OUTRIGGER-BRACED SYSTEMS IN TALL BUILDINGS

Vaidas Razgaitis

ABSTRACT
In building design, an outrigger denotes a rigid horizontal structural element used to
improve strength and resistance to overturning. Because they don’t require closely spaced
perimeter columns and spandrel beams, outrigger-braced systems have replaced the framed tube
as the premier lateral system in tall buildings (Choi, 2012). These outriggers, comprised of multiple
story deep trusses or girders, connect a building’s lateral-load resisting core to perimeter columns
to prevent free cantilever rotation (Taranath, 1998). Outriggers can connect to perimeter columns
on mechanical floors, saving the structural system from intruding on rentable floor space. An
outrigger-braced building has less story drift and will have 30 to 40 percent less overturning
moment in the core when compared to a free cantilever (Lame, 2008). In the case of supertall
towers with mega columns designed for drift control, the reduction in core overturning can be up
to 60 percent (Choi, 2012). Outriggers are an efficient structural solution that dramatically
increases architectural possibilities. However, their use ushers in new space-planning
complications and structural challenges associated with differential shortening between core and
perimeter column members (Nair, 1998). This paper will highlight the benefits, challenges, and
solutions associated with outrigger-braced systems in tall buildings.

INTRODUCTION
The function of an outrigger is to tie together core and perimeter structural systems to
achieve building performance that is superior to either individual system. Outrigger systems have
been utilized for centuries in building design for various purposes. Flying buttresses, a staple of
gothic church architecture, were often employed by church architects to stabilize walls against
lateral forces. Examples of outrigger systems can be seen even earlier in non-building structures;
early Polynesian boats used wooden beams linked to the hull of canoe to prevent overturning
(Choi, 2012).
Load Transfer
In a tall building, an outrigger system connects a central core lateral system to perimeter
columns through horizontal trusses or girders. These horizontal elements engage windward
perimeter columns in tension and leeward columns in compression (Figure 1). This coupling

Figure 1: Tension compression couple (Lame, 2008)

action reduces bending moments in the core, leading to reduced story drift (Taranath, 1988). While
gravity columns can typically handle this increased compressive loading, tension capacity should
always be verified. Concrete columns under tension are susceptible to cracking and reduced axial
stiffness, which can lead to unfavorable alternative load paths (Choi, 2012).
Each location where an outrigger connects to the core can be modeled as a rotational spring.
The restoring couple creates an inflection point in the shape of the deflected curve of a building’s
core- this leads to a drop in the core bending moment at each outrigger location (Figure 2).
Additionally, outriggers reduce the internal moment at the base, due to a greater portion of the
lateral load being carried axially by perimeter columns. (Taranath, 1988).

Figure 2: Outrigger force couple’s impact on deflection

and moment (Taranath, 1998)

The magnitude of drift reduction and overturning moment in the core is dependent on
several building properties: the outrigger flexural and shear stiffness, core flexural stiffness,
outrigger locations along the building height, plan dimensions between core-and-outrigger
centroids, and axial stiffness of the outrigger columns (Lame, 2008).
Types of Outriggers
A conventional outrigger is made up of a truss that directly connects perimeter columns to
the central core, with diagonals spanning multiple floors (Figure 3). Core rotation is resisted by
the vertical force couple carried by perimeter columns.

Figure 3: Elevation view of a typical outrigger truss with

“X” diagonals spanning three floors (Nair, 1998)
 An alternative way to achieve the same effect is to create a “virtual” outrigger, which has
no direct connection to the core wall. In place of this core connection, a virtual outrigger uses stiff
floor diaphragms to resist a portion of the core moment. This share of the core moment is resisted
by a horizontal couple (Figure 4) that is carried through the floor diaphragms to a ring-like truss

Figure 4: Floor diaphragm load path in a virtual outrigger (Nair, 1998)

in a vertical plane around the building perimeter, known as a belt truss (Figure 5). A belt truss
evenly distributes lateral load to be carried axially by perimeter columns. (Nair, 1998).

Figure 5: Belt truss load path in a virtual outrigger (Nair, 1998)

Force transfer between floor slab diaphragms and the belt truss can be achieved with shear
studs. Floor diaphragms will typically have to be additionally reinforced, as they carry in plane
shear in addition to typical vertical live and dead loads (Nair, 1998).
When running an analysis of an outrigger-braced structure with virtual outriggers, it is
critically important to use accurate values for the stiffness of the floor slab. While it is
computationally efficient to model floor slab diaphragms as infinitely rigid, doing so will falsely
report zero force members in outrigger truss chords, as well as obscure values for forces required
for force resolution in the floor diaphragm (Choi, 2012).
Building Geometry
Consider the lateral aspect ratio of a building, defined as the ratio between building height
and core width. In a shorter building, an outrigger system would be impractical. The sizes of
outriggers and columns would be much larger than those required for strength considerations
(Choi, 2012). Outriggers are much more efficient at a higher lateral aspect ratio. For this reason,
towers with wider cores typically need to be much taller than towers with slender cores for an
outrigger system to be efficient. An outrigger systems functions most effectively when it is
 symmetrically distributed about a building’s core. Such an orientation provides the largest distance
for a force couple between perimeter columns. This lessens core overturning without contributing
additional axial loads into the core. In an unsymmetrically braced core, force transfers due to
differential shortening can create additional moment in the core from gravity loading alone (Choi,
2012).
Optimizing Outrigger Elevations
A simplified case study (lateral roof load only, uniform core flexural stiffness, very stiff
outriggers and outrigger columns) performed by Thornton Tomasetti demonstrates how changing
outrigger locations influences story drift (Figure 6). The exaggerated deflected building shapes
show improved performance with multiple outrigger levels at maximum vertical spacing. In
choosing outrigger locations, the designer must consider four key issues: the number of outrigger
sets, the type of outriggers used, the stiffness of trusses and columns, and the availability for
outriggers in the architectural building layout (Choi, 2012).

Figure 6: Simplified effects of outrigger location on roof level drift (Choi, 2012) ©Thornton Tomasetti

Each set of outriggers provides an additional rotational restraint and associated drift
reduction. However, each outrigger level brings with it increased erection time, and often
interrupts workflow compared to that of a typical floor. Minimizing outrigger levels can speed up
construction time, but excessively cutting down on outriggers will lead to heavier members that
require specialized erection equipment. A careful cost-benefit analysis is required to arrive at the
most economical solution (Choi, 2012).
The type of outrigger used has a large influence on the building’s stiffness, which can be
seen as a measure of an outrigger system’s efficiency. Because of the less efficient load path,
virtual outriggers are required on multiple floors to achieve the same stiffness provided by a single
direct outrigger level (Choi, 2012). Both types of outriggers can be employed in the same building.
A designer might chose to utilize belt trusses at elevations where differential shortening between
core and perimeter elements is likely.
 In order to function effectively, outrigger trusses and columns must meet strength and
stiffness demands. Given a target displacement value, optimization techniques such as the unit
load method (Wada 1990) can reveal the relative influence of a member on an outrigger’s stiffness
value.
Most tall buildings reserve the top floor for HVAC and mechanical systems, making this a
preferable location for an outrigger level. Given that an outrigger level is designed at the
mechanical penthouse, studies (Gerasimidis, 2009) have shown that the optimum location to place
a second equally stiff outrigger would be at mid-height. If architectural features prohibit this
location, the outrigger stiffness should be adjusted accordingly to achieve the same efficiency
(Choi, 2012). Stiffness of an outrigger can be adjusted by resizing members of the outrigger truss
or by changing the cross sectional areas of columns that the outrigger connects to. Because the
outrigger stiffness depends on the cross section of columns it connects to, the optimum location
for an outrigger elevation is largely dependent on column size reduction with height (Choi, 2102).
Studies by Brian Smith and Alex Coull have shown that for the optimum performance of an n-
outrigger structure, the outriggers should be designed at the 1/(n+1) up to the n/(n+1) height
locations (Smith, 1991). It is important to understand that these general guidelines have been
derived under the assumptions that the structure is linearly elastic, the columns carry only axial
force, the outriggers are rigidly attached to the core and the core is rigidly attached to the
foundation, and the cross-section properties of the core, columns, and outriggers are uniform
throughout their height.
Smith and Coull found that the lowest outrigger level in a structure carries the highest
bending moment, with each outrigger level above it carrying progressively less moment. Their
studies proved that in a structure with optimally placed outrigger levels, an added outrigger at the
top floor carries only a sixth of the moment of the outrigger below, proving that a mechanical
penthouse is not an optimal outrigger location (Smith, 1991).
Space planning that satisfies the needs of architectural, mechanical, and leasing needs
rarely allows outriggers to be designed at their optimal locations. Outriggers typically end up
running through mechanical floors. If the mechanical floor is not a viable option, a “super
diagonal” strategy can be employed. This deep diagonal truss system was implemented in One
Liberty Place in Philadelphia (Figure 7). The diagonals span four floors, meaning that the
obstructed width is less than a fourth of the clear floor span (Choi, 2012)

Figure 7: One Liberty Place “super diagonal” (Choi, 2012) ©Thornton Tomasetti
Differential Shortening
The improved lateral performance offered by outriggers comes with associated detrimental
effects under gravity loading. Consider the common case of a reinforced concrete core, with rolled
steel perimeter columns. The core will shrink due to inelastic creep, while perimeter columns
change length due to elastic shortening and thermal strains. This differential shortening between
core and perimeter elements will induce additional stresses in the outriggers.
Differential shortening between core walls and perimeter columns can be mitigated through
construction sequencing. A typical solution is to delay final outrigger connections to the core until
the structure has topped out (Chung, 2008). Doing so almost completely eliminates the issue of
elastic shortening; once a structure has topped out, up to 95% of elastic shortening has already
occurred (Choi, 2012). However, this strategy can only be employed if the structure’s core is
capable of resisting lateral construction loads singlehandedly.
Certain situations can make connections between the outriggers and perimeter columns
inevitable; a client might require early move-in on lower floors, or a lateral system might rely on
perimeter columns to resist construction-period lateral loads. One possible solution is the Shim
Plate Correction Method, described and illustrated by Hi Sun Choi, Goman Ho, Leonard Joseph,
and Neville Mathias in the Outrigger Design for High-Rise Buildings technical guide. In this
technique, steel plates are inserted between the surfaces of the outriggers and perimeter columns
(Figure 8 a,b). A connection is then made to achieve composite action between the core and
perimeter columns. As shortening differentials develop between core and perimeter members,
shim plates are removed and added as needed (Figure 8 d,e) to maintain the gaps within a specified
range (Choi, 2012). The Shim Plate Method has significant challenges associated with it. Failure
to control the gap can lead to significant stresses developed in members of the outrigger system.
Maintaining the joint gap in the specified range with shim plates is a very difficult process,
requiring extra manpower, measuring, and monitoring devices (Chung, 2008).

Figure 8: Shim Plate Correction Method (Choi, 2012) ©Arup

 Daewoo Engineering and Construction has developed a new technique to engage perimeter
columns for construction-period lateral loads called the Oil Jack Outrigger System. The system
alleviates challenges associated with the Shim Plate Correction Method, and was used successfully
in the North-East Asia Trade Tower in Korea. The arrangement consists of a pair of interlocking
oil jacks connected by a pipe with an orifice. Oil pressure is then used to press the rams of both oil
jacks to pinch the outrigger at bearing points to perimeter columns (Figure 9). During quasi-static
vertical displacements such as those induced by column shortening, oil slowly flows through the
pipes and orifice. Because the oil is in a closed system, the oil will flow between the two pairs of
cylinders during differential movement, meaning that no additional stresses are induced in
members of the outrigger system. In the case of dynamic lateral loading from winds or earthquakes,
oil movement is resisted by the orifice. The oil jacks sustain pressures, eventually transmitting
them to perimeter columns as axial loads (Chung, 2008). While this system requires less
monitoring than required by the Shim Plate Correction Method, it still requires an agenda of
periodic inspection and maintenance, as well as an acceptable level of redundancy to account for
the potential failure of jack sets (Choi, 2012).

Figure 9: Oil Jack Outrigger Joint System (Chung, 2008)

A third proposed method to mitigate differential shortening while engaging outrigger

action is the Cross Connected Jack System. The technique involves a system of two hydraulically
tied oil-filled flat jacks; one on the top face of an outrigger, the other on the bottom face of the
outrigger on the opposite end of the building (Figure 10). This system does not require the orifice

Figure 10: Cross Connected Jack System (Kwok & Vesey, 1997)
characteristic to the Oil Jack Outrigger Joint System. The pressurized hydraulic system resists
stresses from differential shortening. Once construction approaches completion, the jacks are
grouted for a permanent seal (Kwok, 1997).
Because virtual outriggers don’t directly connect to the core, differential shortening
between the central core and perimeter columns does not induce additional stress in the outrigger.
However, differential shortening between adjacent perimeter columns will induce additional stress
in a belt truss linking them together (Choi, 2012). It is rare for columns in a tall building to
experience significant differences in axial loading in adjacent perimeter columns.
Thermally Induced Strains
Structures that contains a mix of interior and exterior exposed members require special
attention. Thermally induced strain differentials can influence member and connection forces,
floor deflections, and local joint behavior. A comprehensive thermal investigation should consider
thermal material properties, realistic heat flow path, and the ratio of exposed to interior surface
areas. ASCE 7-10 references self-strain loads as T, stating that structural effects of T should be
considered in combination with other loads. ASCE 7-10 specifies that the load factor on T should
never fall below 1.0 (ASCE 2010). Because of the extremely low probability of simultaneous
occurrence of extreme temperatures during wind storms or earthquakes, a load factor of 1.0 on T
is acceptable in load combinations involving wind or earthquake loading. For other combinations,
a higher load factor should be used (Joseph, 2012).
The New York Times Building, a 52-story office tower designed by Thornton Tomasetti
(TT) in New York City, features exterior columns exposed to the environment. The building
contains a braced steel core, with two-way outrigger trusses running through mechanical floors on
the 28th and 51st floors (Figure 11).

Figure 11: New York Times Building steel framework

(Scarangello, 2008) ©Thornton Tomasetti
In their thermal investigation, design engineers followed recommendation from the National
Building Code of Canada (NBC). A temperature differential of +70° F to -80° F was chosen based
off of historical daily maximum and minimum values for New York City, modified to reflect
effects from heating and cooling. In their analysis, engineers calculated that an unrestrained 650-
foot steel column supporting the top office floor would expand 3.5 inches for a 70° F temperature
change. To reduce these thermally induced elongations, TT’s engineers designed “thermal
outriggers”, distributing strains of exposed perimeter columns to interior columns, while at the
same time employing perimeter columns to carry a portion of lateral loads. Engineers called for a
moment connections between outer columns and beams to provide adequate strength for gravity
loading and thermal movement (Scarangello, 2008).
CONCLUSION
Outrigger-braced systems are an efficient, relatively non-intrusive technique to reduce
lateral drift in a tall building. Their use requires close collaboration and coordination between the
client, architect, and contractors involved in the project. Space planning, local building practices,
and site specific challenges often prevent outrigger levels from being placed at their optimal
locations. Structural designers are challenged to meet code requirements on drift limits and
occupant comfort within the constraints created by other interest groups. While each project has
its own specific challenges, most outrigger-braced systems will require the engineer to develop a
method to mitigate the effects of differential shortening between lateral core and perimeter
columns.
REFERENCES

American Society of Civil Engineers (ASCE). (2010) ASCE 7-10: Minimum Design Loads for
Buildings and Other Structures. ASCE: Reston, USA.
Chung, K., Scott, D., Kim, D., Ha, I. & Park, K. (2008) “Structural System of North East Asia
Trade Tower in Korea.” Proceedings of CTBUH 8th World Congress. Council on Tall Buildings
and Urban Habitat: Chicago, pp. 425-432.
Choi, H., Ho, G., Joseph, L. & Mathias, N. (2012) Outrigger Design for High-Rise Buildings: An
output of the CTBUH Outrigger Working Group, Council on Tall Buildings and Urban Habitat:
Chicago.
Gerasimidis, S., Efthymiou, E. & Baniotopoulos, C. (2009) “Optimum Outrigger Locations of
High-rise Steel Buildings for Wind Loading.” Proceedings of 5th European-African Conference
on Wind Engineering (EACWE). International Association for Wind Engineering (IAWE): Tokyo.
Joseph, L. & Choi, H. (2012) “Outrigger System Design Considerations” International Journal of
High-Rise Buildings. Council on Tall Buildings and Urban Habitat: Chicago, Volume 1 Number
3
Kwok, M. & Vesey, D. (1997) “Reaching for the Moon - A view on the Future of Tall Buildings.”
Structures in the New Millennium, Proceedings of the Fourth International Kerensky Conference.
A.A. Balkema: Amsterdam, pp. 109-205
Lame, Ali. "Optimization of Outrigger Structures." Thesis. Massachusetts Institute of Technology,
Dept. of Civil and Environmental Engineering, 2008. Web. 5 Nov. 2015.
Nair, R. (1998) “Belt Trusses and Basements as ‘Virtual’ Outriggers for Tall Buildings.”
Engineering Journal, Fourth Quarter, pp. 140-146.
Scarangello, T., Krall, K. &Callow, J. (2008) “A Statement in Steel: The New York Times
Building.” Proceedings of CTBUH 8th World Congress. Council on Tall Buildings and Urban
Habitat: Chicago, pp. 654-659
Smith, Bryan Stafford, and Alex Coull. Tall Building Structures: Analysis and Design. New York:
Wiley, 1991. Print.
Taranath, Bungale S. Steel, Concrete, and Composite Design of Tall Buildings. 2nd ed. New York:
McGraw-Hill, 1998. Print.
Taranath, Bungale S. Structural Analysis and Design of Tall Buildings. New York: McGraw-Hill,
1988. Print.
Wada, A. (1990) “How to Reduce Drift of Buildings.” ATC-15-3 Proceedings of Fourth US-Japan
Workshop on the Improvement of Building Structural Design and Construction Practice. Applied
Technology Council: Redwood City, USA, pp. 349–365.