Capacity Assessment of RCFT Members and Frames forPerformance-Based Design (PBD) - Phase I

Principal Investigator:
Jerome F. Hajjar

Sponsors:
National Science Foundation
American Institute of Steel Construction
Northeastern University
University of Illinois at Urbana-Champaign
University of Minnesota

Graduate Students:
Mark Denavit
Jie Zhang
Cenk Tort
Steven M. Gartner
Brett C. Gourley
Jorge Grauvilardell
Narina Jung
Alexander O. Molodan
Paul H. Schiller
Cenk Tort
Jose Zamudio

Undergraduate Students:
Mahmoud Alloush
Tarik Ata Rafi
Rachel Back
Brian Beck
Wilfred Chan
Mark Chauvin
Kathryna Clarke
Steve Earl
Ryan Hopeman
Ezra Jampole
Saif Jassam
Michael Kehoe
Sarah Keenan
Angela Kingsley
Tyler Krahn
Susan K. La Fore
Gregory S. Lauer
Elisa Livingston
Brett Mattas
Steven Palkovic
Matthew Parkolap
Jill Pinsky
Alston Potts
Abdulrahman Ragab
Alexandra Reiff
Katherine A. Stillwell
Zhuanqiang Tan

Loading Schemes for RCFT Beam-Column Tests

Top

• A database was created that brings together data from worldwide experimental and computational research performed on Rectangular Concrete-Filled Steel Tube (RCFT) members

The database includes experimental results from worldwide literature covering the following RCFT members: Columns Beam-Columns Panel Zones Pinned Connections Moment-Resisting Connections Frames For each member, the database includes: Description of experimental setup Material properties Geometric properties Experimental results The database includes members subjected to both monotonic and cyclic loading conditions During testing, the specimens experienced several local damage states, including: Yielding of steel tube (dy , Ey) Local buckling of steel tube (dlb, Elb) Crushing of concrete core (dc, Ec) Overall flexural buckling The occurrence of damage levels were recorded in the database using one of the following methods Data recorded by researchers Data extracted from load-deflection curves through analysis

Histogram of the Number of Specimens in the RCFT Database

Top

• Damage indices were created for monotonic and cyclic loading to assist in formulating performance-based design provisions used to correlate damage levels with performance levels in RCFT members

Deformation-Based Damage Index D = dcurr/do where: dcurr= deflection of structural member at the point in the loading history at which damage is being assessed do = deflection attained when the peak load is reached

Graphical Depiction of dcurr and do

Energy-Based Damage Index E = Ecurr/Etotal where: Ecurr = energy absorption at the point in the loading history at which the damage is being assessed Etotal = total energy absorption at the end of the test

Graphical Depiction of Ecurr and Etotal

Damage Indices for Cyclic Loading

• Energy-Based Damage Index

E = E_cb/E_m

where:

E_cb = area under the cyclic back-bone curve until the point in the load history at which damage is assessed

E_m = area under the complete cyclic back-bone curve

• Experimental back-bone curve derived by comparison with experimental results and used to determine E_cb
• Analytical back-bone curve derived through regression analysis and used to determine E_m

Representation of Ecb [after Varma et al. (2002)]

Representation of Analytical M vs. R curve for RCFT Beam-Columns

Results of Development of Damage Indices

• Once the values for specific damage levels were obtained and recorded in the database, equations were then derived to estimate the damage index values at occurrence of local damage
• The damage indices were evaluated for specimens in the database for each local damage level, providing:
  • Expected time of occurrence of local damage
  • Expected sequence of occurrence of local damage
  • Amount of reserve strength of specimens after local damage occurs

The following figures depict variations of deformation-based (dlbw/do) and energy-based (Ety/Etotal) damage indices with geometric properties (D/t) of the test specimens as well as the ratio of axial load (P) at occurrence of local damage to nominal cross-section strength (Po) of the RCFT members

Comparison of D/t and P/Po vs. dlbw/do for Monotonically-Loaded Beam-Column Tests

Comparison of D/t and P/Po vs. Ety/Etotal for Monotonically-Loaded Beam-Column Tests

dlbw= deflection at local buckling of the steel tube web Ety= energy absorption until steel yielding in tension

Top

• A parametric study was conducted to examine the effect of local damage states on behavior of RCFT members

Multiple specimens were generated based on the ranges of geometric (e.g. D/t, L/D) and material (e.g. fy , fc') properties within the scope of the database For each individual specimen, the damage indices were evaluated at each local damage state using the equations derived for the damage indices The occurrence of local damage was compared and contrasted Damage levels were associated with performance levels to facilitate PBD

After (Mehanny and Deierlein, 2000)

Comparison of Deformation-Based Damage Indices
for Monotonically-Loaded Column Tests

Proposed Design Methodology