Reproduced from PCI Journal, July/August 1996, Vol. 41, no.4

Wiltern Center Parking Structure, Harry H. Edwards Award Winner

An Innovative Design Solution for 

Precast Prestressed Concrete Buildings in High Seismic Zones

Robert E. Englekirk, Ph.D., S.E.

The author describes the development of an imaginative new structural system for precast, prestressed concrete buildings in high seismic zones. The key element in the system is the beam-to-column connection that is comprised of high performance ductile rod connectors. Fully tested, the design concept was applied to a four-story parking structure. The design features of the building are described together with the fabrication and erection highlights of the precast components. With the success of this new structural system, the future prospects of precast, prestressed concrete buildings in high seismic zones look particularly bright.

Until now, the West Coast of the United States has not been favorable to the construction of precast, prestressed concrete structures. To aggravate matters, the 1994 Northridge earthquake induced building officials to tighten codes of practice and thereby make it even more difficult for design professional as well as owners and developers to specify precast concrete products in their projects.

Part of the problem is that in the high seismic zones of the United States, structural systems are governed by the Uniform Building Code (UBC). Because the provisions of the UBC are very prescriptive, the detailing requirements for concrete systems limit the ability of an engineer to develop a precast frame system that does anything other than emulate the physical layout and performance of a comparable cast-in-place structure. Frequently, cast-in-place seismic systems are used to provide lateral support for precast concrete buildings, causing structural design problems as well as coordination and scheduling difficulties at the jobsite.

To alleviate this situation, the Precast/Prestressed Concrete Institute (PCI) together with the National Science Foundation (NSF) have sponsored the Precast Seismic Structural Systems (PRESSS) program. Also, the Precast/Prestressed Concrete Manufacturers Association of California, Inc. (PCMAC) is carrying out a vigorous marketing program in the area.

All these efforts, however, will come to no fruition unless individual design engineers and owners/developers take it upon themselves to develop new structural solutions and then courageously implement them in real world precast-prestressed concrete structures.

This article describes how, in only 4 years, an "idea" for a new structural system was transformed into an actual building (see Fig. 1). The objective was to create a plant cast building that could provide functional and aesthetic freedom and yet survive, virtually without damage, an earthquake of catastrophic magnitude. The design features of the building are described together with the production and erection aspects of the precast concrete components. The construction schedule and cost of the project are analyzed. Lastly, the problems encountered on the job are discussed as well as how a prospective user might effectively exploit the attributes of the system.

  Figure 1

System Objectives  An idea usually evolves from convictions. For more than 20 years the author has been convinced that the post-yield behavior of concrete frame beams could be significantly enhanced if the shear transfer mechanism and the inelastic behavior region were separated. The author has never liked to rely on welding to establish a load path and is convinced that the attainment of optimal construction speed requires the use of dry connectors. Given these beliefs, the objectives of the system that emanated from the idea were first defined and then vigorously pursued by the development team.

Among the most important of these objectives were:

The attainment of these objectives was only possible because many talented people (see Acknowledgement) were challenged by the concept and had the courage to think with absolute freedom, proffering how best a specific objective might be attained. Without perseverance and creativity, the optimal attainment of one objective usually compromises the attainment of other objectives. The end result was that all objectives were attained and the mission was accomplished.

The Challenge  Shortly after the Northridge earthquake occurred, The Ratkovich Company, a Los Angeles based developer, approached the author’s firm with a request to design a parking structure.

The developer had the following requirements:

The Solution  In 1992, Englekirk Systems Development, Inc. initiated development of a connector, based on a completely different concept than traditional emulation design. The connector assembly was produced by Dywidag Systems International (DSI). It is called a Dywidag Ductile Connector© (DDC). The DDC allows precast concrete beams to be bolted to a precast column, simplifying construction while at the same time improving seismic behavior. Like a surge protector for a computer system, this connector contains a "capacitor" that limits the loads to the balance of the system, while maintaining its integrity through very large earthquakes.

The "capacitor" is the ductile rod. This rod is made of a very high quality steel material, with controlled post-elastic properties. The balance of the connector components are designed and manufactured using capacity design procedures so that they remain elastic even as the ductile rod yields. Thus, all post-elastic behavior occurs in the rod itself, protecting the beam and column from any possible damage.

The system was tested in April 1993 at the University of California at San Diego. The DDC system sustained more than 25 cycles of large displacement, including story drifts of 4.5 percent without damage.

A major hurdle was overcome when the International Conference of Building Officials (ICBO) gave formal approval for the new DDC connector assembly to be used. This was a key victory because the new connector maximizes the inherent advantages of precast concrete. Now, there was a precast concrete frame that not only satisfies code requirements, but will perform better than competing materials when subject to very large seismic displacements. Given this general approval, the designers proceeded to develop specific precast frame details.

Structural System  A structural system must provide a load path for both vertical and lateral loads. Functional and aesthetic freedom increasingly demand that the lateral and vertical load paths be combined in a frame. When the lateral load path is likely to be subjected to earthquake-induced loads and deformations, it must also be capable of absorbing earthquake-induced energy. This requisite "toughness" is referred to as ductility, or the ability to bend and yet not break.

The structure pictured in Fig. 1 was built at 3780 Wilshire Boulevard in Los Angeles, California. The plan dimensions of the building are about 191 x 210 ft (64 x 58 m) for a total floor area of 160,000 sq ft (14900 m2). The lateral load path for this building is provided by eight ductile frames evenly distributed throughout the 40,400 sq ft (3753 m2) floor plate.

A typical structural floor plan of the building is shown in Fig. 2. Precast, prestressed girders span 32 ft (9.76 m) and double tees span 60 ft (18.3 m).

Figure 2

The uniqueness of the system stems from the fact that the entire lateral load resisting system can be fabricated offsite and assembled merely by bolting the beams to the columns. Accordingly, it is the DDC connector that creates the uniqueness, because the components were all standard precast concrete members.

The beam-to-column connector hardware is developed from a standard DDC assembly shown in Fig. 3. In this structure, two standard assemblies (see Fig. 4) were placed at each end of each frame beam and the base of each frame column. This doubling of the connector assemblies is not a system requirement, but it was cost effective despite the fact that it required the use of relatively large frame beams and columns. Note that the beam and column sections were 36 x 42 in. and 36 x 36 in. (915 x 1067 mm and 915 x 915 mm), respectively.

                                                                                                        Figure 3

The key item governing behavior is the button head (ductile) rods that are cast in the columns and footings (see Fig. 3), for it is these ductile rods that ensure superior earthquake performance by absorbing earthquake-induced energy with each post-yield cycle. As shown in Fig. 3, the steel transfer block is secured to two high strength Threadbars by anchoring nuts. This subassembly is then cast into the frame beam and column base.

The precast beam and column are connected by 1 ½ in. (38 mm) diameter high strength bolts (ASTM-A490). These bolts transfer tension loads and clamp the transfer block to the heads of the ductile rods. Shear is transferred by friction. Construction dead load shear transfer is activated by the pretensioning of the high strength bolts.

Subsequent shear demands are developed from passively activated compression loads that are flexurally induced; accordingly, one need not be concerned with the quality of the assembling process once the system is in place. This is an especially important feature in earthquake prone areas because intended performance is assured. Note that a detailed development of the load path is contained in Refs. 3 and 4.

The shear-friction transfer mechanism makes it possible to provide erection tolerances. Such tolerances normal to the column face are provided by the oversized bolt holes in the transfer block. Longitudinal tolerances are attained through the inclusion of 1 in. (25.4 mm) of shim pack placed on either the column or beam side of the transfer block.

The initial thinking of the design team envisioned the need for a movable erection corbel for the beam but Spancrete of California, who fabricated and erected the building, felt that this step was unnecessary. Spancrete, a PCI-plant certified producer member, headquartered in Irwindale, California, was selected for this project because they have for many years been known for the high quality of their products and superior service.

Fabrication  The team approach allowed Spancrete to begin fabrication of the precast frames far ahead of the site preparation work. This saved the owner significant costs because the offsite replacement parking did not have to be provided until Turner Construction mobilized onto the site, while giving Spancrete the freedom to match cast the frames one at a time.

The column cages were tied with the ductile rods in place (see Fig. 5). The columns were poured after the ductile rods were bolted to the column form. After they had cured sufficiently to allow them to be moved, the columns were placed on a flat casting bed, with the beam forms between them. Bolts were placed, but not tightened, through the beam connector block and into the ductile rods within the columns. A planned field tolerance of ½ in. (13 mm) between the beam and column was accommodated in the formwork. The beam cage was set in the form, and the beam was cast. The entire process took just 4 to 5 days per three-story frame.

   Figure 5              Figure 6                Figure 7

   Figure 8               Figure 9                  Figure 10

The UBC requires that special moment-resisting concrete frames be inspected by a special inspector (Section 1701.5, ICBO, 1994). This requirement can be met during fabrication by specially trained quality control personnel employed by the producer, resulting in significant savings to the owner for it eliminates the need to hire an independent inspection agency at the plant.

Erection  The erection process used by Spancrete was simple. Precast seismic columns were dropped over temporary guide studs (see Fig. 5) inserted into the ductile rods that were cast in the footing. Shims were set to plumb the column at the appropriate height then the bolts were manually tightened (see Fig. 6).

Once the bounding columns were in place, precast beams were lowered into position (see Fig. 7). Two bolts were then placed at the bottom of the beam and tightened using a calibrated torque wrench (see Figs. 8, 9 and 10) once the beam was properly aligned. The applied torque need only be enough to support the weight of the frame beam.

The total erection time for each frame was between 5 to 8 hours including initial bolt torquing. Most of the time, it was accomplished in less than 6 hours.

After the frame was assembled, the erector checked the final alignment. Adjustments were not required but, had they been required, they could have been easily accomplished because bolting to this point was minimal. With the installation and tightening of the remaining bolts, the full seismic capacity of the frame was attained. It should be mentioned that grouting the gap between the beam and column or hardware accommodating the blockout shown in Fig. 3 is not a structural requirement.

System Economy  The Ratkovich Company, a creative developer, has for years been intrigued with plant fabricated construction for residential and office building applications. They also believe that the most economical structure is produced by a team that includes the contractor. This design-build approach allows for an exploitation of creative structural systems that are not possible given the traditional design/bid/build format.

Accordingly, The Ratkovich Company first retained the design team, which included architect Greg Petroff and Robert Englekirk Consulting Structural Engineers, Inc. Conceptual plans were developed and given to Turner Construction Company who then established a preliminary budget and schedule. Next the design team, in conjunction with Turner Construction, developed reasonable alternative structural schemes, each of which were then priced.

Spancrete of California proposed two precast schemes, each of which appeared to be less expensive than the cast-in-place alternatives. One system proposed a shear wall braced building while the other incorporated the DDC in a precast ductile frame. Conceptual estimating is difficult when an experience base exists for one system but not for the other. Nevertheless, Spancrete and Turner were able to factor in all elements that would affect construction costs. Unfortunately, and all too often, this project cost approach is abandoned in favor of preparing an estimate from a quantity survey and historical data. This method entirely disregards the economies generated by plant precasting as well as reduction in field labor and other trade activities.

Once Spancrete was on board, another design iteration was undertaken as Spancrete’s cost saving suggestions were incorporated into the design. The DDC/Spancrete proposal turned out to be the most cost effective system even including absent time considerations. Turner’s estimate was confirmed during construction.

The functional constraints imposed on this parking structure by the multi-use (market, theater and office buildings) occupancy requirements significantly affected construction costs. Internal and external speed ramps were required to accommodate the various tenant requirements. Nevertheless, the cost model of Schedule 1 (see Table 1) demonstrates how effective the design/build process can be, especially with complex structures.

Work item Cost
Site work $230,000
Foundations 230,000
Precast concrete 1,580,000
Topping slab 460,000
Stairs/towers/miscellaneous iron 300,000
Two glass enclosed elevators 160,000
Electrical 300,000
Plumbing 105,000
Miscellaneous 230,000
Total Cost $3,595,000

Note: Each line item includes the contractor's general conditions, overhead, and profit.           Table 1

The cost per gross square foot of parking including that which is on-grade under the structure was $22.30 per sq ft. The multiple ramping program and other project specific requirements significantly distort unit cost figures. If undistorted unit prices were applied to a more conventional parked ramp, four-level structure, the price of a provided parking space would be on the order of $5000.

Most impressive is the construction schedule. Schedule 2 (see Fig. 11) describes the working weeks required to complete each of the independent construction operations. The erection of the precast concrete system required only 16 working days.

Figure 11

Preconstruction activities can also be accelerated by the design/build process because precast concrete shop drawings can be developed simultaneously with construction documentation as can the precasting process itself.

Design Simplicity and Acceptance  Strength methods are easily applied in the design process. A single assembly (see Fig. 3) is capable of developing a reversible tensile load (Tn) of 282 kips (1254 kN). The distance between the force couple is adjustable and can accommodate any beam depth. The load path on either side of the DDC is designed to sustain DDC loads at an overstrength to ensure their sub-yield behavior. Earthquake deformations and energy absorption are thereby confined to the ductile rods shown in Fig. 3.

This scheme differs from cast-in-place ductile frame construction where these actions must occur in the beam adjacent to the face of the column. The DDC is superior for two reasons:

First, the material from which the rods are made is more ductile than reinforcing bars, and second, buckling of the rods during the compressive load cycle is prevented by the confining concrete in the beam-to-column joint.

The approval and acceptance process for a proposed design is facilitated by the existence of comparative tests and a quality control program, both of which have been accepted by ICBO and the City of Los Angeles, where the building described in Fig. 1 was built. The superiority of the DDC system is best exemplified by comparing Figs. 12 and 13.

   Figure 12               Figure 13

Fig. 12 describes the behavior of a prototypical cast-in-place ductile beam-to-column subassembly, which has been subjected to only three cycles of post-yield deformation at a story drift angle of 3.5 percent. The test was conducted at the National Institute of Standards and Technology (NIST).1 This test is used as a baseline standard for the NSF/PCI/PCMAC funded PRESSS research program whose goal is the development of precast concrete in all areas where seismicity is a consideration.

Compare Fig. 12 to Fig. 13, which describes the condition of two precast beams and a column connected by a DDC assembly after it has been subjected to 16 cycles of post-yield behavior, the last three of which were at drift angles of 3.5 percent. The subassembly test of Fig. 13, privately funded by the author, was performed at the University of California at San Diego (UCSD) under the supervision of Professors Priestly and Seible.4

The test specimen differs from that described in Fig. 3 only in that button head forgings replace a machined rod which was screwed into an anchoring block embedded in the column.4 Further, the minor spalling of the column shell has been mitigated by increasing the size of the threaded head and the inclusion of a larger confining shim plate at the column face.

From a connector reliability perspective, it is important to realize that the bolts in the test program were not pretensioned. The DDC connection system can survive virtually any number of severe earthquakes. If required, it can even be replumbed by manipulating the shim packs. The ductile frame system shown in Fig. 13 is obviously a superior system, especially from a damageability perspective.

Suggestions to Prospective Users  The DDC was developed for the high rise frame building market. Any building that can be braced by a cast-in-place concrete or steel ductile frame is a candidate. A distributed bracing program would also allow the facile inclusion of entirely precast floor systems. Imagine how much time could be saved through the use of an entirely plant cast vertical and lateral load carrying system, especially if the skin is also precast and preglazed. Accordingly, the horizons are unlimited and this bodes well for precast concrete in the next century.

Like every system, the DDC system requires attention to certain details. In the design process, some attention must be paid to how and when the loads are imposed on the connector both in shear and flexure. In the described building, for example, four bolts had to be pretensioned in order to deliver the shear capacity required to support the girder and precast double tees.

Initially, only the bottom bolts were pretensioned so as not to impose flexural loads generated by the erection of the precast component on the upper set of ductile rods. Once the precast components were set, the upper assembly was tightened and the topping placed. Flexural loads imposed on the upper ductile rods by the topping and live load need not be considered in a first yield determination of seismic strength.2, 4

Frame girder-to-column connections should be square. Accordingly, frame girders should not be sloped unless tolerances are carefully checked. Precast concrete erection crews and inspectors are not trained to install or inspect high strength bolts. Preconstruction meetings and dry runs will ensure a smooth first time assemblage. Care must be taken not to overtighten the bolts.

This requires that AISC Specifications be followed and that the torquing wrench be calibrated. Load washers are effectively used to allow for the subsequent verification of the desired level of bolt pretensioning. The erection process on the prototype went smoothly and Spancrete crews were able to assemble an entire frame usually in less than 6 hours.

Need for New Building Systems  American businesses are continually challenged by society to produce better and yet more economical products. Bureaucratic controls seem to thwart reasonable efforts yet they will continue to be a part of the process. Fortunately, the control process is becoming more cognizant of society’s needs and their role in the satisfaction of these needs. Academia is also responding by developing the tools necessary to the confirmation of new products.

In the construction area, this is in the form of laboratories that are capable of testing full or large scale subassemblies and in the development of computers and software packages that predict or confirm the behavior of structural systems. Effectively used, these tools can significantly shorten the idea-to-realization time frame. It is, however, imperative that the development of an idea proceed in a logical manner.

Ideas abound but do not always result in real life structures. To be successful, the path to realization should start with the involvement of a company that is interested in and capable of pursuing the idea. The idea development stage must carefully consider the cost effectiveness of the product, because without cost effectiveness the idea will not be realized.

Component testing and system analysis must be carefully undertaken because large scale testing is an extremely expensive way to refine a design. Testing the system is essential but not easily done. Behavior objectives must be confirmed not only be the loading program but also by the appropriateness of the instrumentation and the documentation.

A protocol for the approval of a system may need to be developed, as it was in this case, to establish a pass/fail criterion. The DDC moved from an idea to a constructed reality in less than 4 years, but the process required the dedicated efforts of a company created specifically to develop systems. The success of this venture can be attributed to the close collaboration between the owner/developer and the design and construction teams.

Completed Building  The Wiltern Center Parking Structure was essentially complete at the end of May 1996. Landscaping and other fixtures are yet to be fully realized. However, the facility is now being fully utilized.

The developer, design team, precaster and contractor were very happy with the completed structure (see Fig. 14). The finished structure satisfied the functional and aesthetic objectives, as well as the cost and schedule constraints.

   Figure 14

The functional constraints imposed on this parking structure by the needs of the diverse users significantly affected construction cost as did the aesthetic amenities which included two glass enclosed elevators. Of the $22.30 per sq ft required to complete the building, structural costs were a very cost effective $14.10 per sq ft. The construction schedule extended over an 18-week work period, from contractor mobilization to final completion. The structural shell, including foundations, accounted for just 7 weeks of this time (see Fig. 11).

Concluding Remarks  PCI and the industry have been preparing diligently for the next century. A program supportive of this action plan is already in progress in California. The successful completion of the Wiltern Center Parking Structure is a major step in solidifying this effort. It should be emphasized that the new precast system is not limited to parking structures but can be applied to many types of high rise buildings.

Tough, inexpensive plant cast buildings are now a reality. Precast concrete buildings no longer need to be viewed as "houses of cards," but can be recognized as the buildings most likely to survive earthquakes of catastrophic proportions without damage. This structural superiority is due to the fact that the DDC system is tougher than comparable cast-in-place concrete and structural steel systems.

New construction methods and systems will be developed in unprecedented numbers as the construction industry addresses the needs of an ever-expanding urban population. Accepted procedures for realizing ideas are now in place and this should allow our construction industry to meet the demands of the 21st Century.

Acknowledgement  The Wiltern Center Parking Structure would not have been successfully completed without the input of numerous individuals. In particular, the author wishes to express his appreciation to Suzanne Dow Nakaki, former president of Englekirk & Nakaki, Inc., for her valuable ideas in the design and execution of the project; Juergen Plaehn, former vice president of Dywidag Systems International, for taking the initiative in the development of the Dywidag Ductile Connector; Nigel Priestley and Frieder Seible, professors of structural engineering, University of California at San Diego, for carrying out the full-scale test assembly program.

Credits  The following companies were responsible for the design and construction of the Wiltern Center Parking Structure:

                           Owner: Wiltern Associates, Los Angeles, California

                           Developer: The Ratkovich Company, Los Angeles, California

                           Architect: Greg Petroff, Los Angeles, California

Engineering of Record: Robert Englekirk Consulting Structural Engineers, Inc. (REI), Los Angeles, California

System Developer: Englekirk & Nakaki, A Systems Development Corporation, Inc. (ENI), and Dywidag Systems International (DSI)

General Contractor: Turner Construction, Los Angeles, California

Precast Concrete Manufacturer: Spancrete of California, Irwindale, California

References

1. Cheok, G. S., and Lew, H. S., "Performance of 1/3-Scale Model Precast Concrete Beam-Column Connections Subjected to Cyclic Inelastic Loads," NIST 4433, National Institute of Standards and Technology, Gaithersburg, MD, October 1990.

2. Paulay, T., and Priestley, M. J. N., Seismic Design of Reinforced Concrete and Masonry Buildings, John Wiley & Sons, Inc., New York, NY, 1992.

3. Nakaki, S. D., Englekirk R. E., and Plaehn, J. L. "Ductile Connectors for a Precast Concrete Frame," PCI JOURNAL, V. 39, No. 5, September-October 1994, pp. 46-59.

4. Englekirk R. E., "The Development and Testing of a Ductile Connector for Assembling Precast Concrete Beams and Columns," PCI JOURNAL, V. 40, No. 2, March-April 1995, pp. 36-51.