Characterization and Simulation of Fracture in AASHTO M180 Guardrail Steel
RESEARCH PROBLEM STATEMENT
Recent policy decisions to allow the use of simulation to evaluate the effects of modifications to roadside hardware make the need for development of accurate metal fracture modeling techniques even more critical. In the development of roadside safety devices, many components are designed to absorb energy through large plastic deformations. Other components are designed specifically to fracture at certain loads in order to absorb additional energy or to ensure proper release from the system at critical locations. Similar deformations and fracture are often observed during impact events with respect to vehicle components, such as suspension components. In addition to intentional fracture of barrier components, the high, dynamic impact loads imparted on roadside barriers can potentially lead to catastrophic failure of barrier components. This type of failure mode is not desired and can completely compromise the function and safety performance of these barrier systems.
Steel is one of the most commonly used structural materials in roadside safety. It is used for rail components, posts, and fasteners. Many roadside safety features rely on the toughness of ductile steel to resist rupture that would lead to a potentially catastrophic failure. For example, W-beam guardrail, the most commonly used roadside safety feature, is designed such that the rail element often reaches full cross-section yield during high energy impacts. In this situation, the ductility and toughness of the W-beam rail are critical to the barrierís ability to maintain continuity and successfully redirect an impacting vehicle. W-beam and thrie beam guardrail, the most predominant barrier systems on the federal highway system, are fabricated with steel beam conforming to the AASHTO M180 steel specification. As such, an† increased understanding of the fracture behavior of AASHTO M180 steel would have the potential to provide vast improvements in the safety of barriers using this guardrail steel.
The ability to characterize and analyze steel fracture is critical to the proper design and analysis of guardrail steel. However, design and analysis of steel components for fracture and the corresponding prediction of the onset of the steel fracture is non-trivial. Difficulties include strain rate effects, definition of the proper failure criteria, element mesh size effects, determination of incorporation of initial strains and stresses in components, and modeling of crack initiation and propagation among others. Current non-linear finite element analysis techniques used in the design and evaluation of roadside hardware have not been fully developed to address all of these issues and accurately simulate and predict metal fracture. Thus, the current capability for modeling of metal fracture is limited. This limitation can lead to over-designed and inefficient structural design of roadside hardware and the inability to identify critical failure modes in roadside systems. In addition, inaccurate simulation of fracture could lead to potentially hazardous errors in the use of computer simulation to evaluate the effects of design modifications to roadside hardware.
LITERATURE SEARCH SUMMARY
The importance of metal fracture in the design of roadside hardware has been a common issue. Several past studies have investigated the modeling of W-beam fracture including the simulation of the weak post W-beam by Ray and Plaxico (1). The design of many guardrail systems have also focused on preventing rail fracture when the systems are used in special applications such as long-span guardrail over culverts (2) and guardrail over curbs (3-4). New 31-in. tall guardrail systems have been specifically designed to reduce the potential for rail fracture (5). Crash cushion and end terminal systems such as the thrie beam bullnose (6) and the BEAT (7-8) incorporate or rely on the fracture of steel components as part of their function. Finally, breakaway supports often use steel fracture to disengage critical components from their fixed bases. Thus, steel fracture is employed successfully in a wide range of roadside barriers. However, the ability of designers and modelers to accurately simulate this behavior has been limited.
LS-DYNA (9) is the current state of the art software used for computer simulation of roadside hardware, and was used to design and develop many of the metal fracture critical systems noted above. LS-DYNA has many material models with methods for incorporating metal fracture. The primary limitation of LS-DYNA is relatively poor accuracy in predicting dynamic load conditions at which metals fracture. The primary issue is that procedures for measuring and/or predicting a materialís dynamic fracture resistance have not been satisfactorily described. The constraining effects of dynamic, three-dimensional stress states greatly complicate efforts to predict fracture. This problem is inherent to all dynamic finite element analysis codes. Thus, the ability for analysts to predict the onset and location of failure is not accurate. This shortcoming can lead to catastrophic problems if the simulation does not identify critical failure modes in a system or incorrectly models the response.
Efforts to overcome finite element analysis shortcomings have led to the use of failure bracketing techniques that incorporate two different algorithms, one that tends to predict premature failure and the other that tends to overestimate the toughness of a given component. These procedures allow analysts to bracket barrier performance. When a component incorporates a fracture sensitive part, the simulations cannot precisely predict performance, but it can limit the possible range of outcomes to a narrow band which may be sufficient to evaluate the system performance. However this methodology is largely tied to the experience of the analyst and prior knowledge of the expected loading and fracture modes and thus cannot be relied upon to accurately predict fracture.
There have been many recent advances and improvements in the material models available in LS-DYNA as they relate to metal fracture. However, no concentrated effort has been made to investigate these models and apply them to roadside hardware in order to develop effective methods for simulation of metal fracture. This effort is critical if the roadside safety community intends to continue to expand the use of simulation in the design and evaluation of roadside hardware.
1.††††††††††††††††† M.H. Ray, K. Engstrand, C.A. Plaxico, and R.G. McGinnis, Improvements to the Weak-Post W- Beam Guardrail, Transportation Research Record 1743, pp. 88-96, 2001.
2.††††††††††††††††† Bielenberg, R.W., Faller, R.K., Sicking, D.L., Rohde, J.R., and Reid, J.D., Midwest Guardrail System for Long Span Culvert Applications, Paper No. 07-2539, Transportation Research Record No. 2025, Journal of the Transportation Research Board, TRB AFB20 Committee on Roadside Safety Design, Transportation Research Board, Washington D.C., January 2007.
3.††††††††††††††††† Polivka, K.A., Sicking, D.L., Rohde, J.R., Faller, R.K., Crash Testing of Michiganís Type B (W-Beam) Guardrail System, Final Report to the Michigan Department of Transportation, Transportation Research Report No. TRP-03-90-99, Midwest Roadside Safety Facility, University of Nebraska-Lincoln, Lincoln, Nebraska, November 10, 1999.
4.††††††††††††††††† Polivka, K.A., Sicking, D.L., Rohde, J.R., Faller, R.K., and Holloway, J.C., Crash Testing of Michiganís Type B (W-Beam) Guardrail System - Phase II, Final Report to the Michigan Department of Transportation, Transportation Research Report No. TRP-03-104-00, Midwest Roadside Safety Facility, University of Nebraska-Lincoln, Lincoln, Nebraska, December 13, 2000.
5.††††††††††††††††† Faller, R.K., Polivka, K.A., Kuipers, B.D., Bielenberg, B.W., Reid, J.D., Rohde, J.R., and Sicking, D.L., Midwest Guardrail System for Standard and Special Applications, Paper No. 04-4778, Transportation Research Record No. 1890, Best Paper Award - TRB AFB20 Committee on Roadside Safety Design, Transportation Research Board, Washington D.C., January 2004.
6.††††††††††††††††† Bielenberg, R.W., Reid, J.D., and Faller, R.K., NCHRP Report No. 350 Compliance Testing of a Bullnose Median Barrier System, Paper No. 01-0204, Transportation Research Record No. 1743, Transportation Research Board, Washington, D.C., January 2001.
7.††††††††††††††††† J.D. Reid, J.R. Rohde and D.L. Sicking, "Box-Beam Burster Energy Absorbing Single-Sided Crash Cushion," Transportation Research Record 1797, TRB, National Research Council, Washington, D.C., November 2002, pp. 72-81.
8.††††††††††††††††† J.R. Rohde, D.L. Sicking and J.D. Reid, "Box-Beam Burster Energy Absorbing Tube - Bridge Pier Protection System," Transportation Research Record 1851, TRB, National Research Council, Washington, D.C., November 2003, pp. 74-82.
9.††††††††††††††††† Hallquist, J. O., LS-DYNA Keyword Users Manual: Version 971, Livermore California, Livermore Software Technology Corporation, December 2011.
†††††††††††† The objective of this study would be to improve the accuracy of steel fracture models now utilized in the LS-DYNA non-linear, explicit finite element analysis code as applied to AASHTO M180 guardrail steel. The research effort should focus on the AASHTO M180 steel specification due to its use in the majority of our roadside barrier inventory including W-beam and thrie beam guardrail, approach guardrail transitions, end terminals, crash cushions, and many bridge rails. This objective would include the following tasks:
1.††††††††††††††† Literature Review: The literature review will focus on identifying existing methods for predicting fracture in steel within LS-DYNA. The goal will be to identify the best methods now available and examine the strengths and weaknesses of each approach. A small number of the best available methods will be selected for further analysis.
2.†††††††††††††††† Component Testing: Component testing of desired metal fractures scenarios common to beam guardrail should be conducted and used as a baseline for simulation comparisons. These component tests should focus on the critical and common guardrail failure modes observed in current barrier systems.
3.††††††††††††† Evaluation of the Fracture Simulation Techniques: The best available methods will be applied and utilized to predict material behavior in roadside safety applications. The strengths and shortcomings of each method will then be examined to determine what characteristics of material failure can currently be modeled successfully and the types of dynamic failure that remain outside the capability of existing models.
4.†††††††††††††††† Development of Improved Fracture Methodologies: Techniques will be investigated for improving upon areas where current fracture simulation techniques fall short in Task 3. This effort may include additional material testing of steel components to provide improved material model input and implementation of improved finite element analysis techniques such as implementation of initial stresses, advanced element formulations, advanced constitutive models, etcÖ
5.†††††††††††† Comparison to Full-Scale Crash Testing: The most promising methodologies should be further compared against existing full-scale crash testing of beam guardrail testing that demonstrated metal fracture in order to determine their effectiveness. This would include simulation of full-scale crash tests with critical metal fractures in order to determine the effectiveness of the most promising fracture modeling methodologies.
6.††††††††††††††††† Reporting: A final report will be prepared to document the findings of the study and provide recommendations on the use and limitations of the fracture simulation methods investigated.
ESTIMATE OF PROBLEM FUNDING AND RESEARCH PERIOD
Recommended Funding:† $280,000.
Research Period: 24 months.†
URGENCY, PAYOFF POTENTIAL, AND IMPLEMENTATION
Development of improved techniques for the accurate simulation of AASHTO M180† guardrail steel fracture in LS-DYNA will provide many direct benefits. First, the more accurate simulation of fracture is a critical requirement if simulation is to be effectively used to evaluate roadside hardware in order to ensure that potential failures are identified. Second, improved simulation of metal fracture could potentially lead to the design of more efficient structural design and improved energy absorption techniques as compared to current barrier systems. Finally, these improved techniques will allow for improved identification and analysis of existing roadside barrier failures and determination of the performance limits of existing barrier systems. This information would provide guidance for identifying potential vulnerabilities in existing systems and determining proper maintenance and replacement schedules for damaged systems. Thus, development of improved capabilities for simulation of metal fracture in roadside hardware has the potential to provide a significant leap in safety to the motoring public and reduce accident costs and fatalities.