Antiquated Structural Systems (All)

Antiquated Structural Systems Series
By D. Matthew Stuart, P.E., S.E., F.ASCE, SECB his article is the rst in a series The purpose of this series is to compile that is intended to provide a and disseminate information that will resource of information to structural enable structural engineers to share their engineers that they can refer to for pro- knowledge of existing structural systems jects that involve the repair, restoration that may no longer be in use but are or adaptive reuse of older buildings for capable of being adapted or reanalyzed which no drawings exist. for safe reuse today and in the future. As developable land becomes more difThe Circumferential or cult to nd, particularly in densely popS.M.I System of Reinforced ulated urban cities or suburban areas in Concrete Flat Slabs which open space cannot be used, owners and developers are increasingly turning to The S.M.I. System of designing reexisting facilities to convert into new uses. inforced concrete flat plate slabs was t igh If no drawings are available for an older developed by Edward Smulski, a conpyr building, a structural engineer will typi- sulting engineer from New York City, Co cally rst turn to industry resources to try prior to the 1920s. The system was and determine the nature and capacity of unique in that the primary exural reinthe existing structural system. Any avail- forcement consisted of concentric rings able information is then used to conrm of smooth reinforcing bars supplementthat the facility meets the current build- ed with diagonal and orthogonal trussed ing code requirements, or determine what bars placed between the supporting colstrengthening or remediation must occur umns and radial hairpin bars located at to accommodate the new use intended by the columns. the architect or owner. The author rst encountered this type If no information is available, the of system while evaluating an existing structural engineer must resort to either structure in Philadelphia that had at one expensive non-destructive testing or ex- time been used as an enclosed parking ploratory demolition methods to try and garage, but was being used as an ofce ascertain the nature and capacity of the building in the late 1990s. No drawstructure. In some cases, it becomes ings were available for the structure, but necessary to abandon parts of the build- small openings cut in the slab revealed ing in place and construct independent portions of the internal reinforcement structures around the existing one in and slab thickness to enable an analyorder to support any new imposed loads sis of the load carrying capacity of the or uses safely.
ENGINEERS NOTEBOOK
aids for the structural engineers toolbox
Rmag T S
208 Smulski on the S-M-I System of Flat-Slab Construction.
Unit A Unit A
Unit C Unit T Unit C Unit T Unit C
Truss Bars
Cazi U
Unit
Ue Tn
Unit A
B Unit
Unit A
Unit
Unit Ce
Unit T
Unit Ce
Unit T
Unit Ce
Unit
Be
Unit
Be
Unit Ae
Unit Ae
Unit Ae
Cw
Unit
Tw
Cw
Unit
Tw
Cw
Note: North half of Unit Ce same as Unit CNote: Heavy lines indicate top steel Light lines indicate bottom steel
Section through column head. STRUCTURE magazine
framed oors. However, rather than the typical orthogonal reinforcing bars, the exploratory demolition discovered rings of smooth bars. A subsequent investigation of the available literature on at plate construction from the approximate time period during which the structure had been built revealed that the slab was very likely designed and constructed using 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 independent 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 xed at the column and loaded uniformly around its circumference by reactions transmitted to it by the adjacent surrounding components. 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 exural 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
E R
Unit T
Unit T
Unit A
Unit A
Unit T
Unit T
Unit T
Unit T
Radials
Top Rings
Bottom Rings
Unit Aw
Unit Aw
45 September 2007
Section thru Column Head
Center Rings
226 Smulski on the S-M-I System of Flat-Slab Construction.

I
1/5 1
Actual cantilever
Theoretical Cantilever
Di a
go na lc
II Line of III inflection I II
lea
rs pa n
II
1
1/ 5 1
Circular Cantilever
r
Span
Suspended Slab
1/5 o 3/5 o
Cop Circular
Cantilever
h yrig
1/5 o
Flat Slab Divided into Simple Parts.
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, indicates that the S.M.I System required 20 to 24% less reinforcing than comparable twoway and four-way at slab systems designed during the same historical time period. Comparisons between weights of reinforcing for different two-way and four-way at 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 appeared in the 1918 ACI Journal Proceedings. An extensometer developed by Professor Claude Berry of the University of Pennsylvania 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
Rmag T S
Clear Span o
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 circumference As = 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 inection 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 inection 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
Cazi U
r
Ue Tn
References
typically based on diagonal clear span between the inection 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 Ofce 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 difculty associated with properly fabricating and placing perfectly round and concentrically positioned bars in overlapping top and bottom layers. 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 ofce of Schoor DePalma Engineers and Consultants, located in New Jersey, and may be contacted at mstuart@schoordepalma.com.
E R
46
September 2007
Concrete 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 & Sons, 1947 Graphics reprinted courtesy of ACI Journal Proceeding, 1918. A Test of the S-M-I System of Flat-Slab by Edward Smulski.
STRUCTURE magazine
Figure 1: Reprinted from Tile Engineering Handbook of Design.
Antiquated Structural Systems Series t

h By D. Matthew Stuart, P.E., S.E., F.ASCE, SECB rig opy C
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). 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.
ENGINEERS NOTEBOOK
A R
T F
Clay Tile Arched Floor Systems

Concrete and steel framed oors 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 soft of the exposed tiles. These types of oor 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 re protection for the structural steel. There are two basic types of clay tile arched oor systems; segmental and at. Both systems were constructed using hollow clay tiles of varying sizes and shapes, with internal open cells similar to todays hollow masonry blocks. The typical web and face shell thickness was -inch, and all four sides of the closed
faces of the tile were also typically scored. The blocks were manufactured by a number of different companies, including: National Fireproong Corporation, Pittsburgh; Henry Maurer & Sons, New York; Whitacre-Greer Fireproong 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 prole 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 at 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 at 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 exural member rather than as an arch. This system was
STRUCTURE magazine
18
December 2007
yr Cop
igh
also referred to as the Natco New York reinforced at 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 largespan domes. As this type of construction was not used in conjunction with steel oor framing, it will not be discussed as a part of this article. Standard at arches can be classied 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 soft tiles beneath the beam bottom ange for re protection. The tie rods used to resist the arch thrust forces were generally placed approximately 3 inches from the bottom of the beams in
at arches. In the case of segmental arches, placing the tie rod near the bottom of the beams resulted in the tie continued on next page rod being exposed across the Table 1 horizontal spring line of the arch. Rod Diameter 5/8 inch 3/4 inch 7/8 inch 1 inch If re resistance of the tie rod was Net Area (a) .202 .302 .420 .550 required, it was more often than
A R
T F
not placed higher up from the bottom of the beam as required to avoid exposure. Typically tie rods were -inch in diameter and were spaced as required to resist the specic thrust of the given arch span, although a minimum spacing of fteen 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 spans rods should be veried 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 at and segmental arch can be found as indicated on the following page.
h i g h c a pac i t y
MiNi caiSSONS
As seen in use on a recent NYC project.
ADVERTISEMENT For Advertiser Information, visit www.STRUCTUREmag.org
SAS Stressteel salutes its partners Hayward Baker and Heller & Johnsen for their assistance in another successful project application.
Felix Ferrer, P.E., President 100 New Dutch Lane Fairield, NJ 07004 973-244-5995
STRUCTURE magazine
19 December 2007
Total Thrust (in pounds) per Arch Panel: P = (3wD2/2R)L Where; 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 at arches) L = length of the oor beam supporting the arch in feet Total net area of tie rods per panel (square inches): A = P/f Where; 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) 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.
Load tables published in Kidder and Parkers 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 ton published tables. If, h however, this rig py simplied approach indicates o that theCallowable load carrying capacity is not sufcient 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.
A R
T F
The principal disadvantage of tile arch oor construction was the difcultly of adapting standard sizes to irregularly shaped spaces. In addition, tile arches are more easily weakened by holes and penetrations than a monolithic oor system. Furthermore, it was difcult 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 (sideconstructed 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 conguration. 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. 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 ofce 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 oneand 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).
STRUCTURE magazine
20
December 2007

Part 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.
One- and Two-Way Clay Tile and Unit Masonry Joist Systems
In 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 re resistance, and to provide a at ceiling surface. Proprietary one-way oor systems included the Natcoor and Republic Slagblock systems. Proprietary two-way oor 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 SmoothCeiling 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 1 inches to 3 inches. Joists were typically analyzed as T-beam sections when a monolithic topping was used. With the
exception of the Natcoor system, joist t igh widths typically rvaried from 4 inches to py 6 inches. Co Typically, -inch clear cover was provided between the square or round deformed reinforcing bars and the adjacent 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 temperature/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 efcient for spans over 12 feet, and were used very frequently for spans up to 24 feet with
A R
Flashing
T F
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 lled with cardboard or metal inserts to prevent concrete from owing into the voids, in order to minimize the dead load of the slab. The Natcoor system used specially manufactured clay tiles with curved anges that allowed only the bottom of the tiles to be exposed as the ceiling soft. 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 soft pieces along the bottom of the trenches that resulted in a uniform tile ceiling soft. The Natcoor 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,
ENGINEERS NOTEBOOK
Ground
Plaster
Base Flooring
Wood Sleepers Cinder Concrete Fill
Finished Top
Wood Forms
Figure 1.
STRUCTURE magazine
45 March 2008
Figure 2.
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 installed in either a one-way or two-way conguration. 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-, 4-, 6-, 7- and 8-inch depths. The concrete ribs or joists were typically 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 patented in the 1930s, and the similar Sandberg system both eliminated the need for beams or drop panels by employing embedded internal steel shear reinforcement around either structural steel or right reinforced concrete columns. As with Copy two-way systems, standard tile other units were placed in a modular layout in order to establish a uniform two-way grid of concrete 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 considerable 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 coefcients were typically employed to establish the maximum positive and negative moment
A R
T F
and end shear design envelopes; however, continuous beam analysis was also used to establish the required design parameters. Moment and shear coefcients 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 difcult 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 abovedescribed systems are encountered in an existing building. 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 ofce of CMX located in New Jersey and may be contacted at mstuart@CMXEngineering.com. The online version of this article, www.STRUCTUREmag.org, contains detailed references.
STRUCTURE magazine
46
March 2008

Part 4
By 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. span The prefabricated clay tile systems span tile tile t included both one-way beam and slab (a) exposed (b) flat ceiling igh construction oandr one-way slab conpy beam ceiling C struction. The one-way beam system involved the placement of prefabricated beams spaced parallel to each other at regular intervals between already Concrete Topping constructed load-bearing walls or steel beams. The areas between the beams were then inlled with tiles that were Precast T beams Filler Tile capable of spanning between each adjacent beam. The one-way slab system Figure 1. involved prefabricated slab units that were placed directly adjacent to each other, spanning between previously Prefabricated Clay Tile constructed load-bearing walls, steel & Concrete Block beams or joists. Both the beam and Framing Systems slab systems included a site-cast conAs discussed in the last article in this crete topping, which was poured over series, one and two-way joist framing the beams and ller tiles or one-way systems were constructed using clay tile slabs. The prefabricated beam and and masonry units, which were rst ar- slab systems offered the advantage of ranged and supported on formwork to not having to construct supporting enable placement of internal reinforce- formwork before the framing could be ment and inll and topping concrete erected; however, shoring in the center in situ. In addition, as will be discussed of the span was sometimes employed in this article, similar modular clay tile to increase the clear span capability of and masonry units were also constructed the members through composite acoffsite into prefabricated beams and tion with the site-cast topping. Examples of the clay tile systems (Figslabs that could be delivered to the job site. This method of construction ulti- ures 1-6) included the T Beam Floor, mately progressed to solid precast concrete the U Beam System, the Joistile System, the Shefeld Floor System, the Adel Joistile 282" 282" 2" 2" 12" 2" System, the Kalex Floor System, the United Floor 212" 212" System and the Tilecrete Section A-A Section B-B Floor System. Some of these systems required lling the joints between the adjacent Concrete Topping ends of the clay tile units with mortar, while other Fille systems allowed the ends rT ile Precast Economy of adjacent tiles to butt U beam up against each other. All of Perspective view of Precast U-Beam Floor and sectional details the prefabricated beam and showing optional erection methods with span tile webs up or slab systems, except for the down as desired. inll tiles, included internal Figure 2.
123
1 12 or 2 ,
ENGINEERS NOTEBOOK
4", 5" or 6"

4", 5" or 6"
12
12
Concrete Topping
Precast tile joists
Rmag T S
92" 82" 6" 6"
Cazi U
83"
Ue Tn
Figure 3.
Concrete Topping Precast tile sections
83"
Perspective view of standard Precast Joistile floor system and typical sectional detail.
E R
Perspective view of Sheffield precast tile floor system and sectional detail.
Figure 4.
72"
1, 1 or 2 concrete slab
Typical slab section Concrete Topping
Precast tile oor slab

Perspective view of Adel precast joistile floor system and typical slab section.
Figure 5.
(a) 4-in. Kalex floor section.
(b) 6-in. Kalex floor section.
Concrete Topping
Precast Kalex slab section

Perspective view of precast Kalex tile floor system and typical floor sections.
Figure 6. longitudinal exural reinforcement for positive moment resistance. Negative moment reinforcement for continuity across a supporting wall, beam or joist
STRUCTURE magazine
49 June 2008
TOPPING AS REQD 1 5/8 MIN
1 1/4
Any type floor covering can be used Cement topping increases strength Steel reinforcing rods provide structural strength.
3 7/16
15
1 1/4 1 1/2
13/16
15
3/16
1 3/8
3/16
3/16
Figure 7. 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 conventional internal reinforcement for exural strength; however, a few were developed using prestressed bars and strands. Probably the most widely used masonry block product in the eastern U.S. during the 1950s was the Dox Plank system developed by NABCO in Washington, DC (Figure 7). This product was manufactured with recessed slots in the bottom of the
Table 1.
3/4 R
5/8
3/16
2 11/16
1 9/16
Figure 8.
8 3/16 15 15/16
1 9/16
block to allow for the exural reinforcement to be grouted into the bottom of the plank. There was no mortar required between the adjacent ends of each block.ht yrig Other cross-sectional variations of the Dox Cop 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
System
T Beam
U Beam Joistile Shefeld
R T S
Figure 1 2 Texas, Oklahoma East, Southwest 3 4 5 6 None None NOTES:
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
Prefabricated Clay Tile One-Way Beam and Slab Systems
Regional Use
Southwest, Midwest
g a m
Mortared Joints Yes No Yes Yes Unknown No No No
Cazi U
Beam Spacing or Slab Width 18 to 30 182 to 222 282 12 8 12 12 Unknown 16
Ue Tn
Beam Depth (including topping) x Width or Slab Depth 92 x 8 72 to 8 x 8 82 to 12 x 7 to 8 4 and 5 6 5 5 4 and 6 Unknown 4 and 6
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.
E R
1 15/16
Typical Span 24-ft. 14-ft. 8-ft. 12-ft. 18-ft. to 22-ft. Unknown Unknown 30 12
Notes 1, 3 2, 3 4, 5 4 6 4 7, 8, 9, 10 11, 12 13, 14
Midwest, North Central Midwest Ohio, Pennsylvania, New York, Illinois, Wisconsin New York, New Jersey Missouri
Adel Joistile Kalex United Tilecrete
Included the use of unreinforced, 4 thick Filler Span Tile for one-way span between beams. Drop-in Filler Tile was used for ush ceiling applications. System load tested at Iowa State College; Engineering Experiment Bulletin No. 286. Longer spans were possible with the use of center span shoring. Included the use of unreinforced, 2 thick ribbed Filler Tile for one-way space between beams. Patented June 1936 by Professor Walter M. Dunagan, Iowa State College. Patented 1937 by D.D. Whitacre, Waynesburg, Ohio. System also used as vertical wall element. Reinforcement included bolted rods, which implies an applied pretensioning force. Load tests of 4 slabs conducted by Professor George E. Large, Ohio State University, for the Rochester, NY Building Board in 1939. Unreinforced slab system used in conjunction with open web steel joists. System used in conjunction with a topping slab that provided composite action with steel joists. 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.
STRUCTURE magazine
50
June 2008
1 5/16
1 3/8 4
1/4 R
3 1/2
11/16
30
1 5/16
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 rm of Bryan and Dozier and the Nashville Breeko Block Company designed and constructed prefabricated, post-tensioned concrete block beams. This method of construction resulted in the rst linear prestressed structure to be built in the US the Fayetteville Tennessee High School Stadium and the rst 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 expensive end anchorages. The Breeko Block system was further rened through the use of external, deected tendons. However, by the late 1950s, this system was replaced by precast, pretensioned concrete members. Other, more obscure examples of prefabricated modular concrete block beams and slabs include a prestressed bar system developed 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 patented 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 published 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 difcult to determine the internal reinforcement and, subsequently, 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 another, will assist the readers in their research of an antiquated or archaic system when it is encountered in an existing structure. 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 ofce 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 addressed, please send your suggestions and any relevant information that you have to the author.
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.
Rmag T S
SOLUTIONS FOR:
Ca U
Low Concrete Strength Honeycomb / Void Repair Structural Failure / Damage Missing or Misplaced Reinforcement New Slab Penetrations Increased Loads Change in Code Seismic Upgrade Blast Mitigation
We can e fix it zin
Cop
h yrig
U T
E R
ADVERTISEMENT - For Advertiser Information, visit www.STRUCTUREmag.org
STRUCTURAL STRENGTHENING SOLUTIONS

FROM AMERICAS LARGEST SPECIALTY REPAIR CONTRACTOR
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.
Thats where we come inStructural 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 websitereview 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 800899-1016.
Offices Nationwide
www.spsrepair.com/fix2 800-899-1016
STRUCTURE magazine
51 June 2008

Part 5
ENGINEERS NOTEBOOK
By D. Matthew Stuart, P.E., S.E., F. ASCE, SECB This article is the fth 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. used conventionally reinforced precast concrete inverted T-joists spaced at 28 inches on center, which supported concrete block ller slabs. The entire assembly then received a 2-inch, castin-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 t ighto 36 feet for load capacities from 30 to 900 pounds per square pyr foot, depending on the span, depth of joist and reinforcement. Co The ends of the precast joists could be cast integral with a site Figure 1: F&A System. cast concrete beam, or bear directly on either precast concrete he purpose of this series is to compile and disseminate girders or steel beams. The F&A System included ller blocks a resource of information to enable structural engi- that could be either placed ush with the bottom of the joist or neers to share their knowledge of existing structural recessed at the same level as the bearing ledge of the joist. Similar precast concrete systems that were in use during the systems that may no longer be in use, but are capable of being adapted or reanalyzed for safe reuse in the marketplace of today same approximate time period included Tee joists and Keystone joists. Tee joists (see Figure 3) came in depths of 16 and and the future. 20 inches and were typically prestressed. Keystone joists (see Precast Concrete Figure 4) were available in 8- and 12-inch depths and could
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)
Rmag T S
Framing Systems
16
Cazi U
12 4 2 Up to 15 cantilever with high steel locations Cantilevers can be tapered
Ue Tn
Rein. concrete poured in place tie beam on block wall Flange bearing optional
E R
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
Ledger Beam Block Wall
Figure 3: Tee Joist.

grout 5 built-up roofing
prefabricated roof deck 12 C. G. C.
6.47
PHYSICAL PROPERTIES A = 51 in.2 I = 601 in.4 Sb = 92.9 in.3 St = 108.7 in. 3 Yb = 6.47 in.
Figure 2: Rutten System.
Figure 4: Keystone Joist.
STRUCTURE magazine
25 September 2008
be either conventionally reFLOOR TILE CEILING TILE inforced or prestressed. Tee joists ultimately evolved into Single Tees, Quad Tees and 24 24 the Double Tee member that is still in common use today. 29 Channel Slabs were another prevalent precast concrete 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 3-inchdeep by 2-inch-wide, downt turned edge anges. This Figure 5: Gypsteel Floor & Ceiling Slab.righ py product was capable of spanCo ning up to 9 feet and supporting up to 60 pounds per square foot of superimposed load.
Examples of other proprietary precast systems that are no longer in use include: Gypsteel Floor and Ceiling Slabs (Figure 5): The oor 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 lled with grout for a smooth top nish. 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
Concrete Beam Covering
Figure 6: Waites I Beam.
Rmag T S
Wire Lath
on FilCement T l to c il over e or Tera zz Pipes & cona duits
g on Sl Floorin
eepers
Cope on Concrete I
Rod Reinforcement
Channel Reinforcement
Cazi U
Factory made Unit Concrete Floor Beam
Field Concrete Fireproofing
Ue Tn
F C ield Fire oncret pro e ofin g
Stee l
I Be
am
E R
TYPE A
Suspended Ceiling
lI Stee
m Bea
Fac to
dard Stant Slab ni U
ry m
ade
Uni t
Con cret e
Suspended Ceiling TYPE B
Floo r Be
am
Figure 7a & 7b: Watson Floor System.
STRUCTURE magazine
26
September 2008
8 10 12 Depth , , Fin. Floor to Ceiling
12 Mortar Topping & Finish

Std Floor Units
Negative Bar Girder
Mortar Topping & Finish Var ies 9
Column
86
A A
Section A-A Plaster Locking Joint & Cross-Bridge Poured on Job

9
Firepf g Variable End Unit Column Tie
Var i
Standard Center Unit
es
de Gir
Figure 8: Miller System.

Hy-Rib Metal Lath Electric Welded Steel
Standard Joist Spacing 18 or 24 Other Spacing Optional Depending on Loads 1 or 2 Concrete or Haydite Slab
Cop
h yrig
Figure 9: Lith-I-Bar System.
with at steel bars that projected from the ends of the panels to act in conjunction with hangers suspended from the top ange of the supporting steel framing. The Gypsteel system was manufactured in New Jersey and used extensively in New York City. Waites Concrete I Beams (Figure 6): This system was used in a number of buildings constructed by The Standard Concrete Steel Company of
Rmag T S
Cazi U
New York City. The oor 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 ange of steel beams that were spaced 5 to 7 feet apart. A eld-cast concrete topping was then placed on top of the I-beams, while the spaces between the lower anges were inlled, as well, to provide a at 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 rst conguration 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 ll in the 9- to 10-inch-long gaps between the segments. The center segment was produced in a standard xed 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 eld-cast topping for continuity across the supporting steel beams. The voided, 12-inchwide, 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 Ishaped cross-section mold and compacted with cast-iron rollers. The
Ue Tn
E R
STRUCTURE magazine
27 September 2008
units were typically spaced at 18 or 24 inches on center with a 2Cross Channels Filled inch eld topping cast on either removable or stay-in-place metal 60 with Cement Mortar lath formwork for composite action with the members. Porete Floor System (Figure 10): This system was manufactured 4 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 Formwork A A at the job site. The gap between the abutting aligned units and 14 Precast the continuous pocket along the sides of each unit, in which Units eld-positioned bottom reinforcement was rst placed, were then lled in with a mortar/grout. All of the units were closed at each end to prevent the mortar/grout from owing into the Section A-A Reinforcing Bars hollow voids of the precast member. These units were typically supported by steel beams over which top reinforcement was po14 sitioned in the continuous member pockets to provide continu- Figure 10: Porete Floor System. ity of the oor slab system. This precast system was capable of ht spans from 10 to 25 feet. yrig 1 Space Filled with Grout Cop Extending Wires Overlap Tee Stone System (Figure 11): This precast member could be used as a oor beam, roof beam or wall panel and was origi16 Mesh Wires Project nally manufactured in New York. The T-section was 8 inches at Least 2 on Each Side Flange Reinforcing Rods deep, with a 16-inch-wide ange and a 1-inch-wide stem, and 1 was manufactured in standard lengths of 8 and 16 feet. For oor Mesh Wire Extending construction, the T could be installed in either a ange up or From Flange Down 8 ange down position. The units were placed in the eld with into Stem Forming a 1-inch gap between the edges of the anges, which was lled a Loop Stem with grout. The ange mesh reinforcement extended into these continuous gaps to produce a monolithic slab. Stem Reinforcing Rod Varies 1 Pyrobar Precast Roof System (Figure 12): This cast gypsum system was manufactured for use as a roof slab and was available in both 3-inch-deep solid and 4-inch-deep hollow-core sections Figure 11: Tee Stone System. for short-span applications, as well as 5- and 6-inch hollow-core sections for long-span applications. The short-span sections were Short Span Roof Tile - 30 Type made in 12-inch widths and 30-inch lengths. The long-span secDepth 3 Solid 4 Hollow tions were made in 18-inch widths and lengths from 4 feet to Length 30 30 Wt per Sq Ft 15 lbs 15 lbs 62 feet. The short-span members were typically supported by Grout 30 12 steel bulb tees, while the long-span members were supported by underslung steel wide ange and channel beams. 3 Solid Not to Exceed 2 Reinforcing All of the above precast systems were designed based on the Mat 30: for 30d 30d For I Beams basic reinforced concrete beam analysis theories of their era. Tees Notch can be furnished Load tables were also commonly developed and published by on 3 solid only 4 Hollow most of the manufacturers. The problem with all of the above Details of Short Span Pyrobar Roof-Tile. systems when one encounters them in a building today is that, All Tile Reinforced with Electrically Welded Galvanized Steel Mat. in the absence of existing drawings, it is difcult to determine 18 Standard Long Span Hollow Tile the internal reinforcement and, consequently, the load-carrying 5 capacity of the system. However, it is hoped that this article, by Reinforcing identifying the many different types of products that were in Length Mat use at one time or other, will assist readers in their research of 6 an antiquated or archaic system when it is encountered in an existing structure. Core
Rmag T S
Cazi U
Lap Joint Butt End
Ue Tn
1 3
12
E R
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 ofce 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. STRUCTURE magazine
Not to Exceed 2
Note: Notch can be Furnished on Both 5 & 6 Long Span Hollow Roof Tile
Length Depth
Wt per Sq Ft
40 to 66 40 to 66 5 6
20 lbs 25 lbs
Figure 12: Pyrobar Roof System.
References
Kidder, Frank E. and Harry Parker, Architects and Builders Handbook, 18th Edition, John Wiley and Sons, 1956 Specication and Load Table Archives, Nitterhouse Concrete Products, Inc., Chambersburg, PA
28
September 2008

ENGINEERS NOTEBOOK
Structural Steel Composite Stub-Girder Construction Part 6

By D. Matthew Stuart, P.E., S.E., F. ASCE, SECB or this series of articles, antiquated has been dened as meaning outmoded or discarded for reasons CORE 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 1 w BEAMS innovative or more economical methods of construction. Most of the antiquated systems discussed FLOOR PLAN ht so far have been out of popular use for a yrig 10 - 0 p Co considerable number of years, with some dating back to the rst part of the last cenDUCT tury. However, the subject of this article deals with a system that was still in use less CEILING than 20 years ago. SECTION 1 The purpose of this series is to compile CONVENTIONAL FLOOR FRAMING SYSTEM and disseminate a resource of information (a) to structural engineers for projects that involve the repair, restoration, or adaptive reuse of older Figure 1: Floor buildings for which no drawings exist. It is hoped that this Framing Systems. will enable structural engineers to share their knowledge of Courtesy AISC. existing structural systems that may no longer be in use, but are capable of being adapted or reanalyzed for safe reuse in system went on the marketplace of today and the future. to be used in a large number of high-rise buildings in North America The Stub-Girder up through the 1980s. However, the Composite System system was eventually abandoned because A stub-girder is a composite system of increased labor cost associated with constructed with a continuous structural both fabrication and the need for shoring steel beam and a reinforced concrete slab until the eld-cast concrete slab attained separated by a series of short, typically sufcient strength. wide, ange sections called stubs. Stubs Advantages of the stub-girder system are welded to the top of a continuous that led to its use during the time period beam and attached to the concrete slab in which it was popular included: by shear connectors. Spaces between 1) Reduction in steel tonnage by as ends of stubs are used for installation much as 25% over conventional of mechanical ducts and other utility composite oor framing due to: systems and for placement of transverse a) Improved structural efciency oor beams that span between stubas a result of the greater depth girders. Ideally, the depth of stubs and of the stub-girder compared to oor beams are identical to allow for a conventional system; and transverse framing to support a concrete b) Improved structural efciency slab deck, which spans parallel to the due to the ability of transverse stub-girder, and to facilitate composite oor framing members to act action between the oor beam and slab as continuous beams through (Figure 1). openings between stubs. Stub-girder construction was rst used 2) Reduction of overall depth of the in 1971 at the 34-story One Allen Censtructural oor framing system by ter ofce building in Houston, Texas. as much as 6 to 10 inches over a The system was developed by Joseph P. conventionally framed composite Colaco, Ellisor Engineers, Inc., to faoor system, which allowed for a cilitate integration of mechanical ducts reduced oor-to-oor height and into steel oor framing of repetitive, overall height of the building and multistory high-rise construction. This associated cladding.
4 - 0
CORE
FLOOR PLAN 10 - 0 DUCT
Rmag T S
Cazi U
Ue Tn
W14 STUB GIRDER SECTION 2 STUB GIRDER SYSTEM (b)
E R
CEILING
3 - 6
Testing
Prior 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 oor beams, and a 5-foot-wide, 3inch-deep, lightweight concrete slab over a 2-inch metal deck ange, 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 delamination of the web from the ange. Application 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 localized crushing of the concrete slab. Web stiffeners added to the failed stub allowed the system to achieve a nal failure load that was 2.2 times greater than the calculated design load. The methods of design used to determine capacity of the section included a non-prismatic beam analysis, a Vierendeel girder/truss analysis, and a nite element analysis. For the Vierendeel analysis, stubs and transverse oor 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
STRUCTURE magazine
53 November 2008
of analysis provided a close representation of actual behavior of the stub-girder; however, the Vierendeel and nite element methods more closely identied 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 nite 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 rst exterior stub as observed in the test specimen. Additional tests of stub-girders were performed in the late 1970s in Canada. The primary purpose of these tests was to determine effects of changes in spacing and depth of stubs, and to establish failure modes of a stub-girder. Results conrmed 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 deection of the stub-girder, but did not inuence 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 signicant impact on elastic stiffness of the girder.
Rmag T S
58-9 2-1 9-7 5-0
14 SHEAR CONN. (2 ROWS)
A
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 contraexure within t open panels.h y ig pofrshear connectors resulted 6) Failure Co 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 identied: 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.
Cazi U
9-7 3-0 3-3 3-3 9-7 3-0 3-3
W16x26 3/16 W16x26 3/16 W14x48 (36,000 PSI YIELD) STRAIN GAGE B DEFLECTION GAGE
Ue Tn
2-6
LOAD PT.
W16x26 W16x26
Further research in Canada revealed additional insights into behavior, design, and economical construction of stub-girders. This research indicated that only partial end plate stiffeners, rather than traditional tted stiffeners, were required to reinforce stub webs. Furthermore, web stiffeners were not always required at interior stubs. In addition, cona tinuous perimeter weld between the base of the stub and the top of the continuous beam was not required. Tests also conrmed that rolled wide ange 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) Deection 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 stubgirder 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.
E R
2-6
2-6
3-3
LOAD PT.
10-0 5-0
14 SHEAR CONN. (2 ROWS)
6 SH. CONN. STRAIN 6 LOAD 6 SH. CONN. (2 ROWS) GAGE (2 ROWS) PT. B W16x26 W16x26
W16 x 26 STUB (BOTH SIDES) WELDED TO W10 W10 x 12
W16x26
3/16
3/16
3/16
4x4 - 4/4 WWF DRAPED
3-7/8 BOLTS 2L - 4 x 5 x 5/8 2L - 4 x 5 x 5/8 w/2 - 7/8 & BOLTS/LEG L P- 11 x 1 x 0-11 WELDED TO W SHEAR CONNECTORS
W16x26 STUB (BOTH SIDES) WELDED TO W14 W14 x 103
L P - 15 x 1 x 1-3 WELDED TO W14
ELEVATION
SLAB WIDTH 5-0 53 LT. WT. CONCRETE 54 LT. WT. CONCRETE OVER CORRUGATION C2 SHEAR CONN. EVERY CORRUGATION
W16 x 26
16 GA. C12 x 1 CONTINUOUS 5/16 FILLET WELD
22GA. C2 COFAR W16 x 26
W14 x 48
W14 x 48
SECTION A-A
Figure 2: Stub Girder Load Test. Courtesy of AISC. STRUCTURE magazine
SECTION B-B
54
November 2008
Design Guidelines
Further recommendations and guidelines emerged throughout the 1980s for the stubgirder 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 ange 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 nal design guidelines that were established for stubgirder 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 oor 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 oors 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 deection of the member.
Rmag T S
Cazi U
16) Modeling of stubs as verticals of 9) Overall strength of a stub-girder is not the Vierendeel truss/girder involves controlled by compressive strength dividing the stubs up into vertical of the concrete slab, therefore use of elements equal to one-foot lengths high-strength concrete mixes provides of the section, spaced at one foot on no advantage. center from one end of the stub to the 10) It is typical for ribs of a metal oor other. Vertical stub elements should deck to run parallel to the span of a be modeled as xed at the top and stub-girder. This orientation of ribs bottom, at the top chord (concrete therefore increases the area of the top slab) and bottom chord (continuous chord slab and also makes it possible beam) of the truss/girder, respectively. to arrange a continuous rib or trough 17) Transverse oor beams should be directly above the stubs, which in modeled as single vertical web turn improves composite interaction of members/elements of the truss/girder. the slab with the stub-girder. The top and bottom of the member 11) Welds between the bottom of stubs should be modeled as pinned at top and the top of the continuous bottom t and bottom chords. beam should be concentrated at the igh yr endspof the stubs Co where forces between (CONTINUOUS CONCRETE SLAB) these two elements are TOP CHORD greatest. 12) Internal longitudinal slab reinforcement to add strength, ductility and stiffness to the VERTICAL WEB stub-girder should (STUB & TRANSVERSE be provided in two BEAMS) layers, one just below and one just above the heads of shear studs. 13) Internal transverse slab reinforcement BOTTOM CHORD should be provided (CONTINUOUS BOTTOM BEAM) to add shear strength TYPICAL WIDE FLANGE VIERENDEEL TRUSS and ductility. (EQUIVALENT STUB GIRDER COMPONENTS) Placing transverse reinforcement in a Figure 3. herring bone pattern In conclusion, the stub-girder method of i.e., diagonal to the construction was and still is an innovative direction of the stub-girder span will solution to multi-story, framed steel oor also increase the effective width of the construction. However, as deeper wide concrete ange/top chord. ange sections became more available in 14) Flexural stiffness of the top chord slab of a stub-girder should be based on the the marketplace and design engineers became more accustomed to analyzing web conventional effective width allowed holes in wide ange beams, the use of stubby standard composite beam design girder construction waned. In addition, criteria, except that the transformed because of extra labor costs associated with section should include contribution fabrication of stub-girders and the necessity of both the metal deck and internal to construct stub-girders as shored composite longitudinal reinforcement. construction, the system priced itself out of 15) It is not proper to include the top the industry. ange of stubs in the calculation of moment of inertia of the top chord slab element.
Ue Tn
E R
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 ofce 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. The online version of this article contains references. Please visit www.STRUCTUREmag.org.
STRUCTURE magazine
55 November 2008
References
A Stub-Girder System for High-Rise Buildings Joseph P. Colaco AISC National Engineering Conference New York, NY; May 1972
An Experimental Investigation of Stub-Girders Y.W. Lam, T. Rezansoff and M.U. Hosain University of Saskatoon, Canada Behavior of Building Systems and Building Components Conference Vanderbilt University Nashville, TN; March 1979 Some Aspects of Stub-Girder Design t Reidar Bjorhovde and T.J. Zimmerman igh pyr University of Alberta and the CISC Co Canadian Structural Engineering Conference Montreal, Quebec; February 1980 Reprinted in the AISC Engineering Journal; Third Quarter 1980 Structural Engineering Handbook Chapter 18; Stud Girder Floor Systems Reidar Bjorhovde University of Pittsburgh CRC Press LLC 1999
Rmag T S
Cazi U
Ue Tn
E R
STRUCTURE magazine
November 2008

Part 7
ENGINEERS NOTEBOOK
By D. Matthew Stuart, P.E., S.E., F. ASCE, SECB For this series of articles, antiquated has been dened 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 posttensioning 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 fullled, which is to compile and disseminate a resource of information to enable structural engineers to share their knowledge of existing t structural systems that may no longer be in use but are capablegh being adapted or i of pyrfuture. reanalyzed for safe reuse in the marketplace of today and the Co
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 constructed 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 rst 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 deections, as well as the stresses associated with the stripping process, cracking of the slabs was very common. In order to correct these inherent deciencies, the industry turned to the available European BBRV (button-headed) post-tensioning system (Figure 1). This system of post-tensioning allowed the reduction of slab thicknesses to as little as 8 inches, along with a reduction in
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
Rmag T S
The button-headed system involved the use of parallel, -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 tendons. To anchor the tendons at each end, the wires were threaded through holes in a steel bearing plate and then an externally 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 eld 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 construction site.
Cazi U
Ue Tn
Courtesy of Larry Krauser, General Technologies Inc.
Figure 1. STRUCTURE magazine
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 washer. 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 awed because if the actual elongation did not match the length of the shim plate, eld 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 ll 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 restraint 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 introduce an intermediate stressing point coupler. The couplers usually consisted
E R
25 December 2008
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 specializing 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 companies and the emerging ying form industry that formed-in-place, post-tensioned construction 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 multistory, 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 construction from the precast/prestressed concrete industry in the early 1960s. By the end of the same decade, the post-tensioning strand system 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 exclusively used for greased, unbonded strand. The use of strand was identied 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 exibility of the steel anchor shims. The rst 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 inlled stressing pockets. The wedge anchor assembly was installed with a small, two-piece, round rubber grommet that was positioned between the anchorage and the inside face of the edge form. The grommet allowed the anchorage to be recessed 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 lled in the space between the coil and the strand, preventing cement paste from entering the assembly during concrete placement. After tensioning was completed
Cop
Figure 2.
h yrig
Rmag T S
Methods of Analysis
and any excess strand was removed beyond the outboard end of the anchorage, the remaining hole was lled with grout for a ush nish with the edge of the slab. The seven-wire strand system was also more conducive to intermediate stressing locations, eliminating the nal 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 intermediate 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 ...when ACI 318-77 was published, banded tendon distribution in two-way slab construction was recognized by the Code.
Cazi U
Ue Tn
the structural supports, the analysis of posttensioned structures was originally difcult. The analysis of cast-in-place, post-tensioned structures became much easier, however, after a simplied 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 tendons 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 simplied the process of post-tensioned design, making it as easy as designing a conventionally reinforced structure, thereby helping to increase the popularity of post-tensioned construction from the late 1960s into the 1970s. Two-Way Construction Methods Originally post-tensioned, two-way slab framing systems were constructed with column and middle strips, similar to those used in conventionally reinforced two-way slabs. However, because the continuous two-way tendons had to be placed in a draped parabolic prole 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 specic sequence. This was difcult, 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 conrmed 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.
E R
inside of the coil. As a result, localized failure of the concrete, particularly with lightweight mixes, was common. This inherent deciency of the coil anchor resulted in the development of a ductile iron anchorage plate in the mid1960s. The ductile iron assembly consisted of a at 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.
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
STRUCTURE magazine
26
December 2008
Building Codes and Industry Organization

Up through the ACI 318-71 Code, castin-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 Institute (PTI), which was established in 1976.
Conclusion
The post-tensioning industry is likely to continue to thrive worldwide for both building and bridge construction. Further advancements 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 developed for mono-strand construction that will allow for conventional unbonded installation 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 ofce 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.
Common Problems
The two biggest problems faced by the posttensioning 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 specications in the mid 1970s. Concrete volume change occurs in both nonprestressed 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 signicant 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 signicant 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, particularly 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 reveal itself until about 10 or 15 years after the earliest post-tensioned concrete structures were constructed. The problem was corrected after the PTI developed specications that became enforceable through certication programs created by the organization. The improvements established through the PTI specications included sheathings, coatings and encapsulation systems for use in aggressively corrosive environments.
Cop
h yrig
Rmag T S
Cazi U
Durability 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
Ue Tn
References
E R
STRUCTURE magazine
27 December 2008

Part 8
By 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.
Figure 1a: Courtesy of Peabody Essex Museum.
Wrought and Cast Iron

Ferrous metals used in construction can be categorized into three principal ironcarbon 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) brous 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 phosphorus. Cast iron, while molten, is easily poured into sand molds, making it possible to create unlimited forms, which also results in mold lines, aws and air holes. Cast iron elements are commonly bolted or screwed together, while wrought iron is either riveted or welded.
yr Cop
igh
ENGINEERS NOTEBOOK
A R
Figure 1b.
T F
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 railing and balconies. Around 1850, the structural use of wrought iron became more prevalent as rail beams, bulb-tees, channels and I-beams became commercially 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 perforated 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 manufactured the rst rolled wrought iron beam in the U.S., a 7-inch-deep, bulbtee rail beam. The following year, the Trenton Iron Works also manufactured the rst I-beam in the U.S., the Cooper beam, although I-beams had previously 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
Figure 2.
shipbuilding industry, and prior to the development of the Cooper beam, channel beams had been manufactured. Figure 3 (page 10) shows the evolution of all of these rolled wrought iron sections. Rolled wrought iron beams continued to be used for several decades after 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 provided in the Pencoyd Iron Works manual indicate that the ultimate tensile strength of wrought iron was as high as continued on page 10
STRUCTURE magazine
March 2009
NCSEA Webinars
Wind Load Examples
8th 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 Simplied 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. Dan Werdowatz, S.E. is Vice President of Josephson-Werdowatz & Associates and has over 23 years of experience. He holds both bachelors and masters 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 t review of existing structures. righ 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 bachelors 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.
*The three volume set of 2006 Structural Seismic Design Manuals are availableat a discounted price to registrants.
March 10th
Cop
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.
A R
T F
Strength Design of Masonry

Part 3 of the Masonry Series
March 5th
This webinar will review the strength design requirements of Chapter 3 of the MSJC for reinforced masonry and it will highlight modications 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.
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.
Masonry Tallwalls
Part 4 of the Masonry Series
March 19th
Strength design procedures for masonry have been under development for many years. Tall slender walls benet greatly using strength techniques resulting in taller, thinner construction with less reinforcement than those designed using allowable stress techniques. This presentation will provide the benets 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. 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.
All NCSEA Webinars: 10:00 Pacic, 11:00 Mountain, 12:00 Central, and 1:00 Eastern
UR AL
GI EN
ST RU CT
NCSEA
NG UI IN NT CO
Diamond Reviewed
ED UC AT IO
E NE RS
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.
Register at www.ncsea.com
50,000 psi. The allowable extreme 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, uctuating or impact) and extent of lateral support.
Building Construction Overview
The use of prefabricated cast iron faades, wrought iron beams and cast iron beam and column components served Deck 7-inch Rail Channell Compound I Beam as a forerunner to the multi-story, steelBeam Beam Beam Beam framed buildings of today. In addition, Figure 3. specialty wrought iron structures constructed in New York City in the mid Cast Iron painted with pitch, which was later replaced 1850s, such as rewatch towers and gunshot Cast iron is very hard and resists compression by galvanizing. manufacturing towers that employed drilled forces very well but, because of the carbon Corrugated iron was also manufactured in and socketed rock foundation anchors and content, it is also very brittle. Because of arched sheets for oor construction. Typically, this, cast iron was used primarily for the the ends of the arched sheets were supported inll masonry walls, served as a precursor to construction of columns and compression by the bottom anges of wrought iron I-beams, early steel skyscraper construction. t It is also interesting to note that the manufacigh elements of composite wrought iron girders. with concreter then cast on top of the sheets. turers of cast iron faades and other exposed py o C Shortly before the 1800s, cast iron was used This type of construction replaced brick arch for columns in multi-story, wood-framed oor construction, which had been used exten- components often identied the source of the material by either afxing foundry labels factory buildings in England. Cast iron was sively with wrought iron I-beams previously. to the building or by using trademarks in the used because of its strength and perceived ornamentation of the decorative portions of re resistance. The combination of cast iron the faade. This information can sometimes columns and wood framing continued to be used to assist in the structural evaluation be widely used for the next half century; of an existing building. however, by the mid-1800s, wrought iron beams began to replace wood. Cast iron used for structural purposes during the 1800s typically had a compressive strength of 80,000 psi. Although there was no clearly dened 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 aws, 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 faades of buildings (Figure 4). The cast iron faade was used for both decorative and structural purposes, as the adjacent oor framing was supported by the cast iron. Cast iron was also used in conjunction with masonry faades 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.
A R
T F
Deterioration
Iron oxidizes rapidly when exposed to moisture and air. The product of the oxidation process 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 lm of rust starts to develop, the natural porosity of the corrosion byproduct tends to act as a reservoir for moisture, resulting in even further acceleration of the deterioration process. Cast iron will develop a somewhat protective 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 metal galvanic corrosion can also occur between iron and copper, chromium, lead, stainless 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, porous graphite residue is impregnated within the surface corrosion byproduct. The cast iron element retains its original appearance
Rolled Sheet and Corrugated Iron

The rst sheet iron in the U.S. was rolled in Trenton, New Jersey in the late 1700s. Sheet iron was used as ooring and roong material up until the end of the 1800s. Corrugation of sheet iron, patented in England in 1829, was rst used in the U.S in the 1830s. Corrugated sheet iron was typically
Figure 4.
STRUCTURE magazine
10
March 2009
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 identied in the eld by scraping the surface of the cast iron with a knife to see if deterioration of the iron is revealed beneath the surface.
Repair and Restoration

There are a number of methods available to remove paint and corrosion from cast and Figure 5. wrought iron, including manual scraping, chipt ping or wire brushing; low-pressure grit or sandigh blasting; ame cutting; and chemical removal. pyr and bolt or screw holes with joints, cracks Co Once any existing paint and corrosion have sealant to prevent moisture intrusion and been removed, the most common method of freeze/thaw damage; and, following the paint protecting cast iron from further deterioration manufacturers specications. is to repaint the surface. Prior to repainting, In all cases, it is recommended that a test it is rst necessary to prepare or repair the area be used to confirm that the selected surface. Proper preparation includes elimi- cleaning, preparation and painting techniques nation of crevices or pockets that can collect are effective prior to attempting to remedimoisture to prevent accelerated deterioration; ate the entire restoration area. It is also recremoval or smoothing of sharp corners to ommended that sheltered areas such as eaves, prevent accelerated paint failure; hermetical where evaporation of moisture can be inhibsealing of hollow section to prevent moisture ited, be coated with additional layers of the intrusion and freeze/thaw damage; lling of selected paint or coating.
A R
T F
Additional protection and repair procedures can also include plating with metals, or cladding with plastic or epoxy. Sections or entire portions of a signicantly deteriorated area may be replaced with glass ber reinforced concrete (GFRC), ber reinforced polyester (FRP) or aluminum. If additional structural retrotting is required as a part of an adaptive reuse project, the following remedial work should be avoided if at all possible: welding, burning of holes, the use of impact drills, high strength bolts, and lling of voids or posts with concrete.
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 ofce 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. The online version of this article contains detailed references. Please visit www.STRUCTUREmag.org.
Change your frame of reference.
In the past moment frames were the expensive option when designers wanted small wall sections and open 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) 999-5099 www.strongtie.com
2009 Simpson Strong-Tie Company Inc. SF09
STRUCTURE magazine
11
March 2009

Part 9a Open Web Steel Joists
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 method 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 primary purpose of this series of articles will be fulilled: 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. By 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. Introduction of the H-Series joists, which ht yrig p provided an allowable tensile Co 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 speciication for the J- and H-Series joists by SJI and AISC. 1966 Introduction of the LJ-Series joists, which replaced the LASeries 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 chronological listing of the standard speciications 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
ENGINEERS NOTEBOOK
Rmag T S
Open Web Steel Joists
History 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 Specications 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.
Cazi U
Ue Tn
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); TChord 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 products other than the historical standard 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 members); 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 Ushaped chord members).
E R
STRUCTURE magazine
12
June 2009
Resource Material 80 Years of Open Web Steel Joist Construction; A Compilation of Specications and Load Tables Since 1928; Steel Joist Institute; Myrtle Beach, South Carolina (2009). Catalog of Standard Specications, 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. ht Public domain documents that have limited availability.yrig p o In summary, this is content that is dificult to C ind anywhere else. To access it, visit www.SlideRuleEra.net.
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 ofce 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.
Rmag T S
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
Cazi U
Ue Tn
E R
Watch for Part 9b in an upcoming issue.
STRUCTURE magazine
13
June 2009
Antiquated Structural Systems Series - Table 1

Part 9a Open Web Steel Joists
By D. Matthew Stuart, P.E., S.E., F. ASCE, SECB Table 1: Unique Open Web Joists (Load Tables may be available from SJI)
System Figure 1 Ashland N/A Description HS-Series Joists LS-Series Joists 1952 Structural T Longspan & Standard Joists 1952 to 1962 TChord Longspan Joists Purlin or Steel Joist Massillon Steel Joist Canton Steel Joist Buffalo Steel Joist Special Joists Residence Joist Yield Strength 50 ksi 50 ksi Depth (inches) 8 to 24 Unknown Span (feet) 8 to 48 64 maximum 12-6 to 108 Chords Double angles Unknown Webs Round bars Unknown Notes
Cadmus
See Note 6
10 to 54
Haven Busch
6 7 8 9 10 11 12 N/A
C9 See Note o
Unknown Unknown Unknown
h yrig
18 to 88 8 to 16
25 to 175 10 to 26 4 to 31
Macomber
Rmag T S
Unknown 12 to 20 6 to 10 Unknown Standard Longspan Joist Intermediate Longspan See Note 14 18 to 40 N/A 13 14 See Note 14 Unknown 18 to 22 8 to 24 1955 New Yorker V or Double V Bar Joist V-Girders V-Purlin Allspan V-Lok Purlin V-Lok Girder V-Beam Series 1, 2, 3 & 4 Joists Unknown 8 to 22 N/A 15 16 N/A 17 18 Unknown Unknown Unknown Unknown Unknown Unknown 18 to 48 8 to 60 8 to 76 8 to 36 12 to 40 8 to 28 4 See Note 5 12 to 72 3 Open Web Joists See Note 3 12 to 47 2 Composite Floor Joists Composite Truss Girders 36 & 50 ksi 36 & 50 ksi 8 to 40 16 to 40 N/A
Unknown
Cazi U
8 to 16 8 to 16 Unknown 8 to 40 8 to 16 Unknown 6 to 20 24 to 72 20 to 44 7 to 48 4 to 44 13 to 96 8 to 120 8 to 152 8 to 72 15 to 50 8 to 56 12 to 80 16 to 59
Ue Tn
Split T Angles Split T Angles See Note 10 Round bars Double angles See Note 11 Unknown See Note 12 Double angles See Note 10 V shaped plates V shaped plates V shaped plates V shaped plates V & Double V shaped plates V & Double V shaped plates See Note 18 See note 19 V shaped plates
E R
6, 7 8, 9 10 11 12 13, 14 10, 14 15 15 16, 17 16, 18 19
Round bars
Round bars Round bars Round bars Unknown
Round bars
Angles & bars Round bars Round bars Round bars Round bars See Note 15 See Note 15 Round bars or round pipes Round bars or Angles Round bars Square bars & round pipes
Northwest
4, 5
Ridgeway
V shaped plates
Square bars & round pipes
20 to 48 20 to 50
Double angles Double angles
Round bars Angles
1 2
Vescom
STRUCTURE magazine June 2009 Antiquated Systems Series - Part #9a - Table 1
Notes: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.
Top chord included deformed, extended vertical leg of one angle for composite action with surrounding concrete slab. Top chord included deformed, extended vertical plate in addition to double angles for composite action with surrounding concrete slab. Web allowable stress: 36 ksi (bars) & 50 ksi (pipes); Chord allowable stress: 54 ksi. Joist designs over 80 feet spans were available upon request. Web allowable stress: 33 & 44 ksi (bars), 50 ksi (pipes); Chord allowable stress: 55 ksi. 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. Chord tees cut from standard wide lange or junior beams. Available as parallel chord, single or double sloped top chord or hipped end conigurations. 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). Double angle top chord; Round bars bottom chord. Inverted double angle top chord; Round bars bottom chord. Single steel angle and wood nailer top chord; Round bars bottom. ht Available as parallel chord or single or double sloped topyrig p chord. ksi. Co Allowable combined direct and bending stress in top chords = 20 Sizes #2 - #9: Round bars; Sizes #10 up through #22: Angles. Included proprietary stud and slot end bearing connection See Figure 17. Round bars, round pipes or angles. V & double V shaped plates or double angles. V shaped top chord & U shaped bottom chord plates.
Rmag T S
Cazi U
Ue Tn
E R
Table 2: Unique Open Web Joists (Load Tables may be available from SJI) System Figure 24 25 26 Bethlehem 26 27 28 29 29 Gabriel 30 19 & 20 Truscon 21 21 Description KalmanTruss Joists MacMar Joists BLJ Series BLH Series Standard Open Web Joist Longspan Open Web Joist BJ Series BH Series Long Span Joist O-T (Open Truss) Joists Series AS Joists Yield Strength See Note 8 See Note 10 See Note 11 See Note 12 See Note 13 See Co 11 Note Depth (inches) 8 to 16 8 to 16 52 to 60 52 to 60 8 to 16 24 to 30 24 to 30 18 to 32 8 to 20 Span (feet) 4 to 32 4 to 32 89 to 120 89 to 120 4 to 32 25 to 64 Chords T shape Angles Structural Tee Structural Tee Angles Angles Webs Rectangular Round bars Angles Angles Notes 7, 8, 9 10 11
See Note right 18 to 32 14
py
24 to 60 24 to 60
Notes: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
Web allowable stress: 19,000 psi - 100(l/r); Chord allowable stress: 16,000 psi. Cold formed chord allowable tension: 25 ksi; Hot rolled web members allowable compression: 17,000 psi - 100(l/r). Cold formed chord allowable tension: 28.5 ksi; Hot rolled web members allowable compression: 19,000 psi - 100(l/r). Available as parallel chord, single or double sloped top chord conigurations. Chord angles were some times arranged toe to toe for channel coniguration. Allowable combined top chord compressive stress: 15 ksi; Allowable bottom chord tensile stress: 18 ksi. Manufactured by punching web opening in blanks such that chords and webs do not have to be welded together. Allowable tensile stress: 16 and 18 ksi. Also marked as Kalman Joist. Allowable tensile stress: 18 ksi. Maximum tensile working stress: 22 ksi. Maximum tensile working stress: 30 ksi. Design tensile stress: 18 ksi. 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.
Rmag T S
Series BB Joists See Note 3 22 & 23 Clerespan Joists See Note 6
Cazi U
See Note 12 24 to 64 7 to 40 See Note 1 See Note 2 8 to 24 7 to 48 8 to 24 7 to 48 18 to 32 26 to 64
Ue Tn
See Note 15 Angles See Note 15 Tee & M shaped plates U shaped U shaped Tee & angles
Round bars Angles
Round bars
E R
12 13 4, 14 11, 15 4 1 2 3 4, 5, 6 12, 15
Round bars Round bars Round bars
Round bars
Round bars
Angles & bars
Figure 2
Cop
Figure 1
h yrig
Figure 3
Rmag T S
Figure 5 Figure 6
Cazi U
Ue Tn
E R
Figure 4
STRUCTURE magazine June 2009 Antiquated Systems Series - Part #9a - Figures 13
Cop
h yrig
Figure 7
Rmag T S
Cazi U
Ue Tn
E R
Figure 8
STRUCTURE magazine June 2009 Antiquated Systems Series - Part #9a - Figures
Cop
h yrig
Rmag T S
Figure 9 Figure 10
Cazi U
Ue Tn
E R
Cop
h yrig
Figure 11
Rmag T S
Cazi U
Ue Tn
E R
Figure 12 STRUCTURE magazine June 2009 Antiquated Systems Series - Part #9a - Figures
Cop
h yrig
Figure 13
Rmag T S
Figure 14
Cazi U
Ue Tn
E R
Cop
h yrig
Figure 16
Figure 17
Rmag T S
Cazi U
Ue Tn
Figure 19
E R
Figure 18

Figure 20
Cop
h yrig
Figure 21
Figure 22
Rmag T S
Cazi U
Figure 24 Figure 25
Ue Tn
E R
Figure 23

Figure 27
Cop
h yrig
Figure 28
Rmag T S
Figure 29 Figure 30
Cazi U
Ue Tn
E R

Part 9b Open Web Steel Joists
By D. Matthew Stuart, P.E., S.E., F. ASCE, SECB For this series of articles, antiquated has been dened 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 fullled, which is to compile and disseminate a resource of information to enable ht structural engineers to share their knowledgeig existing pyr of structural systems that may no longer be inCo but are capable use of being adapted or reanalyzed for safe reuse in the marketplace of today and the future.
TYP. CONCENTRATED LOAD L2x2x REQD IF LOAD DOES NOT OCCUR AT JOIST PANEL POINT OR > 6 AND EXCEEDS 150#
Evaluation and Modication of Existing Joists
ENGINEERS NOTEBOOK
Rmag T S
The author would rst 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 modication of existing joists and Joist Girders in Technical Digest No. #12. The rst 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 construction 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 eld. However, if tags can be found, more often than not the tag only identies 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 documentation. 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 denitive distinction is required, and it is possible to secure a material sample in order
Cazi U
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 possible to establish the approximate capacity of the member by eld measuring the chord and web member sizes, as well as the overall conguration 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 lling out the Joist Investigation Form located on the SJI website. SJI has indicated 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 location 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
Figure 1: Typical concentrated load on joist detail.
Ue Tn
JOIST
HANGER ROD
E R
joist is capable of safely supporting the imposed loads. However, if the maximum moment is greater than one foot from the midspan point, the capacity of the joist may not be sufcient even if the applied moment is less than the specied 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 compression) for which the original member was not designed or manufactured. A similar, although typically more advantageous, condition 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 sufcient 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 sufciently 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
STRUCTURE magazine
18
November 2009
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 loadcarrying capacity of the member. The key to the successful use of load redistribution 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 distribution member is less than the inverse of the calculated relative stiffness, then the distribution member may be considered as rigid enough to calculate the static load reactions to the affected joists. For load redistribution solutions, it is the authors 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 conguration 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 ve joists be engaged by any one redistribution member. In addition, the use of pipes for the continuous redistribution truss chord members can be advantageous, as this type of section ts 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 difcult 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 support 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 eld splices. However, it is possible to install new full-length manufactured
joists by means of loose end bearing assemblies. In this scenario, the joists are rst 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 provided for ease of installation and then shimmed once the new joist is in its proper position. When new beams or other similar members 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 envelope of the joist. As with load redistribution solutions, both of the above new framing ap proaches may be difcult 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
Cop
h yrig
Rmag T S
Cazi U
Ue Tn
E R
Figure 2.
STRUCTURE magazine
19
November 2009
framing can involve beams located either beneath or above the impacted existing framing and supported by new columns and foundations, or beams that frame between existing columns. This type of solution can also involve new beam frames supported from posts located directly above existing beams or columns. The above solutions are typically more adaptable to the presence of existing MEP systems, 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 approaches typically involve signicantly more labor costs than material costs because of the expense associated with eld welding. The author prefers to avoid the use of eld reinforcement for the following reasons. A manufactured open web steel joist is basically a pre-engineered product; however, when an engineer involved with the modication of an existing joist species new eld installed reinforcement, that same engineer assumes the responsibility for the overall adequacy of the joist. This liability extends to not only the reinforcing modications but also, inherently, to any pre-existing, unknown conditions or deciencies in the joist. In addition, eld welding associated with the installation of reinforcement also poses concerns for the design engineer. Problems associated with eld welding are discussed in Technical Digest No. #12 and include temporary localized loss of the material 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 (Figure 1, page 18), and reinforcement designed by the original manufacturers engineer. The rst exception is the authors rule of thumb and is not formally endorsed by SJI, because it is not applicable in all cases; for example, it may be ne 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 specic existing joist under investigation. However, it is hoped that the reference 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 existing 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 responsibility and liability for modifying a preengineered 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 modication process, the author has included a t copy of a owchart (Figure 2, page 19) that igh pyr as result of numerous projwas developed Co 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 ofce 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.
Rmag T S
80 Years of Open Web Steel Joist Construction A Compilation of Specications and Load Tables Since 1928 Steel Joist Institute Myrtle Beach, South Carolina 2009
Catalog of Standard Specications, Load Tables and Weight Tables for Steel Joists and Joist Girders 42nd Edition Steel Joist Institute Myrtle Beach, South Carolina 2007 Technical Digest No. #12 Evaluation and Modication of Open Web Steel Joists and Joist Girders Steel Joist Institute Myrtle Beach, South Carolina 2007 Designing with Steel Joists, Joist Girders and Steel Deck James Fisher, Michael West, Julius Van de Pas Nucor Corporation 1991 (Second Edition 2002) Miscellaneous Steel Joist and Joist Girder Specications and Load Tables SJI Archives Steel Joist Institute, Technology, Engineering, and Education Center Myrtle Beach, South Carolina Joist Investigation Form www.steeljoist.org/investigation Steel Joist Institute Myrtle Beach, South Carolina
Cazi U
Ue Tn
References
E R
STRUCTURE magazine
20
November 2009

Part 10
By D. Matthew Stuart, P.E., S.E., F. ASCE, SECB For this series of articles, antiquated has been dened 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 includes a compilation of miscellaneous systems and information for use by the practicing engineer. It is hoped that this nal 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.
ENGINEERS NOTEBOOK
Rmag T S
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 allowable values that were common in the 20th Century.
Additional Antiquated Systems
Cazi U
Building codes in New York City rst addressed masonry walls in 1830. The code provisions for brick became ht ig more complicated with r py each revision and, by Co the portion of the 1892, code dealing with masonry Draped Mesh Floor. was its most complex part. The NYC although partially embedded in the exterior code, like many other codes from different masonry walls, was only clad with what major cities, specied the minimum wall amounted to a brick curtain wall. All three thickness for varying heights of buildings. of these forms of construction co-existed The 1892 NYC code generally called for between 1880 and 1900. an increase of 4 inches (i.e., one wythe of brick) in wall thickness for each 15 Floor Framing feet down from the top of the building. Draped mesh slabs became popular in The minimum thickness for curtain the 1920s. Draped mesh construction masonry brick walls was generally 4 is a type of reinforced slab framing that inches less than that required for load- involves the use of wires that drape between bearing walls at the same height of the the tops of adjacent beams. The types of building. As it is sometimes difcult to mesh used included triangular wire mesh, ascertain the thickness of brick masonry ordinary wire mesh, expanded metal walls in existing buildings, a listing of sheets, plain round and square rods and the various minimum wall thicknesses twisted square rods. The use of wire mesh is provided here (Table 1, page was actually preceded by expanded metal 26) for a number of major sheets. Welding of wires together to form cities from the 1920s. the mesh did not begin until the 1930s. Load-bearing brick masonry Prior to that, the wires were attached walls were eventually replaced at the intersection points by staples or by cage and skeleton wrought washers, or by wrapping the transverse iron and steel frame con- wires around the longitudinal wires. struction, often using cast iron In a draped mesh slab, the concrete columns. Cage construction serves only as the wear surface and as the involved the use of brick faade mechanism by which the imposed loads walls that were as thick as those are transmitted to the mesh. The mesh used for load-bearing construc- alone is what physically spans between tion; the only difference was the beams by means of catenary action. that the frame and supporting Because the concrete is not structurally columns, including those that stressed in this type of system, the comwould eventually be embedded position and quality of the concrete is not in the brick masonry faade as important as in a true exural slab. As a wall, were rst erected ahead of result, it was common to use cinder conthe masonry. Skeleton framing, crete with compressive strengths under
Ue Tn
E R
Web Resources
Additional 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
20 th Century Loadbearing Masonry Building.
STRUCTURE magazine
24
January 2010
1,000 psi. The use of cinder concrete, however, 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. Brick arch oor 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 of the span. The spring line of the arch was constructed on top of the bottom ange of the supporting beams. The space above the arch was lled in with concrete, which sometimes had wood nailer strips embedded in the top of the slab. Tie rods were commonly placed about 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 PSF L = span length of the arch in feet R = rise of the arch in inches
CONCRETE
FINISHED FLOOR
TIE ROD TYPE (A)

TERRA-COTTA SKEW-BACK FLANGE PROTECTION
TYPE (B)
WITHOUT SKEW-BACK FLANGE PROTECTION
BRICK FLOOR-ARCH
Brick Arch Floor.
Cop
h yrig
Other common oor systems included: Fawcett System and Acme Floor-Arch clay lateral cylindrical tile at-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.
Rmag T S
The Kahn System.
Cazi U
Ue Tn
E R
A series of deformed bars.
Roebling Floor Arch arch of dense wire mesh supported on the top of the bottom anges of the beams covered with concrete. Manhattan System and Expanded Metal Company (EMC) Floor at 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 oor 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 at cross-sections with diamondplate-type deformations; the Rib bar, which included a hexagonal cross-section with cupshaped deformations; the American bar with square and round cross-sections and low circumferential depressions; the Scoeld bar with an oval cross-section and discontinuous circumferential ribs; the Corrugated bar with at, round and square cross-sections with cup deformations; the Slant bar with a at cross-section and low projecting diagonal ribs on the at 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 available in prefabricated trussed bar units. A truss
STRUCTURE magazine
25
January 2010
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.
Table 1: Minimum Building Code Thickness of Brick Masonry Walls Inches.
Total Stories
City Boston New York Chicago Philadelphia Denver San Francisco
Floor
1
st
nd
rd
th
5th
6th
7th
8th
12 12 12 13 12 17 12 16 16 18 16
12 12 12 13 12 13 12 16 12 12
Conclusions
Engineers involved with renovation and rehabilitation projects need to be aware of the specics of antiquated structural systems in order to develop non-destructive and unobtrusive 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 understanding of the system involved. Information concerning antiquated structural systems provided by this series of articles, and the referenced source material, has been compiled 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 publics lack of awareness about the hidden structural components of a building, which are typically enclosed after erection by the architectural nishes and therefore of less interest to a casual observer. This general lack of readily available information 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 construction 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 specic 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.
Boston New York Chicago t righ Denver
y Cop Philadelphia
San Francisco Boston
R T S
g a m
San Francisco Boston
Cazi U
12 12 17 17 13 16 12 12 New York Chicago 16 16 16 20 16 16 Philadelphia Denver 18 18 13 16 16 12 17 17 17 16 20 20 22 20 21 16 24 20 22 20 21 20 28 20 26 24 20 32 24 26 24 16 16 20 18 20 17 16 20 20 22 20 21 16 24 20 22 20 20 28 24 26 24 12 16 16 18 16 17 16 20 20 18 20 17 16 24 20 22 20 16 24 20 22 20 New York Chicago Philadelphia Denver San Francisco Boston New York Chicago Philadelphia Denver San Francisco Boston New York Chicago Philadelphia Denver
Ue Tn
12 12 13 13 12 12 12 13 12 13 12 16 16 13 16 17 12 16 16 18 16 17 16 20 20 18 20 16 24 20 22 20 12 16 16 13 12 13 12 16 16 13 16 17 12 20 16 18 16 16 20 20 18 20
E R
12 16 16 13 12 13 12 16 16 13 16 12 20 16 18 16 12 16 16 13 12 12 16 16 13 16 12 16 16 13 12
Boston New York Chicago Philadelphia Denver
STRUCTURE magazine
26
January 2010
Table 2.
Minimum Building Code Live Load - PSF Building Type Residential Hotels, Hospitals Ofce Buildings: First Floor 100 Upper Floors 60 Classrooms
19 th Century Train Station (30 th Street Station, Philadelphia, PA ).
New York Philadelphia 1927 40 40 1929 40 40
Boston Chicago Denver 1926 50 50 1928 40 40 1927 90
San Francisco 1928
40 & 60 40 40
100 60 50
125 60
125 40
ht Public Seating:rig py Co
75
Without Fixed Seats 100 Garages:
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 determine 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; exploratory 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 signicant 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 sometimes possible to determine the original live
Rmag T S
Warehouses 120 Manufacturing: Heavy 120 Light 120 Stores: Wholesale 120 Retail 120 Sidewalks 300
Cazi U
100 Public 120 100 Private 120 100 75 150 200 120 110 110 120
Fixed Seats 100
60
Ue Tn
50 75 100 75 100 125 150 100 100 125250 125250 250 125 250 125 250 125 250 250 125 150
E R
125 125 70 & 90 40 75 75 90 75 120 125 150 100 100 125-250 150 200 250 120 250 125 120 120 150 125 100 150
FYFE
Co. LLC Ty fo Fibr wrap Systems
Over 20 years ago we created the industry... today we set the standard
Structural Strengthening FRP Installation Seismic Upgrade Blast Mitigation Concrete Retrot 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
NSF
R
Tel: 858.642.0694
Fax: 858.444.2982 www.fyfeco.com
Certied to NSF/ANSI 61
STRUCTURE magazine
27
January 2010
Cop
Current Day, Jim Thorpe, PA.
h yrig
load for comparison to the proposed adaptive reuse. Individual building codes were commonly developed by different cities before the advent of national codes. These local codes often reected 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 approach 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 scientic experimentation and tests. Structural engi-
Rmag T S
Early 20 th Century Retail Arcade (Downtown Nashville Arcade).
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 because 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 records 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 eld 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 conrmed that the existing structural system does not have sufcient 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 independently or to provide supplemental support of the existing structure; and/or internally or externally reinforcing the existing system.
Cazi U
Strengthening of existing slab (Slag Block System).
Ue Tn
GPR printout.
E R
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 ofce of CMX located in New Jersey. Mr. Stuart may be contacted at mstuart@CMXEngineering.com.
Reference Material
Architects 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., McGrawHill Book Company, Inc., 1920
STRUCTURE magazine
28
January 2010

Antiquated Structural Systems (All)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Antiquated Structural Systems (All)

Uploaded by

Copyright:

Available Formats

Antiquated Structural Systems Series

aids for the structural engineers toolbox

Section through column head. STRUCTURE magazine

Section thru Column Head

226 Smulski on the S-M-I System of Flat-Slab Construction.

II Line of III inflection I II

Flat Slab Divided into Simple Parts.

Figure 1: Reprinted from Tile Engineering Handbook of Design.

Antiquated Structural Systems Series t

Clay Tile Arched Floor Systems

aids for the structural engineers toolbox

Figure 2: Reprinted from Tile Engineering Handbook of Design.

Figure 3: Reprinted from Tile Engineering Handbook of Design.

ADVERTISEMENT For Advertiser Information, visit www.STRUCTUREmag.org

Antiquated Structural Systems Series

aids for the structural engineers toolbox

Wood Sleepers Cinder Concrete Fill

ADVERTISEMENT For Advertiser Information, visit www.STRUCTUREmag.org

Antiquated Structural Systems Series

aids for the structural engineers toolbox

4", 5" or 6"

Precast tile joists

Typical slab section Concrete Topping

Precast tile oor slab

(a) 4-in. Kalex floor section.

(b) 6-in. Kalex floor section.

Precast Kalex slab section

TOPPING AS REQD 1 5/8 MIN

U Beam Joistile Shefeld

Prefabricated Clay Tile One-Way Beam and Slab Systems

Notes 1, 3 2, 3 4, 5 4 6 4 7, 8, 9, 10 11, 12 13, 14

Adel Joistile Kalex United Tilecrete

We can e fix it zin

STRUCTURAL STRENGTHENING SOLUTIONS

Antiquated Structural Systems Series

aids for the structural engineers toolbox

Ledger Beam Block Wall

Figure 3: Tee Joist.

prefabricated roof deck 12 C. G. C.

Figure 2: Rutten System.

Figure 4: Keystone Joist.

Concrete Beam Covering

Figure 6: Waites I Beam.

on FilCement T l to c il over e or Tera zz Pipes & cona duits

Field Concrete Fireproofing

dard Stant Slab ni U

Suspended Ceiling TYPE B

Figure 7a & 7b: Watson Floor System.

ADVERTISEMENT - For Advertiser Information, visit www.STRUCTUREmag.org

8 10 12 Depth , , Fin. Floor to Ceiling

12 Mortar Topping & Finish

Negative Bar Girder

Mortar Topping & Finish Var ies 9

Section A-A Plaster Locking Joint & Cross-Bridge Poured on Job

Firepf g Variable End Unit Column Tie

Standard Center Unit

Figure 8: Miller System.

Figure 9: Lith-I-Bar System.

ADVERTISEMENT - For Advertiser Information, visit www.STRUCTUREmag.org

Figure 12: Pyrobar Roof System.

Antiquated Structural Systems Series

Structural Steel Composite Stub-Girder Construction Part 6

FLOOR PLAN 10 - 0 DUCT

W16 x 26 STUB (BOTH SIDES) WELDED TO W10 W10 x 12