This is a paper from the 1960s about building joints. I have read it a bunch of times over the years and always thought it was interesting. But now that I read it again, it is packed full of very useful discussions and understandings.
This includes uncertainty on things that I still don’t think are resolved today.
Building Movements and Joints
This text is intended for the concerned reader-whether engineer, architect, builder, or student-who wants to know more about why and how the parts of concrete buildings move and change shape, and how these changes can be controlled.
Temperature and moisture changes are always at work, changing the length, width, and thickness of concrete, masonry, and other building materials. Gravity pulls, bends, and twists the parts of a building as do wind, rain, snow, and earthquake. Chemical actions sometimes occur and cause movements that must be anticipated. These individual movements are minute, but they can have noticeable effects as the building adjusts itself into balance with its environment. When they are recognized and understood, they are easier to cope with.
Although the causes of the movements, fractures, and distress encountered in buildings can be complex, the remedies are usually quite simple. Ways are shown for controlling these movements within the state of the art. If such preventive measures are included in the planning stages, the builder as well as the user can have confidence that the building will be attractive, comfortable, and safe.
Part One: Movements in Concrete Structures
Chapter 1: Movement, Strains, and Volume Changes - Not Restrained
To understand the behavior of concrete, it is helpful to examine its movements and length changes as they occur under two separate conditions: without restraint and with restraint. This chapter concerns the unrestrained condition.
Concrete, in common with other building materials, will change in dimension a definable amount due to moisture, temperature, and certain chemical changes; and when it is free to move without restraint externally and internally, no stress or damage will result to the concrete itself, though the concrete may intrude on some function of a building.
On the other hand, if concrete is restrained externally or internally, the development of deformations, stresses, and possible cracking in the concrete will result from the
influence of external forces and temperature, moisture, and certain chemical changes.
For convenience in this text, the amount of volume change is generally stated in linear rather than volumetric units. Changes in length are often expressed as a coefficient of the length in parts per million, or more simply as millionths. Thus 500 millionths is 0.000500, which is equal to 0.050%. Percentages are used in this text as the common measure of volume change. They are shown to three decimal points, not as a measure of accuracy, but to make it easier to compare the values and determine their relative importance.
MOISTURE CHANGES
Concrete shrinks as it dries and swells as it is wetted. As a rule, when concrete dries and becomes resaturated, not more than two-thirds of the initial drying shrinkage will be recovered. This is illustrated in Fig. 1, where a cement paste specimen is initially dried to 50% relative humidity and then resaturated, but it does not return to its original length.(1) After this, wetting and drying follows the lower diagonal line. During first drying concrete undergoes structural alteration or stabilization, depending on its porosity. Following this, the reversible shrinkage due to wetting and drying is fairly constant. Materials with high initial irreversible shrinkage require maturing before
use.(2)
Dry Shrinkage of Concrete
Several important factors have separate influences on the amount of shrinkage that accompanies drying: (1) the total amount of water in the concrete as illustrated by Fig. 2;(3) (2) the amount the aggregate shrinks; (3) the elastic modulus of the aggregate; and (4) the total amount of aggregate in the concrete, also called the aggregate volume concentration.(4 5)
If concrete is considered as two parts-cement paste and aggregates-the shrinkage action is more readily understood. The two parts do not shrink to the same extent. The cement paste shrinks more than the aggregate and shrinkage of the paste by itself can be five times or more that of the concrete. The absolute volume of aggregate is two to three times that of the paste, so the aggregate has an important role in restraining shrinkage. (6) The relative elasticity of the paste and aggregate helps determine the net shrinkage of the concrete. Table I lists the shrinkage of concretes made with different aggregates. (7)
Drying shrinkage of structural lightweight concrete and also of high-strength concrete made and cured at normal temperatures may be comparable to or slightly greater than that of normal-weight concrete. The difference in shrinkage is usually less than about 30%, and in some cases there is little or no difference. High-strength lightweight concrete, 7000 to 9000 psi (48 to 62 MPa), has about the same shrinkage as comparable normal-weight concrete.
Normally included in values cited for drying shrinkage is the amount of autogenous volume change, which is occasionally an expansion but usually a shrinkage. This change is caused by hydration of cement and aging, and it may range from 0.001% to 0.015%.(8, 9)
According to research conducted in 1966, both size and shape of concrete sections have an effect on how fast and how much shrinkage develops. Fig. 3 shows the rate of shrinkage with respect to total shrinkage.' 101 Neglecting the ends, each thickness of concrete slab or wall has a distinct volume-to-surface (v-s) ratio as does each size of round or square column. Thus a 6-in. (150-mm) slab has a v-s ratio of 3, as does a 12-in. (300-mm) column (round or square). At an age of 100 days, the shrinkage selected from Fig. 3 for this slab or column would be just under 60% of its total shrinkage. The curves represent conditions at 70°F (21°C) and 50% relative humidity.
Note that volume and surface are measured in inches; measuring in feet or millimeters will produce different numbers.
The v-s ratio for prismatic sections can be expressed as the cross-sectional area divided by the periphery (or that part of the periphery that is exposed to drying). A concrete slab or wall that can only dry from one side would have the same drying characteristics as a slab or wall twice as thick that can dry from both sides. Thus a 6-in.-thick (150-mm) slab drying from one side would shrink the same amount as a 12-in. (300-mm) wall drying from both sides.
Since 1966 researchers have reached a conclusion that "the ultimate shrinkage is broadly independent of specimen size."(12, 13) However, depending on the method used to extrapolate measured data to infinite time, different conclusions can be reached. All authors agree that the rate of shrinkage is size-related. Therefore, from a practical viewpoint, because of the large size and long drying time of most beams, columns, and foundations, it is appropriate to assume that shrinkage during the finite life of a structure is size-related.
Drying Shrinkage of Lightweight Concrete Masonry
Drying shrinkage curves for expanded-shale concrete masonry units are given in Fig. 4. These units were initially cured under high-pressure steam and then cured at four lower humidities for 30 days each. Drying shrinkage would be twice as much for units initially cured at 72°F (22°C) in moist air. The curves show a substantial amount of shrinkage at 85% relative humidity (RH) and further substantial drops at 50% and 25%. The shrinkage is not directly proportional to the moisture loss.
The rate of drying shrinkage in a concrete masonry wall, however, would be slower. Fig. 5 is for an 8-in.thick (100-mm) expanded-shale block wall cured at 73° F (23°C) and 50% external relative humidity. The mortar dried rapidly at first, reaching 94% RH internally in one day, but took seven more weeks to reach 85%. The face shell aged four weeks before reaching 85% internally, and 6 months before reaching 60%.(15)
American Society for Testing and Materials (ASTM) Designation: C90, Standard Specifications for Hollow Load-Bearing Concrete Masonry Units (and Canadian Standards Association [CSA] Al65.l, Concrete Masonry Units), limits the moisture content for Type I (moisture-controlled) concrete masonry units as delivered to the jobsite. The maximum permissible moisture content depends on the average annual relative humidity at point of use and the potential linear drying shrinkage of the concrete used in making the units. At most, the ASTM specification allows only a 45% moisture content for units with low shrinkage in an environment with relative humidity above 75%.
Plastic Shrinkage and Plastic Cracking
Plastic shrinkage and cracking sometimes occur in the exposed surfaces of freshly placed concrete slabs because of a rapid evaporation of water from the surface of the concrete. Plastic shrinkage and cracking are usually associated with hot-weather concreting (high temperatures and low humidity), but can occur at other times as well. The cracks can be as deep as 4 in. (100 mm). They are rather straight, without a pattern, and are generally discontinuous. They usually develop during finishing operations after the water sheen on the surface disappears and indicate that the rate of evaporation exceeds the rate of bleeding.
TEMPERATURE CHANGES
Movements in the category of temperature changes are significant. They depend on the local weather conditions and on heating and cooling within the building. Temperature gradients can cause curling or bending of concrete slabs, as shown in Fig. 6, when the temperature varies from one face to the other. This can occur in precast wall panels and in slabs and pavements on the ground.
An average coefficient of expansion of plain concrete is on the order of 0.055% per 100°F (0.100% per 100°C) or 85% of the coefficient for steel. Thus a rise or fall of 100°F will cause a length change of 0.66 in. in a 100-ft length of concrete ( or 5 mm for a 10-m-long section with a rise or fall of 50°C).
Of course, an average coefficient does not apply to all concrete. Table 2 shows some experimental values of the thermal coefficient of expansion for unrestrained laboratory specimens of concrete made with various aggregates. The coefficient for structural lightweight concrete varies from 0.040% to 0.060% per 100°F (0.070% to 0.110% per 100°C), depending on the aggregate type and the amount of natural sand.
Thermal coefficients for expansion of concrete vary because the coefficients for the aggregates vary widely as shown in Table 3. Aggregates are often complex in terms of type and mineral content and so the thermal coefficients cannot be neatly classified by rock or mineral type. It is considered that "the main factor influencing the thermal expansion of rock, and therefore of concrete, is the proportion of quartz present. Rocks with a high quartz content, such as quartzite and sandstone, have the highest coefficients. Rocks containing little or no quartz, such as limestone, have the lowest coefficient."(19)
The expansion coefficient for a given aggregate may vary with different grain textures and with temperature as shown in Fig. 7. So it can be seen why the coefficient for concrete can also vary considerably with the temperature. This points out the need for using averages judiciously and in critical situations the need for testing the aggregate or the concrete.
Another reason why thermal expansion coefficients cover a wide range is because of the method of reporting. Some research reports compute the coefficient merely on the basis of the length change of moist concrete for a rise in temperature. During this process drying shrinkage takes place. But it is not identified or separated from the temperature expansion, because that particular researcher may only be interested in the total effect and because it is difficult or laborious to isolate the shrinkage effect.
Further, a major influence on the factor is the size of the specimen and the rate of temperature increase. Laboratory specimens are small compared with building components such as foundations, columns, girders, and slabs. This size effect as well as the actual speed of ambient air temperature change strongly influences the actual temperature movements in a building.
Cement paste has a higher expansion coefficient than aggregate; and as aggregate is added to the paste, the coefficient drops. Thus
- Cement paste (cement plus water) has the highest coefficient.
- Mortar (paste plus sand) has a lower coefficient.
- Concrete (mortar plus coarse aggregate) has the lowest coefficient.
The thermal coefficient for cement paste may range from 0.060% to 0.090% per 100° F (0.108% to 0.162% per 100°C).(29) The difference in coefficients of mortar and aggregates is illustrated in Table 4 and further shows how the thermal coefficient of concrete is affected by the type of aggregate. Research has shown some correlation of these differences to the durability of concrete exposed to freezing and thawing( 17 21) but the role of this relationship has not yet been fully established.
Another variation of thermal expansion of concrete relates to the moisture content. Air-dry aggregate by itself may have a 10% higher coefficient of expansion than saturated aggregate. It is believed that the thermal
coefficient of expansion of concrete is about equal to the weighted average of the coefficients of its ingredients. (19)
The thermal coefficient of expansion of concrete will vary for other reasons as well. One is called hygrothermal volume change, in which a change of temperature of concrete causes a migration of water between the gels and capillaries of the concrete. As a result, the damp or wet
expansion of concrete will not be the same as forovendry concrete. The absolute difference may be plus or minus 0.010%. However, the difference disappears within a half hour after the temperature stabilizes. The phenomenon yields its largest apparent thermal coefficients at about
68° F (20° C). High-pressure steam curing eliminates hygrothermal shrinking and swelling.(22)
Some data suggest that the water-cement ratio of the concrete has a strong influence on the coefficient. For example, at an age of two months the coefficient for concrete with a water-cement ratio of 0.50 was 16% more than for concrete with a water-cement ratio of 0.70. On the other hand, the maturity of concrete may cause the
coefficient to drop 25% between the age of two and six months. (23)
The presence of reinforcing steel in concrete will influence the composite thermal expansion, since rarely do both materials have identical coefficients. However, the difference may be no more than 5% to 6% when reinforcement
is placed symmetrically in the concrete. If it is not placed symmetrically, as in a floor slab, the thermal coefficient of the concrete remote from the steel will change very little. In a rigid frame with 6% reinforcement/23l the coefficient near the bars may be 10% different, and thus would contribute to curling or curling stresses.
Temperature movements are most severe at a roof. A black roofing under a hot sun can rise 90° F (50° C) above the ambient air temperature during the day. Even with insulation under the roofing membrane, the temperature will rise in the concrete roof slab, creating a tendency to curl, expand horizontally, and move masonry parapets.
Partitions and facing materials in the top story of a building must be detailed to accommodate these movements.
Heat of Hydration
Heat of hydration is the heat concrete generates as it hardens because of the chemical reaction of the cement and water. Heat of hydration can be another source of temperature-caused movement.
If the heat cannot escape from the concrete, the condition is known as adiabatic. In the real world this situation cannot fully exist, but something close to it does occur. Fig. 8 shows temperature buildup under adiabatic conditions, which for Type I cement at 28 days, for example, amounts to almost 40° F (55°C).
This heat rise would of course be a matter to reckon with in planning a building with massive structural elements or massive foundations. If the temperature increased in a long slab or wall resting on the ground, the concrete would extend its length by sliding on the ground; but on cooling, the friction of the ground could cause
tensile forces high enough to severely crack the concrete. When reinforcing steel is provided, the crack widths are less severe. Another consequence during expansion could be to push other parts of the building out of alignment.
In the slabs, beams, and columns with typical dimensions, heat of hydration creates no problems because it dissipates rapidly. But in very thick or massive concrete sections, the heat does not escape quickly. As an example, consider walls of various thicknesses and the time it takes
for 95% of a fixed amount of heat to escape out the sides when the surrounding air temperature is constant. The time for the heat to dissipate is proportional to the square of the thickness of the wall (25)
Another consideration is the temperature of the concrete, where heat is dissipating at the same time that heat of hydration is being created. It can be shown that if the temperature of the air is constant and the same as that of the concrete as delivered, the heat rise in the concrete after 24 hours will be only a few degrees Fahrenheit for
walls as thick as 24 in. (610 mm).
If the air temperature has normal weather fluctuations and the concrete is allowed to lose heat at the same time it is creating heat of hydration, the resultant temperature
of the concrete will tend to follow the air temperature, though with some timelag.
The differential temperature between the surface and the interior of the concrete is decisive, as it may cause cracking. Studies indicate that this difference in the temperature gradient should not exceed 36°F (20°C) for concrete with siliceous gravel if subsequent cracking is to be avoided. (26, 27) For concrete with granite aggregate a temperature differential of 45°F (25°C) is considered a limit, and with limestone aggregate the limit is 56°F (3l°C).(27) Other methods for computing maximum temperature
rise of uninsulated sections appear in Reference
30 and sample curves appear in Reference 28.
CHEMICAL CHANGES AND EFFECTS
Certain movements of concrete result from chemical reactions of the cement and aggregates with the mixing water or later reactions within the hardened concrete in the presence of water or moisture.
Carbonation
Hardened concrete reacts with the carbon dioxide and moisture present in the air and develops a slight shrinkage. The effect is not destructive but actually increases the chemical stability of the concrete. The reaction is very
shallow in dense concrete, and penetrates deep in porous concrete. Some concrete masonry units are deliberately exposed to carbon dioxide during manufacture after reaching 80% of their rated strength to induce shrinkage and make the units more dimensionally stable. Future shrinkage is reduced 30% or more. (29)
On hardened concrete, long-term carbonation may reach a depth of an inch (25 mm) but only over a period of many years. One of the causes of surface crazing of concrete, however, is the shrinkage that accompanies natural air carbonation of young concrete.
Carbonation of another kind also can occur to fresh, unhardened concrete and cause a soft chalky surface. This carbonation usually takes place during cold-weather concreting when there is an unusual amount of carbon dioxide in the air due to direct-fired heating or gasoline powered
equipment. It is not accompanied by significant
movements or cracking, and after the concrete is 24 hours old this danger no longer exists.
Sulfate Attack
Sulfate attack on concrete can occur where soil and groundwater have a high sulfate content. The attack is confined to concrete that is cool and moist, such as foundations and slabs on ground. It usually causes an expansion of the concrete because of the formation of solids as a result of the chemical action. The amount of expansion
in some circumstances has often been higher than 0.100%, and the disruptive effect within the concrete can result in extensive cracking and disintegration. The amount of expansion cannot be accurately predicted; it is not useful in the design of concrete units or structures and cannot
be controlled by jointing. Rather than try to accommodate the expansion, the designer should find a way to avoid it altogether.
Alkali-Aggregate Reactions
In most parts of North America, concrete aggregates are considered more or less chemically inert in concrete. However, in some areas the aggregates definitely react with the alkalis in cement, causing expansion and cracking over a period of years. The reaction is limited to those parts of a structure that are cool and moist. A knowledge of the characteristics of local aggregates is helpful. There are two types of alkali reactive aggregates: siliceous and carbonate.
Alkali-silica reaction occurs in certain localities from coast to coast. In addition to expansion and cracking, the reaction causes a gel, which may be soft or hard, to ooze from pores and cracks and collect on the surface. Cracking usually is random but it may have a pattern, sometimes
called map cracking. Sometimes popouts occur.
Most noncrystalline silica minerals, such as opal, are potentially reactive with the alkali in cement. Some igneous and metamorphic siliceous rocks are also potentially reactive. The expansion may exceed 1.500% in mortar or 0.500% in concrete and can cause the concrete to fracture and break apart.
Alkali-carbonate reaction is less widespread but has been encountered in the Middle West and on the East Coast and could be found anywhere that carbonate rocks are found. Only a few of the alkali-reactive rocks are expansive-mostly some of the various kinds of dolomitic limestones. The characteristics of reactive carbonate aggregate are a relatively high clay content, a finegrained
matrix, and a 1:1 approximate dolomite-to-calcite ratio. The amount of expansion has been measured as high as 0.180% in 9 months for 3x4-in. (75x100-mm) prisms tested at 100% relative humidity and 73°F (23°C). (30)
Knowing the amount of alkali-aggregate expansion is not useful in the design of concrete structures, and the expansion cannot be controlled by jointing. Rather than trying to accommodate the expansion, the designer
should find a way to avoid it altogether.
Shrinkage-Compensating Cement
Shrinkage-compensating cement was developed to produce a concrete without the net drying shrinkage associated with portland cement concrete. Expansion takes place during the first week and is on the order of 0.030% to 0.100%, depending on the amount of restraint present
and the mix design. Fig. 9 shows the early expansion and later shrinkage upon drying. The expansion more than offsets the shrinkage so there is no net shrinkage in the concrete.
By far, the greatest use of shrinkage compensating cement is in concrete floor slabs on the ground. Restraint by reinforcing steel-properly selected and placed steel-is essential to the effective performance of shrinkage compensating concrete. The elasticity of the reinforcing steel restores the concrete to almost its original length as shrinkage progresses, and cracking is avoided. Polyethylene sheets are sometimes used under the slab to reduce
subgrade friction; a 2-in. (50-mm) sand cushion on top of the sheets is recommended if the slab is cast under the open sky.
The simplicity of shrinkage compensation can be appreciated when three types of slabs on ground are compared (Fig. 10). They are relatively long slabs, all cast to the same length. The action of the first (Fig. 10a), with shrinkage-compensating cement, has been discussed. The second slab (Fig. 10b), with normal portland cement, has a control joint. The slab developed a microscopic crack at the control joint because of a temperature drop during the first 24 hours after placement when the concrete had very little strength. Later, drying shrinkage opened the crack.
The third slab (Fig. 10c) also developed a microscopic crack from a temperature drop. A day or two later the slab was post-tensioned, closing the crack. As drying shrinkage developed, the slab shortened. The tendon coupled the two halves together and the crack remained closed.
Corrosion of Metals in Concrete
Occasional corrosion of reinforcing steel has been observed in the form of rust stains, cracking, and spalling of the concrete. The products of corrosion occupy a greater volume than the metal that has corroded. They
may cause internal forces of expansion that can crack and loosen the surrounding concrete. Engineers are aware of the generally excellent service record of concrete structures, determined by long-time studies of specimens in
widely varying exposures. Not so well understood is the remarkably effective protective mechanism afforded to steel encased in concrete or mortar. The protection provided by concrete and mortar is comparable to that of other types of coatings. When corrosion does occur, it
can often be traced to factors such as insufficient cover over the steel, physical damage to concrete that results in steel exposure, or poor-quality concrete that is porous or contains large voids around the steel.
Calcium chloride used as an admixture has been the cause of corrosion problems, and this is a subject in itself.
Designs that place dissimilar metals in contact in moist or saturated concrete, particularly if the concrete contains calcium chloride, have produced corrosion of at least one of the metals. Various nonmetallic coatings have been applied to reinforcing steel to reduce corrosion. An epoxy coating is now in use that transmits bond
to the bar and allows the full structural value of the steel. Cathodic protection has been used successfully to counteract corrosion, particularly in bridges.
Nonferrous metals such as copper, zinc, aluminium, lead, and their alloys may be subject to corrosion when embedded in concrete or in surface contact with it. Copper and its alloys are practically immune to action from fresh concrete and mortar, except in the presence of soluble
chlorides that may cause corrosion.
Zinc (galvanizing) is susceptible to attack by fresh or moist concrete and mortar. Galvanized coatings are so thin that their expansive
pressures generally do not cause any damage to the surrounding concrete, and galvanizing does furnish sacrificial protection to steel. However, calcium chloride admixtures corrode galvanized steel and may lead to severe cracking and spalling of the surrounding concrete.
When aluminum is placed in fresh concrete, a reaction occurs and aluminum oxide and hydrogen are formed. The greater volume occupied by these products causes expansive pressures around the embedded metal and may
lead to serious damage to the surrounding concrete. Aluminium should not be used in concrete containing admixtures with chlorides, nor in or near seawater. Lead is also attacked by fresh or moist concrete, and destruction of an
embedded lead pipe may occur in a few years. When lead is partially embedded in concrete and partially exposed to air, a condition known as differential aeration occurs. The embedded lead has a different electrical potential than that in the air and an electrical current is created that will cause corrosion and gradual disintegration of the
embedded lead. Generally no damage will be observed in the concrete because of the softness of the lead, which will absorb expansive pressures caused by the formation of corrosion products.
Nickel-plated steel will not corrode in moist concrete, but pinholes in the coating may lead to accelerated corrosion in the presence of chlorides. Cadmium coatings will satisfactorily protect steel embedded in concrete, even in the presence of moisture and normal chloride
concentrations.
Nonshrink Grouts
Nonshrink grouts and mortars are regularly used under columns, precast elements, and equipment to avoid settlement due to drying shrinkage. These grouts and mortars are made with an addition such as aluminium powder
that expands when water is added to the mix. The amount of shrinkage that can develop in a 1-in. (25-mm) layer of regular grout is very minute-perhaps only 0.0005 in. (0.012 mm)-so it would be insignificant compared to the shortening of the columns or other precast elements resting on it. Nevertheless, a reason for using nonshrink
grout is to avoid loosening the nuts on the anchor bolts. However, the shrinkage may take years, or forever, to develop because the mortar is capped by a steel plate and because of the thick concrete foundation below that is perhaps in a moist environment, thus inhibiting drying shrinkage.
Chapter 2 - Movements, Strains, and Volume Changes - Restrained
If concrete were free from all restraints against its movement, there would be no cracking. In service, however, concrete is usually in a restrained condition and cracking occurs either because of too much imposed load or because of the development of greater volume changes (movements) than the concrete can withstand.
A simple example of external restraint is a concrete test cylinder in a testing machine (Fig.11). A load is applied at the top and the bed of the machine restrains the concrete from moving down. The load is increased until it is more than the concrete can withstand and the cylinder cracks and fails.
An example of tensile stresses caused by the restraint of a subgrade under a slab is in Fig.34 in this chapter. Internal restraint occurs when the surface of a mass of concrete is restrained from shrinkage during drying because the interior is wetter than the surface (Fig.12). Tensile stresses that may be large enough to cause surface cracking will be created on the surface.
ELASTIC AND INELASTIC STRAINS*
Axial
A basic movement in all materials is the strain or deformation that develops with axial stress. A generalised stress-strain curve for concrete is shown in Fig.13. Elasticity prevails up to the elastic limit and then a plastic deformation occurs. When the load is removed, the recovery line usually is not parallel to the elastic and plastic curve; so the amount of permanent set differs from the amount of plastic deformation.
A stress-strain curve is customarily drawn with stress on the vertical axis and strain on the horizontal axis. Graphs of strain from other causes, however, usually position strain on the vertical axis, as will be seen in this text. When a series of graphs for strain are to be compared, the strain should be on the same axis in each graph. Thus the stress-strain curve in Fig. 13 can be shown as it is in Fig.14, where there are curves for specific water cement ratios by weight. These curves result from tests wherein the strain is constantly increased and the concrete offers first more and then less resistance.
With moderate-to-high-strength concretes, water-cement ratios of 0.50 or less and up to strains of 0.150%, the strain is closely proportional to pressure or stress, and so
is almost elastic. The upper parts of the curves are inelastic or plastic. The maximum usable strain at the extreme Fibre of concrete in compression is assumed for structural
design to be 0.300%. The maximum measured strains shown for the lower water-cement-ratio concretes are smaller than those for the higher water-cement ratios because of the physical problems of testing. The term "elastic" is not favoured for general discussion of concrete behavior because very frequently the strain may be in the inelastic range. For this reason, a term such as "instantaneous strain" is often used.
Modulus of Elasticity
A stress-strain curve for a five-unit-high prism of lightweight concrete masonry with full mortar bedding appears in Fig. 15. (
JI) To compute the modulus of elasticity, the strength of these prisms was adjusted to the strength of a two-unit-high prism. In this case it was found that the modulus of elasticity was much higher than the commonly assumed value, which is 1000 times the ultimate compressive strength of the masonry test prism.
Normal-weight concrete has a modulus of elasticity of 2,000,000 to 6,000,000 psi (14,000 to 41,000 MPa), depending
on the compressive strength. The modulus of
elasticity of structural lightweight concrete is generally 20% to 50% lower than that for normal-weight concrete of equal strength, with the greater difference in the lower weight range. An approximate relationship can be used to estimate modulus of elasticity, Ec, for normal-weight concrete:
Ec= 57,000 sqrt(f’c) for customary units
Ec = 15,100 sqrt(f’c) for SI (metric) units
in which f’c is the compressive strength determined by test cylinders. This empirical formula is reasonably reliable for concretes with compressive strengths of 3000 to 5000
psi (20 to 35MPa). Fig. 16 illustrates this relationship. Note that the modulus of elasticity for most lightweight concretes is between 1,000,000 and 2,500,000 psi (7000 and 17,000 MPa). For important work the modulus of elasticity should be determined by laboratory tests in accordance with ASTM C469, Method of Test for Static Modulus of Elasticity and Poisson's Ratio of Concrete in Compression.
Flexural Strains (Deflections)
Deflection of concrete beams and slabs is one of the most common examples of building movements and its effects are conspicuous. As floors deflect, the partitions they support may separate from the ceiling above, cracks may develop in the wall, and doors may not close properly. Furniture may stand at an angle away from the wall, and need levelling devices to avoid rocking.
A closely allied topic is cambering or building in an upward bowing to anticipate and offset the eventual deflection. The movement of a floor or beam is the same regardless of whether it has been cambered, but the visual effects are improved. Cambering serves a definite purpose,
and if it can be accomplished economically, the owners and tenants will be more satisfied that it was done.
Deflections are the result of flexural strains that may result in cracking of the tension surface of a beam or slab. Indeed, reinforced concrete structural design has always been based on sections that will crack when loaded. The current practice of deflection calculations for reinforced concrete limits the cracking by using an effective moment of inertia. A sample deflection calculation is illustrated in Fig. 17 and further refinements are given in References 32, 33, and 34.
In the past, cracking in concrete beams was rather inconspicuous. In recent times, however, the use of high strength steel has increased the size of the cracks and they are more apparent. Experimental work has shown that crack widths can be roughly predicted. A sample calculation appears in Fig. 18 using the same beam as in Fig. 17. The limitations on crack widths are 0.016 in. (0.4 mm) indoors and 0.013 in. (0.33 mm) outdoors. A crack width as narrow as 0.002 in. (0.05 mm) is apparent to the naked eye as a very fine hairline crack-the approximate diameter of a human hair. However, concrete cracks with widths up to 0.020 in. (0.50 mm) are often called hairline cracks.
Shear Strain
Concrete, like other materials, deforms under shear forces (Fig. 19). The movement or strain is important in determining the load paths or distribution of forces in indeterminate structures-for example, where shear walls and columns both participate in resisting horizontal forces. Whether the load is applied parallel or perpendicular to the wall, the formula for the displacement at the top of the wall is the same and the answer is the same. The amount of movement, while not large, is significant in short stubby members, but is overshadowed in longer members by flexural strains. For example, a concrete shear wall will have a horizontal shear strain about equal to the horizontal flexural strain shown in Fig. 20 at the 10-ft (3-m) height. At a height of 30 ft, the shear strain at the top is only 10% of the total, which is
The shear modulus, G, varies with the strength and temperature of the concrete. When the compressive strength is between 4000 and 5000 psi (27.6 to 34.5 MPa) the shear modulus at 75°F (24° C) will average 42% of the elastic modulus, E. (35)
Poisson’s Ratio
For a typical hollow concrete masonry wall of the same nominal dimensions as in Fig. 20, shear and flexural strains could be seven times that for the concrete wall, but if the wall were filled with grout, the strains would be only two and a half times that of the concrete wall.
When a block of concrete is loaded as in Fig. 21, it will shorten and at the same time develop a lateral strain or bulging. The ratio of the lateral to axial strain is known as Poisson's ratio. The value is generally between 0.15 and 0.25, depending upon the aggregate, moisture content, and age of the concrete; and it is thought to be lower for higher strength concrete. A value of 0.20 or 0.21 is commonly used.
In ordinary building work, Poisson's ratio is of no concern to the designer or contractor; but it is used in the advanced structural analysis of flat-plate floors, roof shells, and mat foundations. As an example, take the case of a concrete slab that is square in plan, fixed along all four sides, and uniformly loaded. With a Poisson's ratio of 0.20, the bending moment at the center of the panel is 20% higher than if Poisson's ratio were zero. (36)
CREEP
When concrete is loaded, the deformation caused by the load can be divided into two parts: a deformation that occurs immediately (such as an elastic strain), and a time dependent deformation that begins immediately but continues at a decreasing rate for as long as the concrete is loaded. This latter deformation is called creep.
When the load is removed, there is an instantaneous recovery from strain that is always less than the elastic strain, and then a gradual recovery from strain (or creep recovery). As a result, there is an irrecoverable creep as well as a permanent set. This is shown in Fig. 22 for a 16-in.-diameter (400-mm) cylinder of plain concrete. The amount of creep is dependent upon (1) the magnitude of stress, (2) the age and strength of the concrete when stress is applied (the higher the strength of the concrete, the less the creep), and (3) the length of time the concrete is stressed. It is also affected by factors related to the quality of the concrete and conditions of exposure such as type, amount, and maximum size of aggregate; type of cement; amount of cement paste; size and shape of the concrete mass; amount of steel reinforcement; and curing conditions. For loads that are applied slowly at very early ages, it is difficult to separate instantaneous deformation and creep.(37) For the usual concrete strengths, creep is proportional to stress when stress is less than 40% of ultimate strength.
The range of creep for lightweight concrete is about the same as the range for normal-weight concrete. The average ultimate creep of lightweight concrete, however, is generally slightly greater than that of normal-weight concrete. When precise knowledge of creep is required, tests should be performed on the concrete in question. A very effective method of reducing creep in lightweight or normal-weight concrete is to use higher strength concrete but keep the stress the same.
Curves illustrating creep appear in Fig. 23. These are based on tests conducted under laboratory conditions in accordance with ASTM Designation C512, Standard Test Method for Creep of Concrete in Compression. Six inch-diameter (150-mm) cylinders were loaded to something less than 40% of the ultimate compressive strength. Companion cylinders not subject to load were used to measure shrinkage, which was then deducted from the total deformation of the loaded specimens to determine creep. Cylinders were allowed to dry, except those marked "sealed." The curves at 28 days in Fig. 23 illustrate the two parts of creep-the deformation accompanied by drying and the deformation that occurs without drying, sometimes called basic creep. Concrete specimens loaded at a late age will creep less than those loaded at an early age. This is especially significant when developing
schedules for floor form removal.
Fig. 24 gives the creep strain curve for the same lightweight concrete masonry five-unit-high prism shown in Fig. 15. The specimens were loaded at the age of seven days to 40% of ultimate strength, and projected final creep was calculated to be 109% of the instantaneous strain.
Both size and shape are believed to affect the total amount of creep during the first three months; afterwards, the rate is close to that for sealed specimens.
poi Large members have less creep, approaching the value of basic creep.A combination of strains is illustrated in Fig. 25. The curves represent the deformations or volume changes in the fourteenth-story column of a 76-story reinforced concrete building while it was being constructed. The column size was 16x48 in. (400x 1200 mm). The column contained just over 2% vertical reinforcement. Concrete design strength was 9000 psi (62 MPa) at 56 days. The method of curing prior to loading has a marked effect on the amount of creep in concrete specimens. Three different methods are charted in Fig. 26. Note that very little creep occurs in concrete that is cured by high pressure steam (autoclaving) and that atmospheric steam cured concrete (13 hours) has considerably less creep than 7-day-moist-cured concrete (16 hours). These two methods of steam curing reduce drying shrinkage of concrete about half as much as they reduce creep. The practical effects of creep can be a time-dependent deformation due to continual movement, a relaxation of
stress, or an intermediate condition. The situation that will exist depends upon whether the applied load, or displacement of part of the structure, is resisted wholly or only in part by the strength and stiffness of the structure. Examples of the two conditions follow.
Continual Movement
Continual movement occurs where there is no restraint, as in a simple post. An instantaneous strain is followed by and increased by creep although the stress or load on the column remains unchanged.
Continual movement from creep and drying shrinkage can cause additional deflection of beams and slabs. For example, it is often predicted that for a beam or slab with no compressive reinforcement
(A;), additional deflection after five years may be equal to twice the original amount, thus tripling the immediate deflection (Fig. 27). This of course assumes that the same load is on the member for the full time.
This multiple of the deflection will vary depending on the age of the concrete when initially loaded. More creep and shrinkage will develop if concrete is loaded at an earlier age, as in Fig. 23. When Fig. 23 is compared with Fig. 14, other values of the multiple can be derived (assuming both figures are based on the same concrete mix). For example, in Fig. 14 for a water-cement ratio of 0.50 the instantaneous strain would be about 0.025% at a stress of 1000 psi (6.9 MPa) in the elastic range. Then if a 0.50 water-cement ratio represents 4000 psi (27.6 MPa) concrete, Fig. 23 for concrete initially loaded at 28 days shows a total creep of over 0.100% per 1000 psi (6.9 MPa) at 1000 days (2.75 years). This is four times the instantaneous strain, making a total multiple of five for the creep effect. If concrete is loaded before the age of 28 days, the multiple is especially pronounced, and deflection of floors and beams is apparent.
Relaxation of Stress
Relaxation of stress may occur either after an instantaneous initial strain or with gradual strain. In either case, creep causes a readjustment of bending moments and
stresses if further strain is prevented or if further movement of some part of the structure is prevented. An example of stress relaxation after an instantaneous initial strain is illustrated by Fig. 28, which represents a huge, stiff compressor supported on a center pier and two
unyielding abutments. The compressor was lowered onto the supports after the concrete pier had hardened and strengthened. The pier at first supports almost all of the
load because it was accidentally constructed slightly higher than the abutments. As creep develops with time, the stress in the pier is relaxed as shown in Fig. 29 for curve I (instantaneous).(39)
Stress relaxation with gradual strain is illustrated in Fig. 29 by curve G (gradual) as if the compressor were being very, very slowly lowered onto the pier. Stress increases
during the strain time and then relaxes during the following years as creep continues. The peak of curve G occurs at the end of the strain time, the length of time when strain is being increased.
The two curves in Fig. 29 show that a delayed or gradual strain does not create as high a stress as does an instantaneous
strain. The maximum stress on curve G will
be an increasing percentage of the maximum stress on curve I as the strain time is decreased. H the strain time is 30 days or more, the percentage will be about 50.
An example of stress relaxation of a concrete slab appears in Fig. 30. Discussion of two types of movements follows. They both involve cases where part of a structure
is displaced, then that part is prevented from further movement, after which there are changing strains and movements in other parts of the structure.
First, consider that the right support beam is displaced downward instantaneously by an outside force. The induced shear in the slab takes the shape of curve I in Fig.29, and the shear created by the displacement relaxes as
the concrete creeps, adjusting and relaxing into its new position.
If the dead and live loads are included, the net shear in the slab at the right support would vary as in Fig. 31a, because the total shear is immediately reduced, and then
as creep (plastic flow) develops, the concrete slab relaxes, tending to resume its originally suspended configuration; so the shear increases with time.
Second, consider that the right support beam is gradually displaced downward by an outside influence during the strain time. The induced shear in the slab takes the
shape of curve G in Fig. 29 because the stress increases gradually, but not in a straight line, increasing less gradually
as creep develops until the end of the strain time when the stress starts to decrease and approaches curve I.
If the dead and live loads are included, the net shear in the slab would vary as in Fig. 31b, because the shear drops gradually as the support drops; but at the end of the
strain time when displacement of the support ceases, creep still continues-the sustained loads cause time dependent
deformations in the form of bending changes
in the slab-resulting in an increasing reaction at the right support until the creep stops.
Relaxation of stress also occurs in prestressed concrete. It has six causes: (a) anchorage seating loss, (b) elastic shortening of concrete, (c) creep of concrete, (d) shrinkage of concrete, (e) relaxation of tendon stress, and (f) friction loss due to intended or unintended curvature in post-tensioning tendons. A report of American Concrete Institute Committee 423 in 1958 stated lump sum losses of 35,000 psi (241 MPa) for pretensioning and 25,000 psi
(172 MPa) for post-tensioning, and these generally give satisfactory results for many applications.
To quote from Commentary on Building Code Requirements. for Reinforced Concrete, ACI 318-77:(33)
Actual losses, greater or smaller than the computed values, have little effect on the design strength of the member, but affect service load behavior (deflections, camber, cracking load) and connections. At service loads, overestimation of prestress losses can be almost as detrimental as underestimation, since the former can result in excessive camber and horizontal movement. Data have been assembled and analyzed to permit computation of the stress loss due to relaxation of tendons composed of stress-relieved wires. Subsequent work on stress-relieved strand conforming to ASTM A416 indicates relaxation losses of about the same magnitude.
Stabilized strand or wire has smaller relaxation losses than conventional stress-relieved tendons. While the strand is at the elevated temperature used for the stress-relieving operation, it is subjected to a high tensile force which produces a specific amount of permanent elongation, thus resulting in low relaxation losses after the tendon is put into service. For specific relaxation values of a particular steel, the engineer should consult the steel manufacturer.
RESTRAINED EXPANSION AND CONTRACTION
Volume changes that are restrained may create large stresses.
When nonuniform temperature or moisture conditions are created within concrete and it is unable to adjust to these variations, internal stresses exist (Fig. 12). In such a
case there is restraint. Where there is no restraint, there is no stress and no cracking. There is movement, but if it is not restrained it will do no structural harm. As part of
a building, though, unrestrained volume changes may result in leakage of moisture, heat, and sound.
External restraint of concrete often develops from friction with the ground. If a long concrete wall is cast without any joints and if after moist curing it is allowed to dry, it tends to shorten but the friction of its footing on
the ground restrains the shrinkage. As drying progresses, tensile stresses build up in the wall, and when they equal the tensile strength of the concrete, the wall cracks. A
uniform temperature is assumed in the wall.
This phenomenon is illustrated by Fig. 32 for concrete cured 7 days under laboratory conditions and then allowed to dry at a uniform temperature with the ends of
the wall fully restrained from moving. The curve for drying stress peaks out much higher than the curve for tensile strength, showing the certainty of cracking when the
concrete is highly restrained against shrinkage. The concrete wall cracked at the age of 40 days, but the time would vary with different concrete mixtures, different
aggregates, different wall thicknesses, and different ambient humidities. Perhaps more important is the duration of good curing conditions, without which the strength
curve will rise more slowly, cracking occur at a lower stress, and more cracks develop in a long wall.
Temperature changes likewise will crack concrete members that are highly restrained by another part of the structure or by ground friction. Consider a long cantilevered
concrete canopy cast without joints that after
moist curing is allowed to drop in temperature. As the temperature drops, the canopy tends to shorten, but it is
restrained longitudinally from doing so by the interior of the building and the resulting forces cause the concrete to crack. Since tensile strength and modulus of elasticity
of concrete both may be assumed proportional to the square root of the concrete strength, calculations show
(Fig. 33) that a temperature drop of a certain number of degrees (depending upon type of aggregate) will crack the concrete regardless of its age, provided that the coefficient
of expansion does not vary with the temperature.
However, in Fig. 7 it is shown that the coefficient may vary considerably. At lower temperatures a larger temperature
drop, perhaps 30°F (17°C), would be required to crack the canopy. At higher temperatures, perhaps only a 20°F (11°C) temperature drop would crack the canopy. The same principles apply to any slab, beam, or wall that is prevented from shortening.
Another instance of restrained temperature expansion or contraction occurs during winter construction. Assume that a structural concrete floor in a multi-storey building is cast in cold weather, and the floor temperature is maintained at a minimum of 55°F (13°C). The lower floors of the building, however, are exposed to weather that might be well below the freezing mark. The 55°F temperature is supplied for several days, and then the temporary heat is stopped. The temperature drop that follows may cause a measurable change in length and can be easily calculated. Such changes in length of floors and roofs during the winter construction season cause secondary
effects, but current design practice does not treat this problem separately for ordinary structures.
For a slab resting on the ground, Fig. 34 illustrates the subgrade drag formula for stresses due to shortening caused by temperature or drying shrinkage.(40) In this case the slab is not fully restrained against moving. The friction force is calculated and equated to the tensile stress in the concrete or reinforcement.
Example problem 1 shows the common use of the formula-that is, to determine the amount of reinforcement to be used. Reinforcing steel or mesh will not stop a crack, but it can hold it together so that it will not be so wide and conspicuous. The coefficient of subgrade friction may vary from 1 to 2.5 and 1.5 is a commonly used value.
Example problem 2 in Fig. 34 is for a slab with no steel and determines the spacing of cracks that might develop if the slab were protected until an age of 28 days from drying
shrinkage or temperature contraction. It is based on concentric axial contraction and will not explain the common crack spacing in plain concrete floors and exterior slabs. An explanation of this cracking follows:
During the first night after placing an unjointed plain concrete slab, the temperature of the concrete declines
due to both reduced hydration activity and lower air temperature. Contraction of the slab from these sources is resisted by subgrade friction and tensile stresses are
created that cause transverse cracks. Spacing of the initial cracks varies from about 40 to 150 ft (12 to 46 m) depending on concrete properties, point-to-point variations in
subgrade friction, and climatic conditions during and after placement.
After the concrete floor or slab hardens, stresses are created by temperature and moisture gradients in the concrete. At and near the bottom of the slab, daily changes
in temperature and moisture content are small. The exposed top surface, however, undergoes fairly large daily variations in temperature and moisture. At night the top
of the slab is usually cooler than the bottom, thus the top tends to contract and curl the slab edges upward. This tendency is resisted by the weight of the slab, creating
tensile stresses in the top of the slab and compressive stresses in the bottom. During the day, the stress pattern is reversed.
Differences in moisture content between the top and bottom of a slab produce similar but less severe stresses lowers moisture contents cause contraction and higher moisture contents cause expansion. The influence of
restrained curling stresses on joint design is complicated because moisture and temperature differences often produce
opposite effects. When the top of the slab is warmer than the bottom, causing the top to expand, the bottom of the slab will usually have a higher moisture content causing it to expand. Hence, the amount of restrained
curling stress will be less than stress due to temperature differences alone.
Because of these opposing factors and others, curling stresses computed from formulas that take account only of temperature gradients are higher than actual measured values. Curling stresses measured on one research project were only one-half the values computed on the basis of temperature.
When other calculations of stresses (based on formulas that take account only of temperature gradient) are compared
to flexural strengths obtained at 8, 16,and 24 hours, indications are that there should be transverse cracks at spacings of 15 to 20 ft (5 to 6 m) or less during the first night of the concrete's life. Since this cracking does not
normally happen, it is another demonstration that calculations based on temperature gradients alone do not produce
values that agree with field experience.
In plain floors with joint spacings of 15 to 25 ft (5 to 8 m), cracks do not generally form beneath all joints for a few weeks to several months after the floor is opened to traffic. In floors where distributed steel is used between joints spaced at 40 ft (12 m) or more, intermediate transverse cracks between joints may not develop for several
months to several years after the floor slabs are opened to traffic. When intermediate cracks do occur, they are spaced about 15 to 25 ft (5 to 8 m) apart; and they are the
result of the combined effect of restrained curling and load stresses.
Restrained curling is complex and repetitive loads compound the problem. Performance shows that restrained curling in combination with loads will cause additional transverse cracks between the initial contraction
cracks. The interval between transverse cracks is normally about 15 to 30 ft (5 to 9 m), depending on factors such as shrinkage properties of the concrete, subbase
and subgrade conditions, and climatic conditions.
Chapter 3 - Movements Due To Special Loading Conditions
Unexpected and unconsidered loads also cause building movements, as do certain construction procedures. Building design and construction would be easier if these factors could be fully anticipated, but they cannot because buildings and construction procedures involve many variables.
The design loads for buildings include applied dead, live, wind, and earthquake loads. Loads that occur during construction are sometimes overlooked in design, but they have a bearing on how the building should be built and how it will perform. Errors made in the design and construction process will affect movements, strains, volume changes, stresses, and cracking later on.
Structural engineers are very conscious of the deflections of beams and slabs under certain loads. But they are less conscious of the fluctuations of certain design loads, fluctuations that cause movements where none may be expected.
DEAD LOADS
Dead loads can generally be closely determined and are fairly constant in buildings. There are cases, though, where dead loads may be more or less than the estimated values, and the discrepancies can add to other movements. One example is when a floor topping is superimposed
on a structural floor. If the floor topping is finished with its surface dead level, it will be thicker at midspan because of the initial deflection of the floor plus the added deflection caused by the topping. Assume a 1-in. (25-mm) topping is planned for a one-way structural floor with a 30-ft (9-m) span. The dead-load deflection of that structural floor might be l/2 in. (12 mm). With the weight of the topping, the floor might deflect another 1/16 in. ( 1.5 mm); and altogether, the topping will be 59% thicker at midspan than planned. This added thickness is where it will add the most to bending stresses.
Another example of variable dead load is the weight of the concrete. Published weights for concrete with 3/4-in. (19-mm) coarse aggregate and 6% air entrainment range from 137 to 145 pcf (2192° to 2320 kg/m3), and without deliberate air entrainment, these values could increase to 152 pcf (2435 kg/m3), not including reinforcing steel. Thus there is a range of l l % that can affect the deflections, stresses, and cracking of structures.
The weights of concrete masonry walls are usually based on values calculated at a standard section of the wall and do not take into account extra mortar that oozes from the joints into the cells, mortar thrown into the cells, door and window jambs slushed with mortar, the thick mortar bed in the first course, and so on. Masonry contractors usually order three to five times as much mortar as the typical joint requires, and some of this excess ends up in the walls. The weight of the masonry units is usually estimated by the structural designer. There are no standard weights for concrete masonry units; each producer decides the density of the concrete and the interior dimensions of the mold for his masonry products.
The dead loads of partitions in buildings are usually approximated in pounds per square foot (kilograms per square meter) of floor. To calculate accurately the weight of each partition and take account of its exact position on a floor would overtax structural engineers. Admittedly the effects on the movement of structures have a tolerance for error.
When large-scale alterations are made in buildings, there is a readjustment in the shape of the structure. The amount of movement could only be predicted through
an exhaustive study because so many imponderables are involved.
LIVE LOADS
Live loads can only be predicted within certain ranges, and they vary every day, especially in warehouses (Fig. 35) where they may have a real influence on the cracking of the floor. The shrinkage force in a slab on ground is usually calculated using the weight of the slabs alone. However, as loads are added to the slabs, shrinkage stresses due to ground friction increase proportionately.* For example, a 6-in.-thick (152-mm) slab 20 ft (6 m) long may by itself develop a friction force for which the unit stress would be only 16 psi (110 kPa), certainly not a significant tensile stress. However, if a warehouse load of 500 psf (2400 kg/m2) is superimposed, the friction force is multiplied greatly. The increased stress, added to the nonuniform stresses from temperature, moisture, and flexure due to gravity loads might be more likely to precipitate cracking.
Fig. 36 shows how average live load varied on each floor of an 11-story office building. At the time of the survey there was only a 13 psf (63 kg/m2) difference between maximum and minimum averages. However, the floors had an allowable live load of 70 psf (340 kg/m2), about six times the average load actually present, and there was nothing to prevent any tenant from increasing the loads to that level or more at any time. Increasing the live load to the maximum allowable might increase the existing deflections by 50%, or say 0.20 in. (5 mm), which will easily cause cracking in partitions unless they are built in a manner to accommodate such movements. The state of the art in the design of tall buildings requires that a degree of freedom should be allowed for the partitions to adjust to the movement of the structure.
Table 5 shows how the live load actually varied in another office building, not only by floor but within each floor. The live load was generally well under the allowable load; but on some floors the loads did vary widely and it is there that deflections could cause partition cracking. As the live load varies, deflection increases or decreases, not only from elastic strain but also from creep. As deflection varies, movements in partitions vary; and if the range is wide, there may be doors sticking, door locks not engaging, and cracks forming.
In a multistory building the changes in live load will cause different amounts of column shortening. Measurements of column shortening in one tall concrete building amounted to 3/16 in. (4.8 mm) for every 10 ft (3.05 m) of height during the first three years after construction. This movement could be enough to cause a visible crack in a partition.
Other variable live loads are heavy snows on roofs, auditorium crowds, water levels in tanks and pools, remodelling that changes the usage for a floor, and equipment
changes. The ponding of rain on sagging roofs has received great attention in recent years, to the extent that the phenomenon is now recognized in some building codes.
Water pressure has been known to buckle basement floors. This happens when an ordinary slab on ground has not been designed for hydraulic uplift. With or without a membrane under the floor, a rapid rise in ground water levels may not be relieved by seepage through temperature and shrinkage cracks. Better drainage under the
5 floor and at its perimeter will help avoid water pressure if the drains themselves do not become overloaded.CONSTRUCTION LOADS
During construction, stockpiles of materials may exceed design loads and so cause deformations or structural movements that will leave their marks. These are transitory
loads, but their effects may not be. For example, if one floor is heavily loaded with construction materials, the deflection recovery of the floor may occur after the partitions in a lower floor have been installed and thus upset the tight fit between floor levels.
Heavy bundles of reinforcing bars are usually lifted by crane onto a formwork deck where they may create excessive concentrated loads at the base of the form work on the new concrete floors below, especially if there is a three day construction cycle per story. The same effect results when a bucket of fresh concrete is emptied into the forms. The American National Standards Institute (ANSI) Standard
A !0.9, Safety Requirements for Concrete Construction and Masonry Work, specifies the total live and dead load to be used for formwork design.WIND AND EARTHQUAKE
Wind is the best recognized source of movements in buildings, but for typical low-rise concrete frame buildings the problem is minimal. The perception of wind movements can be divided into two parts: the emotional impact on the occupants and the impact on partitions, veneers, and mechanical equipment.
A common rule is that horizontal wind movement or drift should not exceed 1/500 of the height of the building. One five-hundredth of a 10-ft-high (3-m) story is about 0.25 in. (6 mm). This rule is based on the stability of the structure as a whole. Studies have shown that tenants will not perceive wind movements when this restriction is observed in design.
Concrete buildings, as usually built, have an inherent stiffness and low damping coefficient so that wind movements are very small. Fig. 37 shows lateral displacements that may occur at each story with two framing systems, illustrating the effect they may have on walls, windows, doors, and so on. As might be expected, the movement per story for a frame only is greater in the lower stories than in those above, because of the accumulation of loads. The opposite can be true for shear walls because of more reinforcement in the lower stories.
A study has been made of the wind drift of high-rise concrete office buildings. There is very limited data on actual field measurements of concrete buildings, but calculations bear out the lack of tenant complaints because of the stiffness of the buildings. Of 21 buildings studied (43) the greatest lateral movement was 1/740 of the total height. In general the computed drift ranged from 1/1000 to 1/2500 of the total height, depending on the slenderness of the structure.
Earthquakes have caused the most severe movements of buildings and damage to structural elements, partitions, and architectural features. Ground displacements of earthquakes are measured in inches. The actual displacement per story height is on the order of 0.75 to 1.0 in. (18 to 25 mm), though calculated values are usually much less. Modern building codes (44) limit the calculated drift to 1/200 of the story height, unless it can be demonstrated that greater drift can be tolerated.
The intensity of earthquakes is measured by the horizontal acceleration of the ground and values may reach 50% of gravity acceleration. In addition, damage from earthquakes depends on the duration of the large-amplitude acceleration pulses, the frequency of the ground motion, and the dynamic properties of the structure. (45)
The seismic risk maps of the United States (Figs. 38 and 39) are based on the severity of ground motion and divide the United States into Zones 0, l, 2, 3, and 4. Fig. 40 is a corresponding map for Canada.
FLOOR VIBRATIONS
Cast-in-place concrete floors in office buildings are not prone to objectionable vibrations though they have been known to develop. Long spans of 40 to 60 ft ( 12 to 18 m), characteristic of prestressed or precast construction, are more likely to develop noticeable vibrations, depending upon the type of occupancy. In a factory, strength to resist vibrations could be a consideration.
The perception threshold for vibrations has been reported at different levels by different authors. Values are in the range of a one vibration per second at 0.001-in. (0.025-mm) displacement amplitude to 20 vibrations per second with zero amplitude. At the same time, the peak acceleration is something less than 0.01 g.(43)
One of the more serious cases of floor vibrations developed in a precast concrete stadium with simply supported floor units weighing 100 psf (490 kg/m2). During a rock concert, there was enthusiastic clapping and stomping. Resonant vibration developed and a peak acceleration of 34% of gravity was measured. For special structures where crowds assemble, the natural frequency should be calculated and compared with published data. (49 50)
Chapter 4 - Movements Due to Foundations, Construction Procedures, and Combined Effects
FOUNDATION MOVEMENTS
The welfare of any structure depends upon its foundation. A certain amount of foundation movement is to be expected in all buildings; and if all the movement is equal,
little harm is done. Where there is more settlement in one part than another, partitions will suffer, tenants will be aware of high spots in the floors, tables may not be level, and other maladies may develop to make tenants dissatisfied with the building. Careful attention to the weight of the building and its occupancy loads and help from experienced geotechnical engineers are needed to assure acceptable foundation movements. Several basic types of foundation situations are illustrated in Fig. 41:
a. Many one- and two-story buildings are constructed without the benefit of soil borings and it is later discovered that they are resting on two or more different
types of soil.
b. Sloping sites may require cutting and filling. The fill must be well compacted so that the foundation and floor will settle no more there than at the cut end.
Even if deep foundations are used, the floor needs support; if it does not rest on a well-compacted fill, it may have to be built as a supported or structural floor. This would be the case for a building to be located over an old dumping ground.
c. When groundwater levels drop, uneven settlements occur. Fast-growing trees are known to desiccate the soil, causing settlements at one side or corner of a
small building. The process may develop over a period of 5, 10, or 20 years; and if the trees are removed, the soil may take another 5 or 10 years to rebound. (51)
d. Another case of settlement is from mining subsidence, where the ground surface may settle while mining occurs below. Structures in the Midlands of England are designed to anticipate subsidence. In one case there, a school settled two ft (0.7 m) while the subsidence occurred in wave fashion, but the
teachers and students in the school were not aware of any movement.
A foundation problem not shown in Fig. 41 is that of expansive soils taking on water and causing considerable heaving in small buildings. In house construction, differential
heaving up to 6 in. (150 mm) is not uncommon. In small concrete buildings the hazard would be to slabs on ground and would result in unsightly cracks in floors
and partitions and disruption of doors and plumbing.
While all clays undergo a certain amount of expansion and contraction during wetting and drying, those with higher plasticity indexes are most troublesome. The percentage of clay in the soil is also important. Soils with a plasticity index of 30 and a clay fraction above 30% are
known to cause serious problems. (52)
There are three general methods of coping with expansive soils: (1) Excavate the soil or build around it, seat foundations below the bad soil and elevate structural floors and grade beams so that they will not be affected
by any soil expansion. (2) Use a mat foundation on expansive soil or a ribbed waffle slab. These should support the entire structure rigidly despite nonuniform support, and lift or tilt it in one plane. (3) Articulate the structure so that relative movement and damages will occur only at predetermined panels or joints.
Frost Heave
While it is commonplace to build foundations below the frost line, they will sometimes be heaved by frost for these
reasons: (1) During a very cold winter with little snow to insulate the ground, frost will penetrate much deeper than the building codes recognize. (2) During construction
when shallow footings in the interior of the building are exposed to subfreezing weather, they can suffer frost heave. (3) When foundations for a part of the building
are outside the heated area, frost will penetrate deeper there than at the walls. These footings should be placed deeper.
What is not common knowledge is a phenomenon called ad freeze. This is the adhesion of freezing ground to a pile,
pier, or wall with subsequent heaving. Thus, "placing footings below the depth of frost penetration does not protect foundation structures from heaving unless adfreezing
of the soil to the structure is prevented or the
load on the structure exceeds maximum uplift forces. "(53)
An adfreeze stress of 25 psi (173 kPa) on a nominal 12-in.-diameter (300-mm) pile has been found by field tests. On an 8-in. (200-mm) concrete masonry wall, the adfreeze
stress was 3.5 psi (25 kPa). Simultaneous tests on a 12-in.diameter (300-mm) horizontal plate anchored rigidly at
the ground surface yielded a heaving pressure well above 21,000 psf (1000 kPa) during the winter when the average
frost heave for the site was 3.5 in. (90 mm). This pressure could readily lift many buildings and clearly demonstrates that frost penetration cannot be permitted in frost-susceptible soils where heaving cannot be tolerated. Frost-susceptible soils are those with a high percentage of silt that do not drain well. If water can continue to enter the soil, there will be a buildup of ice lenses with increasing amounts of heaving. Without physical restraint, there is no apparent limit to the amount of heaving that can occur.
Frost heave can also develop in cold storage warehouses and under artificial ice rinks if proper precautions are not taken. Fig. 42 is a plan for an indoor rink showing its cracking. The rink was well constructed with 3 ft
(1 m) of drained pea gravel for a subgrade. Every summer the rink was shut down and thawed for several months for use as a roller skating rink. However, after six years of
use and while the slab was still in excellent condition, the ice rink was used through the summer because of its popularity. The subgrade froze very deeply and the cracking
shown in Fig. 42 developed during a heave of 6 in. (150mm). The slab continued to heave another 2 in. (50 mm) before the freezer was shut down. Fortunately the slab contained a high percentage of reinforcement, which
helped to close the cracks as the slab settled back to near its original level.
The subject of foundation movements on permafrost is beyond the scope of this text.
CONSTRUCTION PROCEDURES
Movements are affected by various construction procedures such as reshoring, by the sequence of erecting various
features, and by prestressing.
Reshoring
A common method of reshoring is to lengthen the original shores or posts a fraction of an inch (a few millimeters).
This is done with a wrench or wedges and a reassuring pat to see if they are tight. Using this concept, a labourer or carpenter depresses the lower floor a certain
amount at the same time that they lift the upper floor to "take the load off it" until the concrete has gained sufficient strength to carry itself. Several tall buildings have
had severe failures because the reshores were not handled properly.
The reshoring operation is a crucial step in constructing multi-storey concrete buildings and has a far-reaching effect on the deflection of the floors as well as on their
strength. If the reshoring does not prevent or relieve the stresses in the new concrete, the creep effect will multiply the eventual deflection because, as has been shown, the
sooner a concrete slab is loaded, the greater will be its total deflection. If the reshoring is tightened too much, it may create harmful stresses in the new concrete floor and
it could throw excessive loads onto the lower floors and amplify the added creep deflection.
Thus it can be seen that there is a real need for some sort of test device to determine the loads in reshores. Also, a careful study is needed to determine how many lower floors can safely carry the loads, and for what
length of time.
Form removal may result in more flexural cracking than should develop. This in turn will alter the stiffness of a beam or slab, leading to even greater deflections.
Prestressing
Prestressing builds in compression stresses that are partially relieved by creep and drying shrinkage over a long period of time.
Prestressing concrete beams and slabs causes three movements that differ in amount and extent from those found in reinforced concrete construction: elastic and
creep strains that develop over the full length of the member (or the length of the tendons) and drying shrinkage. Drying shrinkage is slightly greater in prestressed concrete than in reinforced concrete because the tendons do not interfere with the shortening, whereas reinforcing bars resist the shrinkage of ordinary reinforced concrete. The ACI Building Code states: "Provisions shall be
made for effects on adjoining construction of elastic and plastic deformations, deflections, changes in length, and rotations due to prestressing. Effects of temperature and
shrinkage shall also be included."(32)
In some prestressed buildings excessive camber of floors has caused partition cracking and bowed floor surfaces.
The problem can be partly corrected by better positioning of the tendons, by close control of the concrete strength (and its modulus of elasticity), by control of the
moisture condition or rate of evaporation, and by proper timing of the prestressing operation.
For prestressing to be effective, the ends of the units, or floors, must be free to move. The concrete cannot be
prestressed unless it can shorten when the tendons are
tightened. This means that walls and columns away from
the center of movement must be flexible or the floor or
roof unit must be free to slide. Prestressed floors and
roofs also need flexible or sliding supports that will yield
to the very strong forces created during prestressing.
These forces are intended to go into the prestressed unit,
and supports that offer too much resistance may be severely
fractured.
Construction Sequence
Craftsmen and builders have known for centuries that
the sequence of installing or assembling the parts of a
building makes a difference. Movements may take place
that interfere with a tight fit. This is very apparent in remodeling
work where it is found that joints have opened,
beams have sagged, and walls have racked because of
foundation settlement. Wherever different materials
meet and wherever temperature and humidity conditions
change, the builder can expect to find movement.
The sequence of installing exterior masonry curtain
walls in a two-story building makes a difference in how
much load goes onto the ground story. Being a curtain
wall and not a bearing wall, it is not intended to support
much load. If the wall is built snug under the second floor
spandrel, it will have to take the weight of the wall above
and the floor live load near the wall (plus some roof live
load). To avoid this, many builders will lay up the wall in
the second story before the wall in the first story.
A buoyant basement will also have movement unless
the contractor first builds enough superstructure above
it before letting the water table rise around the basement.
On the other hand, to prevent flotation, earth or rock
anchors can be provided as they sometimes are for tanks
that are set into the ground.
In tall concrete buildings with a mixture of columns
and concrete walls, close attention must be paid to the
sequence of construction and to the time when the two
types of framing are connected. This is to minimize the
difference in axial shortening of columns and walls so
that beams and slabs will not be subject to large shearing
stresses.
COMBINED EFFECTS
Rarely does one type of volume change or movement
occur by itself. Several of various factors are apt to occur
simultaneously and cause excessive strains and cracking
or cause movements that are incompatible with adjacent
building materials and members.
As an example of the compounding of influences to the
detriment of a structure, consider the open post-tensioned
parking structure shown in Fig. 43. It has unusual
proportions as buildings go, and is exposed to severe
weather fluctuations. Its one-way concrete slabs spanning
about 20 ft (6 m) are supported by prestressed concrete
girders spanning about 60 ft (18 m). The floors in
adjoining bays ramp up so that the girders meet the columns
at staggered levels.
The columns between the ramps are usually quite short
and stiff because of the building proportions and because
they must resist large bending moments and shears. The
moments and shears in the columns are also increased by
frame action normal to the girders, particularly because
of volume changes.
Several deformation conditions of the column-girder
frame are illustrated in Fig. 43. The concrete is constructed
as in (a). Post-tensioning of the girders may
cause a camber of perhaps 0. 75 in. ( 19 mm) at midspan
and a shortening of the span as in (b). The dead load of
the floor is supported by the post-tensioning. A live load
causes a downward deflection of the girders and reverses
the bending of the frame (c). Creep and shrinkage may
again bow the girder upwards 0.75 in. (19 mm) at midspan
reducing bending in the column, but causing additional
shortening of the girders with additionai offsets
and bending of the columns (d). Cold weather will shorten
the girders and further offset and bend the columns (e).
Perhaps the worst condition exists during a very cold
night when there is no live load but some shrinkage and
creep have developed.
At the same time there may be additional strains due
to construction procedures and even some foundation
movements to compound and confound a design problem
that is by nature composed of approximations and
subject to compromises during construction. Conventional
building frames would usually accommodate these
cumulative effects. This unusual structure, however, with
long spans and short stiff columns is vulnerable to combined
effects. In one case significant cracking developed
in the columns and an investigation attributed it principally
to temperature variations.(55)
The combined effects of volume changes in the upper
parts of tall structures has a definite effect on the floors.
Differential shortening of columns will cause warping or
tilting of the slabs, and affect their structural performances.
This problem can be anticipated by careful design
of the structure and careful control of the construction
procedures. The matter is investigated in depth in Reference 54.
Another outstanding example of combined effects is
the stability of wall cladding on concrete frame buildings
(Fig. 44). Very expensive repairs have been necessary on
a number of buildings because provision was not made
for differential movement between the cladding and the
building frame, (56) and the cladding buckled. This is not
a new problem brought on by the advent of high-rise
buildings. The practice of providing an elastic joint under
the shelf angles of multistory buildings goes back many
years and recognizes that the volume changes of the
veneer and frame are different for several reasons, such as
- Elastic shortening of the columns
- Creep
- Shrinkage
- Temperature
- Moisture changes
- Wind effects
The combined effects of various movements are of real significance. If exhaustive, detailed calculations are made for moments, shears, and forces in buildings, it is just as important to study the effects of creep, shrinkage, temperature, and settlements. The difficulty, however, is that much of the technology is based only on laboratory coefficients. Even with elaborate computer facilities, the state of the art for the design and construction of most structures
does not lend itself to precise predictions of movements. Their control then must depend heavily on recognition of their magnitudes (Fig. 45) and a general allowance for them as proposed in the following chapters.
Part Two: Controlling Cracks in Concrete Structures
Chapter 5 -The Need for Crack Control
Because of concrete's low tensile strength, cracking in
some areas of reinforced concrete structures is inevitable.
Structural systems designed with low steel stresses under
service loads may have very limited cracking; and frequently
no cracks are visible. However, with high service-
load steel stresses, common with the use of highstrength
steel and high service loads, some cracking can
be expected at the service load. The cracking at service
loads should not spoil the appearance of the structure or
contribute to corrosion of the reinforcement.
ESTHETIC CONSIDERATIONS
The width of a crack that will not distract from the
appearance of a concrete member depends on the position,
length, width, and illumination of the crack and on
the surface texture of the concrete. The criteria for acceptability
are difficult to set because of differences in
personal opinion. The maximum width of a crack that
will neither impair the surface appearance nor alarm the
viewer is probably in the range 0.010 to 0.015 in. (0.25 to
0.38 mm), but wider cracks may be tolerable.
CAUSES OF CRACKING
Causes of cracking are numerous as stated in Part One,
but most cracks are a result of the following:
Settlement of Plastic Concrete
Cracks are sometimes observed on the top surface of
slabs, in beams, and over the stirrups and other top steel.
These cracks are initiated when the plastic concrete settles
away from the reinforcing bars that are supported
and tied in a fixed position. This type of cracking can be avoided by a good concrete mix design and revibration of
the plastic concrete.
Plastic Shrinkage
Plastic shrinkage cracks appear when water evaporates
from freshly placed concrete faster than the concrete can
bleed or water can rise naturally to the surface. These
cracks appear soon after the concrete has been placed and
while it is being finished. Plastic cracking is usually associated
with hot-weather concreting,* but it can occur at
any time when atmospheric conditions result in an excessive
evaporation rate. When excessive evaporation occurs-
especially in flatwork or where the freshly cast concrete
surface is exposed to the vicissitudes of atmospheric
conditions-rapid drying shrinkage and tensile stresses
in the surface often result in short, tearing cracks.
The following conditions, either singly or collectively,
increase the evaporation rate and the potential for plastic
shrinkage cracking:
- High concrete temperature
- Low humidity
- High winds
- Low ambient temperature
When these factors combine to result in an evaporation
rate that exceeds 0.2 lb per square foot per hour (I kg/
m2 /hr), precautionary measures must be taken to prevent
plastic shrinkage cracks and loss of strength in the
concrete near the surface. Cracking is possible even if the
evaporation rate exceeds only 0.1 lb per square foot per
hour (0.5 kg/m2/hr).
To minimize plastic shrinkage cracking:
- Increase relative humidity
- Reduce concrete temperature
- Reduce wind velocity
Following are specific simple precautions to minimize
the possibility of plastic shrinkage cracking. They should
be considered while planning for hot-weather concrete
construction or while dealing with the problem after construction
has started. Most of them are just good construction
practices that should be used at all times.
- Moisten the subgrade and forms.
- Moisten aggregates that are dry and absorptive.
- Erect temporary windbreaks to reduce wind velocity over the concrete surface.
- Erect temporary sunshades to reduce concrete surface temperatures.
- Keep the fresh concrete temperature low by cooling both aggregates and mixing water.
- Protect the concrete with temporary coverings, such as polyethylene sheeting, during any appreciable delay between placing and finishing.
- Reduce the time between placing and start of curing by eliminating construction delays.
- Protect the concrete immediately after finishing to minimize evaporation. This is most important to avoid checking and cracking. Application of moisture to the surface by fog spray is an effective means of preventing evaporation from the concrete. Fogging should be continued until a suitable curing material such as a curing compound, wet burlap, or curing paper can be applied.
If plastic shrinkage cracks should appear in the fresh
concrete, the cracks can be closed by striking each side
of the crack with a float. However, the cracking will recur
unless the causes are corrected.
Volume Change of Hardened Concrete
Initial drying shrinkage, moisture, and temperature
changes cause volumetric changes that will create tensile
stresses in the concrete if it is restrained, and will lead to
cracking. There are a number of ways that concrete is restrained
against volume change. For example, concrete
near the surface of a member dries and shrinks more rapidly
than the concrete deeper in the member; therefore
the inner concrete will restrain the shrinkage of the outer
concrete creating tensile stresses near the surface that
may cause surface cracking.
Also, shrinkage of structural slabs may be restrained
by their continuity with beams, columns, foundations,
and the presence of reinforcing steel. This restraint can
introduce tension and cause cracks at the junction of
those members with dissimilar ratios of surface area per
unit volume of concrete and at reentrant corners or blockouts
for holes. Temperature changes also will cause tension
in a similar manner if the movements of the member
are restrained.
Cracking due to initial drying shrinkage of the concrete
can be controlled by reducing the potential for
shrinkage with a good concrete mix design that keeps the
total water content as low as possible and by properly
placing the reinforcement. Reinforcement will not prevent
cracking. The restraint added by the reinforcement will tend to encourage it, but the shrinkage strains are distributed by bond along the reinforcing bars and a
number of fine cracks should occur instead of a few wide
cracks.
The minimum amount and spacing of reinforcement
that may be used in slabs and walls is stated in ACI 318,
Building Code Requirements for Reinforced Concrete.
This reinforcement is intended to be adequate to control
the widths of cracks caused by tensile stresses due to drying
shrinkage and temperature change.
Recommendations for the design of temperature reinforcement
usually are based on the total strain caused by
thermal movements (initial and seasonal) plus drying
shrinkage and on the strength of the concrete when mature,
that is, whenf~ equals or exceeds the specified 28-
day strength. This approach, however, overlooks the fact
that the concrete may already have cracked to an unacceptable
degree at an earlier age, particularly if curing
was inadequate.
Control joints purposefully placed in slabs and walls
are intended to eliminate unsightly, uncontrolled, unwanted
cracking in large expanses of concrete. Such
joints normally are grooves in the concrete to cause the
crack to occur there. Sawed joints are common in floors
on ground and in pavements while inserts forming rustications
are common in walls.
Flexural Stresses Resulting from Applied Loads or Reactions
Cracking will occur in the tension zone of members subjected
to flexure from external loads or reactions once
the modulus of rupture of the concrete (its ability to hold
itself together) is exceeded. There are structural design
procedures for dealing with flexural crack control.*
Corrosion
Corrosion of the reinforcing steel in concrete can lead
to volume changes and cause cracking, spalling, and rust
stains in the concrete. The reinforcement is usually protected
against corrosion by the high alkalinity of the concrete.
This corrosion-inhibiting property is reduced if
chemical agents capable of neutralizing the alkalinity
penetrate to the concrete surrounding the reinforcing
steel. Chlorides in deicing chemicals, seawater, admixtures,
and liquid wastes are extremely corrosive chemical
agents. The positioning of reinforcing bars too close
to the surface in a poor-quality, weak concrete of high
permeability is an invitation to corrosion.
The basic protection against corrosion of reinforcement
in concrete is an adequate thickness of protective
cover, a high-strength concrete of low permeability, and
a specified maximum allowance for crack widths.
Chapter 6 - Control Joints in Walls
All concrete and concrete masonry shrinks and swells upon loss and gain of moisture in much the same manner as wood, except to a lesser degree. These volume changes set up stresses of considerable magnitude in any structure because there is always some restraint against free movement between components. These stresses exist in all concrete
buildings and should be taken into account in the design, especially for concrete that will be permanently exposed to view. In walls of large areas it would be impossible to provide sufficient reinforcement to prevent cracking entirely. For all practical purposes, however,
unsightly cracks can be eliminated by controlling their location and making them so inconspicuous that they do not detract from the appearance of the building. This can be done effectively and inexpensively by providing control (contraction) joints in the walls at proper intervals.
Control joints in walls are made by fastening projecting strips opposite each other on the form sheathing to create narrow vertical grooves on the inside and outside of the concrete wall. Strips of wood, plastic, or a piece of noncorrosive metal can be used. The grooves reduce wall thickness at predetermined locations, forcing the cracking at those locations and thereby relieving the stress in the wall. The narrow groove on the outside of the wall is
filled with a nonstaining, concrete-color sealant to prevent penetration of moisture.
LOCATING CONTROL JOINTS
No exact rules can be stated for the location of control joints. Each job must be studied individually to determine where joints can be placed without endangering structural integrity. It has been demonstrated in practice that control joints should be not more than 20 ft (6.1 m) apart in exterior walls with frequent openings. In walls without openings the joint spacing may be a little greater but should never be more than 25 ft (7.6 m) to be most
effective, and it is desirable to have a joint within 10 or 15 ft (3 or 4 m) of a corner if possible. Joint spacing in any exposed cast-in-place interior walls should be identical to joint spacing in outside walls.
Openings in walls are logical locations for control joints. Where small openings are more than 20 ft ( 6.1 m) apart, at the first-story level there should be a joint in line with each jamb below the openings. Above the first-story openings, a single joint at the centerline of each opening will generally suffice, but joints on jamb lines are preferable if shear in the spandrel will permit. Joints should be provided in the solid wall between openings so the maximum
spacing will not be more than 25 ft (7.6 m). In the case of openings less than 20 ft (6.1 m) apart or where several windows are grouped together with narrow concrete mullions between them, joints should be provided as described above at not more than 20-ft (6.1-m) intervals (see Fig. 46).
Sometimes window openings are very wide. If such openings are separated by shallow spandrels and narrow piers so the wall becomes essentially a frame of columns and spandrels, control joints are not necessary. If such an opening is more or less isolated and the height of wall above the opening is at least one-fourth of the width of the opening, or if the spandrel can be satisfactorily designed as a cantilever, a control joint should be located at the center of the opening unless the opening is in the first story in which case there should be joints in line with the jambs below the opening.
Scuppers or other small holes through the walls should be in line with the centers or jambs of window openings where control joints are located. Where steel columns are embedded in the walls and may weaken the wall section even more than nearby openings,joints should be placed in the planes of the columns and at intermediate sections in line with openings if the columns are more than 25 ft (7 .6 m) apart. Bonding of the concrete to the steel columns
at control joints should be prevented by coating the columns with an asphaltic paint and no mesh or other reinforcing steel should pass through the joint. At such locations if there is a pilaster encasing the column, the wall reinforcement should extend into the pilaster as shown in Fig. 47.
Control joints should begin at the top of the wall footing and should extend on the outside of the wall to the top of the parapet, thence over the top of the parapet and down the back of the parapet to the reglet strip. On the inside face the joint extends from floor to ceiling as in Fig. 48. It has not been customary to provide control joints in the floor or roof slabs where they join the walls.
In buildings of several stories with a setback at one or more floors, the joints need not be continuous from one level to another, but may be offset at each roof line in order to locate the joints at the best sections in the respective walls.
CUTTING REINFORCEMENT AT CONTROL JOINTS
Except where needed for structural strength, one-half of the horizontal reinforcement should be stopped off or cut at control joints to further induce cracking at those sections.
When cutting the bars, they should be cut exactly on the line of the joint. It is especially important that there be no lapping of bars at control joints. At openings where control joints are located, extra reinforcement normally is provided to control cracking at the corners. This extra reinforcement should be stopped off at the joints or omitted altogether, depending on whether the control joints are at the centre or at the jambs of the openings. Fig. 49(a) shows the arrangement of reinforcement at an opening when the control joint is at the centre. The
extra bars at sill and head are stopped 2 in. (50 mm) from the joint. Fig. 49(b) shows a control joint at the centre above an opening and in line with both jambs below the opening. Under such conditions the arrangement at the head is the same as in Fig. 49(a) except that the extra vertical bars extend only 2 ft (0.6 m) below the head as well as above. The extra bars are not needed at the sill because the control joints are at the jambs. Fig. 49(c) shows an opening with control joints at the jambs above the head and below the sill so that none of the extra bars normally provided around openings are needed. Fig. 49(d) is similar to Fig. 49(a), but with diagonal bars as the additional corner reinforcement.
FORMING CONTROL JOINTS IN WALLS
Numerous ways have been devised for forming control joints, depending upon materials available and job conditions. Whatever method is used, the thickness of the wall section at the joint should be reduced at least 20% by the depth of the joint; and the sum of the depths of the inside and outside grooves should not be less than 2 in. (50 mm). Where the jointing of the plywood can be arranged so that it is not necessary to cut the plywood, a detail such
as shown in Fig. 50(a) or (b) is satisfactory. These details are also satisfactory where board sheathing is used if there is not too much waste entailed from cutting the boards.
Where it is not feasible to cut the sheathing, the strips forming the joints must be fastened to the surface of the sheathing in a way that will hold them securely against impact from placing the concrete and from vibrators. Fig. 50(c) shows a satisfactory method of attaching a
wood strip to the sheathing if the joint is to be located in a flute. If a soft, close-grained wood is used for the strip it can be nailed to the sheathing as shown in Fig. 50(d), and it will remain in proper position during the placing of concrete if reasonable care is used in consolidating the concrete to avoid hitting the strip a hard blow from one side.
A method for inducing cracks in thick walls is shown in Fig. 51.
At window sills and the tops of parapets it is advisable to seal the control joints to be sure they will not leak. A line across the sill or parapet in the pla1;1e of the control joint should be tooled and filled with calking compound. Only sealants definitely known to be nonstaining, to retain their plasticity, and to adhere tightly to the sides of the joint should be used.
JOINTS IN TlLTUP WALLS
Their performance over the past 5 to 10 years has demonstrated conclusively that concrete tiltup wall panels can completely fulfil their structural function without being rigidly tied together. By keeping the panels separated, which is normal practice in nonearthquake zones, each panel is unrestrained and free to accommodate movements without cracking from drying shrinkage or thermal expansion and contraction. Preventing independent panel movement by interlocking the panels with cast-inplace columns or pilasters or by welding the panels together at metal inserts can be the cause of unsightly cracks and leaks in many wall panels. When all the wall panels are locked together any movement in length is restrained and cracks are inevitable.
Two systems that have been successfully used in earthquake zones for controlling cracking (there may be other variations) are described here. Both systems can be designed to withstand wind and earthquake lateral forces as well as light or heavy roof loads.
In the first system (in addition to the regular horizontal and vertical reinforcement) the outer one-third of the top chord reinforcing steel bars are enclosed in cardboard sleeves or plastic tubes (Fig. 52). The panel is then free to move from either end toward the middle without being restrained by bonding to this reinforcement. A blockout at each end of the reinforcement permits welding it to a steel angle, which then provides a continuous chord
around the top perimeter of the structure. This continuous chord is essential for the roof to act as a horizontal diaphragm to transmit lateral loads to the shearwall elements in the building. The wood ledger is not continuous and cannot serve as a chord member for the horizontal roof diaphragm.
In the second system, a steel channel (or angle) is used as a horizontal ledger to support the roof framing members (Fig. 53). This ledger serves the same structural purpose as the chord reinforcing steel used in the first system. The steel channel is rigidly attached in the centre
portion of the wall panel by welded stud connectors or similar devices embedded in the concrete. Slotted holes kept completely free of cement paste are provided in the end portions of the channel through which the ledger is bolted to inserts in the wall panel. The rigid connection at midlength of the steel channel must transfer in-plane diaphragm forces. The bolts in the slotted holes transfer vertical and transverse loads from roof to wall. As in the first system, the wall panel is free to move from either end toward the middle without restraint. At the joint between panels, the adjacent steel channels are rigidly connected by welding on a steel splice plate to provide a continuous chord for the horizontal diaphragm action of the roof.
In some tiltup buildings, a floor closure strip of concrete 3 to 4 ft (0.9 to 1.2 m) wide is placed around the perimeter of the slab to lock the walls to the floor slab by embedding the reinforcing steel protruding from both elements. Weakened-plane control (contraction) joints should be placed in this closure strip opposite the joints in the wall panels to reduce the incidence of cracking in the floors.
ISOLATED FOOTINGS WITH TILTUP WALLS
Although continuous footings are more commonly used to support tiltup walls, some walls are designed as deep beams and are supported at each end on isolated footings. Shrinkage and thermal movements then tend to produce diagonal cracks as shown in Fig. 54(a). The use of an elastomeric pad between the isolated footing and one end of the panel as shown in Fig. 54(b) has been effective in reducing this cracking.
Chapter 7 - Expansion Joints in Concrete Buildings
Most buildings of simple rectangular shape that are relatively short-200 to 300 ft (60 to 90 m) in length-do not require expansion joints. If, however, it is decided that the extent of movement for the whole building could cause cracking in the structural frame, consideration
should be given to the possible need for an expansion joint or joints at some point in the building.
The purpose of expansion (isolation) joints in reinforced concrete buildings is to permit the separate segments of the structural frame to expand and contract with temperature and moisture changes without adversely affecting the building's structural integrity or serviceability. If the structural frame is free to expand and contract, no stress capable of cracking the concrete will develop from such movement; in actual practice restraint is present in all structural frames. For the most part buildings of ordinary size and regular in plan can be designed to resist the stresses caused by volume change without recourse to expansion joints, but control joints should be provided in all buildings as described in the chapters "Control Joints in Walls" and "Joints in Slabs on Ground." Under certain conditions of size, shape, and plan, however, it is desirable to provide joints that actually separate ( or isolate) a building into independent units so that the stresses developed will not cause damage detrimental to the structure's utility or appearance. {The movement of masonry buildings is explained in a separate chapter.)
Differential foundation movement and dimensional changes due to applied load are added factors that influence the need for expansion joints, so definite rules cannot be established as to either their size or location. Consideration of the causes of volume change and observation of buildings in service should guide the designer.
CAUSES OF VOLUME CHANGES
Moisture Changes
In general, most buildings continue to dry out throughout their lives, so the volume change due to variation in moisture content is shrinkage only. Rain on the walls of a building, unless of very long duration, will not cause an appreciable gain in moisture content of the concrete.
There may be a slight gain or loss of moisture because of seasonal changes in atmospheric humidity, but for practical purposes concrete in buildings may be assumed to shrink and not swell.
Temperature Changes
Temperature variations cause buildings to expand and contract, on a daily cycle or a seasonal one. The expansion and contraction of a building from any cause is restrained and affected by many conditions and therefore cannot be computed accurately. A rough indication of the amount of movement caused by temperature changes is obtained by multiplying the average thermal coefficient of expansion of concrete (0.0000055 to 0.000006) by the length of the structure and by the degrees change in temperature. In other words, movement that might be be anticipated at expansion joints in a building divided into 200-ft (61-m) sections for a 25°F {I3.9°C) change in temperature will be on the order of ¾ in. {9.5 mm). If the 200-ft (61-m) wall section is exposed to a thermal variation from summer to winter of 50° F (27.8° C), the total thermal movement will be about 0. 720 in. or ¾ in. (19 mm).
DRYING SHRINKAGE EFFECT ON TOTAL MOVEMENT
The shrinkage of concrete due to loss of moisture varies with the concrete mixture and with atmospheric conditions. In actual practice, because of incomplete drying and the restraint offered by reinforcement, the shrinkage that occurs is materially less than that determined by laboratory experiments on plain concrete. A typical coefficient of shrinkage determined in the laboratory may be about 600 millionths.
Thus, for a 100-ft-long (30.5-m) unrestrained wall, the contraction could be about ¾ in. (19 mm). The shortening also varies with the mass of the concrete. Mass is not usually considered in calculating expansion-joint movement and partially explains why observations of buildings in service indicate the total movement at less than half that anticipated by combining shrinkage with the contraction due to temperature drop. The restraining effect of the reinforcing steel also and sructural framing also play a major role in the reduction of this movement. It appears safe to consider that the maximum movement at joints located 200 ft (61 m) apart will not exceed l in. (25 mm) under most unfavorable conditions.
FOUNDATION MOVEMENTS
Major differences in foundation loading at different parts of a building can produce a tendency for significant parts of the building to move in relation to each other. This kind of movement will determine the need for, and location of, major joints dividing the building into blocks. Each block can move as a separate unit, Fig. 55(a); or foundation movement can
be prevented by founding the building on an uncompressible material, Fig. 55(b); or the added stress can be relieved by the weight of the earth removed from a basement excavation, Fig. 55(c).
SPACING EXPANSION JOINTS
Buildings under 200 ft (61 rn) long are seldom provided with expansion joints.Buildings of more than 600 ft (183 m) have been constructed and performed satisfactorily without expansion joints.
The possible need for thermal expansion joints in long buildings may be determined initially using the empirical approach described in the following section.
Previously developed empirical rules for expansion joint spacing are not necessarily compatible with modern construction. Therefore, effects of thermal and other volume changes should be determined as part of the structural analysis. If results of the empirical approach indicate an expansion joint may be needed, a more comprehensive analysis can be done to determine if use of expansion joints can be avoided.
As a minimum, each of the following factors should be taken into account
for expansion joint location and design:
- Dimensions and configuration of the building
- Design temperature change
- Provisions for temperature control
- Type of structural frame, connection to the foundation, and symmetry of stiffness against lateral displacement
- Materials of construction
Empirical Approach for Determining Need
The following criteria taken from Reference 58 may be used to determine if a more comprehensive analysis is needed as described, in Section III B of Reference 58:
a. For buildings with a beam-and-column or slab-and column structural frame,* the maximum length of the building** without expansion joints should be determined in accordance with Fig. 56 on the basis of the design temperature change (Llt) in the locality
of construction.
b. For buildings supported by continuous exterior reinforced masonry, expansion joints should be placed at intervals not exceeding 200 ft (61 m). In addition, intermediate subjoints should be positioned and spaced in accordance with the recommendations of the Brick Institute of America and the National Concrete Masonry Association.
In order to be effective, expansion joints should extend entirely through the building, completely separating it into independent units. Column footings that are located at expansion joints need not be cut through unless differential settlements or other foundation movements are anticipated; and in that case, separate smaller buildings should be built on separate foundations to allow for those movements. Expansion joints should be carried down through foundation walls; otherwise the restraining influence of the wall below grade, without a joint, may cause the wall above to crack in spite of its joint. Reinforcement must never pass through an expansion joint.
The architectural treatment of the exterior walls will determine to some extent where joints can be located without marring the appearance of the building. It is advisable to place expansion joints where they can be obscured by a pilaster or other architectural detail. When
so located two purposes are served: the joint is less conspicuous, if visible at all, and it is easier to make the joint weathertight because of the protection afforded.
Expansion joints should always be made as simple as possible without sacrificing effectiveness. They must be designed to move freely. Care should be taken in detailing
and installation to avoid possible damage to the building or to the joint itself through failure to operate properly. The flow of wind and water through the joint must be stopped by weatherproofing. An example of a simple, weathertight joint is shown in Fig. 57. ln this joint, (1)the ribs deflect rainwater away from the joint, and (2) the air seal forces any water that leaks into the joint to run down within it and outside through an open horizontal joint below.
Many gasket and sealant products are available to maintain a complete seal while accommodating movement of the expansion joint. Success will be realized when the joint detail includes all the features of the joint in Fig. 57.
Expansion joints usually are shown fully open on construction drawings. The temperature at the time of joint construction, however, will determine the width of the joint opening. Allowances for temperature variations as shown on the drawings must be made when orming the joint. The actual width of an expansion joint must be greater than the computed maximum joint closing to provide for construction tolerances and for the width and
compressibility or expandability of the joint sealant and compression seal.
Expansion joints in structural floors or suspended slabs must be designed to prevent water from leaking to the floor below, and to provide for a smooth traffic surface. Water seals should be included unless it is definitely known that there will be no water on the floor. Sliding plates flush with the floor surface will interfere least with traffic. Thresholds that project above the floor surface except in doorways are objectionable. An example of a floor joint is shown in Fig. 58.
The roof at an expansion joint is always a critical place because of the possibility of leakage. Details such as are illustrated in Fig. 59 are positive in action and do not depend on the integrity of the roofing for success. Extruded- metal and sheet-metal expansion cover assemblies also are available.
Wall and roof joints must be continuous over parapets. The water seal or flashing must be continuous so there will be no place for leakage. Various modifications can be adapted to parapets and copings of different types. If the computed expansion joint width exceeds 2 in. (50 mm), special consideration should be given to the materials and methods of joint construction to ensure that the joint itself will be able to withstand the distress caused by such a substantial movement.
Expansion joint design should permit fully free movement of the abutting building segments, prevent the entrance of water or debris, and allow for easy maintenance and inspection.
CLOSURE STRIPS
Instead of joints that may be difficult to install because of structural requirements for reinforcing, some designers use closure strips in concrete buildings. These extend
completely through floors and walls and are about 3 to 4 ft (1 m) wide. Concrete for the building component is placed on each side of this closure strip and after proper curing is allowed to air dry and shrink as long as practical-at least 30 days is suggested. The normally required reinforcing steel protrudes from each side into the closure strip where it is suitably spliced. The slabs on each side of the closure strip are free to contract independently. At the
end of the waiting period, concrete is placed in the closure strips to complete the floors and walls of the building, as shown in Fig. 60.
Chapter 8 - Construction Joints in Wall
A true construction joint is the surface where two successive placements of concrete meet, across which it is desirable to develop and maintain bond between the two placements and through which any reinforcement is not interrupted. Stopping places where the process of placing concrete is suspended for 30 minutes or more are called cold joints and should be handled as construction joints. A true construction joint provides full restraint against movement.
Construction joints are necessary because it usually is impractical to build forms for an entire structure and place the concrete in one continuous operation. They should not be confused with expansion joints and control joints, which are placed in large structures to allow for movement due to expansion and contraction.
By the nature of its construction, an expansion joint also acts as a control joint and can function as a construction joint (stopping place). Control joints also can be located to coincide with construction joints.
LOCATION OF JOINTS
Construction joints should be located where they will least impair the strength of the finished structure, will be suited to the architectural design, and will facilitate the construction of forms and placing of concrete. To leave their location to chance or convenience with little consideration for their effect on the strength or appearance of the completed structure is bad practice. From the point of view of strength in floor systems, desirable locations for construction joints are at points of minimum shear or points of contraflexure. This applies only to construction joints perpendicular to the main reinforcement in floor slabs. Joints parallel to the main reinforcement may be made anywhere in the slab except within the section that is considered to be the flange of an L-or T-beam.
Construction joints in columns and bearing walls should be located at the undersides of floor slabs, beams, or girders and at the tops of footings and floor slabs. Beams, girders, haunches, column capitals, and drop panels are placed monolithically with the slab system and
should not be cut through by construction joints. Joints should be indicated on the plans or their location decided upon by conference of architect, engineer, and contractor, and prohibited at any other place.
If the placing of concrete is interrupted for a time and the batch begins to set before the next batch is added, the surface between the two should be considered a construction joint and treated as such.
HORIZONTAL CONSTRUCTION JOINTS
Many factors contribute to the practical heights to which concrete can be placed. Prominent ones include the size of formwork materials, the repetitive use of formwork (particularly in multistorey buildings), placing the reinforcing steel, placing the concrete, vibration, and setting time. It is general practice to limit concrete placements to a height of one story. With concrete placement in one storey increments, horizontal bands, grooves, or ledges
can be useful architectural features for concealing construction joints. Horizontal joints can also be aligned with window heads or sills or both. The joints are shortened by the width of the openings and can be architecturally developed to continue the window lines across the
facade.
If V-shaped grooves are used, the joints should be made at the point of the V. If rectangular or slightly beveled grooves are formed, the joints should be made at the top edge of the inner face of the groove, as illustrated in Fig. 61, to avoid a ragged, irregular arris (or edge). Horizontal joints should be straight and level.
VERTICAL CONSTRUCTION JOINTS
The distance between vertical construction joints in castin-place concrete buildings depends on the amount of form work in place or the production capacity of the placing crew and equipment. In walls, a horizontal length of placement in excess of 40 ft (12.2 m) is not normally recommended. Vertical joints can often be located at or near reentrant corners of walls alongside columns or pilasters, or at other places where they become part of the
architectural treatment of the building. In long sections of walls and other places, it might be necessary to form vertical joints in the flat wall surface of the concrete. These joints may also be required to allow movement (as a control joint) and it would be prudent to include a waterstop to ensure a weatherproof joint. Essential details for the construction of horizontal and vertical construction joints are shown in Fig. 62.
Chapter 9 - Joints in Slabs on Ground
A portland cement concrete slab on the ground can move
horizontally because of a change in temperature or moisture;
it can curl (warp) because of a difference in the temperature
or moisture gradient between the top and bottom
of the slab; it can deflect under traffic loads; or differential
settlement can occur. Because the slab is restrained
in various ways against moving freely in response
to these influences, stresses are induced that can cause
cracking. Joints therefore should be designed and spaced
in the slab to control cracking principally due to drying
shrinkage, to permit relative movement (expansion),and
to aid construction.
The designer must have knowledge about the magnitude
and direction of slab movement in order to design
the joints properly. The effect on movement of the following
significant items should be considered: type and
size of joint, slab length, slab thickness, type of subgrade,
type of load-transfer device (if used), material properties,
reinforcement (if used), type and amount of applied
load, and the environment in which the slab is placed and
will serve.
TYPES OF JOINTS
Control joints in floors (also called contraction joints)
are designed to control random cracking attributable to
initial drying shrinkage, thermal and moisture changes,
frictional stresses, warping, and load stresses by providing
a deliberately weakened plane so that the crack
forms along a line at the preselected location.
Isolation joints for floors (also called expansion joints)
are designed primarily to allow for expansion and contraction
attributable to thermal and moisture changes
by providing a complete separation for the full depth of
the slab to allow free movement between abutting surfaces.
It is advisable to provide an isolation joint around
the perimeter of the slab where it abuts on walls, grade
beams, or any structural element that would restrict
movement of the slab in a horizontal plane.
Construction joints are located to establish the boundary
of the concrete placement. They are designed and
located to align with and move the same as contraction
or expansion joints.
Details of each type of joint are illustrated in Fig. 63.
SLAB LENGTH
The amount of opening and closing of a crack or joint is
a direct function of the distance (length or width) of the
slab between joints. Long slabs result in larger openings
at joints than short slabs. Slabs longer than 20 ft (6. I m)
also tend to develop midslab cracks. Wider openings
more easily allow entrance of incompressibles. Openings
wider than 0.035 in. (0.90 mm) have a marked loss in efficiency
of the aggregate interlock to transfer load, thus
increasing the likelihood that traffic will cause faulting
and spalling of the concrete edges at the opening. Using
shorter slabs will avoid these undesired consequences.
Maximum spacing between joints depends primarily on
- Slab thickness
- Shrinkage potential of the concrete
- Curing environment
- Absence or presence of distributed reinforcement
- Desirable width of joint opening
Slabs made of high-slump concrete improperly cured
in a dry environment, with or without reinforcement,
will shrink excessively and crack extensively. Slabs made
of low-slump concrete properly cured in a moist environment,
with or without reinforcement, will have minimum
shrinkage and few cracks. Jobsite practices are somewhere
between these extremes.
Joint spacing* too often is determined by structural
bay size, with the joints placed on column lines only. This practice usually results in large panels with intermediate cracking within the panel and in poor joint performance.
A rule of thumb for plain (no-reinforcement) slabs is
that joint spacing should not exceed 24 slab thicknesses
for concrete made with less than ¾-in. (19-mm) top-size
coarse aggregate; 30 slab thicknesses for concrete with
greater than ¾-in. (19-mm) coarse aggregate, or 36 slab
thicknesses for low-slump concrete. Suggested joint spacings
are given in Table 6.
REINFORCED SLABS ON GROUND
For reinforced slabs on ground, joint spacing normally
varies from 30 to 80 ft (9 to 24 m). The percentage of reinforcement
increases with an increase in joint spacing. In a
continuously reinforced concrete slab that is without
joints except for construction and expansion joints where
needed, the percentage of distributed reinforcing steel is
0.6 and higher.
Slabs with joint spacings longer than those in Table 6
usually will crack in the middle with the cracks generally
held tightly closed by the reinforcement. The amount of
steel (Fig. 64) needed to hold a crack tight can be calculated
using this subgrade-drag-method equation:
where
A = required cross-sectional area of steel in square inch per foot of width of slab (mm2/ m)
W = weight of slab only, pounds per square foot (kg/m3)
F = coefficient of resistance to movement (generally assumed to be l .5 but can vary from about 1.0 to 2.5)
fs = allowable stress in steel, commonly assumed to be 30,000 psi for welded-wire fabric (200 MPa)
L = length of slab between free ends (A free end is any joint where horizontal movement is permitted.)
To be effective in controlling cracks, the steel must be
positioned at or above the middepth of the slab. Distributed
steel placed in a slab on ground for crack control does
not increase significantly the load-carrying capacity of the
floor.
CURLING (WARPING)
In addition to horizontal movement caused by a change
in moisture and temperature, curling of the slab can be
caused by differences in moisture and temperature between
the top and bottom of the slab.
The edges at the joints tend to curl up when the surface
of the slab is dryer or cooler than the bottom. The slab will
assume a reverse curl when the surface is wetter or warmer
than the bottom. When the edges are curled upward, industrial
truck traffic passing over the joint causes a repetitive
vertical movement that creates a great potential for
fatigue cracking in the slab. The amount of vertical movement
(curling) is small with a short, thick slab.
Based on the geometry of the curved slab, the following
rational formula for curling can be derived:
The linear shrinkage of concrete varies between 400
and 600 millionths or ½ to ¾ in. per 100 ft (12 to 18
mm/ 30 m). These values are for typical job concrete, although
higher or lower values can be determined under
certain conditions. Neglecting the effects of temperature,
if the base of the slab is moist it will have less shrinkage
than the top, which is usually drier. As an example, let it
be assumed that the top of the slab has attained a full
shrinkage of 0.0006 while the bottom has shrunk only
half of this amount or 0.0003. The difference in linear unit
shrinkage is then 0.0006 - 0.0003 = 0.0003 in. per inch
(l.062 mm/mm). Under these conditions, a 4-in.-thick
(100-mm) slab, 15 by 15 ft with diagonal of 21.21 ft (5 by 5
m, diagonal= 7.06 m), would have comer curling calculated
by the formula as follows:
Even though designers are aware of moisture effects
and the existence of curling slabs, they tend to use only
temperature-range changes as the design factor when
determining slab length and the size of joint openings.
Chapter 10 - Masonry Materials
Concrete masonry and clay masonry units have been used together for many years. Concrete masonry began as a backup material for brick masonry in the l940's and has grown to the extent that concrete block has almost totally replaced structural clay tile as the backup material in masonry walls. Concrete masonry and brick construction is popular for curtain and panel walls in low-rise and multistory buildings as well as loadbearing masonry walls in both low- and high-rise buildings.
MATERIAL PROPERTIES
Concrete and clay masonry materials exhibit different properties. Each has its own moisture, thermal, elastic, and plastic flow (creep) properties that must be recognizedand taken into account in design, or the masonry wall may not give satisfactory performance. (59)
MOISTURE MOVEMENT
Building materials, except for the metals, tend to expand with increases in moisture content and contract with drying. For some materials these movements are reversible; for others they are not-or are only partially reversible.
Fired-clay products such as brick expand upon contact with water or humid air, and this expansion is not reversible by drying at normal atmospheric temperatures.
Currently the Brick Institute of America recommends a coefficient of moisture expansion of 0.02% (2.0 X 10-4).
Concrete masonry units such as concrete block expand with a moisture gain and contract with a moisture loss. Of greater immediate concern is the initial drying shrinkage of these units. Many factors affect the volume change of concrete masonry. The major controllable ones are type of aggregate and method of curing. Standard units made with normal sand and gravel aggregate generally have less shrinkage than those produced with various
lightweight aggregates. Units cured by high-pressure steam (autoclaving) have one-third or less shrinkage than units cured by various commercial methods in atmospheric pressure. The presence of high-strength-wire joint reinforcement does not eliminate shrinkage, but does distribute the shrinkage stresses and helps control the number and distribution of drying shrinkage cracks. Values for shrinkage of concrete masonry can vary from a low of
0.01% (1.0 X 10-4) to as high as 0.1% (1.0 X 10-3).
THERMAL MOVEMENTS
Theoretically, thermal expansion and contraction movements are reversible if the member is unrestrained. The coefficients for thermal movement can be determined from measurements of length change in prisms by laboratory tests. The generally accepted coefficient for clay masonry is 0.036% (3.6 X 10-4) per l00°F (55.6°C). Thermal movements occur both horizontally and vertically. Concrete masonry also undergoes external expansion
and contraction. The coefficients are 0.043% (4.3 X 10-4) per 100° F (55.6°C) for lightweight concrete masonry and 0.052% (5.2 X 10-4) per I00°F (55.6°C) for dense, normal-weight units.
ELASTIC DEFORMATION
For stresses permitted in brick masonry, the relationship between deformation of the structural elements and their stress is approximately linear. The reduction of length
of axially loaded masonry elements due to design loads are seldom in themselves critical; however, because these dimensional changes are in addition to those caused by other factors, they must be considered.
Elastic deformations for concrete masonry are similar to those found in brick masonry. They are important and must be considered in the design to assure the proper performance of the masonry wall.
PLASTIC FLOW (CREEP)
Some materials when continuously stressed gradually yield in the direction of the stress application. This is referred to as creep or plastic flow. In clay masonry construction,
the clay units themselves are not subject to flow although the mortar joints are. The joints, however, seldom comprise more than 15% to 27% of the volume in compression. A design value of 2.0 (10-7) per unit of length per pound per square inch (6895 N/m2) is suggested
for clay masonry plus mortar.
Creep in concrete masonry is much larger than in clay brick construction. For high-strength concrete masonry units, the design value for creep will be somewhat lower than the value for conventional cast-in-place concrete. The ultimate magnitude of creep of plain concrete ranges from 2.0 X 10-7 to 2.0 X 10-6 but is ordinarily about 1.0 X 10-6) per unit of length per pound per square inch (6895 N/m ). Creep is not only a function of stress and time, but
is also affected by the physical properties of the concrete and the conditions of exposure.
COMPOSITE WALLS
The use of clay and concrete units is a popular combination in composite masonry walls. As pointed out, they respond in different ways to moisture content, temperature, elastic deformation, and plastic flow, with the movement in concrete units greater and generally opposite to that in clay units. Differential movements within the wall can cause cracking when the stresses created by the wall movements exceed the tensile strength of the
masonry. The designer can control these stresses and minimize the incidence of cracking by the use of bond breaks, flexible anchorage and control joints.
DIFFERENTIAL MOVEMENT
A building is a dynamic structure and its successful performance will depend on the designer's (and builder's) understanding of how all the separate parts interact. The
performance of the walls, for example, is dependent on the materials of construction. All of these must be taken into account and a realistic assessment made of their relative
effects in service.
In the case of cavity walls, any restraints of the differential movement between the exterior and interior wythes could lead to stresses and strains causing distress in the masonry system or lateral deflection (bowing or increased curvature) of the walls. Some type of metal ties connecting the two wythes should be used to accommodate differential
movement.
BOND PREVENTION
Masonry walls are usually supported by concrete foundations. If the walls are of clay brick and the bottom wythe is fastened to the foundation wall with strong mortar but with no provision for the opposing dimensional movement of the two materials, the result will inevitably be cracking at the foundation corners. In other instances, shrinkage cracks that sometimes originate in the foundation wall can extend up into the masonry wall. These
problems can be minimized by a bondbreaker between the foundation and the masonry wall. This can be a layer of building paper or smooth flashing placed under the bottom wythe of bricks in buildings where it is not necessary to anchor the walls to the foundation, as in Fig.
65(a). The detail shown in Fig. 65(b) is suggested for grouted masonry walls.
FLEXIBLE ANCHORAGE
Masonry walls tied rigidly to the structural frame for lateral support often crack because of differential movements between the two components. These movements can be controlled by flexible anchors that will resist tension and compression but not shear. This flexibility will
permit the wall and the structural frame to move independently of each other, within certain limits.
CONTROL JOINTS
The term "expansion joint" is used in the clay brick industry because clay units expand. "Control joint" in concrete terminology is the means to control shrinkage cracking
in concrete masonry units. "Control joints" are used in this text for both masonry types. "Movement joints" would be an appropriate descriptive term.
No single recommendation can be made for the location of control joints. Each building must be studied on its own to determine potential movements and the location of control joints to relieve the excessive stress that might result from such movements. Masonry handbooks contain formulas for calculating total movement and methods for forming the control joints.
Vertical Control Joints
Vertical control joints are used to divide long walls or to separate changes in height or thickness at the junctions in L-, T-, or U-shaped buildings (Fig. 66), at the abutment
of walls and columns, and at one or both sides of all wall openings.
For buildings supported by continuous exterior unreinforced masonry, control joints should be placed at intervals not exceeding 200 ft (61 m). In addition, intermediate-movement joints (control joints) should be positioned and spaced in accordance with the recommendations of the Brick Institute of America and the National Concrete Masonry Association.
Horizontal Control Joints
Horizontal control joints (Fig. 67) are especially important in high-rise buildings. Their absence can cause problems in buildings with reinforced concrete frames and exterior clay-brick wythes supported on shelf angles at spandrel levels. The combined effect of drying shrinkage and plastic flow in the structural frame will reduce the floor-to-floor height. Any expansion of the clay-brickfacing wythe will add to the problem. In the absence of a horizontal control joint between the bottom of the shelf angle and the top course of the masonry panel below it, cracking or breaking can occur. Examples of this construction
are shown in American Concrete Institute's Commentary on Building Code Requirements for Concrete Masonry Structures (ACI 531-79).
BOND BEAMS AND HORIZONTAL JOINT REINFORCEMENT
Bond beams are horizontal structural elements that bind the components of a wall into a structural unit with extra strength and stiffness. They are constructed of special units filled with concrete or grout and reinforced with embedded deformed steel. The bond beam offers resistance to horizontal movement for 24 in. (600 mm) above and below its location in the wall. For this reason they usually are located 4 ft ( 1.2 m) apart to offer the entire wall
height crack control. They also are located above or below openings and in other vulnerable positions.
Horizontal joint reinforcement was developed primarily for crack control due to drying shrinkage and temperature change and for this purpose it functions much the same as a bond beam. It usually is a prefabricated (welded) arrangement of two or more longitudinal and cross wires that is embedded in horizontal mortar joints at a minimum vertical distance of 8 in. (200 mm) and a maximum of 24 in. (600 mm) apart, depending on wall height and the spacing of any control joints.
SUMMARY OF DESIGN STANDARDS FOR COMPOSITE MASONRY WALLS
Until recently, load bearing masonry walls were designed to conform to American National Standards Institute (ANSI) A4I.I and A4l.2. The procedures in these two standards are based on a combination of experience and observation of the performance of masonry buildings.
Building Code Requirements for Masonry (ANSI A41. l) covers requirements for the design and construction of nonreinforced masonry in buildings. The standard covers design of masonry of various materials and is based essentially on minimum wall thicknesses for bearing walls and height-to-thickness ratios that must be strictly adhered to.
Building Code Requirements for Reinforced Masonry (ANSI A4I.2) provides requirements for the design and construction of reinforced masonry in buildings. It includes various masonry materials and provides a more engineered approach to the design.
RATIONAL DESIGN PROCEDURES
The Brick Institute of America in the I 960's developed a rational standard for the design of brick masonry. The current version of this standard, Building Code Requirements for Engineered Brick Masonry, was published in 1969.
In I 968, the National Concrete Masonry Association published its Specifications for the Design and Construction of Loadbearing Concrete Masonry. This standard applies only to concrete masonry. Rather than the arbitrary limits on maximum wall heights of the ANSI standards, in these rational procedures wall thicknesses are determined by analysis. Lateral loads can be determined and shearwalls provided in the building.
These two rational design procedures provide methods for solid brick masonry walls and for solid or hollow concrete masonry walls, but not for the two acting in combination in a composite masonry wall. Therefore, until recently there was no accepted national standard
for the rational design of composite brick and block walls.
In 1979, American Concrete Institute (ACI) Committee 53 I on Concrete Masonry Structures published Building Code Requirements for Concrete Masonry Structures
(ACI 531-79), which provides minimum requirements for the engineered design of concrete masonry and composite (brick and block) masonry elements of structures.
The American Society of Civil Engineers and American Concrete Institute 530 Joint Committee on Masonry is currently in the process of developing a design standard for masonry that will cover both empirical and rational methods for the design of all masonry products, including composite masonry.
Chapter 11 - Cracking in Partitions
A partition normally is assumed to be uniformly supported by a structural floor. This holds true when the partition is first built; it does not continue true if the floor subsequently deflects, unless the partition is able to adjust itself to the new floor shape.
If the partition is of rigid construction such as concrete masonry or portland cement plaster, as the floor deflects the points of support for the partition will shift to the ends and the partition must span them as an arch or a simple beam, see Fig. 68(a). When arch action occurs, the area below the arch may detach itself, causing horizontal cracking in the partition. The cumulative width of the cracks will equal the amount of the floor deflection, Fig. 68(b).
If beam action rather than arch action occurs or if the upper floor deflects more than the lower, then vertical cracking is likely at the bottom of the partition, as in Fig. 68(c). Reverse conditions apply if the effect of prestressing causes upward movement in the floor slab.
Such unsightly cracking and damage can be prevented in one of two ways: the amount of floor deflection can be limited to an amount the partitioning system can accept; or a more flexible partitioning system can be selected that will accommodate itself to the floor deflection.
Where a rigid partition is used, it can be designed with suitable reinforcement to span the ends after the floor has deflected. The partitions should be built as late as possible. Gaps can be left during construction and filled in later after most of the dead load deflection has taken
place. Alternatively, control joints can be incorporated in the wall.
Chapter 12 - Earthquake Movements
The general philosophy of earthquake design for buildings is well established and seeks to (l) prevent nonstructural damage in minor earthquakes, which may frequently occur in the service life of a building; (2) prevent structural damage and minimize nonstructural damage in moderate earthquakes, which may occasionally occur; and (3) avoid collapse or serious damage in major earthquakes, which may rarely occur.
Under current design practice the elements of a structure located in a seismic zone are proportioned to resist internal forces resulting from a static analysis of the structure under the code-specified forces. This procedure does not permit determination of the magnitude and location of inelastic deformations in the individual structural members. Thus, although inelasticity may actually occur only in certain levels and locations, inelastic deformability
must be provided for throughout the entire structure.
The inability of this empirical code approach to give a fairly precise determination as to how much deformability to provide for and where can be overcome by taking advantage of recent advances that permit dynamic inelastic (response history) analyses of structures to be carried
out at reasonable cost.
An economical and structurally efficient earthquake resistant design can be achieved by using carefully selected earthquake accelerograms as loading and dynamic, inelastic response-history analysis to determine member forces and deformations.* The design approach makes it possible to (I) predetermine the sequence in which inelasticity
spreads to various designated structural members, (2) provide ductility details only where required, and (3) balance the strength and deformability requirements of individual members. Efficiency, economy, and desired structural performance are achieved as a result.
Chapter 13 - Precast and Prestressed Structures
This publication has dealt with movements and control of cracking in cast-in-place concrete walls, structural frames, and slabs on ground. Current requirements and design procedures for crack control in precast and prestressed (post-tensioned) structures are available in technical publications issued by other authorities.
Controlling Elastic Shortening from Post-Tensioned Forces {Slabs or Beams)
Phase I. Before Post-Tensioning Vertical steel in wall or column is encased in about 6 in. (300 mm) of foam plastic just below the post-tensioning slab, which is supported by forms and shoring. (The distribution of the load caused by camber of the slabs during the tensioning operation can place very high loads on the shoring.)
Phase II. After Post-Tensioning Tensioning of tendons. produces shortening of the slab and causes vertical bars to bend rather than produce cracks in wall or column. Upon completion
of tensioning, the foam plastic is removed and that space is filled with dry-pack mortar or concrete.
After the dry-pack has hardened, the supporting forms or shores may be removed. There may still be some residual movement in the post-tensioned element due to incremental shrinkage and creep, but allowances for the movement have been provided through bending of the vertical steel.
PRECAST ARCHITECTURAL CONCRETE ELEMENTS
The effect of volume changes within precast elements and the support system is one of the main factors to be considered in the design of connections to transfer loads.
Information sufficient to permit the safe design of architectural precast concrete in accordance with commonly accepted industry practice is contained in PC/ Manual
for Structural Design of Architectural Precast Concrete and Architectural Precast Concrete, Prestressed Concrete Institute, Chicago, Illinois.
PRESTRESSED CONCRETE
Cast-in-place, post-tensioned concrete structures must be designed for the following volume-change movements:
- Elastic shortening resulting from the post-tensioning forces (see box)
- Creep shortening resulting from the post-tensioning forces
- Shortening resulting from shrinkage of the concrete
- Thermal shortening resulting from winter weather and possibly thermal expansion from summer heat
Precast concrete structures with prestressed members in the direction of movement must be designed for movements 2, 3, and 4.
Precast concrete structures with conventionally reinforced members in the direction of movement must be designed for movements 3 and 4. When designing for forces resulting from volume change movements in precast, pretensioned elements, refer to PC/ Design Handbook, 2nd edition, published by the Prestressed Concrete Institute, Chicago, Illinois.
When dealing with volume changes in post-tensioned, cast-in-place concrete construction, refer to Post-Tensioning Manual, published by the Post-Tensioning Institute, Phoenix, Arizona.
Chapter 14 - Conclusion
The following is excerpted from J. W. de Courcy's paper "Movement in Concrete Structures," in Concrete, London, June 1969. De Courcy identified 16 significant factors
that should be considered in the design and construction phases for the control of movements in concrete structures:
SITE CONDITIONS AND ARRANGEMENT OF JOINTS
1. Design temperature
Know or estimate the range of air temperature to be expected
at the site to be used in design for control of
thermal stresses. The designer must estimate where the
temperature of concrete as placed fits into the overall
range. It is helpful to have some idea of the duration
periods for the lowest and highest temperatures. The
probable variation of the actual temperature of the structure
may be greater or less than air temperatures. Temperature
of the concrete may not be constant throughout
a structural member, or assembly of members. This is significant
to warping and curling of slabs. Temperature
differences that may be destructive can occur through differences
in the mass of various parts of the structure. For
example, projecting edges of exterior roofs and balconies
are exposed to temperature variations different from the
main interior supporting structure.
2. Spacing of joints as a function of anticipated longitudinal expansion and contraction
Set the maximum distances to be allowed between joints
of various kinds: expansion joints in walls of long buildings
and retaining walls, and control and construction
joints. Structures not designed to withstand stresses arising
from expansion and contraction should be designed
to allow for some arbitrary total thermal movement, say
I¼ in. per 100 ft (30 mm per 30 m). For ordinary structures
usually no calculations of stress or provision for
movement need be carried out for this purpose. Liquid
retaining structures need special attention.
Expansion of buildings intended for manufacture of
combustible materials or articles, or containing rooms
or storerooms where there is danger of an explosion
should include joints that permit movement but not the
penetration of fire either directly or by overheating.
3. Positioning of joints as a function of the building's shape
Stress concentrations that can accompany changes in section, bulk, exposure, method of construction, or that can be associated with settlement must be taken into account when positioning joints.
Where large changes in plan dimensions take place abruptly is a logical place to provide expansion joints. Steps should also be taken to counteract and minimize
the movement that occurs at the junction of the roof slab
with the wall structure. The location of expansion joints
in controlling cracking in a building will depend upon the
reason for their exact location. This frequently is a matter
of experience and can be characterized as the place
where cracks most probably would occur. In general,
expansion or movement joints should pass through the
whole building in one plane. This is highly desirable, but
by no means always possible.
PROPERTIES OF MATERIALS
4. Initial shrinkage
Initial shrinkage in normal portland cement concrete musCbe accepted as inevitable if expansive cement is discounted and if a I 00% moist environment is not usual for
most concrete in service. Such shrinkage will almost certainly
lead to cracking of the concrete depending on the
type, quantity, and disposition of any reinforcement
used, and whether proper jointing practices have been
designed to control it.
In statically indeterminate structures the effects of
shrinkage on the statically indeterminate magnitudes can
be taken into account by assuming a shrinkage amount.
5. Coefficients of thermal expansion
While there is much unanimity about the figure for thermal
expansion mentioned in item 2, it should be borne in
mind that concretes made with various aggregates can in
fact have coefficients of thermal expansion ranging from
about 4.0 X 10-6 per degree Fahrenheit (0.72 X 10-5 / C) to about 7.1 X 10-6 per degree Fahrenheit ( 1.26 X 10-5 / C). The use of a single recommended value may be imprudent
if the actual value can range as much as indicated.
This also points up the fact that the coefficients for the
concrete can vary considerably from any reinforcement
contained in it.
6. Creep and modulus of elasticity
The effects of creep in controlling structural movement
are the relief of tensile stresses that occur in concrete due
to drying shrinkage or temperature change and the increase
in deflection due to sustained loading. An overprecise
examination of creep seems unwarranted because
of the imprecise manner in which actual loading is imposed
except in structures of great height, long span, or
large dead weight, and in prestressed members generally.
An appreciation of the magnitude of the creep effect
should be borne in mind when considering control of
movements in concrete structures.
The significance of the modulus of elasticity in concrete
design is reduced by the fact that it varies with the age of
the concrete and the magnitude of the imposed load. At
working loads, however, the modulus can be used to estimate
deflection due to short-term loading and elastic
deformation due to prestressing. The modulus of elasticity
is also useful to forecast the total strain due to longterm
or sustained loading, and it is for this reason that it
is considered here in conjunction with creep.
7. Tensile strength of concrete
The tensile strength of concrete is neglected in flexural
calculations of reinforced structural concrete except for
estimating the resistance of prestressed concrete to cracking.
Values for extreme fiber stress are given in standard
documents, for example, Building Code Requirements
for Reinforced Concrete, A CI 3 I 8-77.
The tensile strength of concrete is of paramount importance
in the design and construction of concrete slabs on
the ground (floors and pavements). In Concrete Floors
on Ground, Portland Cement Association publication
EB075D, a minimum compressive strength,!:, equal to
4000 lb per square inch (28 MPa) at 28 days is advisable
for any type of commercial and industrial floor, with the
average flexural strength not less than 600 lb per square
inch (4 MPa) at 28 days. The term "flexural strength" is
synonymous with "modulus of rupture" of the concrete.
MISCELLANEOUS STRUCTURAL CONSIDERATIONS
8. Crack widths
Recognition that cracking will occur in reinforced concrete
structures, coupled with the desire to prevent cracking
where it will mar the appearance of the structure or
where it will lead to corrosion of the reinforcement, leads
to establishing maximum widths of cracks. It should be
noted that crack widths normally relate to cracking in
bending or in tension with the direction of the cracks
normal to the direction of the reinforcement. Suggested
limiting values for crack widths selected from the European
Concrete Committee, Recommendations for International
Code of Practice for Reinforced Concrete are-
- For internal structural parts in a normal atmosphere: 0.012 in. (0.3 mm)
- For internal structural parts in a humid or aggressive atmosphere and external parts exposed to weather: 0.008 in. (0.2 mm)
- For internal or external structural parts exposed to a particularly aggressive medium or where watertightness is needed: 0.004 in. (0. l mm)
9. Conduits and pipes embedded in concrete
The problem of temperature stresses arising from the
passage of hot liquids, gas, or vapor through pipes embedded
in structural concrete is addressed in ACI 3 I 8-77,
Building Code Requirements for Reinforced Concrete,
Section 6.3. Dealing specifically with the problem movements
of heated floors and pavements, British practice
(CP204) states: "It is not necessary to provide more expansion
joints in a heated floor than in one that is unheated."
The British practice cautions that where heating
elements cross an expansion joint, care should be taken
to see that
- The pipes are wrapped and coated with bitumen (to prevent bond) for several inches on both sides of the joint
- Electric cables embedded in the concrete are looped in a layer of sand and wrapped for several inches on both sides of the joint
10. Joint assemblies and jointing materials
Information on the detailing of joints, mainly in bridges,
pavements, and liquid-retaining structures, is contained in many standard publications. By nature of the volume of printed material, any useful extracts would be too
lengthy to include here. For information on joint sealants,
expansion joint fillers, waterstops, types and forms
of joints, load transfer devices, metallic parts for joints,
installation of preassembled joints, metal and elastomeric
bearings, fixed joints, and bearings consult the manufacturer's
literature and the appropriate standards and
specifications.
11. Bearing lengths and friction at bearings
In addition to the detailed information available from the
sources mentioned, some reference should be made to
design problems arising at bearings and supports.
The following extract from ACI 3 I 8-63 refers specifically
to prestressed concrete, but it is broadly applicable
to normal reinforced concrete also.
A. Design and detailing of the joints and bearings shall
be based on the forces to be transmitted and on the
effects of dimensional changes due to shrinkage,
elastic deformation, creep, and temperature. Joints
shall be detailed so as to allow sufficient tolerances
for manufacture and erection of the members.
B. Bearings shall be detailed to provide for stress concentrations,
rotations,and the possibledevelopment
of horizontal forces by friction or other restraints.
The question of friction can be answered by providing
for the longitudinal forced ue to friction at expansive bearings
in the design. Bearing pads are available in various
grades; for proper selection consult the technical brochures
available from most pad manufacturers. The static
coefficients of friction for different materials are given in
the table below extracted from PC! Design Handbook of
the Prestressed Concrete Institute.
12. Further aspects extreme temperature
Temperature conditions outside the range in the general
recommendations can arise in various ways. Structures
that will be in contact with liquids or gases at high temperature
need special design care. Factory chimneys and reservoirs
for hot liquids are examples where allowances
must be made for the additional stresses provided by ternperature
differences between the concrete surfaces. The
effect of sustained high temperatures on the compressive
strength of concrete also should be kept in mind.
13. Insulation
The use of insulation properly placed in the building can
play a large part in minimizing the amount of movement
in concrete members due to temperature variations. The
heating effects of direct sunshine on roofs, walls, and slabs
can be reduced by shading them or by light color reflective
treatments.
14. Vibration. including seismic effects
The installation of industrial machinery and air conditioning
equipment within a building may cause vibrations
sufficient to warrant the use of expansion joints. The proximity
to heavy rail or highway traffic can also cause unpleasant
vibration in a building from the disturbance
transmitted through the ground and foundations.
Earthquake forces usually do not require any jointing
in a building other than that provided to accommodate
other characteristics of the structure. Detailed recommendations
for the handling of seismic forces are contained in
the model building codes and other documents.
15. Design of liquid-retaining structures
Formal recommendations for the design ofliquid-retaining
structures are contained in the American Concrete Institute's
Concrete Sanitary Engineering Structures (ACI
350R-77). This document contains references to joint design
and construction to control movement while preventing
leakage.
16. Some points in construction
At the construction site, formwork is the principal matter that will influence the movement of the structure. The formwork must be designed and erected to compensate for the following:
- Anticipated deflection or settlement of the forms and their supports
- Anticipated deflection of the completed structure under load from the instant it begins to carry its own weight
- Optical sag (the illusion of sagging in a long member that is perfectly flat)
Formwork should not hinder the shrinkage movement of concrete. Cast-in-place slabs that will be post-tensioned present no special problems with respect to camber if they are designed properly. Load-balancing techniques can be used to eliminate any problems with upward camber or downward deflection.
The blind use of any recommendations without consideration of the circumstances in a particular project would be imprudent.
References
- The Reversible and Irreversible Drying Shrinkage of Hardened Portland Cement and Tricalcium Silicate Pastes by Helmuth and Turk, Research Department Bulletin RX215, Portland Cement Association, 1967.
- Thermal and Moisture Deformations in Building Materials by M. C. Baker, CBD 56, Division of Building Research, National Research Council of Canada, 1964.
- Design and Control of Concrete Mixtures, Twelfth Edition, EB0OIT, Portland Cement Association, 1979.
- "Influence of Aggregate Restraint on the Shrinkage of Concrete" by Hobbs, Journal of the American Concrete Institute, September 1974. (Also see Reference 12.)
- "Prediction of Drying Shrinkage" by Hobbs and Parrott, Concrete, Cement and Concrete Association, London, February 1979, page 19.
- "Causes and Control of Cracking in Unreinforced Mass Concrete" by Carlson, Houghton, and Polivka, Journal of the American Concrete Institute, July 1979.
- Creep and Drying Shrinkage of Lightweight and Normal- Weight Concretes by Reichard. National Bureau of Standards Monograph 74, I 964.
- Concrete Manual, Eighth Edition, U.S. Bureau of Reclamation,1975.
- Properties of Concrete by Neville, Pitman Publishing Co., New York, 1963.
- Influence of Size and Shape of Member on the Shrinkage and Creep of Concrete by Hansen and Mattock, Bulletin DXI03, Portland Cement Association, 1966.
- "Control of Cracking in Concrete Structures," Concrete International, American Concrete Institute, Committee 224, October 1980.
- Discussion of Reference 4 by Campbell-Allen in Concrete International, American Concrete Institute, May 1981, page IOI.
- "Influence of Specimen Geometry upon Weight Change and Shrinkage of Air-Dried Concrete Specimens" by Hobbs, Magazine of Concrete Research, Vol. 29, No. 99, June 1977,and Vol. 30, No.103,June
- Shrinkage Characteristics of Concrete Masonry Walls, Paper No. 34, U.S. Housing and Home Finance Agency, Washington, D.C., April 1954. Also Journal of the American Concrete Institute, November 1953.
- Rate of Loss in Humidity in a Newly Constructed Masonry Wall, Technical Services Abstract 17, Portland Cement Association, 1968.
- Dilatometer Method for Determination of Thermal Coefficient of Expansion of Fine and Coarse Aggregate by Verbeck and Hass, Research Department Bulletin RX037, Portland Cement Association, 195 I.
- "Thermal Expansion of Aggregates and Concrete Durability" by Callan, Journal of the American Concrete Institute, February 1952. Discussion is in Part 2, December 1952.
- "Petrography of Concrete Aggregate" by Rhoades and Mielenz, Journal of the American Concrete Institute, June 1946.
- "Thermal Properties of Concrete Under Sustained Elevated Temperatures" by Zoldners, Temperature and Concrete, American Concrete Institute Publication SP25, 1968.
- Thermal Expansion of Certain Illinois Limestones by Harvey, Industrial Minerals Notes No. 24, Illinois State Geological Survey, 1966.
- "Thermal Properties [of Aggregates]" by Cook, Significance of Tests and Properties of Concrete and Concrete-Making Materials, STP169B, American Society for Testing and Materials, 1978.
- The Physical Structure and Engineering Properties of Concrete by Powers, Research Department Bulletin RX090, Portland Cement Association, 1958.
- "The Modulus of Concrete and the Coefficient of Expansion of Concrete and Reinforced Concrete at Below Normal Temperatures" by Berwanger, Temperature and Concrete, American Concrete Institute Publication SP25, 1968.
- Concrete for Massive Structures, IS 128T, Portland Cement Association, 1979.
- "Mass Concrete for Dams and Other Massive Structures," ACI Committee 207, Journal of the American Concrete Institute, April 1970.
- "Large Pours for Reinforced Concrete Structures" by Fitzgibbon, Current Practice Sheets No. 28, 35, and 36, Concrete, Cement and Concrete Association, England, March and December 1976 and February
- "Large Pours" by Bamforth, Letter to the Editor, Concrete, Cement and Concrete Association, England, February 198 I.
- "Effect of Restraint, Volume Change, and Reinforcement on Cracking of Massive Concrete," Report of ACI Committee 207, Journal of the American Concrete Institute, July I 973.
- "Plant Drying and Carbonation of Concrete BlockNCMA- PCA Cooperative Program" by Toennies and Shideler, Journal of the American Concrete Institute, May 1963 (also Portland Cement Association DX064).
- Durability of Concrete Construction by Woods, ACI Monograph 4, American Concrete Institute, 1968.
- "Strength, Elastic and Creep Properties of Concrete Masonry" by Ameny, Loov, and Jessop, The International Journal of Masonry Construction, March
- Building Code Requirements for Reinforced Concrete AC/ 318-77, American Concrete Institute, 1977.
- Commentary on Building Code Requirements for Reinforced Concrete (AC! 318-77), American Concrete Institute, 1977.
- Notes on AC/ 318-77 Building Code Requirements for Reinforced Concrete with Design Applications, EB070D, Portland Cement Association, 1981.
- Elastic Properties of Concrete at High Temperatures by Cruz, Portland Cement Association Research Department Bulletin RXl9I, 1966.
- Moments and Reactions for Rectangular Plates by Moody, Engineering Monograph No. 27, U.S. Bureau of Reclamation, Denver, 1960 (also revised edition, 1963).
- Advanced Topics in Inelasticity and Failure of Concrete by Bazant, Swedish Cement and Concrete Research Institute, Stockholm, 1977.
- Time-Dependent Behavior of Columns in Water Tower Place by Russell and Corley, RD052B, Portland Cement Association, 1977.
- "Time-Dependent Forces Induced by Settlement of Supports in Continuous Reinforced Concrete Beams" by Ghali, Dilger, and Neville, Journal of the American Concrete Institute, November 1969.
- "Design Considerations for Concrete Pavement Reinforcement for Crack: Control" by ACI Committee 325, Journal of the American Concrete Institute, October 1956, page 337.
- "Techniques for the Survey and Evaluation of Live Floor Loads and Fire Loads in Modern Office Buildings," NBS Building Science Series 16, National Bureau of Standards, U.S. Department of Commerce,
- "Design Live Loads in Buildings" by Dunham, Transactions of the American Society of Civil Engineers,
- "Deflections of High-Rise Concrete Buildings" by Fintel, Journal of the American Concrete Institute, July 1975.
- Uniform Building Code, Section 2312(h), 1977; or Blue Book, Section l(H), of the Structural Engineers Association of California.
- Analysis and Design of Small Reinforced Concrete Buildings for Earthquake Forces by Derecho, Schultz, and Fintel, EB004D, Portland Cement Association,
- Building Code Requirements for Minimum Design Loads for Buildings and Other Structures, American National Standards Institute A58. I, 1982.
- The Supplement to the National Building Code of Canada, National Research Council of Canada No. 17724, 1980.
- Vibrations of Concrete Structures, American Concrete Institute Publication SP-60, 1979.
- "Human Perception of Transient Vibrations" by Wiss and Parmelee, Journal of the Structural Division, American Society of Civil Engineers, April 1974.
- "Vibration Criteria for Long-Span Concrete Floors" by Allen, Ranier, and Pernica, Vibrations of Concrete Structures, American Concrete Institute Publication SP-60, 1979.
- "Trees and Buildings" by Legget and Crawford, Canadian Building Digest 62, National Research Council of Canada, February 1965.
- "Some Secrets to Building Structures on Expansive Soils" by Kantey, Civil Engineering, American Society of Civil Engineers, December I 980, page 53.
- Frost-Heave Uplift Forces on Foundations by Penner, Research Paper No. 635, Division of Building Research, National Research Council of Canada,
- Column Shortening in Tall Structures- Prediction and Compensation by Fintel, Iyengar, and Ghosh, EBI08D, Portland Cement Association, 1983.
- Thermal Movements in the Upper Floor of a MultiStory Car Park by Williams and Clements, Technical Report 539, Cement and Concrete Association, England, October 1980.
- "Facades: Errors Can Be Expensive," Engineering News-Record, January 24, 1980.
- "Time-Dependent Deformations of Vertical Members in High-Rise Concrete Buildings" by Russell and Corley, to be published by Portland Cement Association.
- Expansion Joints in Buildings, Technical Report No. 65, National Academy of Sciences-National Research Council, Washington, D.C., 1974.
- Composite Masonry: Research Needs by Wintz and York:dale, American Society of Civil Engineers No. ST6, June 1982.
ADDITIONAL READING
"Cracks, Movements, and Joints in Buildings," NRCC
15477, Proceedings No. 2, Division of Building Research,
National Research Council of Canada, Ottawa, 1976.
Reinforced Concrete Slabs by Park: and Gamble, John
Wiley & Sons, New York, 1980, pages 466-496.
Handbook of Concrete Engineering, Fintel, editor, Van
Nostrand Reinhold Company, New York:, 1974.
Proceedings of Symposium on Design for Movement in
Buildings, The Concrete Society, London, 1969.
Strength and Time-Dependent Deformations of Reinforced
Concrete Masonry, TEK 84, National Concrete
Masonry Association, Herndon, Va., 1977.
"Volume Changes and Plastic Flow of Concrete" by
Davis and Kelly, ASTM Report on Significance of Tests
of Concrete and Concrete Aggregates, American Society
for Testing and Materials, Philadelphia, 1943, page 54.
Differential Movement, Expansion Joints, Technical
Notes on Brick: Construction 18A, Brick Institute of
America, McLean, Va., 1963.
Control of Wall Movements with Concrete Masonry,
TEK 3, 1972; and Design of Concrete Masonry for Crack
Control, TEK 53, 1973, National Concrete Masonry
Association, Herndon, Va.
"Analytical Studies of the Effects of Movements on Steel
and Concrete Bridges" by Ganga Rav and Halverson,
Public Roads, December 1980, pages 103-115.
The Language Movements, Strains, and Volume Changes
Autogenous volume change | Change in volume produced by continued hydration of cement exclusive of effects of external forces or change of water content or temperature. |
Camber | A slight upward curvature intentionally built into a structural element or form to improve appearance by offsetting the deflection of the element. |
Contraction | A decrease of volume occurring as the result of any or all processes affecting the bulk volume. |
Creep | Gradual time-dependent deformation due to sustained load. See also plastic flow. |
Curling | Distortion of an essentially straight or flat member into a curved, warped, or dished shape due to creep or to internal
differences in temperature or moisture content. |
Deflection | Movement from the original position of a structural element because of bending or shear deformation caused
by its weight, applied loads, and temperature or moisture changes. |
Deformation | A change in dimension or shape of a member due to stress, drying shrinkage, and temperature changes. |
Elasticity | That capability of a material to recover its original size and shape after deformation. |
Expansion | An increase in volume occurring as the result of any or all processes affecting the bulk volume. |
Faulting | Differential vertical displacement of a slab or other member adjacent to a joint or crack. |
Flow | Time-dependent irrecoverable deformation. |
Initial drying shrinkage | -The difference between the as-cast length of a specimen and its length when first dried. |
Permanent set | Residual deformation after the removal of all loads (aside from creep effects). |
Plastic deformation | Deformation beyond the elastic limit. |
Plastic flow | Gradual time-dependent deformation due to sustained load. See creep. |
Restraint | Restriction of free movement of hardened concrete; restraint can be internal or external and may act in one or more directions. |
Setting shrinkage | A reduction in volume of concrete prior to the final set of cement,caused by settling of the solids and by the decrease
in volume due to the chemical combination of water with cement when an external source of curing water is not present. |
Shrinkage | Volume decrease caused by drying or chemical changes; a function of time but not of temperature nor of stress due to external load. |
Strain | Deformation of a material resulting from external loading. |
Swelling | Volume increase caused by wetting or chemical changes, or both; a function of time but not of temperature nor of stress due to external load. |
Time-dependent deformation | The deformation of concrete occurring with appreciable time (as days, weeks, or months); includes creep and characteristics affected by age and strength changes such as elasticity, drying, shrinkage, and temperature effects. |
Volume change | -An increase or decrease in volume (length, width, and thickness). |
The Language Cracking
(There are many causes of cracking in concrete structures
and their origin can only be determined by careful
investigation.)
Checking | Development of shallow cracks at closely spaced but irregular intervals
on the surface of mortar or concrete. See also crazing. |
Crazing | Development of fine, random cracks caused by shrinkage. See also checking. |
D-cracking | The progressive formation on a concrete surface of a series of fine cracks at rather close intervals, often of random patterns, but in slabs on grade paralleling edges, joints, and cracks and usually with a radius at slab corners. May be accompanied by deposits of calcium carbonate. |
Delamination | A separation along a plane parallel to a surface. In the case of a concrete
topping slab or wall panel, the splitting, cracking, or separation in a
plane roughly parallel to the surface. Similar to spalling, scaling, or peeling except that delamination affects large areas and can often only be detected by tapping. |
Pattern cracking | - Fine openings on concrete surfaces in a pattern, resulting from a decrease in volume of the material near the surface or increase in volume of the material below the surface, or both. |
Plastic cracking | - Cracking that occurs in the surface of fresh concrete soon after it is placed and while it is still plastic. |
Shrinkage cracking | Cracking of a structure or member due to failure in tension caused by
external or internal restraints as reduction in moisture content develops
or as carbonation occurs, or both. |
Temperature cracking | Cracking due to tensile failure, caused by temperature drop in members subjected to external restraints or temperature differential in members subjected to internal restraints. |