Modelling of the seismic behaviour of TRM-strengthened rammed earth walls
This thesis of the author Reza Allahvirdizadeh was developed within the framework of SAHC Erasmus Mundus Master Course.
Earthen constructions constitute a considerable part of the existing heritage and a large percentage of the World population is still living or working in buildings built with this structural system. Like other types of masonry structures, rammed earth constructions are acceptably stable under gravity loads, although they are significantly vulnerable to earthquakes. Therefore, a precise understanding of their behavior in case of being subjected to ground motions and the proposing of effective strengthening techniques achieved a great interest both in research and practice.
Strengthening methods should be not only to enhance capacity and ductility of the building, but also to satisfy a variety of criteria such as being compatible with the substrate, economical and reversible. Considering all, the low-cost textile reinforced mortar (LC-TRM) strengthening is introduced, and its efficiency on rammed earth walls is studied numerically in this thesis.
In this work, the seismic performance of both unstrengthened and strengthened rammed earth structural components is investigated. In this regard, in-plane and out-of-plane behaviors are studied by means of different constructed nonlinear finite element models. At first, pushover analysis by mass- based lateral load pattern is conducted on unstrengthened walls to evaluate their capacity and understand possible failure mechanisms. Furthermore, the outcomes of these analyses are employed to select the most proper modeling approach from shell or solid elements and the walls with appropriate geometrical dimensions. In the following, pushover analyses are conducted on strengthened walls to choose between different strengthening materials and assess the effectiveness of the adopted strengthening technique. Furthermore, the frequency change of the walls with the damage states (lateral displacement levels) is studied to represent the initiation and propagation of damage in unstrengthened walls and to evaluate the effectiveness of TRM strengthening method.
Finally, an artificially generated ground motion record was applied to both unstrengthened and strengthened walls to perform nonlinear time-history analyses. The outcomes were used to compare the dynamic behavior of the walls against the results of the pushover analyses.
Earthen constructions are all over the world
Earth as the most available material in many regions around the world was probably one of the first building materials used for manmade constructions. In fact, a significant part of the World's heritage and monuments are constituted by structural masonry, where raw earth is a dominant material. It is easy to find worldwide spread historical monuments made of raw earth.
Furthermore, the availability of earth makes this material as the first option for building in rural, isolated and low income societies. On the other hand, and in the past few decades the sustainability and other green inherent characteristics of earthen constructions made them to achieve a considerable attention even in modern developed countries.
Considering all, it is essential to have a precise understanding about their behavior. From a general point of view, these structural systems have a stable behavior under gravity loads, which made them to withstand during centuries, although, they are highly endangered with respect to earthquakes. Hence, investigating their response under earthquake excitations, understanding the most probable failure mechanisms and eventually proposing reliable and efficient strengthening solutions are important concerns of recent research works. The development of this research is important from several aspects, such as preserving the cultural heritage values of many monuments, securing numerous lives and limiting induced economic impacts.
One of the most common construction techniques using raw earth is called rammed earth, in which moist earth is placed between panels and rammed to obtain a compact material. This structural system is widespread all around the World, while limited studies are available in the literature regarding this construction method. Moreover, most of studies are concentrated on the material point of view.
The seismic performance of rammed earth buildings
As stated previously, this study has the purpose of investigating the seismic performance of rammed earth buildings. In this regard, wall components with different configurations are considered to assess the in-plane and out-of-plane behavior of them.
Subsequently, it was aimed to evaluate effectiveness of recently developed strengthening solution for enhancing the seismic performance of these walls. Among all convenient approaches, the low-cost textile reinforced mortar (LC-TRM) was selected, due to its simple application, reasonable costs and compatibility with substrate, which is an especially important concern in historical constructions.
Population living or working in earthen constructions
Furthermore, the low associated building costs led this material to be an appropriate choice for societies with economic issues, as well as for hardly accessible regions and isolated rural areas. The fact is that approximately 30-40% of the world population is estimated to live or work in earthen constructions. Geographical distribution of such constructions is shown in Figure 1.
During last century, traditional earthen constructions are substituted by constructions made of more modern materials, such as steel and reinforced concrete in developed countries. This situation led to the almost disappearing of the use of earth constructions; however the last decades pursue for more sustainable building solutions led to a renewed interest for this type of constructions. An example of a modern earth building is shown in Figure 2.
The fact is that an expensive earthen built heritage exists and that the demand for building with earth is increasing thought these constructions must fulfill modern demands.
On the other hand, many of these buildings are located in regions with medium or high earthquake possibility of occurrence (see Figure 3). Hence, it became necessary to understand the behavior of these constructions not only for modern design objectives but also for strengthening/repair of existing buildings and monuments.
This thesis is focused on study of rammed earth structures. In this technique, earth with adequate moisture content is placed between two parallel panels and is compacted. This practice is well-known in all continents, as it can be understood from its specific name in different languages.
For instance, it is known as pisé in French, tapial in Spanish, taipa in Portugal, terra battuta in Italian, stampflehm in German, hangtu in China, chineh in Iran and pakhsa in Uzbekistan. The rammed earth technique was independently developed in China and in the Mediterranean region; which is later spread by the settlers of the New World. Its development map is depicted in Figure 4.
The weaknesses of earth constructions
To assess, repair and strengthen adequately earth constructions, it is essential to understand their weaknesses. Several factors such as rainwater, soluble salts, and temperature oscillations can lead to occurrence of damage.
Although, these constructions have good behavior under gravity loads, but like other types of masonry buildings they are strongly endangered with respect to lateral loads. For example, many inhabitants, especially in rural regions and historical monuments of Turkey, Iran, Peru and Chile, have been severely affected by the occurrence of recent earthquakes. In this regard, one of the most catastrophic losses is the devastation of the historical citadel of Arg-e-Bam, classified by UNESCO as world heritage.
Analysis of the seismic performance
The present thesis aimed at investigating the seismic performance of selected structural assemblages of rammed earth walls, here termed simply as walls. This objective was achieved by evaluating their in-plane and out-of-plane responses. In addition, the influence of the strengthening with textile reinforced mortars (TRM composites) on their behavior was also assessed.
Firstly, pushover analyses were conducted on the unstrengthened models to understand possible failure mechanisms, evaluate their load and displacement capacities, and select the proper geometrical dimensions and modelling approaches. It was observed that the failure in the in-plane models is mainly due to sway of the wing wall. A shear diagonal crack may also initiate and develop on the web. It was clear, that the shear crack on the web resulted mainly from the in-plane behavior of the wall.
Furthermore, the obtained outcomes of the pushover analyses lead to select the wall with smaller wing walls. With respect to the out-of-plane models, detachment of the web from the wing walls and bending of the web's mid-section were the main reasons of the failure. The asymmetric geometry of the out-of-plane models caused the captured capacities to be different regarding the direction of pushing. It was observed that pushing the out-of-plane model to the outside of the wing walls results in a brittle failure, while the supporting contribution of the wing walls in the reverse pushing direction leads to ductile failure with much greater load and displacement capacities. In the latter case, the arch effect in the web was observed.
Subsequently, the comparison between shell and solid elements highlighted the higher representativeness of the models prepared with solid elements. Disregarding the thickness of elements in the shell element modeling approach, particularly that of the web resulted in greater lateral displacements. However, in most of the cases the load capacity has been decreased. Moreover, unrealistic stress concentrations on the connections of the wing walls and the web, and evaluations of the stress contours led to select the solid element modeling approach for further investigations.
The LC-TRM strengthening technique was implemented in the models to assess the influence of this technique on seismic performance of the rammed earth walls. In this regard, two different LC-TRM solutions with different textile materials (glass and nylon) and relatively similar costs were selected. The glass textile presented higher tensile strength, while the nylon presented an extremely larger deformation capacity. The outcomes obtained from the in-plane model strengthened with each of the textiles showed that the glass mesh leads to a larger improvement on both load and displacement capacities.
The further investigation showed that the previously discussed failure modes were controlled at the strengthened models for a lateral load equal to the peak capacity of the unstrengthened wall. The observed stress and strain states of the adopted strengthening showed its proper and efficient contribution to the behavior of the wall. It should be noted that the applied strengthening technique had no influence on the damage initiation point, but made it more difficult to develop. Moreover, the final failure mechanism of the strengthened model is relatively similar to the unstrengthened one, but it allowed achieving higher capacity and ductility.
In addition to the aforementioned studies, the changes in frequencies and dynamic properties of the both strengthened and unstrengthened models were also investigated. It was observed that the modes with the highest contribution may change during the pushover analysis and at different damage states. Furthermore, a considerable reduction in the frequencies of the modes with highest contribution was observed for the unstrengthened models. The largest reduction belongs to the out-of- plane model pushed to outside of the wing walls.
The aforementioned study was also carried out for the strengthened models. In all cases, the behavior was considerably improved for the same lateral displacement level of the unstrengthened models. In other words, the adopted strengthening technique controls the damage development and softening of the wall.
Finally, an artificially generated code-compatible ground motion record was applied to the models to investigate their dynamic behavior. The generated record satisfies the Eurocode 8 regulations (near- field earthquakes) for the municipality of Odemira in southern Portugal. It was observed that, the in- plane model remains elastic when subjected to that record. The additional induced mass to the wall by the strengthening caused slightly higher loads to be applied, while its higher stiffness leads to smaller lateral displacements. In the case of the out-of-plane model, even such moderate earthquake causes damage to be initiated.
The outcomes of the nonlinear time-history analyses showed that the pushover analyses may precisely predict the vulnerable regions, at least for the level of dynamic displacement imposed. Moreover, it was seen that the pushover analyses can lead to relatively accurate load and displacement predictions for slightly damaged walls.
Opportunities for further research
As part of the present study, a number of areas for further research have been identified. With respect to the modelling part, it may be important to take into account the influence of the interaction between the applied LC-TRM composite and the substrate, and also the interaction between the mesh and the mortar matrix. In the latter case, it is required to obtain previously the bond stress-slip behavior. By considering these behaviors in the numerical analyses, it is expected to allow obtaining a more comprehensive prediction of the capacity and possible failure modes.
The current thesis only focused the seismic performance of two types of structural assemblages, while, more complicated cases may occur by investigating the behavior of the full structure. For instance, the interaction between walls, the influence of existing openings, the effects of lintel and ring beams and also loads from the roof should be further investigated. Furthermore, the efficiency of the adopted strengthening technique on the building should be assessed in comparison to the other available methods.
Moreover, it is recommended to conduct incremental dynamic analyses (IDA) to obtain a more precise load and displacement capacities from the dynamic behavior. Nevertheless, the outcomes obtained may be used for comparison with those reported in the current study from the pushover analyses. Therefore, the reliability of the pushover analysis at different damage states can be assessed. In addition, the outcomes of the IDA may lead to generate fragility curves of the rammed earth walls. It should be noted that several aspects of the nonlinear time-history analyses of the rammed earth components and buildings are still uncertain, such as the cyclic behavior and the damping ratio.