TxDOT’s use of the Overlay Tester

and Hamburg Wheel Test

Dale Rand, P.E.

TxDOT Construction Division

Flexible Pavements Branch

Louisiana

Transportation

Engineering Conference

Baton Rouge, La.

Feb 13, 2007

TxDOT’s Mission Statement

The mission of the Texas Department of

Transportation is to provide safe, effective and

efficient movement of people and goods.

The vision of the Texas Department of

Transportation is to provide transportation systems

and alternatives that are comfortable, safe, durable,

cost-effective, accessible, environmentally

sensitive and aesthetically appealing.

TxDOT’s Vision Statement

Commentary

Rutting and Cracking are two of the biggest enemies of

any flexible pavement

There is a strong need to have quick/simple laboratory

tests that can predict these two distresses

Most of the “good predictor” tests for rutting and

cracking are neither simple nor quick

In many cases designers have to sacrifice crack

resistance to get rut resistance or visa versa

Ultimately, the goal is to design HMA pavements that

are highly resistant to both rutting and cracking

At the HMA Mix Design Stage

Historically, states have used either Marshall or Hveem Stability to predict rutting. These tests are simple but one can argue not strong predictors of rutting.

No DOT’s routinely perform any tests to predict cracking at the mix design stage.

The good tests that predict cracking are not used because they tend to be impractical especially if you have to evaluate hundreds of mix designs.

HMA Rutting and Stripping in Texas

Rutting and stripping were relatively common

failure modes up until the late 1990’s

Over the past 5 years, rutting and stripping

failures have diminished to the point that they

are “non issues”

TxDOT now requires the Hamburg wheel test

on all HMA mixes

Rutting and stripping: Problem solved

HMA Cracking in Texas

Overall, cracking continues to be the biggest

pavement problem that we have.

Reflective cracking is very common.

Relatively thin pavements, multiple overlays.

Addressing the cracking problem is a very high

priority

Thicker pavements?

HMA mixes that are more crack resistant?

Implementation of the

Hamburg Wheel Test

The Hamburg wheel test is a torture test that does a

very good job of identifying mixes that are susceptible

to rutting or moisture damage.

TxDOT maintains a database of over 4,000 Hamburg

tests that have been performed.

Database was used to develop a criteria for all mixes.

Maximum rut depth of 12.5 mm test at 50 C

PG 76-xx………20,000 passes

PG 70-xx………15,000 passes

PG 64-xx………10,000 passes

Hamburg Wheel Test Video

Rutting: 12.5 mm # of Passes: 10,200 Temp: 50C

District: Abilene Mix Type: Superpave Binder: 76-22 (Source 1)

CSJ: 0068-07-046 Aggr.: Limestone Additive: None

ID: 01500318 Lab Mix Notes:

Rutting: 2.8 mm # of Passes: 20,000 Temp: 50C

District: Abilene Mix Type: Superpave Binder: 76-22 (Source 2)

CSJ: 0068-07-046 Aggr.: Limestone Additive: None

ID: 01500380 Lab Mix Notes:

0

5

10

15

20

25

30

35

40

45

50

55

Avg

Ru

t d

ep

th (

mm

)

Lime 20 10 6

Liquid 31 17 12

None 40 18 10

PG 64-22 PG 70-22 PG 76-22

Effect of binder grade and additive typeIncludes all: 50 °C, mix types & aggregate types

What About Cracking

• Do we have to sacrifice cracking resistance

to get rutting resistance?

• Can we quickly and easily predict cracking

resistance?

Designing mixes resistant to thermal

reflection cracking

• Concept over 40 years old

• Early work at Texas A&M and in Europe

• Used extensively in 1980’s to study inter-layers and fabrics

• Currently being used in NCHRP model evaluation study

• TxDOT study (2001) evaluation as mix design tool

Equipment Upgrades

1. Upgrade Hardware and

software

2. Fully computer-

controlled test

Background to the overlay tester

TxDOT’s New Overlay Tester

Use of Overlay Tester in Mix Design Overview of Presentation

• Description of test – Sample fabrication (use routine equipment)

– Quick to run (hours not days)

– Repeatable results

• Related to Performance– Validation studies on in-service pavements

– Sensitivity studies

• Field Evaluation– Test sections

• Implementation as part of design system– Site investigation

– Mix design criteria

– Structural design

Sample Preparation

Mix Design Test Protocol

– Sample size: 6’’ long by 3’’ wide by 1.5’’ high

– Loading: Continuously triangular displacement 5 sec loading and 5 sec unloading

– Standard Test Conditions• Opening displacement: 0.025 in.

• Room temperature: 77 F

– Definition of failure• Discontinuity in Load vs

Displacement curve

• Visible crack on surface

Displacement

Time (s)10 20

Fixed plate

2 mm (0.08 in)

Aluminum plates

150 mm (6 in)

Sample

Movable plate

plate

Ram direction

38 mm (1.5 in)

Statewide Evaluation of Good/Bad

reflection cracking projects

Crack Stops

Latex Modified Asphalt

Binder

US 84 Abilene (6 mo old)

Inplace Recycling

US 175 Dallas

10 year old section

Field Validation Studies

1) does the test rank materials

correctly ?

2) What are acceptable criteria ?

77 Deg F 25 mils opening

0100200300400500600700800

HIPR C76-22 DAC5 + L D64-22 L CM64-22

CR

reps

- fa

ilure

300

US 175

700

IH 20

25

SH 3

2

US 84

520

US 281

IH-10 Type C (PG76-22L), 4.4%AC

Hamburg Test@50C

0

0.5

1

1.5

2

2.5

3

3.5

0 5000 10000 15000 20000 25000

No. of Passes

Ru

t D

ep

th (

mm

)

I-10

Properties Result Target

Cracking(overlay tester cycles to

failure)

2 >200

Rutting(APA rutting after 8000

cycles)

2.6 mm <6mm

Rutting

Hamburg(Hamburg cycles to 0.5

inch rut)

>20K >20K

Rut resistance mix (4 in thick) placed on IH 10 in 2002 heavy traffic

Reflection cracking in 2004

Current Completion Dates

• Rutting testing Jan. 2004

• Fatigue testing (100-mm) Mar. 2006

• Fatigue testing (150-mm) Dec. 2007

Two FHWA ALFs with

12 Pavement Lanes Constructed in

the Summer and Fall of 2002

As-Built Pavement Lanes

CR-AZ----70-22

1

PG70-22Control

2

AirBlown

3

SBSLG

4

CR-TB

5

TP

6

PG70-22

+Fibers

7

PG70-22

8

SBS64-40

9

AirBlown

10

SBSLG

11

TP

12

Lane 1

CR-AZ

300,000

Lane 2

Control

100,000

Lane 3

Air Blown

100,000

Lane 4

SBS LG

300,000

Lane 5

CR-TB

100,000

Lane 6

TP

200,000

Percentage of Area Cracked vs. ALF Wheel Load Passes

0.0

20.0

40.0

60.0

80.0

100.0

120.0

0 50000 100000 150000 200000 250000 300000 350000

Number of ALF Passes

Pe

rce

nta

ge

of

Are

a C

rac

ke

d, %

L2S3 (Control)

L3S3 (Air Blown)

L5S3 (CR-TB)

L6S3 (Terpolymer)

L4S3 (SBS LG)

L1S2 (CR-AZ)

L4S3L6S3

L5S3

L2S3

L1S2

L3S3

OT vs. FHWA-ALF Fatigue Test Results

0.0

20.0

40.0

60.0

80.0

100.0

120.0

0 50000 100000 150000 200000 250000 300000 350000

Number of ALF Passes

Pe

rce

nta

ge

of

Are

a C

rac

ke

d, %

L2S3 (Control)

L3S3 (Air Blown)

L5S3 (CR-TB)

L6S3 (Terpolymer)

L4S3 (SBS LG)

L1S2 (CR-AZ)

18901120

890

60

Did not crack

80

Last update: February, 2007

Summary of

Overlay Test Results

CSTFlexible Pavements

Overlay Test Results

Approximately 250 tests have been performed.

The data includes duplication. After the mix is tested 3 times, the 3 test results are entered in the database.

The summary only includes data where there were more than 5 occurrences of a similar variable (mix type, aggregate type, PG binder)

Test is perform until a 93% reduction or more of the maximum load measured from the first opening cycle. If 93% is not reach, the test is run to 1200 cycles.

Influence of Mix Type

0

200

400

600

800

1000

1200

1400

Nu

mb

er

of

Cy

cle

s

Median 70 286 353 1200

Count 166 30 5 21

Minimum 2 14 258 147

Maximum 1200 1200 696 1200

Dense Graded

Performance-

Designed

Mixtures

Stone-Matrix

AsphaltRBL/CAM

Influence of Mix Type

Dense-Graded Hot Mix Asphalt

0

50

100

150

200

Nu

mb

er

of

Cy

cle

s

Median 22 65 52 127 193

Count 14 33 49 59 10

Minimum 6 5 2 3 14

Maximum 91 1200 880 1200 1200

Type A Type B Type C Type D Type F

Influence of Mix Type

Performance-Designed Mixtures

Crack Attenuating Mixture (CAM)

Stone-Matrix Asphalt

0

200

400

600

800

1000

1200

1400

Nu

mb

er

of

Cy

cle

s

Median 122 919 944 1200 353

Count 11 16 10 11 5

Minimum 84 14 147 202 258

Maximum 387 1200 1200 1200 696

SP-C CMHB-F RBL CAM SMAR-F

Influence of Aggregate Type

All mixtures

0

200

400

600

800

1000

1200

1400

Nu

mb

er

of

Cy

cle

s

Median 1200 32 90 155 455

Count 6 65 32 9 49

Minimum 1200 2 2 86 6

Maximum 1200 1200 1200 387 1200

Sandstone* Limestone Gravel Rhyolite Granite

*Only CAM mixes tested

Influence of Asphalt

0

150

300

450

600

Nu

mb

er

of

Cy

cle

s

Median 66 86 508 207 404

Count 40 59 34 57 10

Minimum 3 2 6 2 14

Maximum 1200 1200 1200 1200 1200

64-22 70-22 70-28 76-22 CRM

All mixtures

Summary and Future Directions

Conclusions

The Overlay tester and Hamburg wheel are effective predictors of HMA cracking and rutting respectively.

They are both practical tests that appear to correlate well with field performance.

They can be performed in hours -Vs- days

The overlay tester has revealed that most TxDOT mixes are relatively susceptible to cracking.

New “Crack Attenuating Mixes” (CAM) are gaining popularity in Texas. CAM mixes are both rut resistant and crack resistant.

Significant HMA changes are very likely in the near future.

Some Changes are Bigger

than Others

Murom Russia – August 2005

Round Rock – Sept. 2006