A Tension-Controlled Open Web Steel Joist

A Tension-Controlled Open Web Steel Joist

DISCLAIMER: No joist will withstand sudden and catastrophic impact forces that exceed system capability. Flex-Joist design offers probability of high ductility and time delay under static gravity overload conditions.

Purpose: Improved Ductility and Reliability under Static Gravity Overload

Flex-Joist Engineered Limit States Intentionally imbalanced member strength ratios Weaker components serve as ductile fuse Initial limit state of ductile yielding in primary tension members Other limit states inhibited until advanced state of collapse

What is so great about ductility? Reduced Probability of Collapse Improved Structural Reliability Reduced Variance in Strength Reserve Inelastic Capacity Load Sharing with Adjacent Joists

What is so great about ductility? Increased Probability of Safe Evacuation Slower Collapse Mechanism Sensory Warning via Large Inelastic Deflections

What is so great about ductility? Improved Structural Reliability: Reduced Variance in Strength Influence of Variance on Reliability Which population has the greatest probability of a value below 1.0?

What is so great about ductility? Improved Structural Reliability: Load Sharing Idealized parallel system sketch Load shared equally between components

What is so great about ductility? Improved Structural Reliability: Load Sharing Sudden Strength Loss (lack of ductile behavior) Load dumps to remaining components (progressive collapse) System strength limited by weakest component System variance equals variance of individual components population

What is so great about ductility? Improved Structural Reliability: Load Sharing Idealized parallel system sketch Load shared equally between components Elasto-Ductile system

What is so great about ductility? Improved Structural Reliability: Load Sharing Ductile behavior Weakest member continues to support plastic capacity after exceeding elastic limit System strength a function of average component strength System Variance: Vs = V n

What is so great about ductility? Slower Collapse Mechanism with Sensory Warning Compressive Buckling Ultimate Strength Compression Element Buckling Design Strength

What is so great about ductility? Slower Collapse Mechanism with Sensory Warning Ductile Tensile Yielding Ultimate Strength Design Strength Tension Element Yield Compression Element Buckling

When Loads Exceed Capacity of a Flex-Joist Flex-Joist Load/Deflection Data Plot

Flex-Joist Design Reliability Study Ratio of Plastic Strength / Experimental Design Load From Villanova Data Plastic Strength Ratio All Average 1.033 Std Dev 0.030 COV 0.029 Qty 18 Series Sample LRFD Design Load (plf) Fy Experimental (ksi) Adjusted Design Critical Load (plf) Plastic Strength (plf) Ratio Plastic / Adj Crit Load J1-1 568 1.01 J1-2 574 1.02 K-Series J1-3 567 1.01 418 60.3 560 J1-4 589 1.05 J1-5 592 1.06 J1-6 582 1.04 J2-1 1878 1.07 J2-2 1882 1.07 LH-Series J2-3 1886 1.07 1303 60.6 1755 J2-4 1852 1.06 J2-5 1868 1.06 J2-6 1855 1.06 Rod-Web- Series J3-1 582 1.01 J3-2 589 1.03 J3-3 567 0.99 420 61.5 574 J3-4 568 0.99 J3-5 572 1.00 J3-6 566 0.99

Flex-Joist Design Reliability Study Steel Dynamics Roanoke Bar Division A529-50 merchant bar May 2008 to October 2012 11546 samples / 4337 batches Stat's Yield Stress (psi) Ratio Yield Stress / 50 ksi min Average 56764 1.1353 Minimum 50000 1.0000 Maximum 76570 1.5314 Std Dev 3415.6 0.0683 COV 0.0602 0.0602

Flex-Joist Design Reliability Study Structural Reliability Analysis: φ = 0.90 Live / Dead Load Ratio = 3 β = ln C φ φ M mf m P m V M 2 +V F 2 +C P V P 2 +V Q 2 β = 3.2

Summary of Flex-Joist Design Characteristics Joist Performance Comparison Criteria Std Joist Flex-Joist % Diff Joist Strength Reliability β 2.6 3.2 22% System Strength Reliability β 2.6 3.4 31% Average ASD Test Strength Ratio 1.8 2.3 29% Average Test Ductility Ratio 1.4 3.2 129% Tension Limit State Probability Low High Electronic Monitoring Suitable Okay Excellent Average Relative Weight 100% 107% System β based on N = 4 statistically unlinked joists working in parallel

Summary of Flex-Joist Design Characteristics Approximately 30% higher Reliability Index (β). Approximately 7% heavier, on average. Clearly room for potentially reducing weight while retaining superior reliability. Subject to justification being provided to support a higher φ y value and/or lower Ω y value, in an ICC Engineering Services Report submittal. Limited applications until fire testing has been performed

Tension-Controlled Joist Limiting Design Factors Conditions preventing the Bottom Chord and End Web from developing their tensile capacity: Unusually high material Fy High compression under net uplift loads, axial loads, or end moments Unusually strict deflection criteria Minimum material size criteria Unnecessarily strict tension member slenderness criteria Uniformly distributed loading on a 20K7 steel joist with a base length of 33 Lowest Stress Highest Stress

Tension Slenderness Ratio Current SJI maximum slenderness ratios are based on the 1946 & 1949 AISC spec s, as follows: For main compression members 120 For bracing and other secondary members in compression 200 For main tension members.240 For bracing and other secondary members in tension...300

Tension Slenderness Ratio Remnants of the 1946 slenderness requirement carried over as far as the 8 th edition (1980) AISC: The slenderness ratio, Kl/r, of compression members shall not exceed 200. The slenderness ratio, l/r, of tension members, other than rods, preferably should not exceed: For main members.......240 For lateral bracing members and other secondary members 300

Tension Slenderness Ratio Current (14 th edition, 2010) AISC states in Section D1: There is no slenderness limit for members in tension. User Note: For members designed on the basis of tension, the slenderness ratio L/r preferably should not exceed 300. This suggestion does not apply to rods or hangers in tension.

www.newmill.com/flex When safe and reliable is not enough Increased reliability Increased probability of time for safe evacuation

Experimental Investigation of Open Web Steel Joists Designed for Tension- Controlled Strength Limit State Joseph Robert Yost, Ph.D., PE Associate Professor, Structural Engineering Department of Civil and Environmental Engineering Villanova University 1

Presentation Overview 1. Introduction and Methodology 2. Experimental Matrix 3. Load and Support Details 4. Test Results 5. Conclusions 2

Research Program Experimental investigation of simply supported uniformly loaded open web steel joists subjected to gravity loading. Top chord in combined compression and bending. Bottom chord and end webs in axial tension. Interior webs alternating tension and compression. 3

Member Limit States and Experimental Objective Member strength limit states Top chord compression buckling Bottom chord and end webs tensile yield Interior webs alternating tension and compression Experimental Objective Load Compression buckling Tension yielding Displacement Design and test series of OWSJ for tension controlled failure limit state. 4

Methodology Design individual members so that tension yield of BC or EW occurs before compression buckling of TC or webs. Call tension-controlled design methodology. Over size compression members relative to strength demand. Define member Demand Capacity Ratio (DCR) as: Tension-Controlled Design Methodology DCR = Required Strength Pr ovided Strength All compression members DCR < 1.0 (reserve strength) Critical tension member DCR = 1.0 (at failure) Other tension members DCR 1.0 (close to failure) Increase slenderness limit on tension members to 300 5

Tension-Controlled Design Term and Member Selection typ.) typ.) typ.) Introduce relative strength term, r: r n = DCR n DCR max-tension =1.0 Relative Strength Ratios Used for Member Selection of Experimental Joists Bottom C. and/or End Webs r = 1.0 (failure) 2P 4.5' 8' 8' 8' 4.5' Interior Tension Webs Top Chord P P P P Compression Webs r 0.95 (5% reserve strength) r 0.90 (10% reserve strength) r 0.80 (20% reserve strength) 2P 6

Presentation Overview 1. Introduction and Methodology 2. Experimental Matrix 3. Load and Support Details 4. Test Results 5. Conclusions 7

(typ.) (typ.) (typ.) Sample Count 2P 33 ft. 2P K-Series x 6 identical samples LH-Series x 6 identical samples K-Series Rod Web x 6 identical samples 8

Experimental Matrix Series N Base SJI Designation Experimental ID Experimental Matrix Maximum Relative Strength Ratio (ρ) Tension Top Chord Webs Bottom Chord & End Webs Compression Webs K 6 20K7 J1-1,2,3,4,5,6 LH 6 28LH11 J2-1,2,3,4,5,6 RW 6 16K9 J3-1,2,3,4,5,6 1.00 0.90 0.95 0.80 All 18 samples Designed for tension control strength limit state Simply supported and subjected to uniform load Monotonically tested to failure Top chord laterally braced at 2 ft. intervals 9

Uniform Load Condition 4.5' 8' 8' 8' 4.5' P P P P P/2 (typ.) P/4 (typ.) P/8 (typ.) 4.5' 8' 8' 8' 4.5' P P P P P/2 (typ.) P/4 (typ.) P/8 (typ.) 2P Cylinder #1 Cylinder #2 1 ft (typ.) Cylinder #3 Cylinder #4 2P 2P 2P 10

.).).) Load Distribution Unit Detail 2P Load Distribution Unit 1 ft (typ.) 2P Cylinder #1 Cylinder #2 Cylinder #3 Cylinder #4 Hydraulic Cylinder Distribution Beam Distribution Unit 11

Presentation Overview 1. Introduction and Methodology 2. Experimental Matrix 3. Load and Support Details 4. Test Results 5. Conclusions 12

700 600 Yield in BC or End Web 500 400 LRFD Design Capacity = 418 lb/ft Load (lb/ft) 300 200 100 DL = 43 lb/ft K-Series Results Unloading to adjust test apparatus. J1-1 J1-2 J1-3 J1-4 J1-5 J1-6 0 0 1 2 3 4 5 6 7 8 9 10 11 12 Midspan Displacement (in) 13

2250 2000 1750 Yield in BC LH-Series Results Strain Hardening 1500 Load (lb/ft) 1250 1000 750 500 250 0 LRFD Design Capacity= 1303 lb/ft Unloading to adjust test apparatus DL = 77 lb/ft 0 1 2 3 4 5 6 7 8 9 10 11 12 Midspan Displacement (in) J2-1 J2-2 J2-3 J2-4 J2-5 J2-6 14

800 700 600 Yield of BC and End Web Rod-Web Series Results Apparent strain hardening Load (lb/ft) 500 400 300 200 100 Unloaded to adjust test apparatus LRFD Design Capacity = 420 lb/ft DL = 45 lb/ft J3-1 J3-2 J3-3 J3-4 J3-5 J3-6 0 0 1 2 3 4 5 6 7 8 9 10 11 12 Midspan Displacement (in) 15

Strength Ratios 1.7 Y/D P/D U/D 1.63 Average Strength Ratio (-) 1.6 1.5 1.4 1.3 1.2 1.52 1.49 1.44 1.39 1.29 1.28 1.26 1.37 D = design strength Y = yield strength P = plastic strength U = ult. strength 1.1 1.0 K (J1) LH (J2) Rod Web (J3) Joist Series 16

Deflection Ratios (U/Y) 4.5 4.0 K-Series LH-Series Rod-Web-Series 3.79 Displacement Ratio U/ Y (-) 3.5 3.0 2.5 2.0 1.5 2.83 3.15 1.0 1 2 3 4 5 6 Average Sample 17

Presentation Overview 1. Introduction and Methodology 2. Experimental Matrix 3. Load and Support Details 4. Test Results 5. Conclusions 18

Conclusions The tension-controlled yield limit state was successfully achieved with all 18 test samples. Relative strength factors of 0.80 for compression web, and 0.90 for top chord was sufficient to prevent primary limit state compression failure. Reserve strength relative to design capacity. Y-to-D strength ratios = 1.30, P-to-D strength ratio = 1.40, and U-to-D strength ratio = 1.50. Significant ductility with average deflection ratios of U-to-Y = 2.8, 3.8 and 3.2 for K-, LH-, and RW-Series. 19