Seismic Studies on Small-Scale Models on Adobe Houses

The purpose of this research was to increase our knowledge of the seismic behavior of low-strength masonry (LSM) buildings, in particular adobe houses. The specific objectives of this study were (a) to explore the possibilities and limitations of reduced-scale model testing of adobe construction, (b) to evaluate the problems of dynamic similitude and material simulation in small-scale models, (c) to study the dynamic response characteristics of simple adobe house configurations and (d) to assess the relative effectiveness of several simple improvement techniques. To fulfill these objectives two series of tests were performed: a materials testing program and a dynamic testing program on six reduced-scale model adobe houses.

The goals of the material tests were to investigate the effects scaling time and size on the properties of the adobe bricks and assemblies. The goals of the model dynamic tests were to study the elastic and post-elastic dynamic behavior of these buildings through shaking table tests. Each model was tested to collapse. The improvement techniques investigated were (1) using a light-weight roof, (2) anchoring the roof beams to the supporting walls, and (3) adding a bond beam.

The shaking table tests provided basic information on the extent of damage and the types of failure that occur in adobe buildings during major seismic events. Wall overturning was the most often observed mode of failure. The lighter roof delayed initial structural damage but had a smaller effect on the model's resistance to collapse. Both anchored roof beams and bond beams tied the walls together, helped prevent wall overturning, and increased the model's resistance to collapse.

It is concluded that life-threatening collapse of adobe buildings can be delayed considerably by adding simple but well-designed details that tie the walls together at the roof level. However, there is no simple means of preventing extensive damage in adobe buildings during severe earthquakes, The design of these details and the extent to which they improve stuctural performance can only be assessed through analytical or experimental means that are capable of predicting dynamic response to the stage of incipient collapse. At the present time, shaking table tests on reduced-scale models may be the most effective means of achieving this goal.