Experimental study of double-panel confined masonry walls under lateral loading

Author’s Accepted Manuscript Experimental Study of Double-Panel Confined Masonry Walls under Lateral Loading Choayb Belghiat, Ali Messabhia, Jean-Patrick Plassiard, Mohamed Guenfoud, Olivier Plé, Pascal Perrotin www.elsevier.com/locate/jobe

PII: DOI: Reference:

S2352-7102(17)30722-2 https://doi.org/10.1016/j.jobe.2018.09.001 JOBE573

To appear in: Journal of Building Engineering Received date: 21 November 2017 Revised date: 31 August 2018 Accepted date: 2 September 2018 Cite this article as: Choayb Belghiat, Ali Messabhia, Jean-Patrick Plassiard, Mohamed Guenfoud, Olivier Plé and Pascal Perrotin, Experimental Study of Double-Panel Confined Masonry Walls under Lateral Loading, Journal of Building Engineering, https://doi.org/10.1016/j.jobe.2018.09.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Experimental Study of Double-Panel Confined Masonry Walls under Lateral Loading a,c* a c a,b Choayb BELGHIAT , Ali MESSABHIA , Jean-Patrick PLASSIARD , Mohamed GUENFOUD , c c Olivier PLÉ , Pascal PERROTIN . a

Applied Civil Engineering Laboratory / University Larbi Tébessi, Route de Constantine 12002-Tebessa, Algeria. University 8 May 1945 Guelma, BP 401 Guelma 24000, Algeria. c University Savoie Mont Blanc / CNRS / LOCIE, Campus Scientifique Savoie Technolac - 73376 Le Bourget du Lac Cedex, Chambéry, France. b

[email protected] [email protected] [email protected] [email protected] [email protected] [email protected] [email protected]

Abstract Confined masonry structures are one of the most widely used construction systems in Algeria, in which both walls and confining elements contribute to carrying the gravity and seismic loads. Several types of confining elements are used (uniform confinements, toothed confinements), each one giving the structure a different mechanical behavior while the wall density is considered as a key safety factor for those structures. The study presented in this paper is an experimental study on two types of confinement. The study originality stems from the size of the tested samples, which included double-panel confined masonry and experimental pushover tests. First, uniform masonry walls and confinements are studied. Second, toothed masonry walls consisting of two panels of Algerian clay bricks are tested. The objectives were on one hand to study the impact of the frame/panel connection on the quasi-static behavior of the structure as well as the loading and unloading behaviors. On the other hand, to highlight the impact of using double-panel walls on resisting mechanism of confined masonry walls. Moreover, the obtained data were used for evaluating the existing analytical models to predict lateral capacities of confined masonry walls. The results indicate that the two sample-types show some similar resisting mechanisms but various capacities under progressive lateral loading, which may have consequences on the seismic resistance of the structure. A mixture of shear-flexural mechanism can be noticed when double-panel walls are used. Keywords: Confined masonry; uniform confinement; toothed confinement; wall density; double-panel; pushover test.

1. Introduction Masonry is a traditional construction method in Algeria. In terms of material, masonry is a composed assembly of different elements (bricks, concrete blocks, etc.) usually bonded by mortar joints, providing a solid assembly. It is commonly used in the construction of walls, such as bearing walls or infill panels between columns and beams. This composite material has appropriate mechanical characteristics, and it is strongly influenced by its environment and how it is constructed. Several factors make it difficult to take the material into account in structural analysis, especially in a dynamic environment (seismic behavior). Therefore, engineers usually consider the masonry panel as a nonstructural element. However, when talking about confined masonry, the frame/panel interaction cannot be ignored. Confined masonry walls (in this study referred to as a CM walls) are common in Algeria, where the confining elements (tie-columns and tie-beams) are cast around the masonry panel. The confined masonry walls are completely different from the RC frame with masonry infill. In terms of construction sequence, masonry panels are constructed first for CM walls, followed by the cast in-place of RC tie-columns and tie-beams. While, the frame is constructed first in case of RC frame with masonry infill, thereafter, panels are constructed within. Under seismic loading, masonry infill prevents the free deformation of the RC frame, the gaps existing due to a relative lack of bond between the masonry infill and the RC frame, makes them clash and drastically minimize their capability from resisting the lateral forces in a seismic event. The masonry panels support small friction of gravity load and act as compressive diagonal struts. While in CM walls, tie-columns are integrated into the masonry panels, due to the straightforward transmission of forces, panels mostly resist the gravity loads and the assembly (CM wall) act consequently as shear walls [1,2]. This totally different system behavior of two types of walls leads to different resisting and failure mechanisms, the usual mode of RC frame with infill failure is either due to tension in the windward column or due to shear on the columns or beams. However, if the frame strength is sufficient to prevent its failure, the increasing racking load eventually produces failure of the infill. In the most common situations, the state of stress in the infill gives rise to a principal compressive stress along the diagonal of the panel and a principal tensile stress in the perpendicular direction [3]. While, CM walls demonstrate a typical damage pattern in the form of diagonal shear cracks, the failure takes place in the form of a single diagonal crack which propagates through the walls and the tie-columns. In most cases, confined masonry panels demonstrate a shear-dominant seismic response. Longitudinal reinforcement in the RC tie-columns provides an adequate flexural resistance, thus the flexural failure mechanism does not govern [2]. Therefore, a great deal of research has been conducted to evaluate the seismic behavior of these structures, but few studies have experimentally examined the influence of frame/panel connection types on the quasi-static behavior of the structure when a load is progressive. Moreover, the majority of design standards and guidelines [4,5,6] consider the wall density as a key indicator of safety for low-rise confined masonry structures. The wall density which is defined as ratio of the cross-sectional area of all walls in one direction to area of the building floor plan, can be increased using double-panel walls. 2 In this study, tests were carried out on large specimens (1.52×2.06 m ) in loading-unloading conditions. Two types of connections were tested, uniform and toothed connections. Starting with a state of the art in the experimental studies of CM walls, this paper describes the samples and constituent materials, and then explains the experimental processes. Thereafter, the paper discusses the impact of using double-panel walls on the resisting mechanism and compares the two types of connections in terms of failure mechanism, stiffness and dissipated energy. Furthermore, assesses analytical available models for strength and stiffness prediction and finally, suggests a future numerical study.

2. State of the art Much research has been conducted to study the impact of different factors on behavior of CM walls in several conditions. Marinilli and Castilla [7] assess the effect of the number of vertical confining elements on the seismic behavior of confined masonry walls. The authors found that the inclusion of additional vertical confining elements in panels of the same nominal transverse area increases the stiffness and lateral strength of the CM wall, and that this accumulation has no effect on the energy dissipation capacity or the damping ratio. Salinas and Lazares [8] studied the effect of the geometry and density of the masonry elements. The authors recommended structural regularity and the use of high-density bricks. They also preferred solid masonry units and limited the use of hollow masonry units to two-storey buildings. In the same context, Okail et al. [9] concluded that the use of resistant masonry units (concrete units) significantly increases the lateral strength of the structure. Subsequently, the influence on the seismic behavior of the different types of frame/panel connections was studied by Wijaya et al. [10]. The authors noted that the toothed connection and the connection with short steel anchors between the columns and the panel do not improve the structure, but continuous anchorage between the columns strengthens the confinement and develops the strut-column

resistance mechanism, which also leads to diagonal cracking. Also, Matošević et al. [11] concluded that details of column/panel connection do not affect the lateral resistance and initial stiffness but improve the nonlinear behavior and the behavior factor as well as the dissipation energy capacity of CM walls. Furthermore, Meli et al. [2] defined the shear and flexural resisting mechanisms as predominant mechanisms of CM walls during earthquake. They also suggested a method to evaluate the maximal resistance capacity for each mechanism. In addition, Ghaisas et al. [12] and Brzev and Pérez [13] proposed a Strut-and-Tie-Model to predict the behavior of confined masonry buildings and to outline a guideline for developing the model for seismic design of CM walls. However, Rai et al. [14] has shown that the toothed confining method prevents the deflection (out of plane failure) of the wall and maintains the structural integrity of the assembly even in the event of severe damage (e.g., earthquake damage). Similarly, Kusumastuti et al. [15] found that the anchorage of column/panel connection limits the damage area on masonry walls and prevents damage in the out-of-plane direction. They subsequently confirmed that the confinement of masonry leads to a significant improvement in strength and structural stiffness and changes the crack propagation pattern. Unfortunately, these previous studies are based on separate observations and few experimental studies are undertaken to compare uniform and toothed frame/panel connections, particularly on double-panel walls.

3. Description of prototypes tested and materials used To represent one of the most widely used building types in Algeria, the prototype studied is adopted to be the middle frame of a two-floor three-bay typical residential building (Fig. 1). The tested CM walls prototypes are designed to be half-scale models while maintaining the height/length ratio. The samples (2.06 × 1.52 m²) comprise two masonry panels separated by a 50-mm air gap (double-panel walls) and confined by a surrounding RC frame casting of same section for practical reasons 0.25×0.25 m² (Fig. 1). Fig. 1: (a) Frame with uniform confining elements, (b) frame with toothed confining elements.

The confining elements were pre-dimensioned according to Algerian seismic regulations RPA2000 V12 [6]. Along with surrounding frame elements, four steel bars (12 mm in diameter) spaced with 20 cm were adopted for the longitudinal reinforcements. Bars of 6 mm diameter spaced with 15 cm were adopted for transverse 3 reinforcements. The walls were made of hollow bricks (see Fig. 2) measuring 300×200×100 mm (from Algeria) and laid with cement mortar. Four specimens were built in the LOCIE laboratory of Savoie Mont Blanc University (France) in partnership with the LGCA laboratory of the University of Tebessa (Algeria). Fig. 2: The used Algerian bricks The first and second specimens, noted PC1 and PC2, represent two uniform confined masonry walls (Fig. 1a), while the third and fourth specimens, noted PH1 and PH2, represent two toothed confined masonry walls (Fig. 1b). All the building materials came from Algeria except the aggregates. They were selected to correspond to the regulations in Algeria. The strength classification of the materials is reported in Table (1). Table 1. Material classes. Material

Class

Cement

CEM I (42.5 N)

Concrete

C25/30

Steel

B500C and FeE 235

Brick

BCR05

Mortar

M15, T 8-12 2

Several series of tests were carried out to check the different materials. A total of twelve 16x32 cm concrete cylinders have been realized in accordance with EN 12390-2 [16]. Eight cylinders were used for compression tests and eight for splitting tensile tests following EN 12390-3 and EN 12390-6 [17,18] respectively. Six tensile tests were performed on six steel bars of 65 cm length (three on 12 mm diameter and three on 6 mm diameter). Moreover, three compression tests for each direction of the bricks (compared to perforations see Fig.2) were performed. In accordance with EN 1015-11 [19], compression tests and bending tensile tests (three 3 point flexural tests) were performed on 4x4x16 cm mortar prisms. The tables below summarize the obtained results.

2

Table 2. Test results for concrete - 16×32-cm concrete cylinders (average value).

Test

Compression tests

Parameter

Splitting tensile strength tests

RC (MPa)

E (MPa)

εPIC

RT (MPa)

Uniform samples

27.93

14 333

2.82E-3

2.933

Toothed samples

27.66

14 740

2.75E-3

2.842

Table 3. Compression and bending tensile strengths of mortar (average value).

Test

Bending tensile strength (MPa)

Compression strength (MPa)

Uniform panels

4.72

3.80

3.89

16.73

16.25

16.26

Toothed panels

3.86

3.70

4.21

15.13

15.50

14.00

Table 4. Brick test results.

Load orientation compared to perforations

Compression strength

Apparent Young modulus

Test no.

Perpendicular direction (MPa)

Parallel direction (MPa)

Vertical direction (MPa)

1

0.42

5.35

0.96

2

0.43

6.15

0.66

3

0.49

4.85

0.86

1

11.64

150

36.67

2

12.07

234

31.67

3

11.64

201

41.11

Table 5. Steel bars test results.

Bar types Test no. Ultimate strength (MPa) 1 495.4 Ф6 2 491.7 3 470.1 1 600.7 Ф12 2 607.0 3 595.3

Elastic limit (MPa) 294.5 288.12 284.00 472.42 492.32 482.67 2

Table 6 Diagonal and vertical compression test results on masonry prisms - 1-m .

Parameter Values (MPa)

Compression strength 1.13

Shear strength (tensile) 0.82

Young module 3 919

Shear module 1 646.5

The results of the analyses reported in Tables (2–5) show that the materials used satisfy the regulatory requirement classes in Table (1) very well. Results shown in table (6) have been obtained from diagonal and vertical compression tests on square masonry panels according to the standard test method for diagonal tension (shear) in masonry assemblages [20] and the British standard methods of test for masonry [21].

4. Experimental tests of confined masonry walls The main objective of this paper is to study the effect of frame/panel connection type on the quasi-static specimen’s behavior subjected to lateral monotonic loading for one sample and to progressive load-unload cycles for the other one. The set of tools and equipment used to conduct the experiments is shown in Fig. (3).

The service live load and the dead load are estimated to be 80 KN (representing the load paths of upper floor) and they are uniformly applied to the top of the beam. The load distribution device used is shown in Fig. (3), it ensures the independence of the vertical loading from horizontal loading and also eliminates friction forces on the top surface of tie-beam. Fig. 3 Test set-up. The vertical load was applied by a load-control electric actuator, which has a load capacity of 120 kN. In the first phase, the vertical loading is applied by 10-kN increments until it reaches the desired vertical load of 80 kN, which is maintained for the remainder of the test. In the second phase, the horizontal load is applied by a displacement-control hydraulic actuator with a load capacity of 300 kN. At the top of the samples, a displacement of ± 1 mm/min is applied to correspond to a quasi-static loading. The displacements are detected using a high-precision camera that takes successive images of the whole wall tested every two seconds and during the full test. To fully monitor the tests, four displacement transducer devices were used to observe the heaving and sliding, the out-of-plane deflection of samples as well as the displacement of the test machine. On the same image series, the sliding and reversal of samples were evaluated using the image correlation technique to correct the evaluated displacement of beams. The 7D software [22] was used to obtain the field of deformations and the crack openings using the Digital Image Correlation technique. This method has already been used in previous studies [23,24,25] at the LOCIE laboratory on structural elements of equivalent size. It detects displacements on the order of 20 μm. The synchronization of all acquisition systems then links local phenomena (cracking, crushing, etc.) to the variations of the loading curve. Curves shown in Fig. (4) for PC1 and Fig. (5) for PH1 are obtained using the image correlation technique. They represent the changes in beam displacement as a function of load (the tie-beam displacement is evaluated at the end point of the beam axe opposite the horizontal actuator). The two curves systematically exhibit a linear initial phase, followed by a second phase for which the stiffness is modified. The first phase from 0 to 86 kN is an elastic phase characterized by a stiffness of 176 kN/mm for sample PC1. For sample PH1, the elastic phase extends up to 99 kN, with a stiffness of 201 kN/mm. Then the samples’ nonlinear responses appear which are characterized by a loss of stiffness. Based on the image analysis (Figs. 3 and 4) provided by the software 7D [22], stiffness degradation occurs because of the crack initiation and propagation. Fig. 4: Response curve of sample PC1.

The images referred to as (a) designate the appearance of a first intra-brick crack in the masonry panel with a staircase pattern, following the path of the mortar along the bricks. The images denoted (b) indicate the appearance of large cracks in the panel, whereas the images denoted by (c) and (d) designate the cracking of the column and the crack propagation pattern under the maximum load attained, respectively.

Fig. 5: Response curve of sample PH1.

The second and fourth cyclic (load-unload) tests were conducted on samples PC2 and PH2, respectively. The response curves of the two samples are shown in Figs. (6) and (7). For specimen PC2, the loading curve is increased linearly to 81 kN, which has an initial stiffness of 200 kN/mm, whereas the linear response of specimen PH2 extends from 0 to 105 kN with an initial stiffness of 198 kN/mm. The various phenomena degrading the stiffness appear as the first crack in the panel (Figs. 6a, 7a), then large cracks in the panel (Figs. 6b, 7b), large cracks in the reinforced concrete frame (Figs. 6d, 7d) and as crack propagation patterns at the beginning of each unloading cycle (Figs. 6c, 6e, 6f, 7c, 7e, 7f).

Fig. 6: Response curve of sample PC2. Fig. 7: Response curve of sample PH2.

5. Test results analysis The present section is devoted to analyzing the obtained results from monotonic and cyclic loading tests performed on four CM walls, the cyclic loading is performed in irreversible way due to lack of necessary set-up, the effect of this limitation is discussed for the parameters evaluated from cyclic tests comparatively with past literature researches.

5.1.

Resisting mechanism

As shown in the (b) images, the cracks appear and propagate, maintaining the same shape on the diagonal area of the panel (staircase form). Due to the deformation direction analysis provided by 7D software [22] and illustrated in Fig. (8a), the crack is caused by the debonding of the brick/mortar interface. The interface between the brick and the horizontal joint exhibited a tensile failure (mode I opening) and the interface between the brick and the vertical joint exhibited a shear failure (mode II in the shear plane). This occurred to shear failure as discussed by Meli et al.[2] and Flores and Alcocer [26]. Otherwise, these authors noted that shear mechanism can be expected when the column depth does not exceed 1.5 times the wall thickness, this is what the study has confirmed through the calculated ratio (25/20 = 1.25). Afterward, the cracks propagate through the panel to the column at the loading level (Figs. 4c, 5c, 6c, 7d). Those cracks correspond to shear cracks through the similar formed angle (θ = 26°) of panel diagonal cracks as shown in Fig. (8b). After this level, cracks that appeared at mid-height of the column alongside the actuator were related to the loading level (Figs. 4d, 5d, 6d, 7e). These cracks correspond to bending cracks that propagate perpendicularly from the outside column surface towards the inside one. As shown in Fig. (8c), the deformation directions pointed out that outside column fibers are in tension and the column is consequently under bending. This combined shearflexural mechanism was previously observed by Matošević et al. [11] on toothed CM walls. An analytical model can be proposed to estimate the resistance of the CM wall. Firstly , using the shear calculations capacity through the initial phase until the severe cracks of panel (Flores and Alcocer [26] and Alcocer et al. [27] models). Secondly , using the flexural calculation capacity (as proposed by Meli et al. [2]) taking into account the residual shear capacity of cracked masonry panel. Furthermore, the same mechanism was established on all samples, diagonal cracks were dominant and no panel/column separation was observed. Yet, it can be concluded that the toothed panel/frame connection does not affect the resisting mechanism. It is worth noting that the same results are found in works of Wijaya et al. [10]. Fig. 8: (a) Cracking of PC1 panel from Fig. 3d; (b) Shear cracking of PC2 column from Fig 5f; (c) Flexural cracking of PC2

column from Fig. 5f.

On one hand, during the four tests, no separation of the panel/frame connection was observed (Figs. 4, 5, 6, 7). It can be pointed out that the samples behaved in perfect coherence through the good connection established by the masonry confinement (Due to characteristics of the brick´s face in contact with the tie-column, and how concrete casting is performed). On the other hand, the cracks’ continuity from the panel towards the RC frame column guarantees that the panels are integrated into the frame elements. As a result, masonry confinement ensures structural integrity as noted in several researches and guidelines as well as design standards ([5,6,7,8,12,28,29]).

5.2.

Stiffness

To prove that the repeatability and reproducibility of the tests are good as well as the presence of envelope curves, the cyclic and monotonic curves of each type of sample are combined as shown in Fig. (9). Moreover, it can be concluded, when the four curves obtained are compared, that the two sample types show a similar initial stiffness. Consequently, as noted by Wijaya et al. [10] and Matošević et al. [11], the toothing of confined masonry walls does not play a major role on the initial sample stiffness. Which means, the masonry panels resist initially the effects of lateral loads on CM wall but the columns do not play a significant role, similar results are pointed out in Meli et al. [2] works. Fig. 9: Envelope curves.

The tendency curves shown in Fig. (10) are deduced from the load-unload cycles carried out during the two tests. These curves represent the changes in stiffness established by the slopes of the loading phases as a function of the displacements corresponding to the beginning of the cycle. The previous curves suggest that the toothing of confined masonry walls do not significantly affect the stiffness degradation comparing with uniform sample results. But slight differences between both types were observed (see Fig. 10). Matošević et al. [11] show that similar stiffness degradation is found when toothed and uniform CM wall are compared. It must be noted that, the slight difference observed between the two types can disappear if reversible tests are performed. Fig. 10: Changes in stiffness as a function of horizontal displacement of the beam for PC2 & PH2.

5.3.

Dissipated energy

A large energy dissipation capacity is required to limit the damage induced by earthquakes. Therefore, the two samples are compared in term of dissipated energy. The area under the cyclic curves of loading and unloading is used to define the cumulated energy dissipated during each cycle Fig. (11). These curves are relatively similar, but the PH2 curve shows slightly higher values than the PC2 curve after 3 mm drift. Those curves then demonstrate that the two sample types (toothed and uniform) show some similar energy dissipation capacity's. However, toothed CM walls are preferred in term of energy dissipation capacity in work of Matošević et al. [11]. it must be noted that the discrepancy of result obtained by irreversible cyclic test performed in the present paper and the reversible cyclic test performed in work [11] suggest that, the irreversibility of cyclic loading have probably an effect on the energy dissipation capacity’s comparing with complete reversible cyclic loading. A sufficient number of cycles must also be performed. Fig. 11: Cumulative dissipated energy for PC2 and PH2 samples.

5.4.

Cracks location and opening

To better compare the two types of sample in terms of crack location and opening, the path distributions of maximum shear strain are used to designate the cracks and with larger image pixels using 7D software [22], as shown in Fig. (12). Fig. 12: Maximum shear strain distributions for sample: (a) PC1, (b) PH1, (c) PC2, (d) PH2.

Regarding image (a) for the PC1 sample and image (b) for sample PH1, it can be noted that the uniform sample distributes the cracks more than the toothed sample. The same result is seen when comparing images (c) and (d) for samples PC2 and PH2, respectively. Moreover, assuming that the value of maximum shear strain indicates the size of the crack, sample PH1 shows an overall shear strain value of 0.017 in the cracks; for the same loading level (250 kN) the value 0.007 is the overall shear strain in the sample PC1 cracks. Similarly, for a 275-kN load, sample PH2 exhibits more open cracks than sample PC2 in terms of their overall values of shear strain in cracks (0.016 and 0.014, respectively). As a result, the toothed samples exhibit more open and localized cracks than the uniform samples. These data results can be useful for the application of the Strut-andTie method on uniform as well as toothed CM walls as proposed by Ghaisas et al. [12] and Brzev and Pérez [13]. Where, the location and orientation of strut and tie elements depend on paths of stress fields.

6. Analytical investigation of lateral capacity of studied specimens 6.1.

In-plane shear strength prediction

In order to predict the cracking load associated to the first significant crack in masonry panel, several researches (A. Matsumura [30] and Moroni et al.[31]) and standards as Chilean Standard [32] proposed analytical models based on friction theory (Mohr-Coulomb theory). Table (7) summarizes the proposed formulae reported form works of Riahi et al. [33]. Table 7: Analytical models for crack strength prediction of CM walls.

Authors

A. Matsumura [30]

Formula (

Comments )

√

: reduction factor 0.64 for partially grouted walls. : is the effective width of wall.

(

Chilean Standard [32]

⁄ )

, , : represent length, width, thickness and cross ear of wall respectively. : width of tension column. : compressive and shear strength of masonry.

Moroni et al. [31]

: vertical applied stress.

Riahi et al. [33] Flores and Alcocer [26]

The formulae mentioned above in table (7) are used to analytically predict the load at the first significant diagonal crack in CM walls. Obtained results have been interpreted as ratio to experimentally obtained results for the four samples (see histogram in Fig. (13)). The predicted loads were designed by (Vxy), where, letters (x, y) denote the first letters of authors names proposing the used formula. As shown in Fig. (12), the models proposed by A. Matsumura [30], Chilean Standard [32] and Moroni et al. [31], highly underestimate the loading causing the first diagonal cracks as the significant low ratio denotes (Vcal/Vexp ≤ 0.82). The underestimated result derived by Chilean Standard [32] can be caused by the implicit factors usually used for design purpose in standard´s formula. It is worth noting that a contradictory result has been obtained concerning those models in the study of Riahi et al. [33]. The model proposed by Flores and Alcocer [26] shows much conservative values (1.12 ≤ Vcal/Vexp < 1.6), but still stands as a good approximation indicator for first significant diagonal cracks of CM walls as noted by Matošević et al. [11]. The best conservative values have been found using the model proposed by Riahi et al. [33] in which various obtained (Vcal/Vexp) ratio still stands in the interval [0.99; 1.41]. As a conclusion, the model proposed in Riahi et al. [33] seems to be the most efficient to predict the load at first significant diagonal cracks for the four CM walls tested. Fig. 13: Ratios obtained from analytical prediction of CM walls strengths at first diagonal crack.

6.2.

In-plane stiffness prediction

For predicting the initial stiffness of studied CM walls, the equivalent strut method of infilled frames is used based on assuming that the stiffness of global system can be expressed as the sum of flexural stiffness of frame bars and the stiffness towards horizontal translation of strut-tie element as mentioned by V. Bergami [34]. Where, the width of the equivalent masonry strut is required. For this purpose, the two analytical formulae proposed by A.W. Hendry [35] and Papia et al. [36] are used to estimate the strut width . Furthermore, based on shear and flexural deformation in lateral loading case, Rai et al. [14], Flores and Alcocer [26] and Riahi et al. [33] proposed analytical models to predict the stiffness of confined masonry walls as explicitly detailed in work of Singhal and Rai [37]. Therefore, the various formulae used in this study are summarized in table (8). Table 8 Analytical models for stiffness prediction at first diagonal crack in CM walls. Authors

Formula

Comments

V. Bergami [34]

( )

+0.567 Papia et al. [36]

(

)

and

√ A.W. Hendry [35]

(

)

(

) √

Riahi et al. [33] (

(

))

Rai et al. [14]

Flores and Alcocer [26]

((

)

)

((

)

)

: inertia moments of column, beam and masonry panel. : distance from bottom section of column to centroid of beam and the distance between column axes. is an empirical constant. , , : represent respectively the height, length, thickness and diagonal length of masonry panel. : cross section of column, beam and masonry panel. : Stiffness parameter and width of equivalent masonry strut. : concrete Young module, Young and shear masonry modulus and poison coefficient of masonry. : compressive strength of masonry. ( : angle made by strut with horizontal. : confinement factor defined as the ratio of the total centerline length of internal grid elements, , to the centerline length of confining elements at the perimeter of the wall, .’ : contact length between wall and column and beam respectively. : cross section and height of wall. : effect of unit material on the panel rigidity, equal to (1,13) for clay brick and (0,72) for concrete block.

In order to assess the efficiency of existing models to predict the CM walls stiffness at the first significant crack in panel, the obtained results using the formulae mentioned in table (8) are illustrated as ratio of calculated stiffness to those obtained experimentally. As shown in Fig. (14), when the formula proposed by Papia et al. [36] is used to evaluate the width of strut. The equivalent strut model underestimates, for all samples, the first cracking stiffness of CM walls. Similarly, the model proposed by Rai et al. [14] also significantly underestimate the stiffness of walls. Riahi et al. [33] model overestimates the stiffness, where the obtained ratio pointed out a double experimental values. Moreover, the model proposed by Flores and Alcocer [26] predict the stiffness of CM walls in a satisfying way in which the ratio values obtained stand within interval [0.97 1.1]. Furthermore, the better predicting model is the equivalent strut method when the width of strut is calculated using the formula proposed by A.W. Hendry [35]. This model has very well predicted, in a conservative way, the first crack stiffness of all CM walls (1 ≤ Kcal/Kexp ≤ 1.15). Fig. 14: Ratios obtained from analytical prediction of CM walls stiffness’s at first diagonal crack.

7. Conclusions This paper reports a full experimental study in monotonic conditions as well as loading-unloading conditions of the influence of panels/frame connection details on the behavior of confined masonry wall commonly used in

Algeria. Series of material characterization test results are also reported in this paper. Based on the analysis of the results and the observations during the tests, the following conclusions can be drawn: 1. The mechanical properties of the materials used in sample design were identified from a series of characterization tests, in which the conformity of materials to the regulatory requirements was demonstrated. 2. When the toothed and uniform connections are compared, it must be noted that the toothed panel/frame connection does not affect the resisting mechanism and initial samples stiffness. This result is consistent with others found in the literature. Both types of system ensure the structural integrity at moderate loading levels. Furthermore, they show similar energy dissipation capacity’s as well as similar stiffness degradation amounts, when irreversible cyclic loading tests were adopted. Otherwise, the contradicting results with literature regarding the energy dissipation capacity can be attributed to the irreversible cyclic tests adopted. 3. Toothed samples exhibit localized damages with one main crack. On the contrary, the uniform confining of masonry walls leads to a more uniform stress distribution. As a consequence, the damage is more diffused, but with less open cracks. This result could have consequences on the seismic resistance where one seeks dissipating energy while keeping the integrity of the structure. Localized and bigger damage are not necessarily safe. However, these data results can be useful for the application of the Strut-and-Tie method on uniform as well as toothed CM walls in which the location and orientation of strut and tie elements depend on paths of stress fields. 4. The current experimental results from double-panel confined masonry walls were used to assess the efficiency of various existing models to predict the shear load and stiffness at the first significant diagonal crack of walls. Most of analytical models proposed failed to capture the shear load at the first diagonal crack in walls. Only reasonable predicted values were obtained for model proposed by Riahi et al. [33]. However, regarding the stiffness prediction case, the models proposed by Rai et al. [14] and Riahi et al. [33] provided respectively significant underestimated and overestimated values. While, the model proposed by Flores and Alcocer [26] predict the stiffness of CM walls in a satisfying way. Furthermore, the better predicting model is the equivalent strut method when the width of strut is calculated using the proposed formula by A.W. Hendry [35]. The model very well predicted, in a conservative way, the first crack stiffness for all CM walls. The two traditional methods therefore show some similar resisting mechanisms under progressive lateral loading. But some various capacities must be noted. It must be noted that, the large variety of existing masonry units, the use of double panels and irreversible cyclic tests performed, are among the most important factors preventing the generalisation of these results. In the near future, these tests will be simulated and used to outline a guideline for seismic design of CM building in Algerian regions.

Acknowledgements This research was supported by LOCIE Laboratory of Savoie Mont Blanc University (France) in partnership with the LGCA Laboratory of the University of Tebessa (Algeria). Funding sources are gratefully acknowledged. The experimental study was possible because of the enthusiastic collaboration of several professors and technicians. Thanks are extended to: Mr. Abdelkrim Nefissi, Mr. Smaali Bouziane, Mr. Goldin Thierry, for their assistance during the experimental study, and to: Mr. Farid Messaoud, Mr. Samir Douza for providing language help.

Declarations of interest none

References: [1]

P. Balandier, Constructions de Maçonnerie, in: DDE La Martinique - SECQUIP, n.d.

[2]

R. Meli, S. Brzev, M. Astroza, T. Boen, F. Crisafulli, J. Dai, M. Farsi, T. Hart, A. Mebarki, A.S. Moghadam, D. Quiun, M. Tomazevic, L. Yamin., Seismic Design Guide For Low-Rise Confined Masonry Buildings, 2011.

[3]

P.. Asteris, D.. Kakaletsis, C.. Chrysostomou, E.. Smyrou, Failure modes of infilled frames, Electron. J. Struct. Eng. 11 (2011) 11–20.

[4]

ACI 530-05, Building Code Requirements for Masonry Structures., 2011.

[5]

FEMA, FEMA 306 : Evaluation of Earthquake Damaged Concrete and Masonry Wall Buildings., (1998).

[6]

RPA99, Regles Parasismiques Algeriennes 99 / Version 2003, 2003.

[7]

A. Marinilli, E. Castilla, Experimental Evaluation of Confined Masonry Walls with Several ConfiningColumns, 13th World Conf. Earthq. Eng. 2129 (2004).

[8]

R. Salinas, F. Lazares, Seismic Performance Of Confined Masonry Buildings With Tubular Bricks In Developing Areas, 14th World Conf. Earthq. Eng. Beijing, China. (2008).

[9]

H. Okail, A. Abdelrahman, A. Abdelkhalik, M. Metwaly, Experimental and analytical investigation of the lateral load response of confined masonry walls, HBRC J. 12 (2016) 33–46. doi:10.1016/j.hbrcj.2014.09.004.

[10]

W. Wijaya, D. Kusumastuti, M. Suarjana, Rildova, K. Pribadi, Experimental study on wall-frame connection of confined masonry wall, Procedia Eng. 14 (2011) 2094–2102. doi:10.1016/j.proeng.2011.07.263.

[11]

Ɖurđica Matošević, V. Sigmund, I. Guljaš, Cyclic testing of single bay confined masonry walls with various connection details, Bull. Earthq. Eng. 13 (2014) 565–586. doi:10.1007/s10518-014-9627-8.

[12]

K. V. Ghaisas, D. Basu, S. Brzev, J.J. Pérez Gavilán, Strut-and-Tie Model for seismic design of confined masonry buildings, Constr. Build. Mater. 147 (2017) 677–700. doi:10.1016/j.conbuildmat.2017.04.200.

[13]

S. Brzev, J.J.P. Gavilán, Seismic Design of Confined Masonry Buildings Using the Strut‐and‐Tie Method, 2013.

[14]

D.C. Rai, V. Singhal, S. Paikara, D. Mukherjee, Sub-paneling of masonry walls using precast reinforced concrete elements for earthquake resistance, Earthq. Spectra. 30 (2014) 913–937. doi:10.1193/102010EQS178M.

[15]

D. Kusumastuti, M. Suarjana, K.S. Pribadi, A. Rildova, T.H. Lie, Experimental Study on Typical Confined Masonry Structure under Cyclic Lateral Load, 15th World Conf. Earthq. Eng. (2012).

[16]

NF EN 12390-2, Essais pour béton durci - Partie 2 : confection et conservation des éprouvettes pour essais de résistance., 2001.

[17]

NF EN 12390-3, Essais pour béton durci - Partie 3 : résistance à la compression des éprouvettes., 2003.

[18]

NF EN 12390-6, Essais pour béton durci - Partie 6 : détermination de la résistance en traction par fendage d’éprouvettes., 2001.

[19]

NF EN 1015-11, Méthodes d’essai des mortiers pour maçonnerie - Partie 11 : détermination de la résistance en flexion et en compression du mortier durci., (n.d.).

[20]

ASTM E 519-02, Standard Test Method for Diagonal Tension (Shear) in Masonry Assemblages, Am. Soc. Test. Mater. (2002) 5. doi:10.1520/E0519.

[21]

EN 1052-1:1998 Methods of test for masonry - Part 1: Determination of compressive strength, 1998.

[22]

P. Vacher, S. Dumoulin, F. Morestin, S. Mguil-Touchal, Bidimensional strain measurement using digital images, Proc. Inst. Mech. Eng. Part C J. Mech. Eng. Sci. 213 (1999) 811–817. doi:10.1243/0954406991522428.

[23]

M. Eymard, J.P. Plassiard, P. Perrotin, S. Le Fay, Mechanical influence of an innovative sprayed coating

for thermal insulation on a shear brick wall, Constr. Build. Mater. (n.d.). [24]

R. Nabouch, Q.B. Bui, O. Plé, P. Perrotin, C. Poinard, T. Goldin, J.P. Plassiard, Seismic Assessment of Rammed Earth Walls Using Pushover Tests, Procedia Eng. 145 (2016) 1185–1192. doi:10.1016/j.proeng.2016.04.153.

[25]

B. Sentosa, Q.B. Bui, O. Ple, P. Perrotin, J.P. Plassiard, Assessing damage of beam-column connection in RC structures from dynamic measurement results, Struct. Eng. Int. (n.d.).

[26]

L.E. Flores, S.M. Alcocer, Calculated response of confined masonry structures., Elev. World Conf. Earthq. Eng. (1996) Paper No. 1830.

[27]

S.M. Alcocer, J. Cesin, L.E. Flores, O. Hernandez, R. Meli, A. Tena, D. Vasconcelos, The New Mexico City Building Code Requirements for Design and Construction of Masonry Structures, Ninth North Am. Mason. Conf. (2003) 656–667, South Carolina, USA.

[28]

T. Schacher, Confined Masonry for one and two story buildings in low-tech environments, 2011.

[29]

Eurocode 8: Design of structures for earthquake resistance-Part 1: General rules, seismic actions and rules for buildings., 2004.

[30]

A. Matsumura, Shear strength of reinforced masonry walls, 9th World Conf. Earthq. Eng. Vol. 6 (1988) 121–126 JAPAN.

[31]

M.O. Moroni, M. Astroza, S. Tavonatti, Nonlinear models for shear failure in confined masonry walls, Mason. Soc. J., 12_2_, 72–78. (1994).

[32]

INN-1997., Albanileria confinada-requisitos de deseno y calculo., Instituto Nacional de Normalizacion, Santiago, Chile, 1997.

[33]

Z. Riahi, K.J. Elwood, S.M. Alcocer, Backbone Model for Confined Masonry Walls for Performance-Based Seismic Design, J. Struct. Eng. 135 (2009) 644–654. doi:10.1061/(ASCE)ST.1943-541X.0000012.

[34]

A.V. Bergami, Implementation and experimental verification of models for nonlinear analysis of masonry infilled r . c . frames, Roma, 2007.

[35]

A.W. Hendry, Structural Masonry, 2nd ed, Ma, n.d.

[36]

M. Papia, L. Cavaleri, M. Fossetti, Infilled frames: Developement in the evaluation of stiffening effects of infills, Struct. Eng. Mech. 16 (2003) 675–693. doi:10.12989/sem.2003.16.6.675.

[37]

V. Singhal, D.C. Rai, In-plane and out-of-plane behavior of confined masonry walls for various toothing and openings details and prediction of their strength and stiffness, Int. Assoc. Earthq. Eng. 44 (2016) 657–675.

Highlights    

The paper presents an original full scale experimental works for one of most building systems used in Algeria. The studied prototypes consist of double panels of bricks embedded in a concrete reinforced frame structure. Different connections wall/frames are tested. Pushover tests are done on four prototypes and the correlation images technic is used in analysis of results. Evaluating of existing analytical models to predict lateral capacities of confined masonry walls.