You are on page 1of 43

Generating Mixture Inputs in

Pavement ME for North Dakota’s


Conditions
Presenter: Duncan Rioba Oteki, PhD. Student
Advisor: Dr. Daba Gedafa, Ph.D., P.E., ENV. SP, F. ASCE
Presentation Layout

• Background

• Literature Review

• Materials and Methods

• Results and Discussion

• Conclusion

• Future Work
Background

Rutting Fatigue Cracking

Thermal Cracking
MEPDG or Pavement ME
Pavement ME Levels

Le • Mixture-specific laboratory tested


ve properties
l1
• Predictive models and laboratory
Le
vel measured binder data
2L
e • Predictive models and default binder
v properties
e
l
3
Problem Statement
The development of durable and long-lasting pavements is
the most effective way to ensure good riding comfort, safety,
and low-cost maintenance.

Accurate and effective characterization of hot mix asphalt


(HMA) properties is vital to this goal

This project seeks to provide a HMA material input data to


support the implementation of Pavement ME in North
Dakota.
Objectives
To measure the dynamic modulus (|E*|) for typical North
Dakota HMA mixes.

To determine the flow number (FN) for typical North Dakota
HMA mixes

To evaluate existing models' capability to predict |E*| for


North Dakota asphalt mixes.
Literature Review
Author/Date Methodology Findings
Timothy et al. 2003 - 5 HMA mixes; -The Original Witczak
(Minnesota) - |E*| testing; model underpredicted |
- Original Witczak model. E*| for all the five
mixtures
Birgisson et al. 2004 - 28 HMA mixes; - |E*| predictions at high
(Florida) - |E*| testing and prediction; temperatures closer to
using the original Witczak measured values than
model for 28 HMA mixtures lower temperature
predictions

Mohammad et al. -13 HMA mixes; - Both models


2007 -|E*| testing; performed well.
(Louisiana) - Original Witczak & Hirsch - Hirsch model was
models. more accurate.
Literature Review Cont'd
Author/Date Methodology Findings
Gedafa et al. 2009 - Laboratory HMA mixes and - Original Witczak
(Kansas) field cores from 5 sites; model
- |E*| testing; underestimates |E*|
- Original Witczak, Modified at low temperatures
Witczak, and Hirsch Models. and overestimates |
E*| at high
temperatures
Martinez and Angelone 2009 -Lab HMA mixes and Field - The Heukelomp
(Argentina) cores; and Klomp model
- Heukelomp and Klomp, was more accurate.
Original Witczak, Hirsch, and - Recalibration to
Alkhateeb Models. improve accuracy

Yousefdoost et al. 2013 -Lab mixes |E*| testing - Original Witczak


(Australia) - Original Witczak, modified was more accurate.
Witczak and Hirsch Models. - It underpredicted |
E*| values
- Not robust enough
Materials and Methods
5 HMA mixes

Measurements
|E|* Predictions
FN |E*|

|E*| models evaluation

Levels 1, 2, and 3 database inputs


for North Dakota
Material Selection
Asphalt Mix NMAS Binder RAP (%)
Grade
HWY 32-Finley 12.5 PG 58H-34 15
HWY 32-Finley 12.5 PG 58S-28 15
HWY 6 - 12.5 PG 58S-34 25
Bismarck
HWY 83 - Minot 9.5 PG58H-34 10
HWY 28 - Minot 12.5 PG58S-28 -
|E*|
|E*| Relevance in Pavement
ME
Climatic Model

Traffic Model

Mechanistic
E = |E*|
Analysis

Fatigue Cracking
Rutting Prediction
Prediction
Specimen Preparation
Asphalt Mix Performance
Tester (AMPT)
• NCHRP- Project 9-29
• Versatile
• Cost
• Time
|E*| Testing

Temperature Frequency
(⁰C) (Hz)
35 (95⁰F) 10, 1, 0.1, 0.01
20 (68⁰F) 10, 1, 0.1
4 (39⁰F) 10, 1, 0.1
Master Curve
Master Curve Interpretation
|E*| Predictions (Level 2)
Model Description Original Modified Hirsch
Witczak Witczak
Binder Properties η √
f √
|G*| √ √
δ √
Mix Volumetric Vbeff √ √
Properties
Va √ √
VMA √
VFA √
ρ¾ √ √
ρ⅜ √ √
ρ4 √ √
ρ200 √ √
|E*| Predictions (Level 3)
Model Description Original Witczak Modified Hirsch
Witczak

Binder Properties PG √ √ √
Mix Volumetric Vbeff √ √
Properties
Va √ √
VMA √
VFA √
ρ¾ √ √
ρ⅜ √ √
ρ4 √ √
ρ200 √ √
Flow Number
Flow Number Interpretation

Traffic level, million ESALs HMA Minimum Average Flow


Number

<3 -

3 to <10 50

10 to <30 190

>=30 740
Results and Discussion
|E*| at 4°C
|E*| at 20°C
E*| at 35°C
Phase angle at 4°C
Phase angle at 20°C
Phase angle at 35°C
Master Curve |E*|
Master Curve Phase Angle
Level 2 Predictions
Original Witczak Model
Modified Witczak Model
Hirsch Model
Level 3 Predictions
Original Witczak Model
Modified Witczak Model
Hirsch Model
Flow Number
Conclusions
• HWY 32_PG58S-28_15% RAP had the highest |E*|.

• Binder performance grade, Aggregate grading, and RAP


significantly affected |E*| and Phase angle values.

• The Hirsch model performed well for Level 2 predictions.

• The Original Witczak model performed excellently for Level 3


predictions.

• HWY 32_PG58H-34_15% RAP had the highest flow number.


Future work
• Incremental Repeated Load Permanent Deformation

(iRLPD) test.

• Simplified Viscoelastic Continuum Damage (S-VECD) test.

• Creep compliance predictions

• 5 HMA mixes from the 2022 construction season.


Thank you.

Questions?

You might also like