sealed, and cured a
t the indicated temperature)	Since strength of concrete depends on both age and temperature we can say that strength  is a function of C(time interval x temperature), and this summation is called maturity.  The temperature is reckoned from an origin found experimentally to be between -12 and  -10 C (11 and 14 F). This is because at temperature
s below the freezing point of water  and down to about - 12 C (11 F) concrete shows a small increase in strength with time  but the low temperature must not be applied, of course, until after the concrete has set  and gained sufficient strength to resist damage due to the action of frost; a "waiting  period" of 24 hours is usually required. Below 
-12 C (11 F) concrete does not appear  to gain strength with time.   Maturity is measured in C-hours ( F-hours) or C-days (F-days). Fig. 5.40 and Fig.  5.41 show that compressive and tensile strengths plotted against the logarithm of  maturity give a straight line. It is, therefore, possible to express strength at any  maturity as a percentage of 
strength of concrete at any other maturity; the latter is  often taken as 19 800 Ch (35 600 Fh), being the maturity of concrete cured at 18 C (64 F)  for 28 days. The ratio of strengths, expressed as a percentage, can then be written  as -    (formula).    Fig. 5.40. Relation between logarithm of maturity and compressive strength of cubes    Fig. 
5.41. Relation between logarithm of maturity and splitting strength (tests carried  out at 2, 13, and 23 C (35, 55, and 73 F) up to 42 days)     Strictly speaking, the linearity applies only above a certain minimum maturity, as  shown in Fig. 5.42.    Fig. 5.42. Relation between compressive strength of ordinary Portland (Type I)  cement concrete a
nd maturity for the data of Gruenwald as treated by Lew and Reichard     The values of the coefficients A and B depend on the strength level of concrete; those  suggested by Plowman are given in Table 5.4. It can be seen that the strength - maturity  relation depends on the properties of the cement and on the general quality of the  concrete, and 
of a period of exposure to a higher temperature are not the same  when this occurs immediately after casting or later in the life of the concrete.  Specifically, early high temperature leads to a lower strength for a given total  maturity than when heating is delayed for at least a week or is absent. This is of  interest in connection with steam c
uring.    Table 5.4: Plowman's Coefficients for the Maturity Equation    Fig. 5.43. Influence of the temperature during the first 28 days after casting on the  strength - maturity relation     Nevertheless, the maturity rule applies fairly well when the initial temperature of  concrete is between 16 and 27 C (60 and 80 F) (although probably a much
ity  rule in the determination of the potential strength of concrete from early  tests.   An excellent review of the maturity concept has been prepared by Malhotra.      Influence of Temperature on Strength of Concrete    We have seen that a rise in the curing temperature speeds up the chemical reactions of  hydration and thus affects beneficially
 the early strength of concrete without any ill-effects  on the later strength. However, a higher temperature during placing and setting,  although it increases the very early strength, may adversely affect the strength from  about 7 days onwards. The explanation is that a rapid initial hydration appears to form  products of a poorer physical stru
cture, probably more porous so that a large proportion  of the pores will always remain unfilled. It follows from the gel / space ratio rule that  this will lead to a lower strength compared with a less porous, though slowly hydrating,  paste in which a high gel / space ratio will eventually be reached. This explanation of  the adverse effects of 
a high early temperature on later strength has been extended by  Verbeck and Helmuth who suggest that the rapid initial rate of hydration at higher  temperatures retards the subsequent hydration and produces a non-uniform distribution of  the  products of hydration within the paste. The reason for this is that at the high  initial rate of hydratio
he subsequent hydration and adversely affects the long-term  strength.   In addition, the non-uniform distribution of the products of hydration per se adversely  affects the strength because the gel / space ratio in the interstices is lower than would  be otherwise the case for an equal degree of hydration: the local weaker areas lower the  streng
th of the paste as a whole .   Fig. 5.44 shows Price's data on the effect of the temperature during the first two  hours after mixing on the developing of strength of concrete with a water / cement ratio  of 0.53. The range of temperatures investigated was 4 to 46 C (40 to 115 F), and  beyond the age of two hours all specimens were cured at 21 C (
asting on the  development of strength (all specimens sealed and after the age of 2 hours cured at  21 C (70 F))     Some field tests have confirmed the influence of temperature at the time of placing on  strength: typically, for an increase of 5 C (9 F) there is a decrease in strength of  1 9 MPa (270 psi).   The influence of curing temperature o
n strength of concrete tested after cooling at 1  and 28 days is shown in Fig. 5.45. However, the temperature at the time of testing  also appears to be a factor, at least in the case of neat (ordinary Portland) cement  paste compacts with a water / cement ratio of 0.14. The temperature was kept constant  from the initiation of hydration. When tes
ted (at 64 and 128 days) at the curing  temperature, the specimens had a lower strength at higher temperatures (Fig. 5.46); but,  if cooled to 20 C (68 F) over a period of 2 hours prior to testing, only temperatures  above 65 C (150 F) had a deleterious effect (Fig. 5.47).    Fig. 5.45. Influence of curing temperature on compressive strength at 1 
and 28 days  (specimens tested after cooling to 23 C (73 F) over a period of two hours)    Fig. 5.46. Relation between compressive strength and curing time of neat cement  paste compacts at different curing temperatures. The temperature of the specimens  was kept constant up to and including the period of testing    Fig. 5.47. Relation between com
pressive strength and curing time of neat cement  paste compacts at different curing temperatures. The temperature of the specimens  was moderated to 20 C at a constant rate over a two-hour period prior to testing  (water / cement ratio = 0.14; Type I cement)     Tests have also been made on concretes stored in water at different temperatures for 
a  period of 28 days, and thereafter at 23 C (73 F). As in Price's tests, a higher  temperature was found to result in a higher strength during the first few days after  casting, but beyond the age of one to four weeks the situation changed radically (see  Fig. 5.43). The specimens cured at temperatures between 4 and 23 C (40 and 73 F) up to  the 
age of 28 days all showed a higher strength than those cured at 32 to 49 C (90 to  120 F). Among the latter, retrogression was greater the higher the temperature, but in  the lower range of temperatures there appeared to be an optimum temperature that yielded  the highest strength. It is interesting to note that even concrete cast at 4 C (40 F)  a
 entrained air. Similar behaviour has been observed when rapid hardening Portland  and modified cement are used. When calcium chloride is added to the mix the adverse  effects of a high temperature during setting are attenuated.    Fig. 5.48. Effect of temperature during the first 28 days on the strength of concrete  (water / cement ratio = 0.41; 
air content = 4.5 per cent; ordinary Portland cement)     The increase in strength caused by the addition of calcium chloride depends on the  temperature of the concrete and is proportionately greater at lower temperatures. For  instance, at 13 C (55 F) the addition of 2 per cent of CaCl2 increases the one-day  strength by about 140 per cent, but 
 case, its influence on the corrosion of  reinforcing steel should not be forgotten (see p. 105).   Klieger's tests indicate that there is an optimum temperature during the early life  of concrete that will lead to the highest strength at a desired age. For laboratory-made  concrete, using ordinary or modified Portland cement, the optimum temperat
ure is  approximately 13 C (55 F); for rapid hardening Portland cement it is about 4 C (40 F).  It must not be forgotten, however, that beyond the initial period of setting and  hardening the influence of temperature (within limits) accords with the maturity rule: a  higher temperature accelerates the development of strength.   The tests described
ed in the desert, strength decreases with an increase in temperature down to a  critical value at about 30 C (86 F), but between 30 C and 45 C (86 F and 113 F)  there may be a slight recovery or no further loss. This behaviour has been observed  using concrete without entrained air and stored at a relative humidity of between 20 and  70 per cent. 
s, however, concrete cast in summer can be expected to have a lower  strength than a similar mix cast in winter. Indeed, on many construction sites the  strength of test specimens has been found to be lower during hot weather even though  from the time of stripping the moulds at the age of 24 hours the specimens were cured in  water at 18 C (64 F)
. Similarly, in tropical countries an apparently lower strength of  concrete has been observed (Fig. 5.49).    Fig. 5.49. Comparison of strengths of concrete with a water / cement ratio of 0.40 cast in  Lagos, Nigeria and in England      Steam Curing at Atmospheric Pressure    Since an increase in the curing temperature of concrete increases its r
ate of  development of strength, the gain of strength can be speeded up by curing concrete in  steam. When steam is at atmospheric pressure, i.e. the temperature is below 100 C (212 F),  the process can be regarded as a special case of moist curing. High-pressure steam curing  (autoclaving) is an entirely different operation and is considered in t
he next section. Steam  curing has been used successfully with different types of Portland cement, but must  never be used with high-alumina cement because of the adverse effect of hot-wet  conditions on the strength of that cement. Concrete with a lower water / cement ratio  responds to steam curing much better than a weaker mix.   The primary ob
ject of steam curing is to obtain a sufficiently high early  strength so that the concrete products may be handled soon after casting: the moulds can  be removed, or the prestressing bed vacated, earlier than would be the case with ordinary moist curing,  and less curing storage space is required; all these mean an economic advantage. For many app
lications,  the long-term strength of concrete is of lesser importance.   Because of the nature of the operations involved in steam curing, the process is used mainly with  precast products. Low-pressure steam curing is normally applied in special chambers or in tunnels  through which the concrete members are transported on a conveyor belt. Altern
atively, portable boxes  or plastic covers can be placed over precast members, steam being supplied through flexible  connections.   Owing to the influence of temperature during the early stages of hardening on the later  strength, a compromise between the temperatures giving a high early and a high late  strength has to be made. Fig. 5.50 shows t
ypical values of strength of concrete made  with modified cement and a water / cement ratio of 0 55; steam curing was applied  immediately after casting .    Fig. 5.50. Strength of concrete cured in steam at different temperatures (water / cement  ratio = 0.50; steam curing applied immediately after casting)     One factor in the retrogression of 
strength of steam-cured concrete is the  pressure developed in the air pores during heating, since air has a higher coefficient  of thermal expansion than the surrounding solid material. Some of this effect can be  avoided by heating in closed formwork or in pressure chambers. With short-term curing  periods (2 to 5 hours) and moderate temperature
s, there is probably little real  retrogression of strength, and the apparent low strength at later ages is due to the  absence of prolonged wet curing.   A related problem is the rate of increase in temperature at the commencement of steam curing. It has  been found that if 49 C (120 F) is reached in a period shorter than about 2 to 3 hours, or 9
9 C (210 F) in  less than 6 to 7 hours from the time of mixing, the gain of strength beyond  the first few hours is affected adversely; such a rapid rise in temperature must not  be permitted. The values quoted are merely a guide, but excessively rapid heating can  cause a loss of strength at later ages of as much as one-third compared with wet cu
ring  at room temperature. The adverse effect of the rapid rise is more pronounced the higher  the  water / cement ratio of the mix, and is also more noticeable with rapid  hardening than with ordinary Portland cement. Saul found that, when the rate of rise  in the temperature of concrete does not exceed the values mentioned earlier, its  strength
 will differ only little from the strength of normally cured concrete, and will  fall within zone A of Fig. 5.51. By contrast, the strength of concrete heated too  rapidly will lie within zone B of the same Figure.    Fig. 5.51 Gain of strength of steam-cured concrete with time (water / cement ratio =  0.50, rapid-hardening Portland cement)     Si
nce it is the temperature at the time of setting that has the greatest influence on  the strength at later ages, a delay in the application of steam curing is advantageous.  Some indication of the influence of the delay in heating on strength can be obtained  from Fig. 5.52 plotted by Saul from the data of Shideler and Chamberlin. The  concrete us
ed was made with modified cement, and had a water / cement ratio of 0.6. The  solid line shows the gain in strength of moist-cured concrete at room temperature  plotted against maturity. The dotted lines refer to different curing temperatures  between 38 and 85 C (100 and 185 F), and the figure against each point denotes the  delay in hours before
 the higher curing temperature was suddenly applied.    Fig. 5.52. Effect of delay in steam curing on the early gain of strength with maturity     From Fig. 5.52 it can be seen that for each curing temperature there is a  part of the curve showing a normal rate of gain in strength with maturity. In other  words, after a sufficient delay, rapid hea
ting has no adverse effect. This delay is  approximately 2, 3, 5 and 6 hours respectively for 38, 54, 74 and 85 C (100, 130, 165  and 185 F). If, however, concrete is exposed to the higher temperature with a smaller  delay, the strength is adversely affected, as shown by the right-hand portion of each  dotted curve; this effect is more serious the
 higher the curing temperature.   Fig.	5.52 shows also that within a few hours of casting the rate of gain in strength is  higher than would be expected from the maturity calculations. This confirms the earlier  observation that the age at which a higher temperature is applied is a factor in the  maturity rule.   Practical curing cycles are chosen
 as a compromise between the early and late strength  requirements but are governed also by the time available (e.g. length of work shifts).  Economic considerations determine whether the curing cycle should be suited to a given  concrete mix or alternatively whether the mix ought to be chosen so as to fit a  convenient cycle of steam curing. Whil
small specimens. On the other hand, with curing at ordinary  temperatures the effects of the heat of hydration are significant only in mass  structures.   It may be worth emphasizing that a longer period of curing at a lower temperature leads  to a higher optimum strength than when high temperature is applied for a shorter time.  For any one perio
d of curing, there is a temperature which leads to an optimum strength.  Also, for a given set of materials, it is possible to draw a line between the optimum  strength at various curing periods and the curing temperature; this is shown in Fig.  5.54.    Fig. 5.54. Strength development of concrete at different curing temperatures for  various peri
t employed in the manufacture of sand-lime brick,  and is still extensively used for that purpose. In the field of concrete, high-pressure  steam curing is usually applied to precast products  (made both of ordinary and lightweight concrete) when any of the following  characteristics are desired:   (a) high early strength: with high-pressure steam
 curing, the 28-day strength on normal  curing can be reached in about 24 hours;   (b) high durability: high-pressure steam curing improves the resistance of concrete to  sulphates and to other forms of chemical attack, also to freezing and thawing, and  reduces efflorescence; and   (c) reduced drying shrinkage and moisture movement.   The optimum
of C3S (see Fig.	5.55). Cements rich in C3S have a greater capacity for developing high  strength when cured at high pressure than those with a high C2S content, although for short periods of  high-pressure steam curing cements with a moderately low C3S / C2S ratio give good results.    Fig. 5.55. Influence of pulverized silica content on the stre
 is no retrogression  of strength. At the age of one year the strength of normally cured concrete is  approximately the same as that of high-pressure steam-cured concrete of similar mix  proportions. The water / cement ratio affects the strength of high-pressure steam-cured  concrete in the usual manner, but the actual values of early strength dif
ica reaction.  Further improvement in sulphate resistance is due to the increased strength and  impermeability of the steam-cured concrete, and also to the existence of hydrates in a  well-crystallized form.   High-pressure steam curing reduces efflorescence as there is no lime left to be leached  out.   On the debit side, high-pressure steam curi
ng reduces the strength in bond with  reinforcement by about one-half compared with ordinary curing so that the application of  high-pressure steam to reinforced concrete members is considered inadvisable.  High-pressure steamed concrete tends also to be rather brittle. On the whole, high-pressure  steam curing produces good quality, dense and dur
uld be applied to concretes made with Portland cement only: high-alumina and  supersulphated cement would be adversely affected by the high temperature.   Within the Portland group, the type of cement affects the strength but not necessarily  in the same way as at normal temperatures; no systematic studies have, however, been  made. It is known, t
hough, that granulated slag may cause trouble if it has a high  sulphur content. High-pressure steam curing accelerates the hardening of concrete  containing calcium chloride, but the relative increase in strength is less than when no  calcium chloride is used.      Variation in Strength of Cement    Up to now we have not considered the strength o
f cement as a variable in the strength of  concrete. By this we do not mean the differences in the strength-producing properties of  cements of different types, but the variation between cements of nominally the same  type: they vary fairly widely, and it is this variation that is considered in this  section.   The scatter of strength of ordinary 
and rapid hardening Portland cements is illustrated  in Fig. 5.56, which shows also that there is a considerable overlap in the strengths of  the two cements. However, these cements nearly always comply with the minimum  requirements of ASTM Standard C 15-78a or of BS 12: 1978, the strength of some of the  cements sometimes reaching values as high
 as twice the prescribed minimum.    Fig. 5.56. Histograms for vibrated mortar cubes made with ordinary and rapid hardening  Portland cements as defined in BS 12: 1958     Although on a site it is difficult to isolate the influence of cement, there is no doubt  that the inherent variation in the strength of cement is reflected in the variation in 
 the strength of concrete. Thus the variability of cement is of considerable practical  importance, particularly since a concrete mix has to be designed so as to give a  satisfactory strength even when a low-strength batch of cement is encountered: no  advantage can be taken of the much higher mean strength. If cement had a lower  variability (but
 the same mean strength) the cement content of the mix could be reduced:  there may thus be some advantage in strength grading of cement even if this were  accompanied by a higher price of the material.   The variation in strength of cement is due largely to the lack of uniformity in the raw  materials used in its manufacture, not only between dif
ferent sources of supply, but  also within a pit or a quarry. Furthermore, differences in details of the processes of  manufacture and above all the variation in the ash content of coal used to fire the kiln  contribute to the variation in the properties of commercial cements.   The magnitude of the variation in the strength of cement can be judge
d from Fig. 5.57,  which shows the results of tests on ASTM standard mortar made with samples of cement  obtained from the same plant in the United States at two-week intervals. The strength is  expressed as a percentage of the mean strength of all the samples from the given works,  and each curve is an average of strength ratios obtained at 3, 7 
and 28 days. The variation in strength  due to testing per se is indicated by the dotted lines which show the strength ratios for tests made at the  same time using a control "stock" cement. The testing error accounts generally  of between 2 1/2 and 4 per cent. ASTM Standard C 917-80 gives a method of assessing the  uniformity of cement from a sin
gle source.    Fig. 5.57 Variation in strength of cement from the same plant     Studies of cement plants in California have shown that the coefficients of variation  of standard mortar cubes made with cement from any one given plant are as follows:-   20 per cent of plants have a coefficient of variation smaller than 6 per cent,   72 per cent - s
lant, the  ordinary Portland cement (minimum 28-day strength of 35 MPa (5000 psi))  had an average 28-day strength of 49.5 MPa (7200 psi) with a standard deviation of 22  MPa (320 psi). This value of scatter includes the testing error, which may well account  for some 1.5 MPa (220 psi). It should be noted that German standards lay down a maximum  
strength of cement, this being 20 MPa (3000 psi) above the specified minimum. An appendix to  ASTM Standard C 917-80 gives data on the typical variability of cement made in U.S. and Canadian  plants.   The standard deviation of the strength of concrete due to the variation in  cement does not increase with age; indeed, German data suggest a  decre
ase of up to 1 MPa (150 psi). Since strength increases with age, the  coefficient of variation of strength becomes smaller the older the concrete at  the time of testing. This behaviour is not surprising because a large part  of the variation in the strength of cement is due to the differences in fineness and in  the C3S content: the effects of th
ese factors are greatest at early ages and with time  cease to be significant. By contrast, the standard deviation within batches increases  with an increase in mean strength. Table 5.5 gives Wright's data on the variability  of cement at different ages at test, and Table 5.6 gives approximate values of the  standard deviation of strength of site-
erent batches of cement from one plant, leads to an  appreciable variation in the strength of concrete. This effect may be of considerable  importance on a large job: the use of cement all from one batch can result in a decrease  in cement content of up to 10 per cent. It must not be forgotten, however, that  variation in cement accounts, at the m
ost, for one-half of the variation in the strength  of site test specimens; U.S. Bureau of Reclamation data suggest a typical value of  one-third. A major part of this variation can be eliminated by the use of cement from  one silo at the plant. The variation in the strength of site cubes is discussed on page  653.   Finally it should be stressed 
that the variation in cement affects to the greatest  extent the early strength of concrete, i.e. the strength most often  determined by test but not necessarily the strength of greatest practical significance.  Furthermore, strength is not the only important characteristic of concrete: from  considerations of durability and impermeability, a ceme
nt content in excess of that  needed for strength may well be required, in which case the variability of cement  becomes unimportant.      Fatigue Strength of Concrete    In this chapter, we have considered so far only the strength of concrete under static  loading. In many structures, however, repeated loading is applied, and when a material  fai
ls under a number of repeated loads, each smaller than the static compressive  strength, failure in fatigue is said to take place. Both concrete and steel possess the  characteristics of fatigue failure but in this book the behaviour of concrete alone is  dealt with.   Let us consider a concrete specimen subjected to alternations of compressive st
atigue limit improves the fatigue strength of  concrete, i.e. concrete loaded a number of times below its fatigue limit  will, when subsequently loaded above the limit, exhibit a higher fatigue strength than  concrete which had never been subjected to the initial cycles. The former concrete also  exhibits a higher static strength by some 5 to 25 p
er cent. It is possible that this  increase in strength is due to a densification of concrete caused by the initial  low-stress level cycling, in a manner similar to improvement in strength under moderate  sustained loading. This property is akin to strain hardening in metals, and is of  particular interest since concrete under static loading is a
 strain-softening rather  than strain-hardening material.   Strictly speaking, concrete does not appear to have a fatigue limit, i.e. a fatigue  strength at an infinite number of cycles (except when stress reversal  takes place). It is therefore to refer to fatigue strength at a very large  number of cycles, such as 10 million.   The fatigue stren
given concrete can withstand a specified number of cycles can  be read off the diagram. For a given ol, the number of cycles is very sensitive to the  range of stress. For instance, an increase in range from 57.5 to 65 per cent of the  ultimate static strength has been found to decrease the number of cycles by a factor of  40.    Fig. 5.63. Modifi
oncrete member  which is to carry a transient load of a certain magnitude.   From the fact that the lines of Fig. 5.63 rise to the right it can also be seen that  the fatigue strength of concrete is lower the higher the ratio oh / ol. Some tests have  shown that lateral pressure increases the fatigue life of concrete, but not at very high stresses
.   The frequency of the alternating load, at least within the limits of 70 to 2000 cycles  per minute, does not affect the resulting fatigue strength, higher frequency is of  little practical significance. This applies both in compression and in flexure-tension,  the similarity between fatigue behaviour in the two types of loading, as well as in 
 splitting tension, suggesting that the failure mechanism is the same. In fact, the  fatigue behaviour in flexure parallels closely that in compression (Fig. 5.64). The  fatigue strength for 10 million cycles is 55 per cent of the static strength. By  comparison, in compression fatigue, the fatigue strength in flexure is between 60 and 64  per cen
t after the same number of cycles. Sufficiently accurate test results are not  available to state with certainty that these two limits (in compression and in flexure)  differ significantly from one another. Tests have shown that the moisture condition of  concrete prior to loading affects its fatigue strength in flexure: oven-dried specimens  show
 the highest strength and partially dried ones the lowest; wet specimens are in  between (Fig. 5.65). The explanation of this behaviour lies in differential strains  induced by the moisture gradients.    Fig. 5.64. Modified Goodman diagram for concrete in flexure fatigue    Fig. 5.65. Effect of moisture condition on fatigue performance of concrete
 specimens     As strength increases with age, fatigue strength both in compression and in flexure  also increases. The important point is that at a given number of cycles, fatigue  failure occurs at the same fraction of ultimate strength, and is thus independent of the  magnitude of this strength (both in compression and in splitting tension) and
oken  aggregate particles than specimens which failed in a static test. Thus failure at the  bond interface is probably dominant in fatigue; in mortar, fatigue failure is believed  to take place at the interface of the fine aggregate particles. It is likely that a  smaller maximum size of aggregate leads to a higher fatigue strength, probably  bec
ause of greater homogeneity of concrete.   Air-entrained concrete and lightweight aggregate concrete have the same fatigue  behaviour as concrete made with ordinary aggregate. Fatigue in concrete  cylinders occurs in the same way as in large specimens subjected to low-frequency  loading.   The fatigue strength of concrete is increased by rest peri
ods (this does not apply when  there are stress reversals), the increase being proportional to their duration between 1  and 5 minutes; beyond the 5-minute limit there is no further increase in strength. With  the rest periods at their maximum effective duration, their frequency determines the  beneficial effect. The increase in strength caused by
der fatigue  loading is not closely related to behaviour under impact loading, although of course  both may occur simultaneously in practice. Failure under fatigue loading occurs at a  strength below the static compressive strength while under impact the static strength is  not impaired.   While this book is not concerned with the fatigue behaviou
r of reinforced and  prestressed concrete, we should note that fatigue cracks in concrete act as  stress-raisers, thus magnifying the vulnerability of the steel to fatigue failure (if the  stress in it is in excess of its critical fatigue stress value).      Impact Strength    Impact strength is of importance primarily in connection with pile driv
ing and with  foundations for machines exerting impulsive loading, but accidental impact (e.g. during  handling of precast members) is also of interest. Extensive tests on impact strength of  concrete were made by Green. As principal criteria he considered the ability of a  specimen to withstand repeated blows and to absorb energy. In particular, 
he studied the  number of blows which the concrete can withstand before reaching the no-rebound  condition, this stage indicating a definite state of damage.   Impact tests, when conducted with a relatively small hammer (25 mm (1 in.) diameter  face) lead to a greater scatter of results than tests on static compressive strength of  the concrete. T
his arises from the fact that in the standard compression test some  relief of a highly stressed weak zone is possible owing to creep, while in the impact  test no redistribution of stresses is possible during the very short period of  deformation. Hence, local weaknesses have a greater influence on the recorded strength  of a specimen .   In gene
ral, Green found that the higher the static compressive strength of the concrete  the lower the energy absorbed per blow before cracking, but the impact strength of  concrete increases with its compressive strength (and therefore age) at a progressively  increasing rate (Fig. 5.66). The  relation is different for each coarse aggregate and  storage
 condition of the concrete. For the same compressive strength, the impact  strength is greater  for coarse aggregate of greater angularity and surface roughness.  This observation was confirmed by Dahms and supports the suggestion that  impact strength is more closely related to the tensile strength of concrete than to its  compressive strength. T
hus concrete made with a  gravel coarse aggregate has a low  impact strength, failure taking place owing to insufficient bond between mortar and  coarse aggregate. On the other hand, when the surface of the aggregate is rough, the  concrete is able to develop the full strength of much of the aggregate in the region of  failure.    Fig. 5.66. Relat
ion between compressive strength and number of blows to  "no-rebound" for concretes made with different aggregates and Type I cement,  stored in water     A smaller maximum size of aggregate significantly improves impact resistance; so does  the use of aggregate with a low modulus of elasticity and a low Poisson's ratio.  Cement content below 400 
kg / m3 (670 lb / yd3) is  advantageous. The influence of fine  aggregate is not well defined but the use of fine sand usually leads to a slightly  lower impact strength. Dahms found a high content of sand advantageous. We could  try to generalize and say that a mix of materials which have a limited variation in  properties is conducive to a good 
impact strength. Extensive tests on the impact  strength of concretes of different properties were made by Hughes and Gregory.   Storage conditions influence the impact strength in a manner different  from  compressive strength. Specifically, the impact strength of water stored concrete is  lower than when the concrete is dry, although the former 
concrete can withstand more  blows before cracking. Thus, as already stated, the compressive strength without  reference to storage conditions, does not give a satisfactory indication of the impact  strength.   There is evidence that under uniformly applied impact loading (a condition difficult  to achieve in practice) the impact strength of concr
ete is significantly greater than  its static compressive strength. This increase in  strength would explain the greater  ability of concrete to absorb strain energy under uniform impact. Fig. 5.67 shows that  strength increases greatly when the rate of application of stress exceeds about 500  GPa / s,  reaching at 4.9 TPa / s more than double the
 value at normal speeds of loading  (about 0 5 MPa / s).    Fig. 5.67. Relation between compressive strength and rate of loading up to impact level     A parallel relation is that between strength and the rate of application of strain  shown in Fig. 5.68. In these tests, the rate of application of stress was lower  than in those shown in Fig. 5.67
;	for instance, the rate of strain of 14 per sec  corresponds to a rate of stress of 400 MPa / s (58 000 psi / s). The tests of Fig. 5.68 give  a lower increase in impact strength than Popp's tests, probably because in the  former the platen friction was greatly reduced by the use of pads    Fig. 5.68. Relation between compressive strength and rat
e	of strain in impact tests for  the data of reference      Quality of Mixing Water    The vital influence of the quantity of water in the mix on the strength of the resulting  concrete has been repeatedly mentioned. The quality of the water also plays its role:  impurities in water may interfere with the setting of the cement, may adversely affec
t	 the strength of the concrete or cause staining of its surface, and may also lead to  corrosion of the reinforcement. For these reasons, the suitability of water for mixing  and curing purposes should be considered. Clear distinction must be made between the  effects of mixing water and the attack on hardened concrete by aggressive waters. Some 
substances are present.  A simple way of determining the suitability of such water is to compare the setting  time of cement and the strength of mortar cubes using the water in question with the  corresponding results obtained using known "good" water or distilled water; there is no  appreciable difference between the behaviour of distilled and or
dinary drinking water.  A tolerance of about 10 per cent is usually  permitted to allow for the chance  variations in strength; an appendix to BS 3148:1980 also suggests 10 per cent. Such  tests are recommended when water for which no service record is available contains  dissolved solids in  excess of 2000 ppm or, in the case of alkali carbonate 
ommended in American literature.   Sea water has a total salinity of about 3.5 per cent (78 per cent of the dissolved  solids being NaCl and 15 per cent MgCl2 and MgSO4), and produces a slightly higher  early strength but a lower long-term strength; the loss of strength is usually no more than  15  per cent and can therefore often be tolerated. So
me tests suggest that sea water  slightly accelerates the setting time of cement, others, a substantial reduction in  the initial setting time but not necessarily in the final set. Generally, the effects on  setting are unimportant if water is acceptable from strength considerations. An appendix  to BS 3148: 1980 suggests a tolerance of 30 minutes
er  organic acids may adversely affect the hardening of concrete; such water, as well as  highly alkaline water, should be tested. The effects of different ions vary, as shown by  Steinour.   It may be interesting to note that the presence of algae in mixing water results in air  entrainment with a consequent loss of strength. The use of algae is,
 of no significance, and any water suitable for mixing,  or even slightly inferior in quality, is acceptable for curing.   It is essential that curing water be free from substances that attack hardened  concrete; these are discussed in Chapter 7.  4 Fresh Concrete    The strength of concrete of given mix proportions is very seriously affected  by 
lting strength. It is convenient to express  the former as a density ratio, i.e. a ratio of the actual density of the given concrete to the  density of the same mix if fully compacted. Likewise, the ratio of the strength of the  concrete as actually (partially) compacted to the strength of the same mix when fully  compacted can be called the stren
gth ratio. Then the relation between the strength  ratio and the density ratio is of the form shown in Fig. 4.1. The presence of voids in  concrete greatly reduces its strength: 5 per cent of voids can lower strength by as much as  30 per cent and even 2 per cent voids can result in a drop of strength of more than 10 per  cent. This, of course, is
 in agreement with Feret's expression relating strength to the sum  of the volumes of water and air in the hardened paste (see p. 268).    Fig. 4.1. Relation between strength ratio and density ratio     Voids in concrete are in fact either bubbles of entrapped air or spaces left after  excess water has been removed. The volume of the latter depend
used for the latter, cement and water are formed into colloidal grout by passage,  at a speed of 2000 rev / min, through a narrow gap, and sand is subsequently added to the grout.  This type of mixer may become of interest in concrete-making because the pre-mixing of cement  and water allows better hydration and may lead to a higher strength at a 
given water / cement  ratio than conventional mixing. For instance, at water / cement ratios of 0.45 to 0.50, a gain in  strength of 10 per cent has been observed. However, two-stage mixing undoubtedly represents  a higher cost and is likely to be justifiable only in special cases.    Uniformity of Mixing    In any mixer, it is essential that suff
 retained on a 4.75 mm  (3/16 in.) sieve   6 per cent  density of air-free mortar  1.6 per cent  compressive strength (average  7-day strength of 3 cylinders) 7.5 per cent.    The U.S. Bureau of Reclamation lays down similar requirements.   In the United Kingdom, BS 3963: 1974 lays down a test of performance  of mixers using a specified concrete m
ortant to know what is the minimum mixing time  necessary to produce a concrete uniform in composition and, as a result, of  satisfactory strength. This time varies with the type of mixer, and, strictly  speaking, it is not the mixing time but the number of revolutions of the  mixer that is the criterion of adequate mixing. Generally, about 20  re
ts, the variability being represented as the range of strengths  of specimens made from the given mix after a specified mixing time. Fig.  4.16 shows the results of the same tests plotted as a coefficient of variation  against mixing time. It is apparent that mixing for less than 1 to 1 1/4 min  produces an appreciably more variable concrete, but 
prolonging the mixing  time beyond these values results in no significant improvement in uniformity.    Fig. 4.15. Relation between compressive strength and mixing time    Fig. 4.16. Relation between coefficient of variation of strength and mixing time     The average strength of concrete also increases with an increase in  mixing time, as shown f
or instance by Abrams' tests (Fig. 4.17). The  rate of increase falls rapidly beyond about one minute and is not significant  beyond two minutes; sometimes, even a slight decrease in strength has  been observed. Within the first minute, however, the influence of  mixing time on strength is of considerable importance. For instance,  Shalon calculat
ed that, for a given required strength, increasing the  mixing time from 30 sec to 1 min permits a saving in the cement content of  as much as 30 kilograms per cubic metre (50 lb /yd3).    Fig. 4.17. Effect of mixing time on strength of concrete     As mentioned before, the exact value of the minimum mixing time varies  with the type of mixer and 
, then the water, and  finally the remainder of the coarse aggregate so as to break up any nodules  of mortar.   Let us consider now the other extreme - mixing over a long period.  Generally, evaporation of water from the mix takes place, with a consequent  decrease in workability and increase in strength. A secondary effect  is that of grinding o
hile a delay in placing without continuous mixing causes a drop in  air content by only about 1/10 per hour. On the other hand, a decrease in  mixing time below 2 or 3 minutes may lead to inadequate entrainment of  air.   Intermittent remixing up to about 3 hours, and in some cases up to 6  hours, is harmless as far as strength and durability are 
concerned, but the  workability falls off with time unless loss of moisture from the mixer is  prevented. Adding water to restore workability, known as re-tempering,  will lower the strength of the concrete.   Some investigators have reported that this loss of strength is  smaller than would be expected from the consideration of the total  water /
hen the added  water forms part of the effective water which governs the strength (see Fig.  4.19).   Re-tempering has been reported slightly to increase the shrinkage but  this probably occurs only if the effective water / cement ratio is increase by  the added water.    Hand Mixing    There may be occasions when concrete has to be mixed by hand 
uring agitating at 4 rev / min     Many specifications impose the same limit on the time of haul of  central-mixed concrete, and it is usual also to limit the total number of  revolutions during both mixing and agitating to approximately 300.  However, agitating up to 6 hours need not adversely affect the strength of  concrete provided the mix rem
	reason, concrete is sometimes re-tempered (see page 233) by the addition  of water immediately before discharge; the workability is thus restored but  it must be realized that the resultant compressive strength will be affected  by the amount of water added to the mix (see Fig. 4.19).    Fig. 4.19. Effect of re-tempering water on the strength of
e	near the  discharge end facilitates placing but increases the friction loss. Aluminium  pipes must not be used because aluminium reacts with the alkalis in cement  and generates hydrogen. This gas introduces voids in the hardened  concrete with a consequent loss of strength, unless the concrete is placed in  a confined space.   The main advantag
cond type of blockage can occur. If the fines  content is too high, the friction resistance of the mix can be so large that  the pressure exerted by the piston through the water phase is not sufficient  to move the mass of concrete, which becomes stuck. This type of failure is  more common in high strength mixes or in mixes containing a high  prop
ortion of very fine material such as crusher dust or fly ash, while the  segregation failure is more apt to occur in medium or low strength mixes  with irregular or gap grading .  The optimum situation therefore is to produce maximum frictional  resistance within the mix with minimum void sizes, and minimum frictional  resistance against the pipe 
own to 0.60  when pressure may also be required). In fact, extremely dry and stiff mixes can be vibrated  satisfactorily so that for a given desired strength concrete can be made with a lower cement  content. This means a saving in cost, but against that we have to offset the cost of the  vibrating equipment, and of heavier and more sturdy formwor
 in an increase in the 28-day compressive  strength of the form shown in Fig. 4.24. The comparison is on the basis of the same  total period of vibration, applied either immediately after placing or in part then and  in part at a specified time later. An increase in strength of approximately 14 per cent has  been reported, but actual values would 
depend on the workability of the mix and on details  of the procedure: other workers have found increases of 3 to 9 per cent. In general the  improvement in strength is more pronounced at earlier ages, and is greatest in concretes liable  to high bleeding since the trapped water is expelled on revibration. For the same reason,  revibration greatly
 improves bond between concrete and reinforcement. It is possible  also that some of the improvement in strength is due to a relief of the plastic shrinkage  stresses around aggregate particles.    Fig. 4.24. Relation between 28-day compressive strength and the time of revibration     Despite these advantages revibration is not widely used as it i
 rate of  evaporation from the fresh mix. These problems concern the mixing, placing and curing  of the concrete.   A higher temperature of fresh concrete results in a more rapid hydration and leads  therefore to accelerated setting and to a lower strength of hardened concrete since a less uniform  framework of gel is established (see p. 318). Fur
 by using larger  quantities of the entraining agent. A related problem is that, if relatively cool  concrete is allowed to expand when placed at a higher temperature,  the air voids expand and the strength is reduced. This would occur, for  instance, with horizontal panels but not with vertical ones in steel moulds where  expansion is prevented. 
Curing also presents additional problems  as the curing water tends to evaporate rapidly. The use of curing compounds is not  entirely satisfactory since it leads to lower compressive strengths than when continuous  water curing is applied; some experimental values are given in Table 4.5. What must be  remembered above all, however, is the overrid
be seen that because of its relatively small quantity  in the mix the temperature of the cement is not important.   The use of hot cement per se is not detrimental to strength but it is preferable not to use  cement at temperatures above about 75 C (170 F). This statement is of interest since hot  cement is sometimes viewed with suspicion and vari
ous ill effects have at times been  ascribed to its use. However, if hot cement is dampened by a small amount of water before  it is well dispersed with other solids it may set quickly and form cement balls.   The influence of the temperature during setting on the strength at later ages is discussed  on p. 318, here it suffices to say that a tempe
different cements     Mather has suggested that a Portland cement content as low as 36 kg / m3 (60 lb / yd3)  is possible, but the content of pozzolanas would be up to twice as high. Such a mix might  have a water content down to 48 kg / m3 (80 lb / yd3). The slump required is about 40  mm (1 1/2 in.) and the resulting 28-day cylinder strength wou
ith a large mass of reinforced  concrete. Here, many of the techniques used with mass concrete are inapplicable because a  medium or high-strength mix is required, early strength may be necessary, and embedding  pipework may not be permitted. The essential problem is, nevertheless, the same, i.e. the  interior of the mass will heat up more than th
l.   Preplaced aggregate concrete is economical in cement, as little as 120 to 150 kg per cubic  metre (200 to 250 lb / yd3) of concrete being used, but the strength of the resultant  concrete is limited by the high water / cement ratio necessary for a sufficient plasticity of  mortar; typical strength is 20 MPa (2900 psi). However, for the usual 
applications of  preplaced aggregate concrete, this strength is generally adequate and a concrete of more  uniform properties is obtained than is the case with conventional methods of placing, as  segregation is practically eliminated. As a result, a dense, impermeable and durable  concrete is produced. No internal vibration is used but external v
before setting is thus reduced, and, as this ratio largely  controls the strength, vacuum-processed concrete has a higher strength and also a higher  density, a lower permeability and a greater durability than would otherwise be obtained.  The magnitude of the decrease in the water / cement ratio due to vacuum-processing is  given in Table 4.8. Ho
wever, some of the water extracted leaves behind voids, so  that the full theoretical advantage of water removal may not be achieved in practice. In  fact, the increase in strength on vacuum treatment is proportional to the amount of water  removed up to a critical value beyond which there is no significant increases, so that  prolonged vacuum tre
atment is not useful. The critical value depends on the thickness of  concrete and on the mix proportions. Nevertheless, the strength of vacuum-processed  concrete broadly follows the usual dependence on the final water / cement ratio, as shown  in Fig. 4.28.    Table 4.8: Water / Cement Ratio and Strength of Vacuum-Processed Concrete    Fig. 4.28
. Relation between the strength of concrete and the calculated water / cement ratio  after vacuum treatment    The vacuum is applied through porous mats connected to a vacuum pump. The mats  consist of an airtight cover, usually made of plywood, with a vacuum chamber formed by  expanded metal. This is faced with a fine wire gauze covered by muslin
s been found to have a somewhat higher strength than vacuum-processed  concrete. This is discernible in Fig. 4.28.   The formation of voids can be prevented if in addition to vacuum-processing intermittent  vibration is applied; under those circumstances a higher degree of consolidation is  achieved and the amount of water withdrawn can be nearly 
ater, an accelerator producing  flash set, such as washing soda, is used. This adversely affects strength but makes repair work  possible. Generally, shotcrete is applied in a thickness up to 100 mm (4 in.).   There are two basic processes by which shotcrete is applied. In the dry mix process  (which is the more common of the two) cement and damp 
et enough to obtain  compaction without excessive rebound. The usual range of water / cement ratios is 0.35 to  0.50; there is generally very little bleeding. In the case of mortar, the usual mix is 1:3.5 to  1:4.5, and 28-day strengths ranging between 20 and 50 MPa (3000 and 7000 psi) are  obtained. Sand with the same grading as for use in conven
 so there would be little  need for testing the strength of hardened concrete. However, in practice, mistakes, errors  and even deliberate actions can lead to incorrect mix proportions, and it is sometimes  useful to determine the composition of concrete at an early stage; the two values of  greatest interest are the cement content and water / cem
 of the resulting hardened concrete, for  example, strength and durability, are adversely affected.   The characteristics of fresh concrete which affect full compaction are its  consistency, mobility and compactability. In concrete practice these are often  collectively known as workability. The ability of concrete to maintain its uniformity is  g
er content while maintaining constant workability. The former may result  in a slight reduction in concrete strength. The water-reducing admixtures, including superplasticizers,  are now more widely accepted as a workability aid. It should also be noted that the use of  pulverized-fuel ash can also result in a significant improvement in concrete w
figure 13.4.  Several methods have been developed for evaluating the shape of aggregate, a  subject discussed in chapter 12. Angularity factors together with grading modulus and equivalent  mean diameter provide a means of considering the respective effects of shape, size and grading of  aggregate (see chapter 15). Since the strength of a fully co
o	unworkable that it cannot be effectively compacted, with the result that its strength and  other properties become adversely affected. Corrective measures frequently taken to ensure that  concrete at the time of placing has the desired workability are either an initial increase in the water  content or an increase in the water content with furth
er mixing shortly before the concrete is  discharged. When this results in a water content greater than that originally intended, some reduction  in strength and durability of the hardened concrete is to be expected unless the cement content is  increased accordingly. This important fact is frequently overlooked on site. It should be recalled that
 transportation, dropping from  excessive heights during placing and over-vibration during compaction should be avoided.  Blemishes, sand streaks, porous layers and honeycombing are a direct result of  segregation. These features are not only unsightly but also adversely affect strength,  durability and other properties of the hardened concrete. I
t	is important to realise that the  effects of segregation may not be indicated by the routine strength tests on control  specimens since the conditions of placing and compaction of the specimens differ from  those in the actual structure. There are no specific rules for suspecting possible segregation  but after some experience of mixing and hand
nt within a concrete mass produce corresponding  changes in its properties. For example, the strength of the concrete immediately  underneath the reinforcing bars and coarse aggregate particles may be much less than the  average strength and the resistance to percolation of water in these areas is reduced. In  general, the concrete strength tends 
o	exist in  cement). They have been found to react with some  aggregates, the  products of the reaction causing distintegration of the concrete, and have  also been observed to affect the rate of the gain of strength of cement. It  should,  therefore, be pointed out that the term "minor compounds" refers  primarily to their  quantity and not neces
ically combined  water; (e) the amount of unhydrated cement present (using X-ray quantitative analysis);  and (f) also indirectly from the strength of the hydrated paste. Recently, thermogravimetric  techniques and continuous X-ray diffraction scanning of wet pastes undergoing hydration  have been successfully used in studying early reactions.    
n when diffusion through the pores in the products of  hydration becomes the  controlling factor.   It is interesting to observe that calcium silicate hydrates show a strength  development similar to that of Portland cement. A considerable strength is possessed long  before the reaction of hydration is complete and it would thus seem that a small 
amount  of the hydrate binds together the unhydrated remainder; further hydration results in little  or no increase in strength.   Ca(OH)2 liberated by the hydrolysis of the calcium silicates forms thin hexagonal  plates, often tens of um across, but later they merge into a massive deposit.    Tricalcium Aluminate Hydrate and the Action of Gypsum 
 contributes little or nothing to  the strength of cement except at early ages, and when hardened cement paste is attacked by  sulphates, expansion due to the formation of calcium sulphoaluminate from C3A may  result in a disruption of the hardened paste. However, C3A acts as a flux and thus reduces  the temperature of burning of clinker and facil
res some strength, for practical purposes it is convenient to distinguish  setting from hardening, which refers to the gain of strength of a set cement paste.   In practice, the terms initial set and final set are used to describe arbitrarily chosen  stages of setting. The method of measurement of these setting times is described on page  50.   It
ous framework and the strength characteristics of the cement  paste would be adversely affected.   Apart from the rapidity of formation of crystalline products, the development of  films around cement grains and a mutual coagulation of components of the paste have also  been suggested as factors in the development of set.   The setting process is 
that no appreciable heat is  evolved, and remixing the cement paste without addition of water restores plasticity of the  paste until it sets in the normal manner and without a loss of strength.   Some of the causes of false set are to be found in the dehydration of gypsum  when interground with too hot a clinker: hemihydrate (CaSO4.1/2H2O) or any
 in the fineness of the cement particles, and  for a rapid development of strength high fineness is necessary (see Fig. 1.4).    Fig. 1.4. Relation between strength of concrete at different ages and fineness of cement     On the other hand, the cost of grinding to a higher fineness is considerable, and  also the finer the cement the more rapidly i
unit weight, would play only a small role in the  process of hydration and development of strength.   However, the sieve test yields no information on the size of grains smaller than 90  um (No. 170 ASTM) sieve, and it is the finer particles that play the greatest part in the  early hydration. Attempts to use smaller sieves, down to 53 um (No. 270
m2 / kg is specified by BS 915: 1972, but slightly higher values are generally  encountered in practice.   In some countries, e.g. Switzerland, fineness of cement is not prescribed in  standards but is indirectly controlled by tests on early strength.      Structure of Hydrated Cement    Many of the mechanical properties of hardened cement and con
n	 other words, 19 per cent of the original weight of cement has remained unhydrated and  can never hydrate since the gel already occupies all the space available, i.e. the gel / space  ratio (see p. 275) of the hydrated paste is 1.0.   It may be added that unhydrated cement is not detrimental to strength and, in  fact, among pastes all with a gel
 / space ratio of 1.0 those with a higher proportion of unhydrated  cement (i.e. a lower water / cement ratio) have a higher strength, possibly because in such pastes the layers of hydrated paste surrounding the unhydrated grains are thinner. Abrams  obtained strengths of the order of 280 MPa (40 000 psi) using mixes with a water / cement  ratio o
ssive strengths up to 375 MPa (or 54 500 psi) and tensile strengths up to 25 MPa  (or 3600 psi) were measured. The porosity of such mixes and therefore the "equivalent"  water / cement ratio are very low.   On the other hand, if the water / cement ratio is higher than about 0.38, all the  cement can hydrate but capillary pores will also be present
hout the progress of hydration. In other words, particles of the  same size are formed all the time and the already existing gel particles do not  grow in size. This is not, however, the case in cement with a high C2S content.      Mechanical Strength of Cement Gel    There are two classical theories of hardening or gain of strength of cement. Tha
ded by W. Michaelis in 1893 states that the  crystalline aluminate, sulpho-aluminate and hydroxide of calcium give the initial  strength. The lime-saturated water then attacks the silicates and forms a  hydrated calcium silicate which, being almost insoluble, forms a gelatinous mass.  This mass hardens gradually owing to the loss of water either b
e. Gel is thus taken to mean the cohesive mass of hydrated cement in its  densest paste, i.e. inclusive of gel pores, the characteristic porosity being about 28 per cent.   The actual source of strength of the gel is not fully understood but it probably arises  from two kinds of cohesive bonds. The first type is the physical attraction between sol
ross-linked by chemical forces. These are much stronger  than van der Waals' forces but the chemical bonds cover only a small fraction of the  boundary of the gel particles. On the other hand, a surface area as high as that of cement  gel is not a necessary condition for high strength development, as high-pressure steam-cured  cement paste, which 
has a low surface area, exhibits extremely good hydraulic properties.   We cannot thus estimate the relative importance of the physical and chemical  bonds but there is no doubt that both contribute to the very  considerable strength of the  hardened paste.      Water Held in Hydrated Cement Paste    The presence of water in hydrated cement has be
ts agree with the specific  surface areas of hydrated neat C3S and C2S. Likewise, the water of hydration agrees with  the additivity of the individual compounds.   This argument does not, however, extend to all properties of hardened cement  paste, notably to shrinkage, creep, and strength; nevertheless, the compound composition  gives some indica
ss restrictive than they used to be (see Table 1.9).    Table 1.9: ASTM Specification C 150-78a: Compound Composition Limits for Cement     The difference in the early rates of hydration of C3S and C2S - the two silicates  primarily responsible for the strength of cement paste - has been mentioned earlier. A  convenient approximate rule assumes th
at C3S contributes most to the strength  development during the first four weeks and C2S influences the gain in strength from four  weeks onwards. At the age of about one year the two compounds, weight for weight,  contribute approximately equally to the ultimate strength. Neat C3S and neat C2S have  been found to have a strength of the order of 7
0	MPa (10 000 psi) at the age of 18  months, but at the age of 7 days C2S had no strength while the strength of C3S was about  40 MPa (6000 psi). The development of strength of neat compounds is shown in Fig. 1.17.    Fig. 1.17. Development of strength of pure compounds     As mentioned on page 14, the calcium silicates appear in commercial cement
s	in  "impure" form. These impurities may strongly affect the rate of reaction and of strength  development of the hydrates. For instance, the addition of 1 per cent of Al2O3 to pure C3S  increases the early strength of the paste, as shown in Fig. 1.18. According to Verbeck, this  increase strength probably results from activation of the silicate 
crystal lattice due to  introduction of the alumina (or magnesia) into the crystal lattice with resultant activating  structural distortions.    Fig. 1.18. Development of strength of pure C3S and C3S with 1 per cent of Al2O3       The influence of the other major compounds on the strength development of  cement has been established less clearly. C
3A contributes to the strength of the cement  paste at one to three days, and possibly longer, but causes retrogression at an advanced  age, particularly in cements with high C3A or (C3A + C4AF) content. The role of C3A is  still controversial.   The role of C4AF in the development of strength of cement is also debatable, but  there certainly is n
o	appreciable positive contribution. It is likely that colloidal hydrated  CaO.Fe2O3 is deposited on the cement grains, thus delaying the progress of hydration of  other compounds.   From knowledge of the contribution to strength of the individual compounds  present it might be possible to predict the strength of cement on the basis of its compoun
d	 composition. This would be in the form of a formula of the type -    strength = (formula),    where the symbols in brackets represent the percentage by weight of the compound, and a,  b, etc. are constant parameters representing the contribution of one per cent of the  corresponding compound to the strength of the cement paste.   The use of suc
h	a formula would make it easy to forecast at the time of  manufacture the strength of cement and would reduce the need for conventional testing. In  practice, however, the influence of different compounds is not always significant and has  been found to depend on age and on the curing conditions. In general terms, an increase in  the C3S content 
increases strength up to 28 days; Fig. 1.19 shows the 7-day strength of  standard mortar made with cements of different composition and obtained from different  works. The C2S content has a positive influence on strength at 5 and 10 years only, and  C3A a positive influence up to 7 or 28 days but a negative influence later on. The  influence of al
kalis is considered on page 48).    Fig. 1.19. Relation between 7-day strength of cement paste and the C3S content in cement  (Each mark represents cement from one plant)     Prediction of the effects of compounds other than silicates on strength is  unreliable. According to Lea, these discrepancies may be due to the presence of glass in  clinker,
 discussed more fully in the succeeding section. In other words, the relations  observed are statistical in nature, and deviations arise from the fact that we are ignoring  some of the variables involved. It can be argued, in any case, that all constituents of  hydrated Portland cement contribute in some measure to strength in so far as all produc
ment of strength. It is not  surprising, therefore, that there is a definite relation between the degree of hydration and  strength. Figure 1.21 shows, for instance, an experimental relation between the  compressive strength of concrete and the combined water in a cement paste with a  water / cement ratio of 0.25. These data agree with Powers' obs
ervations on the gel / space  ratio, according to which the increase in strength of a cement paste is a function of the  increase in the relative volume of gel, regardless of age, water / cement ratio, or compound  composition of cement. However, the total surface area of the solid phase is related to the  compound composition, which does affect t
he actual value of the ultimate strength.    Fig. 1.21. Relation between compressive strength and combined water content	The effects of the minor compounds on the strength of cement paste have not  been thoroughly investigated as these compounds were not thought to be of importance as  far as strength is concerned.  K2O is believed to replace 
one molecule of CaO in C2S with  a consequent rise in C3S content above that calculated. Tests on the influence of alkalis  have shown that the increase in strength beyond the age of 28 days is  strongly affected by  the alkali content: the greater the amount of alkali  present the lower the gain in strength.  This has been confirmed by two statis
tical evaluations of strength of several hundred  commercial statistical cements. The poor gain in strength between 3 and 28 days can be  attributed more specifically to water-soluble K2O present in the cement. On the other  hand, in the total absence of alkalis, the early strength of cement paste can be abnormally  low. Accelerated strength tests
 (see Page 570) have shown that, up to 0.4 per cent of  Na2O, strength increases with an increase in the alkali content (Fig. 1.22). The role of  alkalis is important with reference to cements made by the dry process, which sometimes  contain relatively large amounts of alkalis.    Fig. 1.22. Effect of alkali content on accelerated strength     Th
 completely to yield its equilibrium products, and, as we have seen, the  reactivity of glass is different from that of crystals of similar composition.   It can be seen then that the rate of cooling of clinker, as well as, possibly, other  characteristics of the process of cement manufacture, affects the strength of cement and  defies attempts to
 develop a formula of the type mentioned in the preceding section.  Nevertheless, if one process of manufacture is used and the rate of cooling of clinker is  kept constant, there is a definite relation between compound composition and strength.      Tests on Physical Properties of Cement    The manufacture of cement requires stringent control, an
tures on electrical properties.	It should be remembered that the speed of setting and the rapidity of hardening,  i.e. of gain of strength, are entirely independent of one another. For instance, the  prescribed setting times of rapid hardening cement are no different from those for ordinary  Portland cement, although the two cements harden at di
excess of calcium  sulphate, but its content can be easily determined by chemical analysis.    Strength of Cement    The mechanical strength of hardened cement is the property of the material that is perhaps  most obviously required for structural use. It is not surprising, therefore, that strength  tests are prescribed by all specifications for c
ement.   The strength of mortar or concrete depends on the cohesion of the cement paste,  on its adhesion to the aggregate particles, and to a certain extent on the strength of the  aggregate itself. The last factor is not considered at this stage, and is eliminated in tests on  the quality of cement by the use of standard aggregates.   Strength t
ests are not made on a neat cement paste because of difficulties of  moulding and testing with a consequent large variability of test results. Cement-sand  mortar and, in some cases, concrete of prescribed proportions and made with specified  materials under strictly controlled conditions are used for the purpose of determining the  strength of ce
ment.   There are several forms of strength tests: direct tension, direct compression, and  flexure. The latter determines in reality the tensile strength in bending because, as is well  known, cement paste is considerably stronger in compression than in tension. Since the  flexure test is not used in Great Britain and little used elsewhere it wil
l	not be further  discussed.   The direct tension test used to be commonly employed but pure tension is rather difficult  to apply so that the results of such a test show a fairly large scatter. Furthermore, since  structural techniques are designed mainly to exploit the good strength of concrete in  compression, the tensile strength of cement is 
often of lesser interest than its compressive  strength. For these reasons, the tension test has gradually given way to compression tests.   However, the tension test still exists in some countries as a permitted test for a  one-day strength of rapid hardening Portland cement, and the details of the test as  prescribed in the 1971 edition of BS 12
	temperature between 18 and 20 C (64 and 68 F) in an atmosphere of at least 90 per cent  relative humidity, and tested in direct tension, the pull being applied through special jaws  engaging the wide ends of the briquette. BS 12 : 1958 prescribes the minimum one-day  strength of rapid hardening Portland cement as 2.1 MPa (300 psi), taken as the 
average  value for six briquettes.   There are two standard methods of testing the compressive strength of  cement: one uses mortar, the other concrete.   In the mortar test, a 1:3 cement-sand mortar is used. The sand is again the  standard Leighton Buzzard sand, and the weight of water in the mix is 10 per cent of the  weight of the dry materials
n water until tested in a wet-surface condition. The BS 12 : 1978  requirements for minimum strengths (average values for three cubes) are given in Table  1.10. It is expected that the mortar test will be deleted when BS 4550 is next revised.    Table 1.10: BS 12: 1978 Requirements for Strength of Cement     The vibrated mortar test gives fairly r
eliable results but it has been suggested that  mortar made with one-size aggregate leads to a greater scatter of strength values than  would be obtained with concrete made under similar conditions. Moreover, the values of  strength obtained in a test should approximate the level of strength generally found in  concrete, and this would require the
erature and humidity conditions  of the mixing room, curing chamber, compression testing room, and the temperature of  the water curing tank are specified. The BS 12: 1978 requirements for the minimum values  of the average strength of three cubes at each age are given in Table 1.10. Apart from  satisfying these minima, the strength at later ages 
has to be higher than at an earlier age, as  strength retrogression might be a sign of unsoundness or other faults in the cement. The  requirement of strength increase with age applies also to the vibrated mortar cubes.   The influence of cement on the properties of mortar and concrete is qualitatively  the same, and the relation between the stren
gths of corresponding specimens of the two  materials is linear. This is shown, for instance, in Fig. 1.26: mortar and concrete of fixed  proportions, each with a water / cement ratio of  0.65 were used. The strengths are not the  same for the specimens of each pair, at least in part because specimens of different shape  and size were used, but th
ere may also be an inherent quantitative difference between the  strengths of  mortar and concrete due to the greater amount of entrapped air in mortar.    Fig. 1.26. Relation between the strengths of concrete and mortar of the same  water / cement ratio     It may be interesting to know to what extent commercial cements comply with  the standard 
requirements. Data on samples of cement which failed to comply in  Switzerland in the period 1965-1975 are given in Table 1.11; of the total of over 7000  samples tested, 1.1 per cent failed.  We can see thus that failure is rare and, with respect to  strength, virtually non-existent.    Table 1.11: Distribution of Samples of Cement Failing to Com