Mechanical behavior of rammed earth walls under Pushover tests

ranime el nabouch

Mechanical behavior of rammed earth walls under Pushover tests

ranime el nabouch

2017

visibility

…

description

217 pages

link

1 file

Sign up for access to the world's latest research

checkGet notified about relevant papers

checkSave papers to use in your research

checkJoin the discussion with peers

checkTrack your impact

Abstract

Determining the mechanical behavior of rammed earth walls is a highly important feature giving that there is a renewed interest in the use of the natural material in building construction. On one hand, it contributes to finding suitable and appropriate ways for the conservation and restoration of these building and on another hand, it draws new recommendations for the conception of new rammed earth structures with respect to new building regulations (earthquake building guideline, new thermal regulation).In this context, this thesis investigates the behavior of rammed earth walls subjected to lateral loading. Four unstabilized rammed earth walls were tested under a combination of vertical and monotonic pushover tests up to failure. The shear behavior of these walls is studied along with the failure modes by means of digital image correlation. Results showed that the response of the rammed earth is characterized by significant non-linear behavior with a remarkable ductility. In gener...

Figures (192)

Figure 1.3 Production of Adobe in Ecuador [6] Adobe is also an ancient construction technique that consists of filling molds with moist earth to obtain finally the desired shape (Figure 1.3); the adobes are then left in the sun to dry. They are ready to be used as masonry units. The applied mortar is usually made from the same earth used in the production. Many examples can be found for this type of construction in rural and urban buildings as shown in Figure 1.4.

Figure 1.4 Examples of existing adobe constructions in Aveiro district, Portugal [7]

form above is begun. Rammed earth walls are buil usually consists of two para Figure 1.8(a). The principa by compacting soil between temporary formworks. The formwork lel surfaces separated and interconnected by spacers as shown in binder of the grains is the clay. The mixture of the earth is compacted into layers of approximately 15 cm by the use of a rammer. The average thickness of the wall is 50 cm. As each form is filled, another form is placed above it, and the process is carried on until achieving the desired wall height. Forms can be removed directly as soon as the form above is begun.

Figure 1.9 The use of pneumatic rammer and metallic formwork in modern techniques, (Photo from the laboratory of ENTPE, Lyon)

Figure 1.17 Renovation of a part of rammed earth house in “Pont de beaux voisins”, Rhéne-Alpes,

Figure 1.18 Stabilized Rammed earth construction in Australia

Figure 1.19 Church in Margaret River, Australia made from stabilized earth constructed in 1983

Figure 1.20 Stabilized rammed earth in Domaine de la Terre, l’Isle d’ Abeau, France (from

Figure 1.21 Pre-fabricated walls being craned into position during construction at Montbrison, Loire (Photo: N. Meunier, 1995)

Table 1.1 Energy consumption for different type of construction types [28] The second environmental advantage lies in the demolition phase of the buildings. In fact, the demolition and destruction phase of a material also contributes to the energy consumption of the construction sector [29]. Unstabilized rammed earth is easily recyclable with no need of special treatment or specific storing techniques. The earth can be reused for construction without any environmental impact and therefore no problem of material waste as is the case of conventional materials [30].

Table 1.2 Diffusivity of different building material according to [37] and [38] It is, therefore, interesting to have a comparison of the coefficient of diffusivity between different building materials (Table 1.2). According to Table 1.2, earth material (dry or in brick form) offers the lowest values of the material diffusivity which enables to provide thermal comfort during dwelling.

The study showed that the mean erosion depth was about 0.5% of wall thickness (about 2mm) for stabilized and approximately 1.6% (6.4 mm) for unstabilized materials. The study indicated that rammed earth walls can exceed a life more than 100 years for non-stabilized rammed earth walls. On the other hand, many historical examples of rammed earth structures could represent a good example of the material durability like the temple Japan that was built approximately 1300 years ago. Yet, many uncertainties remain in this field. years ago. Yet, many uncertainties remain in this field.

Figure 2.1 Envelopes for rammed earth construction recommended by Houben and Guillaud [36]

Figure 2.2 Relationship between dry density and compressive strength [53]

A great dispersion is noted in the materials properties in Table 2.1, particularly for Young’s modulus. This is related to many factors as the testing procedures, the workmanship and the type of the soil used. In most of these studies as in [12], [59], the samples are small (10 cm cubes or cylinders 20 cm high x10 cm in diameter) and are manufactured in the laboratory, hence these samples are not very representative of in-situ material. Table 2.1 Summary of material properties for rammed earth in the literature

Figure 2.4 The determination of the modulus of elasticity following the ASTM standard

Figure 2.9 Compaction curves from Proctor tests [68]

Table 2.2 Characteristics of the specimens

Figure 2.13 Cylindrical specimen (20-cm diameter, 40-cm height) during the test setup The prismatic and cylindrical specimens were tested in unconfined compression with displacement control of 0.02 mm/s using a hydraulic jack. Figure 2.14 shows a cylindrical specimen during the test setup. The Digital image correlation (DIC) was used as a measurement technique to capture the failure process of rammed earth walls and specimens and to provide all the needed information about the displacement field for the corresponding structure. the needed information about the displacement field for the corresponding structure.

The images for he DIC were recorded using high definition (3248 x 4872) with almost 16 nillion megapixels white and black camera system. During the tests, a projector was installed. Thus, it is possib kandom speckle ammed earth wa e to guarantee a constant illumination on the specimen (as seen in Figure 2.14). pattern (Figure 2.14 (a)) was applied on the surface of all the specimens and the Is using black paint type mat to avoid any reflection of light. The diffusion was lone using an airbrush gun where the pressure was controlled with compressors machine, herefore the pressure was regulated while applying the paint in order to give the best pattern. The quality of the random grey level was analyzed in term of histogram. Figure 2.14 (b) presents he histogram of grey level within a region of interest on a rammed earth specimen surface.

Figure 2.15 (a) Central part used to calculate the Young modulus from the DIC; (b) the vertical displacement fields for the corresponding figure Thanks to the DIC, the displacements during the loading were recorded which enabled to determine Young’s modulus of the tested specimens. The Young’s modulus was calculated from the displacements of the middle part (following the height) of each specimen as shown in Figure 2.15.

In the current case of rammed earth house of two stories, walls (50 cm thick) are supposed to be loaded with a normal stress of about 0.3 MPa. In practice, concerning the values obtained of the compressive strength, a structure made with this soil would resist to vertical loads in normal conditions. Table 2.4 Comparison with other studies in the literature

Figure 2.16 State of loading (represented by numbers) corresponding for the calculation of the The software displays the results of the principal strain that are derived from Green-Lagrange's strain tensor and the maximal strain denoted ‘emaxi’. Figure 2.17 displays the principal strain along with the crack pattern (through ‘emaxi’) for one of the cylindrical specimen and for different states of loading that correspond to Figure 2.16.

Figure 2.18 Variation of the density within the layer shown on a rammed earth wall due to the compaction energy that is greater at the layer top. Since the energy is distributed within the layer, the density at the top of the layer is higher than in the bottom, therefore it is assumed that the density varies linearly inside the layer as shown in Figure 2.18. This subject was addressed in the work of Bui [60] that revealed the heterogenic aspect of the rammed earth due to the variation of the density inside one layer.

Figure 2.19 shows the DIC results for the vertical strain ‘eyy’ that reflect more clearly the variation of the deformation in the direction of the loading at a certain point of loading that corresponds to 17 kN. This figure shows undoubtedly the difference in the variation of the deformation between the earthen layer and the interface. Therefore, there is a difference in the correspondent Young modulus of earth layer.

Figure 2.20 Failure pattern of the cylindrical specimens (C1) and (C2) During the compression tests, it was observed that the interface may interfere to initiate the cracking in the sample (as in the case of state 4 in Figure 2.17) where the direction of the principal vector initiate from the interface of the first layer and continue along the height of the specimen. Therefore, the interfaces between the layers can be considered as weak points for this material especially in the case of horizontal loading where the interface is more solicited. At the end of the test, all the cylindrical specimens failed in a similar manner as shown in Figure 2.20. Vertical cracking was obtained through the full height of each specimen. These cracks crossed he compacted earth layers and tend to align in parallel to the applied compression direction. In he case of the prismatic specimens, cracks mostly appeared in the central part propagating oward the top and bottom edges. Inclined cracks were also obtained at the corners with signs of crumbling from the sides during the loading phase. After reaching the ultimate load, brittle failure occurred and some pieces from the specimens fell out. Figure 2.21 shows the crack pattern at the end of the test for the prismatic specimens.

Figure 2.24 Principle of the direct shear test The Casagrande shear test is a classical test in soil mechanics. Figure 2.24 shows the principle o: the shear. The soil specimen is placed in a shear box that consists of two independent half boxes The plane located at the separation of the two half-boxes is called sliding plane and correspond: to the specimen shear plane. The test consists of applying a vertical force (N) that is maintainec constant during the test and then applying a horizontal shear force (F) on one of two half-boxe: as in (Figure 2.24).

equation t=c+o.tan(gy) can be seen in Figure 2.25.

Figure 2.28 The shear box installed on the loading frame

Figure 2.32 Unmolding of the upper part of the shear box after the test

Figure 2.34 Plot of the determination of c and 9 for the large shear box inside one layer Results of the large shear box at the interface:

In a study conducted by [88] within the Project of Primaterre, the same material that originates from “Dagneux“ was used to investigate the shear strength at interface of several specimens at different condition’s: oven dry (0% moisture at test); ambient state (w=1-2%) and moist state (water content >4%). The experimental procedure is performed using the uniaxial press and inclined metallic wedges that can help inducing shear stress and normal stress on the specimen interface. Figure 2.38 shows the variation of the shear strength of the rammed earth interface with the variation in moisture state at the testing time for specimens tested at an inclination of 30°. The results conducted in this study showed that the specimen at the dry state had the higher mechanical strength. The shear strength (at the interface) was lower in the case of the specimen with higher moisture content at the testing time. different condition’s: oven dry (0% moisture at test); ambient state (w=1-2%) and moist state

Figure 2.38 Variation of shear strength with moisture state at test — 30° inclination [88]

Effect of time and consolidation of rammed earth wall: In general, the final goal is the determination of the parameters (cohesion and friction angle) of he walls tested under pushover test (in Chapter 3). It is worth to mention that the walls have been submitted to consolidation for more than two/three months, indeed the wall’s high weight (approximately 1 tons) and the weight of the concrete beam (300 kg) that have been placed on he wall (see Chapter 3) contribute to the consolidation effect and could have an impact on the parameters that we aim to determine. In geotechnical investigation on normally consolidated clay, it has been shown that for measurement of the undrained cohesion, the consolidation pressure has to be taken into account for the interpretation of the results [89].

Figure 2.39 Properties in function of depth for soil normally consolidated: (a) consolidation pressure: (b) undrained cohesion Figure 2.39 Properties in function of depth for soil normally consolidated: (a) consolidation

Figure 2.40 (a) A standard direct shear test; (b) the rammed earth specimen inside the box

Table 2.6 Results of the parameters: large shear box and small shear box Indeed, the difference of the water content at the testing time between the two specimens is significant. This supports the previous idea mentioned in section (2.5.4) in which the impact of water content (in fact suction) was discussed. The small specimens were left to cure in enough ime to achieve (2-2.5%), moreover they had small dimensions that ease a greater evaporation, which could explain the difference. On the other hand, the large specimen had more than 4% of water content at the testing time hence their mechanical parameters were directly influenced and he cohesion had drastically decreased. The small shear box shows that the mechanical properties of the rammed earth wall can improve in function of time and that the cohesion and the friction angle for the rammed earth walls are higher than the one obtained by the large shear box.

Table 2.7 Synthesis of friction angle and cohesion. The results show important variations of the friction angle and cohesion in function of each study (difference of material used, of the approach used). Globally friction angle varied between 35- 45° but great dispersion was noted for the cohesion which proves the sensibility of this parameter (suction effect). Yet, more investigation for these parameters at the interface and in the layer is still needed as these parameters should be included in the design principles.

Figure 3.3 Scheme of the cyclic in-plane shear-compression test [75]

Figure 3.4 Crack pattern under a combination of vertical compression and cyclic shear: (a) RE Wall 1, (b) RE Wall 2, (c) and RE Wall3, [75]

is explained in appendix A, Figure A.5). Figure 3.5 illustrates the experimental setup. The dimensions of the wall were 2.4 m x 2.1 m x 0.6 m and they were cast on a concrete footing. The failure for the unreinforced rammed earth was governed by a shear failure mechanism as shown in Figure 3.6 and the ultimate force reached was about 73 kN. A comparison of the test results for the wall specimens before and after reinforcement was conducted (Figure 3.7) showing higher ultimate load and maximum horizontal displacement for the reinforced structure. Thus, his proposed method was validated by the authors as an effective retrofitting technique. However, the relevancy of these reinforcement techniques in the case of the real structure is still questionable especially due to their influence on the other important aspect of the rammed earth.

Figure 3.5 Experimental setup, from the study of [100]

Figure 3.6 Failure of unreinforced rammed earth, from the study of [100]

* This technique is mentioned more in details in Appendix A

Figure 3.7 Horizontal load—displacement response for the wall specimens before and after reinforcement (solid line) [100] Figure 3.7 Horizontal load—displacement response for the wall specimens before and after

Firstly and according to [101], the behavior of masonry is complex as a result of weakness plans along the vertical and horizontal joints. In addition, the application of a horizontal force on a previously loaded wall causes the contributions of flexural and shear stresses. When the masonry wall is subject to flexure, horizontal cracks appear at the interface brick-mortar due to the low tensile strength of masonry joint plans. According to the author, these are only local cracks and the overall rupture of the wall does not occur in these conditions. These cracks reduce the resistant section subjected to the compression, causing a concentration of compressive stresses and rupture of the wall by compression that affects the bottom corner (Figure 3.8). A similar feature can be found for the rammed earth specimens as in the previous section. Horizontal local cracks were located in the interface between the rammed earth layers for lack of cohesion and friction, therefore cracking can propagate following these interfaces (Figure 3.8).

Figure 3.9 Failure mechanism: (a) Sliding on bed joint; (b) Diagonal cracks formation, [102] In the case of rammed earth, similar failure mechanism to those of masonry can occur. The heterogeneous aspect of rammed earth can induce a mechanism of failure due to the sliding between two earth layers. On the other hand, this type of failure could be avoided by controlling the confinement of the wall.

of seismic resistance of masonry wall [104] As mentioned before the pre-loading is an important parameter that can have a direct effect on the behavior of the masonry during the test. The literature has shown that the test type "cantilever" is widely referenced and used to represent the different failures modes that may occur, depending mainly on the vertical preload applied ([102] and [103], for example). Thus, the cantilever type of test is more adapted because it allows the control of the vertical force applied throughout the test, and therefore the failure mode that wishes to be characterized.

Figure 3.11 Metallic formwork fixed to the ground before the manufacturing

Water was added to the earth to obtain the optimum manufacturing water content (following the optimum proctor done in Chapter 2), approximately 12% by weight in this case. The characteristics of each wall are given in Table 3.2. Specimens were sampled to check the water content repeatedly and the drop test was used along the preparation to control the water addition. The manual method assured the best homogenization of the earth.

Table 3.2 Characteristics of the wall specimens

Figure 3.14 Test setup on a rammed earth wall (1.5 m x 1.5 m x 0.25 m)

Figure 3.15 Vertical tie rod to prevent the beam from rocking

During the tests, the two projectors installed were the only two light sources present. Thus, it is possible to guarantee a constant illumination for the specimen. Both projectors were installed on each side of the camera but sufficiently far so that the camera is not disturbed by the heat emitted by these projectors. To facilitate the processing of the results, a specific box allows the synchronization of all the measurement recorded during the test (the force and displacement of all three actuators and the three displacement sensors) and the images captured with the camera. The acquisition frequency was | image per second. The DIC setup is illustrated in Figure 3.16. apparatus

Figure 3.17 System placed on top of the beam Then the horizontal pushover was carried out by a hydraulic actuator (VH) with displacement control (VH on Figure 3.14). The loading rate was 1 mm/min until failure. The DIC was also used to measure the displacements and inspect the evolution of the cracking in the walls during the test. A summary of the parameters of the test is shown in Table 3.3. the test. A summary of the parameters of the test is shown in Table 3.3. Table 3.3 Summary of loading system parameters

Figure 3.19 Horizontal force in function horizontal displacement provided by the hydraulic

Figure 3.20 Pixels close to the placement of M3 ‘igure 3.21 illustrates this comparison for wall 1 (1 m height) where a good correspondenc« tween the two measurements was observed (Figure 3.21). This proves the precision of th pparatus used and assures the accuracy of the results provided by the software 7D. The sam«¢ omparison is done with the other walls (refer to Appendix C). In general, we can neglect the las ifference for some walls (for example in Figure C.3) because it is mainly due to an error fron he experimental setup of the placement of sensors. We can certainly proceed to use the result: rovided by the DIC.

Figure 3.21 Comparison between the displacement given by the image correlation and the displacement sensor in function of the horizontal force VH for wall 1

Figure 3.23 Vertical strain at the end of the preloading phase (case of wall 3)

Phase 2: Horizontal loading Table 3.4 Young’s modulus of the walls.

Figure 3.24 Horizontal loads on top of the walls in relation to the top horizontal displacements Figure 3.24 shows the horizontal loads in relation to the horizontal displacements on top of the four walls. These displacements were obtained from the DIC that is more accurate than the displacements given by the horizontal actuator (which was relatively influenced by the stiffness of the loading system). A pixel on the top to the left of the wall was chosen for the calculation of this displacement as shown in Figure 3.25.

Table 3.5 Summary of walls results As for the walls 1 and 4 that have a height of 1m, a different behavior compared to walls 2 and 3 is to be expected due to their lower height (lower flexural moment at the bottom section) inducing a better strength capacity. Nevertheless, this was observed only for wall 4 that achieved a maximum horizontal load which was greater than the other walls (53 kN) with important ductility. On the contrary, the structural behavior of wall 1 was the most fragile with a global

Figure 3.25 Point chosen for the calculation of the top displacement The results shown in Figure 3.24 demonstrate that the four walls had similar initial slopes of tt horizontal load — horizontal displacement curves (up to a certain limit 10 kN, which was abo 20% of the maximum loads), and then the behavior d each wall (by losing significant earth pieces) tha observed that the walls did not exhibit any ear y tup iffered after this limit due to the damage « caused the stiffness degradation. It we ure that could have occurred due to slidir between earthen layers. Walls 2 and 3, sharing the same height (1.5 m), exhibited nonline: behavior and similar maximum horizontal load abou 40 kN. The stiffness degradation resulte from the occurrence of micro-cracks inducing important displacement ductility and returnin globally a mechanical behavior less brittle and given in Table 3.5. more predictable. A summary of walls results -

Figure 3.26 Damage of wall 1 after reaching the ultimate load resistance significantly lower than wall 4. The maximum horizontal load was close to that of walls 2 and 3 but no ductile behavior was observed. In fact, the wall was damaged by high local compressive strains to the left side after reaching the ultimate load. Figure 3.26 exhibit the aspect of the wall with the appearance of severe damage to the left side of the wall. DIC in the next section shows more clearly the propagation of the crack. tence it could be suggested that the wall 1 had an inferior quality compared to other walls wher ybserving the exterior aspect of the layers. Figure 3.27 shows the wall aspect directly aftet nanufacturing. This wall was the first wall to be constructed and we assume that it could be less well controlled by the laboratory staff. In fact, the manufacturing process plays an importan ‘ole, any changes in the manufacturing of these layer due to either excessive water content or no nough compaction in some of the layers can impose a lower density, therefore reducing the tota rapacity of the wall. By visualizing the exterior aspect of the wall (Figure 3.28), the layers ippear to be in some areas less dense than the ones for other walls, we can therefore assume tha his wall could correspond to walls having a lower density in real practice.

this wall could correspond to walls having a lower density in real practice.

Figure 3.28 Visualizing less compacted areas in the exterior surface of wall 1 comparing to wall 4

Figure 3.30 In-plane bending failure in a wall

Figure 3.31 Crushing on the left side for different walls during the test

Figure 3.32 Tensile stresses resulting from compressive strut

Figure 3.33 Direction of principal strain vectors that highlight the deformation on the left side of

Figure 3.34 Visualizing the strut direction through the displacement vector for wall 1

Figure 3.35 Expansion of cracking width for wall 1 after attaining the ultimate strength

Figure 3.37 Crack growth due to extension during the pre-loading phase for wall 4

Figure 3.38 Crack closure due to compressive stress during the pre-loading phase for wall 4

Figure 3.41 Cracking evolution of the wall 2, in function of horizontal displacement

Figure 3.42 Cracking evolution of the wall 3 in function of horizontal displacement When the loading increased, layer-interface sliding shear occurred by visualizing a horizontal crack, at the lower-left part of the walls, at an interface between two first earthen layers. Figure 3.44 exhibit the principal minor deformation obtained by the 7D software with the direction of deformation for both walls 2 and 3. It was found that both walls exhibit the same type of horizontal crack that was extended to nearly the center of the wall. These cracks influence the resistant section that is subjected to compression, producing a concentration of the compressive stresses and crushing of the bottom corner. Another local vertical crack below this horizontal crack was also formed in the two walls.

Figure 3.43 Losing pixels due to loss of material observed for wall 3

Figure 3.45 Rocking at the base of wall 3 for 2 (corresponding to 36 KN horizontal loading)

Figure 3.47 Summary of all the failure modes of rammed earth walls subjected to shear Summary of the failure modes obtained for rammed earth walls Figure 3.47 sums up all the types of failure mechanism that were observed during the testing under lateral loading that can be summarized by: corner crushing from one side because of the stress concentration in this area, diagonal cracking that is due to diagonal compressive strut failure, sliding of some layers due to interface shear failure and local uplifting from the base due to tensile stresses.

Figure 3.48 The four walls at the end of the test

Figure 3.49 Visualizing interface failure when destroying the walls after each test

Figure 4.1 Material models adopted in the numerical modeling: (a) stress-strain relationship in

Figure 4.2 Behaviour of the macro- and micro-model: (a) using the initial parameters; (b) after calibration [76] Bui et al. [77] used the discrete element method with Mohr-Coulomb failure criterion for earthe blocks and Mohr-Coulomb interface model for interfaces between earthen layers. The techniqu of DEM was applied on masonry structures but has not been applied on rammed earth. This 1 one of the few studies that exist in the literature that uses DEM to model rammed earth walls The model was tested with several cases: walls’ compression test, walls’ diagonal compressio: test where the experimental results of [99] were used and also walls’ vibration under ligh shocks. This study identified 13 parameters needed for the used model. The main parameter were the cohesion and the friction angle. It was admitted that the cohesion of the earthen layer i 7-10% of the compressive strength and the friction angle varied between 45°-56°. The interface’ characteristics were assumed to be 80-90% of the earthen layers. The results obtained shows the there is no influence of the interface when compared to a homogeneous model (Figure 4.3) in th case of vertical loads.

Figure 4.3 Comparison of the numerical and experimental results for vertical loads [77]

Figure 4.4 Comparison of the numerical results in the case of diagonal loading [77]

Figure 4.5 Experimental setup of a diagonal compression test [99] Numerical Model:

Following the ASTM standard, the shear stress t of the specimens is calculated as follows: Figure 4.6 shows the corresponding numerical model of the present study that was constituted from 4049 tetrahedrons elements.

In this section, the relationship proposed by [77] for the cohesion was used (c = 0.1 f,) and the friction angle varied between 36° to 41° which seems appropriate for a first approach. The results are presented in Figure 4.7. The models could reproduce the initial slope of the experiment. The friction angle of 41° could reproduce the ultimate shear stress and the transition phase from inear to non-linear (Figure 4.7(b)). Although the typical “shear peak” was not reproduced with he numerical model due to the limited number of parameters, however, the last one could reproduce the post-peak softening phase of the material. Hence it reproduced the elastoplastic behavior of rammed earth in this case. This model is adopted in the following for the simulation of the experiments on the scale of the wall. of the experiments on the scale of the wall.

Table 4.1 Lists of the different parameters required for the different material

Figure 4.8 Geometry and boundary conditions of the numerical model

Figure 4.9 Results of the numerical model with the input parameters of large shear boy

The results based on the upper layer data overestimated the maximal horizontal load (Hnun=90 kPa) comparing to the experiments (Hexp=39-42 kN). This is not surprising as the cohesion value adopted in the numerical input for the earth layer (close to 263 kPa) is very high and therefore results of the horizontal capacity are important as well. This value for the cohesion does not seem reliable for the earth layer since it was only measured on a specimen that was taken from very well compacted areas and it doesn’t consider the whole aspect of the layer.

Figure 4.10 Comparison of the numerical and experimental results

Table 4.4 Parameters considered in the parametric study Results of this analysis where the effect of the parameters variation on the horizontal load capacity of the rammed earth wall are reported in Figure 4.11 and Figure 4.12 based on the different sets of layer and interface values obtained by the parametric study in Table 4.4. It is evident that the values of the parameters of the interface have an influence on the ultimate load obtained which justifies the use of a model with an interface.

Figure 5.1 Distribution of earth construction and high seismic areas in the world [113] Despite the attractive aspects of this construction material, some countries where the population still lives in earthen buildings have been influenced by severe and damaging earthquakes causing nany earth buildings to collapse. Moreover, it was found that it exists a high agglomeration of sarth buildings which coincides with the map of high seismic risk areas in the world (Figure 5.1). Rammed earth structures are considered to be vulnerable due to their poor mechanica characteristics (compressive strength, tensile strength as shown in the previous chapters) and leavyweight because of the thickness of the walls required for the resumption of the permanen und live loads. These combined aspects make rammed earth walls particularly sensitive to lynamic effects that can cause brittle failures without any prevention. It’s an important detail in ase of evacuation of the building and eventually in case of high damage, material loss and leath.

Figure 5.2 Earth building map and seismic zones in France [11] In fact, the performance of rammed earth walls in buildings under dynamic loading is a major concern (especially in the regions that can be subjected to high seismicity). When exposed to these loading conditions (combination of horizontal and vertical loading) walls are subjected to in-plane shear and/or out of plane lateral force. This requires experimental testing of the shear response and capacity of rammed earth walls. Therefore, important parameters can be assessed by the experimental tests under static lateral load like obtaining the ultimate drift as well as the ultimate load which enhances for the assessment of the mechanical performance of rammed earth buildings.

‘able 5.1 Summary of some of the earth buildings inspected after the February 2011 Christchurch

Damage State: A-None, B-Slight, C-Moderate with cracking damage throughout the building, D-Extensive crack damage throughout, E-Very extensive damage with collapse or partial collapse of the structure. The digit indicates where in the scale eg B1 low within the B category, B2 mid B category, B3 borderline to C

Figure 5.5 Damage limit states following drifts proposed by Calvi [126] According to Calvi for the LSI and LS2 limit, no damage concentration is yet observed, nevertheless when the structure attains the LS3 limit a damage concentration is assumed at the eround story level. Based on the observation of the behavior of the rammed earth walls, the occurrence of the limit states is proposed based on the behavior of the wall and the occurrence of the local crack pattern which follows the definition of each one suggested by Calvi. Table 5.2 describes these limit states for the rammed earth walls and the corresponding one for the case of masonry walls.

We first note that for the LS1 no visible cracks are obtained. For LS2 we distinguish the first reduction of wall stiffness associated with the first appearance of cracks. LS3 indicate the damage concentration through the deformation in the diagonal part or at the compressed corner with a remarkable loss of earth particles. This limit state is typically associated with the maximum shear capacity of the global force-displacement curve. As for LS4 significance loss of earth can be observed along with the extension of these cracks. Table 5.2 Local deformation measures for characterizing different limit states

Figure 5.6 The force-drift ratio for wall 2 and 3 with the evolution of the limit states Figure 5.7 shows the location of the cracks and the associated width in order to define the limit states.The first signs of cracks that were associated to LS2 (slight) were the one that has between 0.5 and 1 mm as seen in Figure 5.7 (case of wall 2 associated to drift 0.23%). These cracks start to appear in the first non-linear part of the load-deformation curve of the wall. The cracks corresponding to LS2 were not easily visible directly on the wall, they were only explored through the DIC that revealed their location.This is why in the region where the drift exceeded 0.1 % the behavior was associated with slight damage.

Figure 5.8 summarizes the suggested limit states for rammed earth walls through the association of the drift ratio. Only the value of the LS4 seems to be underestimated in the adaptation of the method. Hereafter, the method proposed by Calvi can be adjusted for the case of rammed earth wall as conducted in this study and the values to be retained for the drift are that of Figure 5.8. Therefore these limits will be used in order to analyses the results of the pushover test in the next gection

Figure 5.9 Elastic spectrum from Eurocode 8 Nonlinear Static Analysis Procedure Input earthquake ground motion is defined by a response spectrum. To analyze the performance of the structure, the response spectra are the main data that must be provided; these spectra are called “elastic response spectra” that depends on many parameters like the type of soil, the zone of seismicity of the corresponding structures. The values of the periods Tg, Tc et Tp and the soil parameter S that describes the response spectra depends on the type of the soil. Figure 5.9 shows the elastic spectrum from Eurocode 8.

Figure 5.10 Conceptual explanations for transformation of MDOF to SDOF system

Figure 5.11 Bilinear idealization for the equivalent SDOF system [120] Seismic Demand in acceleration—displacement (S,-S,) format The correlation between the standard acceleration S,- and the natural period T is plotted in acceleration vs. displacement domain to obtain the demand diagram (the following calculation comes from Eurocode 8). Sac is the elastic response spectrum in displacement, expressed as:

Figure 5.12 Demand spectra in S,-S, format for constant ductility [132]

Figure 5.16 (a) Example of a simple plan of a rammed earth house inspired from (www.architecte-

‘able 5.3 Scale effect parameters from experiments to the scale of rammed earth wall building

Figure 5.19 Idealization of the capacity curves for wall B2 and B3

Figure 5.20 Idealization of the capacity curves for wall D1 and D4

Figure 5.21 Capacity spectrum for different seismicity zones, soil A

Figure 5.23 Visualizing the intersection points between the elastic spectrum and the capacity curves An example of this calculation for wall 2 and 3 for the case of medium seismicity is shown below:

Table 5.5 Inter-storey drifts calculated for soil B Table 5.4 Inter-storey drifts calculated for soil A

Figure A.1 The use of vertical rods as reinforcement for rammed earth walls proposed by Minke, 2001 In a study of [6], it was suggested to reinforce rammed earth walls with vertical reinforcement horizontal reinforcement. This is mainly because the shear force cannot be transferred by t rods due the low bonding between these elements and earth material. On the second hand t presence of these elements can initiate the propagation of horizontal cracks and therefore dama he wall, and eventually, it is practically difficult to ram the wall in the presence of horizon bars inside the framework. ike vertical rods or bamboo (Figure A.1) inside the wall, these bars should be connecting the foundation and a ring beam above the wall. [6] also mentioned the disadvantages of using ne ale ge al the wall, and eventually, it is practically difficult to ram the wall in the presence of horizontal

Figure A.2 Typical reinforced wall detail proposed by NZS 2499 Code \.1. Laboratory testing of retrofitting techniques: Some other researchers conducted experimental laboratory tests to prove the effectiveness of the reinforcement techniques. [135] proposed the post-tensioning rods as a technique of vertical reinforcement as shown in (Figure A.3). This method can be applied for new rammed earth building and can be considered easy to apply when compared to other techniques presented. building and can be considered easy to apply when compared to other techniques presented.

Figure A.3 Test setup and placement of post-tensioned rods [135]

Figure A.4 Rammed earth panel without and with reinforcement after the test In his study, the panels were tested under dynamic excitation using shake table (to simulate the seismic excitations). Results demonstrate that the specimen that was reinforced had fewer cracks and performed better in terms of load levels. However, the relevancy of these reinforcement techniques in the case of the real structure is still questionable.

Figure A.5 Retrofitting scheme of the rammed earth wall: (a) horizontal direction; (b) diagonal direction Other methods intended to increase the soil strength by adding reinforcement material which has a high tensile strength as in [100]. In this study, they proposed a new retrofitting technique using externally bonded fiber materials to improve the mechanical behavior of the structure. First, they seek to find a suitable fiber material that is cost-effective and that is compatible with the rammed earth. Three type of fiber material was proposed (canvas, bamboo, and tarpaulin), these Fiber materials were chosen based on important features: the strength, workability, ductility, availability, and cost. Finally, the tarpaulin selected for this study after many laboratory tests. This retrofitting technique was validated through a set of experimental test for rammed earth walls under monotonic lateral loading. The walls were strengthened by bonding tarpaulin strips around the walls using different strengthening schemes, as in Figure A.5.

Figure B.1 Definition of the grid element and of correlation pattern [139] Digital image correlation (DIC) is a method capable of measuring the deformation at a surface element. It allows, during solicitations, the measurement of the displacement field at the surface of a specimen by comparing a deformed image to a reference image [72]. The method consists in tracking a random speckle pattern on top of the surface using the recorded images. Different techniques are used to have a random speckle-like: spray painting or air brush.

Figure B.2 Iterative process used by the 7D software to obtain the displacement field, according to levels in the deformed image are obtained with bilinear or bi-cubic interpolations. An iterative procedure is adopted in the software to determine the field displacement. For the first iteration, an initial solution of each grid point A,B,C,D is estimated without changing the shape of the correlation pattern. These squares A,B,C,D will then become quadrilateral Ao,Bo,Co,Do on the image that will be correlated. For the rest of iterations, the pattern is deformed in order to minimize the correlation parameter. This operation is repeated for each point in the grid. So the displacement field is determined at every point of the final image as in Figure B.2. displacement field is determined at every point of the final image as in Figure B.2.

Figure C.5 Comparison between the displacement given by the image correlation and the displacement sensor in function of the horizontal force VH for wall 3

Figure C.6 Comparison between the displacement given by the image correlation and the displacement sensor in function of the horizontal force VH for wall 4

Tableau 6 Résultats des tests de compression non confinés Dans cette premiére étude, le sol utilisé a été évalué d'abord en réalisant un essai de compression sur des échantillons cylindriques et prismatiques (Figure 2.11 et Figure 2.12). Les échantillons cylindriques ont un diamétre de 20 cm et une hauteur de 40 cm et les échantillons prismatiques mesurent 25 cm x 25 cm x 50 cm. Les dimensions de ces échantillons ont été choisies pour reproduire l'énergie de compaction appliquée sur les murs pendant la fabrication A partir des résultats des tests de compression (Table 2.2), on a constaté qu'il existait une variation significative de la performance du matériau en fonction des géomeétries sélectionnées de l'échantillon (éprouvettes cylindriques ou prismatiques). La résistance a la compression était supérieure dans le cas de l'éprouvette cylindrique.

Tableau 7 Résultat de petite et grande boite

ranime el nabouch

Procedia Engineering

Rammed earth (RE) construction is attracting renewed interest throughout the world thanks to its sustainable characteristics: a very low embodied energy, an advantageous living comfort due to a substantial thermal inertia, good natural moisture buffering, and an attractive appearance. This is why several studies have been carried out recently to investigate RE. However, there have not yet been sufficient studies on the seismic performance of RE buildings. This paper presents an experimental study on the static nonlinear pushover method and its application on the seismic performance of RE structures. Several walls with two height/length ratios were built and tested to obtain the nonlinear "shear force-displacement" curves. By transposing to the "acceleration-displacement" system and by using the standard spectra presented in Eurocode 8, the performance points could be determined which enabled to assess the seismic performance of the studied walls in different conditions (seismicity zones and soil types).

View PDFchevron_right

Assessing the in-plane seismic performance of rammed earth walls by using horizontal loading tests

ranime el nabouch

Engineering Structures, 2017

Rammed earth (RE) construction is attracting renewed interest throughout the world thanks to its sustainable characteristics: very low embodied energy, advantageous living comfort due to substantial thermal inertia, good natural moisture buffering and an attractive appearance. This is why several studies have recently been conducted to investigate RE. However, there have not yet been sufficient studies on the seismic performance of RE buildings. This paper presents an experimental study on the static nonlinear pushover method and its application to the seismic performance of RE structures. Several walls with different height/length ratios were built and tested to obtain nonlinear shear force-displacement curves. By transposing these shear force-displacement curves to an acceleration-displacement system and using the standard spectra presented in Eurocode 8, the performance points were determined, making it possible to assess the seismic performance of the walls studied in different conditions (seismicity zones and soil types).

View PDFchevron_right

Modelling the Structural Behaviour of Rammed Earth Components

Tiago Miranda

The rammed earth technique has a significant presence in the earthen built heritage, where was used to build from simple dwellings to fortresses. However, the high vulnerability of rammed earth construction to decay agents and to seismic events puts at risk their further existence and the lives of millions of people. With respect to the seismic behaviour of rammed earth walls, the understanding and modelling of their shear behaviour are topics rarely approached in literature. Nevertheless, these topics are of significant importance in the preservation and strengthening of rammed earth constructions. This paper presents experimental and numerical work where the shear behaviour of unstabilised rammed earth is analysed. The experimental program consisted in the testing of several unstabilised rammed earth wallets subject to diagonal compression, which allowed a better understanding of the shear behaviour of unstabilised rammed earth. The numerical work consists of the modelling, of the previous tests, using the finite element method and by considering both the macroand micro-modelling approaches. In general, the numerical models showed a good agreement with the experimental results.

View PDFchevron_right

Analysis of rammed earth construction

Christopher Gerrard

Structural analysis of historical constructions.New Dehli 2006. ISBN 972-8692-27-7, 2006

Rammed earth is a widely used historic building material, found in Mediterranean regions, along the Silk Road, and in parts of the Himalayas. While guidelines exist for the construction of new rammed earth structures, there is very little guidance for the structural analysis of historic structures.

View PDFchevron_right

The strength of unstabilised rammed earth materials

Paul Jaquin, David Toll, Domenico Gallipoli

Géotechnique, 59(5): 487- 490, 2009

Rammed earth is a manufactured material comprising sand, gravel and clay, which is compacted between forms to build walls. Primarily a historic method of construction, it is now receiving considerable interest worldwide owing to its zero reliance on materials such as cement, and its potential for recycling. Despite its longevity, the source of its shear strength is poorly understood. This paper presents initial laboratory test results that point to the main source of strength in rammed earth being suction, and indicating that recent advances in unsaturated soil mechanics may also be applied to this material.

View PDFchevron_right

Methodological proposal for rammed-earth wall characterization:Understanding of material in preliminary studies

Francisco Javier Alejandre Sánchez

Rammed Earth Conservation, 2012

Propose is a series of analyses and tests to be carried out in historical studies for wall characterization. There are three types of tests: a) classic methods that are well referenced and widely used for other building materials (mortars, soils, aggregates, etc.); b) instrument techniques such as XRF, SEM-EDX, XRD, etc.; and c) adaptations of normalization methods mainly corresponding to UNE-EN standards for application of these materials. Five large groupings can be established based on wall properties: chemical and mineralogical composition, physical and hydric properties, mechanical properties, granulometry, and dating. A correct interpretation of the results obtained allows non-analytical researchers (e.g., historians, archaeologists, restorers, architects, and so on) to expand their knowledge of the material's culture or its history. be set up taking into account the following points

View PDFchevron_right

Evaluating shear test methods for stabilised rammed earth

Kepa Morgan

Proceedings of the ICE - Construction Materials, 2012

Rammed earth is an accessible, sustainable and increasingly popular building material. Owing to a lack of research, current design standards for rammed earth have taken a conservative stance on material attributes like shear strength. Evaluating the shear strength of rammed earth is particularly important in seismic areas because of the material's high mass, low ductility and propensity to fail in shear. Shear test methods designed for other materials have typically been used in practice to determine the shear strength of rammed earth. In this research the design shear strength guidance available in current earth building standards was compared with experimental shear strength results for stabilised rammed earth. The triaxial (geotechnical) and triplet (masonry) tests were used to evaluate specimens reinforced with natural fibres: sisal and New Zealand flax. Both shear test methods showed that the shear strength capacity of cement-stabilised rammed earth was greater than the current guidance provided in the earth building standards. Recommendations were made to use the triaxial test to evaluate the shear strength of stabilised rammed earth and to allow the use of design shear strength equal to 7% of the compressive strength.

View PDFchevron_right

Experimental and numerical evaluation of the compressive and shear behavior of unstabilized rammed earth

MARIA ESTHER PUERTAS GARCIA

Materials and Structures

This paper presents novel experimental results and observations from three one-quarter scale tests on two-way concrete slabs supported by protected steel edge beams under fire conditions. The sizes of the protected secondary edge beams were varied to study the effect of beam stiffness on the fire behaviour of the assemblies. Test results showed that as the stiffness of the protected secondary edge beams increased, the slab central deflection decreased and failure of the slab occurred later. However, composite action between the edge beams and the concrete slab plays a key role in mobilising this beneficial effect. Once the composite slab-beam action is weakened by cracks in the slab over the main or secondary edge beams, the benefit associated with a greater stiffness of the edge beams is lost. Tensile membrane action was mobilised at a deflection equal to 0.9 to 1.0 of the slab thickness irrespective of the bending stiffness of the edge beams. The commencement of tensile membrane stage was marked by one of three indicators: (a) concrete cracks which formed a peripheral compressive ring in the slab; (b) horizontal in-plane displacements along the slab edges; and (c) horizontal and vertical displacements of four corner protected steel columns. The test results were used to validate a finite element model developed using Abaqus/Explicit. Good correlation between the predicted and experimental results was obtained.

View PDFchevron_right

Strengthening Strategies for Existing Rammed Earth Walls Subjected to Out-of-Plane Loading

Phuntsho Wangmo

CivilEng

The paper reports an experimental campaign to study the effectiveness of strengthening measures proposed for rammed earth (RE) wall in an out-of-plane direction. Two simple and feasible strengthening techniques were explored, namely, mesh-wrapped and timber-framed strengthening techniques. The test involved testing three full-scale U-shaped RE walls in an out-of-plane direction. The first specimen without any intervention served as the reference wall, while the two others were strengthened with two different strengthening methods. It was observed that both proposed strengthening techniques improved the load-carrying capacity of the wall and the maximum displacement and the energy absorption. The mesh-wrapped strengthening technique was found to be more effective than the timber-framed strengthening technique, which disrupted the visual aspects of the wall’s facade and needed proper anchoring to the foundation.

View PDFchevron_right

Durability Assessment Of Monolithic Rammed Earth Walls

Piotr Woyciechowski

Archives of Civil Engineering, 2015

One of the main threats to constructions made from rammed earth is destruction due to exposure to water. The way to limit this dangerous phenomenon is to supplement the local soil mixtures with stabilizing agents. The main component used is Portland cement. This article analyses the results of research which focused on the resistance of rammed earth to water erosion. Because of the lack of national standards regarding the method of examining the durability of rammed earth, the research was based on the New Zealand standard NZS 4298: 1998. The results confirm the possibility of using rammed earth stabilized by cement in a temperate climate.

View PDFchevron_right

Loading Preview

Sorry, preview is currently unavailable. You can download the paper by clicking the button above.

Rakesh Khare

International Journal of Structural and Civil Engineering Research

View PDFchevron_right

Seismic Behaviour of Unreinforced Masonry

nikhil zade

Lecture Notes in Mechanical Engineering, 2019

View PDFchevron_right

The Mechanical Properties of Masonry Walls - Analysis of the Test Results

Radmila Sindjic Grebovic

Procedia Engineering, 2015

View PDFchevron_right

Pushover Analysis of Concrete Shear Walls: Benchmarking of CAMUS Experiment

prabir basu

2005

View PDFchevron_right

Numerical modeling of rammed earth constructions: analysis and recommendations

ranime el nabouch

Academic Journal of Civil Engineering, 2015

View PDFchevron_right

Investigation and Analysis of Wall Cracks in Cement Stabilized Rammed Earth Technology

Tesila Kandamby

Engineering, Technology & Applied Science Research, 2019

View PDFchevron_right

Structural Response of Unreinforced Masonry Walls

Carlos Daniel Frau

Journal of Civil Engineering and Architecture, 2016

View PDFchevron_right

Seismic analysis of buildings on different types of soil with and without shear wall: A review

Oshin Victor

PROCEEDINGS OF THE INTERNATIONAL CONFERENCE ON SUSTAINABLE MATERIALS AND STRUCTURES FOR CIVIL INFRASTRUCTURES (SMSCI2019)

View PDFchevron_right

Pushover analysis and failure pattern of a typical masonry residential building in Bosnia and Herzegovina

Mustafa Hrasnica

Engineering Structures, 2013

View PDFchevron_right

Preservation and repair of rammed earth constructions

Oriol Domínguez Martínez

Thesis

View PDFchevron_right

The use of non-destructive testing to evaluate the compressive strength of a lime-stabilised rammed-earth wall: Rebound index and ultrasonic pulse velocity

Jacinto Canivell

Construction and Building Materials, 2020

View PDFchevron_right

Experimental Versus Numerical Response of R.C. Walls Subjected to Earthquake Loading

senad medic

2021

View PDFchevron_right

Investigation Of Elasticity, Compression And Shear Strength Of Masonry Wall From Indonesian Clay Brick

Wisnumurti Wisnumurti

View PDFchevron_right

Behavior of Reinforced Wall System During the 1999 Kocaeli (Izmit), Turkey, Earthquake

Erol Kalkan

View PDFchevron_right

STABILIZATION OF RAMMED EARTH

Editor IJRET

View PDFchevron_right

The monitoring of rammed earth experimental wallsand characterization of rammed earth samples

Paulina Faria

Rammed Earth Conservation, 2012

View PDFchevron_right

Experimental Seismic Performance Evaluation of Unreinforced Brick Masonry Shear Walls

Assawin Wanitkorkul

Earthquake Spectra, 2013

View PDFchevron_right

Experimental and Numerical Assessment on Seismic Performance of Earth Adobe Walls

Mohammad Noori

Structural Durability & Health Monitoring, 2021

View PDFchevron_right

The monitoring of rammed earth experimental walls and characterizarion of rammed earth samples

Paulina Faria

Rammed Earth Conservation Proceedings of the First International Conference on Rammed Earth Conservation Restapia 2012 Valencia Spain 21 23 De June 2012 2012 Isbn 9780415621250 Pags 91 97, 2012

View PDFchevron_right

Stress-Strain Relationships and Failure Load Analysis of Cement-Stabilized Rammed Earth under Concentric and Eccentric Loading Using Finite Element Modelling

Geetha Kuntoji

Advances in Civil Engineering, 2022

View PDFchevron_right

A Comparative Experimental Study of the Modulus of Elasticity of Bricks and Masonry

Y. Totoev

2003

View PDFchevron_right

Deformation and Failure Mode of Masonry

Fernando Fonseca, A. Vermeltfoort, Gihad Mohamad

View PDFchevron_right