comparing the volumetric and mechanical properties of

Comparing the Volumetric and Mechanical Properties of Laboratory and Field Specimens of Asphalt Concrete

TRB Webinar April 20, 2017

2:00 PM to 3:30 PM ET

Louay N. MohammadSam Cooper, III

Department of Civil and Environmental EngineeringLA Transportation Research Center

Louisiana State University

Webinar Outline Objective / Scope Conduct of Experiment Data Analysis

– Impacts of process-based factors on differences specimen types:

» Design (LL), Production (PL) and Construction (PF)– Compare mechanistic properties of 3 specimen types

» LL, PL, PF» LWT, E* (Axial, Indirect)» Develop shift factors

– Compare volumetric properties of 3 specimen types» LL, PL, PF» Develop tolerance recommendation

Summary

Research Team

Acknowledgements NCHRP

– Dr. Ed Harrigan– Project Technical Review Panel

LADOTD and LTRC Research Staff Participating Contractors/DOTs

– Stark Asphalt Florida DOT– Mathy Const. Co. Iowa DOT– Prairie Const. Co. Michigan DOT– Barriere Const. Co. South Dakota DOT– Diamond B Const. Virginia DOT– Community Asphalt Wisconsin DOT

Texas Transportation Institute University of Wisconsin

Objective Determine the cause and magnitude of differences and variances

in measured volumetric and mechanical properties within and between three specimen types:

– Laboratory-mixed and laboratory compacted (LL)» Design

– Plant mixed and laboratory compacted (PL)» Production

– Plant mixed and field compacted (PF)» Construction

Research Methodology• Task 1: Conduct Literature Review• Task 2: Survey, collect, and analyze data from past

research– Levels of Variability in Asphalt Mixtures– Factors Causing Variability between Sample Types

• Accomplished in Three stages – Stage 1: Conventional Literature Review– Stage 2: Survey, collection and analysis of completed research

projects.– Stage 3: Collection and analysis of additional completed

projects suggested by the NCHRP Panel

6

Outline

Introduction

Task 1: Literature

Task 2: Survey Collect and Analyze Data from Past Research

Task3: Develop Experimental Factorial

Task 4: Execute Experimental Factorial

Task 5: Conduct Data Analysis

Task 6: Develop Specification Recommendations

Literature Review : Stage 1

• Compile relevant literature to determine:• Properties Commonly Tested in Quality Assurance Activities

• Asphalt Content, Air voids, VMA, Gmm, Aggregate Gradation, Field Density

• Allowable Levels of Variability• Test Method• Agency Tolerance

7

Outline

Introduction

Task 1: Literature

Task 2: Preliminary Research

Task3: Develop

Experimental Factorial

Task 4: Experimental

Factorial Execution

Task 5: Data Analysis

(Example)

Project Status

Property No. of States Mean Min Max

Asphalt Content, % 45 0.4 0.2 0.7Air Voids, % 28 1.1 0.2 2.0Field Density, % 21 2.3 1.0 5.0

Bulk Specific Gravity (Gmb) 8 0.030 0.020 0.045

Theoretical Maximum Specific Gravity (Gmm)

10 0.020 0.015 0.026

VMA, % 12 1.2 0.3 2.0

Literature Review : Stage 1

• Compile relevant literature to determine:• 60 research publications were evaluated to determine the state

of the practice• Properties Commonly Tested in Quality Assurance Activities

• Asphalt Content, Air voids, VMA, Gmm, Aggregate Gradation, Field Density

• Allowable Levels of Variability• Test Method• Agency Tolerance

8

Outline

Introduction

Task 1: Literature


Task3: Develop



Factorial Execution


(Example)

Project Status

Property No. of States Mean Min Max

Asphalt Content, % 45 0.4 0.2 0.7Air Voids, % 28 1.1 0.2 2.0Field Density, % 21 2.3 1.0 5.0

Bulk Specific Gravity (Gmb) 8 0.030 0.020 0.045

Theoretical Maximum Specific Gravity (Gmm)

10 0.020 0.015 0.026

VMA, % 12 1.2 0.3 2.0

• Conduct a worldwide survey • 50 state DOTs, District of Columbia, Federal Highway Administration

(FHWA), Canada, Puerto Rico, England, and the Netherlands.• More favorable responses for volumetric data than the mechanistic

data.• In general, the data sets included PL and PF samples• Data analyzed from

• The Netherlands• 13 States• FHWA Mobile Laboratory

• Samples may be tested by different users• Contractor• State• Third Party

9

Literature Review : Stage 2Outline

Introduction

Task 1: Literature


Task3: Develop



Factorial Execution


(Example)

Project Status

• Additional search was conducted:• Lacking data to determine cause and magnitude of differences among

all three specimen types.

• Specific Sources of Data:• Universities

• UNR

• State Dots• LADOTD• AZDOT

• National Database• LTPP• NCHRP Reports

• Targeted identification of process–based factors causing variability

10

Literature Review: Stage 3

Outline

Introduction

Task 1: Literature






• Stage 2:• Could not identify projects with properties of multiple sample

types• Therefore, need to continue the search for projects with

multiple sample types• Stage 3:

• Data sets did not all include three sample types for a specific mixture.

• Process-based factors were not controlled.

11

Literature Review and Preliminary Research -- Limitations

Outline

Introduction

Task 1: Literature






• Candidate Factors available as Sources of Variability• Process-Based Factors

ID Factors Details

Process Ba

sed Factors

1 Compaction methods Difference between field and laboratory compaction methods

2 Silo storage Extended storage time at the plant may harden asphalt in the mix

3 Baghouse fines May affect mix gradation and other volumetric properties

4 Reheating May affect binder properties and thus compacted specimens

5 Aggregate absorption May differ between plant and lab and thus affect variability

6 Plant type and settings May affect mixture properties and thus variability

7 Sampling location Sampling location (plant, behind paver, …) may affect variability

Additio

nal Factors 8 Gradation density Sensitive mixes are more susceptible to mix proportions than non‐sensitive

mixtures

9 MTD Use of MTD and lack of MTD use may affect mix properties and variability

10 Aggregate degradation Mixture production may increase the fines fraction for soft aggregates

11 Aggregate moisture Moisture in the stockpile may affect mix properties

12

Literature Review: Stage 3 -- Methodology

• Process-Based Factors • Present during the production and testing of

a mixture.• Source of the delivery of reclaimed fine

material : baghouse fines, reclaimed asphalt pavement

• Time delay in specimen fabrication: reheating, silo storage, sampling location, plant type, use of material transfer device

• Aggregate properties: Absorption, Degradation, Angularity, Surface Texture

• Aggregate stockpile in situ properties: moisture content

13

Outline

Introduction

Task 1: Literature






Research Methodology

Develop Experimental Plan• Selected Process-based Factors:

• Source of the delivery of reclaimed fine material: baghouse fines

• Time delay in specimen fabrication (PL): reheating• Aggregate properties: Absorption, Hardness• Aggregate stockpile in situ properties: moisture

• Time Delay (Reheating) ≠ Aging• Mix often stored for some time before reheating for

evaluation

14

Outline

Introduction

Task 1: Literature






• Factorial• Each factor varied on two contrasting levels (high vs. low) • Factorial based on 2K design

• K = number of factors (5)• 25 = 32 mixtures• X 3 specimen types = 96 test combinations per property

• Volumetric properties and Mechanistic properties

Factor ID Factor Low Level (‐) High Level (+)1 Use of Baghouse Fines No Yes2 Time delay No Yes3 Aggregate Absorption Low High4 Aggregate Degradation Soft Hard5 Aggregate Stockpile Moisture Low High

15

Develop Experimental Plan

Outline

Introduction

Task 1: Literature






• Volumetric properties – AV, VMA, Gmm, AC (Solvent), gradation, Gsb

• (LL, PL, PF)

• Mechanistic properties– Hamburg LWT

• (LL, PL, PF) – IDT Dynamic modulus

• (LL, PL, PF) – Axial Dynamic Modulus

• (LL, PL)

16


Outline

Introduction

Task 1: Literature






• Full Factorial design– 25 factor combinations x 3 specimen types

• 96 test combinations

• Volumetric Specimens– 96 x 3 x 3 = 864

• Mechanistic Specimens– 96 x 4 x 3 = 1152

• Total number of specimens– 2016

• Fractional Factorial Design– Experimental design chosen to exploit the sparsity of effects

principle– Exposes information about features of interest

17


Outline

Introduction

Task 1: Literature






• Quarter fractional design– 25-2 factor combinations x 3 specimen types

• 24 test combinations

• Volumetric properties– 24 x 3 x 3 = 216

• Mechanistic properties– 24 x 4 x 3 = 288

• Total number of samples– 504

• Evaluate the main effects of the factors selected

18


Outline

Introduction

Task 1: Literature






• Quarter Fractional Factorial

• Responses of each property for each Mixture ID to be evaluated in main effects model

• Only main effects will be evaluated• Interactions may not be quantified

Mixture IDBaghouse Fines

Time DelayAggregate Absorption

Aggregate Degradation

Aggregate Moisture Content

Mix 1 No (‐) No (‐) Low (‐) Soft (‐) High (+)Mix 2 No (‐) No (‐) High (+) Hard (+) Low (‐)Mix 3 No (‐) Yes (+) Low (‐) Hard (+) Low (‐)Mix 4 No (‐) Yes (+) High (+) Soft (‐) High (+)Mix 5 Yes (+) No (‐) Low (‐) Hard (+) High (+)Mix 6 Yes (+) No (‐) High (+) Soft (‐) Low (‐)Mix 7 Yes (+) Yes (+) Low (‐) Soft (‐) Low (‐)Mix 8 Yes (+) Yes (+) High (+) Hard (+) High (+)

19


Outline

Introduction

Task 1: Literature






20

Experimental Plan Execution –Search for Mixtures

• State asphalt pavement association,• DOTs• Contractors• Contacts• Etc.

Outline

Introduction

Task 1: Literature






21

Experimental Plan Execution –Search for Mixtures

• State asphalt pavement association,• DOTs• Contractors• Contacts• Etc.

Outline

Introduction

Task 1: Literature






Willing to assist No able to assist No Response

• LL Specimen Fabrication– Approximately 120 kg of loose mixture required

22

Experimental Plan Execution

Outline

Introduction

Task 1: Literature






• PL Specimen Fabrication– Approximately 120 kg of loose mixture required

23


Outline

Introduction

Task 1: Literature






• PF Specimen Collection

• Roadway cores collected prior to trafficking• Each core is trimmed to required specimen size

for testing 24


Outline

Introduction

Task 1: Literature






• Perform laboratory evaluation

25


Outline

Introduction

Task 1: Literature






=?

=?

Laboratory Design Plant Production Field Construction

• Volumetric properties:Volumetric Property Test MethodGmm AASHTO T 209

Gmb/Air Voids AASHTO T 166AASHTO T 269

Asphalt Content (Extraction) AASHTO T 164Aggregate Gradation AASHTO T 30Aggregate Bulk Specific Gravity/Absorption

AASHTO T 84AASHTO T 85

26


Outline

Introduction

Task 1: Literature






• Mechanistic properties:

Mechanistic Property Test Method

Loaded Wheel Tester AASHTO T 324

Axial Dynamic Modulus AASHTO T 342

Indirect Dynamic Modulus NC State

27


Outline

Introduction

Task 1: Literature






Wheel Diameter: 203.5 mm (8 inch)Wheel Width: 47mm (1.85 inch)

Fixed Load: 703 N (158 lbs)Rolling Speed: 1.1 km/hr

Passing Rate: 52 passes/min

• AASHTO T 324-04• Damage by rolling a steel wheel across

the surface of a sample• Cylindrical, Slab

• 50oC, Wet or dry• Deformation at 20,000 passes is recorded

Mechanistic Test –Loaded Wheel Tracking Test (LWT)

28

Outline

Introduction

Task 1: Literature






• IPC UTM-25, AMPT• AASHTO T-342• Sinusoidal axial

compressive stress is applied to a specimen • temperature and frequency

• Dynamic modulus0

0|*|

E

Frequency (HZ) 25, 10, 5, 1, 0.5, 0.1Temp. (°C) -10, 4.4, 25, 38, 54.4

• Phase Angle 360p

i

TT

Mechanistic Test–Axial Dynamic Modulus |E*|

29

Outline

Introduction

Task 1: Literature






• Material Testing System 810• North Carolina State University• Sinusoidal Bi-axial compressive

stress is applied to a specimen • temperature and frequency

Frequency (HZ) 10, 5, 1, 0.5,0.1Temp. (°C) -10, 10, 30

• Dynamic modulus

0202

122102|*|UVad

PE

• Phase Angle 360p

i

TT

Mechanistic Test –Indirect Tension Dynamic Modulus (IDT |E*|)

30

Outline

Introduction

Task 1: Literature






31

Individual Mixture Statistics

32

Data Analysis Methodology• For Each Parameter (i.e. Air Voids)

• Each Sample Type (i.e. LL, PL, PF)• Descriptive Statistics

• Mean• Standard Deviation• Coefficient of Variation

Quality Control check with test method

Outline

Introduction

Task 1: Literature






33

Data Analysis Methodology

• For Each Parameter (i.e. Air Voids)• Delta Analysis

• Algebraic difference between the means of two specimen types

∆Air Voids,PL-LL =MeanAir Voids,PL-MeanAir Voids,LL

Outline

Introduction

Task 1: Literature






• Each mix is to be evaluated individually to determine magnitude of variabilityand difference among sample types.

• Ttest/Ftest – Two groups

• Analysis of Variance – More than two groups

34

Data Analysis Methodology –Statistical Analysis

Tstat·

http://www.itl.nist.gov/div898/handbook/eda/section3/eda3672.htm

• Several Options for evaluation of individual pairings (i.e. PF vs PL, PF vs LL, and PL vs LL)• Least Squared Difference • Duncan Multiple Comparison Test• Student Newman-Keuls Test• Tukey Multiple Comparison Test

35

Least Conservative

Most Conservative

Data Analysis Methodology –Statistical Analysis

≠

=

≠

Outline

Introduction

Task 1: Literature






Individual Mixture Analysis – Mix 1

36

Mixture

Mixture Conditions

Baghouse

Fines UsedReheating

Aggregate

Absorption, %

Aggregate

Degradation

(LA Abrasion)

Aggregate

Moisture

Content, %

Mix1WI No (–) Yes/No (–/+) 1.7 (–) 38 (+) 4.8 (+)

Outline

Introduction

Task 1: Literature






Provided Mixture Could Not Assist No Response

00.10.20.30.40.5

LL - PL LL - PF PL - PF

NN

N

|∆| A

C, %

Mixture Comparison

Specimen Type AVG ST.DevLL 5.40 0.03PL 5.57 0.07PF 5.64 0.23

AC– Delta Summary, %


-0.17 -0.24 -0.07

Effect of Specimen Type: Asphalt Content Mix 1 -- WI

37

Outline

Introduction

Task 1: Literature






Effect of Specimen Type: Air Voids at Ndesign Mix 1 -- WI

38

Outline

Introduction

Task 1: Literature






0.00.20.40.60.81.0

LL-PL LL-PLR PL-PLR

Y

NN

|∆| A

V, %

Mixture Comparison

AV – Delta Summary, %

LL - PL LL - PLR PL - PLR

0.8 0.3 -0.5

Specimen Type AVG ST. DevLL 4.2 0.6PL 3.4 0.3

PLR 3.9 0.2

0

0.02

0.04

0.06


Y NN

|∆| G

mm

, %

Mixture Comparison


Gmm– Delta Summary, %


0.016 0.013 -0.003

Effect of Specimen Type: Gmm Mix 1 -- WI

39

Outline

Introduction

Task 1: Literature






00.010.020.030.040.05


NN

N

|∆| G

sb, %

Mixture Comparison


Gsb– Delta Summary, %


-0.009 -0.001 0.008

Effect of Specimen Type: Gsb Mix 1 -- WI

40

Outline

Introduction

Task 1: Literature






0

5

10

15

20

0 5000 10000 15000 20000Aver

age

Rut

Dep

th,

mm

Number of Passes

LL PL PF

0

5

10

15

20

0 5000 10000 15000 20000Aver

age

Rut

Dep

th,

mm

Number of Passes

LL PL PF

0

5

10

15

20

0 5000 10000 15000 20000Aver

age

Rut

Dep

th,

mm

Number of Passes

LL PL PF

Effect of Specimen Type: LWT Test

41

Mix 5 SD Mix 8 LA

1.5

15

150

1500

15000

150000

1.0E-05 1.0E-01 1.0E+03 1.0E+07

Com

plex

Mod

ulus

, (M

pa)

Reduced Frequency

PLR LL

1.5

15

150

1500

15000

150000

1.0E-05 1.0E-01 1.0E+03 1.0E+07

Com

plex

Mod

ulus

, Mpa

Reduced Frequency

PLR LL

Effect of Specimen Type:Axial Dynamic Modulus

42

Outline

Introduction

Task 1: Literature






0

5000

10000

15000

20000

25000

1.0E-03 1.0E+00 1.0E+03 1.0E+06

Com

plex

Mod

ulus

, Mpa

Reduced Frequency

LL PL PF

0

5000

10000

15000

20000

25000

1.0E-04 1.0E+00 1.0E+04 1.0E+08

Com

plex

Mod

ulus

, Mpa

Reduced Frequency

LL PL PF

Mix 3MN Mix 5VA

Effect of Specimen Type:IDT Dynamic Modulus

43

Outline

Introduction

Task 1: Literature






44

Combined Mixture Analysis

• Meta-Analysis:• A statistical technique used to combine various data

sets into one meta-data set

• Objective: to determine the magnitude and cause of variability between specimen types

• Combine delta analyses of individual mixtures.

45

Outline

Introduction

Task 1: Literature






Data Analysis Methodology –Combined Analysis

Summary of Differences Among Specimen TypesDesign (LL), Production (PL), Construction (PF), • Volumetric Parameters

46

02468

1012

- = + - = + - = +

Freq

uenc

y

Difference

Asphalt Content

LL - PF LL - PL PL - PF

02468

1012

- = + - = + - = +

Freq

uenc

y

Difference

Gmm


02468

1012

- = + - = + - = +

Freq

uenc

y

Difference

Air VoidsLL - PF LL - PL PL - PF LL - PF LL - PL PL - PF

02468

1012

- = + - = + - = +

Freq

uenc

y

Difference

Gsb

• Mechanistic Parameters (IDT |E*|)

01020304050

- = + - = + - = +

Freq

uenc

y

Difference

Low Temperature (-10°C)

Summary of Differences Among Specimen TypesDesign (LL), Production (PL), Construction (PF),

47

01020304050

- = + - = + - = +

Freq

uenc

y

Difference

High Temperature (25-35°C)

LL - PF LL - PL

01020304050

- = + - = + - = +

Freq

uenc

y

Difference

Intermediate Temperature (10°C)

PL - PF LL - PF LL - PL PL - PF


Summary of Differences Among Specimen TypesDesign (LL), Production (PL), Construction (PF),

• Volumetric Parameters

• Mechanistic Parameters

Statistical Difference / Practical DifferenceComparison VTM VMA VFA Gmm AC Gsb

LL vs PF ---- 50%/20% 10%/0% 20%/20%LL vs PL 60%/20% 30%/10% 80%/50% 70%/20% 20%/20% 40%/10%PL vs PF ---- 30%/20% 10%/10% 10%/10%

Comparison LWT Axial E* IDT E*LL vs PF 38% ---- 57%LL vs PL 18% 55% 29%PL vs PF 52% ---- 61%

48

Outline

Introduction

Task 1: Literature






51

Effect of Process-Based Factors

• Analysis of Co-Variance• Blend of ANOVA and Regression• GLM regression model to evaluate factor with the

most significant effect. – Classification Variables: Baghouse (BH), Time Delay (TD)– Independent Variables: Aggregate Absorption (AA), Aggregate

Hardness (AH), Stockpile moisture (SM).

52

Outline

Introduction

Task 1: Literature






Data Analysis Methodology –Combined Analysis

Specimen MixID Replicate Δ AC Baghouse Reheat ABS Hardness Moisture

LL-PF

1WI 1 -0.25 No No 1.7 38 4.87IA 2 0.00 Yes Yes 1.6 50 3.38LA 3 0.00 Yes Yes 2.1 14 5.03MN 4 -0.30 No Yes 0.8 17 3.5

5LA61 5 0.06 Yes No 1.2 22 5.05LA90 6 0.29 Yes No 0.7 22 5.45WI 7 0.04 Yes No 1.3 18 5.45VA 8 0.20 Yes No 0.5 15 4.55SD 9 0.30 Yes Yes 0.5 15 4.06FL 10 -0.20 Yes No 2.8 37 2.8

LL-PL

1WI 1 -0.11 No No 1.7 38 4.87IA 2 0.30 Yes Yes 1.6 50 3.38LA 3 0.10 Yes Yes 2.1 14 5.03MN 4 -0.70 No Yes 0.8 17 3.5

5LA61 5 0.11 Yes No 1.2 22 5.05LA90 6 0.07 Yes No 0.7 22 5.45WI 7 -0.03 Yes No 1.3 18 5.45VA 8 0.20 Yes No 0.5 15 4.55SD 9 0.40 Yes Yes 0.5 15 4.06FL 10 -0.10 Yes No 2.8 37 2.8

PL-PF

1WI 1 -0.14 No No 1.7 38 4.87IA 2 -0.30 Yes Yes 1.6 50 3.38LA 3 -0.10 Yes Yes 2.1 14 5.03MN 4 0.40 No Yes 0.8 17 3.5

5LA61 5 -0.05 Yes No 1.2 22 5.05LA90 6 0.22 Yes No 0.7 22 5.45WI 7 0.07 Yes No 1.3 18 5.45VA 8 0.00 Yes No 0.5 15 4.55SD 9 0.10 Yes Yes 0.5 15 4.06FL 10 -0.10 Yes No 2.8 37 2.8

53

Effect of Process-Based Factors

Outline

Introduction

Task 1: Literature






• General Linear Model (GLM) Procedure, Asphalt Content– Compare all factors’ main effects

simultaneously

– Repeat Process for all variables evaluated

Source DF MS Treatment MS Error F Value Pr > F

Baghouse 1 0.2003 0.0127 15.77 0.01

Time Delay 1 0.0008 0.0127 0.07 0.81

Agg. Abs. 1 0.0947 0.0127 7.46 0.05

Agg. Hardness 1 0.0052 0.0127 0.42 0.55

Stockpile Moisture 1 0.0357 0.0127 2.81 0.16

54

Effect of Process-Based Factors : Analysis of Co-variance -

Outline

Introduction

Task 1: Literature






• General Linear Model (GLM) Procedure, Asphalt Content– Compare all factors’ main effects

simultaneously Source DF MS Treatment MS Error F Value Pr > F

Baghouse 1 0.2003 0.0127 15.77 0.01

Time Delay 1 0.0008 0.0127 0.07 0.81

Agg. Abs. 1 0.0947 0.0127 7.46 0.05

Agg. Hardness 1 0.0052 0.0127 0.42 0.55

Stockpile Moisture 1 0.0357 0.0127 2.81 0.16

55

Effect of Process-Based Factors : Analysis of Co-variance -

Outline

Introduction

Task 1: Literature






P-value < 0.05 = Significant

http://www.itl.nist.gov/div898/handbook/eda/section3/eda3672.htm

Effect of Process-Based FactorsDesign (LL), Production (PL), Construction (PF),

56

Property Comparison Significant ProcessAV, %

Design (LL) - Production (PL)Stockpile Moisture

VMA, % NoneVFA, % None

ACDesign (LL) - Production (PL) Baghouse fine return and aggregate absorptionDesign (LL) - Construction (PF) Baghouse fine returnProduction (PL) - Construction (PF) None

GmmDesign (LL) - Production (PL)

NoneDesign (LL) - Construction (PF)Production (PL) - Construction (PF)

GsbDesign (LL) - Production (PL)

NoneDesign (LL) - Construction (PF)Production (PL) - Construction (PF)

GradationDesign (LL) - Production (PL) Baghouse fine return and aggregate hardnessDesign (LL) - Construction (PF) Baghouse fine return, aggregate hardness, and stockpile moistureProduction (PL) - Construction (PF) None

• Mechanistic:– No effect of process-based factors for all specimen

comparisons

Effect of Process-Based FactorsDesign (LL), Production (PL), Construction (PF),

57

Outline

Introduction

Task 1: Literature






Why are factors not effecting mixture properties? –Survey

• Industry Contacts• Agency Contacts• Contractors

58

Outline

Introduction

Task 1: Literature






Response No Response

Why are factors not effecting mixture properties? – Contactor Survey

• Do you observe VMA collapse in the HMA production at your plant prior to fine-tuning?– Cause: Aggregate Breakdown and increased fines.

Yes61%

No39%

59

Outline

Introduction

Task 1: Literature






Yes22%

No78%

• Do you observe VMA collapse in the HMA production at your plant after fine-tuning?

60

Outline

Introduction

Task 1: Literature







Yes22%

No78%

• Do you observe VMA collapse in the HMA production at your plant after fine-tuning?

61

Outline

Introduction

Task 1: Literature







Contractors in this study understood the materials being used and tuned the plant accordingly

62

Magnitude of Volumetric Differences

• Compare volumetric properties of three specimen types– LL, PL, and PF

• Determine magnitude of differences among specimen types

• Develop tolerance recommendations• Determined by averaging differences for all

mixtures and applying confidence interval based on combined standard deviation

Volumetric Properties ComparisonTolerance Recommendation

63

Outline

Introduction

Task 1: Literature






Comparison Property Avg Min Max Confidence Limit95%

Design (LL) / Production

(PL)

AV,% 0.6 0.0 1.3 0.8

VMA,% 0.4 0.0 2.1 1.2VFA,% 4.0 0.3 9.9 5.4AC,% 0.2 0.0 0.4 0.2Gmm 0.014 0.002 0.039 0.020Gsb 0.011 0.002 0.025 0.014

Passing 0.075 mm, % 0.4 0.0 0.9 0.5

Volumetric Properties ComparisonTolerance Recommendation – Design vs Production

64

Outline

Introduction

Task 1: Literature








(PL)

AV,% 0.6 0.0 1.3 0.8

VMA,% 0.4 0.0 2.1 1.2VFA,% 4.0 0.3 9.9 5.4AC,% 0.2 0.0 0.4 0.2Gmm 0.014 0.002 0.039 0.020Gsb 0.011 0.002 0.025 0.014

Passing 0.075 mm, % 0.4 0.0 0.9 0.5

Volumetric Properties ComparisonTolerance Recommendation – Design vs Production

65

Outline

Introduction

Task 1: Literature







Design (LL) / Construction (PF)

AC,% 0.2 0.0 0.3 0.2Gmm 0.011 0.000 0.020 0.013Gsb 0.010 0.001 0.033 0.019

Passing 0.075 mm, % 0.7 0.1 1.3 0.7

Volumetric Properties ComparisonTolerance Recommendation – Design vs Construction

66

Outline

Introduction

Task 1: Literature







Production (PL) / Construction (PF)

AC,% 0.1 0.0 0.4 0.2Gmm 0.009 0.001 0.027 0.018Gsb 0.008 0.000 0.031 0.017

Passing 0.075 mm, % 0.5 0.1 0.8 0.5

Volumetric Properties ComparisonTolerance Recommendation – Production vs Construction

67

Outline

Introduction

Task 1: Literature






• Design (LL) vs Production (PL): ± 0.2%• Design (LL) vs Construction (PF): ± 0.2%• Production (PL) vs Construction (PF): ± 0.2%

Above Tolerance At Tolerance Below Tolerance Not Specified

Volumetric Properties ComparisonTolerance Recommendation – Asphalt Content

68

Outline

Introduction

Task 1: Literature






69

Magnitude of Mechanistic Differences

• Compare mechanistic properties of three specimen types– LL, PL, and PF

• Develop Shift Factors– Loaded Wheel Tracking Test– Axial Dynamic Modulus– IDT Dynamic Modulus

• Developed by comparing average values to each specimen type.– For example:

• LL avg / PL avg (for each parameter)• Repeat for other specimen comparisons

70

Outline

Introduction

Task 1: Literature






Mechanistic Properties ComparisonShift Factors

• Recommended Shift–Design (LL) to Production (PL): 1.0–Design (LL) to Construction (PF) : 0.75–Production (PL) to Construction (PF): 0.75

– Specification Requirement: 12 mm – Lab Design ≤ 9 mm

Mechanistic Properties ComparisonShift Factors – LWT

71

Outline

Introduction

Task 1: Literature






• Recommended Shift– Shift may need to be utilized at higher temperatures

» May relate to binder oxidation in plant produced mixtures

Comparison Temperature, °C Average ShiftShift Range

Minimum Maximum


(PL)

-10.0 1.0 0.7 1.1

4.4 1.0 0.7 1.1

25.0 0.9 0.6 1.1

37.8 0.8 0.5 1.1

54.4 0.8 0.5 1.2

Mechanistic Properties ComparisonShift Factors – Axial Dynamic Modulus

72

Outline

Introduction

Task 1: Literature






• Recommended Shift– Need for shift increase with test temperature.

» IDT Dynamic modulus is highly sensitive to binder properties and aggregate orientation at elevated temperatures

Temperature, °C ComparisonAverage

CorrectionCorrection Range

Minimum Maximum

-10Design (LL)/Production (PL) 1.0 0.8 1.1

Design (LL)/Construction (PF) 1.0 0.9 1.3Production (PL)/Construction (PF) 1.1 0.9 1.4

10Design (LL)/Production (PL) 0.9 0.8 1.1




Mechanistic Properties ComparisonShift Factors – IDT Dynamic Modulus

73

Outline

Introduction

Task 1: Literature






• Apply shift factor• PF to LL

1000

10000

100000

1000000

10000000

1E-08 0.000001 0.0001 0.01 1 100 10000

Dyn

amic

Mod

ulus

, PSI

Reduced Frequency, Hz

LL PF

Mechanistic Properties ComparisonUse of Shift Factor – IDT Modulus

74

Outline

Introduction

Task 1: Literature







1000

10000

100000

1000000

10000000

1E-08 0.000001 0.0001 0.01 1 100 10000

Dyn

amic

Mod

ulus

, PSI


LL PF


75

Outline

Introduction

Task 1: Literature






Difference can be reduced with shift factor

1000

10000

100000

1000000

10000000

1E-08 0.000001 0.0001 0.01 1 100 10000

Dyn

amic

Mod

ulus

, PSI


LL PF Shifted PF


76

Outline

Introduction

Task 1: Literature







0

5

10

15

20

PF Shifted PF LL

Alligator Cracking (%)

0

1

2

3

PF Shifted PF LL

Total Rutting (in)

0

0.5

1

1.5

2

PF Shifted PF LL

AC Rutting (in)

100

120

140

160

180

200

PF Shifted PF LL

IRI (in/mi)

Shift Factors ImpactPavement Prediction

77

Outline

Introduction

Task 1: Literature






Summary• Cause and magnitude of ∆ in volumetric & mechanical properties

within and b/w three specimen types– LL, PL, PF

• Ten field projects / mixtures– Varying process-based factors

• Comparison of volumetric properties of three specimen types– Specimen types were statistically different

78

Outline

Introduction

Task 1: Literature









• Comparison of volumetric properties of three specimen types– Specimen types were statistically different

79

Outline

Introduction

Task 1: Literature






What was causing this difference?




• Impacts of process-based factors– Volumetric and Mechanistic

• None: PL vs PF– Volumetric

• LL vs PL, and LL vs PF– Contractor Survey

• Adjustment to account for the process-based factors

80

Outline

Introduction

Task 1: Literature









• Impacts of process-based factors– Volumetric and Mechanistic

• None: PL vs PF– Volumetric

• LL vs PL, and LL vs PF– Contractor Survey

• Adjustment to account for the process-based factors

81

Outline

Introduction

Task 1: Literature






Baghouse, Aggregate Absorption, Aggregate Hardness, and stockpile moisture




• Comparison of mechanistic properties of three specimen types– LWT (LL vs PF and PL to PF)– Axial E* (LL vs PL ) -- >25C– IDT E* (LL vs PF) -- > >10C

82

Outline

Introduction

Task 1: Literature









• Comparison of mechanistic properties of three specimen types– LWT (LL vs PF and PL to PF)– Axial E* (LL vs PL ) -- >25C– IDT E* (LL vs PF) -- > >10C

83

Outline

Introduction

Task 1: Literature






Not influences by process-based factors

84

• Comparison of mechanistic properties of three specimen types– Recommended shift factors were developed

– Effects on pavement performance prediction were evaluated– Use of LL or PL moduli in performance prediction would result in

reduced design of the pavement structure– without local calibration

• Comparison of volumetric properties of three specimen types– Tolerance Recommendation were developed– Comparison to existing State practice

Summary –Combined Analysis

Outline

Introduction

Task 1: Literature






85

Implementation Recommendations

86

• Design Recommendations - Volumetrics• Introduce baghouse fines into laboratory produced

mixtures

Implementation RecommendationsOutline

Introduction

Task 1: Literature






Baghouse fines returned during production?

Yes

Determine quantity returned

0.5 -3%

“Pepper” Laboratory Mixture with appropriate baghouse material

87

• Performance Based Specification• LWT – Consider shift factors when evaluating

expected rut depth of field compacted specimens


Introduction

Task 1: Literature






Comparison Conversion

Design (LL) / Production (PL) 1.0

Design (LL) / Construction (PF) 0.75

Production (PL) / Construction (PF) 0.75

88

• Design Recommendations• Axial Dynamic Modulus – Consider shift factors

when evaluating expected modulus values• May be important for performance prediction and

pavement design


Introduction

Task 1: Literature






Comparison Temperature, °C Conversion

Design (LL)/ Production (PL)

-10.0 1.04.4 1.025.0 0.937.8 0.854.4 0.8

90

• Design Recommendations• Indirect Tension Dynamic Modulus – Consider shift

factors when evaluating expected modulus values


Introduction

Task 1: Literature






Temperature, °C ComparisonAverage

CorrectionCorrection Range

Minimum Maximum

-10Design (LL)/Production (PL) 1.0 0.8 1.1






91

• Design Recommendations• Indirect vs Axial Dynamic Modulus• May be used to predict field core modulus from

laboratory tested material


Introduction

Task 1: Literature






IDT vs. Axial Conversion

Low Temperature Comparison, -10°C 0.81

Intermediate Temperature Comparison, 10°C 0.75

High Temperature Comparison, 25 - 35°C 0.90

comparing the volumetric and mechanical properties of

Documents