15
Load-bearing Masonry R J M Sutherland
FEng, BA, FICE, FIStructE Contents 15.1
Introduction
15/3
15.2
Material properties
15/3
15.3
Codes of practice
15/3
15.4
Limit state principles
15/4
15.5
Unreinforced masonry 15.5.1 The mechanism of failure in compression 15.5.2 Slenderness 15.5.3 Eccentricity of loading 15.5.4 Concentrated loads 15.5.5 Lateral loads on masonry panels 15.5.6 Stability and robustness 15.5.7 Accidental forces
15/5
15.6
15/5 15/8 15/8 15/9 15/9 15/10 15/11
Reinforced and prestressed masonry 15.6.1 General 15.6.2 Structural performance of reinforced masonry 15.6.3 Uses for reinforced masonry 15.6.4 Durability of reinforced masonry
15/12 15/13 15/13
15.7
Dimensional stability of masonry
15/13
15.8
Application of masonry and scope for future use 15/13 15.8.1 High-rise (small-cell) residential buildings 15/13 15.8.2 Low-rise (large-cell) buildings 15/14 15.8.3 Boundary walls 15/14 15.8.4 Retaining walls 15/14 15.8.5 Bridges 15/16
15.9
Conclusions
15/17
15.10 Acknowledgements
15/17
References
15/17
This page has been reformatted by Knovel to provide easier navigation.
15/11 15/11
15.1
Introduction
In this chapter the word 'masonry' has been used to describe either brickwork or blockwork as well as natural stone. Today, natural stone is seldom used except as a decorative facing and the advice which engineers need on masonry relates primarily to brickwork and concrete blockwork. The use of brickwork in the UK has changed quite markedly in the 20 to 25 years following its rebirth as a structural material in the early 1960s. Full exploitation of its strength for slender load-bearing walls in high-rise flats has now been halted by the sharp social reaction against this form of building. In the domestic field what once looked like the greatest justification for 'engineered' brickwork has given way to the limited demands of traditional housing. However, there are now new challenges at least as great as those of high-rise housing. The structural use of concrete blockwork largely dates from the 196Os and its fortunes have followed the same path as those of brickwork. One of the new structural challenges for masonry lies in the construction of buildings for sport, education, manufacturing and storage. Here the economy of masonry is being used increasingly in unframed buildings, often single-storey, with larger spans than in domestic construction, taller walls and few partitions or returns to brace the whole assembly. As a result of these changes, today's engineering problems with masonry in buildings are largely wind resistance, overall stability and composite action with floors and roofs. Crushing strength takes second place. Another field where masonry is finding increasing favour is in the cladding of framed construction especially in large industrial units built in steelwork. In this case not only are there problems of the lateral strength of large thin panels but there are complex questions of movement and of the compatibility of the different materials. All these are very much engineering problems and not matters of architectural opinion. In civil engineering, the once dominant place of masonry was taken about a century ago first by mass concrete and then by reinforced concrete. Concrete may be more in keeping with a mechanized age than a labour-dominated material like masonry but its appearance is increasingly being criticized and now doubts are arising as to its durability, especially when reinforced. What is more, with growing appreciation of the structural performance of masonry, especially when reinforced or prestressed, concrete has a rival both in slenderness and loadbearing capacity. This makes masonry particularly attractive for structural use in retaining walls, bridge abutments and other civil engineering works, particularly in areas which are visually sensitive. There is also a good case for the revival of the masonry arch. Engineers need to keep in touch with developments in the use of masonry. Today it is not just a craft material for houses or decorative facings, as was thought 30 years ago, but a major structural element and one benefiting increasingly from engineering understanding.
15.2
Material properties
Before embarking on any structural design in masonry it is important to distinguish between the physical properties of the different materials of which the units are made and to appreciate the limitations of each. Table 15.11 shows the types of masonry units normally available with their materials, sizes, unit strengths and approximate share of the UK market. It also gives the numbers of the current British Standards which define the acceptable quality of each.
Table 15.2 gives some indication of the dimensional stability of each type of masonry unit, i.e. its response to changes in temperature, load and moisture content. Equivalent figures are also given for other materials commonly used in construction. The most essential factors to note are the initial moisture movements: (1) Clay units are fired at a high temperature and expand, for the most part irreversibly, as they take up moisture from the atmosphere. The expansion is greatest immediately after firing but continues at a diminishing rate for effectively about 10 to 20 years. (2) Concrete units (bricks or blocks) are cast wet and shrink as they dry out, again largely irreversibly. The shrinkage is greatest immediately after casting but continues at a diminishing rate for effectively about 10 to 20 years. (3) Calcium silicate bricks, which are of sand and lime, hydrated, pressed and then autoclaved, behave similarly to concrete units. Not only are the initial moisture movements generally greater than any subsequent cyclic ones due to change of atmospheric conditions, but those of clay and concrete are of comparable magnitude and in opposite directions. This simple distinction between the behaviour of clay and concrete has frequently been ignored in the past with results which have sometimes caused major disruption. Today, now that the different properties of the materials are better understood, there is a tendency to over-react to the problems of movement and sometimes to take precautions which are unnecessary and could even be harmful. The question of precautions against movement is discussed in section 15.7.
15.3
Codes of practice
In the UK, the accepted guidance on the way in which masonry should be designed is given in the British Standard Code of Practice BS 5628. The first part of this Code dealing with unreinforced masonry was issued in 1978.2 This part is the successor to the greater part of the earlier code CP 111 and deals essentially with walls and piers. The second part of BS 56283 which covers the structural use of reinforced and prestressed masonry was not published until 1985. It makes good the wholly inadequate treatment of reinforced masonry in CP 111 and also puts prestressed masonry on an 'official' basis for the first time. This part of the Code covers the design of all types of spanning structures in masonry as well as walls and piers. The third part of BS 5628,4 also published in 1985, gives advice on various aspects of detailing with masonry and on workmanship, durability and similar topics. It could be said to be more architecturally slanted than the first two parts of this Code and is the successor to the earlier Code CP 121. Since the issue of all three parts of BS 5628, masonry in the UK has been on a parallel basis to concrete in up-to-date and officially recognized design methods. This does not mean that all an engineer needs to know about masonry is in the three parts of this Code - far from it. However, anyone designing masonry structures, in the UK at least, should be aware of the contents of this Code and, whether experienced in masonry or a newcomer to it, would do well to consult the handbooks to Parts 1 and 2. References to these handbooks and to a selection of other publications on the structural design of masonry are given at the end of this chapter.5-6 While BS 5628, together with the relevant material standards, will be used as anchor points for the advice in this chapter, this Code should be used for checking design rather than as a
Table 15.1 Types of masonry units normally available. Material and manufacture
Clay brick (BS 3921)
Clay fired generally at > 100O0C to achieve ceramic bond
Calcium silicate brick (BS 187)
Sand and lime; hydrated, pressed and autoclaved
Concrete (BS 6073)
Aggregate and cement hydrated and moulded with pressure and vibration
Aggregate Concrete block (BS 6073)
Autoclaved (aerated) concrete block (BS 6073)
Cement and ground sand or PFA with aerating agent hydrated and moulded in large blocks and then cut
Normal (actual) dimensions of unit (mm)
Standard 215 x 102.5 x 65 high Metric modular (small demand) 190x90x65 (BS 6649)
Varies widely: length 390590 height 140290 thickness 60-250
Type of unit
Characteristic strength of unit (N/mm2)
Approximate share of UK market 1985 (106m2of wall)
Range in British Standard
Range commonly used
7-100
14-100
58*
Solid, or frogged
14^8.5
20.5-34.5
1.75
Solid or frogged
7-40
7^0
4.5
Solid or hollow
2.8-35
3.5-21
Dense 30.4 lightweight 22
Solid only
>2.8
2.8-7.0
23
Solid, frogged or perforated
* Equivalent based on 102.5 mm wall
starting-point. The wide variety of forms of structural brickwork and blockwork make their use even more of a design matter, needing individual judgement, than almost any other material. It is the aim in this chapter to point to these design aspects and to emphasize both the great opportunities for the use of masonry and some of the pitfalls, rather than to provide another handbook to the BS Code. Reference is made throughout this chapter to BS 5628. Readers working in countries other than the UK will need to be aware of the local codes, which may differ quite markedly from BS 5628. The following information may be helpful in this respect. US The most widely used code in the US is the Uniform Building Code. Chapter 24 of the 1985 edition deals with masonry on a linear elastic (working stress) basis. Canada The current code CAN-S304-M84 issued by the Canadian Standards Association covers both design by rules and design by full engineering analysis. This is still on a working stress basis. A limit state code is planned for 1990.
Australia A unified code incorporating AS 1640-1974 (SAA Brickwork Code) and AS 1475 (SAA Blockwork Code, Part 1: unreinforced blockwork and Part 2: reinforced blockwork) is about to be issued. This is written in ultimate strength format and will be converted to a full limit state form in the next edition. New Zealand References to existing codes may be misleading but two new codes are in draft DZ 4229 for masonry not requiring specific design and DZ 4210 for designed masonry. The information given above is considered as a starting-point only. Readers are advised to check directly with the appropriate authority in each country.
15.4
Limit state principles
The design guidance in BS 5628: Parts 1 and 2 for unreinforced, reinforced and prestressed masonry follows the same limit, state principles, with partial safety factors, as are used with reinforced
Table 15.2 Dimensional stability of masonry compared with reinforced concrete and steel
Clay brickwork
Coefficient of thermal expansion (per 0C XlO" 6 )
Movement as result of 2O0C change (%)
Unrestrained drying shrinkage (partly reversible) (%)
5-8
0.010-0.016
4-26 Shrinkage of Depends on mortar type of clay allowed for in and firing expansion temperature. Probably figures (right) 0.02-0.12%. Precise
Unrestrained moisture expansion (%)
Elastic modulus (kN/mm2)
Creep with time: creep factor = final strainj elastic strain (for stress ^0.5 x uh.) 1.2-4.0
figures
uncertain. Too few longterm tests 8-14
0.016-0.028
0.01-0.04 (BS limit 0.04)
-
14^-18
Approximately 2.5
Aggregate 6-12 concrete blockwork*
0.012-0.024
0.02-0.06 (BS limit 0.09 maximum)
—
4-25
2.0-7.0
Approximately 8
0.016
0.02-0.09 (BS limit 0.09 maximum)
-
1.5-4.0
No test results available
Reinforced concrete
7-14
0.017-0.028
0.02-0.10
15-36
1.0-4.0
Steelwork
Approximately 12
0.024
Calcium silicate brickwork
Aerated concrete blockwork
;
175-210
* Concrete bricks similar
and prestressed concrete. Most engineers in the UK are now familiar with these principles but, regrettably, there is still a lack of uniformity in the terminology used in the different BS material codes. In BS 5628, the phrases 'design load' and 'design strength' are used to denote the factored loads and strengths which need to be compared to check adequacy. Thus, for the ultimate limit state, if yf is the partial factor of safety for loading and ym is the partial factor of safety for material strength, adequacy is achieved if: Ultimate (characteristic) strength ^ characteristic load x 7f In all cases (direct load, bending, shear, etc.) the partial factors of safety are expressed separately in BS 5628 and never lost within the characteristic values quoted. The same principle and the same terminology are used in BS 5628 for the serviceability limit states and for precautions against disproportionate collapse following a major explosion or other accident, but in these cases the partial factors of safety are different. Table 15.3 shows the principal factors for each limit state and how these compare with the factors of safety used in BS 81107 for concrete. With unreinforced masonry the serviceability limit states of deflection and cracking are seldom if ever relevant but,
when considering the behaviour of reinforced or prestressed masonry sections in bending, they can be vital.
15.5
Unreinforced masonry
15.5.1
The mechanism of failure in compression
Rather than just accept the characteristic strengths and 'Code' factors of safety for masonry, designers are advised to consider what influences its strength and to try to visualize the actual mechanism of failure. Table 15.4 lists some of the major factors affecting the strength of a masonry wall. The mechanism of failure seems to be generally agreed. Because the mortar is almost always weaker than the masonry units it tends to be squeezed out of the joints. This movement of the mortar is restrained by the bricks or blocks, which are thus subjected to lateral tensile stresses which lead first to splitting and finally to collapse. This mechanism is shown diagrammatically in Figure 15.1. Even under absolutely uniform downward loads, masonry walls - brick or block - fail first due to vertical splitting. This is virtually universal. With brickwork, the wall strength averages about 0.3 times (0.15 to 045 times) the brick strength and with
Table 15.3 Partial factors of safety (material) for masonry compared with concrete (BS 5628 and BS 8110) BS 5628: 1978, Part 1
( Unreinforced masonry)
Y mm (compression)
Control level Manufacturing Manufacturing Manufacturing Manufacturing
Ultimate limit state
Accidental damage
2.5 2.8 3.1 3.5
1.25 1.4 1.55 1.75
y mv (shear)
2.5
1.25
7 m (wall ties)
3.0
1.5
2.0 2.3
1.0 1.15
Y mv (shear)
2.0
1.0
Y mb (bond to steel) 7 ms (steel reinforcement)
1.5
1.0
1.15
1.0
1.0
1.5
1.3
1.05
and site special normal and site special special and site normal and site normal
Serviceability limit state
Notes
BS 5628:1985, Part 2 (Reinforced and prestressed masonry) Y mm (compression)
Control level Manufacturing and site special Manufacturing normal and site special
1.5 1.5
No shear reinforcement assumed
BS 8110:1985 Parts 1 and 2 (Concrete) Y m (compression or bending
1.25
Y m (shear) Y m (bond to steel)
-
1.15
Y m (steel reinforcement)
1.05 1.05
1.4 1.0
1.05
Note: Partial factors of safety for load (y f) with masonry similar to those for concrete (basically 1.4 for dead load and 1.6 for superimposed load with variations for combinations and different limit states)
Mortar squeezed out by vertical pressure but restrained laterally by brick/block
Lateral tensile stress in brick/block balances restraint on mortar but eventually splitting takes place Figure 15.1 Simplified failure mechanism for vertical loads on masonry
concrete blockwork it averages about 0.8 times the block strength.1 The difference between the apparent performance of bricks and blocks in walls compared with their individual strengths is primarily due to the shape of the units. The 'cube' strength of the concrete in the blocks is generally well below the equivalent strength of the fired clay in bricks but, when tested as units, the taller blocks fail at a stress nearer to that in a wall than the squat bricks, which are more fully restrained laterally by the platens of the testing machine. This is shown in Figure 15.2. The logical climax is that a storey-high unit should fail at the same load as the wall into which it is built. Figure 15.3 shows the relationship, as given in BS 5628, Part 1, of the characteristic compressive strength of different types of masonry to that of the individual masonry units. This follows the principles outlined above. Other important factors affecting the capacity of a wall or pier to support vertical loads, apart from those shown in Table 15.4, are slenderness, eccentricity and concentration of loading.
Table 15.4 Major factors influencing the strength of masonry walls Effect on wall strength
Variable factor Brickwork
Concrete blockwork
Strength of masonry unit
Most dominant factor: wall strength proportional to square root of brick strength
Most dominant factor
Strength of mortar
Not very significant: wall strength proportional to cube Little effect on wall strength root of mortar strength for middle range of brick strengths
Thickness of mortar bed
Fairly critical: 17 mm bed instead of 10 mm gives 30% reduction; with ground faces and no mortar, wall strength approaches brick strength
Geometry of masonry units
Ratio of wall strength to brick strength little affected whether wirecut, deeply frogged or perforated
No experimental data; effect probably less significant than with bricks Ratio of wall to block strength about 0.8 Ratio of wall to block strength reduced to about 0.5 with normal bond because cross-webs do not line up; higher with stack bond
Bond
English (50% crossbonded material) Flemish (33% crossbonded material)
Poor filling of bed joints
No noticeable difference in strength
Stretcher (100% crossbonded material)
Up to 40% stronger than English or Flemish
Collar j ointed (steel ties only between skins of stretcher bond but no cavity)
1 0- 1 5 % weaker than English or Flemish
Tests show 30% reduction in strength common, and more possible
Poor filling No reduction found in tests even with wholly of perpendicular unfilled perpendiculars joints
Seldom used other than in stretcher bond (or in stack bond with reinforcement in horizontal joints)
No equivalent tests Effect of not filling perpendicular joints at all is small
Source: Sutherland (1981) 'Bride and block masonry in engineering'. Proc. Instr. Civ. Engrs, 70, Table 2.
Brick shape
Characteristic compressive strength of masonry (N/mm^)
Block shape Figure 15.2 Effect of platen restraint on failure of bricks and concrete blocks under test
Figure 15.4 Maximum slenderness of wall (BS 5628)
Compressive strength of individual masonry unit: N/mm2 Figure 15.3 Relationship of characteristic compressive strength of masonry to the compressive strength of the units (BS 5628) 15.5.2 Slenderness Tests on walls in recent years have shown that slenderness is less of a problem than it was thought to be 30 years ago and, in successive British Codes, reductions in load-carrying capacity for slenderness have tended to become smaller. Reduction factors for slenderness tabulated in BS 5628 permit walls with a slenderness (effective height/thickness) of up to 27. Thus, for instance, allowing for end fixity a half-brick (102.5mm thick) wall could be 3.7 m high and still carry 40% of the axial load appropriate for a low-height wall of the same thickness. Such a slim wall, as shown in cross-section in Figure 15.4, must lack robustness and most designers would prefer not to extend slenderness to this limit even with purely axial loading. 15.5.3 Eccentricity of loading Simple eccentricity can be dealt with conservatively by using the appropriate reduction factor for slenderness and eccentricity in BS 5628 and assuming that the eccentricity at the top of any wall reduces to zero at the next level of lateral support below as shown in Figure 15.5(a). In practice, the eccentricity at the bottom of the wall is likely to be as in Figure 15.5(b) which is normally more favourable as it tends to counteract any further eccentricity applied at that level. As an alternative to this simple procedure, the whole structure can be analysed rigorously as a frame.
Figure 15.5 Line of thrust due to eccentric loading: (a) simplified assumption on line of thrust (BS 5628); (b) more likely line of thrust Unless loaded heavily enough to provide fixity as in Figure 15.6(a), the rotation of the ends of floor slabs can cause unsightly cracking as in Figure 15.6(b). This is particularly likely with long-span concrete roofs or upper-floor slabs supported on masonary walls. In such locations the cracking can usually be eliminated, or at least controlled, without over-stressing the masonry, by allowing rotation as shown in Figure 15.6(c).
Splitting caused by vertical load from end of beam
Splitting dangerously accentuated by horizontal force due to shrinkage or thermal movement
(b) Beam ends carried well back along supporting wall Figure 15.7 Concentrated loads on ends of walls or on stiffening piers shown in Figure 15.7(a). It is something which is not explicitly covered in most Codes of Practice but which designers should consider. With long-span beams especially, the bearing should be carried back as far as practicable or a reinforced padstone should be used as indicated in Figure 15.7(b).
Soft pack
Figure 15.6 Avoidance of cracking of masonry walls due to rotation of floor slabs at supports, (a) Slab rotation resisted by couple Wx. No cracking of masonry, (b) Load W not great enough to resist slab rotation. Upper wall lifted by rotation of slab and crack induced at point (C). (c) Load Wnot great enough to resist slab rotation, but soft pack as shown permits rotation without cracking of masonry walls 15.5.4
Concentrated loads
Traditionally, the need to spread concentrated loads from columns or the ends of beams has been met by using padstones of greater strength than the basic masonry of the walls or by laying local courses of stronger brick or stone. Today, this practice has been formalized in rules such as those in BS 5628. The problem is one of splitting and is most serious at the ends of walls or at piers where the splitting is most likely to lead to failure. Such splitting could also be caused - or accentuated - by horizontal forces due to thermal or other movements. This is
15.5.5 Lateral loads on masonry panels The design of masonry walls so that they can be shown to resist wind forces is one of the major areas of doubt in the treatment of the material. The uncertainty is greatest with panel walls with no vertical load except their own weight, but it is present with most thin or lightly loaded walls. Following an extensive series of tests by the British Ceramics Association8 some partly empirical and partly theoretical guidance has been incorporated in BS 5628, Part 1. This can be used to demonstrate adequacy in many of the most common situations but with vertical spanning in particular it is restrictive compared with what has been common practice for many years. Where it can be shown to be appropriate, design for arching is a very good method of proving lateral stability; it, too, is recognized in BS 5628. Real walls do sometimes blow out, or over, and those which do always seem to have a slenderness or lack of restraint far outside recommended limits. Some unlikely 'successes' may well be due to arching - even unexpected arching - and some to much higher tensile strengths in mortar joints than those generally assumed. In many cases the full 'Code' wind forces may never have occurred and may never occur in the future. Much research is still being carried out on the resistance of masonry panels to lateral loads and it is hoped that further guidance may be given in future amendments to BS 5628. In the meantime, it is worth keeping in touch with the results of this research. Two points of detail which designers would do well to remember are these: (1) Whatever the published bending strengths, bending in a horizontal plane as in Figure 15.8(a) is much more reliable
Failure depends on breaking of the masonry units or shearing of the bed joint mortar
Failure depends only on the tensile strength of the bed joint mortar
Figure 15.8 Resistance of masonry panels to lateral loads: (a) Strength in bending in horizontal plane; (b) greater and more certain than in vertical plane in practice than in a vertical plane as in Figure 15.8(b). Tensile strength across bed joints can easily be disrupted especially during construction after the mortar has set but before it has gained its full strength. (2) The edge support to panels will often be altered - sometimes for better and sometimes for worse - by the deformation of the structure, as can be seen from Figure 15.9. It is important to consider such possible movements and to make sure that adequate fixings are used at all edges assumed to provide restraint, even if there are relative movements due to deflection, shrinkage or settlement. 15.5.6 Stability and robustness The great bulk of the advice in the world's masonry codes is devoted to stresses and compressive strength with only passing exhortations to consider stability. This seems curious when one considers that almost all failures of masonry structures - few in practice - have been due, not to overstressing, but to some form of instability. The reason may be that attempts to codify stability, while helpful in some circumstances, have tended to
Worse than planned: creep deflection of slab at level A reduces panel to one way span
Figure 15.9 Possible effects of deformation on lateral support for masonry cladding lead to anomalies, unreasonable restrictions, or even new dangers, in others. Stability and robustness are best seen as design matters. They need thought rather than rules. Stability is a particularly important factor with masonry because of its low tensile strength. Unless reinforced or prestressed, masonry should generally be planned to rely for stability on gravity forces and on the friction induced by such forces. With today's thin walls, this stability depends on the interaction of walls, floors and roofs. A lateral force acting on the face of a wall, such as that due to wind, is transferred to floors and roofs, which act as rigid horizontal plates and in turn transfer the force to the ground through shear walls running in the direction of the force. This is shown diagrammatically in Figure 15.10. Designers should always follow the forces through this route to make sure that all the connections are adequate, that the floors and roofs are stiff and strong enough and that the 'racking' shear strength of the shear walls is great enough. Even if the structure is fully sheltered from wind it is important to consider a nominal lateral force acting in any direction and to
Roof braced in either horizontal planes (B) and (C) or on slope (A) and horizontal plane (B) to form effectively horizontal plates
Critical connections, or joints, to transfer force to the ground Laterally loaded wall to horizontal plates: Joints 5 and 6 Horizontal plate to shear walls: Joints 1,2 and 4 Shear wall to lower horizontal plate: Joint 3 Figure 15.10 Diagrammatic representation of transfer of disturbing force on masonry structure to provide stability
Better than planned: creep deflection of slab at B loads panel (even say through window) and improves lateral stability
balancing requirements which are often in conflict. This is illustrated in Figure 15.12. There is no virtue in just achieving what appears to be adequate stability if an unnoticeable change can make this much more certain. Accidental forces and material defects are largely unpredictable. 15.5.7 Accidental forces Design to allow for accidental forces is just a particular case of design for stability. Rules introduced following the partial collapse of the large panel concrete structure at Ronan Point in 1968 have largely blinded engineers to the broad issue of accidental forces. These rules, which apply only to buildings of five storeys and above, were made to guard against gas explosions and do not in themselves ensure safety against all hazards. The thinking was rightly to limit damage but the tying forces introduced in the structural codes to satisfy the rules may in some cases actually spread the damage. This is illustrated in Figure 15.13 where continuous vertical ties could actually cause progressive collapse as was shown in tests on a quarter-scale model at the Building Research Establishment.9 Accidental damage can be limited by planned sacrifice or by greater structural strength. The choice is a design matter although the solution must satisfy the broad functional requirements of the Building Regulations in most practical cases.
Figure 15.11 Relative strengths of three possible methods of connecting on to a shear wall
15.6 Reinforced and prestressed masonry
follow this to the ground. British Standard 5628 recommends a nominal lateral force of 1.5% of the dead load above the level being considered. A larger force may sometimes be preferable. One of the most difficult questions with masonry is the assessment of the strength of connections. Figure 15.11 shows three means of connecting on to a shear wall. Designers would do well to opt for the strongest connection which is compatible with cost and other requirements rather than accept the lowest level of apparent adequacy. There are few absolutes in design for stability. The skill lies in
15.6.1 General Much of the previous section on unreinforced masonry still applies but, when reinforced or prestressed, the scope for the use of masonry is greatly extended. Reinforced or prestressed beams can be made, as well as walls, following exactly the same general principles as those used with concrete. This has been demonstrated in tests and in practice.10 The question to be decided is when such forms are advantageous, and this can be done only with some knowledge of the possible structural performance of
Least sturdy
Most sturdy
Increasing sturdiness of floors
Increasing sturdiness of walls
Most sturdy
Least sturdy
In situ two-way reinforced concrete Composite precast (two-way) Precast with lateral and longitudinal ties Timber board and joist with staggered joints Precast planks with no lateral ties and no continuity or tie bars at supports Flooring options
Wall layout options Figure 15.12 Comparison of wall and floor options for simple masonry cross-wall construction
Design resistance moments Brick (BS 5628:Part 2) Mortar (i) : special mfg. control
Explosion
Unit strength2 (N/mm )
Concrete (BS 8110:1985)
DESIGN P.M. Md (N. mm)
Figure 15.14 Comparison of bending strength of reinforced brickwork and reinforced blockwork
Figure 15.13 Diagrammatic representation of how simple code recommendations can spread accidental damage, (a) Horizontal ties broken by explosive force. Continuous vertical ties (as codes) keep well intact which bows out dropping ends of several floors above and below, (b) Independent staggered ties would allow local sacrifice and confine damage
reinforced or prestressed masonry. It is convenient, in this context, to compare the properties of masonry with those of concrete. 15.6.2 Structural performance of reinforced masonry Figure 15.14 shows the relative bending strengths of reinforced brickwork (BS 5628, Part 2) and reinforced concrete in accordance with BS 8110. It can be seen that with moderate brick strengths the bending capacity of brickwork matches that required for most reinforced concrete and that, even with the lowest brick strength likely to be used (20 N/mm2 unit strength), there is a very useful level of bending strength. A similar
comparison can be made with concrete blockwork although the highest practical bending strength with blockwork is not so great. With shear, the comparison is not so favourable to masonry as it is with bending. Figure 15.15 shows - again for brickwork that without shear reinforcement the shear strength of all strengths of brickwork is well below that of reinforced concrete; the same is true of concrete blockwork. It is worth noting that the shear strength is virtually independent of the strength of the masonry units, being dependent on the tensile strength of the mortar and its bond to the units. One advantage of prestressing over reinforcing with any material is that the prestress reduces the principal tensile stress and thus increases the shear capacity. With masonry this is a very real advantage but, curiously, no reference has been made to it in BS 5628, Part 2. The omission does not mean that designers cannot take advantage of this property of prestressing.
Design shear strength Brick (BS 5628:Part 2) Unit strength2 (N/mm )
Design shear strength2 (N/mm )
In some cases with short shear spans this can be increased to a max. of 0.87
Concrete (BS 8110) Char, cube Design strength2 strength2 Notes: (N/mm ) (N/mm ) (1) No shear 'cu steel (2) Shear strength varies with proportion of main steel
Figure 15.15 Comparison of shear strength of reinforced brickwork and reinforced concrete
It has been an accepted feature of prestressed concrete for at least 40 years. Compared with reinforced masonry, there has been little practical experience with prestressed masonry and there is a school of thought which considers that it would have been best to omit it altogether from BS 5628 at this time. Nevertheless, it has been used successfully, as is shown later in this chapter, and may be used more in the future. 15.6.3 Uses for reinforced masonry Not only is the shear strength of masonry low, but it is very difficult to incorporate shear reinforcement in it to increase this. For both these reasons, masonry compares unfavourably with concrete for use in beams except possibly in special cases such as that of a deep beam within the plane of a wall. However, reinforced masonry comes into its own in laterally loaded walls where the low shear strength is seldom a major problem. Masonry is essentially a wall material, or one for arches and vaults. Traditionally, it used to depend for stability on its mass but today, with reinforcement, laterally loaded walls can be made of comparable slenderness to those in reinforced concrete. Further, while concrete walls, if visible, are increasingly being faced with masonry, reinforced masonry has its elegance built into the structure. Figure 15.16 shows a number of ways of reinforcing masonry walls using standard bricks and blocks.
Vertical-spanning reinforced masonry
Modified Quetta bond* Grouted cavityt quetta bond* Bars set in Bars normally Bars set in fine fine concrete set in mortar concrete or 'grout' Brick Brick Brick (or concrete block) Used in First 225 mm wall also air-raid possible with developed for earthshelters bars in mortar quake resist- (stainless steel ance advisable if exposed)
Pocket wall* Bars set in normal concrete Brick Mainly used for earth retaining walls
Filled hollow block* Bars set in fine concrete Concrete block Widely used in the US
* Essentially vertical span but secondary horizontal reinforcement often used in bed joints t Full two-way span possible (or horizontal only). (Grout is the term used in the US for fine high slump concrete.) HORIZONTAL-SPANNING REINFORCED MASONRY Normally used for light loads only (except grouted cavity type). Reinforcement normally small bars or mesh set in bedding mortar. Equally suitable for brickwork and blockwork. Figure 15.16 Typical methods of reinforcing masonry walls
15.6.4
Durability of reinforced masonry
As with reinforced concrete, durability is a factor which is receiving increasing attention today. British Standard 5628, Part 2 gives clear and full recommendations on cover to normal
carbon steel and to galvanized reinforcement in masonry for different levels of exposure. With stainless steel reinforcement there is no need for any cover to the steel specifically for durability. Although the supply cost of the steel is several times that of normal carbon steel, the percentage extra on the whole project tends to be very small once fixing and all the other costs of the construction are included. Stainless steel is being used increasingly for ties and fixings in masonry as well as for reinforcement and is no longer the exorbitantly expensive material it used to be.
15.7
Dimensional stability of masonry
The order of unrestrained movements of clay and concrete products is shown in Table 15.2. Because of the restraints which exist to some extent in all real buildings these movements tend to be less than one would calculate from the tabulated figures. The vital question is how much less, and the answer must be that at present we do not know. British Standard 5628, Part 3 recommends an allowance of 1 mm of movement per metre for clay brickwork, with movement joints a maximum of 15 m apart. For calcium silicate bricks joints are recommended at 7.5 to 10m and at 6m for concrete bricks and blocks. It is worth remembering what these joints are for. In the case of clay brickwork they are needed primarily for expansion and thus, to be effective, must be wide enough and filled with something soft enough to allow the expansion to take place. In the case of calcium silicate bricks and all concrete units the joints are needed mainly for shrinkage and effective sealing becomes most important. As in the case of movement joints in concrete structures, there is some evidence that the recommended cure for movement problems has not always been effective. The cure may even introduce new problems of maintenance. A recent study for the Construction Industry Research and Information Association (CIRIA)11 showed that, both for clay and concrete units, adherence to the spacing then recommended in CP 121 failed to eliminate noticeable defects in a significant number of cases, while in quite a large proportion of other cases no noticeable defects were found even in walls well beyond recommended limits of unbroken length. The problem of movement is too complex for simple rules. What is more, in many cases it may be most economical, and in the long term most satisfactory, to reduce the number of joints, risk some cracking and repoint after a few years. Designers would do well to study case histories, observe real buildings and then try to recognize the situations where movement may be serious and those where damage, if it occurs, is only of a cosmetic nature. Figures 15.17 and 15.18 show some key factors but these should only be considered as examples.
15.8 Application of masonry and scope for future use 15.8.1 High-rise (small-cell) residential buildings Unreinforced masonry has formed the sole vertical support to residential buildings of up to at least 18 storeys while with vertical reinforcement it has been used in blocks of over 20 storeys even in seismic zones. In most cases the design has been dominated by the need for resistance to lateral forces and the assessment of interaction between floors and walls to achieve this. Perhaps surprisingly, tension has often proved more of a problem than compression.
HORIZONTAL MOVEMENT (CLAY UNITS) Movement may cause sliding on d.p.c. (often scarely noticeable.) Movement may cause cracking (if not cyclic can be repaired locally) Restraint at d.p.c. will resist sliding but may accentuate tendency to cracking as in (2) above Short 'returns' in plan a major source of cracking due to expansion (most in need of 'protection' by vertical joints to absorb movement). Movement often reduced or eliminated by restraints of many sorts, but local restraints at movement planes, as due to Crack failure to continue cavity tray DPCs around corners, a frequent Cracking here due to thermal cause of cracking. movements of roof slab can usually be eliminated by sliding joint locally HORIZONTAL MOVEMENT (CONCRETE OR between roof and wall CALCIUM SILICATE UNITS) Concrete With concrete blockwork the slab need for vertical movement Loadjoints is generally greater bearing than with clay brickwork. cross-walls Load-bearing blockwork walls are frequently disrupted by cyclic movements at roof level unless local sliding joints are provided. Marked changes of section as over doorways (A) are a frequent source of shrinkage cracking. Full height door frames as (B) are preferable with concrete brickwork.
Overhang say 5-15mm DETAILX
DPC stops short
Figure 15.17 Typical serviceability problems due to horizontal movements in masonry With the increased development of computer programs in the last 15 years the design of such buildings should be much easier than it was in their heyday in the mid 1960s. There are also some signs of a revival in the popularity of high-rise flats and hostels. 15.8.2 Low-rise (large-cell) buildings The challenge of extending the economic use of masonry, as proved in domestic construction, to larger-cell buildings such as sports halls or warehouses has been mentioned at the beginning of this chapter. Here the downward loads tend to be small but the walls may need to span vertically 2 or 3 times as far as in housing. This has been achieved in frameless construction with deep ribs at regular intervals, emphasized architecturally, or by making the walls cellular. In such cellular walls - popularly known as 'diaphragm' walls - the masonry is placed in the most efficient way to resist lateral forces (Figure 15.19a). The walls are usually designed to span from the ground to a roof which is braced to transfer the load to shear walls as already discussed. Sometimes such walls are prestressed with vertical cables either bonded or in voids as shown in Figure 15.19(b), or they can be reinforced. Care is needed during construction to make sure that the walls do not fail in wind before the bracing of the roof is provided. 15.8.3 Boundary walls Masonry has proved itself for boundary walls over a very long
period. Such walls, either of constant section or with wholly inadequate stiffening piers, frequently defy all probability of stability but have given good service for decades or centuries; some such walls survive even in spite of considerable bulges or tilts. Nevertheless, from time to time they do blow over and there is no justification for building unstable boundary walls today. Stability can be achieved by an irregular planform or by reinforcing vertically or prestressing or by any combination of these. Reliance on the tensile strength of the masonry across the bed joints is unwise except on a very small scale. The forms of reinforcement shown in Figure 15.16 can be used or stable walls of very slender sections can be built with special or cut bricks as shown in Figure 15.20. The use of stainless steel is advisable in most boundary walls because of their extreme exposure. Shear strength is virtually never a problem with boundary walls. 15.8.4 Retaining walls Reinforced masonry has proved to be particularly suitable for retaining walls, either on a small scale associated with housing or on major civil engineering works. It is hard to understand why it has not been used more. Reinforced concrete walls are often faced in masonry for appearance. Why not use the masonry structurally? The answer to this must be that it is largely habit which prevents engineers from thinking of reinforced masonry, or fear which leads to unduly high pricing of
Original level Clay brick outer skin Concrete block inner skin Brick slips Outer skin supported on nib
Brick strong, stiff and with high expansion Load largely transferred to, and carried by, brickwork.
Brick weak and compressible with low expansion Brickwork compressed and load, initially transferred to brick, largely returning to frame. Overall vertical contraction.
Overall expansion possible but unlikely No disruption likely MASONRY WITHIN FRAME
SECTIONAL ELEVATION ASBUILT (Concrete frame and cavity cladding in mixed materials)
Concrete nib
Concentrated and eccentric force on outer skin prior to disruption
Frame and concrete blockwork both shrink while brickwork expands. Most vertical load shifts - eccentrically - on to outer skin only. Forces tend to push off slip bricks, crack nib and bow cladding outwards. Possible severe disruption
As M but cladding all clay brick Behaves as (a) or (b) provided the inner skin can support whole load transferred from frame.
As(c) but cladding all concrete block. All materials shrink and differential movements small.
Disruption uncertain
Disruption unlikely
OUTER SKIN SUPPORTED ON NIBS
Figure 15.18 Relative movements of masonry cladding and concrete frames: diagrammatic representation based on two storeys. Real problems tend to be confined to multi-storey building
the slightly unfamiliar. However, today there is enough evidence of successful construction of retaining walls in masonry to prove that they are easy to build, and perform well. Brick retaining walls have been built successfully in Quetta bond, in grouted cavity construction and using the 'pocket wall' technique. With concrete blockwork the fixing of reinforcing bars in the filled hollows in the blocks has become quite common. The planforms of such walls are shown in Figure 15.16. Unlike boundary walls, retaining walls normally need only resist lateral forces in one direction. Thus, for cantilevers the reinforcement should be as near to the loaded face as possible. The pocket-type of wall is ideal in this respect. Figure 15.21 shows how such walls are formed, with the thickness increasing to match the increasing bending moment and the reinforcement as close to the rear face as practicable. In some cases the steps in thickness have been repeated two or three times. In pocket-type walls the reinforcement is surrounded in dense concrete whose compaction can be checked once the small back shutter to the pocket is removed. Thus, the durability is equiva-
lent to that of reinforced concrete but the compressive force is resisted not by the concrete but by the brickwork. Pocket-type retaining walls have been built in the US with heights of up to 7.3 m.12 There are now quite a few major walls of this type in the UK used for bridge wing walls and similar purposes. One British pocket-type retaining wall approximately 4 m high was monitored for 517 days after which the deflection was only 16mm. As expected, the movement was apparently continuing but tailing off.13 The design of pocket-type walls is covered in BS 5628, Part 2 which deals with concrete cover, pocket spacing and workmanship as well, of course, as with structural design. In some circumstances, the characteristic shear strengths in BS 5628, Part 2 may prove a restriction on the performance of such walls, although tests on actual walls have almost all shown failure in bending. There is a good case for revising the shear clauses in BS 5628 especially in relation to retaining walls. One objection which is sometimes raised to the use of masonry retaining walls on civil engineering projects is speed of building. This objection may or may not be real but with
Wall propped by roof Reinforced or prestressed
Prestressed (cables could be unbonded in the cells but grouted cables built in preferable)
Sectional plan B-B
Sectional plan A—A
Figure 15.19 Typical forms of cellular (diaphragm) wall, (a) Unreinforced (wall spans from ground to roof), (b) Prestressed (or reinforced). Wall acts as vertical cantilever Earth pressure
Sectional plan B-B Sectional plan A—A Stainless bar Detail X Specially moulded brick Standard brick cut Figure 15.20 Thin but stable boundary walls formed with special or cut bricks (concrete blocks also as in Figure 15.16)
Sectional elevation of wall Figure 15.21 Typical pocket-type retaining wall pocket-type walls, prefabrication is a very real possibility. This was demonstrated by a trial some years ago.1 15.8.5 Bridges Masonry bridges have generally proved more durable than
those of iron, steel or concrete. There are thousands of them under roads and railways and over canals which have had minimal maintenance during 100 years or more. They have adapted themselves to settlement without distress and still look attractive. Tests have been made in recent years, both in the laboratory14 and by the Transport and Road Research Laboratory on actual arch bridges. Our understanding of the behaviour of masonry bridges is better today than it ever was but in spite of this we continue to use slab or beam bridges under highways which are subject to corrosion due to rain and de-icing salt. In the future, there seems a clear case for using more masonry arches, possibly combining mass concrete with brickwork. Such arches would be particularly appropriate for medium spans where piped culverts are too small but major long spans are not needed. Useful guidance on the appraisal of existing masonry bridges is given in the Departmental Standard BD 21/84 and the associated Advice Note BA16/84 both issued by the Department of Transport.15 These documents are also relevant, at least in part, to new construction.
15.9
Conclusions
With the publication of all three parts of BS 5628, the UK is probably leading the world in recognized guidance on masonry design. Further, in research, the UK has taken a leading role for 20 years. It is now up to design engineers to make full use of this rediscovered material.
15.10 Acknowledgements Tables 15.1 to 15.3 were first published in the Author's paper1 but have been brought up to date. Figures 15.1, 15.3, 15.16 and 15.21 and parts of Figures 15.19 and 15.21 have been adapted with the permission of Thomas Telford Ltd from those in this paper. Figures 15.10, 15.11, 15.12 and 15.13 have been adapted with the permission of the Institution of Structural Engineers from those already published in the Author's paper.9 The Author wishes to thank those concerned for permission to reproduce this material.
References 1 Sutherland, R. J. M. (1981) 'Brick and block masonry in engineering.' Proc. Instn Civ. Engrs, Part 1, 70, 31-63. [Tables 15.1, 15.2 and 15.3 are updated versions of tables first published in the above paper.]
2 British Standards Institution (1978) Code of practice for use of masonry. BS 5628, Part 1: 'Unreinforced masonry.' BSI, Milton Keynes. 3 British Standards Institution (1985) Code of practice for use of masonry. BS 5628, Part 2: 'Structural use of reinforced and prestressed masonry.' BSI, Milton Keynes. 4 British Standards Institution (1985) Code of practice for use of masonry. BS 5628, Part 3: 'Materials and components, design and workmanship/ BSI, Milton Keynes. 5(a) Haseltine, B. A. and Moore, J. F. A. (1981) 'Unreinforced masonry.' In: R. G. D. Brown (ed.) Handbook to BS 5628: structural use of masonry, Part 1: Unreinforced Masonry Brick Development Association, Windsor, (b) Roberts, J. J., Edgell, G. J. and Rathbone, A. J. (1986) 'Palladian.' Handbook to BS 5628, 'Structural use of reinforced and prestressed masonry.' Viewpoint Publication No. 13.028, London. 6(a) Hendry, A. W. (1981) Structural brickwork. Macmillan, London, (b) Curtin, W. G. et al. (1982) Structural masonry designers' manual. Granada, St Albans; (c) Gage, M. and Kirkbride, T. (1980) Design in blockwork, 3rd edn. Architectural Press, London, (d) Orton, A. (1986) Structural design of masonry. Longman, Harlow. 7(a) British Standards Institution (1985) Structural use of concrete. BS 8110, Part 1: 'Code of practice for design and construction.' BSI, Milton Keynes; (b) British Standards Institution (1985) Structural use of concrete. BS 8110, Part 1: 'Code of practice for special circumstances.' BSI, Milton Keynes. 8 West, H. W. H., Hodgkinson, H. R. and Haseltine, B. A. (1977) The resistance of brickwork to lateral loading; Part 1: Experimental methods and results of tests on small specimens and full-sized walls.' Struct. Engr 55, 10, 411^*21. 9 Sutherland, R. J. M. (1978) 'Principles for ensuring stability.' Symposium on stability of low-rise buildings of hybrid construction. Institution of Structural Engineers, 5 July 1978, London, pp. 28-33. 10 Bradshaw, R. E., Drinkwater, J. P. and Bell, S. E. (1983) 'Reinforced brickwork in the George Armitage office block, Robin Hood, Wakefield.' Struct. Engr. 61A, 8, 247-254. 11 Construction Industry Research and Information Association (1987) Movement and cracking in long masonry walls. CIRIA Practice note. (To be published.) 12 Abel, C. R. and Cochran, M. R. (1971) 'A reinforced brick masonry retaining wall with reinforcement in pockets.' In: H. W. H. West and K. H. Speed, (eds) SIBMAC Proceedings, International Brick Masonry Conference. British Ceramic Research Association, Stoke-on-Trent. pp.295-298. 13 Maurenbrecher, A. H. P. (1977) A pocket-type reinforced brickwork retaining wall. Structural Clay Products, Potters Bar, SCP Publication No. 13. 14 Sawko, F. and Towler, K. (1982) 'Load-bearing brickwork: structural behaviour of brickwork arches.' Proc. Br. Ceramic Soc., 30, 7, 160-168. 15(a) Department of Transport (1984) The assessment of highway bridges and structures. Roads and Local Transport Directorate. Advice Note No. BA16/84. HMSO, London, (b) Department of Transport (1984) The assessment of highway bridges and structures. Roads and Local Transport Directorate. Departmental Standard BD21/84. HMSO, London.
