The structural behaviour of horizontally perforated bricks

Ronaldo Bastos DUARTE 1 and Newton Drassy Romeiro da FONSECA 2

ABSTRACT
Horizontally perforated bricks have become the most popular ceramic unit in the construction of buildings in Brazil. Despite all the discussion at the academic environment on the first papers drawing attention for the weakness of the horizontally perforated ceramic bricks, very little has been concerned about the compressive, flexure and shear performance of this type of unit. Collapses of walls and entire buildings causing casualties and deaths are raising concerns and worries of the authorities and the users, but it is still felt there is a lot of research work to be done in order to collect data for the design of a Brazilian code of practice for the use of these bricks. This paper briefly presents the results of the few experimental investigation carried out by different researchers on such bricks. Comparative tests on compression, flexural and shear behaviour between horizontally perforated, vertically perforated and solid bricks units and wallettes show these bricks are too fragile and inadequate to be used as structural unit for brickwork buildings.

KEYWORDS

Compression, Flexure, Shear, Strength.

1 INTRODUCTION

  For a developing country like Brazil structural masonry buildings have become very attractive to builders and consumers due to the economy resulting from the use of local materials and the construction process. Structural masonry buildings are popular, cheap and offers reasonable good housing conditions and have been accepted in all over the country. Therefore, the cellular structural system has been used in the majority of cases, replacing framed concrete structure for residential purposes for up to 4-5 floors buildings. Of course it provides many economic and structural advantages. Nevertheless, until 2005, the Brazilian Code [ABNT 1992] did not make any recommendation for the use of structural units in loadbearing masonry construction. This situation only changed in 2005 [ABNT 2005], when horizontally perforated bricks were recommended only for use in non loadbearing walls. As a result, thousands of buildings were built using the 4-, 6- and 8-hole horizontally perforated bricks and this inadequate practice still remains.
  Reports of accidents only began to appear during the last decade when a serious collapse occurred near Recife causing many casualties and deaths, but, nowadays, more accidents have been related throughout all over the country.
  What makes the accidents in Recife look odd is the fact that the collapsed buildings were 15-20 years old while in the rest of the country collapses use to happen during the final stage of the construction process. Because of that, only in few cases the latter are reported in the media.
  All documented collapses happened in Brazil are related to the use of horizontally perforated bricks as structural units in loadbearing walls [Duarte 2003]. Although these bricks are still widely used there is not enough information available about their mechanical properties, especially about the flexural and shear behaviour. Therefore, in this paper some values obtained from the quite few experimental tests available are presented in order to show how fragile these type of units can be and the danger involved in their use in loadbearing walls of brickwork buildings.

Figure 1. Collapse of one of the blocks of Bosque das Madeiras Buildings (1994), city of Recife [Oliveira and Sobrinho 2005].		Figure 2. Collapse of Érika Building (1999) in the city of Olinda, close to Recife [Oliveira and Sobrinho 2005].

 

Figure 3. Crushing of the lower bricks in a wall. Collapse of a podium three storey residential building in the city of Porto Alegre (1999)		Figure 4. Collapse of a single storey school building in the city of Cacequi (1982).

 

Figure 5. Collapse of a residential building under construction in Porto Alegre (1987).		Figure 6. Collapse of the upper floor of a residential building in the city of Gramado (2005).

  In the tests shown in this paper, the structural behaviour of the horizontally perforated bricks are compared with the solid and the 21-hole vertically perforated bricks, as these units are currently used as structural units in load bearing walls.

Figura 7. Types of bricks. : Bricks (a) and (b) are for structural purposes and brick (c) is only for non-loadbearing walls. Type (c) is also produced having circular and square holes and its structural performance is the main object of this paper.

2 COMPRESSIVE STRENGTH TESTS

  In all cases, the failures shown on Fig 1 to 6 are related to the use of loadbearing walls built with the horizontally perforated bricks similar to that shown in Fig 7 (type c). The weakness of this type of brick in supporting the vertical compressive stress in a loadbearing wall acting perpendicular to the hole direction is shown on Tables 1 to 3.
  On Table 1 compressive tests on horizontally perforated bricks were carried out by Abiko [1984] in order to measure the compressive strength according to the brick face. He tested horizontally perforated bricks with round and square holes. Bricks in rows 1 and 2 were made by the same producer. Similarly, bricks in rows 3 and 4 were produced by another manufacturer (the author did not present the amount of test specimen ). It was found that for the bricks having square holes the failure compressive stress is very similar to directions 1 and 2, because the compressive load is applied perpendicular to the hole directions. Bricks having round holes had an unusual performance on directions 1 and 2 because they also had a very small round hole at the solid centre line (see Fig 8 and 9), affecting the stress distribution at this area, thus reducing even more the load capacity along direction 1. When loading was applied along direction 3 the compressive strength is a lot higher because the compressive stresses follow parallel to the hole directions. This very simple finding was not accounted for the Brazilian Code until 2005, resulting in a large amount of unsafe buildings.
  On Fig 8 and 9 the same kind of compressive tests are shown but this time also performed on wallettes [Roman 1993]. This investigation was carried out into bricks and wallettes specimens, setting the round holes in the horizontal direction (the usual way to lay down these bricks) and in the vertical direction, also in order to test the influence of the loading and the hole direction on the compressive strength. The tests were done on the traditional 4-hole horizontally perforated extruded bricks and wallettes built with these units. The wallettes had different aspect ratio according to the hole direction as they had only 3 bricks each. Tables 2 and 3 show the average results of eight test bricks and five test wallettes and the coefficient of variation (CV). Again it can be seen from the test results that the hole direction has a major influence on the compressive strength of bricks and wallettes. Laid in the usual manner (having the holes along the horizontal direction) the compressive strength is very low. The compressive strength of bricks and wallettes is a lot higher when the bricks are laid having the holes set in the vertical direction.
  It was also noticed that the collapse of the wallettes specimens built with the holes horizontally set is sudden and no further reserve of strength exist after the initial cracking. This may be the reason why frequently people had no time to leave the building safely once the collapse had started causing many deaths (in solids and vertically perforated bricks the first cracks use to appear at about 50-60% of the ultimate load).

Table 1 – Compressive tests along the three main different directions on 8-hole horizontally perforated bricks [Abiko 1984].

 

1.48 1.47 5.90

0.2 1.05 7.44

1.90 1.88 5.27

0.4 0.93 6.48

Compressive strength [N/mm2]

Length

Height

Width

Direction of loading

Dimensions [mm]

Bricks

 

 

Figure 8 Compressive tests on bricks (4-hole horizontally perforated bricks) according to hole direction [Roman 1993].		Figure 9 Compressive tests on wallettes (4-hole horizontally perforated bricks) according to the hole direction [Roman 1993].

   

 

It is reported that only in the city of Recife and surroundings there is about 6,000 brickwork buildings built using horizontally perforated bricks as structural units. This is a very serious problem and the study of alternatives to reinforce these building has become a major research objective in that region.

 

3 FLEXURAL STRENGTH TESTS

  The flexural strength of external brickwork cladding walls has been neglected by most of the Brazilian structural engineers because it could not be related to the reported collapse of entire 4-5 floor buildings. Although there has been an increase in the measured wind velocity in Brazil, it still only has affected the safety of small and popular dwellings houses. But it is important to determine the flexural strength of brickwork because of the action of wind pressure on external cladding walls and external infilled walls of framed construction. Therefore, flexural test were performed on wallettes built with the three different types of bricks in order to compare the influence of the hole direction in the two main directions: parallel and perpendicular to the bed joints.
  The wallettes were built with a 1:1:6 mortar by volume (Portland pozzolan cement:lime;sand) and tested after 28 curing days as simply supported beams laid flat (Fig 10). The imposed load was applied along two lines loading, subjecting the central part of the test wallettes (at least two mortar joints) to an uniform bending moment. The dead weight of the test beams has been included as an additional uniformly distributed load to calculate the modulus of rupture and six specimens of each type were used.

Figure 10. The set up of a flexural test. If the specimen was tested vertically, the dead weight varies from zero to maximum from top to bottom supports, causing some rotational restraint at the bottom. Tested in the horizontal plane this rotational restraint is eliminated.		Figure 11. Failed specimens after flexural tests. It shows specimens 1, 3 and 4 failed along a line parallel to a brick horizontal perforation (not by bond failure at the mortar-brick interface

  Although there is no direct correlation between the flexural strength and the compressive strength of bricks, compression tests were also performed. The average compressive strength of the solid bricks was 47.0 N/mm2 (CV=9.5%); of the 21-hole vertically perforated bricks was 21.2 (CV=14.6%) and of the 6-hole (round) horizontally perforated bricks was 4.6 (CV=10%). The total water absorption by weight was 11.9%, 10.5% and 13.7%, respectively. Again it is possible to see the huge difference between the compressive strength of the solid and of the horizontally perforated test bricks.
  The average, coefficient of variation (CV) and characteristic flexural strength of the wallettes parallel and perpendicular to the bed joints is shown on Table 4.
  From Table 4, it is possible to see that: some test results probably were affected by poor workmanship, producing experimental characteristic flexural strength (fv) lower than the recommended values of BS 5628. Nevertheless, it was observed that four in six horizontally perforated bricks test wallettes, subjected to bending normal to the bed joints (along the vertical direction), failed along a line parallel to a core perforation, and not by bond failure at the mortar/joint interface. This may indicate that the average flexural tensile strength of these bricks in the vertical direction probably were lower than the tensile strength of the mortar/brick bond interface.

 

 

Table 4. Flexural strength of walletttes [Eidelwein and Duarte 1998].

 

Walletes	Plane of failure
	Average flexural strength(N/mm2)	CV (%)	fv (N/mm2)
Parallel to bed joints
Solid bricks	0.21	33.3	0.11
21-hole vertically perforated	0.77	15.6	0.59
6-hole horizontally perforated	0.55	27.3	0.34
Perpendicular to bed joints
Solid bricks	1.43	22.4	0.96
21-hole vertically perforated	1.39	16.5	1.05
6-hole horizontally perforated	0.99	16.2	0.75

 

 

4 SHEAR STRENGTH TESTS

  The transverse walls of structural brickwork buildings are subjected to shear stresses due to wind pressure acting on the longitudinal facade walls. Even in framed buildings the infilled walls may also contribute to the building shear strength. Theoretical methods have been presented in order to take into account the shear strength of both types of walls, the structural and the non-loadbearing walls [Stafford Smith and Riddington 1977]. In all cases, the brickwork shear strength is a necessary parameter to be known for designing the building shear strength. Nevertheless, shear test results of horizontally perforated bricks with precompression are not usually available. Because of that, shear strength tests comparing the structural performance of the 6-hole horizontally perforated bricks, the 21-hole vertically perforated bricks and the solid bricks are presented [Oliari and Duarte 2000 and Oliari 2002].
  There are several different methods to determine the shear strength of masonry. The method used here was presented by Riddington and Jukes [1994] and has many advantages compared with others test methods because it is simple, easy to perform and allows the application of pre-compression in addition to the shear stress, simulating the real behaviour of a wall subjected to both pressures: wind and dead weight. Figure. 12 shows the set up of a shear test.

Figure 12. The set up of a shear test showing two hydraulic jacks mounted to apply horizontal and vertical load.		Figure 13. A vertical crack shows the line of rupture along a mortar joint (6-hole horizontally perforated bricks).

  The experimental program was carried out using solid, 21-hole vertically perforated and 6-hole horizontally perforated bricks (with round holes) performing shear tests on 114 specimens in total. All test specimens were built using 1:1:6 mortar (Portland pozzolan cement : lime : sand) by volume. The average compressive strength of the solid bricks is 44.4 N/mm2 (CV=18.6%), of the 21-hole vertically perforated bricks is 22.7 N/mm2 (CV=18,8%) and of the 6-hole horizontally perforated is 2.0 N/mm2 (CV=30.6%). The total water absorption by weight was 11.64%, 12.54% and 12.65% respectively.
Table 5 shows the shear test results. In this table the compressive stress level (sc), the amount of the shear test specimens (N), the average shear strength (t,), the standard deviation (Sd), the coefficient of variation (CV) and the characteristic shear strength (fv) are also shown. The compressive stress level was kept low due to the low compressive strength of the 6-hole horizontally perforated bricks.

  The shear characteristic strength of the bricks prescribed in BS 5628: Part 1:1992 is 0.15 N/mm2 and 0.21 N/mm2, for mortar designation (iii). We can see that all test results are compatible with the Code.

 

Table 5 – Shear strength of triplets (Oliari and Duarte 2000; Oliari 2002)