Three Dimensional Finite Element Modeling of Flexible Pavements

1
Three Dimensional Finite Element Modeling of
Flexible Pavements

• Michael Willis
• Dr. Beena Sukumaran
• The 6th Annual NJDOT Research Showcase

2
Outline

• Introduction
• Objectives
• Verification
• Models
• Results
• Future Work

Picture Courtesy of NAPTF
3
Introduction

• Purpose of the NAPTF
• Design of pavement sections capable of
  withstanding the loads from newly designed
  heavier aircraft
• Comprised of
• 900 by 60 foot (Test Pavement)
• 600-Ton test Vehicle (75,000 pounds/wheel)
• 1,000 Sensors
• Capable of modeling the design life of a pavement

4
Testing Facility
Picture Courtesy of NAPTF
5
Testing Facility

• Pavement Structures
• Rigid/Flexible Pavement
• Conventional/Stabilized Bases
• Subgrade
• Low Strength Sub-Grade (CBR 3-4)
• Medium Strength Sub-Grade (CBR 7-9)
• High Strength Sub-Grade (CBR 30-40)
• Modeling of the MFC section
• Medium Strength-Flexible Pavement-Conventional
  Base
• Found to have the most severe deformations

6
6-Wheel Configuration
Picture Courtesy of NAPTF
7
Objectives

• Simulation of aircraft loading gear
• Material Model Verification
• Finite Element Model Verification
• Calibration of Material Properties
• Testing and Analysis
• Verify current FAA design procedures
• Use if elastic strain to predict permanent
  deformation

8
Material Model Verification

• Determine if the Drucker-Prager model is capable
  of accurately predicting soil response
• Methodology
• Use plasticity models in quasi three-dimensional
  finite element analysis
• To predict CBR values for the medium strength
  sub-grade (DuPont Clay)
• For cohesive soils the D.P. model refines to von
  Mises
• Compare predictions with known experimental
  results

9
Material Model Verification

• Verification Model
• 185 Elements
• 6,260 Nodes
• CAXA Elements
• Quasi 3-d Fourier
• Reduced Integration

10
Verification Studies

• Soil Properties Medium Sub-grade
• Moisture Content 30.5
• Undrained Shear Strength 13.3 psi
• Dry Density 90.5 pcf
• Elastic Modulus 12,000 psi
• Analysis 1
• von Mises model with ultimate shear strength as
  the yield strength

11
Verification Studies
(a)

• Analysis 2
• von Mises model with unconfined compression
  stress-strain data
• As shown in adjacent figure, zone of plastic
  strain increases with penetration depth
• Analysis 3
• Utilized instantaneous elastic modulus calculated
  from the unconfined compression stress-strain data

(b)
Plastic strain distribution at a) 0.1 piston
penetration (b) 0.2 piston penetration
12
Verification Results
13
Verification Results

• Load-Displacement response remarkably similar to
  field data
• Prediction of CBR consequently improves
• Three-dimensional finite element model accurately
  captures stress-strain response of subgrade

Analysis 3
Stress vs. Displacement plot for the various
verification studies compared with field test
data

• Therefore the Drucker-Prager model may be used in
  further studies

14
Basic Model Design

• Design Considerations
• Element Size
• Element Types (C3D20R / C3D8R)
• Boundary Conditions
• Material Model
• Elastic
• Plastic (Drucker-Prager / Mohr Coulomb)
• Loading Types
• Static vs. Dynamic Loading
• Computational Time

15
Initial Boundary Condition

• Initial Conditions
• Geostatic stress
• Boundary Conditions
• Act as Confining Soil
• Bottom
• Sides
• Perfectly Bonded Layers

(P-401)
(P-209)
(P -154)
16
Material Model

• Non-Cohesive Soils
• Drucker-Prager Model
• Elasto-Plastic Model
• Simplicity
• Frictional Properties (Shear Failures)
• Cohesive Soils
• Drucker-Prager Model
• Refined down to von Mises criterion
• Asphalt is modeled as a very cohesive and stiff
  clay

Asphalt Surface
Crushed Aggregate
Uncrushed Aggregate
Dupont Clay
17
Model Verification

• Compare Boussinesqs vs. Model Prediction
• To validate model geometry and boundary conditions

20
Node
6
6
Tire Imprint and Element Size
18
Model Verification Results
19
Static Punch Test

• Purpose
• Model Verification
• Compare our models ability to predict failure
• FAAs Static Punch Test
• 6-Wheels of 55 kips each
• Static Punch Test (ABAQUS)
• 6 Pressure Loads of 218 psi each

Picture Courtesy of NAPTF
20
Static Punch Test

• Geometry
• 426010
• Basic Trench
• 2 Deep
• 3.5 Wide
• Load Applied
• 21 away
• Wheel Spacing
• 54 Dual
• 57 Tandem

Picture
21
Static Punch Test
22
Static Punch Test
Picture Courtesy of NAPTF
23
Static Punch Results
Load (kips)
Displacement (inches)
24
Moving Wheel Model

• Characteristics
• Velocity 8.8 ft/s
• Wheel Imprint 21x12
• Rest Time 8.5 seconds
• Static Loading
• 218 psi for 0.21 seconds/Element

25
Rut Depth Profile
26
Moving Model Results

• Predicted deformation was considerable more then
  measured.
• Found that shear strength of the base and subbase
  materials were too small
• Material properties were modeled in CBR model
  used earlier (20 to 30 lower then measured
  values)
• Therefore another verification step is necessary
  in calibrating the material properties and model
• Instead of using Static-Punch try using MDD or
  FWD Data

27
Future Work

• Material Models
• Mohr-Coulomb vs. Drucker-Prager
• Visco-Elastic (Asphalt Layer)
• Calibration of Material Properties
• Triaxial Data
• Bearing Capacity Analysis
• Plate Load Testing
• Calibration of Entire Model
• MDD Data
• Strain Measurements
• FWD Data

28
QUESTIONS