Aust PADS Finite Element Method Based Pavement Response

Aust. PADS Finite Element Method Based Pavement Response to Load Model 1

Outline • • • Introduction Finite Element Method Material characterisation APADS - Austpads & Hosted service Worked examples Making sense of the results 2

INTRODUCTION 3

Background • Current designs use CIRCLY to calculated critical strains • CIRCLY is a – layered linear-elastic modelling of materials – cross-anisotropy – GUI actively developed 4

Background • Austroads PTF want greater flexibility – future design tasks – non-linear modelling of materials • Finite Element Method framework – provides headroom to grow – start a journey • Austroads developed FEM tool – – linear-elastic materials cross-anisotropy nonlinear-elastic materials simple interface 5

Schedule • Transitioning from CIRCLY to FEM – The journey started • Official implementation – Not before some years • Staged implementation 1. Linear elastic 2. Nonlinear elastic 6

FINITE ELEMENT METHOD OVERVIEW 7

Pavement model: what for? Objective: calculate the critical responses to be used for performance prediction (performance relationships) Pavement model = multi-layered structure + axle load Critical strains locations Current pavement model • Multilayered • Infinite in plane • Subgrade semi-infinite • Wheel-load = circular 8

Finite Element Method: Quick Overview • Finite element method (FEM) in pavement engineering – Available finite element packages (ABAQUS, …) are very general – Program developed by academics (Universities, Research organisations…) 2 D-axi. FEM pavement model 3 D FEM pavement model 9

Linear vs nonlinear analysis Nonlinear elastic material Modulus Linear analysis Linear elastic material E(σ) E 1 1 E(σ) 1 1 1 E(σ) Stress State Stiffness matrix is CONSTANT E(σ) 1 Stress State σ Stiffness matrix varies with the stress state (i. e. load) Iterative process 10

LABORATORY MATERIALS CHARACTERISATION 11

Presumptive model parameters Austroads project TT 1452 developed presumptive model parameters: Report AP-T 199 -12 (Austroads, 2012) – Base materials (High and normal quality crushed rock) Material High quality base Normal quality base – Subbase materials – Typical subgrades 250 220 1. 0 -0. 25 Material Upper granular subbase Lower granular subbase CBR (%) Material Silt (ML) Highly plastic clay (CH) Silty/sandy-clay (CL/SC) Sand (SW, SP) 2 … 5 3 … 10 10 … 15 175 0. 9 150 0. 8 10 … 35 15 … 70 70 … 85 0. 0 … 0. 10 0. 0 … 0. 15 -0. 25 -0. 50 … -0. 35 … -0. 35 12

Overview of the GUI

Overview of the GUI on initi def d a Lo c& e i f f Tra manc for ships r e P tion a l re t m Pave e tur c u r st ics t s i r cte hara c r s Laye ckness ameter n i r io • Th terial pa in locat a a • M tical str i • Cr

WORKED EXAMPLE 15

Unbound granular pavement: inputs Sprayed sealed surfaced unbound granular pavement Subgrade design CBR = 5% Material Sprayed seal surface Unbound granular Subgrade Thicknes s (mm) Sub-layers thickness (mm) - 475 Semiinfinite Design modulus (Mpa) Ev EV/EH Poisson’s ratio V = H (-) na - - - 95 500 95 314 95 198 2 0. 35 95 125 95 79 na 50 2 0. 45

Unbound granular pavement: inputs Linear elastic Thicknesses Moduli Poisson’s ratio

Unbound granular pavement: outputs � � The calculation is running in the backgr

Unbound granular pavement: outputs Critical strain (CIRCLY output +/- 0 Thicknesses Moduli problem (being fixed) Austroads method (AGPT Part 2 – Appendix K. 1) Critical strains from CIRCLY output: • Subgrade 906 μm/m midway between the loaded wheels

MAKING SENSE OF THE OUTPUTS LINEAR-ELASTIC 20

Unbound pavement 21

Asphalt surfaced unbound 22

Asphalt surfaced unbound 23

MAKING SENSE OF THE OUTPUTS NONLINEAR-ELASTIC 24

Full depth asphalt 25

Analysis types • Linear–elastic – Results very similar to CIRCLY • Nonlinear-elastic – Results different to CIRCLY – Need updated/calibrated performance relationships 26

Further information Seek me out today. 26 th ARRB Conference paper (Bodin et al). www. arrb. com. au/ARRB-Conferences Austroads Report AP-T 199 -12 www. arrb. com. au Thank you