Cable-net facades with novel glass nodes: development, design, and testing

This study focuses on designing an innovative Glass Node for cable-net glass facade structures. The realization process of the Glass Node was done by establishing conceptual relationships between architecture and facade structures, conducting a literature review on previous studies that include technological and philosophical aspects, analyzing design principles of cable-net facade systems, and developing a point fixing component for cable-net facade. Literature review in this research consists of architectural discourses, which are generated by the philosophers and researchers, as well as it includes the development of facade technologies, and its relations with architectural design. In the context of this review, the history of lightweight structures, which are essentially based on Frei Otto's design approach of cable structures, classification of structural glass facades and cable-net facade component analyses according to related standards are also briefly displayed. As a result of the literature review, a research method is developed, which is de-recomposition of the elements of the target component in order to analyze and design an innovative point-fixing component. After the literature review, design of the cable-net facade was performed in four stages which are architectural design of cable-net facade, structural design of cable-net structure, steel frame and insulated glass unit (IGU). The cable-net facade was designed as 5,136 m x 8,812 m which consists of 20 insulated glass units (IGU), Ф14 mm steel wire ropes, anchors, the Glass Node as point fixing components, sealants, and steel frame. After architectural design stage, the cable-net structure was designed, based on the Allowable Stress Design (ASD), ASCE19-16, AISC360-10 and ASCE 7-10. According to the analysis results, the maximum reaction force and deflection were observed at the combination of positive wind load, positive temperature load, dead loads, and prestress load of the designed cable-net. While the maximum reaction force at the pin support is 43,29 kN near the middle axes of structure, the maximum deflection is calculated to be 97,95 mm, observed at the fixing point located at midpoint of structure. If the structure is tested for strength and deflection limits according to related standards (59,95 N and 11,36 mm), it can be seen that the cable-net facade ensures the strength and deflection limits. After the cable-net facade design, a steel frame prototype testing was designed by using SAP2000 according to ASD and AISC360-10. As a result, the steel frame consists of 260x260x12 mm steel box profile beams with 100x100x10 mm steel braces and 260x260x12 mm box profile steel columns. Furthermore, as a final component, insulated glass units were analyzed at SJ Mepla Software. The maximum strength and deflection were determined to be 33,34 MPa and 9,20 mm respectively which also ensures the strength limit (80 MPa) and deflection limit (10 mm) of IGU. Afterwards, the design of the cable-net facade, architectural and structural design, prototyping and testing process of the Glass Node is executed. Design of the Glass Node was performed in two main steps; the architectural design and the structural design. The architectural design and dimensioning process were carried out according to the cable-net facade that was previously designed. The Glass Node was configured as a 170x170x10 mm prism-shaped steel/glass composite component. Moreover, it consists of 2 layers of laminated and tempered glass cap (TG103-1), 1 layer cross shaped stainless steel core (SS104), 4 layers tempered and laminated glass, (TG103-3), 2 layers tempered and laminated glass in two types (TG103-3 and TG103-4), two types of stainless steel bedding element (SS101 and SS103), 4 units stainless steel roll pins (SS102), 6 units M12 countersunk bolts (SS105) and POM bedding elements (PM101-1, PM 101-2, PM 101-3, PM 101-4, PM 101-5, PM 101-6). In accordance with the architectural design, the stainless steel and glass elements are bonded to each other via Sentryglass lamination and SikaFast-3131 methacrylate based adhesive . Following the architectural design, structural analysis of the Glass Node was performed by using the Solidworks Software. During the structural design process, effective loads of the Glass Node were determined to be 2,4 kN wind load and 1,65 kN dead load of IGU. Furthermore, the analysis was conducted by making the model of the Glass Node as three separated parts which are the cap of Glass Node (TG103-1), the stainless steel cross shaped core (SS104) and the glass-stainless steel composites. From analysis; the most critical component was determined as the cap of the Glass Node of which the maximum stress is 29 MPa. This is less than the limit strength of a tempered glass (80 MPa) and it can be seen that all elements of the Glass Node ensure the design strength limits. Nevertheless, structural performance of the bonding materials was to be determined by way of testing. Therefore, a prototyping process and a manufacturing process were executed and afterwards, primarily bonding testing was performed in order to determine adequate bonding materials. Prototyping process of the Glass Node was executed in two steps; design prototyping and engineering prototyping . The design prototyping was carried on by creating a 3D model of the Glass Node and printing all the elements via 3D printers. Afterwards, the 3D-printed Glass Node was assembled, and connections were evaluated in terms of their applicability. As a result, some revisions were added to the architectural design of the Glass Node, such as extra bedding materials, in order to prevent interaction between metal-glass elements. Afterwards, the process was carried out by manufacturing the Glass Node with real materials. Except the bonding of 2 units TG 103-3 and SS104, all the components were bonded via Sentryglass lamination material. However, 3 different adhesive materials (Sentryglass, Sikasil SG500 and SikaFast-3131) were used for bonding for TG103-3 and SS104. After the bonding procedure, primarily tension test process was performed in order to determine the adequate adhesive for bonding of SS104 and 2 units TG103-3. Preliminary test procedure was executed according to DIN18008-3 and additionally the target strength values of Glass Facade were determined to be 2,4 kN (yield strength, Fa) and 7,2 kN (ultimate strength, Ft). All the test procedures were performed by using a tension test machine located at the Facade Testing Institute Laboratories at Çatalca, Istanbul and the results were video recorded. According to the test results, the most adequate adhesive is determined as SikaFast-3131 of which the yield (Fa) and ultimate strength (Fu) are respectively 10,36 kN and 11,54 kN. However, Sentryglass (lamination sheet) and Sikasil SG-500 (structural silicone) could not reach the target strength values. Hence, SikaFast-3131 was used for the main test procedure of the Glass Node. When the test results are analyzed according to the failure type, it is seen that all samples fail due to bolt stripping. Therefore, M10 bolts were replaced by M12 bolts in order to reach higher tension strengths. After revising the bolt types, the main test procedure was also executed according to DIN18008-3. 5 samples are tested, and the test data are recorded by using a software for test tension machine and a video camera. The test was performed in a stabilized environment of 30 °C and 80% humidity. According to the test results, the average ultimate breaking strength and maximum displacement were measured to be 19,22 kN and 10,22 mm, respectively. Moreover, while the maximum ultimate breaking strength (21,04 kN) and maximum displacement (12,68 mm) were observing Sample 5 and Sample 1 respectively. The minimum ultimate breaking strength (16,30 kN) were observed for Sample 2 and minimum displacement were determined to be (7,93 mm) for Sample 2 and Sample 3. Furthermore, average ductility ratio of test samples is calculated to be 1,33. The maximum and minimum ductility ratios were determined to be 1,59 for Sample 4 and 1,0 for Sample 2 respectively. As a conclusion, the strength of the Glass Node is 8 times greater than Fa (2,4 kN) and approximately 2,5 times greater than Fu (7,2 kN). Furthermore, when the failure types of the samples are analyzed; while Sample 1,2,3 and 5 fail by breakage of exterior cap of Glass Node (SS103-1), Sample 4 fails because of bolt stripping. According to the test results, Glass Node ensures the design limits determined according to international standards and it could be assumed as an innovative cable-net facade component which can be integrated to full scaled cable-net facades. The same results show that the strength of the Glass Node is 8 times greater than Fa value and 2,7 times greater than Fu value, which means that the design of the Glass Node can be optimized in terms of architectural and structural design. In conclusion, the study fulfills the objectives of the research by means of increasing transparency ratio of cable-net facade systems by %17 percent with satisfactory structural performance.