Finite Element Simulation of Permanent Deformation on Flexible Pavements
Zhong Wu, Ph.D., P.E. XingWei Chen, Ph.D.
2009 Louisiana Transportation Conference Feb. 8-11, Baton Rouge
Outlines
Background Objective Proposed Material Model Finite Element Simulation Summary and Conclusion Acknowledgement
Background
Rutting is a common distress found in flexible pavements. Prediction of permanent deformation developed in base and beneath layers is equally important as the prediction of rutting in the asphalt layers. Conventional plasticity models (e.g. Elastic-plastic model, Drucker-Prager model, and Mohr-Coulomb model etc.) are suited for the prediction of permanent strain of materials under a monotonic load application.
Background (cont.)
Advanced modeling concepts, such as multisurface plasticity, bounding surface plasticity, Disturbed State Concept, etc. were introduced to model repeated loading conditions in addition to monotonic ones. Their mathematic terms tend to become much more complicate than traditional ones. In terms of a practical application, the parameters used in those models are often considered difficult to obtain and usually require performing multiple complicated laboratory tests.
Objective
To develop a permanent deformation model in the use of numerical simulation of a pavement structure under repeated loading for pavement materials, specifically for different base and treated subgrade materials.
Proposed Material Model Loading: Haversine load pulse of 0.1-second loading and 0.9-second, 10,000 Cycles; Base Materials: 6-in Diameter, 12-in Height 5 psi Confining pressure 15 psi Vertical Stress Subbase&Subgrade Materials: 2.8-in Diameter, 5.6-in Height 2 psi Confining pressure 6 psi Vertical Stress Setup of permanent deformation test
Mohammad, L. N., et al.(2006). “Laboratory Evaluation of Untreated and Treated Pavement Base Materials from a Repeated Load Permanent Deformation Test.” Journal of Transportation R esearch R ecord .
Proposed Material Model (cont.) At each repetition, a small amount of permanent strain will be resulted in accompany of elastic, recoverable strains.
Example of Stress-strain Curve in the Permanent Deformation Test
Proposed Material Model (cont.)
Schematic Stress-strain Curve under Repeated Loading Schematic of Proposed Model
Proposed Material Model (cont.)
Flow Chart of Proposed Model
Determining Input Parameters
First Cycles of Stress-strain Curve
Subsequent Cycle of Stress-strain Curve
Determining Input Parameters (cont.)
Loading Modulus EL
Moduli Ratio Function dN
Finite Element Simulation of Permanent Deformation Test
Base Materials: BCS/Slag, BSC/Fly ash, Limestone, Foam Asphalt (50% RAP), and Foam Asphalt (100% RAP). Subbase&Subgrade Materials: Lime Treated Soil Cement Treated Soil Untreated Subgrade
Permanent Deformation Test
Finite Element Simulation of Permanent Deformation Test (cont.)
Measured Stress-strain Behavior
Predicted Stress-strain Behavior
Finite Element Simulation of Permanent Deformation Test (cont.)
Permanent Strain of Base Materials Permanent Strain of Base Materials (I) (II)
Finite Element Simulation of Permanent Deformation Test (cont.)
Permanent Strain of Treated Soils
Permanent Strain of Subgrade Soil
Finite Element Simulation of Accelerated Pavement Test (APT) Sections
Mesh of APT Test
Finite Element Simulation of Accelerated Pavement Test (APT) Sections (cont.)
Multi Depth Deflectometer (MDD) measured APT rut depth
Predicted APT test rut depth
Summary and Conclusion
The proposed model was successfully implemented into a commercial finite element program, ABAQUS, through a user-defined UMAT subroutine. All of the model parameters can be derived from a repeated permanent deformation test. Two finite element examples were presented: one was used to simulate a permanent deformation test and the other was for rutting prediction of a APT test section.
Summary and Conclusion (cont.)
Satisfactory results were obtained from both simulations. It is concluded that the proposed model seems to have a great potential to be used in numerical simulation of permanent deformation for various pavement materials. Future research is underway to further investigate the proposed model in a three-dimensional finite element model, and verify and/or calibrate material parameters based on field tests results.
Acknowledgement
This study was supported by the Louisiana Transportation Research Center and the Louisiana Department of Transportation and Development.
Questions ?