- Email: [email protected]
Masonry infill construction and retrofit technique for the infill-frame interaction mitigation: Test results
Engineering Structures 132 (2017) 597–608
Contents lists available at ScienceDirect
Engineering Structures journal homepage: www.elsevier.com/locate/engstruct
Masonry infill construction and retrofit technique for the infill-frame interaction mitigation: Test results Marco Preti ⇑, Valentino Bolis Department of Civil, Environmental, Architectural Engineering and Mathematics, Università degli studi di Brescia, Via Branze 43, 25123 Brescia, Italy
a r t i c l e
i n f o
Article history: Received 23 March 2016 Revised 3 November 2016 Accepted 21 November 2016
Keywords: Seismic downgrade Damage control Infill-frame interaction Vertical sliding joints Out-of-plane strength Sliding panels Masonry partitioning Masonry infill Existing infill Infill testing
a b s t r a c t The paper describes the set-up and testing of an innovative construction technique for masonry infill, which can provide a flexible and predictable in-plane response to the infill inside the frame, together with a stable and reliable out-of-plane response. The design strategy is to downgrade the infill reaction inside the structural frame thanks to a dramatic reduction of the masonry in-plane stiffness. The infill is partitioned by vertical planks (or equivalent beams) into sub-panels, free to relatively slide and rock on their toes. The planks connected to the beams provide the necessary out-of-plane stability. The solution was tested for application in both new and existing infills and construction details are discussed. A comparison is also presented with the performance of two infills, one continuous and one with horizontal subpanels, previously tested under the same conditions. The observed infill downgrade makes practically negligible the infill-frame interaction and the post-earthquake masonry damage. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction The research work here presented tries to address the wellknown issues related to the infill-frame interaction in multi-story buildings. The paper describes the set-up and testing of an innovative infill construction technique, which can provide a deformable and predictable in-plane response to the infill inside the frame, together with a stable and reliable out-of-plane response. The design strategy is to downgrade the infill contribution to the building seismic response thanks to a dramatic reduction of its in-plane stiffness. In fact, the uncertainty of the traditional masonry infills in- and out-of-plane response [1] and their possible irregular distribution in the structure jeopardize the building safety and resilience. The post-earthquake damage associated to a poor infill performance highly contributes to the cost and duration of the reconstruction process and activity recover [2,3], even after moderate intensity earthquakes [4]. Despite the possible detrimental effects of the infill-frame interaction, the post-earthquake damage survey showed also in some cases their contribution in preventing the collapse of poorly detailed buildings, not designed to withstand ⇑ Corresponding author. E-mail addresses: [email protected] (M. Preti), [email protected] (V. Bolis). http://dx.doi.org/10.1016/j.engstruct.2016.11.053 0141-0296/Ó 2016 Elsevier Ltd. All rights reserved.
seismic actions; but this contribution is not always reliable due to the possible activation of undesired collapse mechanism in the structure. However, because of their efficiency in terms of construction ease, internal climate control and low building costs, traditionally constructed infills remain widely used, even if in the last decades, a large number of infilled frame buildings have performed poorly during earthquakes [5,6]. The increasing demand for post-earthquake damage control justifies the development of infill typologies for new buildings, capable to survive moderate to intense earthquakes without damage. In the last decade, several authors have proposed engineered masonry infill solutions to address this issue [7–11]). Mohammadi et al. [7] and Preti et al. [8] proposed the horizontal partitioning of the infill. Misir et al. [9] investigated the response of infills made of blocks without mortar, providing out of plane stability by a particular interlocking between blocks of adjacent rows (locked brick infill). Markulak et al. [10] proposed the use of weaker masonry blocks located close to the columns to accommodate the frame deformations. Vailati and Monti [11] substituted the mortar joints with plastic ones to be used with hollow blocks. Other ongoing research projects, aimed at optimizing the design of earthquake resilient infills, were presented in [12–16]. The construction technique here presented stems from the research work presented by Preti et al. [8,17], inspired by historical
598
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
partitioned masonry structures [18] and recently updated towards industrialization by Morandi et al. [19]. The technique adopted the partitioning of the infill in masonry sub-panels connected to the columns for ensuring out of plane stability, but free to move relatively along horizontal planks, embedded in mortar beds. The performed tests showed a ductile infill response provided by the horizontal partitioning, without damage and with a significant stiffness reduction with respect to a traditional masonry infill. The innovation of the here presented solution consists in the configuration of the sliding planks (or equivalent beams), which have vertical instead of horizontal direction, and in their use as out of plane retaining elements. The main aim of the vertical configuration is to reduce the infill shear transfer to the columns of the surrounding frame. This need is justified by the parametric study described in [20,21], that quantified the significant increment of shear demand on the columns due to the interaction of the frame with the horizontally partitioned infills. Such shear increase occurs because of the concentrated frame-infill contact forces located at the masonry sub-panel corners and may affect the side elements of a possible opening (window or door), as well. In new structures, adequately detailed columns can support such shear action, even if a reduction of the demand will simplify the design. On the other hand, in existing buildings, for which the downgrade of infills could be beneficial [22], a reduced shear demand can prevent the columns shear failure, thus avoiding the need for their strengthening. Moreover, in existing buildings, the possible insertion of the sliding elements in vertical cuts operated in the masonry, makes possible their preservation and the consequent saving in terms of material disposal. The retrofit can be worked from the infill outside, limiting the building downtime. The paper presents the test of two masonry infills built with vertical sliding planks and hollow clay blocks, and compares the results to previous tests on similar solid or horizontally partitioned infills. The first specimen (Specimen A) was tested in- and out-ofplane, proving the solution efficiency in reducing the lateral infillframe interaction and in protecting the infill from the out-of-plane collapse. The feasibility of the solution for existing infill walls was studied on a second prototype (Specimen B). The in- and out-ofplane performance was tested under quasi static cyclic loading, also in this case, together with an operational procedure and the specific detailing for the insertion of the vertical planks in the existing infill.
relative masonry sub-panels sliding along the vertical planks (Fig. 1a). The free rocking of each masonry sub-panel produces their corner uplift, allowed by the top gap, and provides a limited resistance (RR) to the infill deformation (in the order of few kilo-newtons), according to the mechanism schematically described in Fig. 1b-i. Such infill resistance can be quantified as the sum of the lateral overturning loads applied to each sub-panel. Assuming, in first approximation, the lateral load located at the top of the subpanels, RR is obtained with Eq. (1), where Wi, hi, zi and N are the weights, heights, internal lever arms and number of the subpanels, respectively.
RR ¼
N X Wi i¼1
hi
zi
The most significant contribution to the in-plane resistance (RF-Fig. 1b-ii) depends on the friction activated by the relative vertical sliding of the sub-panels on their lateral interfaces (of width ti), which counteracts the sub-panels rotation. Assuming, for the sake of simplicity, a constant friction coefficient ðlÞ and average normal stress ðrn Þ on the sub-panel sides, the theoretical value of RF can be evaluated according to Eq. (2):
RF ¼
N X ðrn lÞ t i bi
The proposed construction technique consists in partitioning the infill wall by means of vertical elements, connected to the frame beams and working as sliding joints and retaining elements for out-of-plane actions. In the prototypes under testing, such vertical elements are made of shaped planks, pinned-end restrained to the frame beams by steel plates. Two vertical elements are also located adjacent to the columns and a gap remains between the infill sub-panels and the frame top beam. For thermal and acoustic performance, the gaps are meant to be sealed with a soft material. The planks stay in the thickness of the infill, so they can be covered by plaster to obtain a homogeneous facing. Additional internal or external insulating layers can be added, provided that they are sufficiently flexible in their plane.
However, such contribution is hardly predictable for the difficulty in quantifying the friction stresses in a such statically undetermined structure and because of: (i) the mortar shrinkage allows gaps between the sub-panels and the vertical planks, which may delay the contact and reduce the friction stresses intensity and modify their distribution; (ii) the sub-panel uplift and rotation induce, by compatibility, a geometrical interference between the masonry sub-panel and the windward vertical plank (Fig. 1c), which increases with the drift and tends to increase the normal ðrn Þ and friction ðrn lÞ stresses on the vertical sliding surfaces. In addition, the effect of the geometrical interference is reduced when the infill dilatation is not confined by the frame columns. Depending on the geometry of the sub-panels and the drift level, such geometrical interference, that would theoretically induce an interpenetration, can be quantified according to Eq. (3).
The vertical configuration of the joints imposes an in-plane deformation mechanism characterized by the alternate rigid rotation (rocking) of the masonry sub-panels around their toes, and the
ð3Þ
The interference grows with the width of the sub-panels (b) and their rotation (a), however it ranges in the order of decimals of millimeters and a small gap can completely change its effect, by delaying or nullifying the friction mechanism. Accordingly, for the aim of this structural application, a certain mortar shrinkage is desirable in order to limit the in-plane stiffness and strength of the infill, as it occurred in the experimental tests presented in the following. To enlarge the scenario of the possible mechanisms of an infill with vertical sliding joints, Fig. 1b-iii describes a third resisting mechanism activated by the sub-panels, in the case their rocking mechanism is confined by the top beam (no top gap). The contribution to the infill resistance of the vertically confined rocking (RV) would be given by Eq. (4), where Fv is the resultant of the confining stress acted by the beams and z0i their lever arm.
Rv ¼ 2.1. In-plane mechanism
ð2Þ
i¼1
Dinterference ¼ b ðtana sena þ cosa 1Þ 2. Construction technique
ð1Þ
N X F V z0i hi n¼1
ð4Þ
The investigation of the role of each mechanism in the infill response requires a detailed modeling, which is out of the scope of this paper, but it can benefit from the test here presented for calibration.
599
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
(a) in-plane mechanism
(f) tributary mass of the central vertical plank
(b) sub-panel in-plane resisting mechanisms: i) free rocking; ii) friction on the lateral inferfaces; iii) rocking with vertical confinement
(d) out-of-plane mechanism (section)
(c) ideal uplifting of the sub-panel and interference with the planks
(e) out-of-plane actions on the vertical planks
Fig. 1. In-plane (a–c) and out-of-plane (d–f) static scheme in the infill with vertical panels.
In the proposed technique, the in-plane stress transfer between the frame and the infill (Fig. 1a), is ensured partly by the connections of the vertical planks to the frame beams (F1), and partly by direct contact between the infill and the columns (F2). For Specimen A, the central joints are connected to the beams while the lateral ones are dowelled on the columns. Therefore, the force transfer proceeds from both the columns and the beams. In Specimen B, on the contrary, due to the inherent difficulty in connecting the lateral planks to the columns after their insertion in the masonry thickness, the lateral planks are connected only to the beams, as for the central ones. At the end of the insertion, a little gap (about 3 mm) was left between them and the columns; therefore, in this second case, the infill reaction is transferred to the frame beams, without any overload on the columns. The lateral gap is meant for two reasons: (i) it allows a certain in-plane dilatation of the infill, so limiting the above recalled friction increase on the vertical sliding interfaces; (ii) it mitigates the interference between the linear deformation of the infill and the deformed profile of the columns under bending in a moment resisting frame. 2.2. Out-of-plane mechanism The out of plane resistance of the infill is ensured by the vertical planks, which bring the seismic load from the masonry sub-panels to the frame, acting as simply supported elements on the frame beams (Fig. 1f). The out-of-plane sub-panels seismic action is transferred to the vertical planks by the masonry bridging between them, by beam or arching mechanism in the sub-panel thickness (Fig. 1e). The central planks tributary mass corresponds to one
sub-panel (Fig. 1d), while it is halved for the lateral elements. It is worth noting that the presence of the gap between the infill and the top frame beam nullifies the possible development of vertical arching mechanism in the masonry. In Specimen A, the lateral vertical planks are dowelled to the columns in 3 points and their static scheme differs from that of the central ones, as shown in Fig. 1f. 3. Specimens geometry and test set-up 3.1. Specimens description Both the tested specimens were built inside a steel frame with hinges at the columns ends (Fig. 2a), the same adopted for the tests on infills with sliding joints reported in [17]. The hinges were modified passing from Specimen A to B (Fig. 2c and d) in order to reduce the frame resistance and better capture the infill net contribution. The steel frame in-plane behavior in the two configurations (Fig. 2b) were tested before the construction of the infills. For Specimen A, the bare frame is characterized by a hysteretic curve with a plastic resistance of about 20 kN, developed by the yielding of the four elasto-plastic hinges, while for Specimen B the low friction hinges practically nullify the frame in-plane stiffness. As previously explained, Specimen A reproduced a new constructed infill wall. The infill was partitioned into four masonry sub-panels by means of three vertical planks (Fig. 3a), properly detailed in order to behave as sliding joints during the in plane deformation and ensure the infill out-of-plane stability. Similar planks were dowelled to the columns. First the wooden frame
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
Lateral Load [kN]
600
25 20 15 10 5 0 -5 -10 -15 -20 -25
S Specimen A Bare Frame
S Specimen B Bare Frame
-4
-3
-2
-1
0
1
2
3
4
Drift [%]
(a) geometry
(b) bare frame experimental force-vs.-drift curve
(c) hinge for Specimen A
(d) hinge for Specimen B (front and lateral view)
Fig. 2. Steel frame adopted in the tests on Specimen A and B (dimensions in mm).
(a) geometry of the specimens (dimensions in cm)
(b) block for Specimen A
(c) block for Specimen B
Fig. 3. Geometry and masonry units of the two tested specimens.
was created and afterward it was filled with the masonry (Fig. 4a). The masonry was made of hollow fired-clay blocks (Fig. 3b) with dimensions 250 200 190 mm (L T H), characterized by a 50% voids ratio and a density of 810 kg/m3, and M5 resistance class lime-cement mortar bed and cross joints (15 mm thick). Table 1 reports the materials’ mechanical properties, which are further described in [18]. In the first part of the test, a 5 cm gap was left between the masonry sub-panels and the upper beam; while in the second test sequence such gap was filled with high strength mortar to simulate a no gap configuration. To protect the masonry from local crushing after the filling of the gap, wooden boards were placed horizontally in contact with the beams, acting as soft cushions.
In order to ensure the out-of-plane stress transfer from the masonry to the planks, the latter were equipped with continuous ‘‘shear keys” (Fig. 4b), which were embedded in the masonry mortar during the infill construction phase. A slat inserted and screwed into a longitudinal groove in the center of the plank created the shear key. The central vertical planks were connected to the upper and lower beams of the frame by means of steel plates (Fig. 4c), doweled to the beams, while the lateral vertical planks were connected to the columns by six dowels. The second test (Specimen B) addressed the feasibility of the proposed construction technique on an existing infill. For this reason, the masonry adopted was different from Specimen A, and consisted of hollow clay blocks (Fig. 3c) typical of the Northern Italy
601
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
Shear key
(a) construction phase of the sub-panels
(b) vertical plank cross-section
(c) detail of the plankbeam connection
(d) shear key in the masonry-plank joint
Fig. 4. Details of the vertical planks system.
Table 1 Average mechanical properties of the infill materials. Specimen A
Masonry prisms -holes parallel to the load -holes perpend. to the load Mortar Cement grout for injection Wood perpend. to the grain Wood parallel to the grain
Specimen B
Compressive strength (MPa)
Elastic modulus (MPa)
Compressive strength (MPa)
Elastic modulus (MPa)
7.28
16,148
4.57
8321.5
2.4 12.24 / 2.56 /
4408 18,619 / 255 /
2.15 7.68 41.82 3.06 34.63
2520 6129 14,752 264 3333
tradition, with dimensions 235 235 115 mm (L T H), 65% voids ratio and a density of 600 kg/m3. Also the detailing of the intervention was calibrated for this specific masonry, characterized by a particularly weak and brittle behavior. The Specimen B had the same geometry of Specimen A, but the planks were inserted in a previously constructed solid infill. The blocks were laid in running bond, with the holes in horizontal direction, on layers of a weak lime-cement mortar (15 mm thick). The masonry mechanical properties have been described in [23] and are reported in Table 1. The insertion of the planks in the wall required the execution of vertical cuts in the infill (Fig. 5a and b) by means of a saw and the successive integration of the vertical elements and their connections to the frame beams (Fig. 5c and d). Such connections were obtained by first dowelling ‘‘U” shaped steel plates to the beam and then inserting and screwing the planks. The details of the connection are reported in Fig. 6a and c. Thereafter, grout injection of the remaining gaps between the masonry sub-panels and the planks restored the continuity of the infill. Fig. 6b shows the vertical planks detailing. Pairs of ‘‘L” shaped steel profiles were screwed on both sides of the planks, in order to optimize the depth of the masonry out–of-plane arching and serve as a mold for the subsequent grout injection. A sock prevented the grout leakage, also in the block holes. The sock elasticity allowed for a certain degree of interlocking of the grout in the block holes (Fig. 6d). At the end, a horizontal cut ensured a gap between the resulting sub-panels and the upper frame beam. The entire procedure was operated from one side of the wall, simulating the intervention from the building outside to limit its downtime. 3.2. Test set-up and loading protocol The test protocol on each specimen included different phases, alternating in-plane and out-of-plane excitations (Fig. 7a). For Specimen A, after Phase 1 and 2, when the in-plane and out-of-
plane tests were performed, respectively, the beams to sub-panel gaps were filled and the test repeated with a simplified protocol (Phase 3 and 4) to explore the stiffness reduction provided by the gaps. For Specimen B, after Phase 1 and 2 the beams to subpanel gaps were filled and Phase 1 loading protocol repeated to evaluate the damage associated to the in the in-plane response without the top gaps. The in-plane tests were carried out with the same loading system adopted for the tests described in [17] (Fig. 7b). The in plane horizontal cyclic load was applied by means of a hydraulic jack hinged to the upper beam of the steel frame and to the reaction frame. During the in-plane tests, gauges monitored the specimen inter-story drift, the load applied by the jack, the relative subpanels vertical sliding and their uplift. After the in-plane test, the specimens were tested out of plane, in order to evaluate their stability in presence of the possible damage suffered during the previous test phase. The out-of-plane loading system was designed in order to approximate the effect of a uniformly distributed transverse load on the infill. One-way cyclic loading of increasing amplitude, under force control, was applied. The set-up (Fig. 8) applied eight equal point loads to the wall, two per each masonry sub-panel mid-axis, at ¼ and ¾ of the infill height, in order to approximate the parabolic profile of bending moment diagram on each vertical plank (Fig. 8c). The action was applied by a hydraulic jack (3 in Fig. 8a) supported on a reaction beam (2). The latter was connected to the steel frame columns (1) to obtain a self-balanced system. The jack transferred the load to the three-dimensional distribution frame (4) shown in Fig. 8b, which applied the load in the eight points of the specimen. The system consisted of three levels of isostatic beams, each of them transferring the load from the center to the symmetric pinned-end supports. Gauges lined up at the infill mid-height and along vertical axis of planks and sub-panels monitored the out-of-plane relative infill-frame displacement.
602
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
3
Sub-panel #4
1
Sub-panel #3
2
Sub-panel #2
3
Sub-panel #1
4
2
(a) phase 1: cut size and sequence
(b) phase 2: infill cut operation
(c) insertion of the vertical joints
(d) downgraded infill before testing
Fig. 5. Procedure for the insertion of the vertical joints in the existing infill (Specimen B).
(a) exploded view of the plank-beam connection
(b) cross-section of the vertical element
(c) U shaped plate dowelled to the upper beam
(d) interlocking between the hollow clay block and the injected grout Fig. 6. Details of Specimen B.
603
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
Phase 2
Phase 1
Phase 4
Phase 3
(a)
(b)
Fig. 7. Loading protocol (a) and set-up (b) for the in plane test. The out of plane test was performed in Phases 2 and 4 after the in plane loading of Phases 1 and 3, respectively.
(a) picture of the out of plane loading system
(b) load distribution system (4, in figure a)
(c) experimental vs. ideal moment profile on the central vertical planks
Fig. 8. Out of plane test set-up.
4. Test results 4.1. Specimen A Fig. 9a and b shows the force-vs.-drift in-plane response of Specimen A in the test Phases 1 and 3, respectively. In Phase 1, the specimen highlighted a ductile cyclic behavior with negligible damage, characterized by a developed resistance equal to 68 kN at 2.5% drift, corresponding to about 48 kN infill net contribution. The alternate rocking of each sub-panel around its own toes and the sliding along the vertical planks governed the infill deformation. Recalling the resisting mechanism presented in Section 2.1, the free rocking of the sub-panel provides a limited in-plane reaction (2 kN according to Eq. (1), only about 4% of the infill net contribution at 2.5% drift), while the friction contribution governs the response. Subsequent cycles at the same drift amplitude showed a stiffness degradation, probably due to the local abrasion of the sliding surfaces and gaps adaptations that reduce the effect of the friction mechanism on the response. Sub-panels toes suffered only limited damage at the end of the test Phase 1. The results of the out-of-plane test (Phase 2) highlighted a high resistance of the infill. Thanks to the retaining planks, the infill resisted a total load equal to 72 kN, corresponding to an equivalent uniform acceleration of 4.2 g (calculated as the lateral force divided by the mass of the infill quantified in 1.7t), before the test was
stopped to continue with Phase 3. Fig. 10a shows the out of plane deformation, measured at mid-height of the central vertical planks, for the applied load cycles. The specimen exhibited a bilinear envelope curve with a stiffness reduction at the first application of a load equal to 30 kN, possibly related to a settlement in the planks-to-beam connection, testified by the residual deformation and a creaking perceivable at this load step. No cracks were detected in the masonry. Fig. 10b reports the shape of the wall out of plane deformation, for increasing values of the applied action. The profile results by interpolating the measures of the gauges lined up at mid-height, whose horizontal position along the wall length is indicated on the x-axis. For each loading step, the maximum out-of-plane deformation occurred in the two central masonry sub-panels, indicating the triggering of a relative horizontal sliding with respect to the central vertical plank, whose amplitude was limited below 1 mm by the shear keys connecting the masonry to the vertical planks. After the filling of the top gap with high strength mortar, the inplane response of Specimen A was significantly stiffened (Phase 3). As shown in Fig. 9b, a hardening behavior in both the directions occurred, with a remarkable increase of the resistance with respect to Phase 1 for drift higher than 1%. At 3% drift no damage occurred in the masonry. For each sub-panel, a beam contact length of about one third of the width was measured by means of additional local gauges. Such contact occurred alternatively along the sub-panel
604
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
80
Lateral Force [kN]
60 40 20
Bare Frame Max Resistance
0 Bare Frame Max Resistance
-20 -40 -60 -80
-3.5 -3 -2.5 -2 -1.5 -1 -0.5 0
0.5
1
1.5
2
2.5
3
3.5
Drift [%]
(a) Phase 1 (adapted from [23]) 160
Lateral Force [kN]
120 80
Phase 1 peak
40 0 -40
Phase 1 peak
-80 -120 -160
-3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Drift [%]
(b) Phase 3 Fig. 9. In-plane test results on the original Specimen A (a) and after filling the top gaps (b).
diagonals and a local permanent compressive deformation was observed in the horizontal planks adjoining the beams, after demolition. Fig. 10c highlights a stable out of plane behavior up to the maximum imposed action of 102 kN, exhibiting higher stiffness than that obtained in Phase 2 for deflection higher than six millimeters. Horizontal cracks in the sub-panels occurred during the last run (Fig. 11), at 93 kN applied, and no significant stiffness degradation was observed up to the maximum load of 102 kN, after which the test was stopped. The filling of the gap ensured therefore a higher out-of-plane capacity to the infill, most probably thanks to the activation of the vertical arching mechanism spanning in between the frame beams. 4.2. Specimen B Fig. 12 shows the test results for Specimen B. The in-plane response (Fig. 12a) was symmetric and ductile, with very low resistance and no damage. Reloading cycles showed a certain stiffness and strength degradation. Also in this case, the sliding mechanism governs the infill in-plane resistance, since the net calculated contribution of the rocking mechanism is lower than 1 kN. However, in this case the infill reaction was significantly smaller. A possible explanation, which requires further investigation, is the presence of the little gap between each column and the adjacent vertical planks. Such gap was monitored during the test and its partial closure was measured, testifying a horizontal dilatation of the infill during the in-plane response. Closure or opening of possible gaps
on the two sides of the central vertical planks were monitored, as well. In this case, values of gap closure and opening up to 0.6 mm were measured, suggesting the presence of an initial gap left by the grout shrinkage and a tendency to dilatation of the infill. Regarding the masonry sub-panel toes, note that the disruption visible in Fig. 12b is the result of the necessary local demolition for the insertion of the vertical planks in the construction phase. The results of the out-of-plane test (Phase 2) highlighted a significant resistance of Specimen B, capable of bearing a maximum applied load equal to 40 kN, equivalent to a 4 g transversal acceleration (infill mass equal to 1t). Fig. 12c shows the load vs. out-ofplane maximum displacement in the three central planks. The stiffness progressively reduced by increasing the load. The reduction of the stiffness was directly connected to the progressive triggering of cracks in the masonry sub-panels (Fig. 12b). Fig. 12d shows the profile of out-of-plane deformation of the infill at mid-height, for the different load steps and the corresponding equivalent acceleration values. Fig. 13 shows the profile of deformation along the central vertical plank and the adjacent sub-panel. The graphs show a practically cylindrical infill deformation, governed by the vertical plank flexibility. The infill did not develop a continuous resisting arch in the masonry thickness along the horizontal direction between the frame columns, as testified by the central plank deforming less than the others. Therefore, the sole vertical planks supported the total out-of-plane action. Fig. 13 testifies also an out-of-plane sliding of the masonry sub-panels that
605
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
x
5
80
Cyclic response
4
60 3
50 40
2
30 20
1
10 0
0 0
2
4
6
8
10
Out of plane displacement [mm]
Out of plane Load [kN]
70
Equivalent acceleration [g]
Envelope
0 1 2 3 4 5 6 7 8 9 10
12
0.3g 0.7g 1.4g 2.2g (1) 2.2g 2.9g 3.6g 4.2g
0
100
Out of plane displacement [mm]
200
300
Position X [cm]
(a) Phase 2: out of plane load vs. maximum displacement (M4 gauge)
(b) Phase 2: out-of-plane displacement profile increasing the equivalent acceleration
Phase 2 envelope
5 4 3 2 1 0
0
2
4
6
0
6
Phase 4 (filled gap)
8
10
12
Out of plane max displacement [mm]
(c) Phase 4: out of plane load vs. maximum displacement (M4 gauge)
Out of plane displacement [mm]
110 100 90 80 70 60 50 40 30 20 10 0
Equivalent Acceleration [g]
Out of Plane Load [kN]
x
0.7g
2
1.4g
4
2.1g
6
2.8g 3.5g
8
4.2g
10
4.9g
12
6.1g
14 0
100
200
300
Position X [cm]
(d) Phase 4: out-of-plane displacement profile increasing the equivalent acceleration
Fig. 10. Out-of-plane test results on the original Specimen A (a and b) and after filling the top gaps (c and d).
Fig. 11. Crack pattern at the end of the out of plane test on the specimen A at the end of Phase 4.
deformed more than the retaining planks and a local deformation of the plank support on the top beam; however, the response was stable without strength degradation.
Considering the limited damage of the specimen B after the outof-plane test, the top gaps were filled and the same in-plane loading history of test Phase 1 was repeated twice: first the filling was made with well compacted rock wool (meant for thermal performance) to evaluate its effect on the structural response; then high strength mortar was used in order to explore the efficiency of the intervention in the event of no gap is created between the masonry sub-panels and the frame top beam. Fig. 14a compares the cyclic in-plane responses of the Specimen B with the two types of gap filler. The rock wool filler left practically unaltered the response, which resulted weaker than the first in-plane test (Phase 1) following the same trend of progressive strength reduction along the reloading cycles. The mortar filler produced a stiffer and stronger response. In this last case, damage of the sub-panels corners occurred starting with local cracking at 0.66% drift and evolved into crushing after 2% drift, after which a strength degradation occurred. Fig. 14b shows the damage pattern at the end of the test. Despite some significant damage limited to the sub-panels corners, the response was stable up to very large drift values, unlike typical existing infills performance. The infill resistance remained low (about 30 kN), consistently with the low masonry compressive strength, thus promising a limited infill-frame interaction.
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
Lateral Force [kN]
606
10.0 8.0 6.0 4.0 2.0 0.0 -2.0 -4.0 -6.0 -8.0 -10.0
31kN
32kN 18kN
-3
-2
-1
0
1
2
20kN
27kN
20kN
24kN
3
Drift [%]
(a) in-plane test (Phase 1)
(b) crack pattern in the infill sub-panels at the end of the out-of plane test (Phase 2) x
Out-of-plane load [KN]
35
4
30
3
25 20
2
15
Design acceleration
10
1
5 0
0 0
0
Out of plane displacement [mm]
H3 H5 H7
40
Equivalent Acceleration [g]
45
0.4g
2
1.2g
4
1.6g 2g
6
2.4g
8 3g
10 12
4g
14
2 4 6 8 10 12 Out of plane displacement [mm]
0
(c) out of plane test (Phase 2)
100 200 Position X [cm ]
300
(d) infill displacement profile along the horizontal direction, during the out-of-plane test
Fig. 12. Experimental results for the existing infill with vertical sliding joints (Specimen B).
0.4g
250
200
200 2g 1.6g
150
2.4g
Position Y [cm]
250
4g
3g
100 50
0.8g 1.2g
4g 2g
150
2.4g
3g
8
10
1.6g
100 50
0
Y
Y
Position Y [cm]
0.4g 0.8g 1.2g
0
0
2
4
6
8
10
Out of plane displacement [mm]
(a) central vertical plank
0
2
4
6
12
14
Out of plane displacement [mm]
(b) infill sub-panel #2
Fig. 13. Out of plane deflection along the vertical direction, increasing the equivalent acceleration.
5. Comparison of vertical versus horizontal sub-panel performance The performance of the prototype infill built with vertical sliding joints is compared in Fig. 15 with those of other infills [24] with same material properties and boundary conditions of Specimen A, but different construction technique: (i) a solid infill, built in
running bond and in adhesion to the frame bay, according to tradition; (ii) an infill with horizontal sliding joints. The net infill responses are compared, as envelopes of the in-plane hysteretic curves subtracted of the bare frame contribution. Fig. 15 shows the dramatic reduction of the infill strength obtained thanks to the insertion of sliding joints. The vertical panel configuration obtained the most efficient reduction. Its response however is
607
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
Lateral Force [kN]
30 20 10 0 -10 -20
Mortar Filling
-30
Rock Wool Filling
-40 -3
-2
-1
0
1
2
3
Drift [%]
(a) force-vs.-drift experimental curve
(b) damage pattern at the end of the test
450 350 250 150 50 -50 -150 -250 -350 -450 -3.5
150
Solid Infill Horizontal Joints Specimen A Specimen A filled gap
-2.5
-1.5
-0.5
0.5
1.5
2.5
3.5
Infill net reaction [kN]
Infill net reaction[kN]
Fig. 14. In plane test on Specimen B after the top gapfilling.
Horizontal Joints Specimen A Specimen A filled gap
100
50
0 0
1
2
3
Drift [%]
Drift [%]
Fig. 15. Comparison of the experimental in-plane response for different infill configurations: envelope of the net reaction of the specimen under cyclic loading.
significantly affected by the gaps allowed in the infill. For Specimen A, after filling of the top gaps (Phase 3) a force-drift response similar to that of horizontal joints was obtained, except for a very low initial stiffness, which was affected by the loading history on the same specimen before the gap filling. As a consequence, the filling of such gap for thermal, acoustic and permeability issues deserves particular care, and a soft material may be preferred, as demonstrated in the test performed on Specimen B filling the gap with rock wool.
6. Conclusions The paper describes an innovative construction technique for masonry infill, based on the wall partitioning in vertical subpanels. The experimental results on two prototype infills are described in detail. A dramatic downgrade of the in-plane response of the infill inside the frame was obtained for the Specimen A, if compared to the stiff reaction typical of traditional continuous infill. The out-of-plane response was stable and reliable, thanks to the connection of the vertical planks, partitioning the infill, to the frame beams, and to the negligible damage suffered by the specimen under in-plane loading. The feasibility of the application of the downgrade technique to an existing hollow block masonry infill was tested with the second specimen (Specimen B). Specific procedure and detailing for the insertion of the vertical planks in the masonry and their connection to both the existing frame and masonry infill are described. The entire intervention is operated from one side of the specimen, to simulate the retrofit of an enclosure infill wall from the outside. Specimen B showed an in-plane net infill contribution of few kilo-newtons, practically nullifying the interaction with the frame. Such response, significantly weaker
than Specimen A (16% resistance ratio), cannot be explained only with the different type of masonry adopted for the two specimens (about 60% of both thickness and mass ratio); most probably it depends on the small gaps left in the insertion process between the lateral vertical planks and columns, and on the shrinkage of the grout injection, used to restore the walls continuity, that reduced the friction mechanism. A secondary role in the infill resistance is ascribable to the masonry mechanical properties, as no damage was activated in the sub-panels. A detailed model of the structure is needed to explore the role of the single infill design parameters (particularly the bay and sub-panels geometry, the sliding interfaces friction coefficient and contact conditions) and to extend the results of these first prototypes. The intervention on the existing infill showed to be particularly efficient for its out-of-plane resistance; in fact, the specimen sustained a transverse load equivalent to a 4 g acceleration after the application of several in-plane load cycles, up to 2.5% drift. The infill downgrade is beneficial in terms of post-earthquake damage reduction, but it also simplifies the prediction of the building seismic response with respect to the well-known uncertainties when dealing with traditional infill walls. From the designer point of view, because of the reduced infill frame interaction, the modeling can reliably assume, in a first step, the structure as a simplified bare frame, so neglecting the stiffening effect of the infills. Then few adjustments could take into consideration the infill contribution, dramatically reduced by the proposed technique.
Acknowledgements The authors gratefully acknowledge Malek Neffati, Carlo Piacentini, Michele Moretti, Alessia Ghisla and the technicians of
608
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
the P. Pisa Lab of the University of Brescia for their assistance in the experimental testing. The presented study was partly developed in the research program financed by the ‘‘Presidenza del Consiglio dei Ministri Dipartimento della Protezione Civile” within the Reluis research program; the present publication, however, does not necessarily reproduce the Department position and judgments. A special thanks to Prof. Ezio Giuriani for all the suggestions and fruitful discussions. References [1] Shing PB, Mehrabi AB. Behaviour and analysis of masonry-infilled frames. Prog Struct Mat Eng 2002;4:320–31. [2] EERI. The MW 6.3 Abruzzo, Italy, Earthquake of April 6, 2009; 2009. [3] FEMA. Reducing the risks of nonstructural earthquake damage - a practical guide; 2012. [4] Sucuog˘lu H. Implications of masonry infill and partition damage in performance perception in residential buildings after a moderate earthquake. Earthq Spectra 2013;29:661–7. http://dx.doi.org/10.1193/1.4000147. [5] EERI. 1994 Northridge earthquake reconnaissance report. Earthq Spectra 1996:12. [6] EERI. 1999 Kocaeli, Turkey earthquake reconnaissance report. Earthq Spectra 2000:16. [7] Mohammadi M, Akrami V, Mohammadi-Ghazi R. Methods to improve infilled frame ductility. J Struct Eng 2011;137:646–53. http://dx.doi.org/10.1061/ (ASCE)ST.1943-541X.0000322. [8] Preti M, Bettini N, Plizzari G. Infill walls with sliding joints to limit infill-frame seismic interaction: large-scale experimental test. J Earthq Eng 2012;16:125–41. http://dx.doi.org/10.1080/13632469.2011.579815. [9] Misir IS, Ozcelik O, Girgin SC, Kahraman S. Experimental work on seismic behavior of various types of masonry infilled RC frames. Struct Eng Mech 2012;44:763–74. http://dx.doi.org/10.12989/sem.2012.44.6.763. [10] Markulak D, Radic´ I, Sigmund V. Cyclic testing of single bay steel frames with various types of masonry infill. Eng Struct 2013;51:267–77. http://dx.doi.org/ 10.1016/j.engstruct.2013.01.026. [11] Vailati M, Monti G. Earthquake-resistant and thermo-insulating infill panel with recycled-plastic joints. In: D’Amico S, editor. Earthquakes and their impact on society. Springer International Publishing; 2016. p. 417–32. [12] Da Porto F, Verlato N, Guidi G, Modena C. The INSYSME project: innovative construction systems for earthquake resistant masonry infill walls. In:
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
Proceedings of the 16th international brick and block masonry conference (IBMAC 2016), Padova, Italy. Tasligedik AS, Pampanin S. Rocking cantilever clay brick infill wall panels: a novel low damage infill wall system. J Earthq Eng 2016:1–27. http://dx.doi. org/10.1080/13632469.2016.1190797. Vintzileou E, Adami CE, Palieraki V. In-plane and out-of-plane response of a masonry infill divided into smaller wallets. In: Proceedings of the 16th international brick and block masonry conference (IBMAC 2016), Padova, Italy. Wang Z, Totoev Y, Lin K. Non-linear static analysis of multi-storey steel frame with semi-interlocking masonry infill panels. In: Proceedings of the 16th international brick and block masonry conference (IBMAC 2016), Padova, Italy. Silva L, Vasconcelos G, Lourenço P, Akhoundi F. Experimental evaluation of a constructive system for earthquake resisting masonry enclosure walls. In: Proceedings of the 16th International Brick and Block Masonry Conference (IBMAC 2016). Padova, Italy; 2016. Preti M, Migliorati L, Giuriani E. Experimental testing of engineered masonry infill walls for post-earthquake structural damage control. Bull Earthq Eng 2014;13:2029–49. http://dx.doi.org/10.1007/s10518-014-9701-2. Langenbach R. Learning from the past to protect the future: armature crosswalls. Eng Struct 2008;30:2096–100. http://dx.doi.org/10.1016/j. engstruct.2008.04.028. Morandi P, Milanesi RR, Magenes G. Innovative seismic solution for clay masonry infills with sliding joints: principles and details. In: Proceedings of the 16th international brick and block masonry conference (IBMAC 2016), Padova, Italy. Preti M, Bettini N, Migliorati L, Bolis V, Stavridis A, Plizzari GA. Analysis of the in-plane response of earthen masonry infill panels partitioned by sliding joints. Earthq Eng Struct Dyn 2016;45:1209–32. http://dx.doi.org/10.1002/ eqe.2703. Bolis V, Stavridis A, Preti M. Numerical investigation of the in-plane performance of masonry-infilled RC frames with sliding subpanels. J Struct Eng 2016;0. http://dx.doi.org/10.1061/(ASCE)ST.1943-541X.0001651. Marini A, Belleri A, Feroldi F, Passoni C, Preti M, Riva P, et al. Coupling energy refurbishment with structural strengthening in retrofit interventions. SAFESUST workshop. A roadmap for the improvement of earthquake resistance and eco-efficiency of existing buildings and cities, Ispra; 2015. Facconi L, Minelli F, Giuriani E. In-plane cyclic tests on hollow clay brick masonry infills retrofitted by glass fiber mesh reinforced mortar coating. In: Proceedings of the 16th international brick and block masonry conference (IBMAC 2016), Padova, Italy. Preti M, Bolis V, Stavridis A. Design of masonry infill walls with sliding joints for earthquake structural damage control. In: Proceedings of the 16th international brick and block masonry conference (IBMAC 2016), Padova, Italy.
Contents lists available at ScienceDirect
Engineering Structures journal homepage: www.elsevier.com/locate/engstruct
Masonry infill construction and retrofit technique for the infill-frame interaction mitigation: Test results Marco Preti ⇑, Valentino Bolis Department of Civil, Environmental, Architectural Engineering and Mathematics, Università degli studi di Brescia, Via Branze 43, 25123 Brescia, Italy
a r t i c l e
i n f o
Article history: Received 23 March 2016 Revised 3 November 2016 Accepted 21 November 2016
Keywords: Seismic downgrade Damage control Infill-frame interaction Vertical sliding joints Out-of-plane strength Sliding panels Masonry partitioning Masonry infill Existing infill Infill testing
a b s t r a c t The paper describes the set-up and testing of an innovative construction technique for masonry infill, which can provide a flexible and predictable in-plane response to the infill inside the frame, together with a stable and reliable out-of-plane response. The design strategy is to downgrade the infill reaction inside the structural frame thanks to a dramatic reduction of the masonry in-plane stiffness. The infill is partitioned by vertical planks (or equivalent beams) into sub-panels, free to relatively slide and rock on their toes. The planks connected to the beams provide the necessary out-of-plane stability. The solution was tested for application in both new and existing infills and construction details are discussed. A comparison is also presented with the performance of two infills, one continuous and one with horizontal subpanels, previously tested under the same conditions. The observed infill downgrade makes practically negligible the infill-frame interaction and the post-earthquake masonry damage. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction The research work here presented tries to address the wellknown issues related to the infill-frame interaction in multi-story buildings. The paper describes the set-up and testing of an innovative infill construction technique, which can provide a deformable and predictable in-plane response to the infill inside the frame, together with a stable and reliable out-of-plane response. The design strategy is to downgrade the infill contribution to the building seismic response thanks to a dramatic reduction of its in-plane stiffness. In fact, the uncertainty of the traditional masonry infills in- and out-of-plane response [1] and their possible irregular distribution in the structure jeopardize the building safety and resilience. The post-earthquake damage associated to a poor infill performance highly contributes to the cost and duration of the reconstruction process and activity recover [2,3], even after moderate intensity earthquakes [4]. Despite the possible detrimental effects of the infill-frame interaction, the post-earthquake damage survey showed also in some cases their contribution in preventing the collapse of poorly detailed buildings, not designed to withstand ⇑ Corresponding author. E-mail addresses: [email protected] (M. Preti), [email protected] (V. Bolis). http://dx.doi.org/10.1016/j.engstruct.2016.11.053 0141-0296/Ó 2016 Elsevier Ltd. All rights reserved.
seismic actions; but this contribution is not always reliable due to the possible activation of undesired collapse mechanism in the structure. However, because of their efficiency in terms of construction ease, internal climate control and low building costs, traditionally constructed infills remain widely used, even if in the last decades, a large number of infilled frame buildings have performed poorly during earthquakes [5,6]. The increasing demand for post-earthquake damage control justifies the development of infill typologies for new buildings, capable to survive moderate to intense earthquakes without damage. In the last decade, several authors have proposed engineered masonry infill solutions to address this issue [7–11]). Mohammadi et al. [7] and Preti et al. [8] proposed the horizontal partitioning of the infill. Misir et al. [9] investigated the response of infills made of blocks without mortar, providing out of plane stability by a particular interlocking between blocks of adjacent rows (locked brick infill). Markulak et al. [10] proposed the use of weaker masonry blocks located close to the columns to accommodate the frame deformations. Vailati and Monti [11] substituted the mortar joints with plastic ones to be used with hollow blocks. Other ongoing research projects, aimed at optimizing the design of earthquake resilient infills, were presented in [12–16]. The construction technique here presented stems from the research work presented by Preti et al. [8,17], inspired by historical
598
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
partitioned masonry structures [18] and recently updated towards industrialization by Morandi et al. [19]. The technique adopted the partitioning of the infill in masonry sub-panels connected to the columns for ensuring out of plane stability, but free to move relatively along horizontal planks, embedded in mortar beds. The performed tests showed a ductile infill response provided by the horizontal partitioning, without damage and with a significant stiffness reduction with respect to a traditional masonry infill. The innovation of the here presented solution consists in the configuration of the sliding planks (or equivalent beams), which have vertical instead of horizontal direction, and in their use as out of plane retaining elements. The main aim of the vertical configuration is to reduce the infill shear transfer to the columns of the surrounding frame. This need is justified by the parametric study described in [20,21], that quantified the significant increment of shear demand on the columns due to the interaction of the frame with the horizontally partitioned infills. Such shear increase occurs because of the concentrated frame-infill contact forces located at the masonry sub-panel corners and may affect the side elements of a possible opening (window or door), as well. In new structures, adequately detailed columns can support such shear action, even if a reduction of the demand will simplify the design. On the other hand, in existing buildings, for which the downgrade of infills could be beneficial [22], a reduced shear demand can prevent the columns shear failure, thus avoiding the need for their strengthening. Moreover, in existing buildings, the possible insertion of the sliding elements in vertical cuts operated in the masonry, makes possible their preservation and the consequent saving in terms of material disposal. The retrofit can be worked from the infill outside, limiting the building downtime. The paper presents the test of two masonry infills built with vertical sliding planks and hollow clay blocks, and compares the results to previous tests on similar solid or horizontally partitioned infills. The first specimen (Specimen A) was tested in- and out-ofplane, proving the solution efficiency in reducing the lateral infillframe interaction and in protecting the infill from the out-of-plane collapse. The feasibility of the solution for existing infill walls was studied on a second prototype (Specimen B). The in- and out-ofplane performance was tested under quasi static cyclic loading, also in this case, together with an operational procedure and the specific detailing for the insertion of the vertical planks in the existing infill.
relative masonry sub-panels sliding along the vertical planks (Fig. 1a). The free rocking of each masonry sub-panel produces their corner uplift, allowed by the top gap, and provides a limited resistance (RR) to the infill deformation (in the order of few kilo-newtons), according to the mechanism schematically described in Fig. 1b-i. Such infill resistance can be quantified as the sum of the lateral overturning loads applied to each sub-panel. Assuming, in first approximation, the lateral load located at the top of the subpanels, RR is obtained with Eq. (1), where Wi, hi, zi and N are the weights, heights, internal lever arms and number of the subpanels, respectively.
RR ¼
N X Wi i¼1
hi
zi
The most significant contribution to the in-plane resistance (RF-Fig. 1b-ii) depends on the friction activated by the relative vertical sliding of the sub-panels on their lateral interfaces (of width ti), which counteracts the sub-panels rotation. Assuming, for the sake of simplicity, a constant friction coefficient ðlÞ and average normal stress ðrn Þ on the sub-panel sides, the theoretical value of RF can be evaluated according to Eq. (2):
RF ¼
N X ðrn lÞ t i bi
The proposed construction technique consists in partitioning the infill wall by means of vertical elements, connected to the frame beams and working as sliding joints and retaining elements for out-of-plane actions. In the prototypes under testing, such vertical elements are made of shaped planks, pinned-end restrained to the frame beams by steel plates. Two vertical elements are also located adjacent to the columns and a gap remains between the infill sub-panels and the frame top beam. For thermal and acoustic performance, the gaps are meant to be sealed with a soft material. The planks stay in the thickness of the infill, so they can be covered by plaster to obtain a homogeneous facing. Additional internal or external insulating layers can be added, provided that they are sufficiently flexible in their plane.
However, such contribution is hardly predictable for the difficulty in quantifying the friction stresses in a such statically undetermined structure and because of: (i) the mortar shrinkage allows gaps between the sub-panels and the vertical planks, which may delay the contact and reduce the friction stresses intensity and modify their distribution; (ii) the sub-panel uplift and rotation induce, by compatibility, a geometrical interference between the masonry sub-panel and the windward vertical plank (Fig. 1c), which increases with the drift and tends to increase the normal ðrn Þ and friction ðrn lÞ stresses on the vertical sliding surfaces. In addition, the effect of the geometrical interference is reduced when the infill dilatation is not confined by the frame columns. Depending on the geometry of the sub-panels and the drift level, such geometrical interference, that would theoretically induce an interpenetration, can be quantified according to Eq. (3).
The vertical configuration of the joints imposes an in-plane deformation mechanism characterized by the alternate rigid rotation (rocking) of the masonry sub-panels around their toes, and the
ð3Þ
The interference grows with the width of the sub-panels (b) and their rotation (a), however it ranges in the order of decimals of millimeters and a small gap can completely change its effect, by delaying or nullifying the friction mechanism. Accordingly, for the aim of this structural application, a certain mortar shrinkage is desirable in order to limit the in-plane stiffness and strength of the infill, as it occurred in the experimental tests presented in the following. To enlarge the scenario of the possible mechanisms of an infill with vertical sliding joints, Fig. 1b-iii describes a third resisting mechanism activated by the sub-panels, in the case their rocking mechanism is confined by the top beam (no top gap). The contribution to the infill resistance of the vertically confined rocking (RV) would be given by Eq. (4), where Fv is the resultant of the confining stress acted by the beams and z0i their lever arm.
Rv ¼ 2.1. In-plane mechanism
ð2Þ
i¼1
Dinterference ¼ b ðtana sena þ cosa 1Þ 2. Construction technique
ð1Þ
N X F V z0i hi n¼1
ð4Þ
The investigation of the role of each mechanism in the infill response requires a detailed modeling, which is out of the scope of this paper, but it can benefit from the test here presented for calibration.
599
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
(a) in-plane mechanism
(f) tributary mass of the central vertical plank
(b) sub-panel in-plane resisting mechanisms: i) free rocking; ii) friction on the lateral inferfaces; iii) rocking with vertical confinement
(d) out-of-plane mechanism (section)
(c) ideal uplifting of the sub-panel and interference with the planks
(e) out-of-plane actions on the vertical planks
Fig. 1. In-plane (a–c) and out-of-plane (d–f) static scheme in the infill with vertical panels.
In the proposed technique, the in-plane stress transfer between the frame and the infill (Fig. 1a), is ensured partly by the connections of the vertical planks to the frame beams (F1), and partly by direct contact between the infill and the columns (F2). For Specimen A, the central joints are connected to the beams while the lateral ones are dowelled on the columns. Therefore, the force transfer proceeds from both the columns and the beams. In Specimen B, on the contrary, due to the inherent difficulty in connecting the lateral planks to the columns after their insertion in the masonry thickness, the lateral planks are connected only to the beams, as for the central ones. At the end of the insertion, a little gap (about 3 mm) was left between them and the columns; therefore, in this second case, the infill reaction is transferred to the frame beams, without any overload on the columns. The lateral gap is meant for two reasons: (i) it allows a certain in-plane dilatation of the infill, so limiting the above recalled friction increase on the vertical sliding interfaces; (ii) it mitigates the interference between the linear deformation of the infill and the deformed profile of the columns under bending in a moment resisting frame. 2.2. Out-of-plane mechanism The out of plane resistance of the infill is ensured by the vertical planks, which bring the seismic load from the masonry sub-panels to the frame, acting as simply supported elements on the frame beams (Fig. 1f). The out-of-plane sub-panels seismic action is transferred to the vertical planks by the masonry bridging between them, by beam or arching mechanism in the sub-panel thickness (Fig. 1e). The central planks tributary mass corresponds to one
sub-panel (Fig. 1d), while it is halved for the lateral elements. It is worth noting that the presence of the gap between the infill and the top frame beam nullifies the possible development of vertical arching mechanism in the masonry. In Specimen A, the lateral vertical planks are dowelled to the columns in 3 points and their static scheme differs from that of the central ones, as shown in Fig. 1f. 3. Specimens geometry and test set-up 3.1. Specimens description Both the tested specimens were built inside a steel frame with hinges at the columns ends (Fig. 2a), the same adopted for the tests on infills with sliding joints reported in [17]. The hinges were modified passing from Specimen A to B (Fig. 2c and d) in order to reduce the frame resistance and better capture the infill net contribution. The steel frame in-plane behavior in the two configurations (Fig. 2b) were tested before the construction of the infills. For Specimen A, the bare frame is characterized by a hysteretic curve with a plastic resistance of about 20 kN, developed by the yielding of the four elasto-plastic hinges, while for Specimen B the low friction hinges practically nullify the frame in-plane stiffness. As previously explained, Specimen A reproduced a new constructed infill wall. The infill was partitioned into four masonry sub-panels by means of three vertical planks (Fig. 3a), properly detailed in order to behave as sliding joints during the in plane deformation and ensure the infill out-of-plane stability. Similar planks were dowelled to the columns. First the wooden frame
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
Lateral Load [kN]
600
25 20 15 10 5 0 -5 -10 -15 -20 -25
S Specimen A Bare Frame
S Specimen B Bare Frame
-4
-3
-2
-1
0
1
2
3
4
Drift [%]
(a) geometry
(b) bare frame experimental force-vs.-drift curve
(c) hinge for Specimen A
(d) hinge for Specimen B (front and lateral view)
Fig. 2. Steel frame adopted in the tests on Specimen A and B (dimensions in mm).
(a) geometry of the specimens (dimensions in cm)
(b) block for Specimen A
(c) block for Specimen B
Fig. 3. Geometry and masonry units of the two tested specimens.
was created and afterward it was filled with the masonry (Fig. 4a). The masonry was made of hollow fired-clay blocks (Fig. 3b) with dimensions 250 200 190 mm (L T H), characterized by a 50% voids ratio and a density of 810 kg/m3, and M5 resistance class lime-cement mortar bed and cross joints (15 mm thick). Table 1 reports the materials’ mechanical properties, which are further described in [18]. In the first part of the test, a 5 cm gap was left between the masonry sub-panels and the upper beam; while in the second test sequence such gap was filled with high strength mortar to simulate a no gap configuration. To protect the masonry from local crushing after the filling of the gap, wooden boards were placed horizontally in contact with the beams, acting as soft cushions.
In order to ensure the out-of-plane stress transfer from the masonry to the planks, the latter were equipped with continuous ‘‘shear keys” (Fig. 4b), which were embedded in the masonry mortar during the infill construction phase. A slat inserted and screwed into a longitudinal groove in the center of the plank created the shear key. The central vertical planks were connected to the upper and lower beams of the frame by means of steel plates (Fig. 4c), doweled to the beams, while the lateral vertical planks were connected to the columns by six dowels. The second test (Specimen B) addressed the feasibility of the proposed construction technique on an existing infill. For this reason, the masonry adopted was different from Specimen A, and consisted of hollow clay blocks (Fig. 3c) typical of the Northern Italy
601
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
Shear key
(a) construction phase of the sub-panels
(b) vertical plank cross-section
(c) detail of the plankbeam connection
(d) shear key in the masonry-plank joint
Fig. 4. Details of the vertical planks system.
Table 1 Average mechanical properties of the infill materials. Specimen A
Masonry prisms -holes parallel to the load -holes perpend. to the load Mortar Cement grout for injection Wood perpend. to the grain Wood parallel to the grain
Specimen B
Compressive strength (MPa)
Elastic modulus (MPa)
Compressive strength (MPa)
Elastic modulus (MPa)
7.28
16,148
4.57
8321.5
2.4 12.24 / 2.56 /
4408 18,619 / 255 /
2.15 7.68 41.82 3.06 34.63
2520 6129 14,752 264 3333
tradition, with dimensions 235 235 115 mm (L T H), 65% voids ratio and a density of 600 kg/m3. Also the detailing of the intervention was calibrated for this specific masonry, characterized by a particularly weak and brittle behavior. The Specimen B had the same geometry of Specimen A, but the planks were inserted in a previously constructed solid infill. The blocks were laid in running bond, with the holes in horizontal direction, on layers of a weak lime-cement mortar (15 mm thick). The masonry mechanical properties have been described in [23] and are reported in Table 1. The insertion of the planks in the wall required the execution of vertical cuts in the infill (Fig. 5a and b) by means of a saw and the successive integration of the vertical elements and their connections to the frame beams (Fig. 5c and d). Such connections were obtained by first dowelling ‘‘U” shaped steel plates to the beam and then inserting and screwing the planks. The details of the connection are reported in Fig. 6a and c. Thereafter, grout injection of the remaining gaps between the masonry sub-panels and the planks restored the continuity of the infill. Fig. 6b shows the vertical planks detailing. Pairs of ‘‘L” shaped steel profiles were screwed on both sides of the planks, in order to optimize the depth of the masonry out–of-plane arching and serve as a mold for the subsequent grout injection. A sock prevented the grout leakage, also in the block holes. The sock elasticity allowed for a certain degree of interlocking of the grout in the block holes (Fig. 6d). At the end, a horizontal cut ensured a gap between the resulting sub-panels and the upper frame beam. The entire procedure was operated from one side of the wall, simulating the intervention from the building outside to limit its downtime. 3.2. Test set-up and loading protocol The test protocol on each specimen included different phases, alternating in-plane and out-of-plane excitations (Fig. 7a). For Specimen A, after Phase 1 and 2, when the in-plane and out-of-
plane tests were performed, respectively, the beams to sub-panel gaps were filled and the test repeated with a simplified protocol (Phase 3 and 4) to explore the stiffness reduction provided by the gaps. For Specimen B, after Phase 1 and 2 the beams to subpanel gaps were filled and Phase 1 loading protocol repeated to evaluate the damage associated to the in the in-plane response without the top gaps. The in-plane tests were carried out with the same loading system adopted for the tests described in [17] (Fig. 7b). The in plane horizontal cyclic load was applied by means of a hydraulic jack hinged to the upper beam of the steel frame and to the reaction frame. During the in-plane tests, gauges monitored the specimen inter-story drift, the load applied by the jack, the relative subpanels vertical sliding and their uplift. After the in-plane test, the specimens were tested out of plane, in order to evaluate their stability in presence of the possible damage suffered during the previous test phase. The out-of-plane loading system was designed in order to approximate the effect of a uniformly distributed transverse load on the infill. One-way cyclic loading of increasing amplitude, under force control, was applied. The set-up (Fig. 8) applied eight equal point loads to the wall, two per each masonry sub-panel mid-axis, at ¼ and ¾ of the infill height, in order to approximate the parabolic profile of bending moment diagram on each vertical plank (Fig. 8c). The action was applied by a hydraulic jack (3 in Fig. 8a) supported on a reaction beam (2). The latter was connected to the steel frame columns (1) to obtain a self-balanced system. The jack transferred the load to the three-dimensional distribution frame (4) shown in Fig. 8b, which applied the load in the eight points of the specimen. The system consisted of three levels of isostatic beams, each of them transferring the load from the center to the symmetric pinned-end supports. Gauges lined up at the infill mid-height and along vertical axis of planks and sub-panels monitored the out-of-plane relative infill-frame displacement.
602
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
3
Sub-panel #4
1
Sub-panel #3
2
Sub-panel #2
3
Sub-panel #1
4
2
(a) phase 1: cut size and sequence
(b) phase 2: infill cut operation
(c) insertion of the vertical joints
(d) downgraded infill before testing
Fig. 5. Procedure for the insertion of the vertical joints in the existing infill (Specimen B).
(a) exploded view of the plank-beam connection
(b) cross-section of the vertical element
(c) U shaped plate dowelled to the upper beam
(d) interlocking between the hollow clay block and the injected grout Fig. 6. Details of Specimen B.
603
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
Phase 2
Phase 1
Phase 4
Phase 3
(a)
(b)
Fig. 7. Loading protocol (a) and set-up (b) for the in plane test. The out of plane test was performed in Phases 2 and 4 after the in plane loading of Phases 1 and 3, respectively.
(a) picture of the out of plane loading system
(b) load distribution system (4, in figure a)
(c) experimental vs. ideal moment profile on the central vertical planks
Fig. 8. Out of plane test set-up.
4. Test results 4.1. Specimen A Fig. 9a and b shows the force-vs.-drift in-plane response of Specimen A in the test Phases 1 and 3, respectively. In Phase 1, the specimen highlighted a ductile cyclic behavior with negligible damage, characterized by a developed resistance equal to 68 kN at 2.5% drift, corresponding to about 48 kN infill net contribution. The alternate rocking of each sub-panel around its own toes and the sliding along the vertical planks governed the infill deformation. Recalling the resisting mechanism presented in Section 2.1, the free rocking of the sub-panel provides a limited in-plane reaction (2 kN according to Eq. (1), only about 4% of the infill net contribution at 2.5% drift), while the friction contribution governs the response. Subsequent cycles at the same drift amplitude showed a stiffness degradation, probably due to the local abrasion of the sliding surfaces and gaps adaptations that reduce the effect of the friction mechanism on the response. Sub-panels toes suffered only limited damage at the end of the test Phase 1. The results of the out-of-plane test (Phase 2) highlighted a high resistance of the infill. Thanks to the retaining planks, the infill resisted a total load equal to 72 kN, corresponding to an equivalent uniform acceleration of 4.2 g (calculated as the lateral force divided by the mass of the infill quantified in 1.7t), before the test was
stopped to continue with Phase 3. Fig. 10a shows the out of plane deformation, measured at mid-height of the central vertical planks, for the applied load cycles. The specimen exhibited a bilinear envelope curve with a stiffness reduction at the first application of a load equal to 30 kN, possibly related to a settlement in the planks-to-beam connection, testified by the residual deformation and a creaking perceivable at this load step. No cracks were detected in the masonry. Fig. 10b reports the shape of the wall out of plane deformation, for increasing values of the applied action. The profile results by interpolating the measures of the gauges lined up at mid-height, whose horizontal position along the wall length is indicated on the x-axis. For each loading step, the maximum out-of-plane deformation occurred in the two central masonry sub-panels, indicating the triggering of a relative horizontal sliding with respect to the central vertical plank, whose amplitude was limited below 1 mm by the shear keys connecting the masonry to the vertical planks. After the filling of the top gap with high strength mortar, the inplane response of Specimen A was significantly stiffened (Phase 3). As shown in Fig. 9b, a hardening behavior in both the directions occurred, with a remarkable increase of the resistance with respect to Phase 1 for drift higher than 1%. At 3% drift no damage occurred in the masonry. For each sub-panel, a beam contact length of about one third of the width was measured by means of additional local gauges. Such contact occurred alternatively along the sub-panel
604
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
80
Lateral Force [kN]
60 40 20
Bare Frame Max Resistance
0 Bare Frame Max Resistance
-20 -40 -60 -80
-3.5 -3 -2.5 -2 -1.5 -1 -0.5 0
0.5
1
1.5
2
2.5
3
3.5
Drift [%]
(a) Phase 1 (adapted from [23]) 160
Lateral Force [kN]
120 80
Phase 1 peak
40 0 -40
Phase 1 peak
-80 -120 -160
-3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Drift [%]
(b) Phase 3 Fig. 9. In-plane test results on the original Specimen A (a) and after filling the top gaps (b).
diagonals and a local permanent compressive deformation was observed in the horizontal planks adjoining the beams, after demolition. Fig. 10c highlights a stable out of plane behavior up to the maximum imposed action of 102 kN, exhibiting higher stiffness than that obtained in Phase 2 for deflection higher than six millimeters. Horizontal cracks in the sub-panels occurred during the last run (Fig. 11), at 93 kN applied, and no significant stiffness degradation was observed up to the maximum load of 102 kN, after which the test was stopped. The filling of the gap ensured therefore a higher out-of-plane capacity to the infill, most probably thanks to the activation of the vertical arching mechanism spanning in between the frame beams. 4.2. Specimen B Fig. 12 shows the test results for Specimen B. The in-plane response (Fig. 12a) was symmetric and ductile, with very low resistance and no damage. Reloading cycles showed a certain stiffness and strength degradation. Also in this case, the sliding mechanism governs the infill in-plane resistance, since the net calculated contribution of the rocking mechanism is lower than 1 kN. However, in this case the infill reaction was significantly smaller. A possible explanation, which requires further investigation, is the presence of the little gap between each column and the adjacent vertical planks. Such gap was monitored during the test and its partial closure was measured, testifying a horizontal dilatation of the infill during the in-plane response. Closure or opening of possible gaps
on the two sides of the central vertical planks were monitored, as well. In this case, values of gap closure and opening up to 0.6 mm were measured, suggesting the presence of an initial gap left by the grout shrinkage and a tendency to dilatation of the infill. Regarding the masonry sub-panel toes, note that the disruption visible in Fig. 12b is the result of the necessary local demolition for the insertion of the vertical planks in the construction phase. The results of the out-of-plane test (Phase 2) highlighted a significant resistance of Specimen B, capable of bearing a maximum applied load equal to 40 kN, equivalent to a 4 g transversal acceleration (infill mass equal to 1t). Fig. 12c shows the load vs. out-ofplane maximum displacement in the three central planks. The stiffness progressively reduced by increasing the load. The reduction of the stiffness was directly connected to the progressive triggering of cracks in the masonry sub-panels (Fig. 12b). Fig. 12d shows the profile of out-of-plane deformation of the infill at mid-height, for the different load steps and the corresponding equivalent acceleration values. Fig. 13 shows the profile of deformation along the central vertical plank and the adjacent sub-panel. The graphs show a practically cylindrical infill deformation, governed by the vertical plank flexibility. The infill did not develop a continuous resisting arch in the masonry thickness along the horizontal direction between the frame columns, as testified by the central plank deforming less than the others. Therefore, the sole vertical planks supported the total out-of-plane action. Fig. 13 testifies also an out-of-plane sliding of the masonry sub-panels that
605
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
x
5
80
Cyclic response
4
60 3
50 40
2
30 20
1
10 0
0 0
2
4
6
8
10
Out of plane displacement [mm]
Out of plane Load [kN]
70
Equivalent acceleration [g]
Envelope
0 1 2 3 4 5 6 7 8 9 10
12
0.3g 0.7g 1.4g 2.2g (1) 2.2g 2.9g 3.6g 4.2g
0
100
Out of plane displacement [mm]
200
300
Position X [cm]
(a) Phase 2: out of plane load vs. maximum displacement (M4 gauge)
(b) Phase 2: out-of-plane displacement profile increasing the equivalent acceleration
Phase 2 envelope
5 4 3 2 1 0
0
2
4
6
0
6
Phase 4 (filled gap)
8
10
12
Out of plane max displacement [mm]
(c) Phase 4: out of plane load vs. maximum displacement (M4 gauge)
Out of plane displacement [mm]
110 100 90 80 70 60 50 40 30 20 10 0
Equivalent Acceleration [g]
Out of Plane Load [kN]
x
0.7g
2
1.4g
4
2.1g
6
2.8g 3.5g
8
4.2g
10
4.9g
12
6.1g
14 0
100
200
300
Position X [cm]
(d) Phase 4: out-of-plane displacement profile increasing the equivalent acceleration
Fig. 10. Out-of-plane test results on the original Specimen A (a and b) and after filling the top gaps (c and d).
Fig. 11. Crack pattern at the end of the out of plane test on the specimen A at the end of Phase 4.
deformed more than the retaining planks and a local deformation of the plank support on the top beam; however, the response was stable without strength degradation.
Considering the limited damage of the specimen B after the outof-plane test, the top gaps were filled and the same in-plane loading history of test Phase 1 was repeated twice: first the filling was made with well compacted rock wool (meant for thermal performance) to evaluate its effect on the structural response; then high strength mortar was used in order to explore the efficiency of the intervention in the event of no gap is created between the masonry sub-panels and the frame top beam. Fig. 14a compares the cyclic in-plane responses of the Specimen B with the two types of gap filler. The rock wool filler left practically unaltered the response, which resulted weaker than the first in-plane test (Phase 1) following the same trend of progressive strength reduction along the reloading cycles. The mortar filler produced a stiffer and stronger response. In this last case, damage of the sub-panels corners occurred starting with local cracking at 0.66% drift and evolved into crushing after 2% drift, after which a strength degradation occurred. Fig. 14b shows the damage pattern at the end of the test. Despite some significant damage limited to the sub-panels corners, the response was stable up to very large drift values, unlike typical existing infills performance. The infill resistance remained low (about 30 kN), consistently with the low masonry compressive strength, thus promising a limited infill-frame interaction.
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
Lateral Force [kN]
606
10.0 8.0 6.0 4.0 2.0 0.0 -2.0 -4.0 -6.0 -8.0 -10.0
31kN
32kN 18kN
-3
-2
-1
0
1
2
20kN
27kN
20kN
24kN
3
Drift [%]
(a) in-plane test (Phase 1)
(b) crack pattern in the infill sub-panels at the end of the out-of plane test (Phase 2) x
Out-of-plane load [KN]
35
4
30
3
25 20
2
15
Design acceleration
10
1
5 0
0 0
0
Out of plane displacement [mm]
H3 H5 H7
40
Equivalent Acceleration [g]
45
0.4g
2
1.2g
4
1.6g 2g
6
2.4g
8 3g
10 12
4g
14
2 4 6 8 10 12 Out of plane displacement [mm]
0
(c) out of plane test (Phase 2)
100 200 Position X [cm ]
300
(d) infill displacement profile along the horizontal direction, during the out-of-plane test
Fig. 12. Experimental results for the existing infill with vertical sliding joints (Specimen B).
0.4g
250
200
200 2g 1.6g
150
2.4g
Position Y [cm]
250
4g
3g
100 50
0.8g 1.2g
4g 2g
150
2.4g
3g
8
10
1.6g
100 50
0
Y
Y
Position Y [cm]
0.4g 0.8g 1.2g
0
0
2
4
6
8
10
Out of plane displacement [mm]
(a) central vertical plank
0
2
4
6
12
14
Out of plane displacement [mm]
(b) infill sub-panel #2
Fig. 13. Out of plane deflection along the vertical direction, increasing the equivalent acceleration.
5. Comparison of vertical versus horizontal sub-panel performance The performance of the prototype infill built with vertical sliding joints is compared in Fig. 15 with those of other infills [24] with same material properties and boundary conditions of Specimen A, but different construction technique: (i) a solid infill, built in
running bond and in adhesion to the frame bay, according to tradition; (ii) an infill with horizontal sliding joints. The net infill responses are compared, as envelopes of the in-plane hysteretic curves subtracted of the bare frame contribution. Fig. 15 shows the dramatic reduction of the infill strength obtained thanks to the insertion of sliding joints. The vertical panel configuration obtained the most efficient reduction. Its response however is
607
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
Lateral Force [kN]
30 20 10 0 -10 -20
Mortar Filling
-30
Rock Wool Filling
-40 -3
-2
-1
0
1
2
3
Drift [%]
(a) force-vs.-drift experimental curve
(b) damage pattern at the end of the test
450 350 250 150 50 -50 -150 -250 -350 -450 -3.5
150
Solid Infill Horizontal Joints Specimen A Specimen A filled gap
-2.5
-1.5
-0.5
0.5
1.5
2.5
3.5
Infill net reaction [kN]
Infill net reaction[kN]
Fig. 14. In plane test on Specimen B after the top gapfilling.
Horizontal Joints Specimen A Specimen A filled gap
100
50
0 0
1
2
3
Drift [%]
Drift [%]
Fig. 15. Comparison of the experimental in-plane response for different infill configurations: envelope of the net reaction of the specimen under cyclic loading.
significantly affected by the gaps allowed in the infill. For Specimen A, after filling of the top gaps (Phase 3) a force-drift response similar to that of horizontal joints was obtained, except for a very low initial stiffness, which was affected by the loading history on the same specimen before the gap filling. As a consequence, the filling of such gap for thermal, acoustic and permeability issues deserves particular care, and a soft material may be preferred, as demonstrated in the test performed on Specimen B filling the gap with rock wool.
6. Conclusions The paper describes an innovative construction technique for masonry infill, based on the wall partitioning in vertical subpanels. The experimental results on two prototype infills are described in detail. A dramatic downgrade of the in-plane response of the infill inside the frame was obtained for the Specimen A, if compared to the stiff reaction typical of traditional continuous infill. The out-of-plane response was stable and reliable, thanks to the connection of the vertical planks, partitioning the infill, to the frame beams, and to the negligible damage suffered by the specimen under in-plane loading. The feasibility of the application of the downgrade technique to an existing hollow block masonry infill was tested with the second specimen (Specimen B). Specific procedure and detailing for the insertion of the vertical planks in the masonry and their connection to both the existing frame and masonry infill are described. The entire intervention is operated from one side of the specimen, to simulate the retrofit of an enclosure infill wall from the outside. Specimen B showed an in-plane net infill contribution of few kilo-newtons, practically nullifying the interaction with the frame. Such response, significantly weaker
than Specimen A (16% resistance ratio), cannot be explained only with the different type of masonry adopted for the two specimens (about 60% of both thickness and mass ratio); most probably it depends on the small gaps left in the insertion process between the lateral vertical planks and columns, and on the shrinkage of the grout injection, used to restore the walls continuity, that reduced the friction mechanism. A secondary role in the infill resistance is ascribable to the masonry mechanical properties, as no damage was activated in the sub-panels. A detailed model of the structure is needed to explore the role of the single infill design parameters (particularly the bay and sub-panels geometry, the sliding interfaces friction coefficient and contact conditions) and to extend the results of these first prototypes. The intervention on the existing infill showed to be particularly efficient for its out-of-plane resistance; in fact, the specimen sustained a transverse load equivalent to a 4 g acceleration after the application of several in-plane load cycles, up to 2.5% drift. The infill downgrade is beneficial in terms of post-earthquake damage reduction, but it also simplifies the prediction of the building seismic response with respect to the well-known uncertainties when dealing with traditional infill walls. From the designer point of view, because of the reduced infill frame interaction, the modeling can reliably assume, in a first step, the structure as a simplified bare frame, so neglecting the stiffening effect of the infills. Then few adjustments could take into consideration the infill contribution, dramatically reduced by the proposed technique.
Acknowledgements The authors gratefully acknowledge Malek Neffati, Carlo Piacentini, Michele Moretti, Alessia Ghisla and the technicians of
608
M. Preti, V. Bolis / Engineering Structures 132 (2017) 597–608
the P. Pisa Lab of the University of Brescia for their assistance in the experimental testing. The presented study was partly developed in the research program financed by the ‘‘Presidenza del Consiglio dei Ministri Dipartimento della Protezione Civile” within the Reluis research program; the present publication, however, does not necessarily reproduce the Department position and judgments. A special thanks to Prof. Ezio Giuriani for all the suggestions and fruitful discussions. References [1] Shing PB, Mehrabi AB. Behaviour and analysis of masonry-infilled frames. Prog Struct Mat Eng 2002;4:320–31. [2] EERI. The MW 6.3 Abruzzo, Italy, Earthquake of April 6, 2009; 2009. [3] FEMA. Reducing the risks of nonstructural earthquake damage - a practical guide; 2012. [4] Sucuog˘lu H. Implications of masonry infill and partition damage in performance perception in residential buildings after a moderate earthquake. Earthq Spectra 2013;29:661–7. http://dx.doi.org/10.1193/1.4000147. [5] EERI. 1994 Northridge earthquake reconnaissance report. Earthq Spectra 1996:12. [6] EERI. 1999 Kocaeli, Turkey earthquake reconnaissance report. Earthq Spectra 2000:16. [7] Mohammadi M, Akrami V, Mohammadi-Ghazi R. Methods to improve infilled frame ductility. J Struct Eng 2011;137:646–53. http://dx.doi.org/10.1061/ (ASCE)ST.1943-541X.0000322. [8] Preti M, Bettini N, Plizzari G. Infill walls with sliding joints to limit infill-frame seismic interaction: large-scale experimental test. J Earthq Eng 2012;16:125–41. http://dx.doi.org/10.1080/13632469.2011.579815. [9] Misir IS, Ozcelik O, Girgin SC, Kahraman S. Experimental work on seismic behavior of various types of masonry infilled RC frames. Struct Eng Mech 2012;44:763–74. http://dx.doi.org/10.12989/sem.2012.44.6.763. [10] Markulak D, Radic´ I, Sigmund V. Cyclic testing of single bay steel frames with various types of masonry infill. Eng Struct 2013;51:267–77. http://dx.doi.org/ 10.1016/j.engstruct.2013.01.026. [11] Vailati M, Monti G. Earthquake-resistant and thermo-insulating infill panel with recycled-plastic joints. In: D’Amico S, editor. Earthquakes and their impact on society. Springer International Publishing; 2016. p. 417–32. [12] Da Porto F, Verlato N, Guidi G, Modena C. The INSYSME project: innovative construction systems for earthquake resistant masonry infill walls. In:
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
Proceedings of the 16th international brick and block masonry conference (IBMAC 2016), Padova, Italy. Tasligedik AS, Pampanin S. Rocking cantilever clay brick infill wall panels: a novel low damage infill wall system. J Earthq Eng 2016:1–27. http://dx.doi. org/10.1080/13632469.2016.1190797. Vintzileou E, Adami CE, Palieraki V. In-plane and out-of-plane response of a masonry infill divided into smaller wallets. In: Proceedings of the 16th international brick and block masonry conference (IBMAC 2016), Padova, Italy. Wang Z, Totoev Y, Lin K. Non-linear static analysis of multi-storey steel frame with semi-interlocking masonry infill panels. In: Proceedings of the 16th international brick and block masonry conference (IBMAC 2016), Padova, Italy. Silva L, Vasconcelos G, Lourenço P, Akhoundi F. Experimental evaluation of a constructive system for earthquake resisting masonry enclosure walls. In: Proceedings of the 16th International Brick and Block Masonry Conference (IBMAC 2016). Padova, Italy; 2016. Preti M, Migliorati L, Giuriani E. Experimental testing of engineered masonry infill walls for post-earthquake structural damage control. Bull Earthq Eng 2014;13:2029–49. http://dx.doi.org/10.1007/s10518-014-9701-2. Langenbach R. Learning from the past to protect the future: armature crosswalls. Eng Struct 2008;30:2096–100. http://dx.doi.org/10.1016/j. engstruct.2008.04.028. Morandi P, Milanesi RR, Magenes G. Innovative seismic solution for clay masonry infills with sliding joints: principles and details. In: Proceedings of the 16th international brick and block masonry conference (IBMAC 2016), Padova, Italy. Preti M, Bettini N, Migliorati L, Bolis V, Stavridis A, Plizzari GA. Analysis of the in-plane response of earthen masonry infill panels partitioned by sliding joints. Earthq Eng Struct Dyn 2016;45:1209–32. http://dx.doi.org/10.1002/ eqe.2703. Bolis V, Stavridis A, Preti M. Numerical investigation of the in-plane performance of masonry-infilled RC frames with sliding subpanels. J Struct Eng 2016;0. http://dx.doi.org/10.1061/(ASCE)ST.1943-541X.0001651. Marini A, Belleri A, Feroldi F, Passoni C, Preti M, Riva P, et al. Coupling energy refurbishment with structural strengthening in retrofit interventions. SAFESUST workshop. A roadmap for the improvement of earthquake resistance and eco-efficiency of existing buildings and cities, Ispra; 2015. Facconi L, Minelli F, Giuriani E. In-plane cyclic tests on hollow clay brick masonry infills retrofitted by glass fiber mesh reinforced mortar coating. In: Proceedings of the 16th international brick and block masonry conference (IBMAC 2016), Padova, Italy. Preti M, Bolis V, Stavridis A. Design of masonry infill walls with sliding joints for earthquake structural damage control. In: Proceedings of the 16th international brick and block masonry conference (IBMAC 2016), Padova, Italy.