Clay Soil — Damage to Buildings

Introduction

Foundation movements are a major cause of distress to established buildings. The main cause of such movements in Australia is the swelling and shrinking of expansive clays resulting from soil moisture changes. There are two aspects to this problem. Firstly, buildings must be managed in a manner that reduces the possibility of damage. Secondly, if foundation movements do occur, the damage should be repaired and measure taken to stabilize the footing system.

To achieve these aims, a sound knowledge of foundation behaviour is required. For example, unless the cause of the distress is clearly identified and appropriate remedial measures are selected, there is a danger of further failure.

This report is concerned mostly with the damage that may be caused to existing buildings by movements of expansive clays resulting from soil moisture changes. However, in order to give this particular problem its proper perspective, some discussion of other forms of foundation behaviour is appropriate.

The Nature of Foundation Materials

The foundation of a building is defined as the soil or rock upon which the footings are constructed. The term “soil” is used in the engineering sense to mean the sand, silt or clay material below the organic top-soil layer. The various types of soil are distinguished by the size of the particles. Sands comprise material down to 0.06mm. Silts include the range 0.06mm to 0.002mm, and clays consist of particles finer than 0.002mm. When soils contain mixed types, it is usually the properties of the finer particles that predominate. For example, sandy clay behaves more like clay than sand.

Obviously, if the sound rock is present at or near the surface, it provides an ideal stable material upon which to found a building. Sydney sandstone is an excellent foundation. If the depth of sound rock is moderate, the piles or piers can be driven or drilled to the rock, giving a stable base. For small buildings, or when the depth to rock is excessive, the building is commonly founded within the soil layer. The respected performance of such material depends upon whether it is sand, silt or clay, as well as on loading and environmental factors.

Granular materials, including sands and gravels, are usually trouble-free. Loose sands or heavily loaded sands can experience some settlement problems, but usually, the settlement occurs immediately upon loading. Consequently, problems with older buildings are not likely unless a change is made in the loading. One less obvious way in which this could occur is by pumping water from the soil in an adjoining building site excavation. This may result in a significant lowering of the natural water level or water table and will effectively increase the load within the soil above the newly established water level. The additional ‘load’ may initiate settlement in loose sands or soft silts and clays. Other building activities such as blasting or piling can cause settlement due to compaction or vibration-sensitive sand.

It is important that those who are responsible for existing buildings that may be affected by new construction on adjacent sites make a careful prior survey of the condition of the building in order to facilitate claims and repairs if damage eventuates.

Silts are fine-grained soils with particle size and properties intermediate between the two major groups of sands and clays. Such materials do not have the strength that comes from the plasticity of clays or from the particle to particle contact friction in sands.

Moreover, silts can be subject to large settlements occurring over a long period. Loose silts are generally unsuitable as a foundation. Medium to dense silts may behave similarly to either clays or sands depending on the site and shape of the particles.

Clays are the finest grained soils. The upper limit on grain size is 0.002mm, but most of the clay particles will be even smaller. With decreasing grain size, the surface area of the particles for a given volume increases and, consequently, surface effects dominate the physical and chemical behaviour of the clay. The individual particles of clays are plate-like crystals of clay minerals, and the properties of these minerals also influence clay behaviour. Therefore it is not surprising that there is no single laboratory test that fully describes a clay’s characteristics.

When considering clay as foundation material, the following three properties are of particular interest:

Soil strength, more commonly expressed as loading bearing capacity, settlement due to loading and potential expansive movement

The bearing capacity of a clay, defined as the maximum load per unit area the soil can sustain without failing, should not be a problem in an existing building unless major changes in loading occur or the moisture content of the clay is dramatically increased. Protection of the foundation material from excessive moisture, due to either inadequate site drainage or plumbing leaks, is essential to the proper maintenance of any building and should rarely be a cause of a bearing capacity failure.

Load settlement of clays is more likely to be a problem. Clays undergo settlement in two phases. Initial settlement is associated with water slowly squeezing out of a saturated clay as a direct result of the pressures applied by the building. This form of settlement decreases with time and can be readily predicted by standard engineering tests. In some cases this may be reduced by preloading the foundation material.

The second phase of load settlement results from slip of the clay grain-to-grain contact. Secondary settlement can proceed for centuries. As an example, it has been estimated that the London clays are settling at rates up to 300mm per century. The rate of movement can be reduced by lowering bearing pressures (underpinning with wide pads) or transferring building loads to a more rigid foundation material (deep underpinning with piers).

Most people connected with the building industry are well aware of the problems of settlement of foundations under load. In Australia, clays are normally unsaturated (or dry) and load settlement problems are confined usually to relatively small deposits of marine clays and swamp sediment. Consequently, the major cause of footing failures is movements associated with changes in the moisture content of clays. Such changes cause either soil swelling or shrinkage, and the clays that exhibit this behaviour are termed expansive clays. Their distinctive behaviour is more commonly referred to as reactivity. The nature of these clays are described in detail in the next section.

Characteristics of Expansive Clays

All clays can swell or shrink and the amount of movement depends on the moisture change and the nature of the clay. Some of the most important factors that determine the potential reactivity of the clay are given below.

(a) Mineral composition. The reactivity of clay is influenced by the mineral composition of its plate-like particles. Montmorillonite mineral particles can swell substantially as water molecules penetrate the layers of their crystal structure. This effect does not occur with the more stable kaolinite mineral where only inter-particle movements are involved. Illite behaves in an intermediate fashion. Clay minerals can be identified and their relative proportions estimated by X-ray diffraction and spectrophotometric techniques, but rarely would such laborious methods be justified except for major projects.

(b) Particle Size. Most expansive-clay behaviour is attributable to the influence of forces in the thin films of water surrounding and connecting grains of clay. With smaller grain sizes, the number and significance of these connections increase, which likewise increases the soil reactivity. The reactivity of the clay is moderated also by the amount of inert sand of silt particles present with the clay in the soil. The distribution of particle sizes may be determined in the laboratory, but it is a relatively time-consuming test.

(c) Electrolyte composition. The behaviour of clay is influenced by the chemicals or electrolytes dissolved in the soil-water films at the grain interfaces. This dependence is most evident in the influence of lime stabilization on clays, where soil electrolytes are replaced by the calcium in the lime, thereby causing a reduction in the plasticity and expansive behaviour of some clays. However, in practice it is extremely difficult to ensure that the lime is dispersed uniformly throughout the foundation material.

(d) Soil profile. The soil profile is a record of the variation with depth below ground level of the different layers of foundation material, which are distinguishable by their colour, texture and composition. It can influence the reactivity of clay soil foundations in two ways. First, if the clay layer is very shallow, the total expansive movement will be lessened. Second, and more commonly, an upper soil layer of inert sandy or silty material can act as a moisture barrier and so reduce the amount of moisture change in the underlying clay layers.

So the behaviour of an expansive clay depends on a variety of parameters, some of which are difficult to quantify. It can be quite difficult to determine whether a particular clay profile is expansive and some guidance on this problem will be given in the next section.

Identification of an Expansive Clay

In Melbourne, it has been possible to classify the expansive behaviour of the various soil profiles on the basis of geological maps. This has proved satisfactory because of the following special conditions.

(a) Most of Melbourne has residual soils which are formed by gradual weathering of the base rock in situ. Thus their properties can be associated reliably with the underlying rock type. Also, the soils are generally consistent over the large areas of the same rock type. In these circumstances, the classification of expensive behaviour can be based on a geological map. This is not possible for alluvial soils whose properties can vary significantly over a small area.

(b) The local climate consists generally of a dry, hot summer and a wet winter. Moreover, the clays are usually shallow (1 to 3 m depth). Thus, the extremes of seasonal moisture conditions can be used as a guide to the probable moisture changes under a building.

(c) In Melbourne, the clays have been studied extensively. This work has encompassed laboratory tests and field observations of actual houses, experimental footing systems and ground-movement stations. These data are not available for other regions.

On the other hand, in Sydney, much of the inner city area has been build on sandstone or on shallow sandy clay soils over sandstone, and consequently, expansive clay movements are not significant. In the west of Sydney (eg. Penrith, Blacktown and Campbelltown), the clays are sometimes deeper and potentially expensive. Sydney’s climate provides substantial summer rainfall and consequently, the changes in moisture conditions from summer to winter are not usually severe. However, in times of prolonged droughts, significant shrinkage movements, perhaps 10 to 40mm, can occur.

Brisbane’s more humid climate tends to diminish potential expansion clay movements although occasional drought does induce significant movements. Conditions are even more favourable in Perth, with the main city and suburban area being founded on sand. Hobart rarely encounters dry seasons severe enough to cause problems. Deep alluvial soils in the Tamar Valley south of Launceston have been known to present the occasional problem.

The situation is less favourable in Adelaide. The black earths (locally called Bay of Biscay soils) are highly expansive and also quite permeable. Adelaide has a very hot, dry summer and relatively heavy autumn/winter rainfall so that changes in moisture contents due to seasonal surface effects can occur to considerable depths. Moreover, many of these areas are underlain by a highly plastic but almost impermeable clay (‘Hindmarsh clay’). This clay is capable of substantial movements although the moisture changes that induce such movements may take decades to penetrate deeply into the clay.

In the rural areas of Australia, data on engineering properties of soils often do not exist. Invariably the reactivity of clay deposits has to be determined from first principles. Some of the more common methods of achieving this goal are as follows:

(a) Experience and observation. Perhaps one of the simplest methods is the observation of buildings, pavements, and fences in the area. These structures can be distorted by movements of highly expansive soils. Often such soils are well known by the local building department. Where there are no signs of distress to structures in an area where laboratory tests predict large movements, some circumspection is warranted before accepting the laboratory results.

(b) Field experimentation. A variety of field experiments is possible. A simple technique is to estimate the potential movement from the extremes of moisture content in the soil. The moisture content is measured at the end of a dry spell over various depths and again at a similar location in wet conditions. The movement is estimated by a simple formula relating movement to soil moisture change, which has been shown to be applicable for Melbourne Clays. (Refer Appendix A1)

Ground movement stations provide useful information. The movement at the surface and at depth is compared with a deep stable benchmark. Rods are installed in the ground at the required depths for monitoring. The sides of the rods are isolated from the soil. Observations are required over a range of moisture conditions and therefore one or more seasons are required to obtain any meaningful information. Accordingly, this technique may be practicable for major projects.

(c) Simple laboratory testing. The potential reactivity of clays is often related to index tests such as linear shrinkage, plastic limit, or the proportion of clay-sized particles. An example of a simple classification system is given in Appendix A2. The reactivity of a clay certainly increases with these quantities but the relationship is not clear-cut and seems to depend on other clay properties that are not easily measured. Simple correlations with index test values ignore the importance of the depth or wetting can occur. For Melbourne clay types and climate, experience has indicated that fairly stable sites could have clays with linear shrinkage values as high as 18%, although 12% would be more typical. Values above 20% were associated usually with highly expansive clay sites.

(d) Precise laboratory testing. In the laboratory, the expansion or shrinkage of clay samples can be induced by changing the moisture condition over a known range. The clay movement for a given moisture change is then conveniently termed the instability index of the soil. Once this instability index has been determined, ground movements at a particular site may be approximated by multiplying its value by the anticipated changes in soil moisture state with depth.

Measurements of this kind offer the potential for a fairly accurate and rational theory for expansive clays. The more sophisticated tests are slow and expensive and therefore are often more suited to research than an immediate practical design. However, other more simple test methods are available that are most likely adequate for engineering design problems.

Moisture Changes in Clays

Terminology and Definitions

In order to discuss soil moisture changes, it is necessary to introduce the following basic technical concepts:

(a) Soil moisture content. The moisture content of soil is defined as the ratio of the weights of soil moisture and dry soil, expressed as a percentage. The significance of this moisture content value depends upon the type of clay. For example, at 25% moisture content, one clay may seem dry but another clay may be in a very moist state.

(b) Soil Suction. A more useful concept is that of soil suction, which is a measure of the internal stress, caused by the small amounts of water at the particle-to-particle interfaces. The common unit of suction is the pF unit defined in Appendix A3. Generally, soil suctions can vary from pF 2.5 to pF5 under natural climatic conditions. High suction values are associated with dry soils and low suctions with wet soils. Over the normal range of suctions, the moisture content is approximately linearly proportional to suction for a particular soil.

Soil suctions are commonly determined from measurements of the relative humidity of the air within the soil. Generally, good temperature control is necessary for reliable readings and measurements of suction to be conducted in a laboratory rather on-site. (Further information on soil suction is presented in Appendix A3).

(c) Instability index. The instability index, as described earlier, is the percentage change in the height of a clay sample for a unit change in suction (%/pF). Typical values are commonly 3 to 6%, but for highly expansive clays the value may exceed 10%. (Refer to Appendix A4 for the methods of determination.)

Soil suction changes can occur around and under a building as a consequence of the following:

(a) Seasonal climatic effects.

(b) Interaction (the effect of the presence) of the building with natural seasonal moisture changes

(d) Extraction of moisture by trees or recovery of soil moisture subsequent to tree felling.

Each factor is considered in some detail in the following discussion.

Seasonal Climatic Effects

Under natural conditions, the suction in the soil depends on the climate and vegetation. Over much of Australia, summer is a time of moisture loss with hot dry conditions. In winter, rainfall usually exceeds evapotranspiration and wetter soil conditions prevail. These semi-arid conditions give rise to the typical seasonal suction profiles shown in Figure 1, where the variation in suction decreases with depth and finally reaches a stable value.

For more even climates, seasonal changes are not as significant and natural variations in suction are associated with exceptional droughts and wet seasons.

Interaction of the Building with Natural Seasonal Moisture Changes

When a building is first erected, the natural soil suction profile may lie anywhere within the range from the seasonally dry to seasonally wet profiles. The building then interferes with these conditions by sheltering the soil from rainfall and evapotranspiration. If the building has a slab-on-ground or a suspended floor with poor sub-floor ventilation, the soil surface can be considered to be sealed. Consequently, soil suction near the centre of the building will come to equilibrium with the stable soil suction value at depth (refer to Figure 2). The soil remote from the building will still undergo seasonal suction changes.

Below the centre of the building, there would be a long-term soil movement of either swelling or shrinking depending upon the initial soil suction. If the building was erected in the dry season (high suctions) there would be a long-term swelling, and the reverse would apply. This movement may continue for years or decades. Near the edge of the building, the soil suctions would be intermediate between the centre and seasonal values away from the influence of the building, and some seasonal variation would be expected. The resultant building deformation would tend to be as shown in Figure 3a.

This seasonal movement may not be completely reversible. A long-term shrinkage settlement can result from moisture cycling under heavy loads.

Where the building is on a well-drained site and has a well-ventilated sub-floor space, the soil under the building will eventually become very dry as demonstrated by the suction profiles in Figure 2b. The resultant soil shrinkage may cause subsidence of the sub-floor supports as shown in Figure 3b. Such movements have been observed in some small multi-storey buildings as well as in houses.

Soil suctions under a building can be lowered also by plumbing leaks, bad drainage, site aspect, and garden watering. Plumbing leaks can produce disastrous movements and should be remedied as soon as possible. Poor drainage is not uncommon. Particular care should be taken to ensure that surface water does not collect adjacent to footings and that roof drainage is maintained properly. The effects of inefficient drainage are often most noticeable on cut-and-fill sites where water accumulates in the cut area yet drains freely from the filled area on the opposite side of the building. Consequently, the soil heaves along one side but shrinks along with the other, and the building almost leans downhill (refer figure 3c).

Differential drying can arise from the site aspect, with an exposed north side of the building experiencing more severe drying north side of the building experiencing more severe drying cycles than the shaded southern area. Garden watering has produced failures, usually from the excessive use of fixed watering installations. Proper maintenance of moisture conditions in a garden is essential to prevent either excessive soil drying or wetting.

Trees can have an enormous effect on soil suctions and will be discussed separately.

Interaction of the Urban Infrastructure With Natural Seasonal Moisture Changes

For the very deep clays, long-term changes in moisture conditions will occur as a region changes from a natural grassed, wooded area, or market garden with established drainage and seepage areas, to an urban environment with paving, buildings, gardens and efficient stormwater drainage. Changes from septic to sewerage systems will also change the overall soil suction profiles. These changes are superimposed on the seasonal and local effects mentioned in the previous section. They have their most significant influence on the deeper layers of clay, say from 3 to 15m. To some extent, swelling movements are restricted by the soil overburden pressures (or self-weight) at depth, but nonetheless, large movements are possible. Moreover, these movements occur below the level of most footing systems. The movement may proceed very slowly; for example, movements in Adelaide have been recorded over a period of 30 years. Generally, if the deep-seated movement takes place over a wide area, buildings may not be adversely affected unless building sites are located across a boundary of the affected area, or soil reactivity varies sharply within the area allowing differential ground movements to occur.

Extraction of Moisture by Trees or Recovery of Soil Moisture Subsequent to Tree Felling

Trees require substantial amounts of water. During dry spells, tree roots draw moisture from the soil. If soil water storage is not fully replenished, the roots will extend in search of further soil moisture in subsequent dry periods. Clay will shrink with the extraction of moisture (increasing suction) and subsequently, overlying structures may settle.

The removal of large trees poses the converse of this problem. As soil moisture is gradually restored, clays swell and heave. Shallow seated footings may be uplifted by the soil, depending to a large extent on the loads carried by the footings.

There are many factors that determine the extent of moisture removal by trees. Some of the more important factors are as follows:

(a) Soil profile. Obviously, the proportion of expansive clay in the profile is a determining factor in accessing the potential for damage to a building. However, the soil profile can also affect tree root growth patterns and hence the potential zone of drying. For example, the presence of a water table or rock layer may control the extent of root growth since roots will not penetrate either of them. Furthermore, if an expansive clay is covered by non-expansive soil layers, the depth of root penetration relative to the depth of the top layers will determine the overall movement.

(b) Proximity of trees. The lateral root systems of trees act as the primary soil moisture collector. Limited field data (Yeager 1935) indicates that soil moisture conditions and tree species are the main factors in determining lateral root spread. In persistently wet soil, roots tend to be more concentrated and penetrate deeper.

For convenience, the lateral root spread can be related to the height of the tree, H, and may vary from 0.4 to 2.1 H in natural field conditions (Yeager 1935). So the potential zone of soil drying in terms of the tree height can vary considerably.

(c) Number of trees. The competition for soil moisture between the roots of neighbouring trees may extend the normal lateral root spread of individual trees.

(d) Tree species. The species of the tree determines the tree’s potential water uptake, the pattern of root development for a given site, and its ability to survive in dry soil conditions. In other words, the species can determine the zone of soil drying and the extent of drying within that zone.

(e) Age of tree. The age of the tree relative to the building is important when considering whether it is safe to remove a tree. If the tree is much older than the building, a careful analysis of soil conditions will be required to prevent damage by heaving foundations. If however, the tree is younger than the building, then the maximum possible heave after the tree is removed will be less than the soil shrinkage that has already occurred. Damage may still occur if old settlement cracks have been filled with rigid fillers. As the foundations heave back the filler is compressed, causing bulging of wall renders.

To demonstrate the potential extent of suction changes in clays occurring as a result of moisture extraction by trees, a few typical suction profiles have been reproduced from the literature in Figure 4. Generally, the suction data have been determined during investigations of building damage in which a suction profile in the ground between tree and building has been compared to the profile in an adjoining area relatively devoid of trees. Such studies give valuable information regarding the possible depth and extent of the tree drying effect

Response of Buildings to Foundation Movements

The tolerance of a building to foundation movements brought about by moisture changes in clays is determined by the type of construction of its footings and walls and the building materials used.

Footings

The footings of the old buildings are frequently inadequate. Concrete for footings became part of general building practice only towards the end of the 19th century, but steel reinforcement for concrete was not introduced until the 1920s. The addition of reinforcement led eventually to the concept of the footing as a beam with some capacity (or stiffness) to span local ground movements. Prior to this time, footings were used only to transmit building loads to the soil and therefore non-structural footings or rubble, stone, or brick were common.

Reinforced concrete strip-footings are capable of some beam action, which increases with both the level of reinforcement and the depth of the section. Increasing the footing depth is also advantageous in those expansive clay areas where the depth or moisture change is relatively shallow. As the depth of the footing approaches the depth of negligible moisture change, potential movements of the footing base decrease to zero. However, deep footings can interfere with normal soil moisture distribution, thereby creating differential moisture conditions and hence soil pressures on either side of the footing sufficient to rotate it. The increased surface area of the footing can also give rise to high friction forces in swelling soil.

Deep basements or cellars act as an extended footing system and thereby reduce building movements when compared with similar buildings seated on shallow footings.

Deep piers or piles that penetrate the zone of soil suction change can be most effective in resisting the effects of reactive soils. However, problems can still occur in older buildings. First, shrinking soils remove side friction support so the friction piles may become effectively overloaded as the soil dries, and some load settlement may occur. Second, swelling soils can cause uplift of piers (piles), the magnitude of which is controlled by the depth of the pier and the load carried by it. If the pier is not reinforced, uplift forces can cause tensile failure of the pier section.

A further problem exists with the footing beams supported by the piers or piles. In the case of swelling clay soils, the underside of the beam must be isolated adequately from the surface soil movement, otherwise, the beam may be pushed up off its pile supports.

Walls

Walls vary considerably in their tolerance to footing movements. Apart from the type of wall, tolerance to movement can depend upon the number and size of openings, the height of length ratio of the wall, the degree of articulation in the wall, and the pattern of the ground movement. Articulation is the segmentation of long walls into smaller panels with the use of vertical construction joints, full-height windows, and\or independent infill panels over openings. Its effect is to accommodate movements by providing freely opening and closing joints in the wall.

The type of wall material and construction has a considerable effect. Modern brick-veneer construction is thought to be between 2 and 3 times as accommodating as cavity brick. Common wall types in order of increasing tolerance to footing movement are stone masonry, cavity brick, brick veneer and weatherboard.

Summary of Empirical Tree Planting Rules to Avoid Building Damage

Some of the many factors that influence the magnitude of soil suction changes caused by trees have been discussed already. The extent to which subsequent expansive clay movements are translated into building movement and damage is governed by the nature of the particular building.

To minimize the risk of damage, empirical tree planting rules state the minimum safe horizontal distance (D), in terms of a proportion of tree’s maximum height (H), at which a tree can be planted from a building’s perimeter (Figure 5). A brief review of the available literature on tree damage to buildings is presented in Appendix B. Wherever possible, damage observations have been related to D:H ratios.

Building damage is unlikely if D:H exceeds 1 for single trees and 1.5 for dense stands, rows, or clusters of trees. These limits may be relaxed if the soil profile or clay reactivity is such as limiting potential shrinkage settlement, the building is relatively tolerant to distortion, or its footings are able to resist movement. For example, investigations have shown that the risk of damage to conventional brick-veneer dwellings in Melbourne is minimal for D:H ratio greater than 0.5. Unfortunately, there is very little evidence at present to allow variations in the D:H limits for different species of trees, although it should be possible theoretically.

Investigation and Repair of Building Damage

Effective repair of buildings cannot be achieved without a proper investigation of the problem to uncover the cause (or causes) of damage. Further analysis of the site may then be required to determine the feasibility of rectifying the damage. In all circumstances, the initial diagnosis should be approached with an open mind and must be carried out methodically.

Causes of Damage

Most of the causes of wall cracking can be classified according to the stage of planning or construction of the building during which they occur, as described below:

(a) Site Investigation. Either a site investigation was not required or the investigation failed to reveal potential problems. In the case of older buildings, where the emphasis was placed on soil bearing capacity, the possibility of reactive clay movements may not have been explored.

At sites where load settlement could be a problem, important information may not be revealed if the investigation has not been carried out to suitable depths or has not been extended adequately to encompass the whole site. The properties of the soil within a depth below the footing equal to twice the footing breadth influence the load-settlement behaviour of the foundation. The number of exploratory boreholes required is dependent largely on the site conditions. For example, it is well-known in Melbourne that Quaternary basalt flows can vary considerably across a site. As a consequence, one corner of a building may be underlaid by rock and the opposite corner by a considerable depth of residual soil.

(b) Footing design. The designer may misinterpret the site investigation report or, in the absence of a report, wrongfully assume certain soil properties. Again, lack of knowledge of soil moisture movements has resulted in inadequate designs for older buildings. A further problem may exist where footings have been designed structurally, but the designer has failed to consider the susceptibility of the superstructure to cracking, and deflection tolerances has been exceeded.

(c) Site drainage provisions. Drainage is an essential consideration to prevent excessive moisture movements and consequent heave or consolidation of soft or loose soils. Therefore the specification of site drainage should be a design responsibility that must be coordinated with landscaping requirements.

(d) Construction. Faults in construction arise commonly from poor workmanship, misreading of plans, and ignorance of the properties of building materials. Concrete shrinks and subsequently cracks, and often the cracks are taken wrongly as an indication of settlement. Masonry expands and so can induce vertical cracking in long runs of walls at restraints such as building corners. Green hardwood timber in subfloors or wall frames can distort significantly as it dries, thereby cracking the building elements it supports.

(e) Post-construction maintenance. In terms of moisture movements in reactive clays, the more common problems include location of trees too close to the building, and neglect of either site drainage or the maintenance of plumbing lines. A further problem may exist with pier and beam construction where a gap was provided beneath the beam to isolate it from soil heave, but has later been filled during landscaping of the surrounding area.

Other causes of damage do exist which do not come into any of the above categories. For example, the dewatering of foundation excavations in neighbouring properties, vibrations arising from either blasting or traffic, and underground mining, all lie largely outside the direct control of the property owners.

A Routine Investigation Procedure

To ensure a thorough diagnosis of the causing of building movements, each investigation should follow a planned, standardized procedure. One suggested routine is as follows:

Record construction data
Obtain the history of the damage
Record crack locations and widths
Conduct a level survey of the building
Ascertain the typical soil profile of the site
Obtain soil samples at depth across the site
Carry out laboratory or field tests if required
Analyse the above information

Construction data are essential in determining qualitatively the susceptibility to the movement of the building as a whole. The type of footings, their dimensions and reinforcing, as well as the wall construction, are all important factors. In some cases, construction plans may not be available and excavation must be carried out to gather the necessary data.

The history of damage may provide clues to the cause of movement. Cracks that appear in summer or early autumn are unlikely to be caused by soil consolidation, but rather by soil shrinkage. Cracks that open in summer but close to some extent in winter are due probably to reactive clays, while cracks that progressively get larger as time proceeds can be due to either secondary consolidation of soft clays or extended drying of reactive clays by growing trees.

The date when damage was first noticed can verify whether any substantial changes in the environment of the building could have been responsible for the movement. For example, in the case of tree damage, the damage could be expected to appear within 5 to 10 years of tree planting in a semi-arid climate. In more temperate climates, however, damage may not become apparent until the first major drought after planting.

The location, direction and width of cracking provide clues as to the mode of deformation of the footings and the magnitude of the differential movement. Horizontal cracking indicates that a wall has bowed, possibly as a result of footing rotation. Cracks that are wider at the top of a wall at the bottom occur when footings are in the centre-heave or hogging mode. Dishing movements have the reverse effect. The figure illustrates some common crack patterns and their causes.

The widths and frequency of cracking are used to assess the extent of damage (Appendix C), which may help to provide a more rational basis for the selection of appropriate remedial treatment.

Level surveys of damaged buildings are often particularly valuable. With due regard to original building tolerances, they can still provide a reliable picture of differential movements across a floor. Generally, with older buildings, levels may be taken only on masonry walls preferably at or near the damp proof course. Timber floor levels will probably be unreliable due to either previous re-stumping operations or possible fungal decay problems at the time of inspection.

Once the mode of footing deformation has been established, the reasons for the foundation movement must be investigated. A quick check of the soil profile will soon confirm the probability of soil consolidation or reactive clay movement. If the latter is suspected, a comparison of soil moisture conditions beneath the high and low spots of the building will be essential to the investigation.

Remedial Treatments

Knowing the cause of the damage, a remedial treatment may be assigned that takes into consideration the present extent of damage and both the risk and cost of further damage.

In severe cases of load settlement, a form of underpinning is normally recommended to stabilize the movement of the building. Underpinning provides new support to the building and can either be designed to decrease foundation pressures using wide underpins (Figure 7a) or to transfer the foundation loads by means of piers to a deeper and stronger base (Figure 7b). The latter result may be achieved also by small diameter “root” piles (Figure 7c) which are, in effect, concrete piers cast by a pressure injection technique (Koreck 1978).

However, the use of underpinning in cases of damage due to expansive clay movements is limited and needs to be carefully designed to suit the particular situation. In such cases, the strength of the soil is rarely in question and the underpinning is required to stiffen the footing system or carry the footings to a stable depth. If shrinkage settlement has been responsible for the building damage, the latter option by itself may be adequate to stabilize the building. ? Where soil heave is a problem, the existing footing system should be structurally tied to the deep underpins. Roots piles may then be the most satisfactory solution.

Alternative solutions consist of balancing differences in soil moisture distribution about the building. Lime stabilization will not be discussed as an option because of a lack of experience with this technique and the recently reported failures of field trials (Poor 1975, 1976, 1978). Where damage is slight and the expected risk of further damage is low, moisture differences may be corrected by simple watering programmes and rectification of obvious causes of soil drying. Water is often required to penetrate deeply into clay. Therefore, watering of narrow trenches cut into the clay or of regularly spaced, shallow boreholes is preferred to hosing off the soil surface.

Where the damage is more severe, a vertical cut-off wall may be installed beside the building to help redistribute and equilibrate soil moisture conditions. The cut-off wall consists normally of a thin (150mm) but continuous concrete wall to a depth of at least 1.5m (see Figure 8). The actual design depth is governed by particular site conditions. The services of a geotechnical engineer may be required to estimate this depth.