Home Skip to main content

Research Projects (2000-01)

Identifying Number


Project Title

Moment-Rotation Tests of High Performance Steel (HPS) I-Girders


Colorado State University

Principal Investigator(s)

Dr. Bryan A. Hartnagel
Department of Civil Engineering
Colorado State University
(970)491-4660 or Fax (970)491-7727

External Project Contact


Project Objective

The proposed program will provide additional information on the behavior of high performance steel by conducting experimental tests on four girders fabricated from HPS over a two-year period. Objectives of the program are:

  • To investigate the flexural strength and ductility of bridge I-girders fabricated from HPS.
  • To compare the flexural strength and ductility of bridge I-girders made from HPS to that predicted by current AASHTO bridge design provisions.
  • To develop a finite element model of I-girders fabricated from HPS
  • To compare the experimental and analytical test results with analytical results presented in the literature.

Project Abstract

Bridge designers now have a new choice of steel available for consideration when planning a bridge. High-performance steel (HPS) grade 70W is currently available for bridge construction. The HPS70W was developed under a cooperative research program between the Federal Highway Administration (FHWA), the U.S. Navy, and the American Iron and Steel Institute (AISI). However, current bridge design provisions limit the flexural strength of girders with yield strength greater than 50 ksi (350 Mpa) to the yield moment capacity. The flexural capacity of similar bridge girders designed with yield strength less than or equal to 50 ks i (350 Mpa) is equal to the plastic moment capacity if certain restrictions are met. Further, if the designer chooses, an inelastic analysis of the girder is allowed with steel yield strengths less than or equal to 50 ksi (350MPa). Inelastic analysis and design methods offer larger cost savings compared to the elastic analysis provisions. Even with the disadvantage on the flexural capacity, HPS is still competitive with Grade 50 steels because of material savings obtainable with HPS. If these restrictions could be lifted or even relieved, the use of HPS would provide significant cost savings.

The AASHTO Load and Resistance Factor Design (LRFD) Specification considers Any girder cross section with a specified minimum yield strength exceeding 50 ksi to be non-compact. This limits the load carrying capacity of the cross section to a maximum of the yield moment. The capacity of girders with a specified minimum yield strength less than or equal to 50 ksi can be as large as the plastic moment, provided certain cross section properties are met. Research has shown that the capacity of girders fabricated from steel with a specified minimum yield strength greater than 50 ksi can sustain loads larger than the yield moment. However, more research is necessary before changes in the design specification can be implemented.

Before restrictions on the use of HPS can be removed, adequate knowledge of the material behavior must be known. This proposal is intended to provide additional information on the behavior of HPS. This proposal will support the experimental testing of four I-girders fabricated from HPS. Funding is also included to further the analytical modeling of HPS by extending existing studies to include parameters that were held constant in previous research by others.

To accomplish the objectives, listed below in item 5 of the project, a series of tests will be conducted on four I-girder specimens. The I-girders will be fabricated from HPS70W steel. Two different compression flange slenderness ratios, (bf/2tf), and two different depth of web in compression to depth of web, (Dcp/D), ratios will be used to make up the four specimens. Compression flange slenderness ratios that are near the ultra-compact limit and near the compact limit specified in the current bridge design specifications.

Values for Dcp/D will be 0.5 and 0.75. Each of the four specimens will be tested in three point bending, as were all the other specimens in the literature. The simple beam bending of the test specimens represents the portion of a continuous span girder between the inflection points of moment near the interior support, albeit upside down. This length is approximately 20 percent of the span length on each side of the interior support. The condition of moment at the interior support is similar for girders with any number of continuous spans.

The analytical portion of the project will develop a finite element model of HPS I-girders. Results from this model will initially be compared to results from previous research (22) for validation. An identical specimen will be used for this comparison. After a reliable model is developed, it will be used to predict the behavior of the experimental I-girders. Experimental results will provide verification of all finite element models from this research and from previous research conducted by others.

An additional part of the analytical work will be to extend previous finite element investigations to include the effects of varying parameters held constant in previous research. The following parameter was held constant in the previous study:

The depth of web in compression to the total depth of web (Dcp/D) is assumed to be 0.5. In this study, depth of web in compression to the total depth of web (Dcp/D), will be set to 0.5 and approximately 0 .75. For girders symmetrical about the major bending axis (non-composite), (Dcp/D) is equal to 0.5. However, for girders unsymmetrical about the major bending axis (composite) the (Dcp/D) ratio is closer to 0.75. This is significant because most of the bridge girders designed to day act composite with the concrete deck. This will raise the plastic neutral axis and therefore increase the (Dcp/D) ratio.

Previous research found the effects of this parameter to be significant for Grade 50 steel. Material parameters affecting the stress-strain behavior of the HPS will be varied. The separameters include the yield ratio, (Fy/Fu), the ratio of the strain at strain hardening to the yield strain, (est/ey), and the strain hardening modulus (Est). The range of each of these parameters will be the same characterization of HPS stress-strain behavior used in previous studies. Experimental tension test coupons are planned for material taken from the I-girder specimens. They will be used to determine the tensile stress-strain behavior of the HPS. This information will then compared to the parameters used in the analytical study.

Geometric parameters of the cross section will also be varied in the finite element models. The parameters that most affect the moment-rotation behavior of I-girders include the compression flange slenderness, (bfc/2tfc), the web slender ness, (Dcp/2tw), and the cross section aspect ratio, (D/bfc). Two compression flange slenderness values will be used that correspond to ultra-compact flange (bfc/2tfc = 0.288(E/Fy)½) and compact flange(bfc/2tfc = 0.382(E/Fy)½). Significant inelastic bending strains can develop in the compression flange before local flange buckling occurs if the flange is ultra-compact.

The compact flange limit is that specified by AASHTO. This slenderness limit is relaxed compared to the ultra-compact limit but still allows the girder to reach the plastic moment capacity, Mp for girders fabricated from material with a specified minimum yield strength less than or equal to 50 ksi.

In summary, the analytical model will extend the variation of parameters studied in previous research conducted by others. It will also be used to predict the behavior of the experimental I-girders.

Task Descriptions

The tasks and approximate completion dates necessary to achieve the objectives of the research are:

  • Task 1 – Determine size and geometry of second two test girder specimens. (August 31, 2000, B. Hartnagel)
  • Task 2 – Order test girder specimens (2). (October 1, 2000, B. Hartnagel)
  • Task 3 – Develop and refine finite element model. (ongoing through June 30, 2001, B. Hartnagel and Graduate Assistant)
  • Task 4 – Test both I-girder specimens for moment - rotation behavior and strength. (March 31, 2001, B. Hartnagel and Graduate Assistant)
  • Task 5 – Compare test results, finite element model results and previous research results. (April 30, 2001, B. Hartnagel and Graduate Assistant)
  • Task 6 – Prepare final report. (June 30, 2001, B. Hartnagel)
  • Task 7 – Prepare and submit technical paper on experimental tests and analytical models. (June 30, 2001, B. Hartnagel and Graduate Assistant)

Milestones, Dates

  • Starting Date: July 1, 2000
  • Project Milestones: See task descriptions (Item 7)
  • Ending Date: June 30, 2001

Yearly and Total Budget

The attached budget is for the second year of the project. For the second year the requested amount is $21,228 USDOT and is augmented by $45,344 as CSU match. The total budget is $52,566 (year 1) plus $66,572 (year 2) for a total project budget of $119,144.

Student Involvement

One graduate student will complete a Master's thesis or Ph.D. dissertation on the project and several undergraduate students will work on the project. The graduate student will supervise and conduct the experimental tests. They will also be responsible for developing the analytical finite element models. Hourly undergraduate students will help in conducting the experimental tests by instrumenting the test girders and preparing the data acquisition system. This will provide them with an opportunity to learn about experimental research.

Relationship to Other Research Projects

No other related MPC projects to date.

Technology Transfer Attributes

The research will provide information on the behavior of high-performance steel. This information could be used by AASHTO to make changes in the current bridge design specifications regarding the design of HPS. Analytical models will be developed to investigate parameters that were not varied in previous research. Identical analytical models will be compared to previous research results to validate the models. Experimental tests will verify the analytical models presented in this and other research.

Many experimental tests have been conducted on I-girders fabricated from steel with nominal yield strength of 50 ksi (350 MPa) or less, however, this project will be one of only a few experimental tests conducted on I-girders fabricated with 70 ksi (485 MPa) nominal yield strength steel. Therefore, results from this project will be significant in developing design provisions for HPS70W steel.

Potential Benefits of the Project

Current bridge design provisions allow the use of high-performance steel but conservatively restrict the load carrying capacity of HPS. This research could provide information necessary to ease some of the restrictions placed on the use of HPS. These changes could allow more economical bridges to be designed using HPS. Designers are reluctant to use inelastic design provisions for the design of new structures regardless of the specified minimum yield strength. The real benefit of inelastic design provisions would be their use in load rating bridges constructed from any grade of steel. However, before inelastic analysis and de sign can be used to load rate bridges, the inelastic provisions must be available for the design of bridges. A rating method based on inelastic analysis would be beneficial for rural counties without funds necessary to replace all structurally deficient bridges. If load rating base d on inelastic analysis was allowed, some bridges might not be structurally deficient because of the reserve strength inherent in continuous span bridges.

Moment-rotation tests are performed on high performance steel girders to determine the strength and ductility of HPS I-girders. This information is necessary to develop design provisions for HPS. Current design provisions for HPS are in some cases overly conservative because of the lack of research on the material. It is true that most off-system bridges would not immediately benefit from the use of HPS girders for the superstructure because of the short span of many of these bridges. However, this research will be used to assist in the development of comprehensive inelastic design provisions for steel girders applicable to available steel materials. Inelastic design provisions were first permitted in 1986 when AASHTO released Guide Specifications for Alternate Load Factor Design Procedures for Steel Beam Bridges Using Braced Compact Sections. Now, inelastic design provisions for steel girders are incorporated in the first and second edition of the AASHTO Load and Resistance Factor Design Bridge Design Specifications. As engineers realize that inelastic design is an acceptable method to proportion steel girders for new construction, they will also realize that it is an acceptable method to determine the capacity of existing bridges. This is where the benefit of this research pertains to off-system bridges. Many off-system bridges are required to have load postings because rating calculations show inadequate load carrying capacity when in reality they can safely carry a larger load. Continuous slab on girder bridges have reserve strength available from several sources including, longitudinal redistribution of moment, lateral distribution of load, unintended composite action, bearing restraint and additional capacity provided by barrier curbs or rails. The most reliable strength reserve of continuous steel I-girder bridges is from longitudinal redistribution of moment. However, current rating procedures do not account for this reserve strength. Once engineers are aware that inelastic design provisions are available and provide adequate safety for new construction, they will be able to use these provisions possibly to increase load postings on off-system bridges. By safely limiting the conservatism in current rating methods, limited funds for off-system bridges may be prioritized differently.

TRB Keywords

Inelastic design, high performance steel, bridges

NDSU Dept 2880P.O. Box 6050Fargo, ND 58108-6050