Metropolis Mega-Development: A Case Study in Fast-Tracked Performance-Based Seismic Design of High-Rise Concrete Towers in Los Angeles

Transcription

1 Metropolis Mega-Development: A Case Study in Fast-Tracked Performance-Based Seismic Design of High-Rise Concrete Towers in Los Angeles Abstract The Metropolis mega-development is a five-parcel block mixed-use development in downtown Los Angeles, California containing 4.1 million square feet of luxury multi-family residential, hotel and retail space, making it the largest development currently in Southern California. Metropolis is comprised of four high-rise concrete core shear wall buildings including a 19-story 350-room hotel, a 39-story 308-unit residential tower, a 42-story 525-unit residential tower and a 57-story 725-unit residential tower. The 57-story tower is currently the tallest all concrete high rise tower located in the western United States. Of the four towers, only the hotel tower was less than the 240 ft height limit prescribed in the code for pure concrete shear wall buildings. As such, while the hotel tower was designed using the prescriptive code approach, the other three towers were designed using a Performance Based Design Approach. This allowed the towers to rely only on the core walls for its lateral resisting system as opposed to the dual system consisting of shear walls and moment frames that would be required if a prescriptive approach would have been followed. The result, a more efficient structural design which provides significant advantages to the project in the form of reduced construction costs, improved architectural freedom and predictable seismic performance in a major earthquake. The purpose of this paper is to present the design of this extremely fast-tracked megaproject and the challenges that came with the fast-track nature of this project. Saiful Islam, Ph.D. S.E. Sampson Huang, Ph.D. S.E. Shafiq Ibrahim, P.E. Fengshuang (Rex) Zhang Saiful/Bouquet Structural Engineers Pasadena, CA contains approximately 4.1 million square feet of gross building area, making it the largest development currently in Southern California. The project is comprised of 1,560 luxury residential units in three towers, 350 hotel rooms, and approximately 74,900 square feet of restaurant and retail space built in two phases. The project was designed and developed in two phases as shown in Figure 1: Phase 1 of the project included 1.1 million square feet of gross building area built on a 2.3 acre lot. This initial phase included a 350-room 18-story hotel tower and a 310-unit 39-story residential tower, both with two levels of basement. See Figure 2. Phase 2 of the project included approximately 3.0 million square feet of gross building area built on a 4.0 acre lot. This phase is comprised of two residential towers, a 525-unit 40-story 449 foot tall residential tower (R2) and a 725-unit 57-story 656 foot tall residential tower (R3), both with two levels of below grade parking and retail at the ground level. In addition to the towers, Phase 2 also included an approximately 1.5 million square feet nine-story podium structure with an amenities deck on the roof and approximately 1,900 parking stalls. See Figure 3. Introduction The Metropolis mega-development is a five-parcel block on 6.3 acres of mixed-use development in downtown Los Angeles, California. Just two blocks from Staples Center and L.A. Live, the development spans two full city blocks and connects the financial and entertainment districts, while adding to the vibrant skyline of downtown Los Angeles. It 1

2 Figure 1 - Metropolis Two Phase Construction 2

3 Figure 2 - Metropolis Phase 1 Overall Plan Figure 3 - Metropolis Phase 2 Overall Plan 3

4 In Phase 1, the hotel tower was kept below 240 feet (measured from grade) so that a prescriptive code approach could be used for its design. However, the 39-story 465 ft tall residential tower was designed using the Performance Based Approach which falls under the alternate design approach allowed by the Code. Since the performance-based design is outside of the prescriptive requirements of the building code, the City of Los Angeles requires that the design is peer reviewed by a panel selected by the City which includes an academic researcher, a practicing structural engineer and a geotechnical engineer. See Figure 4 for an architectural rendering of the Phase 1 towers. discussed at length. The introduction of the moment frames would not only have significantly increased the building cost but it would have also added significant time to construction, not to mention the architectural and space planning impact (due to very large moment frame columns and beams). In the final analysis, it was clear that it was far better to go with performance based approach and rely only on core walls for lateral resistance as it yielded the most cost-efficient structure which could be built faster and easier and provided the greatest architectural and planning flexibility. In Phase 2, the two towers and the 9-story podium structure are functionally attached. However, structurally they were separated from each other via seismic joints. This allowed the two towers to be designed using a performance based design approach while the podium structure was designed using a code prescriptive approach. This also precluded the podium structure, which supported a very heavy and extensively landscaped amenities deck, from penalizing the two towers. Furthermore, the seismic joints also allowed a clear load path without any heavy transfer diaphragms and reduced the risk category classification. It also allowed the tower design, which was on the critical path because of performance based design, to proceed while the design of the amenities deck on the podium/parking structure was being completed, thus saving months in the design time. Figure 5 shows the Phase 2 R3 Tower structural elements and exterior design. Figure 4 - Metropolis Phase 1 Towers The performance based design and associated peer review and approval process is very rigorous and time-consuming and typically extends the design phase schedule by several months, if not more. This is typically a concern on any fast track project and, in the case of Metropolis which is considered to be on a super fast-track, this concern was further amplified. The pros and cons of going with performance based design with extended design schedule versus going with a prescriptive design approach, which would have cut down the design and approval time but would have required a dual system consisting of shear walls and moment frames, were Figure 5 - Metropolis Phase 2 R3 Tower 4

5 The performance based design of the residential towers in both Phases 1 and 2 were done in accordance with the An Alternative Procedure for Seismic Analysis and Design of Tall Buildings Located in the Los Angeles Region document developed by the Los Angeles Tall Buildings Structural Design Council (LATBSDC). Structural System Description Phase 1 The Residential Tower and Hotel of Phase 1 are reinforced concrete structures with shear walls providing seismic force resistance. See Figure 6 for a three-dimensional view of the Phase 1 structure. column-slab joint. To increase usable space and to reduce material cost, high-strength concrete up to 8,000 psi compressive strength was used for the vertical concrete elements including walls and columns. All concrete slab utilizes 5,720 psi concrete mix so that puddling is not required at the column-slab joint per ACI318. As shown in Figure 7, the lateral system of the Residential Tower included one full-height central core wall and four sixstory concrete shear walls up to the amenity deck level. Three separate mat foundations were introduced under the concrete shear walls and individual spread footings were used to support gravity columns outside of mat foundation. With the 39-story above-grade structure, the Residential Tower also includes a two-story subterranean basement that is encompassed entirely by perimeter basement walls that serves to retain soil and to provide lateral support. Figure 6 - Phase 1 Structure 3D View The gravity system of the Residential Tower consists of 8-inch post-tensioned slabs for all levels with the exception of the below grade levels and the floors supporting either the amenity deck or heavy mechanical equipment where conventionally reinforced concrete slabs are more suitable. As a common practice for flat-plate slab construction, shear stud rails were used to increase the punching shear resistance at the Figure 7 - Phase 1 Residential Tower Shear Wall System 5

6 As a result of three sets of shear wall systems employed with staggered top of wall elevation along the height of the building, two major transfer diaphragms were introduced: one is at the amenity deck where seismic forces start to unload from the central core wall into relatively-stiffer six-story shear walls; the other transfer diaphragm is located at ground-level where a similar mechanism occurs with the rigidity of the basement shear walls being much higher than the other taller shear walls and the core walls. The two transfer diaphragms were delicately designed to remain essentially elastic under an MCE level seismic event and in turn multiple drag beams were introduced at those levels to create a clear load path for load transfer. Phase 2 For the two residential towers in Phase 2, the gravity systems are similar to Phase 1 except that higher strength concrete was used at the gravity columns and slabs, up to 10,000 psi concrete mix for columns and 6,000 psi concrete mix for all slabs. In the nine-story podium structure, a post-tensioned flat slab with drop panels was used at parking garage levels where headroom is not sensitive for the parking spaces, which also helped control the slab deflection. As both residential towers have a significant low-rise wing (19-story in R2 and 25-story in R3) attached to the main tower, as shown in Figure 8 and 9, a very simple and practical structural lateral system was developed for these towers with a main core shear wall for the main tower stack and a smaller core that extends only through the lower stack wing to balance the twisting of the towers. Similar to Phase 1, both towers included two-story subterranean levels, however, with the basement walls only partially surrounding the tower foot print since the two towers share architectural functions with the podium structure. To avoid drastic torsion behavior below grade, individual basement shear walls were added at the perimeter of the two towers where the basement retaining wall did not occur. Different from the foundation system of Phase 1, a continuous mat foundation was provided under each entire building footprint for the two Phase 2 towers and each foundation employed two different thicknesses under the lowrise wing and the main tower with a transition in between. A delay strip, similar to Phase 1, was employed between the side wing and main tower for both the R2 and the R3 towers. Figure 8 - Phase 2 Structure and Seismic Joint Seperation Figure 9 - Phase 2 Tower Architectural Rendering 6

7 Performance Objectives The performance based design of all three residential towers followed the procedure described in the 2014 Edition of An Alternative Procedure for Seismic Analysis and Design of Tall Buildings Located in the Los Angeles Region by the Los Angles Tall Buildings Structural Design Council. Table 1 shows the specific performance objectives used for the design. For the service level earthquake, during which the building is required to remain essentially elastic, linear response spectra analysis was performed using ETABS with torsion and P- Delta effect taken into consideration. For the MCE level eartquake, nonlinear three-dimensional time-history analysis was performed to assess and also to validate the performance of the residential towers. Table 2 below summarizes the elements that were considered as inelastic in the Perform-3D model and those that were treated as elastic elements. The time-history analysis involved analyzing for 7 pairs of ground motion at the MCE level rotated in two orthogonal directions (14 analyses total). Table 1 Earthquake Performance Objective for Performance Based Design Earthquake Intensity Service Level Earthquake (SLE) : 50% probability of exceedance in 30 years (43 year return period); 2.5% damping Maximum Considered Earthquake (MCE) : 2% probability of exceedance in 50 years (2,475 year return period); 5% damping Serviceability: Performance Objectives Building is to remain essentially elastic with minor yielding of structural elements, minor cracking of concrete and minor damage to non-structural elements. Repairs, if necessary, are expected to be minor and could be performed without substantially affecting the normal use and functionality of the building. Collapse Prevention: Building is to have low probability of collapse. Claddings and their connections to the structure must accommodate MCE displacements without failure. Extensive structural damage may occur, repairs to structural and non-structural systems are required and may not be economically feasible. Table 2 Nonlinear Model Elements Inelastic elements Shear walls in flexure Coupling beams Slab beams for outrigger effects (slab-wall, slab-column connections) Elastic elements Shear walls in shear Columns Diaphragm slabs of podium Foundations Slab column punching shear 7

8 For each of the towers, the potential location of the plastic hinge in the core wall was carefully analyzed and special confinement reinforcement was detailed accordingly within this plastic hinge zone to ensure ductile behavior during even the most critical earthquake. panel. The kick of the curve right above the amenity deck also indicated that there is a major force transfer in the diaphragm where the stiffer shear walls starts to absorb seismic forces. Analysis Performed Phase 1 R1 Tower The Service Level Earthquake (SLE) evaluation was performed by linear response spectrum analysis that assessed the building behavior subject to multiple criteria, among which the drift limit and coupling beam shear capacity check are the most essential. Figure 10 shows the drift profile of the residential tower for the service level earthquake with a maximum drift limit of 0.5% that ensures that the building behaves elastically, however the requirement is usually met and does not govern overall structural design. Figure 11 - Phase 1 Residential Tower SLE Coupling Beam Capacity Plots Figure 10 - Phase 1 Residential Tower SLE Drift Plot Figure 11 shows the demand-capacity ratios (DCR) for one coupling beam along the height of the building for the SLE analysis. Since the coupling beam is a deformation controlled member and is expected to yield under strong earthquakes, the DCR limit for coupling beams under the SLE analysis is set to be 1.5 as per the design criteria approved by the peer review In a parallel process with the SEL analysis, the building behavior under the MCE level earthquake was studied using the three-dimensional nonlinear time-history analysis using the Perform-3D software that involves nonlinearity in several types of structural elements as mentioned in Table 2. For the central core wall, energy is dissipated by two critical fuses : primarily via inelastic rotation of the coupling beams and secondarily via flexural yielding of the shear wall vertical reinforcement, thus the two critical inelastic behaviors have been carefully designed and tuned such that design efficiency and compliance to design criteria can be achieved. Figure 12 illustrates the structural fuse that was considered in the design process and Figure 13 shows the typical rebar configuration for the coupling beams. 8

9 Figure 12 - Structural Fuse in Metropolis Phase 1 Residential Tower Figure 13 - Coupling Beam Rebar Configuration 9

10 Figure 14 shows the drift profile of the Phase 1 residential tower for the MCE level earthquake analyses. As the nonlinear computer model is analyzed with a total of 14 ground motion record, the drift limit set for MCE is 3% for the average drift profile of all ground motions and 4.5% for any one individual ground motion. Coupling beam deformation is the most critical criteria that determines the behavior and how efficient the energy dissipation of the building and the maximum rotation occurred near the amenity deck. As indicated in Figure 15, the maximum rotation limit for average coupling beam rotation from the 14 ground motions is 6%. Tensile yielding of wall vertical reinforcement is the second fuse that dissipates energy during a seismic event. Where the tensile strain exceeds two times the yielding strain, special confinement would be required to ensure ductility. With the amenity deck as the major transfer diaphragm, Figure 16 illustrates the drastic increase of tensile strain near that level illustrated the backstay effect that led to tremendous force transfer between lateral resisting systems. Wall shear stress check under MCE, in Figure 17, is another critical behavior that needs to be fine-tuned. Shear failure is often considered brittle and may cause catastrophic results. Thus in the design criteria, wall shear was deemed to be force critical and all wall shear demands were amplified by a 1.5 factor to ensure that shear failure does not occur. Figure 17 shows the average wall shear stress for all piers and for the 14 ground motions. Figure 14 - Phase 1 Residential Tower MCE Drift Figure 15 - Phase 1 Residential Tower MCE Coupling Beam Rotation 10

11 Figure 16 - Phase 1 Residential Tower MCE Wall Tensile Strain Profile 11

12 Figure 17 - Phase 1 Residential Tower MCE Wall Shear Stress Profile 12

13 Phase 2 Towers SLE and MCE level earthquake analysis for the two Phase 2 towers were performed in a similar manner to the Phase 1 Residential Tower. However, with the additional low-rise wing attached to the main tower, the Phase 2 towers show different behavior as it relates to the deformation and stress distribution in the lateral system that in turn resulted in a different design. As shown in Figure 18, the two curves represent the drift profile at opposite corners of the entire building and due to the difference of stiffness in the major and minor shear cores, the building underwent slight torsion behavior, but was still within the acceptable limits. In Figure 19, the major tower drift profile showed a set back at the lower stack roof which possessed the similar trait to the Phase 1 Residential Tower. However, to avoid the stress concentration issue from Phase 1, the major and minor cores in Phase 2 had been fine-tuned so that the backstay effect at the roof of the lower stack is minimzed. Refer to Figure 20 and Figure 21 for the MCE level coupling beam rotation and shear stress profile, respectively. Figure 18 - Phase 2 R3 Tower SLE Drift Profile Figure 19 - Phase 2 R3 Tower MCE Drift Profile 13

14 Figure 20 - Phase 2 R3 Tower MCE Major Core Coupling Beam Rotation Profile Figure 21 - Phase 2 R3 Tower MCE Major Core Wall Shear Profile 14