Development of Innovative Long-Span Floor Systems for Multi-Story Residential Steel Construction...

2007 NASCC Proceedings New Orleans, LA – April 18-21 Session E10 –

DEVELOPMENT OF INNOVATIVE LONG SPAN FLOOR SYSTEMS FOR MULTI-STORY RESIDENTIAL STEEL CONSTRUCTION

Amit H. Varma is an Assistant Professor of Structural Engineering at Purdue University. He

received his B.S. of Civil Engineering in 1994 from The Indian Institute of Technology. In 1996 he received his M.S. in civil engineering from the University of Oklahoma and his PhD in civil engineering from Lehigh University in 2001. Dr. Varma’s research interests include seismic and fire behavior, analysis, and design of steel-concrete

composite structures and repair and retrofit of damaged or deteriorating infrastructure.

Devin H. Huber is a PhD student and NSF Graduate Fellow in Structural Engineering at Purdue

University. He received his B.S. in Civil Engineering in 2003 from Michigan State University. Mr. Huber has been working on the AISC fellowship project focusing on the development of long span floor systems in residential steel construction since Fall

2003.

ABSTRACT

Typical steel deck-concrete slab floor systems used in multi-story construction require intermediate filler or support beams (W14 or deeper) and are limited to spans of 8-12 feet depending on deck depth. The goal of this AISC fellowship project is to conceive, develop, and validate innovative long span floor systems capable of spanning up to 30 feet with total depths up to 12 inches or less. This paper highlights three systems that are strong candidates to achieve these goals of long span and reduced depth. These are: (1) the composite deep deck system, (2) reduced weight composite deep deck system, and (3) the pre-loaded self-shoring system. The behavior of these systems in the construction, service, and ultimate phases were modeled and investigated analytically using 3D finite element models and other simpler models. This paper focuses on the analytical evaluation and experimental validation of candidate (1). The analytical and experimental results indicate the composite deep deck system will perform adequately in long span (30 ft) multi-story applications.


INTRODUCTION/MOTIVATION

Conventional floor systems for steel structures in the United States typically consist of trapezoidal steel decks composite with concrete slabs. The steel deck profiles have depths of 1.5, 2, or 3 in. with 24 to 16 gage thicknesses. These steel deck-concrete slab systems behave as one-way slabs with unsupported spans limited to approximately 15 ft (for residential and office live loads). Lightweight concrete and 3 in. deep steel deck with 16-gage thickness are required to achieve the 15 ft span. Spans in the range of 8-10 ft are more common when normal weight concrete and thinner decks are used (Easterling, 2002). This limitation on the span means that intermediate filler beams are required to support the slab systems. These filler or support beams reduce the useable headroom and increase the floor-to-floor (story) height of the structure. This limitation can make steel structures with conventional floor systems less efficient for multi-story residential construction.

AISC has funded a four-year fellowship project focusing on the conception, development, and validation of

long-span floor systems (span up to 30ft. unsupported) for multistory residential construction. This project is currently ongoing, and the paper summarizes the progress and major findings of the project. Over the past three years, several long-span floor systems have been conceived and considered by the research team and the advisory panel. These floor systems were evaluated based on their ability to achieve certain performance objectives that were established by the research team (Huber and Varma, 2005). These include the ability to:

1. Achieve long spans of up to 30 ft. unsupported. 2. Achieve minimum floor-to-floor height requirements 3. Utilize unshored construction (preferably). 4. Achieve strength and serviceability criteria during the construction phase 5. Achieve strength and serviceability criteria during the service phase of the structure 6. Achieve adequate performance with respect to floor vibrations 7. Achieve adequate fire resistance 8. Use relatively simple connection systems 9. Provide cost-effective fabrication and erection 10. Provide adequate diaphragm action for transferring lateral loads This paper does not list all the systems that have been considered. It presents three long-span floor systems that

have been identified as strong and viable candidates by the research team. Further analytical evaluation and experimental validation will focus on these floor system candidates. This paper presents the general description and design layouts of the three floor systems. The analytical evaluation and experimental validation of one of these floor system candidates is presented in detail. Finally, the paper summarizes the future work for the project.

Current Progress / Summary

The fellowship project is being conducted in four tasks. The first task focused on conducting a literature review

and survey of existing long-span slab systems. This task was conducted to determine the state-of-art for composite floor systems. It helped in identifying existing solutions that have been proposed or implemented in steel construction. This task has been completed, and the findings are reported in Huber and Varma (2005).

The second task focused on conceptual development and design of long span slab systems. A suite of different

floor system types were developed and proposed. These systems were ranked based on their technical merit and potential to achieve the performance objectives noted above. The systems were analyzed and designed using analytical tools and methods including the finite element method (FEM), numerical analysis, and existing design codes. Based on the rankings, three floor systems were selected for further development and experimental validation. This task and been completed, and the findings are presented in Huber and Varma (2006).

The third task focuses on experimental validation of the floor system candidates. This task is currently in

progress. The experimental validation of one of the floor system candidates will be presented later in the paper. The details of the experimental setup, testing procedures, instrumentation etc. will also be included.

The fourth task focuses on numerical investigations and analytical parametric studies of the long-span floor

system candidates. This task is also currently in progress. The experimental and analytical results from tasks three and four will be used to develop: (a) design guidelines and recommendations for the floor systems, (b)


recommendations for implementing the floor systems in multi-story residential structures, and (c) guidelines for modeling these systems for future research and engineering development if needed.

DESCRIPTION OF DEVELOPED FLOOR SYSTEMS This section presents the descriptions and design details of three floor system candidates that were developed in Task 2 and identified as strong viable candidates by the research team and the advisory panel. Experimental validation of these systems is currently ongoing. The structural performance of the first system has been validated experimentally. It will be presented in the section focusing on experimental investigations. Composite Deep Deck (CDD) System

The composite deep deck (CDD) system consists of a deep deck profile composite with the concrete slab on

top. The CDD system can achieve spans of up to 20-30 ft. by using deck profiles with depths ranging from 6.0 to 7.5 in. The deeper steel decks provide more stiffness in the construction (non-composite) phase, which typically governs the overall design. Figure 1 shows the cross-section of the CDD system designed for achieving the 20 ft. span. The system consists of a 6 in. deep deck with 14-gage thickness and 2.5 in. thick concrete slab on top. Figure 2 shows the cross-section of the CDD system designed for achieving the 30 ft. span. The system requires a 7.5 in. deep cellular deck profile with 14-gage thickness, bottom plate with 16-gage thickness, and 2.5 in. thick concrete slab on top to achieve the 30 ft. span.

Figure 1: Composite Deep Deck System – 20 ft Span

Figure 2: Composite Deep Deck System – 30 ft Span

Both CDD systems were designed with simply supported end conditions. Normal weight concrete (NWC) was

assumed for the 20 ft. span, and lightweight concrete (LWC) was used for the 30 ft. span CDD system. Lightweight concrete was used to reduce the self-weight of the system and to achieve the deflection requirements in the construction phase. However, the use of lightweight concrete is also desirable from a fire resistance standpoint. The 20 ft span can also be used with lightweight concrete if desired.

The connections between the CDD system and the supporting (end) girders were designed to be similar to

existing deck to beam connections. One possible connection type is shown in Figure 3. For this connection, the flutes of the deck will be aligned and butted against one another. Holes will be fabricated in the bottom flutes of the deep or cellular deck profile because they are thicker than the deck profiles for conventional floor systems. The


shear studs will be welded to the supporting girder flanges through these holes. The shear studs are provided to delay the onset of end slip between the deck and the concrete. The desired composite action will not be realized fully after end slip occurs. Researchers have quantified the ability of the shear studs to prevent localized end slip and enhance the performance of composite slab systems (Widjaja, 1997). Monitoring end slip is a key consideration in the development and testing of these CDD floor systems.

Figure 3: Possible End Detail for Composite Deep Deck System

The major advantage of the CDD system is that it eliminates the need for intermediate floor beams for spans up

to 30 ft. This reduces the total depths of the floor systems. For example, the total depths of the CDD systems designed for spanning 20 ft. and 30 ft. are 8.5 in. and 10 in., respectively, which is remarkable. The drawback of the CDD system is that it uses more weight of steel deck per foot as compared to conventional floor systems. The weights of the deep deck profiles are approximately 2.5 to 3.0 times the weights of steel deck profiles (1.5 - 3.0 in. deep) used for conventional floor systems. Hence, their associated costs will also be higher than those for conventional deck profiles. Some of these costs will be offset by: (a) eliminating intermediate filler beams and the shear studs associated with them, and (b) reducing floor-to-floor heights.

Another advantage of CDD systems is that the total weight of the floor system (with concrete) is comparable to

the total weight of conventional floor systems. The total weight of the CDD floor system is much lower (approximately 47%) than the weight of floor systems using pre-cast planks that span up to 30 ft. (8 in. plank with 2 in. topping). Additional details of the weight comparisons are provided in Huber and Varma (2005).

Reduced Weight Composite Deep Deck (RW-CDD) System

The reduced weight deep deck (RW-CDD) system also uses the 7.5 in. deep deck profile with 14-gage thickness for spanning up to 30 ft. However, this system eliminates the bottom plates, i.e., the cellular option to save costs associated with fabrication and material. This reduces the stiffness of the non-composite deck profile in the construction phase. To offset this reduction in stiffness, the concrete in the lower flutes is replaced by a lightweight filler material such as mineral wool or extruded polystyrene (currently used in acoustical deck profiles). This reduces the concrete weight during construction and the final self-weight of the system.

Two variations of this system have been developed, one using the deck profiles in a standard configuration and

one in which the deck profiles are inverted. The cross sections of the floor systems using both these configurations are shown in Figure 4 and Figure 5. Inverting the deck profile provides more steel area at the bottom, thus making it slightly more efficient. The inverted profile also provides more area for the placing the shear studs at the ends.


Figure 4: Cross-Section of Reduced Weight Deep Deck System – Standard Configuration

Figure 5: Cross-Section of Reduced Weight Deep Deck System – Inverted Configuration

The major advantage of the RW-CDD system is the potential cost saving due to the elimination of the bottom

plate. Another key feature is the reduced weight of the concrete, and thus the reduced weight of the total system. If normal weight concrete is used, the system self-weight will be approximately 40psf, which is much lower than similar CDD systems. The end conditions and connections for the RW-CDD system can be similar to those for the CDD system shown in Figure 3. A slight modification is that concrete will be allowed in the lower flutes over and around (about 2-3 ft.) the connection region to enhance its performance. This additional area of concrete in the lower flutes around the connection region will provide better bearing between the shear stud and the concrete, and delay the onset of slip in the system. This will also help in transferring diaphragm forces from the floor system to other collector elements in the structure.

The RW-CDD system has not yet been validated experimentally. So, the description and designs mentioned

above may change slightly after the experimental investigations are completed. The advantages and drawbacks of the RW-CDD system are similar to those listed for the CDD system. There are some additional concerns regarding the fire resistance of the system and its susceptibility to floor vibration due to the reduction in the concrete weight (mass). These concerns will be addressed during the experimental validation part of the project. Experimental investigations will be performed in the near future to determine the heat transfer behavior of the RW-CDD system subjected to heating following the standard E119 temperature-time (T-t) curve. The natural frequencies corresponding to the first and higher modes will also be determined in the experimental portion of the work. The higher mode effects on the floor vibration of the system will be evaluated analytically.

Analysis and Design of the Deep Deck Systems


Three different approaches were used to analyze and design the deep deck systems (CDD and RW-CDD). The

results from these analyses are summarized in Huber and Varma (2006). The analytical results for the 30 ft. span CDD system will be presented along with the experimental results in a later section.

• The first was a simple approach using the design provisions provided by the Steel Deck Institute (SDI, 2002). SDI provides simple design equations for use in the non-composite phase and recommends the use of reinforced concrete design theory (ACI, 2005) when analyzing the composite sections.

• The second approach conducted nonlinear inelastic analyses to determine the complete load-deflection behavior of the composite floor systems. Fiber models were developed for the floor system cross-section assuming full composite action, i.e., no slip. These models were analyzed to determine the moment-curvature (M-φ) response of the cross-section. The complete load-deflection behavior of the system was calculated by discretizing the length into several section points, satisfying static equilibrium at the section points, and using the central difference method to integrate curvatures and rotations to calculate the deflections at different load levels.

• The third approach conducted 3D finite element analyses to determine the complete load-deflection behavior of the composite floor systems. The 3D finite element models accounted for the complex geometries of the floor systems, nonlinear material and geometric properties, and the simply supported end conditions. The models assumed full composite action, i.e., no slip between the deck and the concrete. The results from the finite element analyses included the complete load-deflection behavior and stress states of the floor systems. The finite element models were modified to account for the effects of slip between the deck and the concrete. Additional finite element analyses will be conducted in the future using the experimental results to define or calibrate the behavior of the slip interface model between the deck and the concrete.

Pre-Loaded Self-Shoring (PLSS) System

The pre-loaded self-shoring (PLSS) system is a simple modification of the conventional steel deck-concrete slab

floor system. It was designed to reduce the overall depth of the conventional floor system. Preliminary analysis indicated that the total depth of the conventional floor systems is governed by the depth of the intermediate filler beams (about 14-16 in. deep). The depth of the filler beams is controlled by the deflection limits (L/180) and stiffness requirements during the construction phase of the floor system. The PLSS system pre-loads the intermediate filler beams during the construction phase to offset the deflections produced by the construction loads and the weight of the concrete. As a result, the deflection limits can be achieved while using shallower beam sections (8 in. deep) with smaller moments of inertia.

Figure 6: Schematic of Pre-Loaded Self-Shoring PLSS System in the Construction Phase

Figure 6 shows a schematic of the PLSS system in the construction phase. As shown in the Figure, a screw jack

or other similar means of applying force is placed at the beam mid-span. The screw jack is placed on a column stub that is supported by threaded rods attached to clevises and connected to gusset plates at the ends of the beam. The screw jack is used to apply an upward force to the beam in the construction (non-composite) phase before placing the deck. This upward force (and the resulting internal moment diagram) will counter the effects of the concrete


self-weight and the construction loads. The upward deflection will offset (subtract from) the total deflections produced by the concrete weight and construction loads.

As shown in the Figure, the pre-load is applied by attachments that are supported by the beam itself. Hence, the

beams are pre-loaded and self-shored during construction. The pre-load, screw jack, and the attachments are removed after the concrete cures. These can be used for the subsequent floor systems being constructed. Figure 7 shows the cross-section of the floor system after removing the pre-load in the composite phase. The PLSS system looks similar to the conventional floor system with the exception that the floor beam depth is reduced by about 50% to approximately 8 in. The self-weight of the floor beam will remain in the range of 30 – 40 lb/ft., which is similar to conventional floor systems.

Figure 7: Schematic of PLSS System in the Composite Phase

The design details of the PLSS have been developed by conducting analytical investigations of the system

behavior: (a) after pre-loading, (b) during construction, (c) after removal of pre-load, and (d) during the composite phase. The analytical investigations included the effects of geometric nonlinearity and the second-order effects of the axial forces and end moments induced by the pre-load. The analytical investigations are beyond the scope of the paper, and are presented in detail in Huber and Varma (2005).

Table 1: Design Table for PLSS System

Beam Spacing (ft)

Span (ft)

Self Weight (psf)

Upward Force (kips)

Ix beam (in.4)

Zx beam (in. 3)

Possible Section

Applied upward displacement (in.)

20 40 3 33 3 W8x18 1.0

30 40 4.5 110 6 W8x31 1.5 6

40 40 6 245 12 W10x45 2.0

20 45 4.5 46 4 W8x21 1.0

30 45 6.6 127 10 W8x35 1.75 8

40 45 8.5 300 17 W10x54 2.25

20 45 5.25 70 5 W8x24 1.0

30 45 8 125 12 W8x35 1.75 10

40 45 10.5 345 20 W12x45 2.5

Table 1 summarizes the design details for the PLSS system. It shows the design values for the PLSS system

with beam spacing equal to 6, 8, or 10 ft. and spans equal to 20, 30, or 40 ft. The design values include the required upward forces and displacements at mid-span, the corresponding moments of inertia (Ix) and plastic section modulus (Zx) required for the filler beams. The Table also includes an example section that can be used with the particular PLSS system. For example, a W8x35 section can be used for the PLSS system with beam spacing equal to 8 ft. and span equal to 30 ft. Figure 7 shows the details of this PLSS system.


The experimental validation of the PLSS system has not yet been completed. The experimental validations will

be conducted in the near future, and the design details may change slightly to reflect the findings.

EXPERIMENTAL VALIDATION OF FLOOR SYSTEMS The experimental validation of the floor system candidates is being conducted in Bowen Laboratory at Purdue

University. The experimental investigations focus on three major issues, which are summarized below:

1) Testing for structural behavior. These tests will be conducted to determine the structural behavior of the floor systems including the deflections, longitudinal strains and stress, and end slip: (a) during construction and after concrete placement, (b) at service-level loads, (c) at ultimate loads, and (d) for monotonically increasing loading up to failure or excessive deflections. Additionally, floor vibration tests will be conducted to determine the fundamental frequency the floor system by subjecting it to excitation and measuring accelerations and displacement-time histories.

2) Testing for heat transfer behavior. These tests will be conducted to determine the heat transfer behavior of the floor system specimens subjected to heating following the ASTM E119 temperature-time (T-t) curve on the bottom face. The heat transfer behavior through the floor system cross-section will be measured using thermocouples as recommended by the ASTM standard.

3) Testing for structural behavior at elevated temperatures. These tests will be conducted to determine the behavior of the floor systems subjected to combined thermal and structural loads. The floor system specimens will be subjected to nominal (service-level) loads and thermal loading in the form heating following the E119 T-t curve on the bottom face.

As mentioned earlier, the experimental validation of the floor systems is currently ongoing. This paper presents the results from the testing for structural behavior of the CDD system. The experimental investigations focused on the structural behavior of the 30 ft. span CDD system using the cellular deck profile shown in Figure 2. The testing for heat transfer behavior for this CDD system is currently being conducted at Bowen Laboratory. Experimental Setup

Figure 8 shows a schematic plan view of the test-setup for determining the structural behavior of the floor system specimens. As shown, the floor system specimens are 4 ft. wide and 30 ft. long. They are simply supported at the ends by the reaction frames identified in Figure 8. The end connections between the floor system specimens and the beams of the reaction frames are shown in Figure 3. Two loading frames are located close to the 1/3 point (9 ft. from each end) of the floor system specimens. Hydraulic actuators are attached to the loading frames and used to apply concentrated loads to the floor system specimens.

The bending moment distribution produced by this loading is trapezoidal with the middle third subjected to

uniform bending moment. This trapezoidal bending moment distribution closely approximates the parabolic moment distribution produced by a uniformly distributed load. The concentrated loading (P) can be used to compute an equivalent uniformly distributed surface loading (w) by equating the mid-span bending moment for both cases. The equivalent w (in psf) is computed as the applied load P (in kips) multiplied by 20.

The typical instrumentation layout for the floor system specimens is also shown in Figure 8. The floor system

specimens were instrumented to measure deflections and longitudinal strains in the construction (non-composite) and ultimate (composite) or monotonically increasing load phases. Most of the instrumentation was attached prior to placing the concrete. Figure 9 shows the location of the longitudinal strain gages in the floor system cross-section at the mid-span and quarter points. As shown, longitudinal strain gages were attached to the steel deck, bottom plate, and concrete slab. Figure 10 shows photographs of the CDD floor system specimen before and after placing the concrete.


Figure 8: Plan View of Test Setup with CDD specimen.

Figure 9: Strain Gage Locations in Specimen Cross-section at Mid-Span

Figure 10: CDD Floor System Specimen (a) Before and (b) After Concrete Placement


Construction (Non-Composite) Phase Test

The construction phase behavior was determined experimentally by: (a) subjecting the non-composite specimen to loads simulating the standard construction loading, and (b) monitoring the behavior of the specimen during construction and concrete placement. The standard construction load (w) specified by SDI (2002) is equal to 20psf. It was applied by subjecting the non-composite specimen to concentrated loads equal to 1-kip at the loading frame locations shown in Figure 8. The deflections and longitudinal strains in the specimens were monitored after applying the loading, and also during construction and concrete placement. Frequency Determination Test

The fundamental frequency of the composite floor system specimen was determined by: (a) subjecting it to impulse excitation using a calibrated impact hammer, and (b) measuring its response using accelerometers. The acceleration response of the specimen was transformed into the frequency domain by performing a Fast Fourier Transform (FFT). The results from the FFT were used to determine the fundamental frequency of the floor system. Ultimate Load Testing

The ultimate load test was conducted by subjecting the floor system specimen to monotonically increasing loading using the hydraulic actuators attached to the loading frames shown in Figure 8. The loading was applied monotonically in load control until the system started deforming non-linearly. Further loading was in displacement control until the mid-span deflections reached about 8-10 in., corresponding to the limiting stroke of the actuators. The behavior of the floor system specimen was monitored carefully during the tests. Concrete cracking was monitored and marked on the specimens throughout the test.

Experimental Behavior and Results – Composite Deep Deck (CDD) System

The CDD floor system designed for spanning up to 30ft. was presented earlier. The floor system cross-section

was shown in Figure 2 and the possible end connection detail was shown in Figure 3. Two similar specimens of this CDD floor system were fabricated and tested using the setup and procedures described in the previous section. The experimental results for these two specimens are summarized in this section.

Construction (Non-composite) Phase Results Table 2 summarizes the maximum deflections and strains measured during: (a) the construction load test, and (b) the concrete placement. As shown in Table 2, the specimens behaved elastically during the construction load tests. All the deflections and strains were recovered after the construction loads were removed. The results in Table 2 indicate that the maximum deflections and strains occur during concrete placement. The final deflections and strains after concrete placement were permanent because the system was un-shored. Both specimens had a permanent deflection of 1.1 in. after the concrete was placed.

Table 2: Non-Composite Phase Results

Specimen 1 Specimen 2

Construction Loading Concrete Pour Construction Loading Concrete Pour

Maximum Mid-Span Deflection During Construction (in.)

0.43 1.5 0.42 1.67

Final Mid-Span Deflection After Construction (in.) 0 1.1 0 1.1

Maximum Tensile Steel Strain (µs)

105 500 (14.75 ksi) 100 520 (15 ksi)

Residual Tensile Steel Strain (µs) 0 400 (12 ksi) 0 455 (13.4 ksi)

Frequency Test Results


The fundamental frequencies for both CDD specimens were found to be equal to 6 Hz. The frequency response spectrum from the impact hammer test of specimen 2 is shown in Figure 11. It identifies the fundamental frequency of the system to be equal to 6 Hz.

Figure 11: Frequency Response of CDD Specimen 2

Ultimate Load Test Results The complete load vs. midspan deflection responses of specimens 1 and 2 are shown in Figure 12. The Figure includes the load-deflection responses predicted by the fiber model analysis and the 3D finite element analysis described earlier. The service load (3 kips or 60psf) and ultimate load (7.75 kips or 155psf) levels for the floor system specimens are also identified in Figure 12. The ultimate load capacity predicted using reinforced concrete beam design theory recommended by SDI (2002) is also shown in the Figure.

Figure 12: Load vs. Mid-Span Deflection Behavior for CDD Specimens 1 and 2


EVALUATION OF RESULTS

Construction Phase (Non-composite) The construction phase results shown in Table 2 indicate that the highest deflections and strains in the non-

composite deep deck occurred during construction when the concrete was being placed by workers standing on top of the deck. The maximum deflections were equal to 1.5 and 1.67 in. for specimens 1 and 2, respectively. The slight difference between the deflections of the two specimens is due to the change in the number and weight of workers on top of the decks. The permanent deflections after construction and concrete hardening were equal to 1.1 in. for both specimens. This corresponds to a deflection-to-span ratio of approximately L/360. The current Steel Deck Institutes specifications limit the deck deflection to L/180 or 0.75 in (SDI, 2002). This limitation was designed by decks that spanned a maximum of 10-12 ft. It was not meant for decks that span up to 30 ft. The permanent deflections of the cellular deep deck profile after concrete placement are reasonable and acceptable. Frequency/Floor Vibration Testing

The fundamental frequency of the CDD floor system specimens was determined to be equal to approximately 6

Hz. This value compares reasonably with the frequency determined by performing simple vibration analysis according to AISC Design Guide 11 (1997). The frequency determined from the vibration analysis was found to be 5.9 Hz. The assumptions involved in this analysis included:

- The boundary conditions are simply supported. - The mass contributing to the floor system comes only from the bay in which is being considered. - The effects of girder stiffness are included by assuming girder deflections of 0.5 in. - The floor system will have 4% damping when used in residential construction.

Continuing with these assumptions, Equation 4.1 from AISC Design Guide 11 can be used to solve for the ratio of peak acceleration (ap) to gravitational acceleration (g). The ap/g value for the CDD system was determined to be equal to 0.45%, which is than the limit or maximum threshold (0.5%) recommended by DG 11. This simple analysis indicates that the CDD floor system would perform adequately for floor vibrations due to walking excitation.

The simple analysis highlights an important insight regarding long span floor systems and floor vibrations. The

analysis shows that as floor systems get longer, floor vibrations do not necessarily get worse. In fact, in the United Kingdom a deep deck system called Slimdek has been in use for several years without floor vibration problems. The Steel Construction Institute (SCI) in the UK has published a new design guide for analyzing floor systems that use these types of systems (SCI, 2004). Using these guidelines, the CDD floor system is also shown to be quite adequate from a floor vibration standpoint (Huber and Varma, 2006).

Composite Phase – Ultimate Load Test Results

The load-deflection responses shown in Figure 12 indicate that the behavior of both CDD specimens is quite

similar and comparable. The deflections of the CDD specimens for the service-level loads (3kips or 60 psf) were approximately equal to 0.35 in., which is less than the acceptable value of L/360 = 1.0 in. The ultimate load capacity of the CDD specimens was approximately equal to 24 kips or 480psf (equivalent), which is greater than the required load capacity of 155psf for residential construction. The 30 ft. long CDD floor system specimens had adequate serviceability stiffness and strength for multistory residential structures.

The load-deflection responses indicate that the CDD specimens have significant reserve load capacity and

ductility when compared with the required ultimate load capacity of 155psf. The tests were not terminated due to possible or imminent failure. They were terminated because the hydraulic actuators reached their stroke limits. The experimental results compare favorably with the analytical results from the three approaches presented earlier. The results from the fiber analyses and the 3D finite element analyses compare reasonably with each other and the experimental results. The fiber model analysis over predicts the ultimate load capacity because it does not account for the initial stresses in the steel deck due to the weight of the concrete slab.


There is some deviation between the experimental load-deflection behavior and the analytical predictions at higher load levels. This deviation occurs around 15 kips or 300psf loading and 3 in. mid-span deflection. The reasons for this deviation are being studied. Some possible reasons include the occurrence of end slip and discrepancy between the assumed and actual material properties.

SUMMARY AND CONCLUSIONS The paper summarized the current progress of the AISC Fellowship project focusing on the development and

validation of long-span floor systems for multi-story residential construction. Several floor systems have been conceived, developed, and evaluated by the research team and the advisory panel. The systems were evaluated and ranked based on their potential to achieve the performance objectives defined for long-span (up to 30 ft.) floor systems. Based on these rankings, three floor systems were identified as strong and viable candidates for further development and experimental validation.

The three floor system candidates were presented in detail. These include the: (i) composite deep deck CDD

system, (ii) reduced weight composite deep deck RW-CDD system, and (iii) pre-loaded self-shoring PLSS system. The 30 ft. long CDD system consists of 7.5 in. deep cellular deck profiles composite with 2.5 in. deep concrete slab on top. The total depth of the CDD system is about 10 in. The 30 ft. long RW-CDD system consists of 7.5 in deep non-cellular deck profiles with mineral wool in the flutes and 2.5 in. deep concrete slab on top. The total depth of the RW-CDD system is also 10 in. The PLSS system consists of conventional (1.5-3.0 in. deep) steel deck-concrete slabs supported by shallower (8 in. deep) intermediate steel beams. The steel beams are pre-loaded at the mid-span during construction by a self-equilibrating attachment, which also serves as shoring for the floor system. The 30 ft. long PLSS system has a total depth of about 13.5 in.

The floor system candidates were designed using simple analytical approaches, and verified analytically using

simple and complex models. The experimental validation of the floor system candidates is currently in progress. The validation of the 30 ft. long CDD system was presented in detail. The experimental results indicate that the CDD system achieves the stiffness and strength requirements for multistory steel construction. The fundamental frequency of the CDD system is 6 Hz., and it satisfies the floor vibration requirements of AISC Design Guide 11. The heat transfer behavior of CDD system subjected to heating following the ASTM E119 T-t curve will be investigated in the near future.

FUTURE RESEARCH

The future work for this project will focus on experimental validation of two remaining floor system candidates. The structural behavior will be validated using the test setups, procedures, and instrumentation layouts presented in this paper. The heat transfer behavior of the three floor system candidates will be investigated experimentally in the near future. The experimental results will be used to validate simple 2D heat transfer analysis models using the finite difference and the finite element methods. The fire resistance of the floor system candidates will also be investigated experimentally by subjecting to nominal structural and thermal loads. The results from the experimental and analytical investigations will be used to develop simple design guidelines and recommendations for the long-span floor systems.


REFERENCES

American Society for Testing and Materials (ASTM) (2000). Standard Test Methods for Fire Tests of Building Construction and Materials, Specification No. E119-00, West Conshohocken, PA.

American Concrete Institute (2005). Building Code Requirements for Structural Concrete, Farmington Hills, MI. AISC (1997). Steel Design Guide No. 11: Floor Vibrations Due to Human Activity, Chicago, IL. Easterling, W.S. (2002). “Long Span Composite Slabs – Extending The Reach of Steel Construction.” Proc. North

American Steel Construction Conference, Seattle, WA, AISC. Huber, D.H. and Varma, A.H. (2005). Development of Innovative Long-Span Slab Systems for Multi-Story

Residential Steel Structures: Progress Report, annual progress report submitted to AISC Committee of Research.

Huber, D.H. and Varma, A.H. (2006). Development of Innovative Long-Span Slab Systems for Multi-Story

Residential Steel Structures: Progress Report, annual progress report submitted to AISC Committee of Research.

Steel Construction Institute (SCI) (2004). SCI Publication P331: Design Guide on the Vibration of Floors in

Hospitals. SCI, Nerkshire, UK. Steel Deck Institute (SDI) (2002). Design Manual for Composite Decks, Form Decks and Roof Decks -No. 30. SDI,

Fox River Grove, IL. Widjaja, B.R. (1997). “Analysis and Design of Steel Deck Concrete Composite Slabs.” Ph.D. Dissertation, Virginia

Poytechnic Institute and State Universtiy, Blacksburg, VA.

ACKNOWLEDGEMENTS

This research is funded by the AISC Fellowship project awarded to Dr. Amit Varma. Mr. Tom Schlafly is the program manager. The advisory panel includes Mr. Lawrence Kruth of Douglas Steel, Mr. Larry Novak of Skidmore Owings and Merill, Mr. Ted Hazeldine of Benchmark Fabricated Steel, and Dr. W. Samuel Easterling of Virginia Tech. The steel decks for the research project were donated generously by Consolidated Systems Inc. The loading and reaction frames were subsidized by Benchmark Fabricated Steel. Equipment and assistance for the floor vibration testing was provided by Dr. Douglas Adams and Nathan Yoder of the Mechanical Engineering Department at Purdue University. Additional funding was provided by the National Science Foundation (NSF) in the form of a Graduate Research Fellowship Award to support the Ph.D. student (Mr. Devin Huber).

Development of Innovative Long-Span Floor Systems for Multi-Story Residential Steel Construction...

Documents

Transcript of Development of Innovative Long-Span Floor Systems for Multi-Story Residential Steel Construction...