Transportation Pooled Fund Project 5(165)

DEVELOPMENT OF DESIGN GUIDE FOR THIN AND ULTRA-THIN CONCRETE

OVERLAYS OF EXISTING ASPHALT PAVEMENTS

TASK 1 REPORT:

COMPILATION AND REVIEW OF EXISTING

PERFORMANCE DATA AND INFORMATION

Prepared by
Manik Barman
Dr. Julie Vandenbossche
Feng Mu
Kerri Gatti

Department of Civil and Environmental Engineering

Swanson School of Engineering
University of Pittsburgh

i
 Technical Report Documentation Page
1. Report No. 2. 3. Recipients Accession No.

4. Title and Subtitle 5. Report Date

Development of Design Guide for Thin and Ultra-thin May 2011
Concrete Overlays of Existing Asphalt Pavements,
Task 1 Report: Compilation and Review of Existing
Performance Data and Information 6.

7. Author(s) 8. Performing Organization Report No.

Julie Vandenbossche, Manik Barman, Feng Mu, Kerri
Gatti
9. Performing Organization Name and Address 10. Project/Task/Work Unit No.
University of Pittsburgh TPF 5(165)/Task 1
Department of Civil and Environmental Engineering 11. Contract (C) or Grant (G) No.
Swanson School of Engineering, Benedum Hall 934
Pittsburgh, PA 15260
12. Sponsoring Organization Name and Address 13. Type of Report and Period Covered
Minnesota Department of Transportation Task 1 Report, 2008-2009
395 John Ireland Boulevard 14. Sponsoring Agency Code
St. Paul, MN 55155
15. Supplementary Notes
First of 8 task reports from this project.
16. Abstract (Limit: 200 words)
The objective of this pooled fund study is to develop a rational mechanistic-empirical based design guide
for bonded whitetopping. In Task 1 of the study, common types of distresses were identified so that the
project team could understand how they initiate and progress over the life of the overlay with respect to
traffic and environmental loads. Information was collected about the design features and performance of
existing bonded whitetopping projects. Findings indicate that the performance of a thin whitetopping
section depends to a large extent upon support conditions provided by the existing asphalt. Successfully
performing projects had an underlying asphalt layer thickness (after milling) of more than 3 inches, unless
a concrete layer was also available underneath. The type of distress that develops in a whitetopping is
primarily a function of the thickness of the PCC, while the extent of the deterioration appears to be
related to the thickness and quality of the existing HMA and the joint layout. If the stiffness ratio of the
PCC and HMA layers falls below one, there is an increase in the potential for reflective cracking into the
whitetopping overlay. Additional test sections need to be constructed to help quantify the affect of fibers
on joint performance.

17. Document Analysis/Descriptors 18. Availability Statement

MnROAD Concrete Overlay No restrictions. Document available from:
Whitetopping Pavement Design National Technical Information Services,
Pavement Rehabilitation Pavement Distress Springfield, Virginia 22161
19. Security Class (this report) 20. Security Class (this page) 21. No. of Pages 22. Price
Unclassified Unclassified 112

ii
 TABLE OF CONTENTS
LIST OF FIGURES ................................................................................................................... v
LIST OF TABLES ................................................................................................................... vii
GLOSSARY OF ACRONYMS ............................................................................................... ix
FOREWARD .............................................................................................................................. x
1 INTRODUCTION................................................................................................................. 1
2 OBJECTIVE OF TASK 1 .................................................................................................... 2
3 FHWA ALF PROJECT IN VIRGINIA ............................................................................. 3
3.1 Summary of distress observations .................................................................................. 7
3.2 Analysis of pavement distresses ..................................................................................... 8
3.2.1 Corner cracking ........................................................................................................ 9
3.2.2 Mid-slab cracking .................................................................................................. 10
3.2.3 Joint faulting .......................................................................................................... 11
3.2.4 Joint spalling .......................................................................................................... 11
3.3 Conclusion .................................................................................................................... 12
4 MINNESOTA ..................................................................................................................... 13
4.1 MnROAD research facility ........................................................................................... 13
4.1.1 Design features....................................................................................................... 15
4.1.2 Distress data ........................................................................................................... 17
4.1.3 Factors affecting performance ............................................................................... 27
4.1.4 Reflection cracking ................................................................................................ 33
4.2 Intersections on US-169, Elk River .............................................................................. 38
4.2.1 Analysis of performance ........................................................................................ 40
4.3 LoRay Drive, North Mankato ....................................................................................... 45
4.4 TH-30, Mankato ............................................................................................................ 45
4.5 Conclusion .................................................................................................................... 48
5 PENNSYLVANIA .............................................................................................................. 48
5.1 Intersection of State Route (SR)-133 and SR-100, Chester County ............................. 48
5.2 SR-30, Lancaster County .............................................................................................. 49
5.3 Conclusion .................................................................................................................... 50
6 TEXAS ................................................................................................................................ 50
6.1 Intersections on LP-250 at Wadley Road, Holiday Hill Road and Midland Drive,
Midland .................................................................................................................................. 50
6.2 Intersection of LP-250 at Midkiff Road and Garfield Road, Midland ......................... 52
6.3 Conclusion .................................................................................................................... 53
7 MISSOURI ......................................................................................................................... 53
7.1 Intersection of SR-291 and SR-78, Independence ........................................................ 54
7.2 US-60, between US-71 and BUS-71, near Neosho ...................................................... 55
7.3 US-169 and SR-YY intersection, St. Joseph ................................................................ 57
7.4 Other projects in Missouri ............................................................................................ 58
7.5 Conclusion .................................................................................................................... 59

iii
8 MISSISSIPPI ...................................................................................................................... 60
8.1 Intersection of SR-15 and US-80, Newton County ...................................................... 60
8.2 Intersection of 22nd Avenue and North Frontage Road, Lauderdale, MS ................... 61
8.3 Intersection of SR-35 and US-80, Scott County ........................................................... 62
8.4 Conclusion .................................................................................................................... 63
9 NEW YORK ....................................................................................................................... 63
9.1 Intersection at Waldon Avenue and Central Avenue, near Buffalo.............................. 63
9.2 NY-408 and SH-622, Rochester ................................................................................... 64
9.3 Conclusion .................................................................................................................... 64
10 ILLINOIS ......................................................................................................................... 65
10.1 IL-146, Anna ............................................................................................................... 65
10.2 Intersection of US-36 and Oakland Avenue, Decatur ................................................ 69
10.3 US-36, Tuscola ........................................................................................................... 71
10.4 Highway-2, Cumberland County ................................................................................ 73
10.5 Highway-4, Piatt County ............................................................................................ 75
10.6 US-36 and Country Club Road, Decatur .................................................................... 76
10.7 Conclusion .................................................................................................................. 78
11 IOWA ................................................................................................................................ 78
11.1 R-16, Dallas County.................................................................................................... 79
11.2 IA- 21, Iowa County ................................................................................................... 80
11.3 Conclusion .................................................................................................................. 83
12 MICHIGAN ...................................................................................................................... 83
12.1 Patterson Avenue, from 44th Street to 36th Street, Kentwood ................................... 83
12.2 Zeeb Road, Ann Arbor, Washtenaw ........................................................................... 85
12.3 Schaefer Highway, Detroit, Wayne County ............................................................... 86
12.4 Conclusion .................................................................................................................. 86
13 OKLAHOMA ................................................................................................................... 86
13.1 US-69, North of Stringtown, Atoka County ............................................................... 86
13.2 US-69, North of McAlester, NB and SB Lanes, Pittsburg County............................ 86
13.3 Conclusion .................................................................................................................. 87
14 OVERALL CONCLUSIONS ......................................................................................... 87
REFERENCES ......................................................................................................................... 99

iv
 LIST OF FIGURES
Figure 1: Schematic of distress in the UTW lanes at the FHWA ALF (a) Lane 5; (b) Lane 6; (c)
Lane 7; (d) Lane 8; (e) Lane 9; (f) Lane 10; (g) Lane 11 and (h) Lane 12. ................................. 5
Figure 2: Typical UTW transverse cracking at the FHWA ALF................................................. 6
Figure 3: Typical UTW corner cracking at the FHWA ALF. ...................................................... 6
Figure 4: Typical UTW longitudinal faulting at the FHWA ALF. .............................................. 6
Figure 5: Comparison of HMA rutting to UTW cracking. .......................................................... 9
Figure 6: Permanent deformation of HMA base........................................................................ 10
Figure 7: Mid-slab cracking mechanism of an UTW slab. ........................................................ 11
Figure 8: Longitudinal joint faulting mechanism in ALF UTW slabs....................................... 11
Figure 9: Transverse joint faulting mechanism in ALF UTW slabs. ......................................... 12
Figure 10: Joint spalling mechanism in ALF UTW slabs.......................................................... 12
Figure 11: Layout of test cells at the MnROAD facility. .......................................................... 14
Figure 12: Test cells constructed during Oct 1997 and 2004. ................................................... 14
Figure 13: Test cells constructed during 2008. .......................................................................... 15
Figure 14: Performance history for (a) Cell 93, (b) Cell 94 and (c) Cell 95 from 1997 to 2001.
.................................................................................................................................................... 19
Figure 15: Longitudinal cracking history for Cells 92, 96 and 97 (11.5 years, 9.8 million
ESALs); 60-63 (4.5 years, 3.8 million ESALs). ........................................................................ 21
Figure 16: Transverse cracking history for Cells 92, 96 and 97 (11.5 years, 9.8 million ESALs);
60-63 (4.5 years, 3.8 million ESALs). ....................................................................................... 21
Figure 17: Corner cracking history for Cells 92, 96 and 97 (11.5 years, 9.8 million ESALs); 60-
63 (4.5 years, 3.8 million ESALs). ............................................................................................ 22
Figure 18: Longitudinal cracks in Cell 97. ................................................................................ 23
Figure 19: Longitudinal cracks in Cells 60 and 62. ................................................................... 23
Figure 20: Longitudinal cracks in Cells 61 and 63. ................................................................... 24
Figure 21: Reflection transverse cracks in Cell 61. ................................................................... 24
Figure 22: Reflection transverse cracks in Cell 63 at two locations. ......................................... 25
Figure 23: Faulting history for MnROAD cells until October 2007. ........................................ 26
Figure 24: IRI history for MnROAD cells until March 2008. ................................................... 26
Figure 25: Corner cracks along the inside wheelpath on Cell 94. ............................................. 28
Figure 26: Distress is reduced with the 5-ft x 6-ft panels because the wheelpath is not located
adjacent to the longitudinal joint (Burnham, 2005). .................................................................. 29
Figure 27: Faulting of panels in Cell 97(undoweled) but no faulting in Cell 92 (1-in dowels). 30
Figure 28: Difference in the amount of water infiltrating the surface for the sealed ................. 31
Figure 29: Strain directly under a 9-kip FWD load applied in the wheelpath adjacent to a
transverse joint on a 3-in overlay with 5-ft x 6-ft panels. (Vandenbossche, 2005) ................... 32
Figure 30: Relative stiffness of PCC and HMA layers for Cells 92 through 97. ...................... 37
Figure 31: Relative stiffness of PCC and HMA layers for Cells 60 through 63. ...................... 37
Figure 32: Layout of the UTW sections at the intersections of US-169. ................................... 40
Figure 33: Transverse crack and corner breaks in the Jackson Street test section (03.30.98).
(Vandenbossche, 2003) .............................................................................................................. 42
Figure 34: Corner breaks in the inside wheelpath at the Jackson Street test section. (07.20.99)
(Vandenbossche, 2003) .............................................................................................................. 42
Figure 35: Typical distress patterns that developed at the 6 ft x 6 ft Main Street test intersection.
(Vandenbossche, 2003) .............................................................................................................. 43

v
Figure 1: Relative stiffness of PCC and HMA layers for US-169 UTW test sections……….44
Figure 37: Plot of IRI data for TH-30 from 1994 to 2001. ........................................................ 47
Figure 38: UTW at the intersection of LP-250 and Midland Drive at Odessa District. ............ 51
Figure 39: UTW at the intersection of LP-250 and Wadley Avenue at Odessa District. .......... 52
Figure 2: Anna project layout (Winkelman, 2005)…………………………………………...66
Figure 41: Distresses with respect to ESALs. ............................................................................ 68
Figure 42: Distressed panels at the intersection. (Winkelman, 2005) ....................................... 69
Figure 43: Movement of the panels toward the intersection...................................................... 71
Figure 44: Blow up of the UTW slabs (Winkelman, 2005). ...................................................... 78
Figure 45: Condition of the original HMA layer at the time of construction. ........................... 84
Figure 46: Condition of the UTW after two years (2008). ........................................................ 85
Figure 47: Transverse crack development in overlays < 4 in. .................................................. 96

vi
 LIST OF TABLES
Table 1: Whitetopping categories. ............................................................................................... 1
Table 2: HMA characteristics at the FHWA ALF. ...................................................................... 4
Table 3: Lane assignments for UTW designs at the FHWA ALF. .............................................. 4
Table 4: Summary of the design features for the MnROAD whitetopping cells....................... 16
Table 5: Concrete mixture design information for MnROAD whitetopping cells. ................... 17
Table 6: Summary of cracks for MnROAD test sections. ......................................................... 20
Table 1: Summary of transverse reflection cracking for1997 and 2004 MnROAD cells..........34
Table 8: Asphalt resilient modulus at different temperatures for MnROAD sections. ............. 35
Table 9: Summary of the design features for the US-169 cells. ................................................ 39
Table 10: Concrete mixture design for the US-169 cells. .......................................................... 39
Table 11: Design features for sections at TH-30. (Burnham and Rettner, 2002) ...................... 46
Table 12: Summary of visual distresses for the sections of TH-30. ………………………….
Burnham and Rettner, 2002) ...................................................................................................... 47
Table 13: Concrete mixture design information. ....................................................................... 49
Table 14: Concrete mixture design information. ....................................................................... 51
Table 15: Concrete mixture design information ........................................................................ 53
Table 16: Layer properties of the project at SR-291 and SR-78................................................ 54
Table 17: Concrete mixture design information. ....................................................................... 55
Table 18: Layer properties of US-60 project. ............................................................................ 56
Table 19: Concrete mixture design information. ....................................................................... 56
Table 20: Layer properties for the US-169 project. ................................................................... 57
Table 21: Concrete mixture design information. ....................................................................... 58
Table 22: Layer properties for the project at MO-5, railroad crossing in Lebanon. .................. 59
Table 23: Layer properties for the project at BUS-13 in Branson West. ................................... 59
Table 24: Layer properties for the project at the Intersection of BUS-13 and MO-14. ............. 59
Table 25: Overlay design features and observed distress type for the project at ....................... 61
Table 26: Overlay design features and observed distress type for the project at ....................... 63
Table 2: Whitetopping project information in Illinois (Winkelman, 2005 and Roesler et al.,
2008)………………………………………………………………………………………...…65
Table 28: Concrete mixture design information. ....................................................................... 67
Table 29: Traffic details for the Anna Project. .......................................................................... 67
Table 30: Distress summary until 2004. (Winkelman, 2005) .................................................... 68
Table 3: Concrete mixture design information (Winkelman, 2005 and Roesler et al., 2008)..70
Table 32: Traffic details for the UTW project (Winkelman, 2005). .......................................... 70
Table 33: Distress summary until 2004. (Winkelman, 2005) .................................................... 71
Table 34: Concrete mixture design information. ....................................................................... 72
Table 35: Traffic details for US-36 in Tuscola (Winkelman, 2005 and Roesler et al., 2008). .. 72
Table 36: Distress summary for the project until 2004 (Winkelman, 2005). ............................ 73
Table 37: Concrete mixture design. (Winkelman, 2005 and Roesler et al., 2008) .................... 74
Table 38: Traffic details for Highway-2. ................................................................................... 74
Table 39: Distress summary for the project. (Winkelman, 2005).............................................. 74
Table 40: Distress summary for the project. (Winkelman, 2005).............................................. 75
Table 41: Traffic volumes for the US-36 and County Club Road project. (Winkelman, 2005) 77
Table 42: Distress survey results for the project. (Winkelman, 2005) ...................................... 77

vii
Table 43: Iowa shear strength for different surface preparation techniques (Grove et al., 2001).
.................................................................................................................................................... 80
Table 44: ESALs from 1995 through 1999 (Cable et al., 2001). ............................................... 81
Table 45: Details of pavement layers at Patterson Avenue, from 44th ..................................... 84
Table 46: Summary of projects evaluated. ............................................................................... 88

viii
 GLOSSARY OF ACRONYMS

(i) AADT: Annual Average Daily Traffic

(ii) AASHTO: American Association of State Highway and Transportation Officials
(iii) ACPA: American Concrete Pavement Association
(iv) ADT: Average Daily Traffic
(v) ALF: Accelerated Testing Facility
(vi) BM: Base mixture
(vii) BUS: Business
(viii) CIPR: Cold in-Place Recycle
(ix) CDOT: Colorado Department of Transportation
(x) DOT: Department of Transportation
(xi) FHWA: Federal Highway Administration
(xii) HMA: Hot Mix Asphalt
(xiii) LP: Loop
(xiv) Mn/DOT: Minnesota Department of Transportation
(xv) MnROAD: Minnesota Road Research Facility
(xvi) MO: Missouri
(xvii) NCHRP: National Cooperative Highway Research Program
(xviii) NY: New York
(xix) PCA : Portland Cement Association
(xx) PCC: Plain Cement Concrete
(xxi) PennDOT: Pennsylvania Department of Transportation
(xxii) I: Interstate
(xxiii) IA: Iowa
(xxiv) IL: Illinois
(xxv) SH: State Highway
(xxvi) SM: Surface mixture
(xxvii) SR : State Route
(xxviii) TH: Truck Highway
(xxix) TS: Test Section;
(xxx) TWT: Thin Whitetopping
(xxxi) US: United States
(xxxii) UTW: Ultra-Thin Whitetopping
(xxxiii) WIM: Weigh-in-Motion

ix
FOREWORD
As the need to rehabilitate asphalt roadways throughout the United States has increased
significantly, so has the need for a more rational design method for one of the repair options
known as “whitetopping.” While a handful of mechanistic and “rule of thumb” based design
methods for whitetopping have been available for many years, the increasing need to work
within ever shrinking budgets requires that those designs better predict long term performance.

This report comprises the first of 8 task reports written as part of the Transportation Pooled
Fund Project 5(165): “Development of Design Guide for Thin and Ultra-thin Concrete
Overlays of Existing Asphalt Pavements.” This project, started in September 2008, was
created to address the urgent need by the pavement design community for a rational
mechanistic-empirical based design method for whitetopping.

Special thanks are given to the states who generously donated toward this effort. Those states
included: Mississippi, Missouri, Minnesota, New York, Pennsylvania, and Texas.

The participating states, as well as the overall pavement design community, are grateful for the
hard work and dedication given by principal investigator Dr. Julie Vandenbossche and her
students.

Tom Burnham, P.E.

Project Manager
Minnesota Department of Transportation

x
 ACKNOWLEDGEMENTS

This report contains the performance review of existing bonded whitetopping projects based on
data available in literature and personal communication with various state agencies. The
Departments of Transportation from Pennsylvania, Minnesota, Texas, Mississippi, Missouri
and New York were extremely helpful in providing performance data. Many local chapters of
the American Concrete Pavement Association (ACPA) also contributed valuable information
regarding the design and performance of these whitetopping projects, especially in Michigan
and Oklahoma. The contributions of these agencies, directly or indirectly involved with this
project are gratefully appreciated.

The authors would also like to acknowledge the contribution of the officials of the Minnesota
Department of Transportation, Mn/DOT, for providing the distress data for the bonded
whitetopping constructed at MnROAD. The time and effort dedicated by Mr. Thomas
Burnham, Mr. Timothy Cline and Mr. Benjamin Worel is gratefully appreciated.

The authors express their sincere gratitude and acknowledge the kind cooperation extended by
the following people:

1) Mr. James Sherwood, Federal Highway Administration

2) Mr. Benjamin E. Singer, Pennsylvania Department of Transportation, District 8
3) Ms. Lorraine D. Ryan, Pennsylvania Department of Transportation, District 6
4) Mr. Alan Cross, Mississippi Department of Transportation
5) Mr. William Barstis, Mississippi Department of Transportation
6) Mr. Hua Chen, Texas Department of Transportation
7) Mr. John P. Donahue, Missouri Department of Transportation
8) Mr. Michael Brinkman, New York State Department of Transportation
9) Mr. Daniel M. Degraaf, Michigan Concrete Paving Association
10) Mr. Steven M. Waalkes, Michigan Concrete Paving Association
11) Mr. Clayton J. Stahl, President, Gulisek Construction, LLC
12) Ms. Lynn Tiernan, Contract Administrator, Gulisek Construction, LLC
Mr. Brent Burwell, American Concrete Paving Association OK/AR Chapter
xi
1 INTRODUCTION
Whitetopping is a rehabilitation technique that consists of a concrete overlay on a distressed
asphalt pavement. Although this kind of rehabilitation technique has been reported in literature
dating back to 1918, very few projects were noted until the 1990s (Rasmussen and Rozycki,
2004). Since 1992, this type of project has gained momentum in the United States with
approximately 500 projects in existence across the country (Roesler et al, 2008). Projects of
this type are also being constructed around the world in countries such as Canada, Brazil and
Taiwan.

Within the category of whitetopping, there are additional classifications historically used based
on the thickness of the Portland cement concrete (PCC) layer, as presented in Table 1. For this
study, the focus will be on bonded whitetopping. This includes whitetopping historically
referred to as thin and ultra-thin. For bonded whitetopping, a good bond must be achieved
between PCC and hot mix asphalt (HMA) layers. This allows for a thinner PCC overlay to be
constructed while still fulfilling the intended service life.

Table 1: Whitetopping categories.

Whitetopping category PCC thickness (in)
Conventional whitetopping >6
Thin whitetopping (TWT) 4–6
Ultra-thin whitetopping (UTW) 2–4

The increasing popularity of this method of pavement rehabilitation has triggered many
agencies to put effort towards the development of a design procedure. Many agencies, namely
the Portland Cement Association (PCA), the Colorado Department of Transportation (CDOT),
the New Jersey Department of Transportation, the American Concrete Pavement Association
(ACPA), the Federal Highway Administration (FHWA), and the Illinois Department of
Transportation (IlDOT) have proposed their own design procedures. Each procedure has its
own advantages and disadvantages in comparison to the others. It is common for these
procedures to have only been calibrated using a limited number of local projects. Based on
that fact, their extensive use remains questionable. As per the National Cooperative Highway

1
Research Program (NCHRP) Synthesis 338 (Rasmussen and Rozycki, 2004), state departments
of transportation (DOTs) like Arizona, Iowa, Illinois, Mississippi, Texas, Missouri, Kansas and
Utah adopted the ACPA design procedure for whitetopping design. Some states used the
CDOT design procedure, while others referred to the 1993 American Association of State
Highway Transportation Officials (AASHTO) Guide for the Design of Pavement Structures,
which was not intended for the design of whitetopping. A considerable amount of projects
were also constructed without following any method. A need for developing a mechanistic-
empirical design procedure is then raised so that the whitetopping projects in the participating
states can be carried out more successfully.

The objective of this pooled fund study is to develop a rational mechanistic-empirical based
design guide for bonded whitetopping. In the first stage of the study, the common types of
distresses were identified so that the project team could understand how they initiate and
investigate their progression over the life of the overlay with respect to traffic and
environmental loads. In general, the objective of Task 1 is to collect information about the
design features and performance of the existing bonded whitetopping projects as described in
the sections below.

2 OBJECTIVE OF TASK 1
The main objective of Task 1 is to identify, collect, compile and review the performance of the
existing bonded whitetopping projects from different states. In this pooled fund study, there
are six different participating states: Pennsylvania, Minnesota, Missouri, Mississippi, Texas
and New York. The Federal Highway Administration also provides a supporting role. Distress
data for bonded whitetopping constructed in these states, as well as a few other states like
Illinois, Michigan and Oklahoma, has been collected. In addition, the distress data from the
MnROAD pavement testing facility has been collected and analyzed in this task. The collected
data has been reviewed and analyzed in an attempt to answer the following questions:
(i) What is the minimum required HMA thickness and maximum allowable distress
level?
(ii) Are there modes of failure other than corner cracks that frequently develop and
what common parameters are present when these additional modes of failures occur?

2
 (iii) Under what conditions does reflection cracking typically occur?
(iv) What surface preparation techniques have been used and what level of performance
was achieved? What is the minimum acceptable level of bonding?
(v) What are acceptable joint patterns?
(vi) Is there evidence from companion test sections that structural fibers help improve
the performance beyond providing additional safety once deterioration begins?
(vii) What factors would contribute to the development of corner breaks?
(viii) Do fibers help to increase the load transfer efficiency for longer periods of time by
holding the cracks together?

The review of the performance of the bonded whitetopping projects presented below begins
with the bonded whitetopping constructed at the FHWA accelerated loading facility (ALF).
Next the construction of the bonded whitetopping projects constructed at MnROAD will be
discussed. Finally, a review of the performance of bonded whitetopping projects constructed
on in-service pavements throughout eight different states will be provided. The information
gathered from all of these projects will then be compiled in an attempt to address the questions
defined above.

3 FHWA ALF PROJECT IN VIRGINIA

A comprehensive review of the performance of the whitetopping pavements at the FHWA
ALF at the Turner–Fairbank Highway Research Center in McLean, Virginia was provided in
Synthesis 338 (Rasmussen and Rozycki, 2004) and Rasmussen et al. (2002). In 1998, eight
lanes of UTW overlays were placed over the existing HMA pavements that were built during
1993. Each of the existing lanes had four test sections. The existing HMA pavements were
subjected to accelerated loadings. Since the underlying HMA pavements were previously
loaded by the accelerated loading equipment, the UTW application over these HMA
pavements more closely represents field conditions. The original HMA pavement structure
consisted of an 8-in HMA layer on top of an 18-in unbound crushed aggregate base and 24-in
of A-4 (AASHTO) granular material. The HMA layer in each test section was constructed
with a different binder and aggregate mixture. See Table 2. Lanes 11 and 12 were constructed
with base mixtures (BM) having a top size of 1½ in, and constructed in two 4-in lifts. The
other lanes were constructed with surface mixtures (SM), with a top size of ¾ in, laid in four 2-

3
in lifts. Table 2 also presents a summary of the performance of different types of HMA
mixtures under the accelerated loading in terms of applications required to achieve a 0.8-in rut
depth. A fairly constant temperature rangeing from 115 to 170 oF was maintained during the
accelerated loadings. The values of rut depth presented in the Table 2 were obtained at 136 oF.
It was observed that for both the AC-5 and AC-20 binders, the service life increased by a factor
of 10 with the increasing aggregate top size.

The original HMA layer was milled to a depth equivalent to the thickness of the UTW layer.
The eight UTW test lanes were constructed at dimensions of 12-ft wide and 48-ft long. Half of
the lanes were constructed using polypropylene fiber and the other half were built with plain
concrete (Table 3). As mentioned earlier, the experimental design variables included a range
of HMA binder types and two different mixtures (aggregate gradation) along with two
different UTW thicknesses and three different joint spacings (Table 2 and Table 3).

Table 2: HMA characteristics at the FHWA ALF.

(Rasmussen and Rozycki, 2004)
Binder
Lane- ALF wheel passes for Percent rut depth in HMA at
type/
section 0.8-in total rut depth 0.8-in total rut depth
mixture
5-2 AC-10/ SM 1,180 78
6-2 AC-20/ SM 2245 75
7-2 Styrelf 16,060 63
8-2 Novophalt 33,750 32
9-1&2 AC-5/ SM 525 77
10-1 AC-20/ SM 2,100 86
11-1&2 AC-5/ BM 5,755 82
12-1 AC-20/ BM 22,100 85

Table 3: Lane assignments for UTW designs at the FHWA ALF.

(Rasmussen and Rozycki, 2004)
UTW built thickness (in) Joint spacing (ft) Fiber concrete Plain concrete
4 Lane 5 Lane 6
3.25
3 Lane 7 Lane 8
6 Lane 9 Lane 10
4.50
4 Lane 11 Lane 12

4
The numbers of load applications for each overlay ranged from 200,000 to more than 1
million. The test load was 12.3 kips on all lanes except for the first 310,000 load applications
on Lanes 11 and 12 where a load of 10 kips was used. It may be noted that the temperature of
the HMA layer was kept constant, at about 80 oF, when loaded. In order to maintain this
uniform temperature, radiant heaters were cycled on and off when the ambient temperature was
lower than 80 oF, which resulted in temperature gradients through the concrete. The exhibited
distress was recorded as it developed through periodical visual observations. Typical
distresses observed were mid-slab transverse cracking, mid-slab longitudinal cracking, corner
cracking, joint faulting (longitudinal and transverse), and spalling. Figure 3 is a schematic of
the distress observed at the end of loading.

Figure 4 and Figure 5 show typical transverse and corner cracks observed in the ALF
experimental study. The most significant faulting was observed along the longitudinal joint
probably due to the channelized nature of the loading. Faulting was also observed in a few
transverse joints but this was less severe compared to the longitudinal faulting. Figure 6 shows
faulting that developed. Most of the observed spalling was of low severity.

Figure 3: Schematic of distress in the UTW lanes at the FHWA ALF (a) Lane 5; (b) Lane 6; (c)
Lane 7; (d) Lane 8; (e) Lane 9; (f) Lane 10; (g) Lane 11 and (h) Lane 12.
(Rasmussen et al., 2002)

5
Figure 4: Typical UTW transverse cracking at the FHWA ALF.
(Rasmussen et al., 2002)

Figure 5: Typical UTW corner cracking at the FHWA ALF.

(Rasmussen et al., 2002)

Figure 6: Typical UTW longitudinal faulting at the FHWA ALF.

(Rasmussen et al., 2002)

6
3.1 Summary of distress observations
The details of the distresses observed in each lane, as discussed by Rasmussen et al. (2002) are
presented in the following subsections. Distress in Lanes 6 and 10 are not presented. These
two lanes were still under load applications for another project.

Lane 5: The panels in this lane were heavily damaged after 194,500 load applications. Every
panel along the wheel path developed cracks except for the southernmost panel. The majority
of the distress was corner cracks. Transverse cracks developed in three panels. Significant
longitudinal joint faulting was also observed.

Lane 7: After 283,492 load applications, the panels in this lane experienced no distress. Only
a few cracks were observed at the transition zone, where the wheel load first exerts an impact
on the pavement at loading.

Lane 8: After 625,838 load applications, some cracks were observed in the lane. This included
corner cracks at the southern end of the lane and a corner crack and a partial longitudinal crack
at the northern end of the lane. In addition to these cracks, spalling was observed at five
locations. All observed spalling occurred at the corner and was approximately 1 to 2 in. A
minor degree of faulting along the longitudinal joint was also present in this lane.

Lane 9: A considerable amount of distress was observed in Lane 9 after 265,913 load
applications. Every panel in the wheel path developed cracks. The majority of the distress was
corner cracking, with at least one in each slab. A moderate degree of joint faulting was also
observed along the longitudinal joint.

Lane 11: Minimal damage was observed Lane 11 after 1,071,302 load applications. In the
transition zone, only three cracks were noted: one corner crack, one partial longitudinal crack,
and one partial transverse crack. However, some joint faulting was observed.

7
Lane 12: Lane 12 was in good condition after 1,071,312 load applications. One full and one
partial transverse crack were observed on the southern end of the lane. No significant faulting
was observed.

Overall distress surveys for the ALF UTW experimental study revealed that the majority of the
distresses were corner cracks followed by transverse and sometimes longitudinal cracks.
Longitudinal joint faulting was also found to be significant in places due to the channelized
loading of the ALF.

3.2 Analysis of pavement distresses

An analysis of the performance of the UTW lanes under the accelerated load applications was
performed by Rasmussen et al. (2002). Links were made between the UTW design features
and the observed distress. Possible hypotheses of the various failure mechanisms were also
developed. The type of distress in the UTW was also evaluated with respect to the type of
HMA used to construct the underlying layer.

To make a rough comparison, the degree of cracking observed in each lane was quantified by
using a cracking index. The cracking index is a weighted average of the number of cracks
observed in each lane and it is calculated by summing up the cracks for each lane with a full-
panel crack weighing 1.0, a partial-panel crack 0.5, and a small chip or break 0.1 (Rasmussen
et al., 2002). This cracking index is then compared with rutting that has been quantified as the
number of ALF load applications to reach 0.8-in of total rutting (Figure 7). The hypothesis of
the cause of various distresses that were observed in the ALF UTW experimental study is
presented in the following sections. The mechanism causing each distress was analyzed by
Rasmussen et al. (2002) and a summary of their findings is provided below.

8
 (Note: 20 mm =0.8 in)
Figure 7: Comparison of HMA rutting to UTW cracking.
(Rasmussen et al., 2002)

3.2.1 Corner cracking

Corner cracking is the most common form of distress typically exhibited by whitetopping
pavements, and this was reflected at the ALF UTW project. This is a fatigue-related distress.
The stress state within the UTW as a result of the applied loads can change as the support
conditions change. The change in support conditions can be due to permanent deformation
accumulated with the numbers of load applications or stripping/raveling of the underlying
HMA layer. Figure 8 shows formation of a void underneath the UTW due to permanent
deformation of HMA layer. The UTW slab then acts as a cantilever under the wheel load and
a corner crack develops. Rasmussen et al. (2002) also hypothesized that the corner cracks
might have initiated at the loaded (longitudinal) edge of the slab and propagated diagonally
toward the closest intersecting joint with each successive wheel load. It will be shown in
subsequent sections that other studies have indicated that these corner cracks can also initiate at
the intersection of the wheelpath and the transverse joint and propagate until intersecting with
the longitudinal joint.

9
 Figure 8: Permanent deformation of HMA base.
(Rasmussen et al., 2002)

3.2.2 Mid-slab cracking

Transverse cracking or mid-slab cracking also develops when the concrete loading exceeds the
fatigue limit. Rasmussen et al. (2002) hypothesized that the mid-slab cracking initiates at the
bottom of the UTW slab. Figure 9 illustrates the initiation of mid slab cracking when wheel
load passes directly over the mid-slab at the edge. It is assumed that stress at the edge will be
the highest. This tensile stress at the bottom of the slab is compounded by the presence of a
void, or soft area beneath the slab. Another hypothesis is that the cracks are initiated at the top
of the slab, induced by tensile stresses at the top as the wheel load rolls onto the slab.

10
 Figure 9: Mid-slab cracking mechanism of an UTW slab.
(Rasmussen et al., 2002)

3.2.3 Joint faulting

At the ALF UTW project, both longitudinal as well as transverse faulting was observed. The
channelized nature of the wheel loading resulted in more faulting along the longitudinal joint
where the wheel load is applied. See Figure 10. Regarding the development of transverse
cracking, Rasmussen et al. (2002) hypothesized that under wheel loading, both normal and
shearing forces are generated in the support layer materials, which results in deformations that
lead to transverse faulting, as shown in Figure 11. But, faulting might also be caused by the
erosion of fines as the HMA layer begins to strip due to water infiltration into the transverse
joint.

Figure 10: Longitudinal joint faulting mechanism in ALF UTW slabs.

(Rasmussen et al., 2002)

3.2.4 Joint spalling

Rasmussen et al. (2002) described two common types of joint spalling. The first type, referred
to as “delamination spalling,” is caused by the combined effect of horizontal early edge
microcracking of concrete and traffic loading that eventually weakens the horizontal crack.

11
This is a flat-bottom spall. The second type of spalling is termed “deflection spalling.” This is
caused when the slab undergoes a substantial amount of deflection under heavy wheel loads.
This kind of spalling is observed in airport pavements where the high deflections in the slabs
cause a localized crushing of the material at the joints (Figure 12). Since the slabs for UTW are
comparatively thin, the deflection spalling might govern the mechanism.

Figure 11: Transverse joint faulting mechanism in ALF UTW slabs.

(Rasmussen et al., 2002; Rasmussen and Rozycki, 2004)

Figure 12: Joint spalling mechanism in ALF UTW slabs.

(Rasmussen et al., 2002; Rasmussen and Rozycki, 2004)

3.3 Conclusion
The FHWA ALF study is a great source of information in regard to the performance of UTW
under accelerated loading conditions. This study provided a reasonable correlation of the
performances of the UTW with respect to the underlying HMA characteristics. The
mechanisms of the different types of distress in whitetopping were summarized well and more
importantly, these hypothesized mechanisms were validated by monitoring strains and
deflections throughout the testing.

12
4 MINNESOTA
The Minnesota Department of Transportation (Mn/DOT) constructed several UTW and TWT
projects. An UTW was constructed on LoRay Drive in Mankato, Minnesota in 1993 and then
a TWT was constructed on Truck Highway (TH)-30 in 1995. In 1997, two whitetopping
projects were constructed. The first project was an UTW constructed on United States (US)-
169 in Elk River and the second project, which consisted of both UTW and TWT, was
constructed on Interstate (I)-94 at the MnROAD Research Facility. Both projects were heavily
instrumented to measure the static and dynamic responses of the pavements under various
applied loadings and environmental conditions. Additional TWT and UTW projects were
constructed on I-94 at MnROAD in 2004. A summary of the performance and findings of
these projects is provided below. The discussion will begin with a review of the MnROAD
and US-169 projects since they were the most fruitful in providing a better understanding of
how these pavements perform. This will be followed by a discussion of the performance of the
LorRay Drive and TH-30 sections.

4.1 MnROAD research facility

MnROAD is a full-scale pavement test facility consisting of a 3.5-mile section of interstate and
a 2.5-mile low-volume roadway near Albertville, Minnesota, approximately 35 miles
northwest of Minneapolis. The MnROAD Facility contains more than 50 pavement designs,
defined as cells. A test cell is typically 100 to 500 ft in length. The layout of the MnROAD
interstate test cells is illustrated in Figure 13. Six different sections (Cell 92 through 97) were
constructed with whitetopping in 1997. Figure 14 shows design details for the test cells.
Among these sections, Cells 93, 94 and 95 were replaced by Cells 60 through 63 in 2004
(Figure 14) because of an excessive drop in serviceability. In 2008, nine more cells were
constructed (Figure 15). The test cells constructed as part of the interstate are referred to as the
mainline and are subjected to live traffic redirected from westbound traffic on I-94. The low-
volume road is subjected to a controlled 5-axle truck with 80-kip total weight loading one lane
and 102-kip total weight loading the other lane.

13
 Figure 13: Layout of test cells at the MnROAD facility.
(Adapted from: http://www.dot.state.mn.us/MnROAD/testsections/pdfs/mainline-profile-ls.pdf,
February 20th, 2009)

Cell 93 Cell 94 Cell 95 Cell 96 Cell 97 Cell 92 Cell 60 Cell 61 Cell 62 Cell 63
3 in 3 in 4 in 4 in
4 in 5 in 5 in
sealed no seal
6 in 6 in 6 in sealed no seal

10 in 10 in 8 in 8 in
9 in 7 in 7 in
58-28 58-28 58-28 58-28
58-28 7 in 7 in 7 in 58-28 58-28
1993 1993 1993 1993
1993 58-28 58-28 58-28 1993 1993
HMA HMA HMA HMA
HMA 1993 1993 1993 HMA HMA
HMA HMA HMA
Clay Clay Clay Clay
Clay Clay Clay Clay Clay Clay

Trans
Trans Trans Trans Trans
Trans Tined
Tined Tined Tined Tined
Tined 10-ft x 12- Turf Turf Turf Turf
4-ft x 4-ft 4-ft x 4-ft 5-ft x 6-ft 10-ft x 12-
5-ft x 6-ft ft Panel 5-ft x 6-ft 5-ft x 6-ft 5-ft x 6-ft 5-ft x 6-ft
Panel Panel Panel ft Panel
Panel Polypro- Panel Panel Panel Panel
Polypro- Polypro- Polypro- Polypro-
Polyolefin pylene
pylene pylene pylene pylene
1-in dowel

Oct-97 Oct-97 Oct-97 Oct-97 Oct-97 Oct-97 Oct-04 Oct-04 Oct-04 Oct-04
Oct-04 Oct-04 Oct-04 Current Current Current Current Current Current Current

Figure 14: Test cells constructed during Oct 1997 and 2004.
(Adapted from: http://www.dot.state.mn.us/MnROAD/testsections/pdfs/mainline-profile-ls.pdf,
February 20th, 2009)

14
 Cell 14
SC* 114 SC* 214 SC* 314 SC* 414 SC* 514 SC* 614 SC* 714 SC* 814 SC* 914
6 in 6 in 6 in 6 in 6 in 6 in 6 in 6 in
6 in
long long long long long long long long
long broom
broom broom broom broom broom broom broom broom
5 in 5 in
6 in 6 in
58-28 58-28 7 in 7 in
58-28 58-28 7.5 in
1993 1993 58-28 58-28 8 in 8 in
1993 1993 58-28
HMA HMA 1993 1993 58-28 58-28
HMA HMA 1993
HMA HMA 1993 1993
HMA
HMA HMA
Clay Clay
Clay Clay
Clay Clay
Clay
Clay Clay

6-ft x 6-ft 6-ft x 6-ft

6-ft x 6-ft 6-ft x 6-ft 6-ft x 12-ft 6-ft x 6-ft
Panel Panel
Panel Panel Panel Panel
1-in 1-in
6-ft x 6-ft 1-in dowels 1-in dowels Flat dowels 1-in dowels 6x6
dowels in 6-ft x 6-ft dowels in
Panel in dr. lane in dr. lane in dr. lane in dr. lane Panel
dr. lane No dowels dr. lane
No dowels No dowels No dowels No dowels No dowels No dowels
No dowels No dowels
in passing in passing in passing in passing
in passing in passing
lane. lane. lane. lane.
lane. lane.
Oct-08 Oct-08 Oct-08 Oct-08 Oct-08 Oct-08 Oct-08 Oct-08 Oct-08
Current Current Current Current Current Current Current Current Current
*SC: Sub Cell, dr. - driving

Figure 15: Test cells constructed during 2008.

(Adapted from:http://www.dot.state.mn.us/MnROAD/testsections/pdfs/mainline-profile-ls.pdf,
February 20th, 2009)

4.1.1 Design features

The design features of the 1997 and 2004 MnROAD whitetopping sections are summarized in
Table 4. The original thickness of the existing full-depth asphalt pavement was 13 in. It was
constructed in 1993 on a silty-clay subgrade. This pavement was serving as a transition zone
that separated the 5- and 10 year design mainline test cells (Vandenbossche and Rettner, 1998).
It should also be noted that the condition of the existing HMA layer was sound with few
transverse cracks. The HMA layer was constructed with a Superpave Grade 58-28 mix. Before
the overlay was placed, the existing HMA layers were milled to an equivalent thickness of the
whitetopping to ensure sufficient bonding between the HMA and PCC layers.

15
Both thin and ultra-thin whitetopping sections were constructed at MnROAD and they
represent a range of thickness and panel sizes. Cells 93, 94 and 95 ranged between 3- to 4-in
thick with 4-ft × 4-ft and 5-ft x 6-ft panels. The joints were sealed but the sections were not
doweled. Cells 92, 96 and 97, which were also constructed in 1997, are 6-in thick with 5-ft x
6-ft to 10-ft x 12-ft joint spacings. These cells were also sealed and undoweled with the
exception of Cell 92. Cell 92 contained 1-in diameter dowel bars. Cells 60 through 63 were
constructed in 2004. These overlays were 4- to 5-in thick with a 5-ft x 6-ft joint spacing and
no dowels. Some of the sections were sealed while others were left unsealed.

The concrete mixture design, as well as the measured plastic and hardened concrete properties,
are given in Table 5. Fibers were included in the concrete mixture used for constructing the
overlay in 1997 but not in 2004.

Table 4: Summary of the design features for the MnROAD whitetopping cells.
Thickness Thickness Slab Sealed Doweled
Cell Type of fiber
Age of PCC of HMA Type size joint joint
No. reinforcement
slab (in) layer (in) (ft × ft) (Y/N) (Size/N)
Oct 97-
92 6 7 TWT 10 × 12 Y 1 in Polypropylene
Current
Oct 97-
93 4 9 UTW 4 × 4 Y N Polypropylene
Oct 04
Oct 97-
94 3 10 UTW 4 × 4 Y N Polypropylene
Oct 04
Oct 97-
95 3 10 UTW 5 x 6 Y N Polyolefin
Oct 04
Oct 97-
96 6 7 TWT 5 x 6 Y N Polypropylene
Current
Oct 97-
97 6 7 TWT 10 x 12 Y N Polypropylene
Current
Oct 04-
60 5 7 TWT 5 x 6 Y N None
Present
Oct 04-
61 5 7 TWT 5 x 6 N N None
Present
Oct 04-
62 4 8 UTW 5 x 6 Y N None
Present
Oct 04-
63 4 8 UTW 5 x 6 N N None
Present

16
 Table 5: Concrete mixture design information for MnROAD whitetopping cells.
(Vandenbossche and Rettner, 1999; Burnham 2006; Snyder, 2008)
Cell 92-94, 96-97 Cell 95 Cells 60-63
Water to cementitious ratio 0.38 0.41 0.40
Cement (lb/yd3) 650 650 400
3
Class C fly ash (lb/yd ) 0 0 170
3
Fine aggregate (lb/yd ) 1,187 1,287 1,206
CA (1.5 in minus) (lb/yd3) 0 0 1,059
3
CA (3/4 in minus) (lb/yd ) 1,600 1,500 866
CA (3/8 in minus) (lb/yd3) 277 277 0
3
Fiber content (lb/yd ), (percent by
3, 0.059 25, 0.53 0, 0
volume)
Conchem Air Conchem Air
Admixtures (oz/100 lb cement) KB-1000, 34.0
Polyheed N Polyheed N
Measured air (percent) 5.75 7.2 7.6
Measured slump (in) 2.5 2 1.5
28-day compressive strength (psi) 6,100 5,300 4,085

4.1.2 Distress data

The periodic performance evaluation of the MnROAD cells showed that TWT and UTW
sections with different design features experienced different types of distresses. The extent of
the severity of the distresses is also a function of the design variables. The performance and
distress data collected and reported [Vandenbossche and Fagerness (2002), Burnham (2005
and 2008) and Snyder (2008)] as well as recently collected distress data is described below.

The distress surveys conducted for the 1997 cells revealed that Cells 93, 94 and 95, which
included overlays that ranged from 3- to 4-in thick, exhibited cracking by June 1998. Most of
the cracks were corner cracks, and very little transverse (mid-slab) cracks developed until
January 1999, when a couple of reflection cracks developed. It can be seen in Figure 16 that
Cell 94 (3-in thick and 4-ft x 4-ft square panels) exhibited a significant increase in cracking
near the end of 2000, which was not exhibited by the other Cells. Similar increases in cracking
occurred in 2003 for Cells 93 and 95 (Burnham, 2005).

Thirty percent of the slabs in Cell 94 developed cracking, out of which seventy percent were
reflected from the underlying HMA layer. Approximately eight to ten percent of the slabs
from Cells 93 and 95 were cracked with an equivalent amount of cumulative traffic (3.7

17
million ESALs, December 2001). The performance of the other cells (6-in thick, TWTs)
continues to be good except for Cell 97, which has significant transverse joint faulting until
2001. There were 19 panels in Cell 93 (4.2 percent of all slabs), 22 panels in Cell 94 (4.9
percent) and six panels in Cell 95 (2 percent) that were repaired in 2001 due to excessive
deterioration.

In October 2004, Cells 93, 94 and 95 were replaced with Cells 60, 61 and 63, due to the
excessive drop in ride quality (Burnham, 2005). The present serviceability rating (PSR) of
Cells 93, 94 and 95 went down below the terminal serviceability (PSR=2.5) in 2003, 2002 and
2004, respectively. Table 6 shows the distress summary for each of these cells. The thinnest
overlay, Cell 94, exhibited the largest number of cracked panels, (94 percent in the driving lane)
some of which were shattered. By the spring of 2004, 6.1 million ESALs (annual average
daily traffic (AADT) of 26,400 with 14 percent heavy commercial traffic) had been
accumulated on these sections, which is a remarkable feat given the fact that the slabs were
only 3- or 4- in thick.

50 35
45
30
40
Number of Cracks

35 25

Cumulative Panels
30 20

Cracked, %
25
20 15
15 10
10
5
5
0 0
M 00
Se '9 9

D 0
N 99
A ’9 9

M '9 9

1
A '0 0
O 98

Fe 9

Ju 0
Ju 9
Ja 8
N 98

'0
'9

'0
'9

'0
’9

'
'
’
’

b
pt
ly

ly
ec
ay
il

il
ar
ov
n
ne

ct
ov

pr

pr
Ju

Corner Breaks T ransverse Cracks

(a) Cell 93

18
 50 35
45
30
40

Number of Cracks

Cumulative Panels
35 25

Cracked, %
30 20
25
20 15
15 10
10
5
5
0 0

M '00
Se '99

D 00
N 99
A ’9 9

M '99

1
A '00
O 98

Fe 9

Ju 0
Ju 9
Ja 8
N 98

'9

'0
'9

'0
’9

'
'
’
’

b
pt
ly

ly
ec
ay
il

il
ar
ov
n
ne

ct
ov

pr

pr
Ju

Corner Breaks T ransverse Cracks

(b) Cell 94

50 35
45
30
40
Number of Cracks

Cumulative Panels
35 25

Cracked, %
30 20
25
20 15
15 10
10
5
5
0 0

98 9 8 9 8 9 9 '9 9 '99 '9 9 '9 9 '99 '0 0 '0 0 '0 0 '0 0 '0 1
ne ’ ct ’ ov ’ an ’ ril ay u ly ep t ov Feb ar u ly ec ril
Ju O N J Ap M J S N M J D Ap

Corner Breaks T ransverse Cracks

(c) Cell 95
Figure 16: Performance history for (a) Cell 93, (b) Cell 94 and (c) Cell 95 from 1997 to 2001.
(Vandenbossche and Fagerness, 2002)

19
 Table 6: Summary of cracks for MnROAD test sections.
Transverse Longitudinal
Corner cracks Panels cracked (percent)
Age cracks cracks
Cell
(yrs/ESALs) Driving Passing Driving Passing Driving Passing Driving Passing Total
lane lane lane lane lane lane lane lane
93 4/6.4 million 43 6 9 4 0 0 23 1
41 27
94 4/6.4 million 391 84 8 8 0 0 941 341 64
95 4/6.4 million 30 16 5 2 0 0 32 1
16 1
20
11.5/9.8
92 0 0 0 0 3 6 17 35 26
million
11.5/9.8
96 0 0 0 0 1 0 1 0 0
million
11.5/9.8
97 0 0 0 0 7 0 42 0 21
million
4.5/3.8
60 0 0 0 0 3 0 3 0 2
million
4.5/3.8
61 0 0 2 0 5 4 7 5 6
million
4.5/3.8
62 0 0 0 22 0 0 1 1 1
million
4.5/3.8
63 7 1 3 0 8 5 15 8 11
million
1
Panels repaired in 2001 are not included in the calculated percentage.
2
Both cracks propagated off the same reflection crack.

The most recent distress surveys conducted for the MnROAD cells were performed in March
2009. The distresses observed were analyzed with respect to the accumulated traffic loads.
Trends between cracking (longitudinal, transverse and corner), faulting and IRI with the
cumulative ESALs were established. By the end of March 2009, Cells 60 through 63 (4.5
years in service) and Cells 92, 96 and 97 (11.5 years in service) had been subjected to 3.8 and
9.8 million ESALs, respectively. Figure 17 through Figure 19 present the history of the
development of longitudinal, transverse and corner cracking, respectively. A summary of the
distresses observed through March 2009 are presented in Table 6. Among the cells constructed
in 1997, the performance of Cell 96 is still the best with only two percent longitudinal cracks.
Cells 92 and 97 developed longitudinal cracking in 27 and 24 percent of the slabs, respectively.
In Figure 20, it shows that the longitudinal cracks are progressive and merge with each other.
It is surprising to see that these cells have not experienced any transverse or corner cracks even
after 11.5 years of service. Significant transverse joint faulting has occurred in Cell 97
(undoweled). This will be discussed later.

20
 30%
Percent slabs cracked
25%
20%
15%
10%
5%
0%
0 2 4 6 8 10
ESALs (million)
Cell 92: 6 in; 12 ft × 10 ft; D; S Cell 96: 6 in; 6 ft × 5 ft; UD; S
Cell 60: 5 in; 6 ft × 5 ft; UD; S Cell 61: 5 in; 6 ft × 5 ft; UD; US
Cell 62: 4 in; 6 ft × 5 ft; UD; S Cell 63: 4 in; 6 ft × 5 ft; UD; US
Cell 97: 6 in; 12 ft × 10 ft; UD; S
D: Doweled; UD: Undoweled; S: Sealed; US: Unsealed

Figure 17: Longitudinal cracking history for Cells 92, 96 and 97 (11.5 years, 9.8 million
ESALs); 60-63 (4.5 years, 3.8 million ESALs).

4%
Percent slabs cracked

3%

2%

1%

0%
0 2 4 6 8 10
ESALs (million)
Cell 60: 5 in; 6 ft × 5 ft; UD; S Cell 61: 5 in; 6 ft × 5 ft; UD; US
Cell 62: 4 in; 6 ft × 5 ft; UD; S Cell 63: 4 in; 6 ft × 5 ft; UD; US
Cell 92: 6 in; 12 ft × 10 ft; D; S Cell 96: 6 in; 6 ft × 5 ft; UD; S
Cell 97: 6 in; 12 ft × 10 ft; UD; S

Figure 18: Transverse cracking history for Cells 92, 96 and 97 (11.5 years, 9.8 million ESALs);
60-63 (4.5 years, 3.8 million ESALs).

21
 5%
Percent slabs cracked
4%

3%

2%

1%

0%
0 2 4 6 8 10
ESALs (million)
Cell 60: 5 in; 6 ft × 5 ft; UD; S Cell 61: 5 in; 6 ft × 5 ft; UD; US
Cell 62: 4 in; 6 ft × 5 ft; UD; S Cell 63: 4 in; 6 ft × 5 ft; UD; US
Cell 92: 6 in; 12 ft × 10 ft; D; S Cell 96: 6 in; 6 ft × 5 ft; UD; S
Cell 97: 6 in; 12 ft × 10 ft; UD; S
Figure 19: Corner cracking history for Cells 92, 96 and 97 (11.5 years, 9.8 million ESALs); 60-
63 (4.5 years, 3.8 million ESALs).

Cells 60 and 62, which were constructed in 2004 and have a 6-ft x 5-ft joint spacing and are 5-
and 4-in thick, respectively, developed a relatively small amount of longitudinal cracks after
3.8 million ESALs. Both cells are undoweled and unsealed. It can be seen in Figure 21 that
the longitudinal cracks in these cells also went through multiple panels similar to those in Cell
97. The severity of the cracks is still low; however, some spalling has occurred in Cell 60
(Figure 21). It is interesting to note that the longitudinal cracking that developed in Cells 61
and 63 are comparable, indicating that the longitudinal cracking is associated with the panel
size for these TWT sections (Figure 22).

22
 March 2009

Figure 20: Longitudinal cracks in Cell 97.

Longitudinal Cracks

March 2009

(a) Cell 60 (b) Cell 62

Figure 21: Longitudinal cracks in Cells 60 and 62.

23
 Longitudinal Cracks

March 2009

(a) Cell 61 (b) Cell 63

Figure 22: Longitudinal cracks in Cells 61 and 63.

Reflection transverse cracks

March 2009
Figure 23: Reflection transverse cracks in Cell 61.

Cell 60 is performing reasonably well, while Cells 62 and 63 developed transverse cracks at an
early age. Cell 61 developed its first transverse crack after 1.8 million ESALs. Until March
2009 (3.8 million ESALs), two out of 176 slabs (1.4 percent) in Cells 61 and 62 were cracked.
Out of 176 slabs, six (3.4 percent) low to medium severity transverse cracks developed in Cell
63 after the same period of time. In Cell 61, both cracks were reflected from the underlying
asphalt layer, while three out of six cracks in Cell 63 were reflected. This can be seen in
Figure 23 and Figure 24. Cell 63 also experienced one corner break after 2.2 million ESALs
(Figure 19).

24
 Reflection transverse cracks

March 2009

Figure 24: Reflection transverse cracks in Cell 63 at two locations.

The faulting data for the MnROAD cells, as presented in Figure 25, revealed that there is still
no faulting in Cells 60 through 63. Among all of the cells still in place, Cell 97 has the most
severe faulting. Although Cell 97 has been in service on the interstate for 11.5 years, its
magnitude has not yet reached 0.25 in, which is still considered to be low severity.

The IRI history of all the cells is presented in Figure 26. The rideability of Cells 93 through 95
experienced a period of rapid decline in the last two years of their service lives. The IRI for
Cells 60 through 63 is still quite low after five years of service on the interstate. Among
remaining sections constructed in 1997 (Cells 92, 96 and 97), Cell 92 provides the best
rideability. Between Cell 96 and Cell 97, Cell 96 exhibited a lower IRI which is a result of the
lower amount of distress observed in this section. Cell 97 has significant joint faulting.

25
 0.3

Faulting (in) 0.2

0.1

-0.1
0 2 4 6 8 10
ESALs (million)
Cell 60: 5 in; 6 ft × 5 ft; UD; S Cell 61: 5 in; 6 ft × 5 ft; UD; US
Cell 62: 4 in; 6 ft × 5 ft; UD; S Cell 63: 4 in; 6 ft × 5 ft; UD; US
Cell 92: 6 in; 12 ft × 10 ft; D; S Cell 93: 4 in; 4 ft × 4 ft; UD; S
Cell 94: 3 in; 4 ft × 4 ft; UD; S Cell 95: 3 in; 6 ft × 5 ft; UD; S
Cell 96: 6 in; 6 ft × 5 ft; UD; S Cell 97: 6 in; 12 ft × 10 ft; UD; S

Figure 25: Faulting history for MnROAD cells until October 2007.

600
IRI(in/mile)

400

200

0
0 2 4 6 8 10
ESALs (million)

Cell 60: 5 in; 6 ft × 5 ft; UD; S Cell 61: 5 in; 6 ft × 5 ft; UD; US
Cell 62: 4 in; 6 ft × 5 ft; UD; S Cell 63: 4 in; 6 ft × 5 ft; UD; US
Cell 92: 6 in; 12 ft × 10 ft; D; S Cell 93: 4 in; 4 ft × 4 ft; UD; S
Cell 94: 3 in; 4 ft × 4 ft; UD; S Cell 95: 3 in; 6 ft × 5 ft; UD; S
Cell 96: 6 in; 6 ft × 5 ft; UD; S Cell 97: 6 in; 12 ft × 10 ft; UD; S

Figure 26: IRI history for MnROAD cells until March 2008.

26
4.1.3 Factors affecting performance
The distress history of the bonded whitetopping sections at MnROAD verifies that design
features such as joint spacing, overlay thickness, HMA thickness, joint sealing and the use of
dowel bars influence the performance. The following subsections describe the effect of each
parameter.

Joint layout
The amount of cracking and the cracking patterns that develop in each section were directly
influenced by the joint layout and the stresses that developed as a function of the joint layout.
For example, one of the longitudinal joints of a 4-ft x 4-ft joint layout (Cells 93 and 94) was
located at the inside wheel path (Figure 27), which resulted in high edge stresses. This resulted
in corner cracking for both the 3-in and 4-in overlays. Figure 27 shows two distress photos in
Cell 94 in 2001 and 2003.

In Cell 95, the performance of the 3-in overlay with 5-ft x 6-ft panels was significantly better
than the 3-in overlay in Cell 94 with 4-ft x 4-ft panels. For the 5-ft x 6-ft panels, the
longitudinal joint was moved outside of the wheel path and the loads were applied in the
interior portion of the slab (Figure 28).

Cells 92, 97 (10-ft x 12-ft panel size) and 96 (5-ft x 6-ft panel size) have not exhibited any
corner cracks after 11.5 years. Cells 60 through 63 (5-ft x 6-ft panel size) have been in service
for 4.5 years and also show excellent performance in regards to corner cracks. A comparable
performance was obtained with the 3-in overlay with 5-ft x 6-ft panels (Cell 95) and the 4-in
overlay with 4-ft x 4-ft panels (Cell 93). This indicates that an optimum joint layout can
provide an increase in the performance of the overlay equivalent to increasing the thickness of
the overlay by 1 in. The 3-in overlay with 5-ft x 6-ft panels is also more economical than the
4-in overlay with 4-ft x 4-ft panels, because it requires less concrete and fewer joints The
curling stress in the slab increases with an increase in panel size; however, the tensile stress
due to the combination of wheel loads and curling in the slab with a 5-ft x 6-ft joint spacing is
still less than that in a slab with a 4-ft x 4-ft joint spacing. A study conducted by
Vandenbossche and Fagerness (2002) also analytically verifies this fact. In that study the finite

27
element program , ISLAB2000 was used to model two panel sizes, 4-ft × 4-ft and 5-ft × 6-ft,
for the 3-in overlay test sections at MnROAD. The maximum tensile stress generated in each
panel size was within 1 psi (0.01 MPa), indicating that the two panel sizes respond similarly to
temperature gradients of the same magnitude.

4 ft x 4 ft
panels

Dashed Lines Indicate

Location of Wheelpath.

(a) Cell 94 during 2001 (b) Cell 94 during 2003

Figure 27: Corner cracks along the inside wheelpath on Cell 94.
(Vandenbossche, 2001 and Burnham, 2005)

28
 Figure 28: Distress is reduced with the 5-ft x 6-ft panels because the wheelpath is not located
adjacent to the longitudinal joint (Burnham, 2005).

Dowel bars
Among all of the MnROAD whitetopping sections, only Cell 92 has doweled joints. The
faulting history of the cells, as mentioned in Section 4.1.2, shows that the presence of dowel
bars helps to reduce the joint faulting in Cell 92. Cell 92 and Cell 97 have a similar
configuration except that Cell 92 contains 1-in dowel bars. Figure 29 clearly shows the
presence of faulting in Cell 97, whereas no faulting is visible in the adjacent Cell 92. Dowels
are not typically used for TWT or UTW since these overlays are commonly constructed on
lower volume roads. The results show that there is potential to extend the range of applications
for these overlays since they have the potential to perform for extended periods of time (11.5
years), even on the interstate. This increase in performance was observed with only a 1-in
diameter dowel bar. It is interesting to see that the performance of the joints in Cell 92
(doweled, 10 ft x 12 ft) and Cell 96 (undoweled, 5 ft x 6 ft) are comparable. Decreasing the
joint spacing exhibited a similar improvement in joint performance as the addition of 1-in
dowels.

29
 Cell 92, doweled

Cell 97, undoweled

March/2009

Figure 29: Faulting of panels in Cell 97(undoweled) but no faulting in Cell 92 (1-in dowels).

Sealing of joints
All of the UTW and TWT cells, except for Cells 61 and 63, were constructed with joints filled
with hot pour sealant. The larger amount of longitudinal and transverse cracks (Figure 17 and
Figure 18) in Cells 61 and 63 compared to Cells 60 and 62 clearly shows that sealing the joints
reduces the amount of cracking. It may also be noted that Cell 63, which is unsealed,
developed a corner crack at 2.2 million ESALs. Figure 30 is a photo of the sealed and adjacent
unsealed sections while it was raining. It can clearly be seen that the unsealed joints allow the
water to enter the pavement structure while the rain puddles on the surface for the sealed joints.
The infiltration of the water into the pavement structure facilitates the degradation of the
asphalt and hence the bond between the HMA and the overlay. This is not as critical for
thicker PCC overlays; but, a good bond is essential for the long-term performance of these thin
overlays. The shorter panel spacing also increases the area (due to more joints) for potential
rain infiltration within the overlay, which also increases the need to seal the joints to maintain
the integrity of the underlying asphalt. The infiltrated water though the unsealed joints may
also increase the probability of durability cracks (due to freeze/thaw) near the corners of the
panels, although signs of this were not observed at MnROAD.

30
 Sealed

Unsealed

March/2009

Figure 30: Difference in the amount of water infiltrating the surface for the sealed
and unsealed sections.

Thickness of the PCC layer

The performance of Cells 93, 94 and 95 compared to Cells 92, 96 and 97 verifies the fact that
thicker PCC slabs enhance the life of the pavement. The performance of Cells 60 and 61
compared to Cells 62 and 63 also indicates that thicker slabs offer better resistance against
fatigue cracking due to increased stiffness, as would be expected. The thickness of the overlay
also affects the potential for reflection cracking, as will be discussed in depth in a subsequent
section.

Thickness of the HMA layer

The 1997 and 2004 MnROAD sections were constructed on a sound full depth asphalt
pavement (no subbase). After milling, the remaining asphalt thickness ranged from 7 to 10 in.
Therefore, the sensitivity of the HMA layer thickness is not clearly understood from the
performance of these sections. A similarly sound HMA layer was present under each cell. The
only fact which is clearly verified is that a stiffer HMA layer contributes more to the
development of reflection cracks (discussed in Section 4.1.4 Reflection cracking).

31
Seasonal variation in HMA resilient modulus
The resilient modulus of the HMA decreases with an increase in temperature. Therefore, the
asphalt below the concrete provides less support when the temperature is higher. The PCC
overlay must then bear a larger portion of the load, resulting in higher stresses.
Vandenbossche (2005) characterized the relationship between changes in strain and changes in
the resilient modulus. See Figure 31. In that study, strains were measured at the very bottom of
a concrete overlay (3-in with a 5-ft x 6-ft joint layout) and 1 in from the top of the surface,
under a 9-kip FWD load in the wheelpath adjacent to a transverse joint. It can be seen that the
tensile strain at the bottom of the PCC layer increased rapidly when the resilient modulus of
the asphalt fell below 435 ksi (at greater than or equal to 64 °F). The strain in the concrete is
close to zero when the resilient modulus of the HMA is between 435 ksi to 580 ksi (60 °F) and
the entire concrete overlay is in compression when the resilient modulus of the HMA is above
580 ksi. The average monthly temperature in Minnesota is greater than 43 °F for seven months
of the year in Minnesota and greater than 51 F for five months of the year. Therefore, it can be
assumed that the bottom of a UTW overlay will be in tension under an applied load for a larger
period of the year. This illustrates the importance of considering seasonal effects when
determining the design life of ultra-thin whitetopping.

Figure 31: Strain directly under a 9-kip FWD load applied in the wheelpath adjacent to a
transverse joint on a 3-in overlay with 5-ft x 6-ft panels. (Vandenbossche, 2005)

32
4.1.4 Reflection cracking
The investigation of the performance of Cells 93 to 95 shows that many of the transverse
cracks were the result of previously existing temperature cracks in the asphalt reflecting up
through the concrete. Reflection cracking is a function of both uniform-temperature-induced
and load-related stresses. The thermal contraction of the asphalt in the winter creates a stress
concentration at the bottom of the concrete layer in the region near the tip of the crack in the
asphalt. The magnitude of the tensile stress at the bottom of the concrete can be further
increased as a result of vehicle loads. The combination of both uniform-temperature-induced
and load-related stresses can cause the crack in the underlying asphalt to propagate up through
the concrete overlay (Vandenbossche and Fagerness, 2002). As the magnitude of stress at the
bottom of the concrete layer is also a function of the design features of the overlay, it is worth
to reviewing the reflection crack pattern of the existing whitetopping projects to understand the
correlation of the design features.

The reflection cracking history for the 1997 and 2004 MnROAD cells is presented in Table 7.
The distress survey conducted for 1997 MnROAD cells in 2001 revealed that the 6-in overlays
(Cells 92, 96 and 97) did not experience any reflection cracking. Thirty-two percent of the pre-
existing cracks in the asphalt propagated up through the overlay in the 3-in section with 5-ft x
6-ft panels (Cell 95). In the 3-in section with 4-ft x 4-ft panels (Cell 94), fifty-six percent of
the pre-existing cracks reflected through the overlay. Fifty percent of the cracks reflected up
through the concrete in the 4-in overlay with 4-ft x 4-ft panels (Cell 93). All but two of the
reflection cracks developed during the spring and winter. It is possible that these two cracks
also initiated during the winter or spring but were not noticed until the following summer.
Reflection cracking typically occurred earlier in the driving lane than in the passing lane
indicating that the accumulated traffic volume affects the development of reflection cracks. In
the most recent distress survey in March 2009, it was found that 4-in sealed (Cell 62) and 4-in
unsealed sections (Cell 63) developed two and three reflection cracks, respectively. It is
interesting to see that Cell 61, which is unsealed and comparatively thicker (5 in), also
developed two reflection cracks.

33
 Table 7: Summary of transverse reflection cracking for1997 and 2004 MnROAD cells.
HMA
Thickness Transverse
Age Slab size Transverse transverse
Cell of PCC cracks that are
(yrs)/ESALs ft×ft cracks cracks
slab (in) reflective (%)
reflected (%)
93 4/3.7 million 4 4×4 27 19 50
94 4/3.7 million 3 4×4 19 47 56
95 4/3.7 million 3 5x6 4 100 32
11.5/9.8
92 6 10×12 0 0 0
million
11.5/9.8
96 6 5x6 0 0 0
million
11.5/9.8
97 6 10×12 0 0 0
million
60 4.5/3.8 million 5 5x6 0 0 N/A
61 4.5/3.8 million 5 5x6 2 100 N/A
62 4.5/3.8 million 4 5x6 2 100 N/A
63 4.5/3.8 million 5x6 3 100 N/A
4

The panel size and overlay thickness also affect the development of reflection cracks. Among
Cells 92 to 97, the section with the shortest joint spacing and the thinnest overlay (3 in with 4-
ft x 4-ft panel spacing) experienced the highest percentage of reflection cracks while no
reflection cracking occurred in the 6-in overlays. The 4-in overlay with the same panel size, 4
ft x 4 ft, had a slightly lower percentage, but this difference was not statistically significant.
The 3-in section with larger panels, 5 ft x 6 ft, had the lowest percentage of thermal cracks
propagating through the overlay among the three UTW test sections. However, the
performance of Cells 60 to 63 indicates that 4-in and 5-in thick slabs with 5-ft x 6-ft joint
spacing are also vulnerable to reflection cracking when the stiffness of the HMA layer
becomes greater than the stiffness of the PCC layer.

The stiffness of the asphalt as well as the quality of the bond between the concrete overlay and
the asphalt has a significant effect on the performance of the overlay. The stiffness of the
asphalt layer changes with temperature. During the winter, the stiffness of the HMA layer
increases until a threshold value is reached. Temperatures ranging between 100 F and 4 F
have been measured using thermocouples embedded in the middle of the asphalt layer during
construction of the MnROAD sections (Vandenbossche and Fagerness 2002). Cores taken

34
from the MnROAD sections were used to determine the resilient modulus of the asphalt layer.
The resilient modulus of the asphalt layer for the MnROAD pavement sections at various
temperatures can be found in Table 8.

Table 8: Asphalt resilient modulus at different temperatures for MnROAD sections.

Temperature
-5 0 10 20 30 40 50 60 70 80 100
(F)

Resilient
modulus
1.78 1.68 1.40 1.04 .80 .60 .48 .32 .20 .18 .168
(million
psi)

HMA layers and the accumulated heavy traffic loads. The relative stiffness of the PCC and
HMA layers can be determined using the equation given below. Based on the investigation of
the reflection cracks that developed in the MnROAD Cells, it was found that reflection cracks
were a function of the relative stiffness of the concrete.
E PCC  hPCC
3
 1   HMA
2
 (i)
DPCC / HMA   
E HMA  hHMA
3 1  2 
 PCC 

Where,
DPCC / HMA is the relative stiffness of the PCC and HMA layer;

EPCC and EHMA are the elastic modulus of the PCC layer and resilient modulus of the HMA
layer, respectively;
hPCC and hHMA are the thicknesses of the PCC and HMA layers, respectively;

 PCC and  HMA are the Poisson’s ratio of the PCC and HMA layers, respectively.

If DPCC / HMA is equal to 1, then the stiffness of PCC layer is equal to that of the HMA layer.

The reflection cracks are expected to occur when the value of DPCC / HMA is below 1. Using the

above equation DPCC / HMA was calculated for the MnROAD UTW and TWT test sections for a
range of HMA temperatures and is presented in Figure 32 and Figure 33. It is interesting to

35
observe that the DPCC / HMA for Cells 94 (PCC=3 in and HMA = 10 in) and 95 (PCC = 3 in and
HMA = 10 in) is below the critical value of 1 for most of the temperature ranges measured in
the HMA layer for the MnROAD sections, which indicates that the PCC slabs were vulnerable
to reflection cracking for most of the year. Cell 93 (PCC= 4 in and HMA = 9 in) has a
DPCC / HMA value less than 1 at a temperature below 60 °F. The occurrence of a significant
amount of transverse reflection cracks in these cells was discussed in Section 4.1.2. Cells 92,
96 and 97, which are 6-in thick, did not suffer any reflection cracks which makes sense
because the DPCC / HMA was always greater than 1. This theory can be validated further by

considering the recent performance data for Cell 61; the DPCC / HMA value for this cell is below 1
at 5 °F. As of March 2009, two reflection transverse cracks developed in Cell 61. Cells 62
and 63, which were constructed with 4-in PCC slabs over 8-in HMA layers, had a DPCC / HMA that
dropped below one at 50 °F. Two and three reflection cracks have developed in Cells 62 and
63, respectively. The performance of the MnROAD sections also revealed that heavy traffic
loads in the driving lane accelerate the progression of these kinds of cracks. Further
verification of this concept will be provided when discussing the performance of the UTW
constructed at Elk River, Minnesota.

36
 16

12
DPCC/HMA

0
-10 0 10 20 30 40 50 60 70 80
Temperature (0F)

Cell 92: PCC = 6 in; HMA = 7 in Cell 93: PCC = 4 in; HMA = 9 in
Cell 94: PCC = 3 in; HMA = 10 in Cell 95: PCC = 3 in; HMA = 10 in
Cell 97: PCC = 6 in; HMA = 7 in Cell 96: PCC = 6 in; HMA = 7 in
Critical value

Figure 32: Relative stiffness of PCC and HMA layers for Cells 92 through 97.

16

12
DPCC/HMA

0
-10 0 10 20 30 40 50 60 70 80
0
Temperature ( F)

Cell 60: PCC = 5 in; HMA = 7 in Cell 61: PCC = 5 in; HMA = 7 in
Cell 62: PCC = 4 in; HMA = 8 in Cell 63: PCC = 4 in; HMA = 8 in
Critical value

Figure 33: Relative stiffness of PCC and HMA layers for Cells 60 through 63.

37
4.2 Intersections on US-169, Elk River
An ultra-thin whitetopping was constructed in 1997 at three consecutive intersections on US-
169 in Elk River, Minnesota in 1997. These overlays were constructed in the outer southbound
lane of US-169 at the intersections of Jackson (Cell 98), School (Cell 99), and Main Streets
(Cell 91). The layout of the whitetopping cells can be found in Figure 34. The southbound
lane on US-169 is used by many commercial trucks coming from gravel pits, concrete plants
and waste disposal facilities located just north of this intersection. The trucks usually brake
heavily as they approach the first traffic signal at Jackson Street. The speed of the traffic is
significantly reduced by the time it approaches the third traffic signal at the Main Street
intersection. The original pavement was constructed in 1961 on a sandy subgrade and consists
of a 4-in HMA surface on top of a 5-in Class 5 densely graded aggregate base and a 6-in Class
4 aggregate subbase. In 1991, 2-in of HMA were milled off and the pavement was overlaid
with 1.5-in of HMA. The total HMA thickness at the time the UTW was constructed was
established to be 6.25 in, based on a total of ten cores pulled on April 8, 1997 between
roadway posts 159.080 and 160.367.

A distress survey was performed prior to the construction of the overlay. It was found that the
existing HMA layer experienced low to medium severity cracks, severe rutting (1.25 in), and
shoving, especially at the start of the intersection due to the stopping of heavy trucks. FWD
testing was performed on the existing HMA layer prior to the construction of the UTW. The
analysis indicated that the subgrade is very strong, as would be expected since the subgrade
consists of sandy gravel. The base was constructed of Class 5 and Class 4 aggregates
containing a large amount of fine material so the stiffness of this material is most likely not as
high as is expected. The deflection data also indicated the presence of stripping in several
locations. This was verified when forensic cores were pulled from the existing pavement just
prior to construction of the overlay. A 3-in fiber reinforced concrete overlay was placed 788 ft
north of each intersection. The first 12 ft of each test section was increased to 8 in to minimize
the initial impact of the wheel load at the transition. The design features of the UTW and the
concrete mixture design can be found in Table 9. Two different types of concrete mixtures
were used, one containing polypropylene and the other polyolefin fibers (Table 10). The

38
overlays at the first two intersections with the polypropylene fibers consisted of 4-ft x 4-ft
panels while the last intersection with the polyolefin fibers had 6-ft x 6-ft panels.

Table 9: Summary of the design features for the US-169 cells.

(Vandenbossche, 2003)
Thickness Thickness Slab Sealed Doweled
Cell Type of fiber
Life span of PCC of HMA Type size Joint joint
No. reinforcement
slab (in) layer (in) (ft×ft) (Y/N) (Dia./N)
Sept 97- Polypropylene
98 3 3 UTW 4 × 4 Y N
Sept 99 (3 lb/yd3)
Sept 97- Polypropylene
99 3 3 UTW 4 × 4 Y N
Sept 99 (3 lb/yd3)
Polyolefin
Sept 97-
91 3 3 UTW 6×6 Y N (25 lb/yd3)
Sept 99

Table 10: Concrete mixture design for the US-169 cells.

(Vandenbossche, 2003)
Concrete with Concrete with
Category of data polypropylene polyolefin fibers
fibers (Cells 98 and 99) (Cell 91)
Water to cement ratio 0.43 0.37
Cement, (lb/yd3) 450 550
Class C Fly Ash, (lb/yd3) 120 100
Fine aggregate, (lb/yd3) 1,287 1,287
CA (19 mm minus), (lb/yd3) 1,552 1,500
CA (10 mm minus), (lb/yd3) 277 277
Fiber content, (lb/yd3) 3 25
Measured air, (percent) 6 6
Measured slump, (in) 2.25 2

39
 N

Figure 34: Layout of the UTW sections at the intersections of US-169.

(Vandenbossche, 2003)

4.2.1 Analysis of performance

The sections on US-169 were in service between September 1997 and September 1999.
During this period, the sections accumulated approximately 670,000 ESALs. The one-way
AADT was 16,000 in 1997 with eight percent trucks. The AADT grew to 17,000 by 1999.
Forty-nine percent of these trucks are categorized as five-axle semis. The correlation between
the performance and the design features of this UTW project is analyzed in the following
subsections.

40
Increased depth of concrete at the approach
Increasing the concrete thickness of the first 12 ft of each test section to 8 in successfully
prevented any distresses from occurring at each of the test sections as the vehicles transitioned
from the HMA onto the UTW. The most heavily distressed area in each of the test sections
was just prior to the approach of the intersection. The change in vehicle speed is the greatest in
this location as the vehicles accelerate and decelerate when the traffic light changes.

Joint layout
Cracks observed in the ultra-thin whitetopping test sections with a 4-ft x 4-ft joint pattern
included corner breaks and transverse cracks. The corner breaks occurred primarily along the
inside longitudinal joint and the lane/shoulder (L/S) longitudinal joint. Many of the corner
breaks that developed along the inside longitudinal lder0 1 90a3( )-49(61 0 0 1 90.024 504.79 Tm[(ile)4( -14

41
 Corner
Breaks

Transverse
Crack
Figure 35: Transverse crack and corner breaks in the Jackson Street test section (03.30.98).
(Vandenbossche, 2003)

Figure 36: Corner breaks in the inside wheelpath at the Jackson Street test section.
(07.20.99) (Vandenbossche, 2003)

42
Figure 37: Typical distress patterns that developed at the 6 ft x 6 ft Main Street test intersection.
(Vandenbossche, 2003)

PCC and HMA layer thicknesses

The performance of the US-169 sections suggests that insufficient thicknesses of the PCC and
HMA layers lead to failure within a short period after construction. A sound HMA layer under
the 3-in thick PCC layer could have provided better support to reduce the flexural stress in the
PCC layer. The combination of thin HMA and PCC layers was insufficient to provide a
suitable service life.

However, comparing the distress surveys prior to and after the overly was placed revealed that
none of the transverse cracks in the HMA layer reflected into the overlay for any of the test
sections, while reflection cracks did develop in the 3-in and 4-in overlays constructed at
MnROAD. The same joint patterns used on US-169 were also employed on I-94. The
difference in the performance can be attributed to the fact that the UTW on US-169 was placed
on top of 3-in of HMA exhibiting signs of raveling and the UTW at MnROAD was constructed
on 10-in or more of high quality HMA. Therefore, at MnROAD, a higher bond strength and
structural rigidity in the HMA layer was experienced. In turn, higher tensile stresses at the
bottom of the UTW in the regions of the cracks in the HMA were produced As previously
discussed, reflection cracking is a function of the relative stiffness of the PCC and HMA
layers. The relative stiffness of the PCC compared to the HMA layer, represented by DPCC / HMA ,
for the US-169 UTW test sections was found to be more than one over the possible range of

43
`temperatures at the project site (Figure 38). As a result, no reflective cracking was observed
for any of these test sections.

28
24
DPCC/HMA

20
16
12
8
4
0
-10 0 10 20 30 40 50 60 70 80

Temperature (0F)
(Jackson Avenue intersection) (School St. intersection)
(Main St. intersection) Critical value

Figure 38: Relative stiffness of PCC and HMA layers for US-169 UTW test sections.

The number of distressed panels in the Jackson Street test section was approximately twice as
high as the number of distressed panels at the School Street intersection and four times as high
as at the Main Street intersection. The difference between the performances of the School
Street and Jackson Street test sections is somewhat surprising because the overlay design is the
same. There are several possible explanations. First, raveling of the HMA layer at the Jackson
Street intersection was greater than that at the School Street intersection. Second, the speed
limit on US-169 changes from 55 mph to 45 mph just north of Jackson Street and the traffic
signal at Jackson Street is the first in a series of traffic signals with the School Street
intersection following the Jackson Street intersection. The commercial trucks traveling on
southbound of US-169 rapidly reduce their speed as they approach the first traffic signal at
Jackson Street and the speed of the traffic would have been significantly reduced by the second
intersection at School Street. Therefore, it is likely that the dynamic stresses at the School
Street test section are significantly lower than the Jackson Street test section.

44
4.3 LoRay Drive, North Mankato
The LoRay Drive project is located immediately south of the bridge over US-14 in North
Mankato, MN. The pavement was rehabilitated with whitetopping in 1995 as part of a
MnROAD research project. The roadway was rehabilitated because of significant rutting on
both sides of the bridge. The pavement was milled to a depth of 4 to 6 in and concrete was
placed to match the existing grade line. Polypropylene fiber reinforced concrete was used on
the north side of the bridge. On the south side of the bridge, 6-in and 4.5-in nominally thick
slabs were constructed on the southbound and northbound lanes, respectively. A field review
was conducted on May 21, 2007. Significant faulting was found on both sides of the bridge.
The cause of the faulting is unknown due to the lack of data. The presence of severe faulting is
surprising because the PCC slabs are supported by 11-in to 15-in HMA layers. The pumping
action from such a thick asphalt layer was not anticipated. However, the percentage of heavy
commercial trucks on this road is quite significant.

4.4 TH-30, Mankato

Background: TH-30 in southern Minnesota is a low traffic volume road, 35 miles southwest of
Mankato, MN. The project is located in a rural area, with flat to slightly rolling topography.
In 1993, the road was overlaid with thin whitetopping in five different sections using various
overlay design features.

Existing pavement structure: The original road was a gravel road constructed in 1934 over a
silty and clayey subgrade. It was reconstructed in 1955 with a 6-in soil cement treated base
and a 1.5-in HMA wearing surface. In 1973, this road was overlaid with a 2.75-in HMA layer.

Construction details: There were five whitetopping and two HMA overlays constructed over
approximately 11 miles. The sections consist of two 12-ft wide PCC lanes with a 2-ft HMA
shoulder on each side. Transverse joints were skewed with 12-ft joint spacing. The first 10
panels of whitetopping Test Section 3 (TS-3) were doweled to reinforce the transition from the
HMA overlay to the whitetopping test sections. Surface preparation was performed only in
TS-5, where the existing HMA layer was milled. The PCC overlay for of all the other sections
was placed directly on the existing HMA. The HMA thicknesses varied from 5.25 in to 9.75 in.

45
Overlay design features: The design features for these sections are provided in Table 11. All
of the sections but TS-6 were bonded TWT. In TS-6 two coats of curing compound were used
as a bond breaker.

Table 11: Design features for sections at TH-30. (Burnham and Rettner, 2002)
Test Thickness Bonded? Location Dowelled? Special features
section of PCC slab (Y/N) (station) (Y/N)
TS-3 5 in min. Y 121.98 - Y* *¾ in dia. dowels in first 10
(6 in avg), 122.98 panels only
TS-4 5 in min. Y 124.00 - Y ¾ in dia. dowels throughout
(6 in avg) 124.48
TS-5 6 in Y 122.98 N placed over milled hot-mix
- 124.30 asphalt
TS-6 5 in min. N 124.48 N bond breaker: two coats of
(6 in avg) -124.58 curing compound
Control 5 in min. (6 Y 124.58 N N/A
section in avg) - 130.40

Traffic: In 1992, the AADT and the average daily truck traffic (ADTT) were 385 and 90,
respectively. The projected 20-year-design average daily traffic (ADT) and ADTT were taken
as 710 and 110, respectively.

Material testing: The concrete used in these sections had lower strength than specified. The
28-day compressive strength (90th percentile) of the concrete was 2,972 psi compared to the
design value of 3,900 psi. The 28-day flexural strength was 507 psi compared to the design
value of 675 psi.

Performance data: The performance of the test sections through June 2002 was reported by
Burnham and Rettner (2002). The observations from the visual distress survey performed in
2002 are presented in Table 12. Test Sections 3, 5, 6 and the Control Section all have very few
visual distresses. In TS-4, several joints were found to have spalling. The most noticeable
distresses in the whitetopping sections are the longitudinal cracks near the centerline in a small
number of panels in the control section. No considerable joint faulting was observed in the 12-

46
ft x 12-ft panels. The ride quality of the sections measured from 1994 to 2001 is shown in
Figure 39. TS-

47
was 86 °F during the time the FWD testing was performed. It is anticipated that these high
LTEs are a function of the high slab temperatures present at the time of testing.

4.5 Conclusion
The whitetopping projects in Minnesota provided a vast amount of information for a large
range of design scenarios. The Elk River project on US-169 has provided the foundation for
the performance analysis of the UTW on thin HMA. On the other hand, the I-94 project
includes both UTW and TWT on the significantly thicker HMA. The TH-30 project provides
information on the performance of whitetopping in rural road applications. A range of design
variables were evaluated at each location. The performance data from these projects was used
to achieve a better understanding of the cause of different types of distresses in the UTW and
TWT. FWD data and distress data from these projects is valuable information and will be used
in the future tasks of the current project.

5 PENNSYLVANIA
The state of Pennsylvania has several whitetopping projects; but many of them, were placed on
an asphalt overlay which was placed over a pavement that was originally concrete. The
following sections describe the background, existing pavement structure, design features,
concrete mixture design, traffic, and performance data for a couple of more standard
whitetopping placed over an asphalt pavements.

5.1 Intersection of State Route (SR)-133 and SR-100, Chester County

Background: The intersection of SR-133 and SR-100 in District 6 of the Pennsylvania
Department of Transportation (PennDOT) was deteriorated due to rutting. In September 1998,
a whitetopping project was constructed at this intersection.

Existing pavement structure: The underlying layers are 10 in of an ID-3 bituminous concrete
base and 6-in of a densely graded 2A subbase. Detailed specifications in regards to these
materials can be found in the PennDOT Materials Specification Manual, Publication 408
(2003).

48
Overlay design features: A 4-in overlay was placed with 4-ft x 4-ft panels. No dowel bars or
tie bars were used. The joints were not sealed.

Concrete mixture design: The 7- and 28-day compressive strengths for the mixture were 4,770
psi and 5,550 psi, respectively. The PCC mixture design is given in Table 13.

Table 13: Concrete mixture design information.

Cement type Type - I (Lafarge)
Cement content (lb/yd3) 650
Water to cement ratio 0.41
Coarse aggregate content (lb/yd

49
5.3 Conclusion
A complete data set for the projects in Pennsylvania is not available. However, the data for
two projects presented in this report show UTW is constructed for a considerably high volume
of traffic.

6 TEXAS
Two UTW projects were constructed in Odessa, Texas. Both projects were at different
intersections on Loop (LP)-250. The following subsections present the details of both projects.
One project consists of UTW at three different intersections and the other project consists of
two different intersections.

6.1 Intersections on LP-250 at Wadley Road, Holiday Hill Road and Midland Drive,
Midland
Background: Three intersections on LP-250, at Wadley Road, Holiday Hill Road and Midland
Drive in the City of Midland, were rehabilitated with UTW in May 2005. The length of the
project was approximately 300 ft.

Existing pavement structure: The original pavement was an 8-in HMA pavement over a base
course. The bituminous layer and base layer were milled to a depth of 9 in. Then, a 7-in HMA
layer was placed in a single lift just prior to laying the whitetopping.

Overlay design features: The design features for the UTW at each of the three intersections are
similar. The overlay is 3-in thick with 3-ft x 3-ft panels. Milling was performed on the newly
overlaid HMA layer. The HMA layer was milled up to a depth of 3 in before placing the PCC
layer. No dowel or tie bars were used. The longitudinal and transverse joints were not sealed.

Concrete mixture design: The concrete mixture design used to construct the overlay is
provided in Table 14, but there is no information regarding the fine and coarse aggregate
contents as well as the water to cement-ratio.

50
 Table 14: Concrete mixture design information.
Cement type Type III
Cement content(lb/yd3) 658
Coarse aggregate type Grade No.4
Grade 1 with fineness modulus of 2.6 to
Fine aggregate type
3.0
Entrained air (percent) 6 +/- 1.5
Fiber type, dosage(lb/yd3) Polypropylene fiber, 3

Traffic: The ADT of the LP-250 in 2001, 2006 and 2007 was 25,000, 26,650 and 31,180,
respectively with 2.1 percent trucks. The 20-year design ESALs for the same is 2.4 million.

Distress data: The panels in this project were reported to have experienced mid-slab and
corner cracks one to two years after construction. Heavy traffic (>25,000 ADT) and the fact
that the wheelpath was directly adjacent to a longitudinal joint are possible reasons for corner
crack development. The corner crack development observed at this project is comparable to
that observed in Cell 93 and 94 at MnROAD.

Figure 40 and Figure 41 are two pictures of the intersections at Midland Drive and Wadley
Avenue obtained from Google Street View in February of 2009. The figures clearly show that
the inside wheel path is along the longitudinal joints.

Inside wheel
path

Figure 40: UTW at the intersection of LP-250 and Midland Drive at Odessa District.
(Google Street view (February/2009))

51
 Inside wheel path

Figure 41: UTW at the intersection of LP-250 and Wadley Avenue at Odessa District.
(Google Street view (February/2009))

6.2 Intersection of LP-250 at Midkiff Road and Garfield Road, Midland

Background: The UTW at the intersection of LP-250 at Midkiff Road and Garfield Road is 3-
in thick. These sections were constructed in September 2001.

Existing pavement structure: The original pavement structure of this project is similar to the
project described in Section 6.1. The original bituminous layer and base layer were removed
to a final depth of 9 in. A 7-in Type-D dense graded bituminous mix was then placed in two
3.5-in lifts as a preparation for the whitetopping. A PG 64-22 asphalt binder was used in the
HMA layer.

Overlay design features: The thickness of the UTW is 3 inches and the panel size is 3-ft x 3-ft.
No dowel or tie bars were used and the joints were not sealed. The HMA layer was milled to a
depth of 3 in prior to placing the PCC overlay.

Concrete mixture design: High early strength concrete with polypropylene fibers was used to
construct the UTW. Some of the details in regards to the PCC mixture design used as well as
its compressive strength are given in Table 15.

52
 Table 15: Concrete mixture design information
Cement type Type III
Cement content (lb/yd3) 658
Coarse aggregate type Grade No.4
Grade 1 with fineness modulus of 2.6 to
Fine aggregate type
3.0
Entrained air (percent) 6 +/- 1.5
Fiber type, dosage (lb/yd3) Polypropylene fiber, 3
24-hour Compressive strength (psi) 2,400

Traffic: The ADT of LP-250 in 2001, 2005, 2006 and 2007 were 25,000, 32,000, 31,470 and
23,460，respectively. The truck traffic is 2.4 percent. The 20-year design ESALs for the same
is 2.0 million.

Distress data: Detailed distress data is not available. However, it is known that the pavement
suffered from mid-slab and corner cracking six to twelve months after construction.

6.3 Conclusion
The UTW projects in Texas were constructed at intersections that are subjected to heavy traffic
volumes. Longitudinal joints along the wheelpath caused corner and mid-slab cracks two years
after construction. High early strength concrete was used on the project that exhibited distress
six months after construction.

7 MISSOURI
In the state of Missouri, both ultra-thin (3.5 in) and conventional (10 in) whitetopping overlays
were constructed at The Spirit of St. Louis Airport in St. Charles County in 1995. A 3.5-in
UTW was constructed in the light aircraft area, which has an area of 14,000 yd2. Another
15,000-yd2 area of the airport was overlaid with conventional 10-in whitetopping, which is
used for parking heavy aircrafts. Several thin and ultra-thin whitetopping projects have been
constructed in Missouri since this original project. This section presents the performance
histories of six different projects constructed in Missouri.

53
7.1 Intersection of SR-291 and SR-78, Independence
Background: This project was constructed in September 2000 on the severely rutted
intersection of SR-291 and SR-78. All of the approaches along with the turning lanes were
overlaid for a total length of 900 ft.

Existing pavement structure: Initially, the center lanes of the intersection were a concrete
pavement, while the other sections were HMA. The intersection had been overlaid many times
with HMA afterwards. Cores were pulled in May 2009 to establish the thickness of each layer.
Laboratory testing on the cores also provided additional information such as the bond shear
strength (Iowa shear test) and the unit weight of the PCC. A summary of this information is
provided in Table 16.

Table 16: Layer properties of the project at SR-291 and SR-78.

PCC Old Iowa Unit
overlay HMA thickness PCC shear weight of
Route Direction Location
thickness (in) layer strength* PCC
(in) (in) (psi) (lb/ft3)
SR- Thin stripped
NB 1+44 4.8 8.5 43 141
291 layer
SR

54
 Table 17: Concrete mixture design information.
Cement type Type I
Coarse aggregate type ½ in crushed limestone
Fine aggregate type Missouri river sand
3
Fiber type and dosage (lb/yd ) Polypropylene, 3
28-day Compressive strength (psi) > 4,000

Traffic: The ADT in both 2002 and 2008 was 30,000 with 10 percent trucks.

Distress data: Only 131 out of 7,786 panels (1.7 percent) have exhibited distress through June
2009. Laboratory testing of the cores indicated that debonding occurred between the PCC and
HMA layers at several locations. The laboratory testing also revealed that any bond that was
present was not very high.

7.2 US-60, between US-71 and BUS-71, near Neosho

Background: A 4,200-ft long UTW project was constructed on US-60, between US-71 and
Business (BUS)-71 near Neosho in 1999. The original pavement was constructed in 1960 and
it was rehabilitated in 1974. Before the construction of the whitetopping, the HMA pavement
exhibited both longitudinal and transverse cracking.

Existing pavement structure: The original thickness of the HMA layers was 7 in. The HMA
binder is AC-20 grade. The base is a 10-in rolled stone layer that lies on a silty clayey
subgrade. Table 18 summarizes the data obtained from cores pulled in 2009.

55
 Table 18: Layer properties of US-60 project.
PCC overlay HMA Unit weight
Iowa shear
Route Direction Location thickness thickness of PCC
strength (psi)
(in) (in) (lb/ft3)
US-60 EB 18+683 4.5 4.5 112 136
US-60 EB 18+695 4.6 4.5 137 137
US-60 EB 18+707 4.6 4.5 115 138
US-60 EB 18+985 4.9 4.5 64 136
US-60 EB 18+998 4.6 4.5 130 133
US-60 EB 19+911 4.6 4.5 158 135
US-60 WB 19+476 5.0 4.5 157 140

56
7.3 US-169 and SR-YY intersection, St. Joseph
Background: This intersection was originally constructed of concrete pavement in 1931 and it
was widened with HMA in 1961. All of the lanes were rehabilitated with HMA overlays in
1977 and 1989. Finally, an UTW was constructed at this intersection in September 2000 after
the flexible layers had exhibited severe rutting and shoving.

Existing pavement structure: The outer lanes consist of full-depth HMA and an unknown base
layer over a silty-clay subgrade. An AC-20 grade binder was used in the HMA. The center
lanes were originally PCC. Survey and laboratory testing was conducted in May 2009. Table
20 contains the thicknesses of the PCC and HMA layers, Iowa shear strength at the interface,
and the unit weight of the PCC.

Table 20: Layer properties for the US-169 project.

PCC overlay HMA Unit weight
Old PCC Iowa shear
Route Direction Location thickness thickness of PCC
layer (in) strength (psi)
(in) (in) (lb/ft3)
US-169 NB 0+33 3.5 11 N/A Debonded 135
US-169 NB 0+69 4.0 11 N/A Debonded 141
US-169 NB 1+08 4.0 11 N/A 31 133
Thin stripped
SR-YY EB 0+51 4.0 6.5 Debonded 139
layer
SR-YY EB 0+96 4.0 10 N/A Debonded 140
SR-YY EB 1+41 4.0 10 N/A Debonded 137
NB: Northbound, EB: Eastbound

57
 Table 21: Concrete mixture design information.
Cement type Type I
Cement content (lb/yd3) 752
Coarse aggregate type ½-in crushed limestone
Fine aggregate type Missouri river sand
Fiber type, dosage (lb/yd3) Polypropylene, 3
28-day Compressive strength (psi) 5,810

Traffic: The ADT in 2000 and 2008 were 22,000 and 24,000, respectively with an eight
percent truck percentage. The linear traffic growth rate is one percent.

Distress data: In one of the turning lanes, where the thickness of the UTW was approximately
1.25 in, the pavement experienced some cracks resembling the shape of a spider web. Similar
to other UTW projects in Missouri, this project also experienced more corner cracks than mid-
slab cracks. Of the 4,520 slabs, 499 slabs (3 percent) are cracked. No other forms of distress
have been observed. One great concern for this project is the occurrence of debonding
between the layers, which was verified by the Iowa shear tests.

7.4 Other projects in Missouri

In Missouri, there are many other TWT projects. This subsection briefly presents the
information collected on the projects at (i) Missouri (MO)-5, railroad crossing in Lebanon; (ii)
BUS-13 in Branson West and (iii) the Intersection of BUS-13 and MO-14. A complete set of
data for these projects is not available; but, the information that is available for all three
projects is summarized in the tables below. All three projects are comparatively new. Cores
were pulled and the thicknesses of the HMA and the overlays are given in the Table 22 through
Table 24 along with the shear strength at the interface, which was measured using the Iowa
shear test. No visible distresses have been observed after five to six years of service.

58
 Table 22: Layer properties for the project at MO-5, railroad crossing in Lebanon.
HMA Unit weight
PCC overlay Iowa shear
Route Direction Location thickness of PCC
thickness (in) strength (psi)
(in) (lb/ft3)
Left turn
MO-5 0+22 4.5 7 Debonded 139
lane
Left turn
MO-5 0+44 4.5 7 148 139
lane
Left turn
MO-5 1+44 4.8 7 Debonded 139
lane
Left turn
MO-5 1+84 4.4 7 Debonded 136
lane

Table 26: Layer properties for the project at BUS-13 in Branson West.
HMA Unit weight
PCC overlay Iowa shear
Route Direction Location thickness of PCC
thickness (in) strength (psi)
(in) (lb/ft3)
BUS-
NB 0+37 5.6 7 150 141
13
BUS-13 NB 0+88 4.4 7 18 141

59
sections, the bond between the HMA and whitetopping was poor; yet, to date, very little
cracking has occurred. This could indicate that the bond was damaged when the core was
pulled or that the overlays will begin to deteriorate quickly.

8 MISSISSIPPI
Several roads and intersections in Mississippi were rehabilitated with whitetopping. The three
which will be discussed in this report include (i) US-80 at SR-15 in Newton (constructed in
2003), (ii) Intersections of 22nd Avenue and North Frontage Road (constructed in 2008) and
(iii) Intersection of SR-35 and US-80 (constructed in 2008).

8.1 Intersection of SR-15 and US-80, Newton County

Background: The thin whitetopping on SR-15 and US-80 was constructed in Newton County
in 2003. The approaches of SR-15 and US-80 had significant rutting, oxidation, and fatigue
cracking before the construction of the thin whitetopping.

Existing pavement structure: On SR-15, a HMA pavement was constructed in 1977 on a 4-in
asphalt stabilized base with a 6-in granular subbase. The subgrade is classified as A-6 and A-7
with the upper 8 in being lime treated. On US-80, the HMA layer was constructed in 1976
with 6-in of asphalt stabilized base and a 7-in granular subbase. The subgrade is also A-6 and
A-7 with a lime treated upper 8 in. The intersection was overlaid many times with HMA.

Overlay design features: The thickness of the overlay varied between 5 and 6 in with 5-ft x 6-
ft panel size. The HMA layer was milled off up to a depth of 6 in. The existing thickness of
the HMA layers (after milling off) on SR-15 and US-80 were 6.25 in and 7 in, respectively.
No dowel or tie bars were provided and the joints were not sealed.

Distress data: The distress survey conducted in December, 2009 revealed that this project is
performing well without any noticeable distress even after 6 years.

60
8.2 Intersection of 22nd Avenue and North Frontage Road, Lauderdale, MS
Background: This project was constructed in 2008. A considerable amount of rutting was
observed in the asphalt pavement before the construction of the whitetopping.

Existing pavement structure: The southbound section of 22nd Avenue consists of a 3.5-in
HMA layer on top of an 8-in PCC slab and a 10-in granular base. The northbound section of
22nd avenue consists of a 6-in HMA layer on top of an 8-in asphalt treated base and a 10-in
granular subbase. Frontage road consists of a 6-in HMA layer at the northeast approach and a
13-

61
8.3 Intersection of SR-35 and US-80, Scott County
Background: The intersection of SR-35 and US-80 in Scott County has recently been
rehabilitated with thin whitetopping during 2008. The asphalt pavement exhibited
considerable rutting, oxidation, and fatigue cracking before the thin whitetopping was placed.

Existing pavement structure: Details of the existing HMA layers are unknown; however, it is
known that the underlying base is asphalt treated and the subbase layers are of granular
material. The northbound approach of SR-35 has a 7.5-in thick base on top of a 6-in subbase.
The upper 8-in of the subgrade is treated with lime and is classified as either A-6 or A-7. The
southbound approach of SR-35 consists of a 4.5-in base with a 6-in thick subbase. The upper
8-in of the subgrade is treated with lime and is classified as either A-6 or A-7. The base of the
pavement of US-80 is 7-in thick with a 7-in subbase. Again, the upper 8-in of the subgrade is
treated with lime and is classified as either A-6 or A-7.

Overlay design features: The design features of this project are given in the Table 29. The
thickness of the overlay varied from 5 to 6 in. No dowel or tie bars were used and the joints
were not sealed. The HMA layer was milled up to a depth of 6 in prior to the whitetopping.

Distress data: The distress survey conducted in December, 2009 revealed that this project
experienced only a longitudinal crack in one of the lanes. Apart from this, a few longitudinal
joints are found to be spalled (Table 29).

62
 Table 29: Overlay design features and observed distress type for the project at
the Intersection of SR-35 and US-80.
Location of Year of Thickness of Size of Panel
Observed distress
section construction PCC (in) (ft x ft)
MS-35-NB-OSL1 Long. crack
MS-35-NB-ISL2 No distress
MS-35-SB-ISL3 Long. jt. spalling
MS-35-SB-OSL4 2008 Long. jt. spalling
5 to 6 6x6
US-80-EB-OSL1 No distress
US-80-EB-ISL2 Long. jt. Spalling
US-80-WB-

63
the UTW overlay to cover up the spots of exposed subbase. Polypropylene fibers were used in
the concrete mixture.

Traffic: In 2002, the ADT was 12,250 with one percent trucks.

Distress data: The distress data is not available in detail; however, there is information that
corner cracks occurred along the free longitudinal edge of the overlay.

9.2 NY-408 and SH-622, Rochester

Background: An UTW project on NY-408 and SH-622 in Livingston County was constructed
in 2002. Four lanes of NY-408, two lanes of SH-622 and partial area on the on and off ramps
were rehabilitated.

Existing pavement structure: Limited information was received concerning the layer details;
but, it is known that the total thickness of the HMA layer was 13.6 in on average.

Overlay design features: The thickness of the whitetopping is 4 in with 4-ft x 4-ft panels. No
dowel or tie bars were used and the joints were not sealed. The existing HMA surface was
milled up to 4 in. Polypropylene fibers were used in the concrete mixture.

Traffic: In 1997, the ADT was 9,350 with 14 percent trucks. In 2007, the ADT increased by
10,500 with 15 percent trucks.

Distress data: Similar to the project discussed earlier in Section 9.1, this project also
experienced some corner cracks at the free longitudinal edge of the pavement.

9.3 Conclusion
These projects are companion sections and are still performing well seven years after
construction. Detailed distress data is not available. The development of corner cracks at the
free longitudinal edge of the pavement may draw attention for the investigation.

64
10 ILLINOIS
The Illinois Department of Transportation (IDOT) started experimenting with thin and ultra-
thin whitetopping starting in 1974 and 1998, respectively (Roesler et al., 2008). Along with
other design features, concrete mixtures were varied in terms of cement content, mineral
admixture content, fiber content, water to cement ratio and air content in the experimental
studies. Table 30 presents a summary of the nine whitetopping projects in Illinois.

Table 30: Whitetopping project information in Illinois.

(Winkelman, 2005 and Roesler et al., 2008)
Construction Overlay
Project location Route number Project length Overlaid surface
date thickness (in)
Asphalt concrete
Anna IL-146 Intersection June 2001 3
and exposed brick
US-36 and
1/3 PCC and 2/3
Decatur Oakland Intersection April 1998 3.5
AC
Avenue
Tuscola US-36 0.80 miles May 1999 4-7.5 Asphalt concrete
Cumberland September
CH-2 3.54 miles 5.75 Asphalt concrete
County 2001 3.54 miles
September &
Piatt County CH-4 4.94 miles 5 Asphalt concrete
October 2000

65
 Figure 42: Anna project layout (Winkelman, 2005).

Overlay design features: The thickness of the bonded overlay is 3 in. The panel size is 3ft x 3-
ft; however, some joints were slightly shifted to accommodate manholes and the geometrics of
the intersection. The existing HMA surface was milled up to 3 in. Hot-poured sealant was
used on the full-depth relief joints.

Concrete mixture design: Proportions of the materials, properties of the plastic and hardened
concrete are given in Table 31.

66
 Table 31: Concrete mixture design information.
(Winkelman, 2005 and Roesler et al., 2008)
Cement type Type I
3
Cement content (lb/yd ) 755
Water to cement ratio 0.36
Coarse aggregate content (lb/yd3) 1805

67
 Table 33: Distress summary until 2004. (Winkelman, 2005)
Year of survey 2002 2003 2004
Number of panels 1,706 1,706 1,706
Number of panels cracked 201 272 340
Percent panels cracked 11.8 16 19.9
Corner cracks 12.70 percent
Diagonal cracks 0.45 percent
Longitudinal cracks 2.74 percent
Transverse cracks, 3.18 percent

14

12
Percent slabs cracked

10

0
0 0.1 0.2 0.3 0.4 0.5 0.6
ESALs

Corner cracks Diagonal cracks Longitudinal cracks Transvesre cracks

Figure 43: Distresses with respect to ESALs.

Figure 44 shows some of the distressed panels with high severity. It seems that the cracks
developed in the shape of a spider web. A little bit of depression is observed as well. The
distress initiation at this location is unknown; but, it was reported that the underlying layers did
not show any considerable defects. The severe distresses indicate that non-uniform support
conditions may have been the trigger. Additionally, the progression of the debonding between
the layers might have caused the distresses to spread across multiple slabs.

68
 Figure 44: Distressed panels at the intersection. (Winkelman, 2005)

Thinner slabs, which were 3-in thick, might be a concern considering the loads of the heavy
commercial vehicles. It can be concluded that 3 in oT-2997t3.342-T1 0 0 1 350.23 427.63 Tm[( )] TJ

69
Overlay design features: The thickness of the overlay was 3.5 in. Panels vary in size with an
average dimension of 3.6 ft x 4.3 ft. Some relief joints were placed to avoid reflection cracks.
Hot-poured joint sealant was used in those joints.

Construction details: The existing bituminous surface had 3.5 in milled off. This was
followed by brooming and high

70
developed in three panels of the whitetopping sections in 2003. A few corner breaks were
observed as well. The debonding of the layers was investigated with a sounding rod, which
indicated that five percent of the slabs debonded after five years of service. Some of the panels
showed movement and as a result, distresses developed. The whitetopping panels in both the
driving and passing lanes shifted and moved uphill towards the intersection and can be seen in
Figure 45. The sounding rod test revealed that the layers were not debonded at those sections,
which would suggest that the HMA layer had been sheared from the brick paver layer.

Table 36: Distress summary until 2004. (Winkelman, 2005)

Year of survey 1999 2000 2001 2002 2003
Number of panels 181 181 181 181 181
Number of panels cracked 4 14 21 26 34
Percent panels cracked 2.2 7.7 11.6 14.4 18.8

Towards intersection

Figure 45: Movement of the panels toward the intersection

(Winkelman, 2005).
10.3 US-36, Tuscola
Background: This project was constructed on the mainline of US-36 near Tuscola. The
project was completed in May of 1999. The rehabilitation was performed east of Tuscola in
the eastbound and westbound lanes of US-36 to the intersection with I-57.

Existing pavement structure: Several layers were present below the HMA layer at various
locations. These layers consisted of brick pavers, full-depth concrete and a granular

71
embankment. The brick pavers and the full-depth concrete were overlaid with 3-in of HMA,
while the granular embankment was overlaid with a 4.25- in of HMA.

Overlay design features: The thickness of the overlay varied with the profile and grades of the
roadway and ranged from 4 to 7 in. The size of the panels varied with the thickness of the
slabs. The average panel dimension was 5.5 ft x 5 ft. The areas with excessive reflection
transverse cracks were milled out and a new HMA layer was placed at 18 locations.

Concrete mixture design: Concrete mixture design information can be found in Table 37.

Table 37: Concrete mixture design information.

(Winkelman, 2005 and Roesler et al., 2008)
Cement type Type I
Cement content (lb/yd3) 755
Water to cement ratio 0.34
Coarse aggregate content (lb/yd3) 1,704
Fine aggregate content (lb/yd3) 1,035
Mineral admixture type/content None
Water reducer Daracem 65
Air entraining admixture Daravair 1400
Fiber type Polypropylene

Traffic: Traffic in this project is comparatively lower than most of the other projects.
Although the ADT is not huge, there are a considerable amount of trucks (11 to 16 percent)
using this particular section of US-36 due to the presence of a stone quarry nearby. Details of
the traffic data can be found in Table 38.

Table 38: Traffic details for US-36 in Tuscola (Winkelman, 2005 and Roesler et al., 2008).
Year 2000 2001 2002 2003 2004
ADT 5,500 4,900 5,050 5,200 5,350
ADTT (11 to 16 percent
600 800 800 800 850
trucks)
Annual ESALs (million) 0.11 0.15

72
Distress data: Since there were no surface preparations other than patch repairs, debonding
was anticipated; however, only one percent of the slabs debonded at the end of five years.
Table 39 shows that 6.1 percent of the slabs developed cracks at the end of five years of
service. At the end of 2005, approximately 3.5 percent of the slabs had corner cracks; 1.26
percent had diagonal cracks; 0.39 percent had longitudinal cracks; 1.31 percent had transverse
cracks, and 0.58 percent had patching.

Table 39: Distress summary for the project until 2004 (Winkelman, 2005).
Year of survey 2000 2001 2002 2003 2004
Number of panels 4,09 4,809 4,809 4,806 4,805
Number of panels cracked 51 96 155 230 292
Percent panels cracked 1.1 2.0 3.2 4.8 6.1

This project also experienced joint blowup along a full-depth relief joint during the spring of
2003 and 2004. A hypothesis for the cause of this blowup is that the PCC slabs expand during
the excessively hot days in spring while some incompressible materials may have clogged up
the joint providing resistance that led to the blowup.

10.4 Highway-2, Cumberland County

Background: The Cumberland County Highway-2 project consists of a 5.75-in thin
whitetopping overlay. In the fall of 2001, Highway-2 was overlaid with whitetopping between
the towns of Bradbury and Janesville in Southeastern Illinois. The existing pavement had
several transverse cracks and exhibited considerable rutting in the wheel path.

Existing pavement structure: Limited information about layer compositions is available. It is

only known that the thickness of the HMA layer was 6.5 in on top of a 10-in aggregate base.

Overlay design features: The overlay is 5.75-in thick with 5.5-ft x 6-ft panels and skewed
transverse joints. The HMA layer was milled off to a depth of 3 in before the construction of
the TWT.

Concrete mixture design: Concrete mixture design is presented in Table 40.

73
 Table 40: Concrete mixture design. (Winkelman, 2005 and Roesler et al., 2008)
Cement content (lb/yd3) 575
Water to cement ratio 0.34
Coarse aggregate content (lb/yd3) 1,836
3
Fine aggregate content (lb/yd ) 1,256
Mineral admixture type/content None
Retarder Daratard 17
Air entraining admixture Daravair 1400
Fiber type Polypropylene

Traffic: The highway passes through a rural area where the ADT is not much higher. Since
the project is located in an agricultural area, the truck traffic mainly consists of single- and
multiple-unit trucks. Nevertheless, the presence of a local stone quarry contributes quite a bit
towards the ADTT. The traffic details are presented in Table 41.

Table 41: Traffic details for Highway-2.

Year 2002 2003 2004
ADT 2,050 2,050 2,150
ADTT (25 percent) 500 500 500
Annual ESALs (million) 0.07 0.07 0.08
Cumulative ESALs
0.07 0.14 0.22
(million)

74
10.5 Highway-4, Piatt County
Background: The project is along Highway-4 in Piatt County and it was constructed during the
fall of 2000. It lies between the eastern city limits of Monticello and the Champaign County
line east of Central Illinois.

Existing pavement structure: The existing HMA layer is 4-in thick after milling and it is on
top of a cement treated aggregate base.

Overlay design features: This project consists of a 5-in whitetopping with sections of different
skewed joint spacings (5.5-ft or 11-ft square slabs).

Traffic: The section has an ADT of 2,000 (20,000 ESALs) per year.

Distress data: The distress data from 2001 to 2004 is presented in Table 43. At the end of
2004, 0.2 percent of the 5.5-ft x 5.5-ft slabs developed cracks, compared to one percent for the
slabs with an 11-ft x 11-ft joint layout.

Table 43: Distress summary for the project. (Winkelman, 2005)

Year of survey 2001 2002 2003 2004
5-in overlay with 5.5-ft x 5.5-ft panels

Number of panels 1,912 1,912 1,912 1,912

Number of panels
0 2 2 4
cracked
Percent panels
0.0 0.1 0.1 0.2
cracked
5-in overlay with 11-ft x11-ft panels

Number of panels 100 100 100 100

Number of panels
0 0 0 1
cracked
Percent panels
0.0 0.0 0.0 1.0
cracked

75
10.6 US-36 and Country Club Road, Decatur
Background: An UTW was constructed at the intersection of US-36 and Country Club Road
during April and May of 1998. Its surface was severely rutted in both directions with the most
severe rutting in the westbound direction. It should also be noted that there was reflection
transverse cracking and a small amount of longitudinal cracking and block cracking.

Existing pavement structure: The existing HMA pavement consisted of a 1.25-in bituminous
concrete overlay on a pervious bituminous concrete surface. Layer details are not available.

Overlay design features: The thicknesses of the PCC layers in the eastbound and westbound
approaches are 3.5 in and 2.5 in, respectively. Joint spacing varied according to the layout of
the transverse and other cracks of the HMA layer to avoid reflection cracks. In the westbound
lane, the spacing ranges from 2.5 ft to 3.75 ft and in the eastbound lane, it was from 3.5 ft to
5.25 ft. The average panel dimension was 2.95 ft x 3.85 ft in the westbound direction and 3.85
ft x 4.5 ft in the eastbound direction.

Construction details: In both directions, the existing HMA was milled off. The thicknesses
milled off were 2.5 and 3.5 in in the eastbound and westbound directions, respectively. The
milling was followed by brooming and high-pressure water cleaning.

Traffic: Traffic volumes for US-36 and County Club Road are presented in Table 44. Since
2004, the section had accumulated 1.36 million ESALs.

Distress data: The distress data until 2003 (Table 45) for the eastbound and westbound
approaches show that more than 70 percent of the slabs had cracked. More importantly, almost
50

76
 Table 44: Traffic volumes for the US-36 and County Club Road project. (Winkelman, 2005)
Single Multiple Cumulative
Passenger Annual ESALs
Year ADT unit unit ESALs
vehicles (million)
trucks trucks (million)
1999 25,150 650 650 23,850 0.21 0.21
2000 24,850 550 550 23,750 0.18 0.39
2001 24,550 725 725 23,100 0.23 0.62
2002 24,075 750 750 22,575 0.24 0.86
2003 23,600 800 775 22,025 0.25 1.11
2004 23,125 775 775 21,575 0.25 1.36

Table 45: Distress survey results for the project. (Winkelman, 2005)
Year of survey 1999 2000 2001 2002 2003
2.5-in overlay with 2.9 -ft x 3.8-ft panels, eastbound
Number of
810 810 810 798 798
panels
Number of
376 393 461 508 562
panels cracked
Percent panels
46.4 48.5 56.9 63.6 70.4
cracked
3.5-in inlay with 3.8-ft x4.5-ft panels, eastbound
Number of
618 618 618 606 606
panels
Number of
376 393 461 508 562
panels cracked
Percent panels
46.4 48.5 56.9 63.6 70.4
cracked

77
 Figure 46: Blow up of the UTW slabs (Winkelman, 2005).

10.7 Conclusion
The performance review of whitetopping projects in Illinois provides a lot of information about
distress trends with respect to different design variables. Performance data from the IL-146,
Anna project verified that placing whitetopping over brick pavers might not be the best
decision. Many projects were constructed with a higher quantity of cement. The distress
patterns indicate that drying shrinkage played a role in the development of cracks. Occurrence
of the blow up in the US-36 project is quite surprising and contrary to the belief. A large
amount of expansion of small size slabs was not expected. However, the projects with thinner
slabs experienced more corner cracks which accord with the traditional belief that the thin
slabs experience more corner cracks.

11 IOWA
Two whitetopping projects from the state of Iowa are included in this task report. The first one
was constructed on a 1-mile section of County Road, R-16 in the Dallas County in 1991. The
other one was constructed on a 7.2-mile section of Iowa (IA)-21 in Iowa County in 1994.

78
11.1 R-16, Dallas County
Background: The primary objective of this project was to investigate the effectiveness of
different surface preparation techniques adopted to enhance the interface bonding between the
old HMA and the new PCC overlay. The project was constructed in Dallas County, on County
Route R-16, from Dallas Center south 4.5 mile to Ortonville. The overlay was laid on top a of
severely distorted HMA layer. The distresses like rutting, transverse cracks and alligator
cracking were observed in the existing HMA layer (Grove et al., 1993).

Existing pavement structure: The original pavement was built in 1959. It was a 22-ft wide
road and was composed of a 2.5-in HMA layer on a 6-in rolled stone base over 4-in soil base.
In 1971, the pavement was resurfaced by a 3-in HMA layer.

Overlay design features: The project consists of 12 sections. Overlay thicknesses were varied
between 4 to 5 in. The design variables were mainly different HMA surface preparation
techniques, such as, brooming, milling, water/air blasting, cement-water grouting and tack-
emulsion coating.

Concrete mixture design: Two types of mix were used in the project. The average 28-day
compressive strength and flexural strength of concrete were 3,850 and 665 psi, respectively.

Traffic: The ADT in this route ranges from 830 to 1,050.

Interface bonding: Detailed distress data is not available for this project although, interface
bond strength data is available. Grove et al. (1993) analyzed the interface bond strengths for
the sections with respect to the different surface preparation techniques used. To determine the
Iowa shear strength at the interface, cores were pulled out from the sections. Then, a
comparative study was conducted to identify the surface preparation technique that resulted in
the highest bond strengths. A statistical analysis was performed on the Iowa shear strength
data measured for the cores taken from different overlay sections and is presented in Table 46.
The two surface preparation techniques that resulted in higher bond strengths were milling, and
brooming + water/air blasting. The milling + cement & water grouting technique also

79
exhibited higher bond strength. In general, all techniques that included milling resulted in
higher bond strengths tha

80
Overlay design features: The project consists of 65 sections; 35 test sections, 27 transition
sections, and three control sections. The design variables are HMA surface preparation (milled,
patch only, and cold in place recycle (CIPR)), PCC thickness (2, 4, 6, or 8 in), synthetic fiber
reinforcement usage (fiber or no fiber), and joint spacing (2-, 4-, 6-, or, 12-ft square panels).
One of the important features of this project was the heavy instrumentation of the test sections
to obtain the strain and temperature measurement over time.

Traffic: In this project, weigh in motion (WIM) data was collected to calculate the ESALs
based on AASHTO damage factors for 6-in PCC. Table 47 shows the ESALs from 1995 to
1999.

Table 47: ESALs from 1995 through 1999 (Cable et al., 2001).
Year Northbound ESALs Southbound ESALs
1995 2,865 9,328
1996 10,468 14,394
1997 12,337 23,394
1998 28,248 41,379
1999 57,410 42,456
Total 111,328 130,744

Distress data: A limited amount of data was collected because there were not that many
distresses that developed after three and five years. A particular section, which was
constructed with 2-in, 2-ft x 2-ft PCC panels on top of a milled HMA surface, experienced
more longitudinal cracks. A section constructed with 2-in, 4-ft x 4-ft PCC panels on CIPR
surface experienced more corner cracks. Since the wheel path coincides with the longitudinal
joints for the 4-ft x 4-ft panels, this may result in more corner cracks and fractured slabs.
Overall, the distresses were found to be localized and some may have occurred due to
construction issues; therefore, it does not seem possible to conclude that there is a strong
relationship between distresses and design variables.

81
 Interface bonding: An elevated average strains were observed for all PCC thicknesses where
)4wo 0BT1 0 0 1 206.45
patch-oo
70375.67a6C
a6C /op]/Type/Paginada0
/op]/Type/surfte15a
TJET0t146.18a6C
T09.06 7(tte11(/op]/Type/wBTsPaginada0
<<-72161)4eBThniques0BT1
TJET0t169.7a6C
0 0 1 206.45 70526(e)
/op]/Type

82
11.3 Conclusion
Both of these studies reveal the effect of some of the most important overlay design variables
for whitetopping pavements. Whitetopping sections that included milling in the surface
preparation achieved a better interface bonding when compared to the sections that were
broomed without milling. Although the minimum required bond strength is not yet known, a
value of 100 psi can be considered as an indication of a good interface bonding (Grove et al.,
1993 and NHRCP 338, 2004).

12 MICHIGAN
There are many whitetopping projects scattered all over the state, out of which data has been
collected for three projects. The data was obtained through communications with the Michigan
chapter of ACPA. Among the three projects, distress data is only available for the project on
Patterson Avenue. However, all three projects are included in this report with the expectation
that a complete data set will be available in the near future.

12.1 Patterson Avenue, from 44th Street to 36th Street, Kentwood

Background: This whitetopping project is 33,000 yd2. The original pavement was constructed
during the 1960’s followed by multiple HMA overlays until 2006 when the pavement was
suffering rutting at the intersection approaches and some low severity thermal cracks. There
may have been some other isolated distresses, but there was no fatigue cracking. The top layer
was also worn and oxidized. Figure 47 shows the HMA layer condition before the
construction of the Patterson Avenue project.

Existing pavement structure: Several overlays were placed before the whitetopping overlay.
The details of the layer composition can be found in Table 48.

83
 Figure 47: Condition of the original HMA layer at the time of construction.

Table 48: Details of pavement layers at Patterson Avenue, from 44th

Street to 36th Street.
Date of
Layers Thickness (in) Description of layers
construction
Layer 1 4 2006 PCC-UTW
Bituminous mixture No. 1500T,
Layer 2 1.5 1987
20AAA, AC10 binder
Bituminous mixture No. 1300L,
Layer 3 1.5 N/A
20AAA, AC5 binder
Bituminous mixture No. 12B
Layer 4 1.75 N/A
modified, AC10 binder
Bituminous base Mixture No. 500
Layer 5 5.5
modified 20C, AC5 binder
22A and 21A aggregate base
Base 7 to 8 1967 course originally used as gravel
road until 1967
Subbase 18 N/A Probably a Michigan Class II
Subgrade N/A N/A Probably a sandy loam

Overlay design features: The overlay is 4-in thick and the panel size is 4ft x 4ft.
Approximately 4 inches of the existing HMA layers were milled before the construction of the
whitetopping. In the concrete mixture, 1.5 lb/yd3

84
Distress data: There are two distressed areas: the intersection at Patterson Avenue and 36th
Street and the southbound lanes of Patterson Avenue at Danvers Drive. The intersection at
Patterson Avenue and 36th Street was widened in 2006 along with the application of
whitetopping. The underlying asphalt edge did not line up properly with the joint between the
whitetopping and the full-depth widening. The southbound lanes of Patterson Avenue at
Danvers Drive were gapped halfway when the whitetopping was placed. This area coincides
with a manhole. There is some distress around the manhole that is indicative of poor
support/poor compaction of the base/subbase/asphalt around the structure, as well as de-
bonding of the PCC from the asphalt. However, the overall condition of the project is good. A
photograph taken in 2008 as shown in Figure 48 presents the condition of the road.

Courtesy: Steven M. Waalkes, 2008

Figure 48: Condition of the UTW after two years (2008).

12.2 Zeeb Road, Ann Arbor, Washtenaw

Only limited information is available about this project. The original HMA pavement was
constructed 8 years before the construction of a TWT in 2000. The thickness of the overlay is 6
in with 6-ft x 6–ft panels.

85
12.3 Schaefer Highway, Detroit, Wayne County
The existing asphalt pavement was in extremely poor condition. The 3-in whitetopping with 4-
ft x 4-ft panels was constructed during 1996. Fibers were used in the concrete mixture, but
details are not available.

12.4 Conclusion
The performance of the Patterson avenue project is well after three years even with a high
traffic volume. Even though the longitudinal joints coincide with wheel path, no information
about the corner cracks was received.

13 OKLAHOMA
The Oklahoma chapter of ACPA was contacted for data regarding whitetopping projects in the
state of Oklahoma. This section of the report includes three different projects on US-69 in
Oklahoma. The first project is located in Atoka County, north of Stringtown. This was newly
constructed in 2007 and has experienced no distress. The second project on US-69, north of
McAlester is performing excellent even after seven years. The other project is also performing
well after 8 years of construction.

13.1 US-69, North of Stringtown, Atoka County

The 47,850-yd2 project was constructed in 2007. The existing HMA pavement experienced
rutting and fatigue cracking before the construction of whitetopping. The overlay thickness is
5 in with 6-ft x 6-ft and 6-ft x 7-ft joint spacings. The concrete mixture has 564 lb/yd3 cement
content with 20 percent fine aggregates. Polypropylene fiber at 3l lb/yd3 was used in the
mix

86
the time the whitetopping was constructed. The total thickness of the HMA layers was 12 in.
The existing layers had been milled and overlaid with HMA approximately every four years
over the last twelve years. The existing HMA pavement experienced severe rutting and
shoving before the construction of the whitetopping. The TWT overlay thickness of this project
is 6 to 7 in. The inside and outside lanes have 6-ft x 6-ft and 6-ft x 7-ft joint spacings,
respectively. Surface preparation was done with cold milling and sweeping of the existing
layers. Synthetic Fiber (polypropylene) was added at a dose rate of 3 lb/yd3 in the concrete
mixture for both of the approaches. The 3-day compressive strength of the mixture from the
northbound was 3,000 psi and for the southbound it was 5,000 psi. In 2007, the ADT was
28,000 with 30 percent truck traffic. Directional and lane distributions were 50 and 80 percent,
respectively.

The northbound approach is performing well without any issue but the southbound approach
developed a few hairline longitudinal cracks within weeks after construction. These non-
working cracks have not spalled over time.

13.3 Conclusion
The US-69 project north of Stringtown is relatively new and no distress data is available. The
performance of the other project is excellent considering the fact that the ADT is 28,000 (in
2007) with 30 percent truck traffic.

14 OVERALL CONCLUSIONS
Data has been collected from many projects in the states included in this pooled fund study.
Although the data received for some of the projects is incomplete, they are still included in this
report with the expectation that in the near future a complete set of data will be available for
further analysis. A summary of the projects reviewed is provided in Table 49. The review of
the performance data of the existing whitetopping projects has been carried out keeping in
mind the objectives of Task 1. The conclusions drawn for each of the questions presented in
the objectives of Task 1 is summarized below.

87
 Table 49: Summary of projects evaluated.
Overlay
HMA
thickness- Bond shear
State Roadway thickness, Observed distress
panel size strength (psi)
(in)
(in:ftxft)
Corner and trans. cracks; long. jt.
FHWA ALF - Lane 5 5.5 3.25:4x4 N/A
faulting.
FHWA ALF - Lane 7 5.5 3.25:3x3 N/A Cracks in the transition zone.
FHWA ALF - Lane 8 5.5 3.25:3x3 N/A Corner cracks; long. jt. faulting.
Virginia
Corner and long. cracks; long. jt.
FHWA ALF - Lane 9 4.5 4.5:6x6 N/A
faulting.
FHWA ALF - Lane 11 4.5 4.5:4x4 N/A Corner, trans. and long. cracks.
FHWA ALF - Lane 12 4.5 4.5:4x4 N/A Trans. cracks.

I-94 (MnROAD) Cell:92 7 6:10x12 N/A Long. cracks in 26 % of the panels.

Corner cracks in 10 % of the panels;

I-94 (MnROAD) Cell:93 9 4:4x4 N/A Trans. cracks in 3 % of the panels;
19 % of trans. cracks are reflective.

Minnesota Corner cracks in 56 % of the panels;

I-94 (MnROAD) Cell:94 10 3:4x4 N/A Trans. cracks in 4 % of the panels;
47% of trans. cracks are reflective.

Corner cracks in 19 % of the panels;

I-94 (MnROAD) Cell:95 10 3:5x6 N/A Trans. cracks in 3 % of the panels;
100 % of trans. cracks are reflective.

I-94 (MnROAD) Cell:96 7 6:5x6 N/A Long. cracks in 1 % of the panels.

88
 Table 50: Summary of projects evaluated (continued).
Overlay
HMA
State Roadway thickness,
(in)

89
 Table 51: Summary of projects evaluated (continued).
Overlay
HMA
thickness- Bond shear
State Roadway thickness, Observed distress
panel size strength (psi)
(in)
(in:ftxft)
TH-30 (35 miles southwest of
5.25 to 9.75 6:12x12 N/A Very low faulting; low spalling.
Mankato) TS:5
TH-30 (35 miles southwest of
Minnesota 5.25 to 9.75 6:12x12 N/A Very low faulting; low spalling.
Mankato) TS:6
TH-30 (35 miles southwest of
5.25 to 9.75 6:12x12 N/A Long. cracks, very low faulting.
Mankato) Control section
Intersection of SR-133 and
10 4:4x4 N/A N/A
SR-100, Chester County
SR-30, Lancaster County 11.5 3:3x3 N/A N/A
Pennsylvania SR-8031(Segment 0510) and SR-
3.5 2:3x3 N/A N/A
0022(WB ramp to SR-0083)
1 corner crack.
SR-30, Uniontown, Fayette County 8 4:4x4 N/A
Spalling in a few trans. jts.
Intersections on LP-250 at Wadley
Road, Holiday Hill Road and Midland 4 3:3x3 N/A Corner and trans. cracks.
Texas Drive
Intersection of LP-250 at Midkiff Road
4 3:3x3 N/A Corner and trans. cracks.
and Garfield Road
Min.: 43
Max.: 137
Intersection of SR-291 and SR-78,
Missouri 3 to 8 4:4x4 Ave.: 90 1.7 % of slabs cracked.
Independence
Stan. dev.: 66
n=2

90
 Table 52: Summary of projects evaluated (continued).
Overlay
Over
ck si.89 zQe/F1 11.04 Tf1 0 0 1 385.13 515.02 Tm[(O)5(v)11(er)- 70u> B 385.13 515.02 T
HMA
thickness
State Roadway thickness,
(in)

91
 Table 46: Summary of projects evaluated (continued).
Overlay
HMA
thickness- Bond shear
State Roadway thickness, Observed distress
panel size strength (psi)
(in)
(in:ftxft)
SR-15:
Intersection of SR-15 and US-80 6.25; US- 5 to 6:5x6 N/A No distress
80:7

SB and NB
section of
22nd
Avenue:

Mississippi Intersection of 22nd Avenue and North 3.5 and 6; 5 to 6:8,9x6 Long., trans., diagonal and corner
N/A
Frontage Road NE and and 9,10x6 cracks; long. jt. spalling
NW
approach of
Frontage
Road: 6
and 13

Intersection of SR-35 and US-80 N/A 5 to 6:6x6 N/A Long. crack; long. jt. spalling

Intersection of Waldon and Central

N/A 4:4x4 N/A Corner cracks along the lane/shld. jt.
New York Avenues, near Buffalo
NY-408 and SH-622, Rochester 9.6 4:4x4 N/A Corner cracks along the lane/shld. jt.
Corner cracks in 13 % of the panels;
Diagonal cracks in 1 % of the panels;
IL-146, Anna N/A 3:3x3 N/A
Long. cracks in 3 % of the panels;
Illinois Trans. cracks in 3 % of the panels.
Intersection of US-36 and Oakland Corner and trans. cracks in 19%
Avenue, Decatur N/A 3.5:3.6x4.3 N/A
panels.

92
 Table 46: Summary of projects evaluated (continued).
Overlay
HMA
thickness- Bond shear
State Roadway thickness, Observed distress
panel size strength (psi)
(in)
(in:ftxft)
Corner cracks in 3 % of the panels;
Diagonal cracks in 2 % of the panels;
Mainline of US-36, near Tuscola 3 to 4.25 N/A:5.5x5 N/A Long. cracks in 1 % of the panels;
1 % of trans. cracks are reflective;
Patching: 1 %.
Highway-2, Cumberland County 3.5 5.75:5.5x6 N/A Reflection cracks in 1% of panels.
Illinois
5:5.5x5.5 <1 % of slabs cracked in the 5.5-ft x
Highway-4, Piatt County 4 and N/A 5.5-ft jt. layout;1 % of slabs cracked
5:11x11 in 11-ft x 11-ft jt. layout.

93
(i) What is the minimum required HMA thickness and maximum allowable distress level?
The performance of the thin whitetopping section depends upon the support conditions to a
large extent. The performance of the UTW sections (Cells 98, 99 and 91) on US-169 clearly
indicates that the minimum thickness of the HMA layer should not be less than 3 in. The
whitetopping sections at the MnROAD testing facility were constructed over a sound HMA
layer underneath. As a result, these sections did not experience any distresses that usually
initiate because of the thinner supporting HMA layer. Reviewing the design features from all
the other projects, the HMA layer thickness (after milling) was more than 3 in unless a
concrete layer was available underneath. Corner crack development in LP-250 whitetopping
sections in Texas within the first year of construction indicates that a thicker (> 4 in) asphalt
layer is required if the traffic volume is higher.

The life of a whitetopping overlay is not only dependent on the thickness of the HMA layer but
also on the remaining life or severity of the distress of the existing HMA layer. Recall, the
performance of the US-169 test sections. The failure of those sections was caused by
debonding at the interface of the PCC and HMA layers and between lifts of the HMA layers.
It can be concluded that the minimum required thickness of the HMA layer should be decided
based on the anticipated traffic, the severity of the existing distresses, the design life of the
whitetopping, and temperature variation in the region.

(ii) Are there modes of failure other than corner cracks that frequently develop and what
common parameters are present when these additional modes of failure occur?

The review of the performance of the existing TWT and UTW sections provided a better
understanding of distress that develop in these types of overlays. The type of distress that
develops is primarily a function of the thickness of the PCC, while the extent of the
deterioration appears to be related to the thickness and quality of the existing HMA and the
joint layout.

Corner cracks are the primary distress of observed in the UTW projects. For example, corner
cracks in the US-169 UTW projects (3-in thick PCC), LP-250 UTW projects (3-in thick PCC),

94
Cell 93 to 95 of MnROAD test sections (3- to 4-in thick PCC), FHWA ALF test lanes (3.25- to
4.5- thick PCC), Anna project in Illinois (3- in thick PCC), US-36 and Country Club Road
project at Decatur in Illinois (2.5- to 3.5- in thick PCC) all have PCC layer thicknesses less
than 4 in and all exhibited corner cracks. Of the 23 overlays that were less than or equal to 4 in,
corner cracking was the primary mode of distress for all but two of the projects reporting
cracking. These overlays represent joint layouts ranging from 3 ft x 3 ft to 5 ft x 6 ft and HMA
thicknesses ranging from 3 in to 11.5 in. The MnROAD cells help to further differentiate
between the effects of overlay thickness and joint layout. For example, the overlay was 3 in
thick for Cells 94 and 95 and the primary mode of distress for both cells is corner cracking, but
Cell 94 exhibited 34 percent more panels with corner cracks. This is the result of the smaller
panels (4 ft x 4 ft) found in Cell 94, which locates the longitudinal joint in the wheelpath
compared to that in Cell 95 (5 ft x 6 ft). Cell 96 consisted of a 6-in overlay and also had 5 ft x
6 ft panels. This Cell exhibited only longitudinal cracks, again indicating that the type of crack
is a function of the overlay thickness.

Many of these sections also exhibited transverse cracks but these cracks were typically either
reflection cracks or secondary cracks that developed from corner cracks. See Figure 47.
Although, there are instances when transverse cracks developed independently of the presents
of other cracks, as shown in Figure 49.

On the other hand, Cells 60, 62, 92, 96 and 97 from MnROAD did not suffer any corner cracks.
Transverse cracks and longitudinal cracks are the dominant distress for these cells. This
indicates that corner cracking is the common mode of distress only in the case of UTW
sections. For TWT sections, transverse cracks and longitudinal cracks are the main concern.

95
a.) Development from corner cracks. b.) Development independent of other cracking.

Figure 49: Transverse crack development in overlays < 4 in.

Along with corner, m[0( )-39(with )-41(c)4(orne)-2(r, )-36o57oET EMC /P9(e)4(lopm)-3(e)4(nt i)-3(n d(nt

96
(iv) What surface preparation techniques have been used and what level of performance
was achieved? What is the minimum acceptable level of bond?
A review of the construction of these projects has shown that the surface preparation typically
consist of milling a portion of the HMA layer. Typically the depth milled was equivalent to
the depth of the overlay. It is widely accepted that milling is the best procedure to ensure the
best possible bond between the PCC and HMA layers. The experimental study conducted by
Grove et al. (1993) on R-16 and Cable et al. (2001) on IA-21 in Iowa also revealed this. It is
difficult to quantify the minimum acceptable level for the shear strength of the bond although it
is realized that debonding is the main reason that corner cracks initiate in the UTW sections.
The performance of the FHWA ALF sections revealed that even though the bond between the
PCC and HMA layer look like intact, corner cracks can still develop. Therefore, it may be
concluded that the better the bond, the more resistance there is against corner cracks and this
bond must be uniform throughout the bottom of the overlay. A high degree of bonding can be
achieved but this is only beneficial when the bond is present throughout the bottom of the
overlay.

(v) What are the acceptable joint patterns?

It appears that joint layout dictates the performance of whitetopping slabs, especially for UTW.
Any joint spacing that ensures the longitudinal joints do not coincide with the wheel path is
acceptable. Cell 95, which had a slab thickness of 3 in (5-ft x 6-ft panel), exhibited
approximately 75 percent less cracks than the 3-in thick Cell 94 (4-ft x 4-ft panel). The
dominant percentage of corner cracks in the whitetopping projects of Missouri, which have 4-
in thick PCC layers with 4-ft x 4-ft joint layouts, also suggests that this joint configuration
should be avoided. Based on the performance of MnROAD Cells 96 (6-in: 5 ft x 6 ft panels)
and 97 (6-in: 10 ft x 12 ft panels) it can be stated that the 5 ft x 6 ft panel size is not just the
best option for UTW but also TWT.

(vi) Is there evidence from companion test sections that structural fibers help to improve
performance beyond providing additional safety once deterioration begins?
It can be assumed that fibers provide additional flexural strength to the slab. The laboratory
study conducted by Roesler et al. (2008) suggests that the modulus of rupture of the concrete

97
increases with the introduction of structural fibers. However, the same conclusion could not be
drawn in the field due to lack of data from companion projects sites.

(vii) What factors contribute to the development of corner breaks?

Corner breaks occur when the fatigue limit (stress to strength ratio) is exceeded in the concrete.
The stress/strength ratio increases with the number of load applications. This is believed to be
the result of a change in support conditions, which causes permanent deformation of the
support layers (Rasmussen and Rozycki, 2004). The repeated load applications create
permanent deformation to the HMA layer beneath the UTW and the voids at the corner of the
slabs would result in cantilever action that increases the tensile stress at the top of the concrete
surface. Once the fatigue limit is reached, the UTW fails due to corner cracking.

Another reason for corner cracks is the debonding at the concrete and asphalt interface or
between lifts of the asphalt layer. Debonding can initiate due to stripping or raveling of the
asphalt layer due to moisture infiltration. Debonding of the layers results in an increased
tensile stress at the top of the concrete surface; this eventually leads to corner breaks.

(viii) Do fibers help to increase the load transfer efficiency for longer periods of time by
keeping the cracks together?
The investigation of the performance of Cell 95, which was constructed with polyolefin fibers,
indicated that these types of fibers seem to hold the pieces of concrete in place. However, due
to the lack of performance data from other projects, a strong conclusion cannot be drawn. The
laboratory study conducted by Roesler et al. (2008) stated that structural fibers are also able to
bridge the cracks or increase the post crack performance.

Additional test sections need to whi3 TJET Tm[( )]2of e

98
REFERENCES

AASHTO Guide for Design of Pavement Structures. American Association of State Highway
and Transportation Officials, Washington, DC, 1993.

Burnham, T. R. Forensic Investigation Report for MnROAD Ultrathin Whitetopping Test

Cells 93, 94 and 95. Report MN/RC-2005-45, Minnesota Department of Transportation, St.
Paul, MN, 2005.

Burnham, T. R. MnROAD Lessons learned: Thin and Ultra Thin Concrete Overlay. Minnesota
Department of Transportation, St. Paul, MN, 2006. http://www.mrr.dot.state.mn.us/
research/lessons/whitetopping.pdf. Accessed January 2009.

Burnham, T. R. Construction Report for MnROAD Thin Whitetopping Test Cells 60-63.
Report MN/RC -2006-18, Minnesota Department of Transportation, St. Paul, MN, 2006.

Cable, J. K., J. M. Hart, and T. J. Ciha. Thin Bonded Overlay Evaluation. Iowa DOT, Project
HR-559, Iowa Department of Transportation, Ames, IA, March 2001.

Grove, J. D., G. K. Harris, B. J. Skinner. Bond Contribution to Whitetopping Performance on

Low- Volume Roads. Transportation Research Record: Journal of the Transportation Research
Board, No. 1382, Transportation Research Board of the National Academics, Washington, DC,
1993, pp. 104-110.

Pavement Policy Manual. Publication 242-September, 2007 Edition. Commonwealth of

Pennsylvania Department of Transportation Bureau of Maintenance and Operations,
Harrisburg, PA, September, 2007. ftp://ftp.dot.state.pa.us/public/PubsForms/Publications
/PUB%20242.pdf. Accessed February 2009.

Rasmussen, A. O., B. F. McCullough, J. M. Ruiz, J. Mack and J. A. Sherwood. Identification

of Pavement Failure Mechanisms at FHWA Accelerated Loading Facility Ultrathin

99
Whitetopping Project. Transportation Research Record: Journal of the Transportation
Research Board, No. 1816, Transportation Research Board of the National Academics,
Washington, DC, 2002, pp. 148-155.

Rasmussen, R.O. and D.K. Rozycki. Thin and Untra-Thin Whitetopping. NCHRP Synthesis of
Highway Practice 338, National Cooperative Highway Research Program, National Research
Council, Washington, DC, 2004.

R&T update, Concrete Pavement Research and Technology. UTW Scores a Million. Number
4.06, American Concrete Pavement Association, Skokie, IL, June, 2003.
http://www.acpa.org/Downloads/RT/RT4.06.pdf. Accessed February 2009.

Roesler, J., A. Bordelon, A. Ioannides, M. Beyer, and D. Wang. Design and Concrete
Material Requirements for Ultra-Thin Whitetopping. Publication FHWA-ICT-08-016. Illinois
Center for Transportation, IL, June 2008.

Snyder, M. B. Lessons Learned from MnROAD (1992-2006). Final Report prepared for
American Concrete Pavement Association, Skokie, IL, October 2008.

Tarr, S. M., M. J. Sheehan, and P. A. Okamoto. Guidelines for the Thickness Design of
Bonded Whitetopping Pavement in the State of Colorado. Report No. CDOTDTD- R-98-10,
Colorado. Department of Transportation, Denver, CO, December 1998.

Vandenbossche, J. M. The Measured Response of Ultrathin and Thin Whitetopping to

Environmental Loads. Presented at 7th International Conference on Concrete Pavements,
Orlando, FL, September 2001.

Vandenbossche, J. M. Performance Analysis of Ultrathin Whitetopping Intersections on US-

169, Elk River, Minnesota. Transportation Research Record: Journal of the Transportation
Research Board, No. 1853, Transportation Research Board of the National Academics,
Washington, DC, 2003, pp. 18-27.

100
Vandenbossche, J.M. Best Practices for the Design and Repair of Thin and Ultrathin
Whitetopping Based on MnROAD Findings. Proceeding of the International Conference on
Best Practices for Ultra Thin and Thin Whitetoppings”, Federal Highway Administration,
Washington, DC, 2005, pp 242-261.

Vandenbossche, J. M. and A. J. Fagerness. Performance, Analysis, and Repair of Ultrathin

and Thin Whitetopping at Minnesota Road Research Facility. Transportation Research Record:

101