Science & Innovation
Engineered solution for increased seismic performance
An Innovative Earthquake Proof AAC Infill Wall System
AAC is used as infill walls all over the world due to its advantages such as being lightweight, providing good insulation, fire-proofing, and high durability. Currently, AAC constitutes more than twenty percent of the infill wall market in Turkey. It is well known that the seismic performance of infill walls in the past earthquakes was insufficient due to both economic and psychological reasons. The infill walls were observed to collapse in the out-of-plane direction in many buildings. In some other cases, despite the framing system being undamaged, it was seen that inhabitants of the buildings were resistant to immediately occupy their apartments if cracking of the infill walls were observed. Therefore, there is a need to devise wall systems that are damage free after earthquakes.
Past researches [1-6] focused mostly on the seismic performance investigation of masonry infilled frames with minor emphasis on AAC infill walls , leaving some gap in the full understanding of the seismic response of AAC infill walls. Few researchers proposed innovative systems [8-13] to enhance the performance of masonry infill walls, again not focusing on AAC solutions. Furthermore, the combined in-plane and out-of-plane response of AAC infill walls was not experimentally studied in the past. In order to address the needs of the AAC community, a comprehensive research program was initiated with the support from Turkish AAC Association and AKG to investigate the seismic response of AAC infill walls. Based on these findings an innovative AAC infill wall system was devised, patented and tested by METU and AKG. The last few years of work on AAC infill walls constitutes the main theme of this paper.
Expected Seismic Performance of AAC Infill Walls
The seismic performance of plastered AAC infill walls was studied by testing four single bay single story half scaled specimens under increasing later cyclic displacement excursions and out-of-plane (OOP) loading. The specimen details and the test setup are shown in fig. 1. An axial load of 200 kN on each column was applied by using hydraulic jacks to simulate the gravity loads from the upper stories of a five-story reinforced concrete frame prototype building. A distributed load of 7 kN/m was placed on the beams by using steel blocks to simulate gravity AAC worldwide 1| 2018 29 loads transferred from slabs based on the analysis of the prototype building. The OOP loading was applied to the specimen by using an airbag loading system. Target concrete compressive strength was 30 MPa for all the specimens, whereas AAC blocks and plaster had a compressive strength of 2.5 MPa and 1 MPa, respectively. Specimen 1 was a bare frame without infill walls, and was tested under in-plane loads to benchmark the load-deformation response of a reinforced concrete frame designed according to the Turkish Earthquake Code . Specimens 2, 3 and 4 had infill walls with 20 mm gap left along the beam interface (fig. 1); except this they had similar frame details to Specimen 1. Specimen 2 was tested under in-plane loads, Specimen 3 was under out-of-plane loading, and Specimen 4 was tested under combined in-plane and out-of-plane loads. The out-of-plane load applied in the testing of Specimen 4 was equal to 30 % of the out-of-plane capacity obtained from the testing of Specimen 3. During the testing of Specimen 3, the OOP pressure on the AAC walls was increased till the failure of AAC infill wall was observed. In contrast, the OOP pressure was constant throughout the testing of Specimen 4 and the same in-plane loading protocol was applied to the specimen till AAC infill wall failed.
Test results including the damage pictures of the specimens and the load-deformation response of the specimens are shown in fig. 2. The lateral load capacity was reached at a drift ratio of about 1.3 %, concrete cover spalling at the column bases was observed at about 2 % drift ratio, and 20 % lateral strength drop was observed at about 3 % drift ratio. Beyond this drift ratio, bar buckling of the longitudinal column reinforcement was observed at the column base plastic hinge regions. This specimen experienced a ductile failure mode with a ductility ratio of about 6. Specimen 2 behaved in a stiffer manner owing to the presence of the plastered AAC wall. No cracking except along the interfaces between infill wall and surrounding frame was observed in this specimen up until a 1 % drift ratio, beyond which strength degradation initiated. The ultimate strength of about 115 kN (-135 kN) dropped to its 80 % at about 2 % drift ratio (fig. 2). The displacement ductility ratio of this specimen was about 4.0, indicating a reduction in deformability compared to the bare frame. Specimen 3 was pushed in the OOP direction by applying uniform pressure until the complete failure of the AAC infill wall occurred in the absence of in-plane loads. A horizontal crack was first observed at the center of the AAC infill wall. Then, this crack slightly inclined with the increase in the pressure along OOP direction. Finally, cracks from both directions converged to each other in a V-shaped cracking pattern resembling a two-way mechanism. This crack caused a total collapse of the AAC infill wall at about 50 kN total out-of-plane load and 17 mm mid-span deflection. The damage pattern obtained from the testing of Specimen 4 (fig. 2) demonstrated that the interaction of the in-plane and the out-of-plane actions caused a significant decrease in the drift ratio capacity of the AAC infill wall. The drift ratio capacity of Specimen 4 was nearly 1 %, at which the total collapse of AAC infill wall was observed (fig. 2). Therefore, the application of nearly 33 % of the OOP pressure capacity resulted in nearly an equal amount of reduction in the ultimate drift ratio.
Turkish Earthquake Code  limits the inter-story deformation to 2 % in frame buildings. This drift limit should at least correspond to the ultimate deformation capacity of a well-designed structure. This expectation was met with sufficient margin of safety upon observing the deformation capacity of Specimen 1. Similarly, the drift capacity of Specimen 2 with the AAC infill wall was found to be sufficient but with less safety margin compared to Specimen 1. However, in the presence of out-of-plane loading simulating the out-of-plane seismic excitations, a significant reduction in the in-plane drift capacity was evidenced in Specimen 4. The findings suggest that the structural engineer may design a reinforced concrete frame building according to a modern seismic code and expect the system to behave in a ductile manner.
However, once the AAC infill walls are constructed, the in-plane deformation capacity of the infilled frame is greatly reduced, even below the code prescribed design levels, due to premature outof- plane collapse possibility. Hence, AAC infill walls should either be considered in the seismic design or relevant measures should be taken to reduce their interaction with the boundary framing members. One solution to mitigate the collapse possibility of infill walls is described below.
Earthquake Proof AAC Infill Wall Performance
Based on the limited test results presented above, AAC infill walls in reinforced concrete frames may not perform satisfactorily during future earthquakes. They are expected to crack extensively not complying with serviceability requirements under service level earthquakes or fail under combined in-plane and out-of-plane seismic demands. In order to allow the RC frame system to behave in a ductile manner similar to the bare frame, and sustain out-of-plane loads without collapse, an innovative system was proposed, patented and tested. The objective was to create an autoclaved aerated concrete block solution for earthquake-safe AAC infill walls. AAC blocks with the slider connectors were manufactured by inserting the metal slider assembly into the AAC blocks (fig. 3). The slider was provided with sufficient tolerance to move in and out of the AAC block depending on the target design drift of the story. For the test specimens, the slider was allowed to move about 80 mm in and out of the block such that cyclic deformations of up to 2 % could be accommodated during the tests. For the beam connections, a similar procedure was used by inserting the slider into the channel connected to the beam in the horizontal direction. For this study, the spacing of slider AAC block was selected as one per three rows of AAC blocks to provide an out-of-plane capacity of 20 kN. The gap at the interface of columns (beam) and the infill wall was filled with stone wool providing insulation and fire resistance (fig. 3). The infill wall was covered with fiber reinforced mesh plaster overlay as a final finish.
Three additional specimens were tested to examine the performance of the proposed system. Specimens, similar to Specimens 2, 3 and 4 in terms of frame properties, but equipped with the proposed slider system, were tested. The observed damage and measured load-deformation responses of the tested specimens, (Specimens S5, S6, and S7), are presented in fig. 4 along with the comparisons with the companion specimens. Cracking of the AAC wall was not
observed up to 2 % inter-story drift ratio during the testing of Specimen S5. The gaps of the slider connectors closed on the loaded wall-frame interface, whereas they opened on the opposite interface. Consequently, the interaction between the frame and the infill was eliminated with the proposed system. The testing of Specimen S5 was stopped when the limit of the drift ratio proposed by the Turkish Earthquake Code  (i.e. 2 %) was attained. At this drift ratio, the lateral strength dropped by about 8 %. This result shows that the proposed system provided sufficient ductility and deformability under in-plane loading. The out-of-plane load carrying capacity of Specimen S6 was measured as 20 kN. This strength was less than that of Specimen S3, however it was deemed sufficient considering the outof- plane capacity to be about 6 times the weight of the AAC wall. It should be noted that the capacity of the AAC infill frames with the proposed system can be increased by using more sliding connectors if needed. The testing of Specimen S6 showed that the proposed system resulted in nearly 1.7 times the out-of-plane ductility measured for Specimen S3. Finally, the load-deformation response of Specimen S7 exhibited only 8 % strength degradation at a drift ratio of 2.5 %. Consequently, the displacement ductility of Specimen S4 increased by a factor of about 2. No sudden out-of-plane collapse of the infill wall was observed during the testing of Specimen S7. The in-plane drift capacity versus the out-of-plane load ratio of the slider block system was significantly improved compared to the conventional AAC infill wall system. Based on the limited test results, it can be stated that the frames with classical AAC infill walls may be unable to meet the code given drift limits, whereas those with the proposed sliding connector system are sufficiently deformable satisfying the drift limit requirements without sustaining heavy damage or collapse.
The seismic performance of AAC infill walls was experimentally investigated in this study. It was found that AAC infill walls may be vulnerable to collapse under combined in-plane and out-of-plane seismic demands and they may not satisfy the drift limits currently given by the seismic codes. In order to overcome these problems, an innovative AAC block wall system was devised and tested. Test results demonstrated that the proposed innovative system is successful to reduce the expected risk imposed by infill walls.
Prof. Dr. Baris Binici obtained his B.Sc. Degree in Civil Engineering at Middle East Technical University and completed his M.Sc. and Ph.D. degrees from The University of Texas at Austin. He joined to Middle East Technical University as a faculty member in 2003. His research interests are in the area of structural and earthquake engineering, reinforced concrete design and modeling, seismic testing and simulations. He has over 100 international publications and acted as a consultant for seismic testing, analysis, and design of structures over 30 projects
Prof. Dr. Erdem Canbay obtained his B.Sc. degree at Istanbul Technical University and completed his M.Sc. and Ph.D. degrees from Middle East Technical University. After completing his postdoctoral study in Purdue University, he joined to Middle East Technical University as a faculty member in 2003. His research interests are reinforced concrete design, seismic testing, and repair and strengthening of RC structures. He has over 50 international publications, and 3 books on reinforced concrete structures, earthquake engineering and strength of materials.
Asst. Prof. Dr. Alper Aldemir obtained his BSc, M.Sc. and Ph.D. degrees from Middle East Technical University. He joined to Hacettepe University as a faculty member in 2016. His research interests are analytical and experimental investigations of concrete and masonry structures. He has over 30 international publications.
İsmail Ozan Demirel obtained his B.Sc and M.Sc degrees from Middle East Technical University (METU). He is currently a PhD candidate at METU Civil Engineering Department Structural Engineering Division. He is also working as a part-time instructor in Bilkent University, Architecture Department. His research interests are experimental and analytical investigation of infilled RC and masonry structures.
Ugur Uzgan obtained his B.Sc. Degree in Civil Engineering from Dokuz Eylül University. He worked as a site engineer in different construction projects, and then he joined AKG Gazbeton. After serving as the Product Director, he has been acting as the Head of Research and Development at AKG. His key interests are research methodology and management for industry and product development.
Zafer Eryurtlu obtained his B.Sc. Degree in Civil Engineering from Dokuz Eylül University, Civil Engineering Department. After graduation, he worked on different construction sites in technical positions. He joined AKG Gazbeton in 2005 and still continues to work as the Product Executive. He coordinates cooperation with academia in developing innovative AAC products.