Interview with Bret Lizundia

Interview Minutes:

Date : 10/17/2003

Present :  Bret Lizundia, Project Manager from Rutherford and Chekene,  

                    Structural Engineers for New de Young

                     Frances Yang, Arch 229b Graduate Student, interned at R&C

                   summer of 2003

Location:  Job Site

 

General Project Info:

$135 million construc cost, $200 mill total, difference is soft cost

Project split up into 2 bldgs.

Main bldg:

3-stories

Base isolated¡ªdecided by COFAM from day one.

270,000SF

Steel¡ªcheaper than concrete system

Ordinary Braced Frames above base isolation (elastic design level)

Tower:

9-stories

Fixed base, 6¡¯ thick concrete mat foundation

35,000SF

Concrete

Exception to city code height limit

Doesn¡¯t hold any art, so not under as strict seismic perf criteria

For public and educational program use

Phased in 3 parts because of size of plan

Had to underpin Asian Art Museum b/c was not demoed in time.

 

The Tower:

Bret explained the change in the tower scheme from H&D¡¯s original design.  It used to be a rectangle that twisted by rotating 37 degrees about its vertical axis.  This was changed so that the axis does not rotate.  Instead, the ¡°twist¡± effect comes from the shifting of the north and south walls, in plane, in opposite directions.  Thus, the rectangle becomes a parallelogram.  He illustrated this with a shoebox model, saying that if you imagine pinching the top diagonal corners of the shoebox, it then looks like it is twisting, but is really only shifting.  H&D liked how this change created a more dramatic tilt.  There were actually 4 different tower versions carried through 50% CD, including one that changed from a rectangle to a triangle.

The effect of the twist in an earthquake is accommodated for by PT cables, positioned vertically, from one side of the base of the wall, to the opposite, far side of the top of the wall.  Post-tensioning these cables pulls down the top of the wall opposite the way that it wants to cantilever over.  In this way, the leaning corner is held back by the post-tensioning force and prevents ratcheting of the towers during an earthquake.  These tendons are grouped in bundles of 12 into galvanized ducts, which are greased and coated with plastic, thereby unbonding it (allowing slip) from the concrete.  Five of these ducts will run up diagonally across the walls.

There was a time of debate between F&C, R&C and H&D over what material to use for the tower.  Designs for both materials were carried for a while in DD.  Once Swinerton received CDs, decision was concrete.  Steel would have req¡¯d waterproofing, also not used as much in Europe.  Concrete naturally waterproofs.  Leaning system not addressed in code.  Would have had to use elastic design for gravity moment, could not use ductile design, thus would req much more steel.  H&D ultimately decided they wanted the look of concrete.   Decision to go with concrete separate from decision on configuration of twist.

80¡¯ of glass runs  along long side of tower.  Copper shroud around tower¡ªveil effect.  Copper expected to turn brown in 6-8 months.  Many different  configs for elevator core with one or two elevators, very tight space.  Tower has only 3000SF floor plates, inefficient use of space, but tower design was understood as architectural statement. 

 

Base Isolation:

The seismic design of the new de Young incorporates the most high-performance, expensive, technically advanced lateral force resisting system in current practice. The main building is completely supported by base isolators under every column, which can be seen in an approx 4¡¯ space between the basement level and the ground (ie., earth, not ground level).   The objective was to bring design accels/forces in this bldg down to levels where art can be braced by conventional means, eg. strapping, nailing, etc.  There are two types of devices used in the base isolation scheme.  One isolator is made up of alternating layers of an elastomeric material, ie. rubber, and thin metal plates, about 4ft in diameter.  The other kind is a slider, which is gigantic metal half sphere that sits on a metal plate coated with Teflon, approx 6¡¯ in diameter.  When the building wants to move in an earthquake, these are designed to allow the building to slide and dissipate the energy of the earthquake through friction between the material surfaces.  The higher the load, the more frictional resistance develops in the devices.  In this way, the building is in essence isolated from the ground shaking when an earthquake occurs.  The ¡°separation¡± changes the bldg period from about 0.5 seconds to 3 seconds, which brings down the accelerations, ie. forces the bldg will experience.  During an earthquake, the earth will move back and forth under it, the friction and damping devices will take out the energy, and very little force will travel up to the structural elements.

The building movement allowed by the isolators must be designed for, as it affects everything along the perimeter of the building.  First, since these isolators allow the building to move up to 3¡¯ laterally in each direction, there is a 3 ft. ¡°moat¡± (waterless) around the perimeter of the main building.  The isolation system has also been augmented by dampers installed across the moat, which also dissipate energy as the building wants to move.  Landscaping was designed to allow movement of the concrete slab that covers and prevents people from falling into the moat.  All electrical, plumbing, mechanical, and other utilities that must cross the moat were also designed for the potential 3¡¯ movement.  The respective consultants did this by replacing standard stiff pipes or sheaves that would break with conduits that are flexible, collapsible, and/or stretchable.

Devices: 76 rubber bearings from Japan, 76 sliders from Richmond, 24 viscous Taylor dampers (4 in crawl space, rest across moat) from NY.  No friction pendulum b/c more expensive.  Dampers, 15¡¯ long, underwent rigorous testing.  $1 million worth of mech caps to hide seismic joint at the tower/main bldg interface.

Reqs hanging elevator  shafts, bumpers under elevators and designing for higher forces in case elevator ¡°falls¡±

 

Other:

Catwalks b/c archs didn¡¯t want holes in ceiling.  Serve as platform access to mech shafts instead of through access doors.  $1 million for catwalks.  $100,000 for access doors.

 

Also below are the minutes from last year, which provided a basis for many of my questions:

Meeting Minutes -- November 3rd, 2002

Attendants:

Andrew Sparks - Structural Engineer, Rutherford and Chekene
The Team

Role in De Young Design:
R&C was selected as the structural engineer for the De Young museum by COFAM, in part due to their work on the structural
bracing of the museum in 1989, and partly due to the marketing of Rich Nierowski.
Andrew was brought on to the project 2.5 years ago to work on the tower when the project was a year into schematic design.
He was proficient in concrete design and because of his architectural interests R&C felt he would bring a good sense of
design to the project.
He was the project engineer for the tower and worked under Brett Lizundia.

Leading to Concrete:
The tower was designed as a concrete structure. H&D wanted concrete as a finish material. R&C felt concrete was the best
option to structurally and seismically support the building given its form. With a 40'x80' footprint and a height of 144',
the tower resembles a small high-rise which would lend itself to steel construction, but the rotating floor plates would
make steel construction difficult. Concrete could be created in the twisting forms that the design required which would
be more difficult in steel.

Contact with H&D:
R&C worked directly with Mark Loughton, Thomas Jacobs, and Chris Haas, part of the design team at H&D on both the building and tower
design.

Contact with F&C:
F&C took control at CD Phase. All discussions with H&C and all decisions were coordinated thru F&C. R&C did not find
F&C overly controlling. They at times had difficulty getting information and dimensions from them that were required to
detail their work. R&C felt this was due to the additional communication and coordination required because of the
presence of both a design and CD architect. F&C was required to provide dimensions to R&C but they were being determined
by H&D. F&C appeared frustrated with the time required to get approval.

R&Cs team structure for the De Young Museum:
R&C had two teams. The Tower Team and the Building Team. Brett managed both teams and coordinated communication with H&D,
F&c, and COFAM. The Tower team consisted of Andrew, and towards the end of the project, a Berkeley PHD student who
performed structural analysis. The Building Team was substantially larger, consisting of 2-3 project engineers at any
one time, 5-7 over the course of the project as well as design engineers working under the project engineers. Each group
worked on all phases of their scope, from schematic through detailing and CDs.

Effect of Phasing on R&C:
R&C felt that the effect of the Phasing, which required the structural to be done ahead of the architectural, and which
compacted the schedule, was minimal. Due to the Design Methodology of H&D only the fine tune details were affected. Once
H&D sets the design of the form it doesn't change. This makes design changes have a minimal effect on structural. R&C
could complete the bulk of the work early, and they did not have to revise any major elements which would have a ripple
effect through the smaller details.

Structural Issues:
H&Ds initial design called for a rotating floor plate that changed from the grid of the park to the grid of the city as it
ascended. There were several structural issues that this raised:
1. The design creates a cantilever at two opposite corners of the tower. Each corner wants to fall in the opposite
direction creating a moment couple. The force on the building would screw it into the ground.
2. The rotation in plan also caused concerns about creep in the concrete which would create cracks over time and weaken
the building in the event of an earthquake.
3. The design creates a unidirectional hinge. This is an unbalanced lateral force that occurs in the event of an
earthquake. This is a ratcheting force that increases in strength.
4. The rotating floor plates (which were designed to be identical to simplify the building) generated walls that contained
hyperbolic curves. The form work to build this was prohibitively expensive.

Solutions:
1. R&C suggested keeping the support walls parallel and shifting them along the wall's axis at each floor. While this made
dissimilar floor plates the walls were easier to form. H&D was originally opposed to the idea formally but once a model
was built they approved the change.
2. To compensate for the moment couple and the unidirectional hinge, R&C design a post tension system in the floor plates
instead of using additional concrete to offset the imbalance. The post tension cables curve through the floor plates and
pull the cantilevered ends back towards the center of the building. The major concern with this method was that the cables
would break in an earthquake due to the substantial stress they would undergo, but because of their length the cables can
stretch to take this force.

This scheme was rather ground breaking, and as part of their original fee they had to support that it would be effective.

 

¡¡