Conference PaperPDF Available

Analysis of steel-structure/masonry-wall interaction in historic buildings

Authors:
  • Old Structures Engineering
Conference Paper

Analysis of steel-structure/masonry-wall interaction in historic buildings

Abstract and Figures

Conservation of historic high-rise buildings requires analysis of the peculiarities of early modem construction. The beams, columns, and facades of large steel-frame, masonry-wall buildings experience complex undesigned interactions. Masonry curtain walls of these buildings are usually 300 mm thick, capable of resisting structural loads. American investigation of interaetions between masonry eurtain walls and steel frames began with facade inspection laws intended to find damaged masonry on high-rise buildings. Structural effects found include cracking, out-of-plane displacement, and spalls resulting from thermal stress, sidesway, and rust-jaeking. This paper compares the effect on masonry curtain walls of various structural mechanisms, specifically for high-rise buildings as constructed in the United States between 1900 and 1930. For each mechanism, the stress and strain interaction of steel and masonry elements is considered, rather than analyzing them separately as in modem design practice.
Content may be subject to copyright.
Structural Analysis
of
Historical Constructions -Modena, Lourenço & Roca (eds)
© 2005 Taylor & Francis Group, London, ISBN 04 1536 379 9
Analysis
of
steel-strueture/masonry-wall interaetion in historie buildings
D.
Friedman
Consulting Engineer. New York, USA
ABSTRACT: Conservation ofhistoric high-rise buildings requires analysis
ofthe
peculiarities
of
early modem
construction. The beams, columns, and facades oflarge steel-frame, masonry-wall buildings experience complex
undesigned interactions. Masonry curtain walls
of
these buildings are usually 300 mm thick, capable
of
resisting
struetural loads. American investigation
of
interaetions between masonry eurtain walls and steel frames began
with faeade inspection laws intended
to
find damaged masonry on high-rise buildings. Structural effeets found
include cracking, out-of-plane displacement, and spalls resulting from thermal stress, sidesway, and rust-jaeking.
This paper compares the effect on masonry curtain walls
of
various struetural mechanisms, speeifieally for high-
rise buildings as eonstrueted
in
the United States between 1900 and 1930. For each mechanism, the stress and
strain interaction
of
steel and masonry elements
is
eonsidered, rather than analyzing them separately as
in
modem
design practiee.
fNTRODUCTION
As
the conservation movement has expanded to
include modem buildings, and as the oldest steel-
frame high-rise buildings have passed one hundred
years sinee their construction, engineers have begun to
study technical issues
in
maintaining and repairing his-
torie high-rises. One
ofthe
most diffieult issues
is
the
interaction under load ofheavy masonry curta in walls
and steel frames, since this was neither anticipated
in
the original designs nor
is
common
in
the construe-
tion
of
eurrent buildings. Much
of
the damage that
is
seen
in
these buildings -ineluding craeking, out-of-
plane displacement, and masonry spalling -
is
similar
in
appearanee to that seen
in
traditional masonry eon-
struetion but results from different causes. The causes
can best be explained
by
analyzing these buildings as
examples
of
a unique type.
New
York
City has been a eenter
of
study
of
his-
torie high-rises, both formally within the eonservation
eommunity and empirically
by
engineers and archi-
teets inspecting and designing repairs for facades. The
historie reason for the study
is
the sheer concentra-
tion
of
such buildings
in
the city: in 1929, there were
2479 buildings
in
New York higher than ten stories
out
of
4829
in
the United States. Chicago had the see-
ond greatest total, 449 (Regional Survey, 1931). The
empirical reason for the study
is
the 1980 city law
that requires periodic inspection and repair
of
ali tall-
building faeades. Recognition
of
New York's unique
tall-building history has not neeessarily translated into
conservation and restoration teehniques specific to the
buildings. Simply put, architectural histories, teeh-
nical deseriptions
of
designated landmarks, and ad
hoc conservation knowledge among designers and
contraetors foeus more on the masonry envelopes
of
those buildings than their structural details. At one
extreme, descriptions
oftall
buildings in architectural
guides may omit criticaI details
of
a building, sueh
as
one deseription
of
the 320 m Chrysler Building
that fails to mention that it has a brick-veneer exte-
rior wall, let alone that it has a steel frame (White &
Willensky, 2000). At the other extreme, technicalliter-
ature specifieally written for eonservation typieally
emphasizes materiais coneems over less-known struc-
tural ones. The New
York
Landrnarks Conservaney, a
non-profit advocacy and technieal expertise organiza-
tion, has created a useful guide
to
the repair ofbuilding
facades that emphasizes work on high-rise buildings
(Meadows, 1986). The guide correctly mentions steel
deterioration and the effects
ofwind
pressure and ther-
mal changes
as
causes
of
damage, but
is
aimed at
identifying and repairing materiais damage. Unfortu-
nately, there
is
no
equivalent guide for identifying and
repairing damage caused by interaction between steel
and masonry elements,
or,
as they are often identi-
fied, structure and architecture. The lack
of
technical
sources
is
exacerbated by the visual similarity between
material deterioration and secondary damage from
structural movement. For example, brick faces may
spall because
of
incompatible pointing (a materiais
problem) or beca use
of
high compressive stress from
103
lack
of
expansion joints (a structural problem). Struc-
tural problems such
as
excessive sidesway a
re
typically
not visible
in
themselves, but become manifest
in
damage to non-structural elements.
2 HISTORlCAL AND ARCHITECTURAL
CONTEXT
The earliest steel-framed buildings
in
t
he
United
States, completed before building codes recognized
the reality
of
curtain-wall construction, had very thick
exterior walls, with 600 mm
of
masonry common
in
the 1890s. A trend towards 400 mm walls shortly
before 1900 was finalized
by
the 1
901
New
Yo
rk
City Building Code, which required minimum 300 mm
walls on ali steel-frame buildings. Walls continued to
be
built thicker than minimum
in
order to carry archi-
tectural ornament. Various facing materiais were used,
most often brick and glazed terra cotta.
Steel-frame buildings constructed before the
1901
code vary greatly
in
structural type, and can
be
considered
as
partly experimental. Those constructed
after
1901
and before the 1929 economic crash are
much more structurally uniform, as described below.
Figure
1.
953 Fifth Avenue is the narrow building with a
three window-wide street facade, directly above the truck.
This latter class
of
building
is
high
ly
concentrated
in
New
York
and Chicago, as
no
other city had a large
numberoftall steel-frame buildings constructed
in
that
era, and construction details from the post-World War
II
period are significantly different.
The character
of
large portions
of
Manhattan
is
marked by
ear1y
twentieth century steel-frame build-
ings: the downtown and midtown business districts
contain hundreds of office buildings ofthis type, while
t
he
north-south avenues
of
the uptown residential
neighborhoods are lined with miles
of
mid- and
high-rise apartment houses. Some
of
these build-
ings, such
as
the 240 m Woolworth Building
of
1913,
were planned as architectural monuments and have
remained so since, but most
of
these buildings were
meant as ordinary "background" architecture (Fig.
1).
Individual buildings have been designated
as
protected
landmarks
by
the New York City Landmarks Preserva-
t
io
n Commission. There are also designated landmark
districts containing many
of
these buildings, but the
significant buildings within the districts (such
as
the
Upper East Side district) are often low-rise, traditional-
construction buildings such
as
churches and private
residences. Despite the fact that so many steel-frame
mid- and high-rise buildings
in
Manhattan are not
considered architecturally distinguished, this type
of
building
is
of
great interest because
it
quantitatively
dominates use and repair.
3 BUILDING DESCRIPTION
The analysis
in
this paper concerns the structural effect
of
outside forces on ali
bu
ildings
of
a given type -
steel-frame high-rises built between 1900 and 1930 -
using numerical examples from one such building.
Generalizing from the analysis
of
ohe
bu
ilding to a
class depends greatly on the assumption that the class
has strongly marked properties that can be defined
in
advance. The author's observations during restora-
tion projects and the historical record both confirm
this assumption -the existence
of
what architectural
historian Carol Willis has called "vernaculars
of
cap-
italism," where building codes, street layouts, and
local economic conditions combined with the standard
building technology
of
the day
to
produce "standard-
ize[d] highrise design" (Willis, 1995
).
The description
that follows
is
based largely on the author's observa-
tions, but
is
similar to those
in
traditional archi
te
ctural
and technological histories such
as
ElIiott (1992) and
Condit and Landau (1996).
3.1
Characteristics
oi
the general type
The defining characteristic
ofthe
type
is
the presence
of
a structural-steel skeleton frame designed
to
carry
ali gravity and lateralloads. Under t
he
building codes
in force during the period
of
interest (the first thirty
104
years
of
the twentieth century), the only lateral load
explicitly used
in
the design
ofmulti-story
apartment,
office, and industrial buildings was wind load; under
current codes
in
the United States, wind load usu-
ally governs for steel-frame buildings except
in
the
high-seismic areas (primarily the west coast) where
relatively little high-rise construction took place until
after 1945. Connections were typically riveted except
for beam-to-girder double-angle connections, which
were either riveted
or
bolted.
The most common lateral-Ioad systems were
moment frames with semi-rigid bracket connections,
often top and bottom stiffened angles. Knee braces
or
more complex moment frames were used on slen-
der or unusually tall buildings; full-bay bracing was
used only
in
the tallest and most slender buildings.
The reliance on moment frames
is
in
part an artifact
of
design methods: the use
of
portal- and cantilever-
frame analyses did not provide accurate lateral drift
results that might have encouraged the use
of
stiffer
frames. Matrix-based analysis was impractical with-
out computers and moment-distribution was not yet
available (Cross, 1930).
Several tloor systems were
in
use simultaneously
during the period
of
interest. In 1900, terra-colta tile
arches were the standard method
of
providing a tloor
between beams. By 1930, the most common system
was the draped-mesh slab, often constructed using
cinder-aggregate concrete. Ordinary bar-reinforced
concrete slabs were sometimes used, although they
were rare
in
New
York
during the 1920s and 30s.
Patented reinforced-concrete slabs, such as the Kahn
System, were most common
in
the 19IOs, but appear
throughout the period.
Building facades were solid masonry, consisting
of
a veneer
of
ashlar, terra colta,
or
face brick over
common-brick back-up. The 190 I New
York
Building
t 2
H II><I
H I
Code cleared the way for walls
of
constant 300 mm
thickness by explicitly recognizing the frame struc-
tural support
of
the walls,
in
place
of
the codes based
on masonry structure previously
in
use. The detai I used
for supporting the back-up masonry nearly always con-
sisted
of
the masonry resting at each floor on either
the spandrel beams
or
the slab above the spandrel
beams. The veneer was commonly supported through
mechanical interlock with the back-up (e.g., headers).
Windows were simple rectangular openings with either
loose lintels or, rarely, hung lintels. Masonry piers built
integrally with the walls were typically used to provide
fire-protection to the spandrel columns.
The most important structural aspect
ofthese
build-
ings
is
not obvious during cursory examination. Unlike
modern construction, where great efforts are made to
structurally isolate facades through the use
of
expan-
sion joints and flexible ties, the exterior
of
these
buildings
is
a system
of
masonry and metal elements
in
continuous contact. More specifically, there are
no expansion joints
of
any kind
in
the curtain walls
and the fire-proofing piers tie the columns to the
walls. The presence
of
these piers and the elose con-
tact between masonry, spandrel beams, and floor-slab
edges makes independent movement
ofthe
walls and
frame impossible, and therefore negates a common
design assumption.
3.2 Case study: 953 Fifth Avenue
The apartment house constructed
in
1924 at 953 Fifth
Avenue
in
Manhattan is typical structurally except for
its bar-reinforced concrete slabs. At fourteen stories
(46 m) above grade, it was not particularly tall when
built, however it was built on a single 7.6-meter-wide
lot and
it
therefore has a fairly high slenderness ratio
of
6 (Fig. 2).
In
reality it receives no wind load
in
the
3
I H
I H
Figure
2.
Typical floor framing plan, 953 Fifth Avenue. I
is
a spandrel column
in
a masonry pier, 2
is
lhe typical floor slab,
and 3 is a typical spandrel beam embedded
in
lhe wall.
105
1 2
6
2
4
5
3 2
Figure
3.
Typical spandrel section, 953 Fifth Avenue. I
is
the veneer masonry, 2
is
back-up masonry, 3
is
the spandrel
beam, 4
is
the floor slab, 5
is
the concrete beam encasement
and fire proofing, and 6 is cement floor topping.
short plan direction because
it
is
sheltered on both the
north
an
d sou
th
sides
by
buildings
of
similar height and
age, but it was,
of
course, designed as free-standing.
953 Fifth has always been a high-end apartment
house. There
is
little difference in structure between
luxury and ordinary buildings; the use
of
limestone
veneer on the west (street) faca
de
being the only
noticeable departure from ordinary materiais.
There
is
a light court at the south-e
as
t comer that is
roughly half the east-west length
of
the building and
one-third its north-south width. The basic frame con-
sists
of
seven single-bay, knee-braced frames spanning
north-south, connected to create three multiple-bay
moment frames running east-wes!.
Two
conditions were analyzed: the north facade,
which
is
a wide and windowless lot-line wall, and the
west facade, which
is
a narrow and slender windowed
wall. Both walls are 300 mm ofsolid masonry, with a
ll
-
brick construction on the north and limestone veneer
over brick back-up on the wes!. The colurnn centerlines
are 356 mm back from the wall faces, with the spandrel
beam locations varying to provide 100 mm
of
cover
between the flange tips and the wall faces (Fig. 3).
Commercial finite-element software
(Dr.
Frame ver-
sion 2.0) was used
to
analyze the bare steel frame and
the frame with masonry shear walls, and classical anal-
ysis was used for the walls as vertically-cantilevered
beams.
3.3 Summary 01 ana/ysis
Table I lists the movements associated with various
forms
of
motion
in
the building. Ali loads are the
maximum design loads (fu
ll
dead and live load on
Table I. CaIculated movements under load.
Cause Location
Sidesway North wall
Sidesway North wall
Sidesway North wall
Thermal North wall
Thermal North wall
Compression North wall
Compression North wall
Compression North wall
Compression North wall
Compression North wall
Sidesway Westwall
Sidesway Westwall
Sidesway Westwall
Sidesway Westwall
Thermal West wall
* Horizontal movement.
**
Vertical movement.
Structure
Frame
Frame &
masonry
Masonry
Masonry
Masonry
Column
(Iive + dead)
Column
(dead load)
Masonry
(self-weight)
Masonry
(Iive + dead)
Masonry (dead)
Frame
Frame &
masonry
Masonry
Masonry
(reduced wind)
Masonry
Movement
(mm)
5.6
0.
05
0.5
4.8*
7.4**
15.8
11.4
3.8
9.7
6.6
15.
7
6.1
36.8
6.1
1.3
*
interior floors,
fuI
I wind load on exposed wall area, full
thermal variation) unless otherwise noted. The details
of
each condition are described in the following tex!.
Several pattems are
ap
parent
in
the calculated
movements and described
in
detail below. First, load
sharing between the frame and walls will occur under
initial dead load and under some wind
lo
ads. Second,
the slendemess
of
a given facade
is
a determin-
ing factor
in
load sharing from diffe
re
ntial stiffness.
Third, significant qualitative differences
in
results
exist depending on whether full code loads are
used or reduced loads that reflect more ordinary
circumstances.
4 EFFECTS OF TEMPERATURE CHANGE
New
York
has large temperature swings every year,
with winter average lows
of
- 3°C
to
- 1°C and a
record low
of
-26°C and summer average highs
of
27°C
to
29°C and a record
hi
gh
of
41
°
C.
The standard
temperatures for
HVAC
design are -
11
°C for winter
and
32
°C for summer. Because
of
direct thermal gain
from sunshine on masonry,
it
is
common for masonry
temperatures during summer
da
ys to exceed
38
°C,
therefore a maximum temperature swing
of32
°C was
used for examination
of
faca
de
therm
al
effects. The
masonry wall is assumed to have been built at the mean
106
temperature
of
11
°C, creating equally large thermal
stresses at the minimum and maximum outside tem-
peratures; for the sake ofsimplicity, the effects will be
discussed for expansion only although equal effects
exist for contraction.
The masonry
of
the facade is directly exposed to
both the overall ambient temperature changes and
(depending on orientation) direct heat gain from the
sun. Buildings
of
this type have no dedicated insula-
tion, but rather rely on the insulating properties
ofthe
masonry and the interior plaster. Since the columns
are encased in masonry piers
of
equal exterior and
interior thickness and the spandrel beams are encased
on the exterior with masonry and on the interior with
concrete fire-protection, the steel temperature is Jikely
to
be
roughly half-way between that
of
the interior
and exterior air. This tends to reduce the effects
of
extreme temperatures but this ameJiorative effect is
not considered here.
A wall plane exposed to a large temperature change
will undergo both vertical and horizontal movement.
The effect
of
vertical thermal expansion
is
limited
by
the geometry
of
the connection between wall and
frame: since most
of
the wall is back-up masonry
that
is
vertically bounded by spandrel beams, and
the veneer
is
regular1y tied to the wall by header
bricks, any vertical expansion
of
the wall will be
resisted by the spandrel beams and the forces trans-
mitted to the columns. Vertical motion
is
ultimately
not damaging beca use, pushing upwards from the
foundation, it can simply lift the unrestrained upper
portions
of
the facade and adjacent steel upwards.
At worst, this creates differential movement between
spandrel and interior columns, but this
is
resisted by
the ductile-metal frame.
Horizontal thermal expansion
is
potentially more
serious, because it
is
not resisted by gravity but
it
is
restrained by structure. Unlike the foundations, which
cannot be moved downward by thermal pressure, or the
roof, which can be moved upwards relatively freely, the
columns and walls that restrain horizontal movement
have limited capacities for this undesigned load. There
are various mechanisms that in theory restrain the hor-
izontal thermal expansion:
(I)
shear at the intersecting
wall at each plan corner, (2) friction between the wall
and the floor structure along the wa
ll
length, and
(3) resistance provided by the wall columns
in
bend-
ing from load transmitted through the fire-proofing
piers. The second and third mechanisms depend on
the geometry
of
construction, since the force transfer
from masonry to steel can only take place where the
materiais are in direct contact. The practice
ofbuilding
the spandrel beams and columns into the wall provides
such contact.
Resistance by the columns in bending cannot take
place since the distributed load that is created along the
columns (up to 200 kN/m) creates moments far larger
than the original wind-Ioad designo The presence
of
intermediate masonry piers at each intermediate col-
umn
is
ofno
additional help, since those piers cannot
resist the lateral load by themselves but will simply
transfer it to the same columns. Resistance to the load
through friction with the floor structure
is
resistance
through horizontal shear
in
the walls; analysis
of
the
long north wall shows a required frictionlshear stress
of
130
kPa on each
of
the two 200 mm by 30,500 mm
contact surfaces per floor, which is realistic under the
conditions described. However, on the short west wall,
the required friction is 510 kPa, which
is
greater than
the allowable shear strength for most clay masonry,
and implies a large compressive force (which may not
be present) in order to develop the friction.
Failure ofall
ofthe
resistance mechanisms results
in
outward movement
of
the ends
of
a wall, cracking the
corner masonry on the intersecting walls in line with
the inside face
of
the wall
in
questiono At an ordinary
exterior corner, both walls are moving from roughly
the same temperature change, and cracks develop on
both faces.
5 EFFECTS OF DIFFERENTIAL STIFFNESS
Differences
in
stiffness between adjacent building
elements
is
one
of
the most potentially damaging
effects that can be found. Modern construction con-
tains numerous provisions for movement to prevent
accidental load transfer to relatively stiff elements,
including slip joints
in
curtain-wall mullions, expan-
sion joints
in
masonry curtain walls, and movement
joints in interior partitions. Buildings
of
the stud-
ied type contain continuous masonry curtain walls in
direct contact with the structural frames and interior
terra-cotta- or gypsum-block partitions solidly built
between floor slabs. The relative fragility
ofthe
parti-
tions makes them less likely to carry structural load,
so the focus here
is
on the exterior walls.
Similar to thermal effects, the effects
of
differential
stiffness can be examined for vertical and horizon-
tal movement. As thermal effects are limited to the
wall plane beca use the amount
of
change
in
wall thick-
ness from temperature variation
is
negligible, stiffness
effects are effectively confined to the wall plane by the
magnitude
ofthe
element stiffnesses. Masonry walls,
subjected to out-of-plane forces are far more
fl
exible
than the structural frame to which they are attached
and therefore act, proper1y, as non-structural elements.
Walls subjected to in-plane loading have stiffnesses
of
the same order
of
magnitude as the frames -often
greater than the frames -and therefore have a ten-
dency to carry load. The simplest example is vertical
movement under floor loading. As shown in Table
1,
the shortening
ofthe
frame columns under dead load
is
greater than the shortening that the adjacent masonry
107
walls would undergo
if
they were carrying the load.
This means that the load will
be
distributed between
the wall and the frame
in
proportion
to
their verti-
cal stiffness until a
nd
unless the wall suffers enough
damage to release its load. A 300 mm wall would not
be designed as the structural support for a
14
-story
building because the compressive stresses are toa high;
assuming that the curta
in
wall was built before all the
dead load
is
in
place implies that as the pressure within
the wall increaseel, slip between wall and frame or local
crushing
of
masonry relieved some
of
the load back
to
the frame. As live loads increased during and after
construction, a similar effect would occur.
The more interesting effect
of
differential stiff-
ness occurs with sideway from lateral load. The steel
frame
is
strong enough to carry the loads, but
is
significantly more tlexible than the combined frame-
and-wall. Wind pressure can move a windward wall
(bending
in
its weak direction) and t
he
frame behind
more than
it
moves the parallel walls. The long north
wa
ll
is
far stiffer
by
itself than the frame, while the
short and fenestrated west wall
is
more tlexible than
the bare frame. The walls can only serve to fully stiffen
the frames
as
long
as
they remain uncracked; each
crack reduces the continuity
ofthe
walls and therefore
their stiffness. This
ad
hoc composite structure has
ra
rely been discussed. Stockbridge, in h
is
paper on the
Woolworth Building, lists a number
of
mechanisms
but looks at ali
of
them
in
terms
of
vertical compres-
sion within the wall (Stockbridge,
1981
). He does not
separately discuss the tlexure and shear caused within
the masonry
by
sidesway.
Unlike thermal changes, wind loads are highly
dependent on conditions immediate adjacent to the
building
in
questiono Tall neighbors can block winds,
as
is
true for the north-south wind at 953 Fifth Avenue,
and a grid layout such as exists
in
midtown and upper
Manhattan and
in
the downtown area
of
Chicago pre-
vents full wind load from being applied
to
buildings
near or below the median height by preventing most
winds perpendicular to facades. Exceptional condi-
tions may allow for full wind load: the west face
of
953 Fifth faces the 800-meter-wide open space ofCen-
trai Park. The likelihood
of
sidesway
as
the dominant
mechanism for cracking can be judged
by
compar-
ing cracking patterns
to
directions
of
possible wind
loading.
As Table I shows, the absolute difference between
frame and wall tlexibility
is
greater for the slen-
der west facade than the stocky north facade. More
importantly, the maximum bending stress
in
the north
wall
is
97 kPa, acceptable for most masonry, while
it
is
455 kPa
in
the west wall under reduced load
and 2760 kPa under full code load. In other words,
while the masonry contributes greatly to the stiff-
ness
of
both stocky and slender walls,
it
is
not strong
enough to carry the loads its stiffness attracts
in
the
Figure 4. Multiple-wing
la
yout at Four Park Avenue.
slender walls. This difference applies in most buildings
of
this type since,
in
the era before air-conditioning
and tluorescent lights, buildings with large overall
plan dimensions had light courts that create multiple
slender wings (Fig. 4).
6 EFFECTS OF RUST-JACKING
Unlike intermittent wind pressures and thermal expan-
sion, pressures caused
by
rusting steel are continuous
and (until repairs are made) ever-worsening. The pres-
sure increases from zero
as
rust builds
up
be
hind
the masonry, forcing it outward. The volume
of
rust
compared to base metal varies with, for example,
Gibbs providing values
of
seven
to
twelve times the
original volume (Gibbs, 2000). These large values rep-
resent free expansion. When masonry
is
solidly built
against a rusting piece
of
steel, the volume increase
is
restrained
by
the strength
ofthe
masonry. In technical
terms, rust-jacking
is
the defiection
of
that masonry
under load until failure (typically through cracking)
and the forced movement
ofthe
masonry after failure,
including bulging, spalling, and collapse.
In buildings
of
this type, t
he
structural steel was
protected with red-lead paint
anel,
by
virtue
of
be
ing
embedded within a masonry wall, has never been sub-
jected to mechanical abrasion, impact, or exposure
to attack by any non-water-soluble chemical agents.
Experience has shown that the combination
of
paint
and masonry skin performs well
in
fiat and ordi-
nary sections
of
wall. Where a greater than normal
source
of
water entry into the masonry exists, such
as parapets and the top surfaces
of
ap
plied ornament
including cornices and water-tables, damage
is
likely
108
to occur. Rust-jacking,
in
other words,
is
the struc-
tural equivalent
of
an opportunistic infection. While
the masonry and paint that provide weathering protec-
tion
to
spandrel beams and columns fail over time, the
most serious rusting occurs in those areas where the
masonry fails first, from poor detailing that aids water
entry, improper construction, or most often, cracks
from thermal and lateral movements. This secondary
effect is the most serious form
of
damage
in
high-rise
buildings since
it
is
the only one that
is
progressive
and reduces structural capacity:
if
rusting
of
structural
steel were not a concem, it might be possible to accept
many
ofthe
cracks formed by movement as "naturally-
occurring" expansion joints that posed no real danger
to either people or property.
Because the effects
of
rust-jacking are non-Iinear -
the restraint provided
by
masonry drops from full to
zero afier cracking -and are highly sensitive
to
voids
in
the column fire-proofing piers and in the veneer
outboard
of
the spandrel beams, it
is
not possible
to
analyze this effect directly.
7 COMMON REPAlR TECHNIQUES
Since the passage
of
Local Law I °
in
1980, the
New
York
City Department
of
Buildings has required
inspection
of
ali building facades higher than six sto-
ries every five years (Local Law
10
, 1980). Conditions
identified as dangerous must be repaired shortly afier
the inspection report
is
fi
led (Prior
to
the 1998 revision
of
the
la
w,
a category
of
"precautionary conditions"
existed between "safe" and "dangerous;" conditions
50
noted had to be repaired before the next inspection.
).
A large industry has developed among architects,
engineers, riggers, and ma sons
in
performing repairs.
Given the physical similarity
of
buildings in the type
under consideration, similar weathering exposure, a
general lack
of
maintenance beyond ordinary joint
pointing, and the small community
of
professionals
and contractors involved in "Local Law
10
repairs
,"
standard methods
of
addressing problems developed
quickly
in
the 1980s.
The most common repair performed
is
rebuilding
masonry at externai corners. Cracks on both wa
ll
faces
are common, highly visible, and suggest a section
of
loose masonry.
In
the majority
of
buildings that
have brick
or
terra-cotta veneer, the corner masonry
units are typically removed, the steel painted, water-
proofed, and repaired
as
required, and a new masonry
corner constructed. Similar repairs are made at hor-
izontal strips over spandrel beams where the beams
have rusted (Figs 5,
6).
Expansion joints may be cre-
ated in conjunction with masonry replacement or by
themselves. The removal
of
comer
masonry provides
an opportunity to create expansion joints at both faces
of
the corner with minimal effort and the removal
of
Figure 5. Old beam and
comer
column masonry repairs at
the rear façade 50 West 34th Street.
Figure 6.
Comer
column masonry removed and column
cleaned and painted at Four Park Avenue.
masonry outboard
of
a spandrel beam provides a sim-
ilar opportunity to create horizontal expansion joints
below the leveI
of
wall support.
It may seem obvious to state that any steel exposed
by the removal
of
masonry should be painted and
waterproofed to prevent future deterioration, but this
work
is
not always performed.
lt
is
not uncommon for
workers to open up a ten-year-old repair
to
masonry
and find unpainted, rusting steel. The use
of
self-
adhering rubberized-asphalt sheet waterproofing has
109
made protecting complex steel shapes fiom water rel-
atively simple as long as there
is
enough
of
a void
between the steel and masonry at complex corners
(such
as
the intersection
of
spandrel beams and comer
columns) to allow for some movement
ofthe
sheets.
Steel repairs may be performed with masonry
repairs, ranging from reinforcing plates welded to
flanges and webs that have lost mater
ial
to rust up to
complete replacement
of
columns and beams. Repair
is
favored over replacement
as
it
is
far less disruptive
to the occupied interiors and more safely performed
fiom scaffolding over occupied sidewalks.
8 CONCLUSIONS
Regardless
of
the causes, damage to masonry cur-
tain walls on high-rise buildings must be repaired to
maintain public safety. However, different causes
of
damage may require different repair details.
At
th
is
time, despite the large and growing body
of
experi-
ence
in
New
York
among designers and contractors
with repairs, analysis
is
rarely performed to discover
the causes
of
observed damage. The mechanism
of
rust-jacking
is
well understood, and the initial cracks
are typically described as "thermal movement."
Two
aspects ofthermal expansion are not obvious: first, that
the force developed
in
restrained thermal expansion
is
independent
ofthe
length
ofthe
element; and second,
that simple friction with the floor structure
is
, for long
walls, more
of
a restraint on thermal expansion t
ha
n
that provided
by
intersecting walls and columns. The
result for this type
of
construction, counter-intuitive
for non-engineers,
is
that the longer a wall is, the
less force it exerts from thermal change at comer
intersections with other walls.
Lateralload
is
rarely considered in damage surveys
ofthese buildings, since the steel frames are typically
adequate for ali lateral loads and current design prac-
tice does not typically treat steel and masonry as a
composite structure. Analyzing this effect can give an
upper bound on stresses
in
the masonry
and,
mo
re
importantly, can provide insight into patterns
of
facade
damage, and therefore into repairs required. For exam-
pie, ifthermal expansion
is
believed to be an important
factor,
ve
rtical expansion joints may be cut into large
flat wall planes as well
as
at corners, while
if
sidesway
is
actually the dominant factor, the comer joints are
ali
that is required.
The analys
is
and repairs techniques that have been
developed
in
New
York
(and similar techniques in
Chicago, since the 1996 enactment of a faça
de
inves-
tigation law) are mostly empirical. The advantage to
this approach
is
that only those techniques that pro-
duce acceptable results become commonly used. The
di
sadvantage
is
that a lack
of
expl icit understanding
of
a given problem leads to an iterative repair approach,
which
is
waste
ful
of
time and money. Analysis
of
facades
as
composite steel-and-masonry structures, as
suggested here,
is
one method towards better focused
repairs.
REFERENCES
Condit,
C.
& Landau, S.
B.
1996.
Ri
se
of
lh
e New York
Skyscrape
r,
1865-1913. New Haven:
Yale
University
Press.
Cross, H. 1930. Analysis
of
rigid frames
by
the distribution
of
f
ix
ed end moments. Proceedings of
lh
e
Am
erican Society
o
fCivil
Engineers 56: 919-928.
ElI
iott,
C.
1992. Technics
and
Architecture. Cambridge,
Massaehusetts: The M.I.T. Pres
s.
Gibbs,
P.
2000. Teehnie
al
Adviee Nole 2
0:
Corrosion
in
Ma
sonry Clad Early 20th Cenlury Steel Framed Buildings.
Edinburgh: Historie Seotland.
198
0
Loeallaw
10
/80 in
Th
e City Reeord 108(32169).
Meadows, R. et
aI.
1986. Historie Building
Fa
e
ad
es: A Man-
ual for lnspeetion and Rehabililalion. New
York:
New
York
Landmarks Conservaney.
1931
Regional Survey of New
York,
vol. VI: Buildings:
Th
eir
Uses and
lh
e Spaees
Ab
out Them. reprinted
19
74, New
York:
Amo
Press.
Stoekbridge, J.G. 1981. The interaetion between exterior
walls and building frames in historie tall buildings. In
Lynn
8eedle
(ed.), Developments in Tall Buildings, 1983.
Stroudsburg, Pennsylvania: Hutehinson Ross Publishing
Company.
White,
N.
& Willensky,
E.
2000.
AJA
Guide lo New
Yor
k, 4th
edition. New
York
: Three Rivers Press.
Willis,
C.
1995. Form Follows Finanee. New
York:
Prineeton
Arehiteetural Press.
11
0
... Prior to 1900 and code developments, however, Transitional wall buildings varied more in their systems and detailing and were somewhat experimental. (Friedman, 2005) Transitional walls are characterized by perimeter walls of masonry laid tightly to the steel structural frame with interior steel framing for floors and roof. They formed the effective link between masonry bearing wall structures and later curtain wall or cavity wall construction. ...
... Buildings greater than 1000 feet tall, such as the Empire State Building and the Chrysler Building, in New York City, NY, were constructed and remain beautiful Transitional building examples today. (Friedman, 2005 ...
Article
Prior to the development of framed structures, the primary building system was loadbearing masonry. With the introduction of structural steel framing in the late 1800s, the logical tendency was to incorporate masonry into the building structures. This resulted in several types of construction starting with caged construction and progressing into transitional building construction. In 1940s, engineers and architects began using veneer construction with steel frames. In each construction type, the structural frame and the masonry interacted differently. This paper will explain the progression of the development of structural steel framing and the interaction of masonry walls. It will also describe how that progression was a necessary step leading to the creation of the new system of hybrid masonry and structural steel framing. Each hybrid type will be described.
Analysis of rigid frames by the distribution of fixed end moments
  • H Cross
Cross, H. 1930. Analysis of rigid frames by the distribution of fixed end moments. Proceedings of lhe American Society ofCivil Engineers 56: 919-928.
Technics and Architecture
  • C Eli Iott
ElI iott, C. 1992. Technics and Architecture. Cambridge, Massaehusetts: The M.I.T. Press.
  • P Gibbs
Gibbs, P. 2000. Teehnieal Adviee Nole 20: Corrosion in Masonry Clad Early 20th Cenlury Steel Framed Buildings. Edinburgh: Historie Seotland.
Historie Building Faeades: A Manual f or lnspeetion and Rehabililalion
  • R Meadows
Meadows, R. et aI. 1986. Historie Building Faeades: A Manual f or lnspeetion and Rehabililalion. New York: New York Landmarks Conservaney.
The interaetion between exterior walls and building frames in historie tall buildings
  • J G Stoekbridge
Stoekbridge, J.G. 1981. The interaetion between exterior walls and building frames in historie tall buildings. In Lynn 8eedle (ed.), Developments in Tall Buildings, 1983. Stroudsburg, Pennsylvania: Hutehinson Ross Publishing Company.
Form Follows Finan ee
  • C Willis
Willis, C. 1995. Form Follows Finan ee. New York: Prineeton Arehiteetural Press.