BME Pál Vásárhelyi Doctoral School in Civil Engineering and Earth Sciences

BME Department of Structural Engineering

Supervisor: Dr. István Bódi

Experimental and numerical investigation of reinforced and plain masonry walls

Introducing the research area

Brief introduction of the research place

    The  Department of Structural Engineering of Faculty of Civil Engineering deals with the structural design and analysis of superstructures, buildings and bridges. Our department disposes the biggest laboratory of the faculty that is appropriate for experimental investigation of structures and provides place for experiments necessary for the research, teaching and design. Our subjects base the professional knowledge of future structural engineers.

History and context of the research

  The international literature describes reinforced masonry systems that place the vertical reinforcement in the middle of special masonry units or in which the bricks are layered in special order to form a reinforced concrete column at the center. In both cases, the pockets or holes are filled with concrete. The advantages of these walls were investigated by numerous researchers. It would be too expensive to analyze every possible loading cases, masonry units and brick bonding possibility. Applying numerical models, those virtual experiments are also executable that do not dispose of technical and financial background for execution. 

    Prerequisite of the modelling is the knowledge of the mechanical characteristics of constituents. Due to the big variety of mortar, masonry units and their manufacturers, differences can not only be observed in the mechanical properties of materials but there is a spread in the literature data available for them. In Hungary, manufacturers do not have information on all features, and standard investigation methods are only available for the investigation of the mechanical features of masonry.

     For these reasons, a lot of researchers deal with material and modelling problems in the field of load bearing masonry. The most widespread numerical modelling methods in this research area are finite and discrete element methods. The latter one was chosen for the analysis because no three dimensional discrete element model applied for masonry walls and backed with experimental research has been published as yet.

The research goal, open questions

   The dissertation is the first part of a year-long research project conducted in cooperation with the biggest Hungarian brick manufacturer and an international engineering consulting and software development firm. The final goal of the research is to construct a building method that ensures bigger deformation capacity for masonry walls adjusted to the Hungarian brick types, building systems and tradition, which enables application of masonry systems to middle-high buildings. In order to make the design easier the following needs to be determined: where to place the reinforcement and what brick bonding technology should be used to attain the most economical implementation. This will enable us making recommendations for brick manufacturers to consider upon compiling their selection of masonry units.

    The primary aim of the research is the determination of the mechanical parameters of brick and mortar missing from the literature. Based on these parameters, a plain masonry modelling system can be developed and the behavior of the model can be confirmed through experiments. One of the aims was developing a new brick bonding method for reinforced masonry, determining its advantages and disadvantages compared to traditional masonry walls and confirming its enhanced resistance against horizontal loads. Based on experiments, models can be developed that prove that buildings of more than 2-3 storeys can be built from reinforced masonry while meeting the design requirements.

Methodology

  The failure modes of plain masonry walls subjected to different loading directions and loading types were collected, grouped and typed. In order to determine the failure modes of reinforced masonry (how reinforcement affects the crack pattern, deformation and load bearing capacity) experimental investigations were carried out. It was stated that the failure occurs mostly at the brick-mortar joint in the case of the plain and reinforced masonry, as well.

Figure 1: Failure modes in case of walls subjected to horizontal loads

   Three methods are frequent for modelling masonry: the first one where both the mortar and the masonry unit are modelled separately (Figure 2 a), a second where the unit is simulated (b) and a third one where the constituents are not shown separately, only one homogenous material symbolizes the behaviour of the wall (c).

Figure 2: The modelling methods of masonry: a) micromodel, b) simplified micromodel, c) macromodel (Lourenco, 1996)

    As failures in plain and reinforced masonry occur at the surface of the unit-mortar joint and not in the mortar itself, if modelling conditions allow, a simplified micromodel can be created for both cases. With the help of a simple 2D discrete element model it was confirmed that all possible failure modes of reinforced masonry can be modelled by adequately choosing the parameters [15].

We have set three conditions for a proper 3D masonry wall model:

1. The model has to show the same joint behavior as described in the literature in details.

2. The model has to provide the same crack pattern as in the experiments using the material characteristics determined according to the experiments with brick and mortar specimens.

3. Any point of the wall should be displaced to the same extent under the same loading both in the model and in the experimental setup.

The key issue of the discrete element modelling is the determination of the mechanical material properties and other parameters that can not be determined by experiments. Because of the antinomic and deficient nature of the information on these parameters in literature, own experimental measurements were necessary.

In the Structural Laboratory of the Department of Structural Engineering and of the Department of Construction Materials and Engineering Geology destructive and non-destructive measurements were carried out in order to determine the size and loading direction dependent material parameters. Additionally, experimental investigations were carried out on plain, horizontally, vertically and in both direction reinforced masonry walls subjected to horizontal loads in order to determine their behavior. (Click here to see a movie of a failure of an experimental wall)

Figure 3: Parts of the experiments carried out     

Results

Figure 4: The new brick laying method developed and a conventional masonry system

2. To prove the beneficial behavior of the new system, a series of experiments were carried out on various masonry walls using two types of mortars. It was proved that in the new system the first cracks appear only after twice as large displacements and at 20% higher load levels than in conventional masonry walls. Before the final failure, the new walls are capable of 10 times higher deformation then conventional masonry walls. This system is beneficial in blast or earthquake hazarded areas, next to temporary vacant building sites or for partition walls or for walls that are loaded with earth pressure. [8, 9]

3. Based on the experiments, characteristic failure mechanisms were analyzed and the influencing parameters were determined. In order to simplify the design work of engineers a design method was developed to analyze the behavior of these new type of walls subjected to shear. [10]

4. In close cooperation with Wienerberger company, mechanical properties of the brick unknown for the manufacturer but necessary for the design have been determined. Experimental investigations were carried out on more than 100 brick specimens and the reasons for contradictions in literature were clarified. It was stated that bricks show different behavior in each loading directions. It was determined how the material features depend on the following parameters: the roughness of surface, the height, length, width and shape of specimen, the type of sampling (bored or cut), and loading direction. The mechanical properties of one Hungarian mortar that is important from the design point of view were measured on more than 50 specimens.  [13]

         

5. A 3D discrete element model was built to simulate the behavior of  the brick-mortar joint based on the parameters determined in the experiments. With the application of the results, a wall model was created for conventional solid brick masonry that governs the same crack pattern (Figure 6 a), deformation and load bearing capacity as that of the experiment. The developed model is appropriate for the simulation of the experiments carried out and it is possible to model the real wall subjected to the same load with adequate choice of parameters. The sizes, layout and the properties of the units as well as the loading direction of the wall are variable arbitrarily in the model. Figure 6 b shows the failure of a fictitious wall in the case of an impact of high velocity on the middle brick in the wall. [14] (Click here to see a movie of a failure of the wall.)