|
The Dana-Farber Research Tower is located in the Longwood Medical
Area at the northwest corner of Binney Street and Deaconess road.
The tower has 14 above-ground stories devoted to office and research
laboratory uses and five underground parking levels. The structure
occupies the entire site, with the building perimeter as close
to the property line as practical. The bottom floor of the garage
is at about EL. -9 feet or about 52 feet below the original grade
(El. 43 ft). The research tower is abutted by existing structures
on three sides of the site and by Binney Street along the southern
edge (Fig. 1).
Two orthogonal cross-sections (A-A, B-B) through the center of the site are
shown in Figure 2. The stratigraphy of the studied sections is typical of the
subsurface of the site. The soil profiles d include the following layers:
1'-14' of miscellaneous fill, over 13' to 22' of sand, 31' to 57'-thick Boston
Blue Clay with a 8'-10' yellow crust at the top of the layer, glacial till, and
conglomerate bedrock. The bedrock was medium to hard, slightly weathered gray
to purple, coarse-grained conglomerate with closely spaced dipping joints. The
Rock Quality Designation (RQD) ranged from 28 to 40%. The depth to bedrock
varied between approximately 66' and 90'. Based on the information form borings
T1-T6 and from boring EC-17 drilled within the limits of the proposed tower and
the slurry wall installation, the bedrock surface drops from east to west and
from north to south. As discovered by the slurry wall installation the actual
bedrock surface between borings was fairly irregular.
The tied back slurry wall provided temporary support during the excavation of
the basement, but is not part of the permanent structure of the research tower.
Instead it acts as a permanent lateral earth support system and a barrier that
isolates the tower from vibrations (mainly from the adjacent MATEP power plant)
in order to protect delicate (and expensive) medical experiments. The research
tower itself is founded on a series of caissons that bear on the underlying
bedrock (Roxbury Conglomerate).
The slurry walls are supported by six- (6) level of tiebacks to bedrock at an
average inclination of 45° to the horizontal. All the tiebacks extend to
bedrock and have a bonded length of about 20'. On the east wall the first two
levels of tiebacks do not exist because of the existence of the MATEP utility
tunnel. A prestressed concrete edge was used to provide the necessary lateral
support for the east wall, supported at the north and south walls. The design
thickness of the slurry wall is 3' feet, and the wall extends a minimum of 2'
feet in the underlying bedrock. The slurry wall/bedrock friction was designed
to be enough to counteract the lateral forces without the need for the 2' key
that provides added safety.
Slurry wall deflections were very small, with a maximum value reached close to
dH=0.7" (in the south wall (Fig. 3). Most slurry walls were slightly
pulled back for major portions of their length although at some construction
stages small inward movements were recorded. Despite the small wall movements
or even the pulled back walls, the settlements were excessive in the eastern
and western sides. Air drilling caused ground softening and ground losses can
explain the occurrence of small wall movements and large settlements (Fig. 4).
As indicated by Figure 4 the settlement troughs were typical of excavation
projects. Maximum settlements were recorded near the wall and diminished with
increased distance. Settlements up to dV=2.8" were measured in the eastern
side within the MATEP utility tunnel. Inward east wall movements in combination
with the soil losses through the tiebacks contributed to the increased
settlements in the MATEP tunnel. Clearly excessive settlements were induced by
soil and water losses through the anchor heads. On the slurry walls that the
losses through the tiebacks were kept to a minimum the settlements were
acceptable and the damage to the adjacent structures was kept to a minimum.
Adjacent buildings (Jimmy Fund, and Brigham and Women's Hospital buildings)
settled far less than the tunnels and with no recorded damage.
Subsequent analyses by
Konstantakos, Whittle, Scharnet and Regalado et. al (2004)
demonstrated that finite element simulations with inclusion of
soil losses could capture the observed wall and surface deformations
consistently. The authors would like to extend their gratitude
to Dave Shields by GEI who were the consultants of this project
and who gave invaluable insight into this project.
The strain gages placed at the bottom of the slurry wall reinforcement showed
that all of the vertical forces induced on the slurry wall by the tiebacks were
transferred to the base of the wall. This could be expected because the slurry
wall was not allowed to settle and thus no side friction developed between the
slurry wall and the retained soil. Induced moments at the bottom of the slurry
wall became significant only after the excavation had reached the 5th level
(out of 6 levels total) .
|
