View
814
Download
23
Category
Preview:
DESCRIPTION
Series of articles published from Structure Magazine featuring Antiquated Structual Systems. Available elsewhere on line for free.
Citation preview
STRUCTURE magazine September 2007
aid
s for the structural eng
ineer’s toolbox
EN
GIN
EER’
S NO
TEBO
OK
45
Antiquated Structural Systems SeriesBy D. Matthew Stuart, P.E., S.E., F.ASCE, SECB
This article is the first in a series that is intended to provide a
resource of information to structural engineers that they can refer to for pro-jects that involve the repair, restoration or adaptive reuse of older buildings for which no drawings exist. As developable land becomes more dif-
ficult to find, particularly in densely pop-ulated urban cities or suburban areas in which open space cannot be used, owners and developers are increasingly turning to existing facilities to convert into new uses. If no drawings are available for an older building, a structural engineer will typi-cally first turn to industry resources to try and determine the nature and capacity of the existing structural system. Any avail-able information is then used to confirm that the facility meets the current build-ing code requirements, or determine what strengthening or remediation must occur to accommodate the new use intended by the architect or owner.If no information is available, the
structural engineer must resort to either expensive non-destructive testing or ex-ploratory demolition methods to try and ascertain the nature and capacity of the structure. In some cases, it becomes necessary to abandon parts of the build-ing in place and construct independent structures around the existing one in order to support any new imposed loads or uses safely.
The purpose of this series is to compile and disseminate information that will enable structural engineers to share their knowledge of existing structural systems that may no longer be in use but are capable of being adapted or reanalyzed for safe reuse today and in the future.
The Circumferential or S.M.I System of Reinforced
Concrete Flat SlabsThe S.M.I. System of designing re-
inforced concrete flat plate slabs was developed by Edward Smulski, a con-sulting engineer from New York City, prior to the 1920s. The system was unique in that the primary flexural rein-forcement consisted of concentric rings of smooth reinforcing bars supplement-ed with diagonal and orthogonal trussed bars placed between the supporting col-umns and radial hairpin bars located at the columns. The author first encountered this type
of system while evaluating an existing structure in Philadelphia that had at one time been used as an enclosed parking garage, but was being used as an office building in the late 1990s. No draw-ings were available for the structure, but small openings cut in the slab revealed portions of the internal reinforcement and slab thickness to enable an analy-sis of the load carrying capacity of the
framed floors. However, rather than the typical orthogonal reinforcing bars, the exploratory demolition discovered rings of smooth bars. A subsequent investi-gation of the available literature on flat plate construction from the approximate time period during which the structure had been built revealed that the slab was very likely designed and constructed us-ing the S.M.I. System.The concentric rings of the S.M.I.
System are located in the top of the slab directly above the columns (referred to as Unit C in the available literature), and in the bottom of the slab at the mid-span of what we would now call a column strip (Unit A), as well as in the bottom of the slab at the mid-span of what is now referred to as a middle strip, or centered in the bay formed by the column grid (Unit B). There is typically no top reinforcing provided in the middle strip at the intersection with the column strips, as is now required by the latest building codes. The concentric rings of bottom reinforcement overlap at the interface zones of Units A and B, while the top reinforcement above the column typically overlaps the Unit A bottom bars below.The slab is separated into three inde-
pendent sections as a part of the design of the system. These parts include the column head section (Unit C), the slab between the columns (Unit A) and the central portion of the slab (Unit B). The column head is analyzed as if it were a circular cantilever fixed at the column and loaded uniformly around its cir-cumference by reactions transmitted to it by the adjacent surrounding compo-nents. The slab between the columns and the central portion of the slab is analyzed for positive bending moments only.The design of the S.M.I. System is
based on the same flexural theory of reinforced concrete used by all other previous methods of analysis, i.e. bending moments are resisted by internal stress in the concrete, compressive on one side of the neutral axis of the section and tension on the other. The primary difference with the S.M.I. System is that the tensile stresses in the structure are offset by the concentric rings of reinforcing bars, which resist the tendency of the concrete within the ring to deform/elongate due to the tensile bending forces. In other words, the rings were subjected to hoop stresses – axial forces acting on the Section through column head.
UnitC
UnitC
UnitC
UnitCe
UnitCe
UnitCe
Cw CwCw
Unit Be
Unit Ae
Unit T
Unit A
Uni
t
T
Unit A
Uni
t
T
Unit A
Uni
t
T
Unit Tw Unit Tw
Unit T
Unit TUnit T
Unit Ae
Uni
t A
Uni
t A
Uni
t A
w
Uni
t A
Uni
t A
Unit Ae
Unit B Unit B
Unit Be
Uni
t A
w
Uni
t
T
Uni
t
T
Uni
t
T
Note: Heavy lines indicate top steel Light lines indicate bottom steel
Note: North half of Unit Ce same as Unit C
Top
Rin
gsR
adia
lsC
ent
er
Rin
gs
Tru
ss B
ars
Bot
tom
Rin
gs
Sect
ion
thru
Col
umn
Head
208 Smulski on the S-M-I System of Flat-Slab Construction.
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine September 200746
rebar perpendicular to the radial direction of the concrete tension. The rings consist of smooth bars. The ends of the rings are lapped to develop their full strength. The laps of the concentric rings are staggered to avoid adjacent laps from occurring at the same radial location within the designated Unit. Comments by one of the authors of the
4th Edition (1925) of Plain and Reinforced Concrete Volume 1, Sanford Thompson, in-dicates that the S.M.I System required 20 to 24% less reinforcing than comparable two-way and four-way flat slab systems designed during the same historical time period. Comparisons between weights of reinforcing for different two-way and four-way flat slab systems provided in the CRSI publication, Evaluation of Reinforcing Steel Systems in Old Reinforced Concrete Structures, does not list the pounds of steel required in a typical interior panel of the S.M.I. System; however, other information concerning this system is provided in the same document.Professor W. K. Hatt conducted load tests
of the S.M.I. System at Purdue University prior to 1920. The results of these tests ap-peared in the 1918 ACI Journal Proceedings. An “extensometer” developed by Professor Claude Berry of the University of Pennsylva-nia measured stresses within the reinforcing rings. The 41 feet x 36.5 feet, 2x2 bay test frame, with cantilevers on three sides and an upturned spandrel beam on the fourth, was loaded using bricks stacked in such a way to prevent arching action of the masonry units. The center-to-center spacing of the columns
was 16 feet. All columns included a capital. The slab thickness was 52 inches. The test frame was loaded from 150 PSF to 950 PSF until failure occurred.The following working stress formulas are
used to analyze S.M.I. slabs and size the required reinforcement:
(Unit C) Column Head 2Asfs = 6.64(M/jd) Where: M = Bending Moment per 2 of the circumferenceAs = Sum of the cross-section of rings (Based on the assumption that the directions of the bending moments are radial. The circumference of the Unit was typically established as the average of the inflection points for the continuous orthogonal and diagonal moment diagrams between the column spacings.) (Unit A) Between the Columns 2Asfs = 2(M1/jd) Where: M1 = Bending Moment on portion covered by the rings As = Area of one section of rings (Based on the assumption that the principal bending moments act primarily in one direction. Span of Unit was typically established as orthogonal distance between the inflection points of the opposing columns.) (Unit B) Center Portion of Slab Asfs = 2(M2/jd) Where: M2 = Bending Moment acting in the distance equal to the diameter of a ring As1 = Area of one section of the rings (Based on the assumption that the bending moments act diagonally. Span of unit was
typically based on diagonal clear span between the inflection points of the opposing columns.) Additionally, F.E. Turneaure and E.R.
Maurer were researching the principles of circumferential and radial bending moment analysis at the University of Wisconsin in the early 1900s. A discussion of their methods of analysis can be found in The Principles of Reinforced Concrete Construction, 3rd Edition (1919).The available literature that deals directly
with the S.M.I. System indicates that the method of construction was patented by Edward Smulski. However, a cursory search through the U.S. Patent Office indicates that there were only two patents granted to Smulski, one for a cast-in-place counterfort system for retaining, reservoir and dam walls, and one for a two-way, orthogonal reinforced slab system that included encased steel beams.
It is not clear how predominant the use of the S.M.I. System was during the early 1900s and later in the century. The number of such structures that were constructed and the number currently remaining are unknown. In the opinion of the author, it is not likely that this system was used to a large degree or was very popular because of the assumed difficulty associated with properly fabricating and placing perfectly round and concentri-cally positioned bars in overlapping top and bottom layers.▪
Graphics reprinted courtesy of ACI Journal Proceeding, 1918. A Test of the S-M-I System of Flat-Slab by Edward Smulski.
Flat Slab Divided into Simple Parts.
Circular Cantilever
Suspended Slab
Clear Span o
Circular Cantilever
II
IIII
I
I
I
I
III
Span
Line ofinflection
Theoretical Cantilever
Actualcantilever
r r
Diagonal clear span
1
1/5 1
1/5 1
1/5 o3/5 o
1/5 o
226 Smulski on the S-M-I System of Flat-Slab Construction.
D. Matthew Stuart, P.E., S.E., F.ASCE, SECB, is licensed in 19 states and has over 30 years of experience as a structural consulting engineer. He currently works as a Senior Project Manager at the main office of Schoor DePalma Engineers and Consultants, located in New Jersey, and may be contacted at mstuart@schoordepalma.com.
ReferencesConcrete Plain and Reinforced Volume 1, 4th
Edition Taylor, Thompson and Smulski, John Wiley & Sons, Inc., 1925
“A Test of the S-M-I System of Flat-Slab Construction”, Edward Smulski ACI Journal Proceeding, 1918
Principles of Reinforced Concrete Construction, 3rd Edition, John Wiley & Sons, Inc., 1919
Evaluation of Reinforcing Steel Systems in Old Reinforced Concrete Structures, 1st Edition CRSI, 1981
Manual of Structural Design, 3rd Edition, Jack Singleton H.M. Ives & Son’s, 1947
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine December 2007
aid
s fo
r th
e stru
ctur
al e
ngin
eer’s
tool
box
EN
GIN
EER’S
NO
TEBO
OK
18
Antiquated Structural Systems SeriesBy D. Matthew Stuart, P.E., S.E., F.ASCE, SECB
This article is the second in a series that is intended to provide a resource of information to structural engineers to which they can refer for projects that involve the repair, restoration or adaptive reuse of older buildings for which no drawings exist. Part 1 of the series can be found in the September 2007 issue of STRUCTURE® magazine (www.STRUCTUREmag.org/archives).
faces of the tile were also typically scored. The “blocks” were manufactured by a number of different companies, including: National Fireproofing Corporation, Pittsburgh; Henry Maurer & Sons, New York; Whitacre-Greer Fireproofing Co., Waynesboro, Ohio; and Fraser Brick Co., Dallas, Texas.Flat arch tile units typically varied in
depth from 6 to 16 inches. The average dead weight of these units varied from 25 PSF to 58 PSF. Segmental arch tile units were provided with radial sides so that each tile acted as a voussoir component of the arch. Segmental tiles typically came in 6- and 8-inch depths. Both types of arches were constructed on timber formwork platforms, which were used to secure the tiles in place during construction. The formwork was typically suspended from timber “jack” beams spanning between and over the tops of the supporting steel beams.
In a segmental arch, clay tiles are arranged in a shallow profile between adjacent parallel beams, as shown in Figure 1. The steel beams were typically held together with tie rods, which helped to resist the outward thrust imposed by the arch on the steel beams, both temporarily during construction and permanently at an end span. The tie rod is not shown in Figure 1. Solid clay bricks were also used in a similar fashion; however, hollow clay tiles typically offered an assembly that was not as heavy as solid brick.The flat clay tile arch, as shown in
Figure 2, transferred the load between the beams acting as a jack arch with a tapered keystone located at the center of the span. Again, the resulting outward horizontal thrust reaction that occurred at the beams was typically resisted via tie rods that were required both temporarily during construction of interior spans and permanently at end spans.Another type of flat clay tile arch was the
reinforced system shown in Figure 3. For this type of “arch” system, closely spaced internal reinforcing rods were embedded between the tiles near the bottom, which allowed for the entire section to function more as a true flexural member rather than as an arch. This system was
Figure 1: Reprinted from Tile Engineering Handbook of Design.
Figure 2: Reprinted from Tile Engineering Handbook of Design.
The purpose of this series is to compile and disseminate a resource of information to enable structural engineers to share their knowledge of existing structural systems that may no longer be in use, but are capable of being adapted or reanalyzed for safe reuse in the marketplace of today and the future.
Clay Tile Arched Floor Systems
Concrete and steel framed floors constructed in the late 1800s and early 1900s often included hollow clay tile arches, which spanned between beams and girders. The arches were typically covered with a concrete topping, and often had plaster applied directly to the soffit of the exposed tiles. These types of floor systems were often stronger and stiffer than that calculated by the simple conventional methods of analysis used at the time. In addition, the clay tiles served two purposes; transferring loads to the supporting beams and providing fire protection for the structural steel. There are two basic types of clay tile
arched floor systems; segmental and flat. Both systems were constructed using hollow clay tiles of varying sizes and shapes, with internal open cells similar to today’s hollow masonry blocks. The typical web and face shell thickness was ½-inch, and all four sides of the closed
D R A F TCopyright©
D R A F TCopyright©
STRUCTURE magazine December 200719
AD
VERTISEM
ENT – For A
dvertiser Inform
ation, visit w
ww
.STR
UC
TUREm
ag.o
rg
also referred to as the Natco “New York” reinforced flat arch. It served as a precursor to one and two-way tile joist systems, which will be discussed in a future article.A third type of clay tile arch construction
includes the Guastavino timbrel arch, which consists of a series of laminated layers of tile slabs that were laid and bonded together with Portland cement mortar to form solid large-span domes. As this type of construction was not used in conjunction with steel floor framing, it will not be discussed as a part of this article.Standard flat arches can be classified
into two groups: end construction, and combination side and end construction. End construction consisted of laying the axis of the tiles’ hollow cells parallel to the direction of the span, except at the center keystone tile, as shown in Figure 2. The combination side and end construction method placed the axis of the tiles’ hollow cells perpendicular to the span of the arch (parallel to the supporting beam), for the majority of the blocks used in any one row. In both cases it was normal for the depth of the tiles, in combination with the concrete topping, to be approximately the same depth as the supporting beam. This method of construction assured that the beam was completely braced for the full depth of the steel section, and also made it easier to install soffit tiles beneath the beam bottom flange for fire protection.The tie rods used to resist the arch thrust
forces were generally placed approximately 3 inches from the bottom of the beams in
flat arches. In the case of segmental arches, placing the tie rod near the bottom of the beams resulted in the tie rod being exposed across the horizontal spring line of the arch. If fire resistance of the tie rod was required, it was more often than
not placed higher up from the bottom of the beam as required to avoid exposure. Typically tie rods were ¾-inch in diameter and were spaced as required to resist the specific thrust of the given arch span, although a minimum spacing of fifteen times the width or eight times the depth of the supporting steel beam was recommended.Tie rods at an end span were required,
as there was no opposing thrust present at the outside face of the spandrel beam. At interior spans, with adjacent arches present on either side, tie rods were only required during construction, but were typically left permanently in place. For this reason, when modifying an existing building constructed with clay tile arches that involves the removal of an interior span, the capacity of the remaining adjacent span’s rods should be verified to assure that the end span conditions created on either side of the new opening will remain stable.The total arch thrust, net area of the tie
rods and maximum spacing for both a flat and segmental arch can be found as indicated on the following page.
Figure 3: Reprinted from Tile Engineering Handbook of Design.
SAS Stressteel
salutes its partners
Hayward Baker and
Heller & Johnsen
for their assistance
in another successful
project application.
high capacity
MiNi caiSSONSAs seen in use on arecent NYC project.
Felix Ferrer, P.E., President
100 New Dutch LaneFairield, NJ 07004973-244-5995
Rod Diameter 5/8 inch 3/4 inch 7/8 inch 1 inch
Net Area (a) .202 .302 .420 .550
continued on next pageTable 1D R A F TCopyrig
ht©
D R A F TCopyright©
STRUCTURE magazine December 200720
D. Matthew Stuart, P.E., S.E., F.ASCE, SECB is licensed in 20 states and has over 30 years experience as a structural consulting engineer. He currently works as a Senior Project Manager at the main office of CMX located in New Jersey and may be contacted at mstuart@CMXEngineering.com.
The online version of this article contains references. Please visit
www.STRUCTUREmag.org.
Future installments of the archaic structural systems series will cover one- and two-way clay tile and unit masonry joist systems; prefabricated clay tile and concrete block framing systems; precast concrete framing systems; antiquated post-tensioning systems; and outdated structural steel stub-girder construction. If there are other topics along these lines that you would like to see addressed, please send your suggestions and any relevant information that you have to the author (mstuart@cmxengineering.com).
®
Total Thrust (in pounds) per Arch Panel:P = (3wD2/2R)LWhere; w = uniform dead + live load on
arch in PSF D = arch span in feet R = effective rise of arch in inches (typically 2.4 inches less than
the depth of the clay tile units for flat arches)
L = length of the floor beam sup-porting the arch in feet
Total net area of tie rods per panel (square inches):A = P/fWhere; f = allowable unit stress (typically
18,000 psi)Maximum spacing of tie rods (feet):
T = (af )/ (3wD2/2R)Where; a = net area (square inches) of tie
rod (see Table 1, page 20)
Load tables published in Kidder and Parker’s Architects’ and Builders’ Handbook for segmental arch construction list spans up to 21 feet, with load-carrying capacities of up to 485 pounds per linear foot. Table 10-32 from the Principals of Tile Engineering Handbook of Design lists spans up to only 10 feet 6 inches.Both of the tables from the Principals of
Tile Engineering Handbook of Design were based on load tests, which were reduced by a substantial safety factor of 7. When evaluating existing clay tile arch systems, it is recommended that the initial load capacity rating be based on published tables. If, however, this simplified approach indicates that the allowable load carrying capacity is not sufficient for the new reuse requirements, then it is possible to calculate an increased strength by taking advantage of the inherent composite capabilities of the clay tiles and the concrete topping, as well as the concrete topping and the steel beam. If this enhanced strength approach does not produce favorable results, then it may be necessary to conduct a load test of the in situ system per ASTM E196, Standard Practice for Gravity Load Testing of Floors and Flat Roofs.
AD
VER
TISE
MEN
T –
For A
dve
rtise
r In
form
atio
n, v
isit
ww
w.S
TRU
CTU
REm
ag.o
rg
The principal disadvantage of tile arch floor construction was the difficultly of adapting standard sizes to irregularly shaped spaces. In addition, tile arches are more easily weakened by holes and penetrations than a monolithic floor system. Furthermore, it was difficult to place mortar in end construction, i.e. when the open cells were placed end to end. Also, for end construction, if a single tile was removed in a row, then the remaining tiles became unsupported unless the scored sides of the tile were mortared in with the adjacent rows of tiles. Arches in which the scored sides of the
tiles were placed adjacent to one another, transverse to the arch span, were more conducive to placing mortar between the tiles. This type of construction (side-constructed arches) had an advantage over end construction. However, tests conducted during the period of time in which clay tiles were used extensively indicated that tiles were much stronger in an end construction application as opposed to a side construction configuration.Finally, tile arch construction was
susceptible to poor workmanship because the quality of the work could only be observed from the top and not from below during construction until after the formwork was removed.▪
Flat arch spans typically varied from 3 feet to 10 feet, and were capable of supporting safe uniform loads between 126 and 1,400 pounds per square foot as indicated in Table 10-31 from the Principals of Tile Engineering Handbook of Design. D R A F TCopyrig
ht©
D R A F TCopyright©
STRUCTURE magazine March 2008
aid
s for the structural eng
ineer’s toolbox
EN
GIN
EER’
S NO
TEBO
OK
45
Antiquated Structural Systems SeriesPart 3
By D. Matthew Stuart, P.E., S.E., F.ASCE, SECB
This article is the third in a series that is intended to provide a resource of information to structural engineers for projects that involve the repair, restoration or adaptive reuse of older buildings for which no drawings exist. Part 2 of the series can be found in the December 2007 issue of STRUCTURE® magazine (www.STRUCTUREmag.org/archives).The purpose of this series is to compile and disseminate a resource of information
to enable structural engineers to share their knowledge of existing structural systems that may no longer be in use, but are capable of being adapted or reanalyzed for safe reuse in the marketplace of today and the future.
exception of the Natcoflor system, joist widths typically varied from 4 inches to 6 inches. Typically, ¾-inch clear cover was
provided between the square or round deformed reinforcing bars and the adja-cent tile or masonry units or the top and bottom of the exposed concrete surface of the joist. It was typical to use straight bottom bars and trussed top bars bent down to align with the bottom bars near the center of the span. When a concrete topping was used, it was typical for tem-perature/shrinkage reinforcement to be provided orthogonal to the joist span. The amount of this steel was typically 0.0025 times the gross cross-sectional area of the topping, and it was spaced at no more than 18 inches on center.One-way systems were very efficient for
spans over 12 feet, and were used very frequently for spans up to 24 feet with
loadings that ranged from 40 to 125 PSF, and up to 18- and 20 foot spans for heavier loadings. For two-way systems, and at the end of the span for one-way systems, it was common for the open webbed ends of clay tiles (or masonry units) to be filled with cardboard or metal inserts to prevent concrete from flowing into the voids, in order to minimize the dead load of the slab.The Natcoflor system used specially
manufactured clay tiles with curved flanges that allowed only the bottom of the tiles to be exposed as the ceiling soffit. Other one- and two-way clay tile systems could be formed and cast either with the bottom of the concrete joist exposed or with tile soffit pieces along the bottom of the trenches that resulted in a uniform tile ceiling soffit. The Natcoflor joists were no more than 2 inches in width, spaced at 13 inches on center, with a depth that varied from 4 inches to 12 inches (Figure 1). The joists were typically cast using cement grout consisting of one part cement and two and one-half parts sand. A composite concrete topping was not required above the tiles, in order to attain the maximum load-carrying capacity of the system. The Schuster two-way system (Figure
2), which was patented in 1915, used clay tiles that were 12 inches x 12 inches, or 16 inches x 16 inches, and had depths of 4, 6, 8, 10 or 12 inches. The joists were typically spaced at 16 inches on center or 20 inches on center; however,
One- and Two-Way Clay Tile and Unit Masonry Joist
SystemsIn one- and two-way clay tile and unit
masonry joist systems, individual units were laid in such a way as to form trenches that allowed reinforcing bars to be placed in the bottom of the resulting joist cross sections. This method of construction is very similar to the more recent pan joist system; however, unlike steel pans, the clay and masonry units were left in place for added strength and fire resistance, and to provide a flat ceiling surface.Proprietary one-way floor systems
included the Natcoflor and Republic Slagblock systems. Proprietary two-way floor systems included the Schuster, Smooth-Ceiling, Sandberg and Republic Slagblock systems. All of these employed regularly shaped units of varying size and depth that resulted in a uniform modulation of joist sizes and spacings. However, during the 1930s, a patented “wide-center” system was introduced for both one-way and two-way framing that allowed for wider clay tile units to be placed at the center of the span and narrower units to be placed at the end of the span. This resulted in wider joists near the supports, which in turn resulted in greater shear capacity at the end of the span, similar to the more recent tapered end pan joist system.With the exception of the Smooth-
Ceiling and Sandberg systems, the clay tile and unit masonry could be constructed to span between steel beams, concrete beams or loadbearing walls. In addition, most of the systems could be placed with or without a concrete topping. When a monolithic concrete topping was used, the thickness typically varied from 1½ inches to 3 inches. Joists were typically analyzed as T-beam sections when a monolithic topping was used. With the
Figure 1.
Wood Sleepers
GroundPlaster
Cinder Concrete Fill
Flooring
Finished Top
Flashing
Base
Wood Forms
D R A F TCopyright©
D R A F TCopyright©
STRUCTURE magazine March 200846
tiles could be doubled up to allow for joist spacings of 28 or 30 inches on center. This two-way system was typically used in square bays or rectangular bays in which the longer span was not more than 50% greater than the shorter span.The Republic Slagblok system could be in-
stalled in either a one-way or two-way config-uration. The Slagblok unit measured 8 inches x 16 inches, and came with one open end and one closed end. Each unit was placed in combination with another Slagblok to form closed cells that were 16 inches x 16 inches. Slagbloks came in 3-, 4½-, 6-, 7- and 8-inch depths. The concrete ribs or joists were typical-ly 4 inches in width and spaced at 20 inches on center. Typical spans for this system var-
ied from 15 to 25 feet. The author has seen similar one-way joist systems constructed as recently as the 1970s using regular concrete masonry units.The Smooth-Ceiling
system, which was pat-ented in the 1930s, and the similar Sandberg system both eliminated the need for beams or drop panels by employ-ing embedded internal
steel shear reinforcement around either struc-tural steel or reinforced concrete columns. As with other two-way systems, standard tile units were placed in a modular layout in order to establish a uniform two-way grid of con-crete joists. Typically, both systems eliminated all tiles from around the column to enable this area to be cast as solid concrete.Although load tables, which included con-
siderable factors of safety, were provided by the manufacturers of most of the above systems, the actual design of the joists was accomplished using conventional working stress methods of analysis that were available at the time. Moment and shear coefficients were typically employed to establish the maximum positive and negative moment
and end shear design envelopes; however, continuous beam analysis was also used to establish the required design parameters. Moment and shear coefficients were also used for two-way analysis.Even though load tables and methods of
analysis are available for all of the above clay tile and unit masonry systems, when one encounters any of these same systems in an existing building, and there are no original drawings available, it is difficult to determine what the internal reinforcement is, and subsequently the load-carrying capacity of the system. However, hopefully this article, by identifying the many different types of systems that were in use at one time or other, will assist readers in their investigation of the structural framing when any of the above-described systems are encountered in an existing building.▪
The online version of this article, www.STRUCTUREmag.org, contains
detailed references.
D. Matthew Stuart, P.E., S.E., F.ASCE, SECB is licensed in 20 states and has over 30 years experience as a structural consulting engineer. He currently works as a Senior Project Manager at the main office of CMX located in New Jersey and may be contacted at mstuart@CMXEngineering.com.
Figure 2.
®
ADVERTISEMENT – For Advertiser Information, visit www.STRUCTUREmag.org
D R A F TCopyright©
D R A F TCopyright©
STRUCTURE magazine June 2008
aid
s for the structural eng
ineer’s toolbox
EN
GIN
EER’
S NO
TEBO
OK
Perspective view of standard “Precast Joistile” floor
system and typical sectional detail.
Concrete Topping
Precast tile joists
123 “ 1”, 12“ or 2”
4", 5" or 6"
4", 5" or 6"
12”12”
Concrete Topping
Filler TilePrecast ‘T’ beams
(a) exposed
beam ceiling
(b) flat ceilingspan tile
spantile
49
Antiquated Structural Systems SeriesPart 4By D. Matthew Stuart, P.E., S.E., F.ASCE, SECB
This article is the fourth in a series that is intended to provide a resource of information to structural engineers that they can refer to for projects that involve the repair, restoration or adaptive reuse of older buildings for which no drawings exist.The purpose of this series is to compile and disseminate a resource of information
to enable structural engineers to share their knowledge of existing structural systems that may no longer be in use but are capable of being adapted or reanalyzed for safe reuse in the marketplace of today and the future.
units, which will be addressed in a subsequent article.The prefabricated clay tile systems
included both one-way beam and slab construction and one-way slab con-struction. The one-way beam system involved the placement of prefabricat-ed beams spaced parallel to each other at regular intervals between already constructed load-bearing walls or steel beams. The areas between the beams were then infilled with tiles that were capable of spanning between each ad-jacent beam. The one-way slab system involved prefabricated slab units that were placed directly adjacent to each other, spanning between previously constructed load-bearing walls, steel beams or joists. Both the beam and slab systems included a site-cast con-crete topping, which was poured over the beams and filler tiles or one-way slabs. The prefabricated beam and slab systems offered the advantage of not having to construct supporting formwork before the framing could be erected; however, shoring in the center of the span was sometimes employed to increase the clear span capability of the members through composite ac-tion with the site-cast topping.Examples of the clay tile systems (Fig-
ures 1-6) included the “T” Beam Floor, the “U” Beam System, the Joistile Sys-
tem, the Sheffield Floor System, the Adel Joistile System, the Kalex Floor System, the United Floor System and the Tilecrete Floor System. Some of these systems required filling the joints between the adjacent ends of the clay tile units with mortar, while other systems allowed the ends of adjacent tiles to butt up against each other. All of the prefabricated beam and slab systems, except for the infill tiles, included internal
Prefabricated Clay Tile & Concrete Block Framing Systems
As discussed in the last article in this series, one and two-way joist framing systems were constructed using clay tile and masonry units, which were first ar-ranged and supported on formwork to enable placement of internal reinforce-ment and infill and topping concrete in situ. In addition, as will be discussed in this article, similar modular clay tile and masonry units were also constructed offsite into prefabricated beams and slabs that could be delivered to the job site. This method of construction ulti-mately progressed to solid precast concrete
Concrete Topping
Section A-A Section B-B
Filler TilePrecast Economy
‘U’ beam
Perspective view of “Precast ‘U’-Beam Floor” and sectional detailsshowing optional erection methods with span tile webs up or down as desired.
282" 282"
92
"
6"
82
"
6"
12"2"
2"2"
212" 212"
Concrete Topping
Precast tile sections
Perspective view of “Sheffield” precast tile floor system
and sectional detail.
83" 83"
72
"
Concrete Topping
Precast tile �oor slab
Typical slab section
Perspective view of “Adel” precast joistile floor system and typical slab section.
1”, 1½” or 2” concrete slab
Concrete Topping
Precast Kalex slab section
(a) 4-in. Kalex floor section.
Perspective view of precast “Kalex” tile floor system and typical floor sections.
(b) 6-in. Kalex floor section.
Figure 1.
Figure 2.
Figure 3.
Figure 4.
Figure 5.
Figure 6.
longitudinal flexural reinforcement for positive moment resistance. Negative moment reinforcement for continuity across a supporting wall, beam or joist
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine June 200850
was also sometimes placed in the site-cast topping. None of the units included shear reinforcement. Table 1 summarizes all of the clay tile systems mentioned above.Similar prefabricated beam and slab systems
were also developed from modular concrete block. Most of these systems used conven-tional internal reinforcement for flexural strength; however, a few were developed using prestressed bars and strands. Probably the most widely used masonry block product in the east-ern U.S. during the 1950s was the Dox Plank system developed by NABCO in Washington, DC (Figure 7). This product was manufac-tured with recessed slots in the bottom of the
Prefabricated Clay Tile One-Way Beam and Slab Systems
System Figure Regional UseMortared
JointsBeam Spacing or Slab Width
Beam Depth (including topping) x
Width or Slab Depth
Typical Span Notes
“T” Beam 1 Southwest, Midwest Yes18” to 30” 92” x 8”
24-ft.1, 3
182“ to 222” 72“ to 8” x 8” 2, 3
“U” Beam 2 Texas, Oklahoma No 282“82” to 12” x
7” to 8”14-ft. 4, 5
Joistile 3 East, Southwest Yes 12”4” and 5” 8-ft.
46” 12-ft.
Sheffield 4 Midwest, North Central Yes 8” 5”18-ft. to
22-ft.6
Adel Joistile 5 Midwest Unknown 12” 5” Unknown 4
Kalex 6Ohio, Pennsylvania, New York, Illinois, Wisconsin
No 12” 4” and 6” Unknown7, 8, 9,
10
United None New York, New Jersey No Unknown Unknown 30” 11, 12
Tilecrete None Missouri No 16” 4” and 6” 12” 13, 14
NOTES:1. Included the use of unreinforced, 4” thick Filler Span Tile for one-way span between beams.2. Drop-in Filler Tile was used for flush ceiling applications.3. System load tested at Iowa State College; Engineering Experiment Bulletin No. 286.4. Longer spans were possible with the use of center span shoring.5. Included the use of unreinforced, 2” thick ribbed Filler Tile for one-way space between beams.6. Patented June 1936 by Professor Walter M. Dunagan, Iowa State College.7. Patented 1937 by D.D. Whitacre, Waynesburg, Ohio.8. System also used as vertical wall element.9. Reinforcement included bolted rods, which implies an applied pretensioning force.
10. Load tests of 4” slabs conducted by Professor George E. Large, Ohio State University, for the Rochester, NY Building Board in 1939.11. Unreinforced slab system used in conjunction with open web steel joists.12. System used in conjunction with a topping slab that provided composite action with steel joists.13. Patented system used in conjunction with open web steel trusses; however, tiles were supported on bottom chord, which allowed a
concrete topping to be placed that encapsulated the trusses, resulting in concrete ribs capable of spanning up to 24 feet.14. Tested in 1939 at the National Bureau of Standards; BMS Report No. 16.
Table 1.
block to allow for the flexural reinforcement to be grouted into the bottom of the plank. There was no mortar required between the adjacent ends of each block.Other cross-sectional variations of the Dox
Plank were developed by members of the Dox Plank Manufacturers Association. Figure 8 shows an example of an alternate block that differed from that originally developed by NABCO. In this case, the internal
Steel reinforcing rods provide structural strength.
Cement topping increases strength
Any type floorcovering can be used
2 11/16” 8 3/16”
1/4” R
3/4” R
15 15/16”
1 9/16” 1 9/16”
3 1/2
4”
2”
1”
1”
3/16”
3/1
6”
13/16”
3 7/16”11
/4”
1 5
/8”
MIN
11
/4”
11
/2”
13
/8”
3/16”
11/16”
3/1
6”
15
/16
”1
5/1
6”
13
/8”
1 15/16”
5/8
30º
15º
15º
TO
PP
ING
AS
RE
Q’D
reinforcement was completely encapsulated by the block by means of a continuous sleeve. It is not clear whether the continuous reinforcement in the sleeves was grouted in place, or if the bars were threaded at each end of the plank so that the modules could be precompressed together via tensioning of the bar as it was tightened against each end of the member using a nut.
Figure 7.
Figure 8.
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine June 200851
D. Matthew Stuart, P.E., S.E., F. ASCE, SECB is licensed in 20 states and has over 30 years experience as a structural consulting engineer. He currently works as a Senior Project Manager at the main office of CMX located in New Jersey and may be contacted at mstuart@CMXEngineering.com.
Future installments of the archaic structural systems series will cover precast concrete framing systems; antiquated post-tensioning systems; structural steel stub-girder; open web steel joist and cast iron construction. If there are other topics along these lines that you would like to see ad-dressed, please send your suggestions and any relevant information that you have to the author.
Flexicore, a product similar to the NABCO Dox Plank, was also available in the 1950s. A hollow core plank is still manufactured today under this same name; however, the current product is a true precast, prestressed concrete member.In the 1950s, the consulting firm of Bryan
and Dozier and the Nashville Breeko Block Company designed and constructed pre-fabricated, post-tensioned concrete block beams. This method of construction re-sulted in the first linear prestressed structure to be built in the US – the Fayetteville Tennessee High School Stadium – and the first prestressed bridge to be built in the US at Madison County, Tennessee. This method of construction was made practical and economical by the Roebling Company through the development of high-quality tendons that could be bonded without ex-pensive end anchorages. The Breeko Block system was further re-
fined through the use of external, deflected tendons. However, by the late 1950s, this system was replaced by precast, pretensioned concrete members.Other, more obscure examples of prefab-
ricated modular concrete block beams and slabs include a prestressed bar system devel-oped by P.H. Jackson of California in 1872; a prestressed wire system developed by C.W. Doering in 1888; a system patented by K.E.W. Jagdmann in 1919; a horizontally draped stressed reinforcement system pat-ented by Albert Stewing and Stefan Polonyi in 1967; and a tensioned, Y-shaped block system patented by Hossein Azimi in 1987.All of the above tile and concrete block
systems were designed based on the basic reinforced masonry and concrete beam analysis theories of their era. Load tables were also commonly developed and pub-lished by most of the manufacturers. The problem with all of the above systems, when one encounters them in a building, is that in the absence of existing drawings it is difficult to determine the internal reinforcement and, subsequently, the load carrying capacity of the system. However, it is hoped that this ar-ticle, by identifying the many different types of products that were in use at one time or an-other, will assist the readers in their research of an antiquated or archaic system when it is encountered in an existing structure.▪
®
SOLUTIONS FOR:
• Low Concrete Strength• Honeycomb / VoidRepair• Structural Failure /Damage• Missing or MisplacedReinforcement• New Slab Penetrations• Increased Loads• Change in Code• Seismic Upgrade• Blast Mitigation
STRUCTURAL STRENGTHENING SOLUTIONS
FROM AMERICA’S LARGEST SPECIALTY REPAIR CONTRACTOR
www.spsrepair.com/fix2 • 800-899-1016Offices Nationwide
A situation like this can cause costly delays. In many cases it can be fixed using proven strengthening techniques and materials that are not common in new construction.That’s where we come in…Structural Preservation Systems (SPS) understands the engineering, construction and economic issues facing a scenario like the one above. We know it requires a balance of technical and contracting expertise. We combine our experience in design support and contracting knowledge to work with and support you, the structural engineer.Go to our website…review the case studies and see how you can benefit from our 30+ years of experience helping structural engineers.Have an immediate need? Call Lisa Hardy at 800-899-1016.
We can fix itWe can fix it
AD
VERTISEM
ENT - For A
dvertiser Inform
ation, visit w
ww
.STR
UC
TUREm
ag.o
rg
References“Principals of Tile Engineering,” Handbook of Design. Harry C. Plummer and Edwin F. Wanner,
Structural Clay Products Institute, 1947. ACI Journal Proceeding, 1918
“Prestressed Concrete Innovations in Tennessee.” PCI Journal January-February 1979. Ross H. Bryan
“Dox Plank for High Speed Floor and Roof Construction,” Design Tables. NABCO Plank Company. Publication Date – Unknown; Made available by the NCMA – Accession No. TF02657.
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine September 2008
aid
s for the structural eng
ineer’s toolbox
EN
GIN
EER’
S NO
TEBO
OK
25
This article is the fifth in a series that is intended to provide a resource of information to structural engineers for projects that involve the repair, restoration or adaptive reuse of older buildings for which no drawings exist.
Antiquated Structural Systems SeriesPart 5
By D. Matthew Stuart, P.E., S.E., F. ASCE, SECB
Figure 1: F&A System.
Figure 2: Rutten System.
Figure 3: Tee Joist.
PHYSICAL PROPERTIES
A = 90 in.2
I = 1790 in.4
Sb = 170 in.3
St = 327 in. 3
Yb = 10.53 in.
Wt = 93 lbs per lin. ft.
Field Pour
Up to 15’ cantileverwith high steel locations
Rein. concretepoured in placetie beamon block wall
Ledger Beam
Block Wall
Cantileverscan be tapered
Flange bearingoptional
12”
4”
16
”
2½”
5¼”
12”
3¼”
6.47”
C. G. C.
PHYSICAL PROPERTIES
A = 51 in.2
I = 601 in.4
Sb = 92.9 in.3
St = 108.7 in. 3
Yb = 6.47 in.
grout
built-up roofing
prefabricatedroof deck
Figure 4: Keystone Joist.
used conventionally reinforced precast concrete inverted T-joists spaced at 28 inches on center, which supported concrete block filler slabs. The entire assembly then received a 2-inch, cast-in-place concrete topping, which acted compositely with the precast joists. Prior to the F&A System, Peter Rutten developed and patented a similar system in the 1930s (see Figure 2).The F&A precast joists were available in depths of 6, 8, 10 and
12 inches, and were capable of spanning anywhere from 6 feet to 36 feet for load capacities from 30 to 900 pounds per square foot, depending on the span, depth of joist and reinforcement. The ends of the precast joists could be cast integral with a site cast concrete beam, or bear directly on either precast concrete girders or steel beams. The F&A System included filler blocks that could be either placed flush with the bottom of the joist or recessed at the same level as the bearing ledge of the joist.Similar precast concrete systems that were in use during the
same approximate time period included Tee joists and Key-stone joists. Tee joists (see Figure 3) came in depths of 16 and 20 inches and were typically prestressed. Keystone joists (see Figure 4) were available in 8- and 12-inch depths and could
The purpose of this series is to compile and disseminate a resource of information to enable structural engi-neers to share their knowledge of existing structural
systems that may no longer be in use, but are capable of being adapted or reanalyzed for safe reuse in the marketplace of today and the future.
Precast Concrete Framing Systems
As discussed in the last article in this series, many contractors prefabricated modular clay tile and masonry units off-site into beams and slabs that could be delivered to the job site. This method of construction ultimately progressed to solid precast concrete units, which is the topic of this article.In the 1950s, one of the most prevalent precast concrete system
in general use was the F&A System. This system (see Figure 1) S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine September 200826
Factory made Unit Concrete Floor BeamSuspended Ceiling
Steel I Beam
TYPE B
Field ConcreteFireproofing
Standard
Unit Slab
Figure 7a & 7b: Watson Floor System.
ADVERTISEMENT - For Advertiser Information, visit www.STRUCTUREmag.org
ConcreteBeam Covering
Cope on Concrete I
Channel Reinforcement
Rod Reinforcement
Flooring on SleepersCement Tile or Terazza
on Fill to cover Pipes & conduits
WireLath
Figure 6: Waite’s I Beam.
Factory made Unit
Concrete Floor Beam
Suspended Ceiling
FieldConcreteFireproofing
Steel I Beam
TYPE A
Examples of other proprietary precast systems that are no longer in use include:Gypsteel Floor and Ceiling Slabs (Figure 5): The floor slabs of this system
consisted of 24-inch-wide, 22-inch-thick molded precast gypsum, reinforced with cold drawn wires that projected from the coped or rabbeted bearing ends of the panels. The wires were twisted together with the adjacent panel end and the slot was then filled with grout for a smooth top finish. These slabs were premanufactured to span both 24 inches and 30 inches between steel support framing members. Ceiling slabs were 24-inch-wide, 2-inch-think molded precast gypsum, reinforced
be either conventionally re-inforced or prestressed. Tee joists ultimately evolved into Single Tees, Quad Tees and the Double Tee member that is still in common use today.Channel Slabs were an-
other prevalent precast con-crete member in the 1950s. These were typically used for roof construction between supporting precast beams or steel members. he slabs were typically 24 inches wide and 1 inch thick, with 3½-inch- deep by 2-inch-wide, down-turned edge “flanges”. This product was capable of span-ning up to 9 feet and supporting up to 60 pounds per square foot of superimposed load.
CEILING TILEFLOOR TILE
24”24”
29¾”
Figure 5: Gypsteel Floor & Ceiling Slab.
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine September 200827
Negative Bar
Girder
Girder
Firep’f ’g
Variable EndUnit
Column Tie
Colu
mn
Standard Center Unit
A
A
¾” Mortar Topping & Finish
¾” Mortar Topping & Finish
Varies
Varies
9”
9”
8’6”
Plaster
Locking Joint & Cross-BridgePoured on Job
St’dFloor Units
12”
8”, 1
0”, 1
2” D
ep
thFi
n. F
loo
r to
Ce
ilin
g
Section A-A
Figure 8: Miller System.
with flat steel bars that projected from the ends of the panels to act in conjunction with hangers suspended from the top flange of the supporting steel framing. The Gypsteel system was manufactured in New Jersey and used extensively in New York City.Waite’s Concrete I Beams (Figure 6): This system was used in a number
of buildings constructed by The Standard Concrete Steel Company of
ADVERTISEMENT - For Advertiser Information, visit www.STRUCTUREmag.org
Hy-Rib Metal Lath Electric Welded Steel Standard Joist Spacing 18” or 24”Other Spacing Optional Depending on Loads
1½ “ or 2” Concrete or “Haydite” Slab
Figure 9: Lith-I-Bar System.
New York City. The floor framing system consisted of precast concrete I-beams spaced at approximately 18 inches on center of either 10-inch or 12-inch depth, which were supported from the bottom flange of steel beams that were spaced 5 to 7 feet apart. A field-cast concrete topping was then placed on top of the I-beams, while the spaces between the lower flanges were infilled, as well, to provide a flat ceiling surface.Watson Reinforced Concrete Floor System (Figures 7a and 7b): This
type of precast construction was installed by the Unit Construction Company of St. Louis and included two types of framing. The first configuration was intended for long spans and heavy loads, and involved the use of precast T-sections placed side by side. The T-sections were supported by steel beams that were encased in concrete. For shorter spans (less than 20 feet) and loads of 200 pounds per square foot or less, precast beams spaced at 5 feet on center were used to support precast channel slabs. The precast beams were in turn supported by steel beams encased in concrete.Miller Precast System (Figure 8): This system was devised by the Precast
Floors Corporation of New York in 1929. The precast units were shipped in three separate segments, which were aligned and supported on temporary shoring at the job site. Projecting reinforcement at the interior ends of the segments was embedded in a dry mix concrete, which was used to fill in the 9- to 10½-inch-long gaps between the segments. The center segment was produced in a standard fixed length, while the end segments were produced in varying lengths to allow for adjustment to accommodate different span lengths. Negative moment reinforcement was then embedded in a field-cast topping for continuity across the supporting steel beams. The voided, 12-inch-wide, box-shaped units were produced in 6-, 8- and 10-inch depths. Lith-I-Bar System (Figure 9): This system was developed in Michigan
and involved a dry-mix, lightweight concrete that was placed in an I-shaped cross-section mold and compacted with cast-iron rollers. The
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine September 200828
units were typically spaced at 18 or 24 inches on center with a 2-inch field topping cast on either removable or stay-in-place metal lath formwork for composite action with the members. Porete Floor System (Figure 10): This system was manufactured
in New Jersey and consisted of precast hollow formed units of 4 to 6 feet in length. This system was similar to the Miller system in that units were aligned and supported on temporary shoring at the job site. The gap between the abutting aligned units and the continuous pocket along the sides of each unit, in which field-positioned bottom reinforcement was first placed, were then filled in with a mortar/grout. All of the units were closed at each end to prevent the mortar/grout from flowing into the hollow voids of the precast member. These units were typically supported by steel beams over which top reinforcement was po-sitioned in the continuous member pockets to provide continu-ity of the floor slab system. This precast system was capable of spans from 10 to 25 feet.Tee Stone System (Figure 11): This precast member could be
used as a floor beam, roof beam or wall panel and was origi-nally manufactured in New York. The T-section was 8 inches deep, with a 16-inch-wide flange and a 1-inch-wide stem, and was manufactured in standard lengths of 8 and 16 feet. For floor construction, the T could be installed in either a flange up or flange down position. The units were placed in the field with a 1-inch gap between the edges of the flanges, which was filled with grout. The flange mesh reinforcement extended into these continuous gaps to produce a monolithic slab.Pyrobar Precast Roof System (Figure 12): This cast gypsum sys-
tem was manufactured for use as a roof slab and was available in both 3-inch-deep solid and 4-inch-deep hollow-core sections for short-span applications, as well as 5- and 6-inch hollow-core sections for long-span applications. The short-span sections were made in 12-inch widths and 30-inch lengths. The long-span sec-tions were made in 18-inch widths and lengths from 4 feet to 62 feet. The short-span members were typically supported by steel bulb tees, while the long-span members were supported by underslung steel wide flange and channel beams.All of the above precast systems were designed based on the
basic reinforced concrete beam analysis theories of their era. Load tables were also commonly developed and published by most of the manufacturers. The problem with all of the above systems when one encounters them in a building today is that, in the absence of existing drawings, it is difficult to determine the internal reinforcement and, consequently, the load-carrying capacity of the system. However, it is hoped that this article, by identifying the many different types of products that were in use at one time or other, will assist readers in their research of an antiquated or archaic system when it is encountered in an existing structure.▪
Details of Short Span Pyrobar Roof-Tile.
All Tile Reinforced with Electrically Welded Galvanized Steel Mat.
30”
12”
12”
3” Solid
4” Hollow
ReinforcingMat
3”
1”
Depth 3” Solid 4” HollowLength
Wt per Sq Ft
30” 30”
15 lbs 15 lbs
Grout
Not to Exceed 2”
Not to Exceed 2”
Notch can be furnished
on 3” solid only
30d” For I Beams30d” 30:” for
Tees
18”
5”
6”
ReinforcingMat
Standard Long Span Hollow Tile
Length
Core
Depth
Length
Wt per Sq Ft 20 lbs 25 lbs
4’0” to 6’6” 4’0” to 6’6”
5” 6”
LapJoint
ButtEnd Note: Notch can be
Furnished on Both 5” & 6” Long SpanHollow Roof Tile
Short Span Roof Tile - 30” Type
Figure 12: Pyrobar Roof System.
D. Matthew Stuart, P.E., S.E., F.ASCE, SECB is licensed in 20 states and has over 30 years experience as a structural consulting engineer. He currently works as a Senior Project Manager at the main office of CMX located in New Jersey, and may be contacted at mstuart@CMXEngineering.com.
Future installments of the archaic structural systems series will cover post-tensioning systems; structural steel stub-girder; open web steel joist and cast iron construction. If there are other topics along these lines that you would like to see addressed, please send your suggestions and any relevant information that you have to the author.
Reinforcing Bars
A A
4”
6’0”
14”
PrecastUnits
Formwork
14”
Section A-A
Cross Channels Filled
with Cement Mortar
Figure 10: Porete Floor System.
Extending Wires Overlap
¼” Flange Reinforcing Rods
16”
8”
1”
1”
1” Space Filled with Grout
Mesh Wires Projectat Least 2½” on Each Side
Mesh Wire ExtendingFrom Flange Downinto Stem Forming
a LoopStem
Stem ReinforcingRod Varies
Figure 11: Tee Stone System.
ReferencesKidder, Frank E. and Harry Parker, Architects’ and Builders’
Handbook, 18th Edition, John Wiley and Sons, 1956
Specification and Load Table Archives, Nitterhouse Concrete Products, Inc., Chambersburg, PA
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine November 2008
aid
s for the structural eng
ineer’s toolbox
EN
GIN
EER’
S NO
TEBO
OK
For this series of articles, “antiquat-ed” has been defined as meaning outmoded or discarded for reasons
of age. In reality, however, most if not all of the systems that have been and will be discussed are no longer in use simply because they have been replaced by more innovative or more economical methods of construction.Most of the antiquated systems discussed
so far have been out of popular use for a considerable number of years, with some dating back to the first part of the last cen-tury. However, the subject of this article deals with a system that was still in use less than 20 years ago.The purpose of this series is to compile
and disseminate a resource of information to structural engineers for projects that involve the repair, restoration, or adaptive reuse of older buildings for which no drawings exist. It is hoped that this will enable structural engineers to share their knowledge of existing structural systems that may no longer be in use, but are capable of being adapted or reanalyzed for safe reuse in the marketplace of today and the future.
53
Antiquated Structural Systems SeriesStructural Steel Composite Stub-Girder Construction – Part 6By D. Matthew Stuart, P.E., S.E., F. ASCE, SECB
system went on to be used in a large number of
high-rise buildings in North America up through the 1980s. However, the system was eventually abandoned because of increased labor cost associated with both fabrication and the need for shoring until the field-cast concrete slab attained sufficient strength.Advantages of the stub-girder system
that led to its use during the time period in which it was popular included:
1) Reduction in steel tonnage by as much as 25% over conventional composite floor framing due to:a) Improved structural efficiency
as a result of the greater depth of the stub-girder compared to a conventional system; and
b) Improved structural efficiency due to the ability of transverse floor framing members to act as continuous beams through openings between stubs.
2) Reduction of overall depth of the structural floor framing system by as much as 6 to 10 inches over a conventionally framed composite floor system, which allowed for a reduced floor-to-floor height and overall height of the building and associated cladding.
TestingPrior to use of the stub-girder system,
a load test was performed at Granco Steel Products Company in St. Louis. The test specimen included a W14x48 continuous bottom beam, W16x26 stubs and floor beams, and a 5-foot-wide, 3¼-inch-deep, lightweight concrete slab over a 2-inch metal deck flange, which was attached to stubs via shear connectors (Figure 2, page 54).The test specimen was loaded beyond
calculated design load, with initial failure occurring at the exterior end of the outermost stub at one end of the stub-girder. The method of failure included web crippling and delamina-tion of the web from the flange. Ap-plication of additional load resulted in crushing of the slab at the inside edge of the same stub. However, separation between the bottom of the slab and the top of the stubs did not occur, which indicated that composite behavior was maintained up to the point of local-ized crushing of the concrete slab. Web stiffeners added to the failed stub allowed the system to achieve a final failure load that was 2.2 times greater than the calculated design load.The methods of design used to deter-
mine capacity of the section included a non-prismatic beam analysis, a Vierendeel girder/truss analysis, and a finite element analysis. For the Vierendeel analysis, stubs and transverse floor beams act as verticals and the concrete slab and continuous beam act as chords. See Figure 3 (page 55) for a comparison of a typical Vierendeel truss and stub-girder components. All three of these methods
The Stub-Girder Composite System
A stub-girder is a composite system constructed with a continuous structural steel beam and a reinforced concrete slab separated by a series of short, typically wide, flange sections called stubs. Stubs are welded to the top of a continuous beam and attached to the concrete slab by shear connectors. Spaces between ends of stubs are used for installation of mechanical ducts and other utility systems and for placement of transverse floor beams that span between stub-girders. Ideally, the depth of stubs and floor beams are identical to allow for transverse framing to support a concrete slab deck, which spans parallel to the stub-girder, and to facilitate composite action between the floor beam and slab (Figure 1).Stub-girder construction was first used
in 1971 at the 34-story One Allen Cen- ter office building in Houston, Texas. The system was developed by Joseph P. Colaco, Ellisor Engineers, Inc., to fa- cilitate integration of mechanical ducts into steel floor framing of repetitive, multistory high-rise construction. This
Figure 1: Floor Framing Systems. Courtesy AISC.
CORE
w BEAMS1
FLOOR PLAN
± 10’ - 0”±
4’ -
0”
DUCT
CEILINGSECTION 1
CONVENTIONAL FLOOR FRAMING SYSTEM
(a)
CORE
2
FLOOR PLAN
SECTION 2STUB GIRDER SYSTEM
(b)
± 10’ - 0”
± 3
’ - 6
”
CEILING
DUCT
W14 STUB GIRDER
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine November 200854
of analysis provided a close representation of actual behavior of the stub-girder; however, the Vierendeel and finite element methods more closely identified secondary moment effects on each side of the openings. The Vierendeel method of analysis also provided a more accurate representation of actual steel stress, while the finite element method provided a more accurate representation of stress in the concrete slab, including high stresses that resulted in crushing of concrete at the inside edge of the first exterior stub as observed in the test specimen.Additional tests of stub-girders were per-
formed in the late 1970s in Canada. The primary purpose of these tests was to deter-mine effects of changes in spacing and depth of stubs, and to establish failure modes of a stub-girder. Results confirmed that behavior of a stub-girder was similar to a Vierendeel girder/truss. Supplementary conclusions of these tests included:
1) The stiffness of the girder increases as the length of the open panel between the stubs decreases.
2) Shear distortions at open panels (as a result of the Vierendeel action) were an important parameter in determining elastic deflection of the stub-girder, but did not influence rotation of solid end sections of the overall girder.
3) Tensile cracking of the concrete slab at the ends of open panels occurred at relatively low loads, but did not have a significant impact on elastic stiffness of the girder.
4) Further extensive cracking of the concrete slab at the ends of open panels occurred in the inelastic range of the girder. It was further determined that ultimate strength and ductility of the girder could be improved through use of internal reinforcement within the slab that was placed to resist the observed cracking.
5) Precision of the Vierendeel method of analysis was dependent on accuracy of distribution of shear forces between the concrete slab and the continuous lower beam across open panels and assumptions made relative to location of points of contraflexure within open panels.
6) Failure of shear connectors resulted as a combination of shearing and prying effects.
7) To prevent premature failure due to web crippling at stubs, stiffeners should be provided.
8) Five different failure mechanisms were identified: buckling of the stub web, concrete failure in the vicinity of shear connectors, diagonal tension failure of the concrete slab, shearing off of headed stud connectors, and combined yielding of the steel beam and crushing of the concrete slab at the ends of open panels due to cumulative effects of primary and secondary (Vierendeel) moments.
Further research in Canada revealed ad-ditional insights into behavior, design, and economical construction of stub-girders. This research indicated that only partial end plate stiffeners, rather than traditional fitted stiff-eners, were required to reinforce stub webs. Furthermore, web stiffeners were not always required at interior stubs. In addition, a con-tinuous perimeter weld between the base of the stub and the top of the continuous beam was not required. Tests also confirmed that rolled wide flange shapes were more conducive to stub-girder construction than split T (WT) or rectangular hollow tube (HSS) sections.Additional conclusions of these later
Canadian tests revealed that:1) Deflection computations using the
Vierendeel method of analysis were typically conservative for service loads, and unconservative for ultimate loading conditions.
2) The amount of internal slab reinforcement, particularly in the direction transverse to the stub-girder span, was established based on Canadian Standard Association (CSA) criteria available at the time of testing.
3) The conventional method of calculating number of shear studs required and application of standard methods of composite design to analysis of stub-girders appeared to provide satisfactory results; however, caution was recommended when specifying closely spaced studs, particularly at the end stub.
9’-7” 9’-7”
58’-9”
9’-7” 10’-0”
2’-1” 5’-0” 2’-6” 5’-0” 2’-6”2’-6”3’-0”3’-3½” 3’-3½” 3’-0”3’-3½” 3’-3½”
14 SHEAR CONN.(2 ROWS)
14 SHEAR CONN.(2 ROWS)
6 SH. CONN.(2 ROWS)
6 SH. CONN.(2 ROWS)
STRAINGAGE
LOADPT. LOAD
PT.LOADPT.
A
A
W16x26 W16x26 W16x26
W16x26 W16x26 STUB(BOTH SIDES)WELDED TO W14
W16x26W16x26
W14x48 (36,000 PSI YIELD)
W16x26
6¾¨
B
BW14 x 103
4x4 - 4/4 WWFDRAPED
W16 x 26W16 x 26
W14 x 48W14 x 48
SHEARCONNECTORS
SLAB WIDTH5’-0”
16 GA. C12” x 1½”CONTINUOUS
5/16” FILLET WELD
22GA. C2 COFAR
5’3
”
LT
. W
T.
CO
NC
RE
TE
C2 SHEAR CONN.EVERY CORRUGATION
5’4” LT. WT. CONCRETEOVER CORRUGATION
W16 x 26 STUB(BOTH SIDES)WELDED TOW10
W10 x 12
3-7/8” BOLTS
2L - 4” x 5½” x 5/8”
2L - 4” x 5½” x 5/8”w/2 - 7/8” & BOLTS/LEG
- 11” x 1” x 0’-11”WELDED TO W
STRAINGAGE
DEFLECTIONGAGE
- 15” x 1” x 1’-3”WELDED TO W14
3/163/163/163/163/16
SECTION A-A SECTION B-B
ELEVATIONPL
PL
Figure 2: Stub Girder Load Test. Courtesy of AISC.
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine November 200855
Design GuidelinesFurther recommendations and guidelines
emerged throughout the 1980s for the stub-girder system. In fact, the American Institute of Steel Construction (AISC) had plans to develop a design guide for stub-girder construction; however, because deeper wide flange sections became more readily available and guidelines for design of reinforced and unreinforced web openings became more established (see AISC Steel Design Guide Series 2; Steel and Composite Beams with Web Openings; 1990), it was never published. In order to document some of the final design guidelines that were established for stub-girder construction, the following list of criteria is provided:
1) Economical spans for stub-girders range from 30 to 50 feet, with the ideal span range being 35 to 45 feet.
2) Transverse floor beams should be spaced at 8 to 12 feet on center.
3) Stubs do not necessarily have to be placed symmetrically about the centerline of the stub-girder span.
4) Use of 3 to 5 stubs per span is the most common arrangement.
5) The stub located nearest the end of the stub-girder (and the surrounding, adjacent truss/girder elements) is the most critical member, as it directly controls behavior of the overall stub-girder. In addition, the end stub may be placed at the very end of the continuous bottom beam, directly adjacent to the support point.
6) Performance of a stub-girder is not particularly sensitive to length of stubs, as long as length of stubs are maintained within the following limits:
a) Exterior stubs should be 5 to 7 feet long.
b) Interior stubs should be 3 to 5 feet long.
However, increasing the length of the open panel between stubs will reduce stiffness of the stub-girder.
7) Stub-girders must be constructed as shored composite construction in order to take full advantage of the concrete slab top chord. Further, because of the additional dead load imposed by shoring from upper floors in multi-story construction, the need for shoring the non-composite section becomes even more critical.
8) Stub-girders should be fabricated or shored to provide a camber that is equal to dead load deflection of the member.
9) Overall strength of a stub-girder is not controlled by compressive strength of the concrete slab, therefore use of high-strength concrete mixes provides no advantage.
10) It is typical for ribs of a metal floor deck to run parallel to the span of a stub-girder. This orientation of ribs therefore increases the area of the top chord slab and also makes it possible to arrange a continuous rib or trough directly above the stubs, which in turn improves composite interaction of the slab with the stub-girder.
11) Welds between the bottom of stubs and the top of the continuous bottom beam should be concentrated at the ends of the stubs where forces between these two elements are greatest.
12) Internal longitudinal slab reinforcement to add strength, ductility and stiffness to the stub-girder should be provided in two layers, one just below and one just above the heads of shear studs.
13) Internal transverse slab reinforcement should be provided to add shear strength and ductility. Placing transverse reinforcement in a herring bone pattern – i.e., diagonal to the direction of the stub-girder span – will also increase the effective width of the concrete flange/top chord.
14) Flexural stiffness of the top chord slab of a stub-girder should be based on the conventional effective width allowed by standard composite beam design criteria, except that the transformed section should include contribution of both the metal deck and internal longitudinal reinforcement.
15) It is not proper to include the top flange of stubs in the calculation of moment of inertia of the top chord slab element.
In conclusion, the stub-girder method of construction was and still is an innovative solution to multi-story, framed steel floor construction. However, as deeper wide flange sections became more available in the marketplace and design engineers be-came more accustomed to analyzing web holes in wide flange beams, the use of stub-girder construction waned. In addition, because of extra labor costs associated with fabrication of stub-girders and the necessity to construct stub-girders as shored composite construction, the system priced itself out of the industry.▪
D. Matthew Stuart, P.E., S.E., F. ASCE, SECB is licensed in 20 states. Matt currently works as a Senior Project Manager at the main office of CMX located in New Jersey, and also serves as Adjunct Professor for the Masters of Structural Engineering Program at Lehigh University, Bethlehem, PA. Mr. Stuart may be contacted at mstuart@CMXEngineering.com.
(CONTINUOUS CONCRETE SLAB)
TOP CHORD
VERTICAL WEB
(STUB &
TRANSVERSE
BEAMS)
BOTTOM CHORD
(CONTINUOUS BOTTOM BEAM)
TYPICAL WIDE FLANGE VIERENDEEL TRUSS
(EQUIVALENT STUB GIRDER COMPONENTS)
Figure 3.
16) Modeling of stubs as verticals of the Vierendeel truss/girder involves dividing the stubs up into vertical elements equal to one-foot lengths of the section, spaced at one foot on center from one end of the stub to the other. Vertical stub elements should be modeled as fixed at the top and bottom, at the top chord (concrete slab) and bottom chord (continuous beam) of the truss/girder, respectively.
17) Transverse floor beams should be modeled as single vertical web members/elements of the truss/girder. The top and bottom of the member should be modeled as pinned at top and bottom chords.
The online version of this article contains references. Please visit www.STRUCTUREmag.org.
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine November 2008
ReferencesA Stub-Girder System for High-Rise BuildingsJoseph P. ColacoAISC National Engineering ConferenceNew York, NY; May 1972
An Experimental Investigation of Stub-GirdersY.W. Lam, T. Rezansoff and M.U. HosainUniversity of Saskatoon, CanadaBehavior of Building Systems and Building Components ConferenceVanderbilt UniversityNashville, TN; March 1979
Some Aspects of Stub-Girder DesignReidar Bjorhovde and T.J. ZimmermanUniversity of Alberta and the CISCCanadian Structural Engineering ConferenceMontreal, Quebec; February 1980Reprinted in the AISC Engineering Journal; Third Quarter 1980
Structural Engineering HandbookChapter 18; Stud Girder Floor SystemsReidar BjorhovdeUniversity of PittsburghCRC Press LLC 1999
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine December 2008
aid
s for the structural eng
ineer’s toolbox
EN
GIN
EER’
S NO
TEBO
OK
25
Antiquated Structural Systems SeriesPart 7By D. Matthew Stuart, P.E., S.E., F. ASCE, SECB
For this series of articles, “antiquated” has been defined as meaning outmoded or discarded for reasons of age. In reality, however, most if not all of the systems that have been and will be discussed are no longer in use simply because they have been replaced by more innovative or more economical methods of construction.However, this article deals with a method of construction that is still very much
in use today. Therefore, this installment will provide a history of how the post-tensioning industry developed in the United States.The historic, original construction practices described in this article may still be
encountered in existing structures. Therefore, the primary purpose of this series of articles will be fulfilled, which is to compile and disseminate a resource of information to enable structural engineers to share their knowledge of existing structural systems that may no longer be in use but are capable of being adapted or reanalyzed for safe reuse in the marketplace of today and the future.
the amount of total reinforcing steel in the concrete section because the tendons were made of higher strength material than conventional reinforcement.
Button-Headed Tendons
The button-headed system involved the use of parallel, ¼-inch-diameter, 240-ksi, cold-drawn wires, each with an approximate tensile capacity of 7 kips. The wires were combined together, generally in groups of seven, to form ten-dons. To anchor the tendons at each end, the wires were threaded through holes in a steel bearing plate and then an exter-nally threaded, circular stressing washer. A “button” was then cold-formed on the end of each wire via impact of a hammer on the end of the tendon. The buttons prevented the wires from passing back through the holes in the circular washer and, as a consequence, allowed them to become anchored during and after the field tensioning operation. The tendon assemblies were then coated with mastic for corrosion protection and wrapped in heavy wax paper to prevent the tendon from bonding to the concrete. The above process was completed in a pre-assembly shop prior to shipment to the construc-tion site.
Once the tendons were at the job site, the assemblies were placed in the forms, the concrete was cast, and the wires were stressed once the concrete achieved a minimum strength. The stressing process was performed with a hydraulic jack, which was attached to the threaded wash-er. A steel shim was inserted between the bearing plate and the washer to prevent the circular anchorage from retracting back against the bearing plate. As the shim plate was prefabricated, the length of the plate parallel to the tendon had to match exactly the calculated elongation. This system of anchoring the tendon was inherently flawed because if the actual elongation did not match the length of the shim plate, field corrections had to be made.A second problem associated with the
button-headed system was the fact that the stressing washer and shim plate had to be external to the concrete in order to facilitate the stressing operation. This resulted in placing this portion of the anchor assembly either outboard of the concrete or within a formed stressing pocket. In either case, it was necessary to cast a continuous pour strip or fill in the pockets after the stressing operation was completed.Another problem presented by the
button-headed system was the need to use bulky couplers when intermediate stressing of the tendons was necessary. Intermediate stressing procedures were required any time a continuous framed slab was longer than approximately 160 feet. This was because the frictional re-straint of the wires during the tensioning operation limited the practical stressing lengths of the tendons to approximately 80 feet. By stressing the tendons at each end anchorage, a maximum of 160 feet could be tensioned without having to in-troduce an intermediate stressing point coupler. The couplers usually consisted
Post-Tensioned Concrete Construction
In the 1950s, post-tensioned concrete was introduced into the United States from Europe, primarily as a part of the lift-slab construction industry. However, the first post-tensioned structure con-structed in the U.S. was the Walnut Lane Bridge in Philadelphia, between 1949 and 1950. This structure consisted of precast girders that were post-tensioned using the Magnel system developed in Belgium; i.e., the tendons were stressed after the concrete was cast, as opposed to pretensioning in which the tendons are stressed between abutments or within self-stressing forms before the concrete is cast. From these beginnings, this system of “prestressing” both cast-in-place and precast concrete structures and members continues to this day.
Lift-Slab Construction
Originally, the lift-slab industry used conventional reinforcing in the framed slabs that were first stack-cast on the ground and then lifted into position via hydraulic jacks located at the tops of the building columns. Typical slab spans and depths ranged from 25 to 30 feet and from 10 to 12 inches, respectively. However, because of these span lengths and the related excessive dead load deflections, as well as the stresses associated with the stripping process, cracking of the slabs was very common.In order to correct these inherent de-
ficiencies, the industry turned to the available European BBRV (button-head-ed) post-tensioning system (Figure 1). This system of post-tensioning allowed the reduction of slab thicknesses to as lit-tle as 8 inches, along with a reduction in Figure 1.
Courtesy of Larry Krauser, General Technologies Inc.
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine December 200826
of large, high-strength, externally threaded steel studs that were screwed into an internally threaded hole in the stressing washer.
Formed-in-Place Construction
At the same time that lift-slab construction was being advanced, separate companies spe-cializing in post-tensioned, formed-in-place construction began competing for the same projects. However, it was not until alliances were made between the post-tensioning com-panies and the emerging flying form industry that formed-in-place, post-tensioned con-struction really began to become competitive with lift-slab construction. By the 1960s, formed-in-place, post-tensioned construction had eclipsed lift-slab construction as the predominant method of constructing multi-story, cast-in-place concrete buildings.
Post-Tensioned Strand
The use of spirally woven, seven-wire strand as post-tensioned tendons was introduced into cast-in-place, post-tensioned construc-tion from the precast/prestressed concrete industry in the early 1960s. By the end of the same decade, the post-tensioning strand sys-tem had replaced the button-headed tendon system. Up until the early 1970s, unbonded, mono-strand tendons were installed greased and spirally wrapped with two layers of heavy kraft paper. By 1975, plastic sheathing was ex-clusively used for greased, unbonded strand.The use of strand was identified because
of the ability of the tendons to be secured with wedge anchors, which eliminated the primary disadvantage of the button-headed system – the lack of flexibility of the steel anchor shims. The first wedge anchor for use in the post-tensioned industry was patented by Edward K. Rice of Atlas Prestressing Corporation (Figure 2). The wedge anchor was constructed with a tapered, high-strength wire coil with no bearing plate. The plate shown in Figure 2 was only provided to secure the anchor to the inside face of edge form. This method of installation also eliminated the other disadvantage of the button-headed system – the need for a continuous pour strip or infilled stressing pockets.The wedge anchor assembly was installed
with a small, two-piece, round rubber grom-met that was positioned between the anchorage and the inside face of the edge form. The grommet allowed the anchorage to be re-cessed a few inches in from the edge form, which created a round hole in which the nose of the tensioning jack could be inserted. The grommet also filled in the space between the coil and the strand, preventing cement paste from entering the assembly during concrete placement. After tensioning was completed
and any excess strand was removed beyond the outboard end of the anchorage, the re-maining hole was filled with grout for a flush finish with the edge of the slab.The seven-wire strand system was also
more conducive to intermediate stressing locations, eliminating the final disadvantage of the button-headed system. With a strand, the cable could be placed as a continuous tendon with a wedge anchor assembly simply slid down the length of the strand to the in-termediate stressing point.The anchorage force associated with the
coil wedge assembly was transferred to the concrete by direct tension resistance to the lateral forces generated by the wedges on the
inside of the coil. As a result, localized failure of the concrete, particularly with lightweight mixes, was common. This inherent deficiency of the coil anchor resulted in the development of a ductile iron anchorage plate in the mid-1960s. The ductile iron assembly consisted of a flat bearing plate combined with a center, open, tapered barrel ring where the wedges were secured. The current end anchorage plates in use today are very similar to these original bearing assemblies.
Methods of Analysis
Because of the indeterminate nature of a continuous, cast-in-place concrete beam or slab and the secondary effects induced by draped post-tension tendon uplift forces at
the structural supports, the analysis of post-tensioned structures was originally difficult. The analysis of cast-in-place, post-tensioned structures became much easier, however, after a simplified method of analysis called “load balancing” was presented by T.Y. Lin in a 1963 ACI Journal paper.The load balancing method of analysis basi-
cally involved idealizing the vertical forces applied to a structure from the draped ten-dons as a series of upward uniform loads on the member. The load balancing method allowed structures to be accurately analyzed using standard structural engineering methods such as moment distribution. This greatly simplified the process of post-tensioned design, making it as easy as designing a conventionally reinforced structure, thereby helping to in-crease the popularity of post-tensioned con-struction from the late 1960s into the 1970s.
Two-Way Construction Methods
Originally post-tensioned, two-way slab fram-ing systems were constructed with column and middle strips, similar to those used in conventionally reinforced two-way slabs. How-ever, because the continuous two-way tendons had to be placed in a draped parabolic profile – near the top of the slab at the column lines, and near the bottom of the slab at midspan – the cables had to be placed in a basket weave pattern. This required that the tendons be numbered and installed in a specific sequence. This was difficult, particularly for structures with irregularly spaced column grids.To avoid the complexities of a basket weave
tendon installation, an alternate method of construction was conceived by T.Y. Lin & Associates and Atlas Prestressing Corporation for the infamous Watergate building located in Washington, D.C. in the late 1960s. The system involved banding the tendons – grouping the strands together within a narrow strip – in one direction along the column line grid, while the tendons in the other direction were spaced uniformly above the banded tendons. The banded method of construction resulted in considerable labor savings over the basket weave system, and has become the predominant method for placing post-tensioning tendons in two-way slabs ever since.The structural adequacy and performance
of the banded tendon layout was confirmed through a number of laboratory tests at the University of Texas, Austin in the early 1970s. In addition, the performance of the banded method of construction has also been proven through the continued serviceability of the many buildings that have been erected to date with this method of construction.
“...when ACI 318-77 was published, banded tendon distribution in two-way slab construction was
recognized by the Code.”
Figure 2.
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine December 200827
Building Codes and Industry Organization
Up through the ACI 318-71 Code, cast-in-place, post-tensioning design was essentially ignored, even though precast/prestressed members were addressed. However, when ACI 318-77 was published, banded tendon distribution in two-way slab construction was recognized by the Code. This improvement came about primarily through the formation and involvement of the Post Tensioning Insti-tute (PTI), which was established in 1976.
Common ProblemsThe two biggest problems faced by the post-
tensioning industry have been restraint cracking and corrosion, with the latter being the most pervasive. Solutions to restraint cracking were discovered fairly soon as the industry learned how to develop construction details to avoid the problem. Addressing corrosion did not start to occur until the PTI developed improved specifications in the mid 1970s.Concrete volume change occurs in both non-
prestressed and post-tensioned structures. Axial post-tensioning compression forces tend to close the restraint cracking associated with concrete volume change within the span of the member. However, the compression forces also result in significant axial shortening of the concrete member at the ends. Therefore, any restraint that is provided at the ends of the member – e.g., at walls or columns – can and does result in significant cracking in the post-tensioned member or the restraining structural support. Solutions to this problem include the development of slip joints and pour strips.The early unbonded tendon coatings (grease)
and sheathings were very inadequate for protecting the cables from corrosion, par-ticularly in harmful environments such as parking garages, where the use of deicing materials created caustic conditions for both the concrete and the internal reinforcement. The actual extent of the problem did not re-veal itself until about 10 or 15 years after the earliest post-tensioned concrete structures were constructed. The problem was cor-rected after the PTI developed specifications that became enforceable through certification programs created by the organization. The improvements established through the PTI specifications included sheathings, coatings and encapsulation systems for use in aggres-sively corrosive environments.
ConclusionThe post-tensioning industry is likely to con-
tinue to thrive worldwide for both building and bridge construction. Further advance-ments within the industry should continue to enhance the design and construction of post-tensioned structures. In fact, the author anticipates that, sometime in the future, an advanced chemical sheathing will be devel-oped for mono-strand construction that will allow for conventional unbonded installa-tion and tensioning, yet subsequently create a long-term bonded condition in the absence of internal ducts or grouting.▪
D. Matthew Stuart, P.E., S.E., F. ASCE, SECB is licensed in 20 states. Matt currently works as a Senior Project Manager at the main office of CMX located in New Jersey, and also serves as Adjunct Professor for the Masters of Structural Engineering Program at Lehigh University, Bethlehem, PA. Mr. Stuart may be contacted at mstuart@CMXEngineering.com.
ReferencesDurability of Post-Tensioning Tendons: Technical Report; November 2001, The International
Federation for Structural Concrete, Status of the Durability of Post-Tensioning Tendons in the United States, Cliff L. Freyermuth
Post-Tensioned Concrete: Five Decades of American Building Construction, Kenneth B. Bondy, Concrete Construction, December 2001
Post-Tensioned Concrete in Buildings, Past and Future, an Insiders View, Kenneth B. Bondy, PTI Journal Vol. 4 No.2, December 2006
AD
VERTISEM
ENT - For A
dvertiser Inform
ation, visit w
ww
.STR
UC
TUREm
ag.o
rg
®
S T R U C T U
R E®
magazi
ne
Copyright
aid
s fo
r th
e stru
ctur
al e
ngin
eer’s
tool
box
EN
GIN
EER’S
NO
TEBO
OK
STRUCTURE magazine March 2009 8
Figure 1a: Courtesy of Peabody Essex Museum.
Figure 1b.
continued on page 10
Antiquated Structural Systems SeriesPart 8By D. Matthew Stuart, P.E., S.E., F. ASCE, SECB
This article is the eighth in a series that is intended to provide a resource of information to structural engineers for projects that involve the repair, restoration or adaptive reuse of older buildings for which no drawings exist.The purpose of this series is to enable structural engineers to
share their knowledge of existing structural systems that may no longer be in use, but are capable of being adapted or reanalyzed for safe reuse in the marketplace of today and the future.
Wrought and Cast IronFerrous metals used in construction can
be categorized into three principal iron-carbon alloys, based on approximate carbon content: wrought iron (0.020% to 0.035%), steel (0.06% to 2.00%) and cast iron (2% to 4%).Wrought iron is almost pure iron and
contains between 1% to 4% slag (iron silicate). The slag is not alloyed into the wrought iron, which gives the material its characteristic laminated (or layered) fibrous appearance. Wrought iron can also be distinguished from cast iron by its generally simpler forms and less uniform appearance.Cast iron contains varying amounts of
silicon, sulfur, manganese and phos-phorus. Cast iron, while molten, is easily poured into sand molds, making it pos-sible to create unlimited forms, which also results in mold lines, flaws and air holes. Cast iron elements are commonly bolted or screwed together, while wrought iron is either riveted or welded.
Wrought Iron
Wrought iron refers to ferrous metals that can be worked or “wrought” on an anvil, or shaped and forged in rolling machines. Wrought iron is tough and stringy, and has an elasticity that was conducive for use in bolts, beams and built-up girders. Wrought iron is also easily welded.Until the mid-1800s, wrought iron in
buildings was used primarily for tie rods, straps, nails, hardware or decorative rail-ing and balconies. Around 1850, the structural use of wrought iron became more prevalent as rail beams, bulb-tees, channels and I-beams became commer-cially available.When wrought iron was employed as
tie rods, the material was typically used in conjunction with cast iron anchor plates shaped as stars, rosettes or an S. In built-up girders, wrought iron was
also used in conjunction with cast iron. Initially, these composite built-up girders were constructed as bowstring trusses with an upper cast iron chord and a lower wrought iron tie rod (Figures 1a and 1b). Later versions included perfo-rated girders generally constructed with cast iron in the top three-quarters of the member and wrought iron in the bottom portion (Figure 2).In 1854, the Trenton Iron Works man-
ufactured the first rolled wrought iron beam in the U.S., a 7-inch-deep, bulb-tee rail beam. The following year, the Trenton Iron Works also manufactured the first I-beam in the U.S., the “Coo-per beam”, although I-beams had previ-ously been rolled commercially in France in the 1840s. Prior to the development of the bulb-tee rail beam, wrought iron deck beams had been rolled for use in the
shipbuilding industry, and prior to the development of the Cooper beam, chan-nel beams had been manufactured. Figure 3 (page 10) shows the evolution of all of these rolled wrought iron sections.Rolled wrought iron beams contin-
ued to be used for several decades af-ter the mid 1850s, even after structural steel became available. However, as the quality and quantity of available steel improved, the use of wrought iron gradually came to an end. In general, the use of wrought iron beam framing in conjunction with cast iron compression components lasted from the mid-1850s until the late 1890s.The earliest known tabulation of wrought
iron structural shapes was published by Carnegie Kloman and Company in 1873. The largest beam listed was a 15-inch-deep I-beam that weighed 67 pounds per foot. This publication is very rare, but an 1876 edition is available in the Library of Congress. The author was able to locate a digitized copy of the 1892 8th Edition of the Wrought Iron and Steel Construction manual (published by Pencoyd Iron Works of Philadelphia) on Google Book Search.Material and design properties pro-
vided in the Pencoyd Iron Works man-ual indicate that the ultimate tensile strength of wrought iron was as high as Figure 2.
D R A F TCopyright©
D R A F TCopyright©
Wind Load Examples8th of 8 presentations on the SEAOC Structural Design Manuals*
This webinar will provide a step-by-step approach to applying the wind loading provisions of the 2006 International Building
Code (IBC), ASCE 7-05 and the new “Simpli ed” method to be adopted in the 2009 IBC. The presentation will include an explanation of the formulas used in the various wind loading methods as well as their application (portions of example 7 in Volume 2 of the Structural/Seismic Design Manual). The design example will cover both the Main Wind Force Resisting System and Components and Cladding provisions, along with a comparison of the results between the various methods.
All NCSEA Webinars: 10:00 Pacifi c, 11:00 Mountain, 12:00 Central, and 1:00 Eastern
March 10th
NCSEA Webinars
NCSEA is pleased to offer, along with the Masonry Society, a series of 6 masonry webinars. The
speakers are all well-respected in their eld; and
the sessions offer courses on masonry that not only help the practicing structural engineer, but are great
preparation for the SE exam in late April, 2009.
March 5th
March 19th
*The three volume set of 2006 Structural Seismic Design Manuals
are availableat a discounted price to registrants.
Strength Design of MasonryPart 3 of the Masonry Series
This webinar will review the strength design requirements of Chapter 3 of the MSJC for reinforced masonry and it will highlight modi cations
by IBC Section 2108. Simple design examples for walls subjected to out-of-plane loads will be reviewed. NOTE: This seminar will not address shear walls.
All webinars will award 1.5 hours of continuing education in all 50 states. Registration is $250 per session and
several people may attend for a single fee. A live question and answer period will follow each session.
Masonry TallwallsPart 4 of the Masonry Series
Strength design procedures for masonry have been under development for many years. Tall slender walls bene t greatly
using strength techniques resulting in taller, thinner construction with less reinforcement than those designed using allowablestress techniques. This presentation will provide the bene ts and theory of tallwall design. Examples will be developed based upon the International Building Code, to illustrate the theory, and computer solutions will be presented for comparison.
Dan Werdowatz, S.E. is Vice President of Josephson-Werdowatz & Associates and has over 23 years of experience. He holds both bachelor’s and master’s degrees, specializing in structural engineering, and is licensed in eight western states. He has worked with NCEES regarding the Structures II exam including updating the exam to the 2006 IBC. He collaborated on updating the wind examples in the Structural/Seismic Design Manual published by ICC. His practice involves design of new structures as well as review of existing structures.
Stephen Kerr, S.E. is a Senior Associate Engineer of Josephson-Werdowatz & Associates and has over 14 years of experience in building structural design. Stephen has a bachelor’s degree in Architectural Engineering from Cal Poly, San Luis Obispo, and is licensed in three states. Stephen has performed numerous wind analyses of new and existing structures to the IBC and ASCE 7 codes, and has collaborated with Dan on updating the Structural/Seismic Design Manual. Currently he serves on the NCSEA General Engineering Subcommittee.
Richard Klingner, Ph.D., FTMS is the Associate Chair of the Civil, Architectural and Environmental Engineering Department, University of Texas at Austin. His eld of specialization is structural engineering, with emphasis on the analytical and experimental investigation of the dynamic response of structures, earthquake-resistant design of masonry and concrete structures, and anchorage to concrete. For the period 2002-2008, he was Chair of the Masonry Standards Joint Committee.
David T. Biggs, P.E. is Principal of Ryan-Biggs Associates, a structural engineering rm in New York. He specializes in the design, evaluation, and restoration of masonry structures, forensic engineering, and the development of new masonry products. Mr. Biggs chairs the Prestressed Subcommittee for the Masonry Standards Joint Committee and is a Distinguished Member of the American Society of Civil Engineers. After the World Trade Center disaster, he investigated the performance of masonry in surrounding buildings.
DiamondReviewed
STRUCTURAL
CONTINUING
EDUCATION
ENGINEERS
NCSEA
Register at www.ncsea.com
D R A F TCopyright©
D R A F TCopyright©
STRUCTURE magazine March 2009 10
Figure 4.
Deck
Beam
7-inch Rail
Beam
Channell
Beam
Compound
Beam
I Beam
Figure 3.
50,000 psi. The allowable extreme fi-ber stress was indicated as 14,000 psi for wrought iron and 16,800 for steel. Limits for safe loads on wrought iron beams provided in the design tables include the character of the service load (i.e. static, fluctuating or impact) and extent of lateral support.
Cast Iron
Cast iron is very hard and resists compression forces very well but, because of the carbon content, it is also very brittle. Because of this, cast iron was used primarily for the construction of columns and compression elements of composite wrought iron girders.Shortly before the 1800s, cast iron was used
for columns in multi-story, wood-framed factory buildings in England. Cast iron was used because of its strength and perceived fire resistance. The combination of cast iron columns and wood framing continued to be widely used for the next half century; however, by the mid-1800s, wrought iron beams began to replace wood.Cast iron used for structural purposes
during the 1800s typically had a compres-sive strength of 80,000 psi. Although there was no clearly defined yield point of the material, the tensile strength ranged between 10,000 and 15,000 psi. The brittleness of the material, in conjunction with its inherent manufacturing flaws, made cast iron highly susceptible to tensile failures; therefore, for most of the 1800s, cast iron was only used to resist compression forces.In the mid-1800s, cast iron began to be
used for the construction of the front façades of buildings (Figure 4). The cast iron façade was used for both decorative and structural purposes, as the adjacent floor framing was supported by the cast iron. Cast iron was also used in conjunction with masonry façades as exposed decorative/structural window and door lintels (Figure 5), as well as for balconies and verandas. Many of these types of exterior structures still exist today, thanks to ongoing painting and maintenance efforts.
Rolled Sheet and Corrugated Iron
The first sheet iron in the U.S. was rolled in Trenton, New Jersey in the late 1700s. Sheet iron was used as flooring and roofing material up until the end of the 1800s. Cor-rugation of sheet iron, patented in England in 1829, was first used in the U.S in the 1830s. Corrugated sheet iron was typically
painted with pitch, which was later replaced by galvanizing.Corrugated iron was also manufactured in
arched sheets for floor construction. Typically, the ends of the arched sheets were supported by the bottom flanges of wrought iron I-beams, with concrete then cast on top of the sheets. This type of construction replaced brick arch floor construction, which had been used exten-sively with wrought iron I-beams previously.
Building Construction Overview
The use of prefabricated cast iron fa-çades, wrought iron beams and cast iron beam and column components served as a forerunner to the multi-story, steel-framed buildings of today. In addition, specialty wrought iron structures con-structed in New York City in the mid
1850s, such as firewatch towers and gunshot manufacturing towers that employed drilled and socketed rock foundation anchors and infill masonry walls, served as a precursor to early steel skyscraper construction.It is also interesting to note that the manufac-
turers of cast iron façades and other exposed components often identified the source of the material by either affixing foundry labels to the building or by using trademarks in the
ornamentation of the decorative portions of the façade. This information can sometimes be used to assist in the structural evaluation of an existing building.
DeteriorationIron oxidizes rapidly when exposed to mois-
ture and air. The product of the oxidation pro-cess is rust. The minimum relative humidity to promote “rusting” is 65%, but humidity levels lower than this can cause oxidization in the presence of pollutants. In addition, if chlorides are present, the corrosion process can become accelerated. Once a film of rust starts to develop, the natural porosity of the corrosion byproduct tends to act as a reser-voir for moisture, resulting in even further acceleration of the deterioration process.Cast iron will develop a somewhat protec-
tive surface scale, which makes it slightly more resistant to corrosion than wrought iron; however, it is still recommended that cast iron be painted to prevent rusting. Iron can also be corroded by acids, magnesium and some sulfur compounds. Dissimilar met-al galvanic corrosion can also occur between iron and copper, chromium, lead, stain-less steel and brass. In general, wrought iron rusts more rapidly than cast iron. However, because of the slag content of wrought iron, the material is more resistant to progressive corrosion than cast iron.Another type of deterioration common to
cast iron is graphitization. This condition can occur in the presence of acidic precipitation or seawater. As the iron corrodes because of exposure to these types of environments, po-rous graphite residue is impregnated within the surface corrosion byproduct. The cast iron element retains its original appearance
D R A F TCopyright©
D R A F TCopyright©
STRUCTURE magazine March 2009 11
Figure 5.
ADVERTISEMENT - For Advertiser Information, visit www.STRUCTUREmag.org
Additional protection and repair procedures can also include plating with metals, or cladding with plastic or epoxy. Sections or entire portions of a significantly deteriorat-ed area may be replaced with glass fiber re-inforced concrete (GFRC), fiber reinforced polyester (FRP) or aluminum.If additional structural retrofitting is required
as a part of an adaptive reuse project, the fol-lowing remedial work should be avoided if at all possible: welding, burning of holes, the use of impact drills, high strength bolts, and filling of voids or posts with concrete.▪
and shape, but becomes gradually weaker internally. Graphitization typically occurs when cast iron is not painted for extended periods or where the sealant has failed at the joints between adjoining components. This condition can be identified in the field by scraping the surface of the cast iron with a knife to see if deterioration of the iron is re-vealed beneath the surface.
Repair and RestorationThere are a number of methods available to
remove paint and corrosion from cast and wrought iron, including manual scraping, chip-ping or wire brushing; low-pressure grit or sand-blasting; flame cutting; and chemical removal.Once any existing paint and corrosion have
been removed, the most common method of protecting cast iron from further deterioration is to repaint the surface. Prior to repainting, it is first necessary to prepare or repair the surface. Proper preparation includes elimi-nation of crevices or pockets that can collect moisture to prevent accelerated deterioration; removal or smoothing of sharp corners to prevent accelerated paint failure; hermetical sealing of hollow section to prevent moisture intrusion and freeze/thaw damage; filling of
joints, cracks and bolt or screw holes with sealant to prevent moisture intrusion and freeze/thaw damage; and, following the paint manufacturer’s specifications.In all cases, it is recommended that a test
area be used to confirm that the selected cleaning, preparation and painting techniques are effective prior to attempting to remedi-ate the entire restoration area. It is also rec-ommended that sheltered areas such as eaves, where evaporation of moisture can be inhib-ited, be coated with additional layers of the selected paint or coating.
The online version of this article contains detailed references. Please visit www.STRUCTUREmag.org.
D. Matthew Stuart, P.E., S.E., F. ASCE, SECB is licensed in 20 states. Matt currently works as a Senior Project Manager at the main office of CMX located in New Jersey, and also serves as Adjunct Professor for the Masters of Structural Engineering Program at Lehigh University, Bethlehem, PA. Mr. Stuart may be contacted at mstuart@CMXEngineering.com.
®
© 2009 Simpson Strong-Tie Company Inc. SF09
In the past moment frames were the
expensive option when designers wanted
small wall sections and open fl oor plans.
The Simpson Strong-Tie® Strong Frame™
ordinary moment frame is changing that
by offering engineered moment frame
solutions complete with anchorage designs.
Now designers can spend minutes
choosing a pre-engineered moment frame
rather than hours designing one. And since
the Strong Frame moment frame installs
with 100% bolted connections, it is easier
for contractors to handle and install.
For more information, see our webinar
at www.strongtie.com/sfwebinar.
To view and request a copy of the
Strong Frame Ordinary Moment Frame
catalog, visit www.strongtie.com or
call (800) 999-5099.
STRONG FRAME™
ORDINARY MOMENT FRAME
C-SF09
(800) 999-5099www.strongtie.com
Change your frame of reference.
Blank.indd 1 1/12/2009 9:14:21 AM
D R A F TCopyright©
D R A F TCopyright©
aid
s fo
r th
e stru
ctur
al e
ngin
eer’s
tool
box
EN
GIN
EER’S
NO
TEBO
OK
STRUCTURE magazine June 200912
For this series of articles, “antiquated” has been deined as meaning outmoded or discarded for reasons of age. In reality, however, most of the systems that have been discussed are no longer in use simply because they have been replaced by more innovative or more economical methods of construction. This article, however, deals with a meth-
od of construction that is still very much in use today. Nevertheless, the historic, original construction practices described in this article may still be encountered in existing structures. Therefore, the pri-mary purpose of this series of articles will be fulilled: to compile and disseminate a resource of information to enable struc-tural engineers to share their knowledge of existing structural systems that may no longer be in use, but are capable of being adapted or reanalyzed for safe reuse in the marketplace of today and the future.
Open Web Steel JoistsHistory
The author would irst like to thank the Steel Joist Institute (SJI) for providing much of the material that was used in the development of this article. In fact, a brief history of open web joists is provided in the Catalog of Standard Specifications and Load Tables for Steel Joists and Joist Girders, published by SJI. A brief summary of this history is as follows: 1923 The irst Warren type, open web
truss/joist is manufactured using continuous round bars for the top and bottom chords, with a continuous bent round bar used for the web members.
1928 First standard speciications adopted after the formation of SJI. This initial type of open web steel joists was later identiied as the SJ-Series.
1929 First load table published. 1953 Introduction of the longspan
or L-Series joists for spans up to 96 feet with depths of up to 48 inches, which were jointly approved by AISC.
1959 Introduction of the S-Series joists, which replaced the SJ-Series joists. The allowable tensile strength was increased from 18 ksi to 20 ksi, and joist depths and spans were increased to 24 inches and 48 feet, respectively.
Ashland Steel Products Co., Inc – Ashland City, Tennessee); Vescom Structural Systems, Inc. – Westbury, New York; Ridgeway Joists (manufactured by Continental Steel Ltd. – Coquitlam, British Columbia); Northwest Joist Limited (a Division of Brittain Steel Limited – New Westminster, British Columbia); Cadmus Long Span and Joist Corporation (afiliated with Alexandria Iron Works, Inc. – Alexandria, Virginia); T-Chord Longspan Joists (manufactured by the Haven Busch Company – Grandville and Grand Rapids, Michigan); and the Macomber Steel Company – Canton, Ohio. Table 1 (please see online version; article end note provides web address) provides a summary description of the joists produced by these manufacturers.In addition, some manufacturers, prior
to becoming SJI members, produced pro- ducts other than the historical stan-dard SJI joist series. Some of these manufacturers include: Truscon Steel Company – Youngstown, Ohio; Macmar and Kalmantruss joists (manufactured by Kalman Steel Corporation, a Subsidiary of Bethlehem Steel Company – Bethlehem, Pennsylvania); and Gabriel Steel Company – Detroit, Michigan. Table 2 (please see online version; artcile end note provides web address) provides a summary description of the joists produced by these manufacturers. In addition to the information provided in Table 2, it should be noted that Bethlehem Steel Company also produced cold formed joists with hat channel sections for the chord members, and Gabriel Steel Company also produced unique V-shaped top chord and single round bar bottom chord members.Additional manufacturers not included
in Tables 1 and 2 include: Berger Steel Company (double V-shaped chord mem-bers); Armco Steel (cold formed hat channel chord members); Raychord Corporation (cold formed hat channel and U-shaped chord members); Republic Steel (cold formed hat channel chord members); and USS AmBridge (cold formed U-shaped chord members).▪
Antiquated Structural Systems SeriesPart 9a – Open Web Steel JoistsBy D. Matthew Stuart, P.E., S.E., F. ASCE, SECB
1961 Introduction of the J-Series joists, which replaced the S-Series joists. The allowable tensile strength was increased from 20 ksi to 22 ksi. Introduction of the LA-Series joists to replace the L-Series joists, which included an allowable tensile strength increase from 20 ksi to 22 ksi. Introduc-tion of the H-Series joists, which provided an allowable tensile strength of 30 ksi.
1962 Introduction of the LH-Series joists, which provided yield strengths between 36 ksi and 50 ksi.
1965 Development of a single speciica-tion for the J- and H-Series joists by SJI and AISC.
1966 Introduction of the LJ-Series joists, which replaced the LA-Series joist. In addition, a single speciication was developed for the LJ- and LH-Series joists.
1970 Introduction of the DLH- and DLJ-Series joists, which included depths up to 72 inches and spans up to 144 feet.
1978 Introduction of Joist Girders, including standard speciications and weight tables.
1986 Introduction of the K-Series joists, which replaced the H-Series joists.
1994 Introduction of the KCS joists, which provided a constant moment and shear capacity envelope across the entire length of the member.
SJI also recently published 80 Years of Open Web Steel Joist Construction. This publication includes a complete chrono-logical listing of the standard speciica-tions and load tables for all of the steel joists, and weight tables for the Joist Girders, previously made available by SJI over the time period from 1928 to 2008. This manual can be an invaluable tool for an engineer involved in the analysis of existing buildings constructed with open web steel joists.In addition, there were also a number of
joists produced by manufacturers that were either never members of SJI or joined it later. Some of these manufacturers in-clude: Ashland Steel Joists (manufactured by
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine June 200913
D. Matthew Stuart, P.E., S.E., F. ASCE, SECB is licensed in 20 states. Matt currently works as a Senior Project Manager at the main office of CMX located in New Jersey, and also serves as Adjunct Professor for the Masters of Structural Engineering Program at Lehigh University, Bethlehem, PA. Mr. Stuart may be contacted at mstuart@CMXEngineering.com.
Watch for Part 9b in an upcoming issue.
Resource Material
80 Years of Open Web Steel Joist Construction; A Compilation of Specifications and Load Tables Since 1928; Steel Joist Institute; Myrtle Beach, South Carolina (2009).
Catalog of Standard Specifications, Load Tables and Weight Tables for Steel Joists and Joist Girders; 42nd Edition; Steel Joist Institute; Myrtle Beach, South Carolina (2007).
Miscellaneous Steel Joist and Joist Girder Speciications and Load Tables; SJI Archives; Steel Joist Institute, Technology, Engineering, and Education Center; Myrtle Beach, South Carolina
Another Resource
Robert Higgins, P.E., maintains a website that provides civil and structural engineering information in the following categories:• Out-of-print material that may be useful when working on existing facilities.• Older, usually conservative methods for solving technical problems.• Public domain documents that have limited availability.In summary, this is content that is dificult to ind anywhere else.
To access it, visit www.SlideRuleEra.net.
Table 1 online version web address: www.STRUCTUREmag.org/archives/2009-6/table-1.pdf
Table 2 online version web address: www.STRUCTUREmag.org/archives/2009-6/table-2.pdf
S T R U C T U
R E®
magazi
ne
Copyright
Table 1: Unique Open Web Joists (Load Tables may be available from SJI)
System Figure DescriptionYield
Strength
Depth
(inches)
Span
(feet)Chords Webs Notes
Ashland
1 HS-Series Joists 50 ksi 8 to 24 8 to 48 Double angles Round bars
N/A LS-Series Joists 50 ksi Unknown64
maximumUnknown Unknown
Cadmus 5
1952 Structural T
Longspan & Standard
Joists
See Note 6 10± to 5412’-6” to
108Split T Angles 6, 7
Haven Busch 6
1952 to 1962 T-
Chord Longspan
Joists
See Note 9 18 to 88 25 to 175 Split T Angles 8, 9
Macomber
7 Purlin or Steel Joist Unknown 8 to 16 10 to 26 See Note 10 Round bars 10
8 Massillon Steel Joist Unknown 8 to 16 4 to 31 Round bars Round bars
9 Canton Steel Joist Unknown 8 to 16 Unknown Double angles Round bars
10 Buffalo Steel Joist Unknown 8 to 16 Unknown See Note 11 Round bars 11
N/A Special Joists Unknown 12 to 20 8 to 40 Unknown Unknown
11 Residence Joist Unknown 6 to 10 6 to 20 See Note 12 Round bars 12
12Standard Longspan
JoistSee Note 14 18 to 40 24 to 72 Double angles Angles & bars 13, 14
N/AIntermediate
LongspanSee Note 14 18 to 22 20 to 44 See Note 10 Round bars 10, 14
13 1955 New Yorker Unknown 8 to 24 7 to 48 V shaped plates Round bars
14V or Double V Bar
JoistUnknown 8 to 22 4 to 44 V shaped plates Round bars
N/A V-Girders Unknown 18 to 48 13 to 96 V shaped plates Round bars
15 V-Purlin Unknown 8 to 60 8 to 120 V shaped plates See Note 15 15
16 Allspan Unknown 8 to 76 8 to 152 V & Double V
shaped platesSee Note 15 15
N/A V-Lok Purlin Unknown 8 to 36 8 to 72V & Double V
shaped plates
Round bars or
round pipes16, 17
17 V-Lok Girder Unknown 12 to 40 15 to 50 See Note 18Round bars or
Angles16, 18
18 V-Beam Unknown 8 to 28 8 to 56 See note 19 Round bars 19
Northwest 4Series 1, 2, 3 & 4
JoistsSee Note 5 12 to 72 12 to 80 V shaped plates
Square bars &
round pipes4, 5
Ridgeway 3 Open Web Joists See Note 312± to
47±16± to 59± V shaped plates
Square bars &
round pipes3
Vescom
2Composite Floor
Joists36 & 50 ksi 8 to 40 20 to 48 Double angles Round bars 1
N/AComposite Truss
Girders36 & 50 ksi 16 to 40 20 to 50 Double angles Angles 2
Antiquated Structural Systems Series - Table 1Part 9a – Open Web Steel JoistsBy D. Matthew Stuart, P.E., S.E., F. ASCE, SECB
STRUCTURE magazine • June 2009Antiquated Systems Series - Part #9a - Table 1
S T R U C T U
R E®
magazi
ne
Copyright
Notes:1. Top chord included deformed, extended vertical leg of one angle for composite action with surrounding
concrete slab.2. Top chord included deformed, extended vertical plate in addition to double angles for composite action with
surrounding concrete slab.3. Web allowable stress: 36 ksi (bars) & 50 ksi (pipes); Chord allowable stress: 54 ksi.4. Joist designs over 80 feet spans were available upon request.5. Web allowable stress: 33 & 44 ksi (bars), 50 ksi (pipes); Chord allowable stress: 55 ksi.6. Allowable compressive stress for top chord or web members = 15 ksi. Allowable combined compressive stress
at top chord panel points and allowable tensile stress = 18 ksi.7. Chordteescutfromstandardwidelangeorjuniorbeams.8. Availableasparallelchord,singleordoubleslopedtopchordorhippedendconigurations.9. Allowable combined compressive stress at mid-panel chord and web = 15 ksi (1952); 20 ksi (1956). Allowable
combined compressive stress at panel points = 24 ksi (1956). Allowable tensile stress = 20 ksi (1952 & 1956).10. Double angle top chord; Round bars bottom chord.11. Inverted double angle top chord; Round bars bottom chord.12. Single steel angle and wood nailer top chord; Round bars bottom. 13. Available as parallel chord or single or double sloped top chord.14. Allowable combined direct and bending stress in top chords = 20 ksi.15. Sizes #2 - #9: Round bars; Sizes #10 up through #22: Angles.16. Included proprietary stud and slot end bearing connection – See Figure 17.17. Round bars, round pipes or angles.18. V & double V shaped plates or double angles.19. V shaped top chord & U shaped bottom chord plates.
STRUCTURE magazine • June 2009Antiquated Systems Series - Part #9a - Table 1
S T R U C T U
R E®
magazi
ne
Copyright
Notes:
1. Web allowable stress: 19,000 psi - 100(l/r); Chord allowable stress: 16,000 psi.
2. Cold formed chord allowable tension: 25 ksi; Hot rolled web members allowable compression: 17,000 psi - 100(l/r).
3. Cold formed chord allowable tension: 28.5 ksi; Hot rolled web members allowable compression: 19,000 psi - 100(l/r).
4. Availableasparallelchord,singleordoubleslopedtopchordconigurations.5. Chordanglesweresometimesarrangedtoetotoeforchannelconiguration.6. Allowable combined top chord compressive stress: 15 ksi; Allowable bottom chord tensile stress: 18 ksi.
7. Manufactured by punching web opening in blanks such that chords and webs do not have to be welded together.
8. Allowable tensile stress: 16 and 18 ksi.
9. Also marked as Kalman Joist.
10. Allowable tensile stress: 18 ksi.
11. Maximum tensile working stress: 22 ksi.
12. Maximum tensile working stress: 30 ksi.
13. Design tensile stress: 18 ksi.
14. Allowable combined compressive stress at panel points and allowable tensile stress = 18 ksi. Allowable combined
compressive stress at mid-panel and compression webs = 15 ksi.
15. Double angle top chord; Round bars bottom chord.
Table 2: Unique Open Web Joists (Load Tables may be available from SJI)
System Figure DescriptionYield
StrengthDepth
(inches) Span (feet)
Chords Webs Notes
Bethlehem
24 KalmanTruss Joists See Note 8 8 to 16 4 to 32 T shape Rectangular 7, 8, 9
25 MacMar Joists See Note 10 8 to 16 4 to 32 Angles Round bars 10
26 BLJ Series See Note 11 52 to 60 89 to 120 Structural Tee Angles 11
26 BLH Series See Note 12 52 to 60 89 to 120 Structural Tee Angles 12
27 Standard Open Web Joist See Note 13 8 to 16 4 to 32 Angles Round bars 13
28 Longspan Open Web Joist See Note 14 18 to 32 25 to 64 Angles Angles 4, 14
29 BJ Series See Note 11 24 to 30 24 to 60 See Note 15 Round bars 11, 15
29 BH Series See Note 12 24 to 30 24 to 60 See Note 15 Round bars 12, 15
Gabriel 30 Long Span Joist 18 to 32 24 to 64 Angles Round bars 4
Truscon
19 & 20
O-T (Open Truss) Joists See Note 1 8 to 20 7 to 40“Tee” & M
shaped platesRound bars 1
21 Series AS Joists See Note 2 8 to 24 7 to 48 U shaped Round bars 2
21 Series BB Joists See Note 3 8 to 24 7 to 48 U shaped Round bars 3
22 & 23
Clerespan Joists See Note 6 18 to 32 26 to 64 “Tee” & angles Angles & bars 4, 5, 6
STRUCTURE magazine • June 2009Antiquated Systems Series - Part #9a - Table 2
S T R U C T U
R E®
magazi
ne
Copyright
13
STRUCTURE magazine • June 2009Antiquated Systems Series - Part #9a - Figures
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine • June 2009Antiquated Systems Series - Part #9a - Figures
Figure 7
Figure 8
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine • June 2009Antiquated Systems Series - Part #9a - Figures
Figure 9
Figure 10
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine • June 2009Antiquated Systems Series - Part #9a - Figures
Figure 11
Figure 12
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine • June 2009Antiquated Systems Series - Part #9a - Figures
Figure 13
Figure 14
Figure 15
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine • June 2009Antiquated Systems Series - Part #9a - Figures
Figure 16
Figure 17
Figure 18
Figure 19
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine • June 2009Antiquated Systems Series - Part #9a - Figures
Figure 20
Figure 21
Figure 22
Figure 23
Figure 24
Figure 25
Figure 26
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine • June 2009Antiquated Systems Series - Part #9a - Figures
Figure 27
Figure 28
Figure 29
Figure 30
S T R U C T U
R E®
magazi
ne
Copyright
aid
s fo
r th
e stru
ctur
al e
ngin
eer’s
tool
box
EN
GIN
EER’S
NO
TEBO
OK
STRUCTURE magazine November 200918
Antiquated Structural Systems SeriesPart 9b – Open Web Steel JoistsBy D. Matthew Stuart, P.E., S.E., F. ASCE, SECB
For this series of articles, “antiquated” has been defined as meaning outmoded or discarded for reasons of age. In reality, however, most of the systems that have been discussed are no longer in use simply because they have been replaced by more innovative or more economical methods of construction.This article, however, deals with a method of construction
that is still very much in use today. Nevertheless, the historic, original construction practices described here may still be encountered in existing structures. Therefore, the primary purpose of this series of articles will be fulfilled, which is to compile and disseminate a resource of information to enable structural engineers to share their knowledge of existing structural systems that may no longer be in use but are capable of being adapted or reanalyzed for safe reuse in the marketplace of today and the future.
Evaluation and Modification of Existing Joists
The author would first like to thank the Steel Joist Institute (SJI) for providing much of the material that was used in the development of this article. The evaluation and strengthening of existing open web steel joists and Joist Girders is often required as a result of equipment upgrades or new installations and adaptive reuse or change in use of a facility. The SJI provides an excellent resource for the evaluation and modification of existing joists and Joist Girders in Technical Digest No. #12.The first step in the process of evaluating
an existing joist is to determine the capacity of the member. Ideally, the best method for doing this is through original construc-tion or shop drawings, which allow the identity of the joist to be established. Similarly, it is also sometimes possible to identify the joist by means of fabrication tags left attached to the joists in the field. However, if tags can be found, more often than not the tag only identifies the shop piece mark number rather than the actual joist designation.In some instances, it may only be possible
to establish the type or series of the joist through the available documenta-tion. In this situation, it is possible to assume conservatively that the capacity of the existing joist is no more than the lightest joist in the corresponding series for the given depth. In addition, if it is not clear whether a J- or H-Series joist is involved, the J-Series joist should always be conservatively assumed because of its lower load-carrying capacity. However, if a definitive distinction is required, and it is possible to secure a material sample in order
to obtain results from a standard ASTM tension coupon test, a determination as to whether the joist is 36 ksi (J- Series) or 50 ksi (H-Series) can be made.If no drawings are available, it is still pos-
sible to establish the approximate capacity of the member by field measuring the chord and web member sizes, as well as the overall configuration of the joist. This information can then be used to analyze the structure as a simple truss. Critical assumptions that must be made with this approach include the yield strength of the members and whether the existing panel point welds are capable of developing the full capacity of the connected component members. An alternate method includes filling out the Joist Investigation Form located on the SJI website. SJI has indi-cated considerable success in identifying the series and designation for many older joists with this resource.The next step in the evaluation process
is to determine all of the existing loads on the joist system. The existing and new loading criteria are then used to establish the shear and moment envelope of the individual joist, for comparison with the allowable shear and moment envelope based on either the historical data provided by SJI or an independent analysis of the member as a simple truss. In the former case, unless the joists were fabricated with a uniform shear and moment capacity over the entire span length (i.e., KCS joists), then it is also necessary to evaluate the lo-cation of the maximum imposed moment.Typically, if the maximum moment is
within one foot of the midspan point and the maximum applied moment is less than the joist moment capacity, the
joist is capable of safely supporting the im-posed loads. However, if the maximum moment is greater than one foot from the midspan point, the capacity of the joist may not be sufficient even if the applied moment is less than the specified capacity. This situation can occur for two reasons. First, the moment capacity envelope of the joist may actually be less in regions of the span that are not within one foot of the midspan point. Second, a shift in the moment envelope from that normally associated with a uniformly loaded simple span (and the prerequisite shear envelope) may result in stress reversals in the web members (i.e., from tension to compres-sion) for which the original member was not designed or manufactured. A similar, although typically more advantageous, con-dition also can occur with J- or H-Series joists because of variations in the uniform shear capacity of these members.When the existing joists do not have
sufficient capacity to support the new loads, one of three methods can be used to rectify the condition: load redistribution, adding new joists or beams, or reinforcing the existing joists. Load redistribution involves the installation of a sufficiently stiff member perpendicular to the span of the joist as required to distribute the applied load to enough adjacent joists such that no one joist is overstressed as a result of the new loading. Adding new joists or beams typically involves the installation of an additional framing member parallel to the joist span, such that all or most of the new applied load is supported by the new framing. New self-supporting beams can also be installed perpendicular to the joist span, as required to reduce the original
TYP.
JOIST
HANGER ROD
CONCENTRATED LOAD
L2x2x¼ REQ’D
IF LOAD DOES NOT OCCUR
AT JOIST PANEL POINT OR
> 6” AND EXCEEDS 150#
Figure 1: Typical concentrated load on joist detail.
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine November 200919
span length of the member. Another alternative consists of new independent, self-supporting beam and column frames that avoid the imposition of any new loads on the existing joist framing system. Reinforcing involves the installation of supplemental material to the original joist as required to increase the load-carrying capacity of the member.The key to the successful use of load redis-
tribution is the installation of a structural member that can adequately and predictably distribute the applied load to enough adjacent joists to justify the safe support of the load. A method of calculating the relative stiffness of a distribution member is available in the reference material noted in the online version of this article. In general, if the spacing of the joists is less than approximately 78% of the calculated relative stiffness of the distribution member and the joists, and the length of the distribu-tion member is less than the inverse of the calculated relative stiffness, then the distri-bution member may be considered as rigid enough to calculate the static load reactions to the affected joists.For load redistribution solutions, it is the
author’s preference to use trussed distribution members, rather than individual beams, to ensure adequate transfer of the applied load. Trussed means continuous members located perpendicular to both the bottom and top chords of the existing joists in conjunction with diagonal web members connected to the continuous members at the intersection of the joist chords. The resulting configuration looks like a truss and provides greater stiffness than an individual beam connected to either the bottom or top joist chords alone. The author also recommends that no more than five joists be engaged by any one redistribution member. In addition, the use of pipes for the continu-ous redistribution truss chord members can be advantageous, as this type of section fits neatly through the V-shaped panel point openings created at the intersection of the existing chords and web members. However, load redistribution solutions may be difficult to install, depending on accessibility and the presence of existing MEP systems, ceilings or other appurtenances.As indicated above, adding new joists or
beams to an existing system can also be used to accommodate new loads on an existing joist structure. When new members are added parallel to the existing joists, the new framing can be used either to reduce the tributary area of the existing joists or to provide direct sup-port of the new loads such that there is no impact on the existing joists. Methods used to install new parallel framing often involve manufacturing, shipping and erecting the new members using field splices. However, it is pos-sible to install new full-length manufactured
Figure 2.
joists by means of loose end bearing assem-blies. In this scenario, the joists are first erected on a diagonal to allow the top chord to be lifted above the bearing elevation. The joist is then rotated into an orthogonal position, with the lower portion of the bearing assembly then dropped and welded into place. Typically, in this situation, a shallower bearing seat is also pro-vided for ease of installation and then shimmed once the new joist is in its proper position.When new beams or other similar mem-
bers are added perpendicular to the joist span, the new framing serves to reduce the span of the existing members, thereby increasing the
load-carrying capacity of the joists. However, it is still necessary to analyze the existing joists to ensure that no load reversals have occurred in tension-only web members, and that the actual applied moment falls within the remaining existing moment capacity en-velope of the joist. As with load redistribution solutions, both of the above new framing ap-proaches may be difficult to install.New framing that involves the installation
of independent, stand-alone beam and column frames is intended to provide direct support of the new loads such that there is no impact on the existing joist framing. This type of new
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine November 200920
framing can involve beams located either be-neath or above the impacted existing framing and supported by new columns and founda-tions, or beams that frame between existing columns. This type of solution can also in-volve new beam frames supported from posts located directly above existing beams or col-umns. The above solutions are typically more adaptable to the presence of existing MEP sys-tems, ceilings or other appurtenances.Procedures for reinforcing joists are expertly
described in SJI Technical Digest No. #12 and involve two basic approaches: 1) ignore the strength of the existing member and simply design the new reinforcement to carry all of the applied load, or 2) make use of the strength of the existing member when designing the reinforcing. Both of the recommended ap-proaches typically involve significantly more labor costs than material costs because of the expense associated with field welding.The author prefers to avoid the use of field
reinforcement for the following reasons. A manufactured open web steel joist is basically a pre-engineered product; however, when an engineer involved with the modification of an existing joist specifies new field installed reinforcement, that same engineer assumes the responsibility for the overall adequacy of the joist. This liability extends to not only the reinforcing modifications but also, inherently, to any pre-existing, unknown conditions or deficiencies in the joist. In addition, field welding associated with the installation of reinforcement also poses concerns for the design engineer. Problems associated with field welding are discussed in Technical Digest No. #12 and include temporary localized loss of the mate-rial strength of the existing steel due to heat generated by the weld, induced eccentricities, inadequate load path mechanisms, and lack of access, particularly at the top chord.The only exceptions that the author makes
include the installation of supplemental web members as needed to transfer concentrated loads greater than 150 pounds on chords that are located greater than 6 inches from a panel point to the closest adjacent panel point (Fig-ure 1, page 18), and reinforcement designed by the original manufacturer’s engineer. The first exception is the author’s rule of thumb and is not formally endorsed by SJI, because it is not applicable in all cases; for example, it may be fine for a 30K12, but not for a 10K1.The analysis of existing open web steel joists
can be a challenging undertaking and often involves a considerable amount of detective work. Unfortunately, there is typically little or no documentation available concerning the capacity of a specific existing joist under investigation. However, it is hoped that the ref-erence information provided in the online
version of this article will assist in increasing the likelihood that the capacity of a joist can be determined using the historical data that is available from SJI.Typically, the investigation of an exist-
ing joist results in the need to modify the structural system to provide for the support of new imposed loads. At this juncture, the engineer must then determine if he or she is more comfortable with assuming the re-sponsibility and liability for modifying a pre-engineered product or employing a possibly less risky option, such as load redistribution or adding new joist or beam framing. To assist structural engineers with the evaluation and modification process, the author has included a copy of a flowchart (Figure 2, page 19) that was developed as result of numerous proj-ects that involved existing joists.▪
D. Matthew Stuart, P.E., S.E., F. ASCE, SECB is licensed in 20 states. Matt currently works as a Senior Project Manager at the main office of CMX located in New Jersey, and also serves as Adjunct Professor for the Masters of Structural Engineering Program at Lehigh University, Bethlehem, PA. Mr. Stuart may be contacted at mstuart@CMXEngineering.com.
References80 Years of Open Web Steel Joist ConstructionA Compilation of Specifications and Load Tables Since 1928Steel Joist InstituteMyrtle Beach, South Carolina2009
Catalog of Standard Specifications, Load Tables and Weight Tables for Steel Joists and Joist Girders42nd EditionSteel Joist InstituteMyrtle Beach, South Carolina2007
Technical Digest No. #12Evaluation and Modification of Open Web Steel Joists and Joist GirdersSteel Joist InstituteMyrtle Beach, South Carolina2007
Designing with Steel Joists, Joist Girders and Steel DeckJames Fisher, Michael West, Julius Van de PasNucor Corporation1991 (Second Edition – 2002)
Miscellaneous Steel Joist and Joist Girder Specifications and Load TablesSJI ArchivesSteel Joist Institute, Technology, Engineering, and Education CenterMyrtle Beach, South Carolina
Joist Investigation Formwww.steeljoist.org/investigationSteel Joist InstituteMyrtle Beach, South Carolina
S T R U C T U
R E®
magazi
ne
Copyright
aid
s fo
r th
e stru
ctur
al e
ngin
eer’s
tool
box
EN
GIN
EER’S
NO
TEBO
OK
STRUCTURE magazine January 201024
Antiquated Structural Systems SeriesPart 10
By D. Matthew Stuart, P.E., S.E., F. ASCE, SECB
This article includes a compilation of miscellaneous systems and information for use by the practicing engineer. It is hoped that this final article – along with the previous nine – has provided a resource of information to structural engineers involved with the renovation of existing structural systems that are capable of being adapted or reanalyzed for safe reuse in the marketplace of today and the future.
Additional Antiquated Systems
Masonry
Masonry bearing walls were rarely, if ever, designed for actual loading conditions. However, analysis of a typical 8-inch, double-wythe brick wall for a three- to five-story building indicates that the compressive stresses are well below the al-lowable values that were common in the 20th Century.
Building codes in New York City first addressed masonry walls in 1830. The code provisions for brick became more complicated with each revision and, by 1892, the portion of the code dealing with masonry was its most complex part. The NYC code, like many other codes from different major cities, specified the minimum wall thickness for varying heights of buildings. The 1892 NYC code generally called for an increase of 4 inches (i.e., one wythe of brick) in wall thickness for each 15 feet down from the top of the building. The minimum thickness for “curtain” masonry brick walls was generally 4 inches less than that required for load-bearing walls at the same height of the building. As it is sometimes difficult to ascertain the thickness of brick masonry walls in existing buildings, a listing of the various minimum wall thicknesses
is provided here (Table 1, page 26) for a number of major cities from the 1920s.Load-bearing brick masonry
walls were eventually replaced by cage and skeleton wrought iron and steel frame con-struction, often using cast iron columns. Cage construction involved the use of brick façade walls that were as thick as those used for load-bearing construc-tion; the only difference was that the frame and supporting columns, including those that would eventually be embedded in the brick masonry façade wall, were first erected ahead of the masonry. Skeleton framing,
although partially embedded in the exterior masonry walls, was only clad with what amounted to a brick curtain wall. All three of these forms of construction co-existed between 1880 and 1900.
Floor Framing
Draped mesh slabs became popular in the 1920s. Draped mesh construction is a type of reinforced slab framing that involves the use of wires that drape between the tops of adjacent beams. The types of mesh used included triangular wire mesh, ordinary wire mesh, expanded metal sheets, plain round and square rods and twisted square rods. The use of wire mesh was actually preceded by expanded metal sheets. Welding of wires together to form the mesh did not begin until the 1930s. Prior to that, the wires were attached at the intersection points by staples or washers, or by wrapping the transverse wires around the longitudinal wires.In a draped mesh slab, the concrete
serves only as the wear surface and as the mechanism by which the imposed loads are transmitted to the mesh. The mesh alone is what physically spans between the beams by means of catenary action. Because the concrete is not structurally stressed in this type of system, the com-position and quality of the concrete is not as important as in a true flexural slab. As a result, it was common to use cinder con-crete with compressive strengths under
20 th Century Loadbearing Masonry Building.
Draped Mesh Floor.
For this series of articles, “antiquated” has been defined as meaning outmoded or discarded for reasons of age. In reality, however, most of the systems that have been discussed are no longer in use simply because they have been replaced by more innovative or more economical methods of construction.
Web ResourcesAdditional information concerning draped mesh construction can be found
in the Practice Points archive of The Association for Preservation Technology website. www.apti.org/publications/PP-archive/Friedman-PPs.pdf
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine January 201025
Brick arch floor construction consisted of a single arch of unmortared brick, typically only one wythe or 4 inches thick, capable of spanning 4 to 8 feet with a center rise of approximately ⅛ of the span. The spring line of the arch was constructed on top of the bottom flange of the supporting beams. The space above the arch was filled in with concrete, which sometimes had wood nailer strips embedded in the top of the slab. Tie rods were commonly placed about ⅓ of the height of the beam and were spaced from 4 to 6 feet on center. The entire system had to be built on formwork, which supported the brick. The thrust (T) on the arch, in pounds per linear foot, can be calculated as follows:
T = (1.5 x W x L2)/R
Where:W = load on the arch in PSFL = span length of the arch in feetR = rise of the arch in inches
Other common floor systems included:Fawcett System and Acme Floor-Arch –
clay lateral cylindrical tile flat-end construction arch.Rapp Floor and McCabe Floor – gauge-steel
inverted tees spaced at approximately 8 inches on center, supporting a layer of brick and upper cinder concrete slab spanning 4 feet between supporting beams.
BRICK FLOOR-ARCH
CONCRETE FINISHED FLOOR
TYPE (A)TYPE (B)
TIE RODTERRA-COTTASKEW-BACKFLANGE PROTECTION
WITHOUT SKEW-BACKFLANGE PROTECTION
Brick Arch Floor.
The Kahn System.
A series of deformed bars.
Roebling Floor Arch – arch of dense wire mesh supported on the top of the bottom flanges of the beams covered with concrete.Manhattan System and Expanded Metal
Company (EMC) Floor – flat and arched (for EMC) expanded metal mesh covered with concrete. Multiplex Steel-Plate, Buckeye and Pencoyd
Corrugated Floor – Riveted steel plates supporting a concrete slab.Thompson Floor – Unreinforced concrete
slab spanning approximately 3.5 feet between beams connected with tie rods.Roebling Flat Slab Floor and Columbian Floor
System – reinforced concrete slab.Metropolitan System – early draped mesh
floor system.Plain round and square bars were typically
used in reinforced concrete buildings built before 1920. Plain bars began to be phased out during the 1910s and early 1920s in favor of deformed bars. The two types of deformations used at that time included longitudinal and radial deformations. In addition, the Ransome bar included deformations induced by twisting square bars.Other forms of longitudinally deformed bars
included: the Thatcher bar, which was a square
bar with cross-shaped deformations on each face; the Lug bar, which was a square bar with small round projections at the corners; the Inland bar, which was a square bar with raised stars on each face; the Herringbone, Monotype and Elcannes bars, which included complex cross-sections similar to radial deformed bars, but with longitudinal deformations; the Havemeyer bar, which included round, square and flat cross-sections with diamond-plate-type deformations; the Rib bar, which included a hexagonal cross-section with cup-shaped deformations; the American bar with square and round cross-sections and low circumferential depressions; the Scofield bar with an oval cross-section and discontinuous circumferential ribs; the Corrugated bar with flat, round and square cross-sections with cup deformations; the Slant bar with a flat cross-section and low projecting diagonal ribs on the flat faces; the Cup bar with a round cross-section and cup deformations; and the Diamond bar with a round cross-section and low circumferential ribs. The modern designation of #3 to #8 round cup or diamond deformed bars was established in 1924.Reinforcing for concrete beams was also avail-
able in prefabricated trussed bar units. A truss
1,000 psi. The use of cinder concrete, how-ever, due to the acidic nature of the clinker (coal cinder) used as the aggregate, resulted in the corrosion of the embedded iron beams and reinforcing mesh. Catenary systems are also vulnerable to collapse as a result of failure of the wire anchorages.
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine January 201026
bar is essentially a top bar at the ends of a beam that is bent diagonally down to a bottom bar position at midspan. Prefabricated assemblies included the Kahn System, the Cumming System, the Corr System, the Hennebique System, the Pin-Connected System, the Luten Truss and the Xpantruss System.
ConclusionsEngineers involved with renovation and re-
habilitation projects need to be aware of the specifics of antiquated structural systems in order to develop non-destructive and unob-trusive solutions. This approach enables the project to be more economically viable because of the extent of structural costs associated with a typical renovation project. In other words, without any knowledge of an existing structural system, it is still possible to develop a structural solution; however, this approach will always be much more intrusive, and therefore more costly, than if the engineer has a sound under-standing of the system involved.Information concerning antiquated structural
systems provided by this series of articles, and the referenced source material, has been com-piled and made available because the history of structural systems is far less documented than the history of architecture. This lack of documentation can be traced to the general public’s lack of awareness about the hidden structural components of a building, which are typically enclosed after erection by the architectural finishes and therefore of less in-terest to a casual observer.This general lack of readily available in-
formation on antiquated structural systems has occurred despite the fact that most of the methods of analysis and materials used in this country, including steel and concrete, are not much older than 100 years. At the same time that new materials, technologies and methods of analysis have become available and readily embraced by design engineers and the con-struction industry, previously used systems were, more often than not, quickly discarded and forgotten.The information that has been presented
in this series is intended to represent the knowledge that has been available at various stages of different methods of construction over the past century or so in the United States. However, this information cannot be used from a perspective in which any framing system can be assumed to correspond precisely to a specific system described in the material presented. As is still the case now, the fact that records indicate that a particular structural component should be able to support a given load does not mean that errors were not made during the original construction or as a part of the initial design.
Total Stories
CityFloor
1st 2nd 3rd 4th 5th 6th 7th 8th
2 Boston 12 12
New York 12 12
Chicago 12 12
Philadelphia 13 13
Denver 12 12
San Francisco 17 13
3 Boston 12 12 12
New York 16 16 12
Chicago 16 12 12
Philadelphia 18 13 13
Denver 16 12 12
San Francisco 17 17 13
4 Boston 16 12 12 12
New York 16 16 16 12
Chicago 20 16 16 12
Philadelphia 18 18 13 13
Denver 16 16 12 12
San Francisco 17 17 17 13
5 Boston 16 16 12 12 12
New York 20 16 16 16 16
Chicago 20 20 16 16 16
Philadelphia 22 18 18 13 13
Denver 20 20 16 16 12
San Francisco 21 17 17 17 13
6 Boston 16 16 16 12 12 12
New York 24 20 20 16 16 16
Chicago 20 20 20 16 16 16
Philadelphia 22 22 18 18 13 13
Denver 20 20 20 16 16 12
San Francisco 21 21 17 17 17 13
7 Boston 20 16 16 16 12 12 12
New York 28 24 24 20 20 16 16
Chicago 20 20 20 20 16 16 16
Philadelphia 26 22 22 18 18 13 13
Denver 24 20 20 20 16 16 12
8 Boston 20 20 16 16 16 12 12 12
New York 32 28 24 24 20 20 16 16
Chicago 24 24 20 20 20 16 16 16
Philadelphia 26 26 22 22 18 18 13 13
Denver 24 24 20 20 20 16 16 12
Table 1: Minimum Building Code Thickness of Brick Masonry Walls – Inches.
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine January 2010
In addition, it is common to encounter some overlap between a previous and more recent method of construction, which has resulted in a blending of two otherwise discrete structural systems. Also, before ASTM began to standardize construction materials in the late 1890s, the quality of irons, steels and cementious products varied greatly. Therefore, when dealing with a building that predates ASTM testing, samples of the existing structural materials should be obtained and tested as a part of the due diligence effort.In the absence of existing drawings, the
methods of evaluating the properties of an existing system include core samples for cementious material strength, depth and/or thickness; coupons to determine iron or steel tensile strength; x-rays to determine internal reinforcement; petrographic analysis to deter-mine the quality, condition and consistency of concrete; ground-penetrating radar (GPR); profometer to determine the location of internal reinforcement; Schmidt hammer to determine in situ concrete compressive strength; explor-atory demolition; and in situ load tests.In some instances, it is not possible or not
practical to obtain material strength properties of an existing system in order to complete an analysis using current methods. However, if the past performance of the structure has been good (i.e., no signs of distress or significant deterioration), then it is very likely that the system is adequate for the same use in the future. In such situations, however, it is helpful to try and determine what the likely original live load designation was for comparison to the planned current use.If the engineer can determine what the likely
original use of the building was and has access to copies of older building codes, it is some-times possible to determine the original live
Minimum Building Code Live Load - PSF
Building Type New York Philadelphia Boston Chicago Denver San Francisco
1927 1929 1926 1928 1927 1928
Residential 40 40 50 40 40 & 60 40
Hotels, Hospitals 40 40 50 40 90 40
Office Buildings:
First Floor 100 100 125 125 125 125
Upper Floors 60 60 60 40 70 & 90 40
Classrooms 75 50 50 75 75 75
Public Seating:
Fixed Seats 100 60 100 75 90 75
Without Fixed Seats 100 100 100 125 120 125
Garages:
Public 120 100 150 100 150 100
Private 120 100 75 100 150 100
Warehouses 120 150 125-250
125-250
200 125-250
Manufacturing:
Heavy 120 200 250 250 250 250
Light 120 120 125 125 120 125
Stores:
Wholesale 120 110 250 250 120 125
Retail 120 110 125 125 120 100
Sidewalks 300 120 250 150 150 150
Table 2.
27
AD
VERTISEM
ENT - For A
dvertiser Inform
ation, visit w
ww
.STR
UC
TUREm
ag.o
rg
19 th Century Train Station (30 th Street Station, Philadelphia, PA).
T y f o® F i b r w r a p® S y s t e m s
FYFECo.
LLC
NSFNSFNSFR
Certifi ed to NSF/ANSI 61
Over 20 years ago we created the industry...
today we set the standard
Structural Strengthening
• FRP Installation
• Seismic Upgrade
• Blast Mitigation
• Concrete Retrofi t• Specialty Gunite
• Underwater & Coastal Repairs
• Expansion & Seismic Joints
Pipe Repair and Renewal
• Large and Small Diameter
• PCCP, RCP, Steel Structural Repairs
• Carbon Fiber Structural Liners
Concrete Restoration
• Epoxy Crack Injection
• Spall Repair
• Corrosion Protection
Advanced Fire Protection
8380 Miralani Drive, San Diego, CA 92126
Tel: 858.642.0694 Fax: 858.444.2982
www.fyfeco.com
S T R U C T U
R E®
magazi
ne
Copyright
STRUCTURE magazine January 2010
D. Matthew Stuart, P.E., S.E., F. ASCE, SECB is licensed in 20 states. Matt currently works as a Senior Project Manager at the main office of CMX located in New Jersey. Mr. Stuart may be contacted at mstuart@CMXEngineering.com.
load for comparison to the proposed adaptive reuse. Individual building codes were com-monly developed by different cities before the advent of national codes. These local codes often reflected different allowable strengths for the same building materials and varying degrees of minimum live loads. Table 2 (page 27) is an example of minimum live loads for a number of major cities.The original criteria for the design of antiquated
structural systems was a performance-based ap-proach grounded in experience, both good and bad (i.e., successes and failures). The transition to the more recent analytical design approach has come about through the development of strength-based formulas derived from scientific experimentation and tests. Structural engi-
neering of buildings as a separate discipline did not exist as late as the 1840s. However, the need for engineers began to grow in the 1850s with the advent of wrought iron beams, which had to be mathematically designed be-cause there was no craft tradition to provide rules of thumb. In addition, the establishment of ASCE in 1852 helped to promote the rapid spread of technical information, such as re-cords of experiments with cast and wrought iron performed in England by Hodgkinson and Fairbairn.It should also be recognized that an existing
structural system can often be found to have two different load-carrying capacities – one found using the original codes and methods of analysis, and another using the current codes and methods of analysis. The differences between these two approaches can typically be explained by the expansion of knowledge in the field of structural engineering. More often than not, comparisons between the original and more current methods of analysis will reveal that the older design was conservative. In any case, if the properties of the materials can be substantiated, it is always possible to analyze an older structure using the latest methods of analysis and most current codes. In most cases, in fact, the current building code will mandate such an approach.In situations in which it is confirmed that
the existing structural system does not have sufficient capacity to support the new loads, there are two basic methods that can be used to rectify the condition: adding new framing members, either to support the new loads inde-pendently or to provide supplemental support of the existing structure; and/or internally or externally reinforcing the existing system.▪
GPR printout.
Current Day, Jim Thorpe, PA. Strengthening of existing slab (Slag Block System).
28
Early 20 th Century Retail Arcade (Downtown Nashville Arcade).
Reference MaterialArchitects’ and Builders’ Handbook, 18th
Edition, Frank E. Kidder, Harry Parker, John Wiley and Sons, 1956
Historical Building Construction, Donald Friedman, W. W. Norton & Company, 1995
Structural Engineers’ Handbook, Milo S. Ketchum, C.E., McGraw-Hill Book Company, Inc., Second Edition, 1918
Handbook of Building Construction, Volume II, George A. Hool, S.B., McGraw-Hill Book Company, Inc., 1920
S T R U C T U
R E®
magazi
ne
Copyright
Recommended