Structural conception and collapse control performance based design of multistory structures under accidental actions

Project Coordinator

Phase: Preliminary investigations

Objectives

• review of existing methods, structural concepts and analytical tools for evaluating the progressive collapse potential of new and existing buildings,
• identification of gaps in knowledge
• evaluation of effectiveness of collapse control based design for protection of building structures in case of accidental actions

Task 1-1: Review of existing methods, identification of research needs

The activity is devoted to a review of research on progressive collapse, related guidelines and recommendations as well as to the gaps in knowledge and research needs following the latest developments in the field. Case studies that can offer some insight into the problems of progressive collapse and lessons to be learned are also investigated. The benefits of seismic design philosophy on the integrity of the multi-storey structures will be identified.

Task 1-2: Preliminary analysis and selection of case study structures

In order to test realistic specimens, they are selected from representative multi-story frame structures, designed according to present code provision but without considering any accidental actions or loss of any element. Moment frames and braced frames are considered, with one way or two way behaviour, and with or without composite floors. Note that two-way structural framing systems have greater robustness than structures that are designed to span one way. The response can be improved also by using a two-way floor slabs, by adding supplementary ties or by allowing the gravity beams (along the simple spans) to develop the catenary behaviour.

Deliverable: A Summary Report: the report describes the present development in the field of progressive collapse and causes of abnormal loading, underlines the gaps and research needs. The main directions of the research are indicated.                (link doc 4.1)

Phase: Design of experimental and numerical program

Objectives

• design of specimens for experimental program, based on case study structures: typology, materials, technology of execution, loading parameters, data acquisition, rigs and test set-up, supply of materials and specimens.
• design of numerical program

Task 2-1: Design of specimens for experimental program

The experimental program consist of test on:

• subassemblies of two-bay two-span frame structures
• beam-to-column joints 
• joint components – T-stubs with different configurations
• welding details, different strain rates, ambient and elevated temperature
• materials (steel, reinforcing, concrete)

The largest specimens are the sub-assembly structures and are two-bay two-span models. Their dimension is 12m x 12m x 2m. Therefore, due to the limitation in the testing capacity of the host institution, the subassemblies specimens are half-scaled to cover two-bay two-span configuration.  Similar tests are not available in the literature. Beam-to-column joints are tested under monotonic load beyond the flexural capacity to initiate the catenary behaviour. Similar tests have not been performed and are of great interest for column loss scenarios, where the verticals deflections can be large enough to develop the catenary tension. Connections must be capable of transferring catenary tension from beam to beam across the column. Different weld details are of concern, i.e. fillet welds, bevel butt welds. Strain rate from cvasi-static to dynamic range are of concern. Both the strength of steel and concrete increases significantly when the strain rate increases. The high strain rate might affect the ductility and failure mode of materials and welds.

There is a large amount of specimens to be produced and supplied for testing program, as follows:

• Half-scale subassemblies of two-bay two-span configuration, steel and composite (steel-concrete) 
• Full scale beam-to-column joints
• Macro-components (T-stubs)
• Small specimens: material specimens, weld details.

The supplier will be selected so as to assure a good quality of materials. In order to guarantee constant quality parameters, the fabrication of all steel specimens will be ordered to the same steelwork shop, with good expertise in the field. For the composite floors of the sub-assembly test, the concrete will be poured in the laboratory.

Task 2-2: Design of numerical program

The structures are designed to resist moderate and severe seismic actions. Different column loss scenarios are considered. Next figure shows the geometry of the reference frames. Sub-assembly and beam-to-column joint specimens will be selected from the reference structures and tested experimentally. The details are given in Phase 4 and Phase 5.

Deliverables: Technical drawings, shop drawings and materials sspecifications for specimens: design of subassembly and beam-to-column joints, design of macro-component (T-stub), weld details and material specimens (main geometric dimensions, size and type of cross sections, connections, quality of materials, weld details and base material specimens), fabrication drawings and list of materials), as well as the testing rigs and set-up in laboratory.

Experimental program (Phase 3, 4 and 5)

- Task 3-1: Experimental test on base materials and weld details

The mechanical properties of materials and the performance of the welds are important for the capacity of structural members to resist extreme load events. Tensile test at room temperature and toughness test at room temperature (20° C) and elevated temperature (540° C), under cvasistatic (0.05 mm/sec) and dynamic loading (10mm/sec) are performed. Impact, blast and explosions ask for high toughness of base material and welds in those members and connection components where a ductile response is expected. The weld details are loaded under monotonic loading.

- Task 3-2: Experimental test on T-stub macro-components

The most important macro-components of bolted end plate beam-to-column connections are the T-stubs. The behaviour of such macro-components depends on the material characteristics, bolt arrangement and type of loading. The design of T-stubs for a ductile failure mode (see EN 1993-1-8) is based on a monotonic loading. The ductile failure mode can be altered in case of a dynamic loading (strain rate effect) or in case of fire (elevated temperature). In addition, the catenary action of beams can only develop if the connections have the capacity to sustain the tensile capacity of the beam. In order to evaluate the influence of parameters presented before, T-stubs are tested at room (20° C) and elevated temperature (540° C), under cvasistatic (0.05 mm/sec) and dynamic loading (10mm/sec).

Deliverables: Test results

Experimental program (Phase 4 and 5)

- Experimental test on joint specimens

Pure beam-to-column steel joints are tested experimentally until the plastic rotation capacity is attained and further until the catenary action develops. Bolted and welded joints, full strength and partial strength are selected (80% of the beam capacity). The tests consist of pushing the middle column down until the failure of the specimen (push-down test) by means of hydraulic actuators. Numerical studies are in progress to evaluate the behaviour and key parameters (yielding, rotation capacity, ultimate capacity, total deformation capacity).

- Testing in blast conditions

In this task, the blast effect on the beam-to-column components is investigated. The steel framing consists of a column and girder stud. Two types of girder stud to column connections, one rigid and one pinned are tested. The response modes and material behavior (fracture) is much different in a blast environment. Moreover, the stand-off distance is important and affects the behaviour. Therefore, different charges at different distances will be performed.

- Experimental test on subassembly specimens

These tests simulate the response of an 3D assembly following a column loss. By using the membrane (catenary) effect of the floor structure or the bridging effect of the frame, the load of the lost column can be transferred to the adjacent members. The test specimens are half-scaled of a typical floor structure. The specimen is a two-bay two-span structure, with the size of 6.0x6.0x1.5m. Two solutions are tested: with and without concrete slab and two different scenarios are considered: middle column and corner column loss. The tests consist of pushing the column by imposing a vertical displacement with hydraulic actuators until the failure of the specimen (push-down test).

Phase 6 -Numerical program

Task 6-1: Calibration and validation of numerical models based on test results

The output of Phase 3, 4 and 5 serve as input for Phase 6. Models for materials and accidental actions, acceptance criteria and relevant limit states for different damage levels are used to calibrate the numerical models.  

Phase 7 - Design guidelines and recommendations

Objectives

• To provide a synthesis of the results obtained in the project, with the relevant conclusions.
• To provide general design guidelines to avoid progressive collapse in case of extreme events.
• To provide general recommendations for multi-hazard mitigation

Task 7-1: Guidelines for the collapse control performance based design of multi-story frame buildings against accidental actions

Designers would benefit from a standardized procedure in order to ensure general structural integrity of multi-story frame buildings. The method can be used for design of new buildings and for the upgrading of the existing ones. A performance matrix for robustness base design will be also provided.

Task 7-2: Recommendations for best practice in selection of structural system, fabrication and material requirements for improving the robustness

Ensuring the structural integrity of new buildings may be provided by following prescriptive minimum requirements for design. This includes minimum requirements for connections, tie force resistance, continuity and ductility. This can be used for low level of consequence buildings (protection).