Masonry facades

Forces applied to a material create stress or pressure within that material; the amount of stress depends on the magnitude of the loads. Stress is expressed as force/area and measured in N/mm2.

Under load, most building materials deform to some degree, although the magnitude of that deformation can be very small. The physical extension or shortening of a material under stress is called strain. Strain (e) does not have units as it is a ratio of change of length to original length, but it is sometimes expressed as a percentage or in units such as mm/mm. Because these movements can be very small the term 'microstrain' (me) is sometimes used, which is e x 10-6.

Within a range known as the elastic limit, materials will return to their original size once the loading is removed. Within this limit, the deformation due to load is proportional to the stress induced by the load. Thus, it is possible to express the properties of a material according to its stress and strain ratio or 'modulus of elasticity' (E):

E = stress/strain i.e. load per unit area/deformation per unit dimension

Strain = α´t

Stress = E x α´t

Where,

E = modulus of elasticity

α´ = coefficient of linear thermal expansion

t = temperature range

So, knowing the properties of the materials in question enables not only the change in length to be calculated for temperature and moisture changes, but also the stresses that will be generated by those changes. If the calculated stresses exceed the ultimate tensile strength of the brickwork, cracking will result.

Long-term loadings on materials can produce additional deformations, i.e. movements that do not return elastically. These deformations are termed creep (or sometimes plastic flow) deformations and characterise many standard building components, particularly concrete.

Creep can be defined as the gradual increase in strain with time. All columns and walls, irrespective of the material from which they are constructed, will shorten due to loading from above. In multi-storey construction this can be a significant consideration. Creep will increase the shortening of concrete columns. For most buildings the elastic shortening of steel columns is sufficiently small to be ignored. In timber columns axial creep is negligible and is not normally considered in design (Better Exploitation Of Materials Research: Interfaces & Connections, BRE). The actual level of creep depends upon many factors, including relative humidity, the ratio of surface area to volume, the type and quality of cement, the amount of water present, the stress in the concrete and the age of loading (Further observations on the design of brickwork cladding to multi-storey rc frame structures, Technical Note 9, BDA, 1979). If loading is applied before it has reached its 28-day strength, creep deflections can be appreciably more.

Early framed buildings tended to be of heavy beam or girder construction which, with heavy brick or stone claddings, were able to withstand the effects of gradual shrinkages in the frame without damaging the fabric. However, advances in design led to progressively stronger forms of construction with the ability to sustain higher stresses. Thinner, more highly loaded members were the result, often with increased flexibility. Such forms of construction revealed the effects of the combined problems of creep and shrinkage both axially in columns and horizontally in the form of increased deflections.

Buildings are not static; they move constantly. Concrete frames in particular do not stop moving once they have been completed, but continue to move for many years as a result of the effects of drying shrinkage and creep. In horizontal members the total deflections will be the sum of creep deflections for permanent loads, elastic deflection for transient loads together with the effects of temperature change and shrinkage.

Creep is not peculiar to concrete frames; masonry can suffer as well, although the effects are not so well understood. Research by CIRIA (Design for movement in buildings, Technical Note 107, 1981) suggests that creep in walls seems to be somewhat greater than in columns.

Unless measures are taken to permit movements, the combined effects of all of these forces may generate failures. If a component is free to move without restraint, all well and good. However, in building terms there is usually some form of restraint, whether intentional or not. In an unrestrained condition, deformations will not result in cracking. Where restraint is present, deformations may exceed critical levels for stress or strain with the result that cracking or other forms of movement occur.

In addition to the natural properties of materials, the question of building inaccuracy must not be ignored. In building terms there is no such thing as perfection. It is simply not possible to construct something to the degree of precision that eliminates the need for constructional tolerance. Many materials are natural products that change shape and size, and the practical difficulties of assembly and achieving exactness cannot be underestimated.

However, while a designer might consider the need to provide for elastic and long-term movements in a building, he/she must also take into account reasonable allowances for building tolerance. A designed gap of 15mm to allow for expansion could turn out to be an actual gap of 20mm or 2mm depending on the materials used. So, the effects of expansion in a clay brick facing could become pronounced if a movement joint did not have the required capability.

Creep in steel-framed structures is usually negligible; the same would apply to concrete. However, for reinforced concrete the movements can be significant.

Drawing illustrating the cumulative effects of frame shrinkage over the height of a building

Depending upon the aggregate used, the shrinkage of concrete can be estimated to be between 0.066% and 0.085% (660-850 microstrain). Loading of around 10N/mm2 produces creep of around 300 microstrain. Therefore, over a 3m high storey height the combined shrinkage could be a shade under 2.9-3.5mm. The effects of temperature and moisture must also be considered. A 3m high panel will exhibit around 1.5mm of vertical movement due to expansion plus around 2.7mm for irreversible and reversible moisture expansion. So, expansion is probably greater than contraction of the frame and while a bitumen dpc has some 'give' it will not in itself be able to take up the difference. Estimating movement for joint width design is notoriously unreliable and usually results in overestimation.

Little data is available on actual in-service measurements of building shrinkages. According to BDA Technical Note 9 (Further observations on the design of brickwork cladding to multi-storey rc frame structures, 1979) one American source quotes shrinkage of 1 inch per 100 feet of height - 25mm per 30.4m. As a rule of thumb, 1mm per metre of height would be a reasonable estimation.

BDA Design Note 10 (Designing for movement in brickwork, 1986) provides a list of useful rules of thumb that will cover a majority of situations, while various recommendations are also given in CIRIA Practice Note 44 (Movement and cracking in long masonry walls, Special Publication 44, 1986). Unfortunately these documents provide conflicting advice. BS 5628 Part 3 (Code of practice for use of masonry. Materials and components, design and workmanship, 2005) also gives guidance as to the location of movement joints. Taken together a fairly conservative set of rules of thumb are shown below:

• For restrained walls of clay brick, allow a minimum width of movement joint of 15mm; joint spacing should never exceed 12m.
• For clay brick walls that are lightly restrained or unrestrained (for example spandrel panels or parapets) the distance between joints should not exceed 7-8m.
• For calcium silicate bricks, allow shrinkage control joints at 7.5-9m centres; the aspect ratio of panels must never exceed 3. If greater than 3 provide joints to form smaller panels.

The expansion and contraction of brickwork is similar in both the vertical and horizontal dimensions. Horizontal movements can be reduced by providing bed joint reinforcement.

For cladding (whether brick or other forms) the effects of long-term creep must be taken into account during design. Typical problem areas can be illustrated by an examination of:

• brick support knibs or shelf angles;
• brick slips; and
• masonry infill panels.

Brick support knibs

In multi-storey construction it is usual to support the external leaf of a cavity wall at each storey, usually by means of a shelf angle or (less common in contemporary construction) a concrete knib designed to carry loads back to the structural frame. Without this support, the weight of each storey would be carried by the masonry below it, inducing excessive compressive loads.

Although simple in theory, the design and installation of support angles has, over the years, provided fertile ground for failures and disputes.

The supported brick panels will be subjected to certain movements - expansion or contraction due to moisture and temperature as well as long term creep. The loads so generated must also be taken by the support. However, the structure to which the support is fitted will (assuming it is of concrete) also suffer from long term creep as well as deflections. Expansion of brickwork and contraction of the frame may well work against each other. The forces so generated can be sufficient to shear concrete knibs as illustrated below.

Shear forces resulting in knib failure

In the drawing above, a lack of provision for movement at the top of each masonry panel has also led to load transfer between panels.

The alternative to an in-situ concrete knib is a steel angle, usually of stainless steel to prevent corrosion problems. While it may seem obvious, the angle must be properly supported to reduce deflection and to transfer loads to the frame. However, it is not unknown to find deficiencies with support, and seemingly innocuous evidence of local disturbance in brickwork can indicate serious problems. The case study illustrates a particular set of defects with significant repercussions - the need to rebuild the entire brick skin.

Careful attention must be given to shelf angles, they must be capable of adjustment to make up for construction tolerances. In the case study, the combination of large radius to the angle and deviations in the line of the ring beam meant that the bottom course of brickwork could not fit properly.

Case study

The building is a 1980s office in the south-west of England, comprising around 7,000m2 of accommodation arranged on a ground and 6 upper floors. The frame is of in-situ reinforced concrete. The external walls are clad with clay brickwork constructed in faceted panels and with a corbelled parapet at roof level.

An inspection from ground level revealed isolated cases of brick spalling below horizontal movement joints at each floor level. The movement joints appeared to be pointed with sealant in the usual way and so the managing agent's initial conclusion was that there may be localised bridging of the soft joints with mortar. Although not an unreasonable suggestion, further investigations were undertaken. These revealed a series of potential problems best illustrated in the drawing shown below.

Steel shelf angle detail

The construction exhibited several faults, some of which were the accidental consequence of poor design, while others were more wilful.

• First, the soft joints were found to be nothing more than a layer of mastic sealant applied over a mortar joint, the arrangement permitting load transfer to adjoining panels of brickwork and the immediate cause of the spalling edges of individual bricks.
• However, and more significantly, the angle supports themselves had not been attached to the frame properly. The angle was designed to be bolted back to the concrete at regular centres, but intermediate 'bolts' were found to be nothing more than the head of a bolt carefully cut off and stuck onto the steel with epoxy resin. Deflections in the angle could therefore transfer loads to the masonry below causing overstressing. In one location the brickwork was effectively unsupported to virtually full height.

The inner leaf of the enclosing cavity wall was of lightweight concrete blockwork built off a ring beam at the perimeter of the building. The blockwork was rigidly pointed to the underside of the ring beam so that any shortening of the frame would place the blockwork under compression, potentially causing the inner leaf to bow.

The designer selected a thickness of stainless steel for the angle support that was of unusual thickness. The angle had been roll formed with the radius of the roll being greater than the 50mm cavity width. As a result the bottom course of brickwork would not sit on the angle properly, so the bricks had been hand cut with the edge of a trowel leading to an irregular bedding width. The rough cutting weakened the brickwork and made the bottom course vulnerable to rotation and spalling. Poor positioning of the support meant that the bricks also had to be cut to achieve proper coursing.

Given the severity of the support problems, the only practicable solution was the removal and replacement of the entire outer skin of brickwork.

A prism of cracked brickwork at an external corner resulting from support angles being stopped short of the corner

Problems can occur if the shelf angles do not extend to the full width of the masonry panel and are cut short of the corners of the building. If this happens, the masonry then bears onto both the angle and the masonry below. The forces generated by expansion of the brickwork are then carried partly by the support (and thence back to the frame) and partly by the masonry. This condition creates a shear force which, if greater than the strength of the masonry will cause vertical shear cracks at the corner - a problem that could be confused with expansion movement. Given that the positioning of ties at corners is undesirable, the shear cracks could create a tall, unbonded prism of brickwork to become unstable.