DRAFT
Constructability and Productivity Analysis for Long Life
Asphalt Concrete Pavement Rehabilitation Strategies
Report Prepared for
CALIFORNIA DEPARTMENT OF TRANSPORTATION
By
E. B. Lee, C. W. Ibbs, J. T. Harvey, J. R. Roesler
June 2001
University of California at Berkeley
Institute of Transportation Studies
Pavement Research Center
ii
iii
ACKNOWLEDGEMENTS
The University of California Berkeley research team would like to acknowledge the
information and work contributed by the California Department of Transportation and the
Southern California Asphalt Pavement Association ( SCAPA), especially Mr. Jim St. Martin for
his work in coordinating with industry representatives.
iv
v
TABLE OF CONTENTS
Acknowledgements ........................................................................................................................ iii
Table of Contents ............................................................................................................................ v
List of Figures ............................................................................................................................... ix
List of Tables......................................................................................................................... ...... xiii
Executive Summary ........................................................................................................................ 1
1.0 Introduction ............................................................................................................................. 3
1.1 UCB Previous Research and Future Plan for LLPRS......................................................... 5
1.2 Scope and Objective of Research........................................................................................ 8
1.3 Research Approach ............................................................................................................. 9
2.0 Experiment Design for the AC Constructability Analysis .................................................... 13
2.1 Assumptions.................................................................................................................... . 13
2.2 Hierarchical Structure of the Analysis Options................................................................. 15
2.2.1 Construction Window ............................................................................................... 16
2.2.2 Pavement Design Profiles ......................................................................................... 17
2.3 Rehabilitation Options for CSOL ( Crack Seat and Overlay)............................................ 18
2.3.1 Paving of Shoulders for the CSOL Option................................................................ 20
2.3.2 Layer Profiles for CSOL ........................................................................................... 21
2.3.3 Lane Closure Tactics for CSOL................................................................................ 22
2.4 Rehabilitation Options for Full- Depth AC Replacement.................................................. 29
2.4.1 Layer Profiles for Full- Depth AC Replacement ....................................................... 31
2.4.2 Number of Lanes Rehabilitated During the Weekend Closure................................. 32
2.5 Construction Resource Constraints ................................................................................... 36
2.6 AC Rehabilitation Constructability Analysis Process....................................................... 38
vi
3.0 Cooling Time Simulation...................................................................................................... 41
3.1 Program Inputs and Outputs.............................................................................................. 43
3.2 Experimental Validation of CalCool................................................................................. 44
3.3 Validation of CalCool with Field Data ............................................................................. 47
3.3.1 Temperature Data Collection for CalCool Validation and Calibration in Lompoc,
CA ............................................................................................................................... ... 47
3.3.2 Temperature Data Collection for CalCool Validation and Calibration in San
Leandro, CA .............................................................................................................................. 48
3.3.3 Comparison of CalCool and Field Measurements .................................................... 49
4.0 Results of the AC Constructability Analysis......................................................................... 53
4.1 CSOL Production Capability ............................................................................................ 53
4.1.1 Deterministic Analysis .............................................................................................. 53
4.1.2 Stochastic Analysis ................................................................................................... 58
4.1.3 Production Comparison of the Rehabilitation Options for CSOL ............................ 63
4.2 Full- Depth AC Replacement Production Capability......................................................... 64
4.2.1 Deterministic Analysis .............................................................................................. 64
4.2.2 Stochastic Analysis ................................................................................................... 68
4.2.3 Productivity Comparison of Full- Depth AC Replacement ....................................... 70
5.0 Validation of the AC Constructability Analysis Model ( I- 710 Project) ............................... 73
5.1 Background of the I- 710 Project ....................................................................................... 73
5.2 Predicted Production Capability for the 710 Project......................................................... 78
6.0 Effects of Construction Windows and Comparison of Paving Materials ( Concrete and AC)..
............................................................................................................................... ............... 83
vii
6.1 Effects of Changing Construction Window ...................................................................... 83
6.2 Effect of Paving Shoulders for CSOL............................................................................... 85
6.3 Comparison of Concrete and Asphalt Concrete Rehabilitation ........................................ 86
7.0 Conclusions and Recommendations of the AC Constructability Analysis ........................... 89
7.1 Conclusions ....................................................................................................................... 89
7.2 Recommendations from the AC Constructability Research ............................................. 92
8.0 Glossary and Nomenclature .................................................................................................. 95
8.1 Terms.......................................................................................................................... ...... 95
8.2 Abbreviations .................................................................................................................... 97
9.0 References ............................................................................................................................. 99
viii
ix
LIST OF FIGURES
Figure 1: Overall research structure for the constructability analysis of Caltrans LLPRS............ 4
Figure 2: Research structure for constructability analysis of Caltrans Concrete LLPRS. ............. 6
Figure 3. The input screen of the prototype analysis software for estimating asphalt concrete
constructability. ..................................................................................................................... 11
Figure 4. Typical plan view of one direction of the freeway and lane numbering. ..................... 13
Figure 5. Overlap of longitudinal joints on multi- lift AC paving of adjacent lanes. ................... 15
Figure 6. Hierarchical research structure for study of Caltrans LLACPRS................................. 17
Figure 7. Two layer profiles for CSOL ( Crack Seat and Overlay). ............................................. 19
Figure 8. CSOL lane closure for CSOL Half Closure Full Completion option........................... 25
Figure 9. CSOL lane closure for CSOL Half Closure Partial Completion option. ...................... 28
Figure 10. Layer profile of Full- Depth AC Replacement option................................................. 30
Figure 11. Work plan and lane closures for Full- Depth AC Replacement option. ...................... 34
Figure 12. Typical AC pavement cooling curve for single lift paving. ....................................... 41
Figure 13. CalCool main input window. ...................................................................................... 43
Figure 14. CalCool tabular and graphical output window. .......................................................... 45
Figure 15. Comparison of Webster experimentally observed and CalCool predicted cooling
times. ............................................................................................................................... ..... 46
Figure 16. Cooling curve for a single lift of rich bottom AC placed on granular base ( Lompoc
project). ............................................................................................................................... . 50
Figure 17. Cooling curve for a double lift of AC placed on rich bottom AC layer ( Lompoc
project). ............................................................................................................................... . 50
Figure 18. Cooling curves for a three lift AC layer placed on granular base ( San Leandro
project). ............................................................................................................................... . 51
x
Figure 19. Cooling curve for a three lift AC layer placed on existing PCC ( San Leandro project).
............................................................................................................................... ............... 51
Figure 20. Deterministic analysis of CSOL production in centerline- meters as a function of semi
bottom dump truck cycle time............................................................................................... 55
Figure 21. Deterministic analysis of CSOL production in centerline- meters as a function of
rehabilitation option and number of semi bottom dump trucks/ hour.................................... 55
Figure 22. Deterministic analysis of CSOL production in lane- meters as a function of semi
bottom dump truck cycle time............................................................................................... 57
Figure 23. Deterministic analysis of CSOL production in lane- meters as a function of
rehabilitation option and semi bottom dump trucks/ hour. .................................................... 57
Figure 24. Forecast of production for CSOL from stochastic analysis ( CSOL Half Closure Full
Completion Layer Profile “ A”). ............................................................................................ 59
Figure 25. Resource sensitivity for CSOL stochastic analysis ( CSOL Half Closure Full
Completion Layer Profile “ A”). ............................................................................................ 60
Figure 26. Stochastic analysis of CSOL production in centerline- meters as a function of
rehabilitation option. ............................................................................................................. 61
Figure 27. Stochastic analysis of CSOL production in lane- meters as a function of rehabilitation
option......................................................................................................................... ........... 62
Figure 28. Deterministic analysis of Full- Depth AC Replacement production as a function of
Single- or Double- Lane Rehabilitation, and type and number of trucks per hour. ............... 67
Figure 29. Forecast of production for Full- Depth AC Replacement from stochastic analysis
( Full- Depth Double- Lane Layer Profile “ B”). ...................................................................... 69
xi
Figure 30. Resource sensitivity for Full- Depth AC Replacement stochastic analysis ( Full- Depth
Double- Lane Layer Profile “ B”). .......................................................................................... 69
Figure 31. Stochastic analysis for Full- Depth AC Replacement production, Single- versus
Double- Lane Rehabilitation. ................................................................................................. 71
Figure 32. Site layout of the LLACPRS demonstration project on I- 710.................................... 74
Figure 33. Proposed pavement profiles for I- 710 project. ........................................................... 75
Figure 34. I- 710 rehabilitation stage construction schedule. ....................................................... 76
Figure 35. Schematic of the stage construction for the I- 710 project. ......................................... 77
Figure 36. Asphalt constructability stochastic analysis for the I- 710 project. .............................. 79
Figure 37. Comparison of the effect of different construction windows. .................................... 84
xii
xiii
LIST OF TABLES
Table 1 Major Factors Affecting the AC Rehabilitation Productivity....................................... 16
Table 2 Number and Capacity of Resources Used in the Deterministic Analysis ..................... 37
Table 3 Comparison of Predicted Cooling Time using CalCool and Observed Cooling Time . 46
Table 4 Deterministic Analysis Results for CSOL Production per 55- Hour Weekend Closure,
Four- Lane Rehabilitation. ..................................................................................................... 54
Table 5 Deterministic Analysis Results for CSOL Production, Four- Lane Rehabilitation ....... 56
Table 6 Example of Random Variables for the CSOL Half Closure Full Completion Layer
Profile “ A” Option, Stochastic Analysis ............................................................................... 58
Table 7 Stochastic Analysis Results for CSOL Production, Four- Lane Rehabilitation............. 60
Table 8 Stochastic Analysis Results for CSOL Production, Four- Lane Rehabilitation............. 62
Table 9 Production Comparison for CSOL Rehabilitation ........................................................ 63
Table 10 Deterministic Analysis Results for Production of Full- Depth AC Replacement, Single-
Lane Rehabilitation ............................................................................................................... 65
Table 11 Deterministic Analysis Results for Production of Full- Depth AC Replacement,
Double- Lane Rehabilitation .................................................................................................. 66
Table 12 Deterministic Analysis Results for Full- Depth AC Replacement Production, Single-versus
Double- Lane Rehabilitation....................................................................................... 67
Table 13 Example of Random Variables for the Full- Depth AC Replacement, Double- Lane,
Layer Profile “ B,” Stochastic Analysis ................................................................................. 68
Table 14 Stochastic Analysis Results for Full- Depth AC Replacement Production. .................. 70
Table 15 Production Comparison for Full- Depth AC Replacement, Four- Lane Rehabilitation.. 71
Table 16 Asphalt Constructability Stochastic Analysis for Proposed I- 710 Case Study............. 78
Table 17 Comparison of the Effect of Different Construction Windows .................................... 84
xiv
Table 18 Comparison of Production Capability of CSOL Rehabilitation Option During a 55-
hour Weekend Closure, Effect of Paving Shoulders............................................................. 86
1
EXECUTIVE SUMMARY
A large portion of the highway system in the United States has exceeded its design and its
service life. Deterioration of the existing highway system adversely affects the safety of road
users, ride quality, the operational cost of vehicles, and the cost of highway maintenance. This
report presents the results of a constructability and productivity analysis for the Caltrans Long
Life Asphalt Concrete Pavement Rehabilitation Strategies ( LLACPRS), focusing on optimizing
the maximum production capability within a 55- hour weekend closure.
With the assistance of California asphalt concrete paving contractors, the constructability
analyses explored the effects of the following parameters: rehabilitation materials, design profile
[ Crack Seat and Overlay ( CSOL) and Full- Depth Asphalt Concrete ( AC) replacement of
different thickness], cooling time, number and capacity of construction resources, and alternative
lane closure strategies. The experiment design consisted of a hierarchical structure of
rehabilitation options based on consultation with industry and Caltrans.
Prototype constructability analysis programs running on commercial spreadsheet
software were developed to interactively link all factors involved in the rehabilitation processes.
The analysis programs were designed to help road agencies and paving contractors determine
which rehabilitation and construction strategies were the most feasible in an urban environment
with the underlying goal of balancing the maximization of production capability and
minimization of traffic delay. The asphalt constructability analysis procedure has been
implemented for both deterministic and stochastic analyses.
The asphalt concrete constructability analyses indicate that the proposed objective of
Caltrans to rebuild 6 lane- kilometer of truck lanes within a 55- hour weekend closure has a low
probability of success. Material delivery resources, especially dump trucks for demolition and
delivery trucks for asphalt concrete supply, were the major constraints limiting the production.
2
The total layer thickness for asphalt concrete proved to be a major determining element on the
production capability. For example, the production capability of Full- Depth AC Replacement is
just about 60 percent of CSOL production within a weekend closure for a scenario in which the
two truck lanes need to rehabilitated. However, CSOL requires rehabilitation of all lanes
including shoulders on both sides, thereby limiting its effective productivity. Different
rehabilitation working methods, determined by the construction access, lane closure tactics, and
paving procedures, also have a significant effect on the production capability of the
rehabilitation.
The comparison of different construction windows, ( i. e., a weekend closure versus
continuous closure) was also examined to see the effect of different construction windows on
production capability. Continuous closure/ continuous operation enables the CSOL project to be
finished 15 percent faster and the Full- Depth AC Replacement project to be finished 12 percent
faster compared to weekend- only closures. However, the total duration of the closure for
continuous closure/ daytime operation was longer than that for the weekend- only closure.
This study concludes that efficient lane closure tactics designed to work with the
pavement profile can minimize non- working time, such as the time waiting for the AC to cool,
and increase the production capability of the project. The constructability analysis for AC
developed in this study will aid transportation agencies in their decision- making processes for
prioritizing the number of rehabilitation projects on their backlogs, selecting optimal strategies,
and effectively communicating project duration with the public and other project stakeholders,
such as local governments.
3
1.0 INTRODUCTION
The “ 1995 State of the Pavement Report” indicated that 22,500 lane- km out of 78,000
lane- kilometers in the state highway system required corrective maintenance or rehabilitation,
with 7,000 lane- km needing immediate rehabilitation.( 1) Caltrans has identified 2,800 lane- km
of California urban freeway as candidates for rehabilitation; most of the candidates are in urban
corridors of Southern California and the San Francisco Bay Area. The criteria for long- life
pavement rehabilitation candidate projects are poor structural condition and ride quality and
150,000 ADT ( Average Daily Traffic) or 15,000 Average Daily Truck Traffic.
In order to complete the desired 2,800 lane- km of long- life pavement in ten years,
Caltrans needs to rehabilitate approximately 6 lane- km of pavement every weekend. Initially,
Caltrans developed LLPRS ( Long Life Pavement Rehabilitation Strategies) for rehabilitation of
existing portland cement concrete ( PCC) pavement that met the following objectives: provide
30+ years of service life, require minimal maintenance, and have sufficient production capability
to rehabilitate about 6 lane- km within a weekend construction window of 55 hours. Caltrans
proposed the short construction window of 55 hours per weekend, i. e., 10 p. m. Friday to 5 a. m.
Monday to minimize traffic disruptions during pavement rehabilitation.( 2)
Caltrans LLPRS consists of two sub- categories: LLCPRS ( concrete) and LLACPRS
( asphalt concrete). In this report, PCC pavement rehabilitation with asphalt concrete is referred
to as AC Rehabilitation; PCC pavement rehabilitation with concrete is called Concrete
Rehabilitation.
For both strategies, the assumed existing pavement to be rehabilitated is the same: 200 to
225 mm of plain, jointed PCC; 100 to 150 mm of cement treated base ( CTB); some type and
thickness of aggregate subbase; and the compacted natural subgrade. The AC Rehabilitation
strategies currently included under LLACPRS are: crack, seat, and overlay of the existing
4
pavement; and removal of the concrete pavement structure at least to the aggregate subbase and
replacement with an asphalt concrete structure. The crack, seat, and overlay LLACPRS strategy
has a thicker overlay and different materials from the typical Caltrans crack, seat, and overlay
strategy. Rehabilitation strategies currently included under Concrete Rehabilitation include
removal of the concrete slabs and potentially removal of the CTB and replacement with new
slabs and base ( if required), as shown in Figure 1.
LLPRS
Construction
Window
Continuous
Closure Weekend Closure
Paving Material Concrete Asphalt Concrete
( AC)
Curing or Cooling
Time Curing Time Cooling Time
Design Profile 203- mm
Slab
254- or
205- mm
Slab
CSOL
Full-
Depth
AC
Construction
Analysis
Figure 1: Overall research structure for the constructability analysis of Caltrans LLPRS.
5
1.1 UCB Previous Research and Future Plan for LLPRS
The research described in this report for AC Rehabilitation is a part of the five- stage
study of constructability analysis of LLPRS conducted by the research team at the University of
California at Berkeley ( UCB). According to the Construction Industry Institute ( CII),
“ Constructability is the optimum use of construction knowledge and expertise in planning,
design, procurement, and field operations to achieve overall project objectives.”( 3) Developing
a constructability analysis tool that addresses methodology, processes, and analysis models for
pavement rehabilitation is a challenging task for both transportation agencies and pavement
contractors, as they must consider many input variables and options involved in the rehabilitation
process. Without well- developed tools for pavement rehabilitation process, transportation
agencies are in a difficult situation in their decision- making processes for prioritizing the
backlogged rehabilitation projects, selecting optimal strategies, and effectively communicating
with the public and other project stakeholders. Consequently, the need is growing for a
constructability analysis tool that can assist departments of transportation and pavement
contractors in the implementation of rehabilitation strategies with multiple rehabilitation
alternatives. The construction analysis tool also needs to be integrated with construction and
user- delay costs in order to select the optimal rehabilitation strategy in terms of pavement design,
construction schedule, and minimum inconvenience to the public. Figure 1 shows the basic
structure of the Caltrans LLPRS for both concrete and asphalt concrete materials. The following
list describes the previous UCB LLPRS research work, including future plans:
1. Concrete Constructability Analysis. The first stage of the LLPRS research, the
constructability analysis for LLCPRS ( Concrete Rehabilitation) was completed and
reported to Caltrans.( 4, 5) Figure 2 shows the hierarchical structure of analysis
options for the concrete constructability analysis model.
6
LLCPRS
Construction
Window
Continuous
Closure Weekend Closure
Paving Material Concrete
Curing Time
Design Profile 203- mm Slab 254- or 205- mm
Slab
Construction
Analysis
FSHCC: 4 Hours PCC: 8 or 12
Hours
Working Method Concurrent Sequential
No. of Lanes
Rehabilitated Single- Lane Double- Lane
Figure 2: Research structure for constructability analysis of Caltrans Concrete LLPRS.
2. Case Study for the Concrete Constructability Analysis. As the second stage of the
research, a case study for LLCPRS ( Concrete Rehabilitation) was implemented with a
Caltrans concrete demonstration project on the I- 10 freeway in Pomona, California.
A technical report documenting the research was submitted to published by the
Innovative Pavement Research Foundation ( IPRF) and the Federal Highway
Administration ( FHWA).( 6, 7) The case study played an important role in the
7
validation and calibration of the concrete constructability analysis model developed
by the UCB team.
3. Asphalt Concrete Constructability Analysis. The third stage of the research, the
constructability analysis for LLACPRS ( AC Rehabilitation) is developed and
presented in this report.
4. Case Study for Asphalt Concrete Constructability Analysis. In the fourth stage of
this research, a case study for AC Rehabilitation is underway with a Caltrans AC
demonstration project on the I- 710 freeway ( Long Beach Freeway) for validation and
calibration of the asphalt constructability analysis model. The initial planning of the
I- 710 project is covered in this report with predicted production capability from the
UCB asphalt constructability analysis model. A detailed technical report documenting
the results of this case study will be published separately when the case study is
completed.
5. Knowledge- base Simulation Software for Constructability Analysis. The final
objective of the LLPRS constructability analysis research is professional- level
knowledge- based simulation software to be used as an estimating and analysis tool.
The proposed simulation software will integrate both hydraulic cement concrete and
asphalt concrete models with deterministic and stochastic analysis modules. This
specific research task is sponsored by four state departments of transportation
( California, Minnesota, Texas, and Wisconsin). At the time of this writing,
programming for the software has already begun with a tentative completion date of
March 2002. The simulation software will be used by the road agencies in the
construction planning of pavement rehabilitation projects.
8
1.2 Scope and Objective of Research
This report describes the details of the constructability analysis for Caltrans Long Life
Asphalt Concrete Pavement Rehabilitation Strategies ( LLACPRS), sometimes referred to as AC-Long
Life Strategies, in a similar fashion to the concrete constructability analysis described in
the previous report.( 4) As inputs, the asphalt constructability analysis model used current
asphalt concrete rehabilitation strategies along with typical asphalt concrete construction
processes used in the asphalt paving industry. The desired output from the analysis was the
maximum production capability in terms of lane- km within a short construction window such as
a 55- hour weekend closure. This output was used for comparison of different rehabilitation
strategies, resource constraints, design profiles, and lane closure tactics.
Two different options for AC Rehabilitation were analyzed in terms of design profile:
CSOL ( Crack Seat and Overlay) and Full- Depth Replacement. The analysis model developed in
this research can, with slight modifications, easily be applied to other types of asphalt concrete
rehabilitation. The asphalt constructability analysis procedure has been implemented for both
deterministic and stochastic analyses. In the deterministic constructability analysis, input
parameters involved in the rehabilitation processes, such as resource constraints, are fixed with
representative values. In the stochastic approach, input parameters are treated as random
variables. In addition, a 55- hour weekend closure was compared with two additional
construction windows ( continuous closure with continuous operation and continuous closure
with daytime- only operation) to see the effect of different construction windows on production
capability.
The constructability analysis is limited to the scheduling aspects of pavement
rehabilitation to determine the maximum production capability. The construction scheduling
analysis is a baseline for further consideration of direct construction costs and indirect costs from
9
user delay. Long term pavement performance and then life cycle cost analysis can be evaluated
in the future when the scheduling and cost aspects are integrated.
An initial part of a case study for the LLACPRS on Interstate 710 is included in this
report for the validation of the asphalt constructability analysis model. The predicted maximum
production capability for both CSOL and Full- Depth AC Replacement for the I- 710 project are
presented. The predicted production capability can be used as a guideline for the road agency
and contractor to check their initial rehabilitation scheme and plan. The predicted production
capability from the asphalt analysis model will be compared with the actual performance of the
demonstration project when the project is completed in 2002. The details of the case study are
covered in Section 5.0.
1.3 Research Approach
The asphalt constructability analysis was conducted with processes and methodology
very similar in principle to those used for the concrete constructability analysis,( 4) with some
modifications to accommodate the different characteristics of asphalt materials, such as cooling
time and multi- layer paving.
The basic elements of the constructability analysis, such as construction windows, paving
materials, and design profiles were identified by Caltrans and experienced staff at UCB. These
elements were checked and adjusted through a series of technical meetings with the Southern
California Asphalt Pavement Association ( SCAPA) and Caltrans pavement and material
engineers.( 8– 11) A number of field trips were made to construction sites in Southern California
to gather field data, especially resource constraints, scheduling aspects, and cooling time
information.
10
Based on the information gathered from the industry ( SCAPA), Caltrans, reference
information from the concrete constructability analysis, and a comprehensive literature review, a
hierarchical structure for the analysis options was developed. The structure included a number
of options at each level of analysis. The following options are considered for the asphalt
constructability analysis:
• Design profile
• Layer ( paving lift) profile
• Lane closure tactics
• Completion of paving ( stage construction)
A prototype simulation program linking all parameters interactively in the hierarchical
structure of the analysis options was developed, which is running on commercially available
spreadsheet software ( Microsoft ® Excel). The software was designed to determine the maximum
production capability of the rehabilitation in tables and graphs. An example of the main input
window of the simulation program is shown in Figure 3.
An accurate prediction of the cooling time ( the time to cool the single hot mix asphalt
layer to the required stop temperature) is an essential element in the scheduling of the paving
operation. A cooling time simulation software was used to identify the number of hours required
between paving of lifts of asphalt concrete and opening to traffic of the final lift. The cooling
time analysis software used in the research was validated through a number of field calibration
studies. The details of the cooling simulation program are described along with the validation
results in Section 3.0.
11
Figure 3. The input screen of the prototype analysis software for estimating asphalt
concrete constructability.
12
13
2.0 EXPERIMENT DESIGN FOR THE AC CONSTRUCTABILITY ANALYSIS
This section details the experiment design for the constructability analysis for asphalt
concrete rehabilitation options.
2.1 Assumptions
The following assumptions were made to decrease the number of independent parameters
in the asphalt constructability analysis process:
a. As was used in the Concrete Rehabilitation constructability analysis, the weekend
closure was a 55- hour construction window starting Friday at 10: 00 p. m. and ending
on Monday at 5: 00 a. m.
b. Moveable concrete barrier ( MCB) was used as the safety barrier system between
traffic and the construction zone.
c. The freeway has four lanes in each direction with shoulders. The lane numbering
scheme is shown in Figure 4.
Traffic Flow
Shoulder
2 ( S2)
Shoulder
1 ( S1)
Passenger
Lane 1
( P1)
Passenger
Lane 2
( P2)
Truck
Lane 1
( T1)
Truck
Lane 2
( T2)
Figure 4. Typical plan view of one direction of the freeway and lane numbering.
14
d. For Full- Depth AC Replacement, only the truck lanes ( in most cases two lanes) were
replaced.
e. For Crack Seat and Overlay ( CSOL), one direction of the freeway ( in most cases two
truck lanes and two passenger lanes) including shoulders on both sides was subjected
to crack seat and overlay.
f. The outer shoulder could not be used as a major construction access lane because a
sound wall was adjacent to the shoulder. The shoulder could be used as a main
access lane if the width was greater than 3 meters.
g. Before the paver can begin to place a subsequent lift of asphalt concrete, the current
lift must cool to a maximum temperature of 74° C ( 165° F).
h. The cooling time of each layer for multi- lift paving was estimated by a numeric
cooling time simulation program called CalCool.( 12)
i. Prior to the weekend closure, the existing PCC pavement was pre- cut and ready for
removal for the Full- Depth AC Replacement case. The PCC slab was cracked and
seated prior to the weekend closure for the CSOL case.
j. Daytime and nighttime operations during the weekend closure had the same
productivity, except for the impact of the AC cooling time.
k. Only one paving team was used for the AC paving operation for simplicity.
Consultation with the SCAPA and initial calculations indicated that it would not be
practical to use multiple paving teams working simultaneously because the number of
delivery trucks, the capacity of the AC plant, and construction access were
maximized for a single- paving team. One AC plant was also assumed, due to
conflicts between the delivery trucks, different criteria for material testing from
15
different mixing plants, and the fact that coordination of AC cooling times between
paving crews were major obstacles to manage a multi- plant team. In practice,
multiple crews and plants may be used for some projects.
l. Multiple demolition teams could work simultaneously for Full- Depth AC
Replacement only if enough construction access lanes were provided so that conflicts
between demolition trucks could be minimized. This scenario was possible because
the paving operation was planned to start only after the demolition work was
completed.
m. For interlock between asphalt concrete lifts, longitudinal joints between adjacent
lanes should be offset, as shown in Figure 5.
Cracked and Seated PCC Slab
1st Lift 1st Lift
2nd Lift
3rd Lift
Final Lift
2nd Lift
3rd Lift
Final Lift
Fabric Layer
Figure 5. Overlap of longitudinal joints on multi- lift AC paving of adjacent lanes.
2.2 Hierarchical Structure of the Analysis Options
Through a comprehensive literature review and consultation with Caltrans engineers and
SCAPA, the potential elements most likely to govern the production capability of an AC
Rehabilitation project were identified and summarized, as presented in Table 1. Based on these
elements, an experimental design for the asphalt constructability analysis was schematically
16
developed, as shown in Figure 6. The following sections describe the factorial design that was
developed and give details about each factor level.
Table 1 Major Factors Affecting the AC Rehabilitation Productivity
Factor Options
Construction Window 5C5o- nhtoinuur oWuse eCkleonsdu rCe losure
Paving Material Asphalt Concrete ( AC)
Design Profile FCuSlOl- DL e( pCtrha cAkC, SReeapt laancde mOevnetrlay)
Cooling Time Governed by the layer profile type
Layer Profile Type LLaayyeerr PPrrooffiillee ““ BA””
Lane Closure Type FHualllf CClloossuurree
Full Completion
Affects CSOL only
Paving Completion Type Partial Completion
Layer Profile Type LLaayyeerr PPrrooffiillee ““ BA””
Single- Lane Replacement
Affects Full- Depth AC
Replacement only Number of Lanes Replaced Double- Lane Replacement
2.2.1 Construction Window
Caltrans initially set the weekend closure time of 55 hours to avoid construction delays
and traffic interruptions during weekday hours. The majority of the asphalt analysis was focused
on the weekend closure construction window, although the comparison of different construction
windows, ( i. e., a weekend closure versus continuous closures) is also covered in this report. As
concluded in the concrete analysis, a weekend closure strategy has some disadvantages,
including repeated mobilization/ demobilization and securing of resources on weekends.( 4) The
major advantage of a continuous closure is that working hours
17
LLACPRS
Construction Window Continuous Closure Weekend Closure
Paving Material Asphalt Concrete ( AC)
Cooling Time
Design Profile CSOL Full- Depth AC
Construction Analysis
Cooling Time
Lane Closure Full
Closure
Single-
Lane
Half
Closure
Double-
Lane
Completion Full
Completion
Layer Profile
Layer
Profile
" A"
Layer
Profile
" B"
Layer
Profile
" A"
Layer
Profile
" B"
Partial
Completion
Note: Layer Profiles " A"
and " B" are different
depending on the Design
Profile choice ( CSOL or
Full- Depth AC)
Figure 6. Hierarchical research structure for study of Caltrans LLACPRS.
are maximized without lost time for mobilization/ demobilization, which may or may not reduce
inconvenience to the traveling public.
2.2.2 Pavement Design Profiles
The two design profile options analyzed for rehabilitation of deteriorated PCC with
asphalt concrete were:
18
• Crack Seat and Overlay ( CSOL), and
• Full- depth replacement with asphalt concrete ( Full- Depth AC Replacement)
As the choice of the pavement design profile determines the main components of AC
Rehabilitation, the detailed layer profiles and work plans for each option are fully described
separately in Section 2.3 for CSOL and Section 2.4 for Full- Depth AC Replacement.
2.3 Rehabilitation Options for CSOL ( Crack Seat and Overlay)
Figure 7 shows the proposed pavement profile used for the CSOL option. The AC
overlay is 200 mm ( 8 in.), which is broken down into four lifts of hot mix asphalt concrete. The
existing concrete pavement was assumed to be 200 mm ( 8 in.), which is typical of most Caltrans
rigid pavements. The major advantage of the CSOL option is that it does not require removal of
the existing PCC slab, unlike PCC pavement reconstruction or the Full- Depth AC Replacement
option. Consequently, with CSOL the majority of the working hours during the weekend closure
can be exclusively assigned to the placement of the asphalt concrete overlay. This should result
in more production capability ( lane- km) relative to the other rehabilitation methods.
The disadvantage of the CSOL option from an overall production capability point of view
is that the net centerline- meters of freeway that can be rehabilitated within a single weekend
closure is less than half of the total rehabilitation work that could be completed if only the truck
lanes required rehabilitation. This is because the shoulders ( S1 and S2) and passenger lanes ( P1
and P2) have to be overlaid simultaneously with the two truck lanes ( T1 and T2) ( Refer to Figure
4). This constraint of the CSOL option will significantly reduce its overall production capability
because the other options only require replacement of the truck lanes.
1st Lift
Existing Pavement Layer Profile " A"
Cracked and Seated PCC
203 mm ( 8 in.)
Cement Treated Base ( CTB)
102 mm ( 4 in.)
Aggregate Base ( AB)
305 mm ( 12 in.)
Layer Thickness Cooling
OR
Layer Profile " B"
50 mm 2 hours
1st Lift
3rd Lift
Retained Fabric AC
Subgrade
2nd Lift 75 mm 4 hours
3rd Lift 75 mm 4 hours
Final Lift 25 mm 0.5 hours
2nd Lift 75 mm 4 hours
Final Lift 50 mm 2 hours
Layer Thickness Cooling
Portland Cement Concrete ( PCC)
203 mm ( 8 in.)
Cement Treated Base ( CTB)
102 mm ( 4 in.)
Aggregate Base ( AB)
305 mm ( 12 in.)
Subgrade
Cracked and Seated PCC
203 mm ( 8 in.)
Cement Treated Base ( CTB)
102 mm ( 4 in.)
Aggregate Base ( AB)
305 mm ( 12 in.)
Subgrade
Total Thickness = 230 mm ( 9 in.) Total Thickness = 200 mm ( 8 in.)
55 mm 2 hours 25 mm 0.5 hours
Figure 7. Two layer profiles for CSOL ( Crack Seat and Overlay).
19
20
Another limitation of the CSOL option is that the overlay cannot be placed underneath
bridge overpasses unless there is adequate clearance between the freeway and the bridge to
accommodate the overlay. For pavements under a bridge overpass where adequate clearance
cannot be achieved with the CSOL option, either Full- Depth AC Replacement or concrete slab
removal and replacement( 4- 7) must be used.
2.3.1 Paving of Shoulders for the CSOL Option
The main disadvantage of the CSOL option is that the entire freeway in one direction has
to be overlaid to meet adjacent lane grade criteria including the shoulders. The maximum
allowable height difference between lanes is 50 mm, although differences of less than 25 mm are
desirable. The shoulders outside of lanes P1 and T2 must be overlaid in addition to all of the
traffic lanes ( P1, P2, T1, T2), otherwise the shoulders would be 200 mm below the mainline
highway elevation.
Two options are available for the overlay of the shoulders for CSOL:
• Pre- paving. The shoulders can be overlaid in a series of nighttime closures prior to
the 55- hour weekend closure for the overlay of the main traffic lanes.
• Simultaneous paving. The shoulders can be overlaid at the same time as the main
traffic lanes during the 55- hour weekend closure.
In the case of the pre- paving option, K- rails or Moveable Concrete Barrier ( MCB) should
be installed as a safety barrier between the traffic zone and the shoulders after the shoulder
overlay until the weekend closure for the main traffic lane overlay.
In the case of the simultaneous paving option, the shoulders are paved at the same time as
the rehabilitation of the main traffic lanes and the limited resources and limited accesses are
21
shared among all the paving operations during the weekend closure. Accordingly, the
production capability of this option in terms of centerline- meters will reduced by as much as 40
percent, assuming the width of the shoulder is 3 m and the overlay thickness is the same as the
main traffic lanes.
For more direct comparison of the rehabilitation production capability of CSOL with that
of other rehabilitation methods, the shoulders on both sides are assumed to be paved
simultaneously with the main traffic lanes during the construction window for the CSOL
analysis. More detailed production comparison of pre- paving and simultaneous paving options
are covered at the end of the report ( Section 6.2).
2.3.2 Layer Profiles for CSOL
After cleaning, sweeping, and tacking the concrete pavement, four lifts of hot mix asphalt
will be placed on a cracked and seated existing PCC pavement surface. The following are two
options for the CSOL in terms of the pavement layer profile, as shown in Figure 7:
• CSOL Layer Profile “ A”
• CSOL Layer Profile “ B”
Both layer profiles were selected as spanning a typical range by the UCB Pavement
Research Center ( PRC). The main purpose of comparing the CSOL Layer Profile “ A” with the
CSOL Layer Profile “ B” was to evaluate the impact of different layer profiles as a sensitivity
comparison on the rehabilitation production capability. Actual structural sections must be
designed for each project location.
The cooling hours in the right hand column of each layer profile option in Figure 7 were
calculated from a numerical cooling simulation program, CalCool.( 12) The assumed
22
environmental condition of the pavement before running the cooling time analysis was based on
typical summer weather in the hotter climate regions of California: ambient temperature of 37° C
( 100° F); surface temperature of 43° C ( 110° F); wind speed of 5 kph, paving start time July 1,
10: 00 a. m.; stop temperature 74° C ( 165° F).
In Figure 7, the interface between the first and second AC lift is a fabric helping to
minimize reflective cracking in the AC overlay. The fabric is installed and compacted while the
first AC lift is still hot enough to bond to it.
2.3.3 Lane Closure Tactics for CSOL
Efficient lane closure tactics are the biggest concern for any state department of
transportation ( DOT). The agency needs to balance inconvenience to road users and production
capability of the rehabilitation. Two lane closure tactics were considered for the CSOL analysis:
• CSOL Full Closure
• CSOL Half Closure
2.3.3.1 CSOL Full Closure
In the case of CSOL Full Closure, one direction of the freeway is completely closed for
rehabilitation by switching the traffic to the other side, utilizing counter- flow traffic. All four
lanes of the designated segment of the freeway together with shoulders on both sides ( refer to
Figure 4) will be overlaid completely within the 55- hour weekend closure, lane- by- lane and
layer- by- layer, sequentially.
The sequence of the operations for the CSOL Full Closure option starts with one paving
machine beginning to place the first lift of hot mix asphalt from the far right lane, Truck Lane 2
23
( T2) ( Figure 4). When the paving team completes the first lift of the overlay in lane T2, the
paving team travels back to the starting point to place the first lift of the next lane, Lane 3 ( T1).
This process continues until the leftmost lane, Passenger Lane 1 ( P1), has been paved with its
first lift of AC. As soon as the first lift for all the traffic lanes are completed, the paving team
begins placing the second lift at the start of lane T2. This paving process is repeated until all
four AC lifts have been paved on all four traffic lanes. As mentioned previously, the shoulders
on both sides are assumed to receive the overlay simultaneously with the main traffic lane
overlays ( simultaneous paving, as described in Section 2.3.1).
The temperature of the previously placed lift should be measured before the next lift is
placed to make sure the specified stop temperature is reached. In most cases for the Full Closure
option, there was no waiting time caused by slow cooling of the AC lift, even in the scenario
least conducive to AC cooling ( i. e., hot summer and daytime paving). The main reason for this
is that the sequence of paving the large number of lanes ( typically four) provides adequate
cooling time for a given lane before the paving team is ready to begin the next lift. In addition,
because AC delivery trucks ( semi bottom dump) will use a lane next to the paving lane as the
access rather than drive on the hot lane, the concern about the cooling time for construction
delivery vehicles is eliminated for the Full Closure option.
One of the benefits of the CSOL Full Closure option is that it maximizes paving
production without wasting time for AC lifts to cool enough to receive additional lifts. However,
state DOTs are unlikely to completely close one direction of an urban freeway for rehabilitation
for a 55- hour weekend.
24
2.3.3.2 CSOL Half Closure
Another closure option would be to close two out of four lanes in one direction while
completing the CSOL rehabilitation. This would allow for two lanes to be opened to traffic in
the direction of the rehabilitation and four lanes of traffic open in the opposite direction. The
traffic would be separated from the construction zone by a MCB between Passenger Lane 2 ( P2)
and Truck Lane 1 ( T1), as shown in Figure 8.
The process for the AC overlay construction would be to place the first two lifts in lanes
T1 and T2. Traffic would then be switched to the paved lanes ( T1 and T2), and the rehabilitation
work would move to the remaining two lanes ( P1 and P2). The traffic switch from T1 and T2 to
P1 and P2 is needed either one or two times, depending on the CSOL paving completion option
( discussed subsequently in Sections 2.3.3.2.1 and 2.3.3.2.2).
The primary negative aspect of this option is the delay caused by switching traffic. As
the maximum temperature for allowing traffic on the newly paved lane is typically 50° C, which
is lower than the maximum temperature for placement of the next lift [ typically assumed to be
74° C ( 165° F) in the analysis], additional cooling time is needed before traffic can be allowed on
the hot lanes.
There are two sub- categories for the CSOL Half Closure option for weekend closure
construction:
• CSOL Half Closure Full Completion
• CSOL Half Closure Partial Completion
The CSOL Half Closure Partial Completion option paves two of four AC lifts over the
entire four lanes of traffic in one direction of the freeway while the CSOL Half Closure Full
Completion option finishes all four lifts of AC on all four lanes during the weekend closure.
25
Open to Traffic Construction Paving
Access
Moveable Concrete
Barrier ( MCB)
Traffic Flow
Shoulder
2 ( S2)
Shoulder
1 ( S1)
Passenger
Lane 1
( P1)
Passenger
Lane 2
( P2)
Truck
Lane 1
( T1)
Truck
Lane 2
( T2)
Figure 8a. Plan view of first and final stages.
Cracked and Seated PCC
1st Lift 1st Lift
2nd Lift
3rd Lift
Final Lift
2nd Lift
3rd Lift
Final Lift
3rd
Stage
Final
Stage
1st
Stage
2nd
Stage
Lane P1+ P2 Lane T1+ T2
Fabric Layer
Figure 8b. Paving sequence ( traffic must be switched twice during paving).
Figure 8. CSOL lane closure for CSOL Half Closure Full Completion option.
26
Therefore, the CSOL Half Closure Full Completion option would not finish as many
centerline- km of paving as the CSOL Half Closure Partial Completion option during a given 55-
hour weekend closure.
2.3.3.2.1 CSOL Half Closure Full Completion Option
The main feature of the CSOL Half Closure Full Completion option is that it completes
the four- lift overlay for all four lanes of the segment being rehabilitated during one weekend
closure. Figure 8 shows a schematic of the CSOL Half Closure Full Completion work plan.
Some of the advantages of this option are that two out of four lanes in one direction will
always be open to traffic during the rehabilitation process and that the entire AC overlay
thickness will be completed on all four lanes by the end of the weekend closure. A ramp down
from the height ( 200 mm or 230 mm) of the overlay must be completed at the end of the
weekend closure.
The first stage of this method is to overlay the first two lifts of the two truck lanes ( T1
and T2). While the first lift of the Truck Lane 2 ( T2) is being overlaid, the adjacent lane ( T1)
provides construction access. The first lift on lane T1 is then placed after completion of the first
lift on lane T2. The second lift on lane T2 is then placed followed by the second lift on lane T1.
When the second lift on T1 has cooled to the required temperature ( i. e., 50° C maximum
to allow traffic), the two traffic lanes ( P1 and P2) will be closed and the two partially overlaid
lanes ( T1 and T2) will be opened to traffic. In the second stage, the first two lifts on the two
inner lanes ( P1 and P2) will be placed with same procedure as the first stage. The third and
fourth lift on the two inner lanes ( P1 and P2) will be placed immediately after the second stage is
done, without any traffic switch. Traffic must be then switched again to move to the traffic back
27
to lanes P1 and P2. Finally, the fourth stage of construction completes lifts three and four on
lanes T1 and T2.
Some potential problems with the CSOL Half Closure Full Completion option is that
there is the possibility for wasting time during the paving operation from waiting for the AC to
cool and switching the traffic flow lanes twice. In order to overcome these limitations, one
alternative solution is the CSOL Half Closure Partial Completion.
2.3.3.2.2 CSOL Half Closure Partial Completion Option
The main difference between CSOL Half Closure Partial Completion and CSOL Half
Closure Full Completion is that in the first weekend closure, only the first two AC lifts are
placed on all four lanes. This requires only one traffic switch from lanes T1 and T2 to P1 and P2
during the weekend closure. The remaining two lifts of AC are completed during the second
weekend closure with a similar single traffic switch, as shown in Figure 9.
Open to Traffic Construction Paving
Access
Moveable Concrete
Barrier ( MCB)
Traffic Flow
Shoulder
2 ( S2)
Shoulder
1 ( S1)
Passenger
Lane 1
( P1)
Passenger
Lane 2
( P2)
Truck
Lane 1
( T1)
Truck
Lane 2
( T2)
Figure 9a. Plan view of first and fourth stages.
28
1) First two lifts are paved during the first weekend closure:
2) Last two lifts are paved during the second weekend closure:
1st Stage 2nd Stage Next Week
Previous 3rd Stage 4th Stage
Week
Cracked and Seated PCC
1st Lift ( P1+ P2) 1st Lift ( T1+ T2)
2nd Lift ( P1+ P2) 2nd Lift ( T1+ T2)
Cracked and Seated PCC
1st Lift 1st Lift
2nd Lift
3rd Lift ( P1+ P2)
Final Lift ( P1+ P2)
2nd Lift
3rd Lift ( T1+ T2)
Final Lift ( T1+ T2)
Fabric Layer
Fabric Layer
Figure 9b. Paving sequence ( traffic must be switched once during each of the two weekend
closures.
Figure 9. CSOL lane closure for CSOL Half Closure Partial Completion option.
29
The first stage of this method is to place the first two AC lifts on lanes T1 and T2 and
then to switch traffic from the two inner lanes ( P1 and P2) to the newly overlaid lanes ( T1 and
T2). The second stage paves the first two lifts on lanes P1 and P2. After the first two lifts have
been completed, the two- lane freeway closure is opened to traffic until the following weekend.
During the second weekend closure, the remaining two lifts are placed on the inner lanes ( T1 and
T2). In the final stage, the traffic is switched over to lanes T1 and T2 and the inner lanes ( P1 and
P2) are paved with their final two lifts.
Compared with the CSOL Half Closure Full Completion option, the potential benefit of
the CSOL Half Closure Partial Completion option is to minimize the waiting time for asphalt
concrete cooling and switching traffic compared. However, the concern with this method is the
structural performance of the first two AC lifts under traffic loading for one week.
2.4 Rehabilitation Options for Full- Depth AC Replacement
In the Full- Depth AC Replacement option, the existing PCC truck lanes ( T1 and T2) are
replaced with new asphalt concrete. The old PCC slab and CTB will be demolished and hauled
away, and part of the aggregate base ( AB) will be trimmed to accommodate the required depth of
the new asphalt concrete pavement, as shown in Figure 10. The first lift of asphalt concrete will
be a 76- mm ( 3- in.) rich bottom AC layer placed on the top of the re- compacted AB. Four or five
additional lifts of AC will be paved sequentially depending on the pavement profile selected.
The profile of the existing PCC and new asphalt pavement ( Full- Depth AC Replacement) with
typical AC cooling times during summer weather in California are shown in Figure 10.
Layer Profile " B"
AC
OR
Layer Profile " A"
1st Lift
Existing Pavement
Layer Thickness Cooling
Total Thickness = 330 mm ( 13 in.)
Layer Thickness Cooling
Total Thickness = 406 mm ( 16 in.)
Removed Retained
Aggregate Base ( AB)
279 mm ( 11 in.)
Subgrade
76 mm 1 hour
2nd Lift 76 mm 2 hours
3rd Lift 76 mm 6.5 hours
4th Lift 76 mm 1.5 hours
Final Lift 25 mm 0.5 hours
Aggregate Base ( AB)
203 mm ( 8 in.)
Subgrade
1st Lift 76 mm 1 hour
2nd Lift 76 mm 1.5 hours
3rd Lift 76 mm 6 hours
4th Lift 76 mm 1.5 hours
5th Lift 51 mm 1.5 hours
Removed for Final Lift 51mm 1 hour
Layer Profile " A"
Removed for
Layer Profile " B"
Aggregate Base ( AB)
305 mm ( 12 in.)
Portland Cement Concrete
( PCC)
203 mm ( 8 in.)
Cement Treated Base ( CTB)
102 mm ( 4 in.)
Subgrade
( Note: Assumed paving start time is 1: 00 AM. Longer cooling time may be required for some lifts due to midday paving).
Figure 10. Layer profile of Full- Depth AC Replacement option.
30
31
The disadvantage of the Full- Depth AC Replacement option is that the production
capability of this option within one weekend closure will be the least among other AC
Rehabilitation options. The Full- Depth AC Replacement option is the most work intensive
process, although it may provide the DOT with a better performing rehabilitation scenario
compared to the CSOL options.
The following two sub- options are analyzed for the Full- Depth AC Replacement option
with respect to pavement profile selection as shown in Figure 10:
• Full- Depth Layer Profile “ A”
• Full- Depth Layer Profile “ B”
Both layer profiles were selected as spanning a typical range by the UCB Pavement
Research Center ( PRC) for the purpose of checking the impact of different layer profiles on the
production capability of the Full- Depth AC Replacement. This does not mean that either profile
is more structurally desirable; they are considered only a sensitivity comparison. Actual
structural sections must be designed for each project location.
2.4.1 Layer Profiles for Full- Depth AC Replacement
In the case of the Full- Depth Layer Profile “ A” option, 330 mm ( 13 in.) of new asphalt
concrete will replace the existing PCC slab, CTB, and 25 mm of AB. The profile has five lifts, a
76- mm ( 3- in.) rich bottom AC lift, three 76- mm lifts, and a 25- mm AC surface course ( typically,
open graded asphalt rubber), as shown in Figure 10.
The Full- Depth Layer Profile “ B” option is a total of 406 mm ( 16 in.) of AC, consisting
of six lifts. The six lifts are a 76- mm ( 3- in.) rich bottom AC lift, three 76- mm AC lifts, a 51- mm
32
AC lift, and a 51- mm top lift. The old PCC and CTB will be removed along with the top third
( 102 mm) of the aggregate base.
Similar to the CSOL case, the cooling hours in the right hand column of each layer
profile option shown in Figure 10 were calculated from a cooling simulation program,
CalCool.( 12) The assumed environmental condition of the pavement before running the cooling
time analysis was the same as for CSOL— typical summer weather for a hot climate region in
California: ambient temperature of 37° C ( 100° F); surface temperature of 43° C ( 110° F); wind
speed of 5 kph; paving start time July 1, 1: 00 a. m.; stop temperature 74° C ( 165° F).
For both layer profiles the following two additional sub- options were analyzed to take
into account the number of lanes rehabilitated during a single weekend closure:
• Full- Depth Single- Lane Rehabilitation
• Full- Depth Double- Lane Rehabilitation
2.4.2 Number of Lanes Rehabilitated During the Weekend Closure
Through communications with asphalt concrete paving contractors ( SCAPA), two
alternative lane closure tactics were defined to carry out the Full- Depth AC Replacement option:
• Full- Depth Single- Lane Rehabilitation, as shown Figures 11a and b, and
• Full- Depth Double- Lane Rehabilitation, as shown in Figure 11c.
In the Full- Depth Double- Lane Rehabilitation scheme, the two truck lanes ( T1 and T2)
are demolished and rebuilt during one weekend closure, while in the Full- Depth Single- Lane
Rehabilitation, only one truck lane is rehabilitated during the first weekend closure and the other
truck lane is completed during the second weekend closure. The single- and double- lane
rehabilitation concept for AC Rehabilitation is similar to the lane closure tactics for Concrete
33
Rehabilitation described in Reference ( 4). Note that the double- lane rehabilitation option for
Full- Depth AC Replacement does not specify paving both lanes simultaneously.
Of the two working methods used for concrete rehabilitation, only the sequential method
is applicable for the Full- Depth AC Replacement option. In the sequential method, the paving
operation starts only when demolition of the existing PCC pavement is finished.
The concurrent working method, in which paving and demolition activities are
progressing simultaneously, is not practical for the Full- Depth AC Replacement option because
placement of one AC lift ( especially the first lift) only requires several hours, as shown in Figure
11. Consequently, the demolition team working in front of the pavement team would easily be
caught by the paving operation if a concurrent working method were employed.
2.4.2.1 Work Plan for Full- Depth Single- Lane Rehabilitation
During the first weekend closure, two truck lanes ( T1 and T2) will be closed to rebuild
Truck Lane 2 ( T2). Truck Lane 1 ( T1) is used as the construction access for demolition and
paving activities, as shown in Figure 11a. On the following weekend, T1 will be rebuilt with T2
serving as the construction access lane.
The use of one demolition team was assumed because only one construction access lane
is available. In theory, multiple demolition teams can work simultaneously ahead of the first
demolition team if they are properly spaced.
In the case of multiple demolition teams with one access lane, the demolition trucks from
different crews will probably interact negatively if there are not multiple entrances and exits to
the construction site— this is supported by observations made during the I- 10 project for
Concrete Rehabilitation.( 6) In a scenario without multiple entrances and exits, the average cycle
( a) Single- lane ( T1: 1st weekend)
S1 P1 P2 T1 T2 S2
Traffic Flow
Open Access Paving
S1 P1 P2 T1 T2 S2
Open Access Paving
( c) Double- Lane ( T1+ T2)
S1 P1 P2 T1 T2 S2
Open Access Paving
Moveable Concrete
Barrier ( MCB)
( b) Single- lane ( T2: 2nd weekend)
Traffic Flow
Traffic Flow
( d) Linear Scheduling
Schedule ( hour)
0
1
2
3
4
0 10 20 30 40 50 60
Progress ( lane- km)
Mobilization Demolition Paving
Demobilization Weekend Closure
Figure 11. Work plan and lane closures for Full- Depth AC Replacement option.
34
35
time of the demolition trucks increases significantly, and the benefits of multiple demolition
teams diminish significantly.
As soon as the PCC slab and CTB are removed and the AB is trimmed, five or six lifts of
asphalt concrete are placed sequentially lift by lift with a single paving team. During the
following weekend closure, Truck Lane 1 ( T1) will be rebuilt using the same procedure— two
truck lanes ( T1 and T2) will be closed and Truck Lane 2 ( T2) will be used as the construction
access, as shown in Figure 11b.
A negative structural aspect of Full- Depth Single- Lane Rehabilitation is that the
interlocking of AC lifts by overlapping of longitudinal joints between adjacent rehabilitated lanes
( T1 and T2), as shown in Figure 5, is not possible. In addition, safe movement of the asphalt
delivery trucks from the delivery lane to the paving lane has to be resolved because the initial
elevation difference between the demolished lane and the access lane is between 330 and 406
mm. This discharging constraint is more serious with the semi bottom dump truck, which has no
side dumping feature.
2.4.2.2 Work Plan for Full- Depth Double- Lane Rehabilitation
For the Double- Lane Rehabilitation option, both truck lanes ( T1 and T2) will be rebuilt
during one weekend closure, which requires closing three lanes ( P2, T1, and T2). Passenger
Lane 2 ( P2) is assigned as the construction access for demolition and paving, as shown in Figure
11c. Only one demolition team and one paving team are assumed to be used in a sequential
construction operation due to the availability of only one access lane. Truck Lane 2 ( T2) will be
used as access for paving Truck Lane1 ( T1), and Passenger Lane 2 ( P2) will be used as access
for paving Truck Lane 1 because Truck Lane 2 will not be cool enough for delivery trucks.
36
Double- Lane Rehabilitation enables interlocking AC lifts along the joints between
adjacent lanes. However, the Double- Lane Rehabilitation scheme also causes more traffic
interruption because three lanes in one direction must be closed to traffic for the 55- hour
weekend closure.
2.5 Construction Resource Constraints
In order to achieve a realistic production capability for urban freeway rehabilitation, the
proper resource constraints must be recognized and established from a practical point of view.
This is a slightly different approach from that used for the concrete constructability analysis. In
the case of the concrete constructability analysis, maximum resource availability was initially
assumed to be the maximum theoretical production capability. This was done to check whether
the Caltrans production objective of 6 lane- km within one weekend closure is achievable. More
realistic resource constraints were then used in the sensitivity analysis to identify the impact of
resource limitations on the construction productivity.
As observed in the Caltrans LLCPRS demonstration project ( I- 10) case study, the
maximum resource constraint assumed for the concrete constructability analysis appears too
optimistic.( 6) Accordingly, a more practical and realistic resource constraint is assumed for the
asphalt constructability analysis. The following equipment resources are the major constraints
limiting the production capability of AC Rehabilitation:
• Production capacity of the asphalt concrete mixing plant
• Number and capacity of hauling trucks ( dump truck: DT) for demolition ( for Full-
Depth AC Replacement only)
37
• Capacity and number of asphalt concrete delivery trucks ( semi bottom dump truck:
SBT)
• Speed of asphalt concrete paving machine
• Speed of asphalt concrete compaction rollers in achieving required compaction
Table 2 summarizes the number and capacity of resources used in the deterministic
constructability analysis. The values shown in Table 2 were used to calculate the range of the
production capability of AC Rehabilitation within a 55- hour weekend closure. Based on the
experience of several AC contractors, the asphalt delivery and demolition hauling trucks were
found to be the primary constraints while the mixing plant and paver were the secondary
constraints. The AC compaction rollers were not a major constraint for AC Rehabilitation.
Table 2 Number and Capacity of Resources Used in the Deterministic Analysis
Resource Quantity Production
Capacity Units Remarks
AC Mixing Plant 1
Min.: 100
Max.: 200
Avg.: 150
m3/ hour
Dump Truck ( Demolition
for Full- Depth AC
Replacement option)
Min.: 8
Max.: 12
Avg.: 10
25 tons each trucks/ hour Efficiency = 0.6
No. of Teams = 1 to 2
Semi Bottom Dump
Truck ( Asphalt
Placement)
Min.: 9
Max.: 20
Avg.: 12
25 tons each trucks/ hour Efficiency = 0.95
No. of Teams = 1
Paver 1
25 mm: 7.5
50 mm: 6.0
75 mm: 4.5
km/ hour
Production Capacity is
inversely proportional to
AC lift thickness
Similar to the Concrete Rehabilitation scenario, a major concern for increasing the
production capability of the project is the total number of trucks that can be mobilized. For
example, if 10 demolition trucks were required every hour, approximately 45 demolition trucks
38
would need to be mobilized for every weekend rehabilitation project ( i. e., 45 trucks = 10 trucks
per hour per demolition team × 1.5 demolition teams × 2 shifts × 1.5 hours per truck turnaround).
Similarly, the total number of asphalt delivery trucks and the supply of aggregate to the mixing
plant would also need to be sufficient to avoid delays on the production side.
The locations of the plant and the demolition dumping area with respect to the
construction site are essential parameters influencing the production capability of the
rehabilitation because they directly affect the turnaround time of the demolition and delivery
trucks. Sufficient space is also needed at the asphalt concrete plant for the aggregate stockpiles.
Although the plant and paver are not the critical resource constraints governing
production capability, contractors believe these two resources are the most crucial pieces of
equipment for the success of the project. If one of these large and expensive pieces of equipment
breaks down during the pavement rehabilitation, the paving operation is suspended until it is
fixed or replaced, thereby causing overall productivity to drop significantly. Therefore,
redundancy in the mixing plant and paving machine is essential to prevent complete loss of
productivity, especially when the contract has severe incentive/ disincentive clauses.
2.6 AC Rehabilitation Constructability Analysis Process
The process used for the AC Rehabilitation constructability analysis is summarized as follows:
1. Set the rehabilitation project length as a production objective: for this study, 6 lane-km.
2. Set up construction window: for this study, 55- hour weekend closure or continuous
closure.
3. Select paving material: asphalt concrete.
4. Choose design profile: CSOL or Full- Depth AC Replacement.
39
5. Decide layer profile: Layer Profile “ A” or “ B” ( see Figure 7).
6. Consider lane closure tactics: Full or Half Closure ( applicable to CSOL option only).
7. Select paving lane strategies: Single- or Double- Lane Rehabilitation ( for Full- Depth
AC Replacement option only).
8. Compare Completion Option: Full or Partial Completion ( for CSOL option only).
9. Introduce cooling time analysis to check waiting time between paving of sequential
AC lifts.
10. Carry out a simple CPM ( Critical Path Method) scheduling to calculate net working
hours. From the CPM scheduling, total non- working hours are calculated first for the
following operations: 1) equipment mobilization/ demobilization, 2) delay for AC
cooling, 3) traffic switch time, and 4) time for paver to travel back to the start point
after completing a lift. The net working hours for demolition ( Full- Depth AC
Replacement case only) and AC paving are extracted by subtracting the total non-working
hours from the construction window length.
11. Calculate quantity of materials: demolition ( Full- Depth AC Replacement) and asphalt
concrete.
12. Determine the required number of resources and capacity.
13. Apply resource constraints. The number of trucks per hour is limited by the
minimum time for loading of old PCC slabs and the unloading of the new asphalt
concrete. For example, the number of demolition trucks showing up per hour for
each demolition team cannot exceed 12 in urban areas, based on the information
gathered from the concrete case study on the recent I- 10 reconstruction project near
Pomona.( 6, 7) The number of semi bottom dump trucks per hour for asphalt concrete
40
delivery is limited to 15, based on field data from several asphalt concrete overlay
projects.
14. Introduce linear scheduling concept. Linear scheduling methods are applied to the
constructability analysis to identify the maximum production capability of the AC
Rehabilitation given the resource constraints and progress of the resources involved.
Linear scheduling especially helps in the allocation of time between the paving and
demolition ( Full- Depth AC Replacement case only) activities. After the total paving
time is calculated from the CPM scheduling ( refer to Step 10 above), the paving
hours for each lift are determined based on the proportion of the thickness of each lift
to the total profile thickness. AC cooling time analysis is then applied to check if the
AC lifts will have cooled to the stop temperature before the paver is ready to place the
next lift. If the AC lift is expected not to have sufficiently cooled, the total number of
working hours is decreased and the linear scheduling process is re- run.
15. Finalize maximum production capability. The prototype software picks out the most
constraining resource at the calculated maximum production capability of the
rehabilitation for different design profiles, lift construction strategies, and lane closure
tactics.
16. Implement a stochastic analysis. Based on the same process used for the
deterministic constructability analysis, a stochastic constructability analysis is run by
varying the resources and scheduling parameters with an assumed Probability
Distribution Function ( PDF). This stochastic analysis gives a range of possible
production capabilities ( i. e., lower and upper bound with average) along with a
confidence level ( typically one standard deviation).
41
3.0 COOLING TIME SIMULATION
The time to cool the asphalt concrete layer to the specified maximum temperature at
which the paving machine or traffic can be placed on it ( cooling time) is considered a critical
component for the compaction of hot mix asphalt. The cooling time permits determination of the
optimal compaction time. The optimal time is between the high temperature “ overstressed
condition” of the mixture at which the asphalt is too soft to support compaction rollers, and the
low temperature “ understressed condition” at which the roller can not create sufficient shear
forces to further increase density ( compact the mix). Figure 12 shows a typical cooling time
curve for a single hot mix asphalt lift and how the optimal compaction time is determined from
the cooling temperatures.( 12)
Time
Average Pavement Temperature
Overstressed
Understressed
Optimal Compaction
Temperature
Optimal Compaction
Time Frame
Figure 12. Typical AC pavement cooling curve for single lift paving.( 12)
In the case of fast- track AC Rehabilitation with multi- lift AC paving, cooling time is
important for a different reason, especially in moderately warm climates such as is typical in
42
many parts of California. In multi- lift ( 4 to 6 lifts) construction, AC paving is scheduled for a
number of lanes ( typically 2 to 4 lanes) within limited weekend closure. To optimize paving
time, the next lift is placed immediately after the compaction of the first lift and therefore the
first lift must cool to the maximum allowable AC temperature before the next lift is placed.
A computer simulation program was used to predict the temperature profiles in multi- lift
AC Rehabilitation. The maximum production capability of the project within a weekend closure
is determined by subtracting waiting time for AC cooling from the total number of available
working hours in the CPM schedule. By optimizing the lift thickness and length of paving, the
number of hours of waiting for AC lifts to cool can be minimized.
A software program called PaveCool was developed and implemented in Minnesota to
estimate the allowable compaction time for single lift paving in cold weather.( 13, 14) The
limitation of PaveCool was that it did not cover multi- lift asphalt concrete and was not designed
for warm weather paving conditions. In 1999, a research team at the University of Minnesota
was contracted by the Pavement Research Center of UCB to develop a new analysis software
( CalCool) to predict the cooling time of multi- lift asphalt concrete pavements.( 12) The
numerical simulation software was developed utilizing Fourier’s Second Law to deal with heat
transfer in a pavement structure and the finite difference methods to solve a series of heat flux
equations. More details about the basic theory of CalCool are described by Timm.( 12) The
cooling time output from CalCool is an estimated solution with some calibration to field test
sections. In reality, the asphalt concrete cooling is very sensitive to the following variables:
cloud cover, wind speed, ambient temperature, material composition, time of placement, and
layer thickness.
43
3.1 Program Inputs and Outputs
As shown in Figure 13, the CalCool main input window consists of four categories as
following:
• Paving starting time
• Environmental conditions
• Existing surface conditions
• Mix specifications
Figure 13. CalCool main input window.
44
The cooling time from CalCool is the average lift temperature for the individual lifts.
The results of cooling time simulation are plotted graphically as cooling time curves or
alternatively can be tabulated to show the predicted cooling time of individual lift to a specified
temperature, as shown in Figure 14. The input and output data can be exported to a text file or a
spreadsheet.
3.2 Experimental Validation of CalCool
CalCool needed to be validated with actual field data before used as a part of the asphalt
constructability analysis model. A validation study of CalCool using experimental data collected
by Pavement Research Center Staff from several AC paving projects in California was
performed.( 15) Both single and multi- lift comparisons were made between CalCool and the
field data. Comparisons were also made with AC cooling data available in the literature from
other field projects.
Table 3 compares the cooling time from CalCool with experimental data where the
delivered temperature of the hot mix asphalt was 149° C ( 300° F) and the stop temperature was
79° C ( 175° F) for two different ambient temperatures.( 16, 17) The predicted cooling time by
CalCool was similar to the test results for both thin and thick asphalt pavement layers except for
one data point. Cooling curves from two experiments were in good agreement with the predicted
cooling curves from CalCool for a single AC lift, as shown in Figure 15.( 17)
Figure 14. CalCool tabular and graphical output window.
45
46
Table 3 Comparison of Predicted Cooling Time using CalCool and Observed Cooling
Time
Single Lift Cooling Time* ( min.)
Layer
Thickness
Ambient
Temperature
Asphalt Institute
Observation
CalCool
Prediction
25 mm ( 1 in.) 32° C ( 90° F) 9 10
51 mm ( 2.4 in.) 32° C ( 90° F) 23 28
76 mm ( 3 in.) 32° C ( 90° F) 45 52
61 mm ( 2.4 in.) 21° C ( 70° F) 78 40
89 mm ( 3.5 in.) 21° C ( 70° F) 77 78
119 mm ( 4.7 in.) 21° C ( 70° F) 110 119
178 mm ( 7 in.) 21° C ( 70° F) 220 237
* Cooling time from 149° C ( 300° F) to 79° C ( 175° F)
0
20
40
60
80
100
120
140
160
0 30 60 90 120 150 180 210 240
Cooling Time ( min.)
Temperature (° C)
Webster ( 178 mm)
CalCool ( 178 mm)
Webster ( 89 mm)
CalCool ( 89 mm)
Figure 15. Comparison of Webster experimentally observed and CalCool predicted cooling
times.
47
3.3 Validation of CalCool with Field Data
CalCool was compared with field data from two construction projects.( 15) The first site
involved daytime construction on a 2.4- km length of Route 1 in Lompoc, CA ( near Santa
Barbara, CA). The second site involved a nighttime construction on a main road in San Leandro,
CA.
3.3.1 Temperature Data Collection for CalCool Validation and Calibration in Lompoc, CA
The Lompoc construction site involved removal of the existing asphalt concrete and
placement of approximately 270 mm of new asphalt concrete in three lifts over the existing
granular base. The first lift of material on the existing granular base was a rich bottom ( 5.8
percent asphalt content, AR- 8000) asphalt mixture with 19- mm maximum size coarse aggregate.
The asphalt content for the subsequent lifts was 5.3 percent. The hot mix asphalt concrete was
placed in windrows by semi bottom dump trucks. For much of the time, the AC paver was
waiting for the delivery of the hot mix asphalt and as a result, delivery temperatures measured in
the windrow were on average 155° C.
An “ anteater” was used to pick up the windrow and transfer it to the paver. The delivery
temperature of the asphalt concrete was taken with a digital thermometer once the bottom dump
truck placed the windrow. AC temperatures were monitored over time at the same locations. At
each location, temperatures were recorded at three spots: near the edge, 1 m from the edge, and
mid- depth in the lift. The air temperature and wind speed were also recorded at each location.
The number and frequency of the measurements varied depending on the number of locations
being monitored. Sampling of temperatures and wind speed continued until the AC temperature
reached 50 or 60° C. At this construction site, the second lift was placed a day after the first lift,
48
while the third lift was placed immediately after the second lift when its temperature reached
60° C.
3.3.2 Temperature Data Collection for CalCool Validation and Calibration in San Leandro, CA
The second site used to calibrate CalCool was on Marina Boulevard in San Leandro, CA.
Unlike the Lompoc site, this project was constructed at night due to its use as a main corridor for
heavy truck traffic off of Interstate 880. Construction involved removing 318 mm of existing
asphalt concrete and replacing it with a 19- mm maximum size coarse aggregate mix with 5.2
percent asphalt ( AR- 8000). The first lift of asphalt concrete was placed over the existing
granular base near the edge and over portland cement concrete on the adjacent lanes. The
existing layers were wet due to heavy mist and rain. Three lifts of asphalt concrete were placed
nearest the edge and four lifts on the adjacent lanes.
This construction can be considered a true multi- lift construction. The lifts were placed
one after the other in the same night similar to the scenarios analyzed in this research and
discussed in Section 2.0. End dump trucks were used to deliver the hot mix asphalt concrete.
Unlike the Lompoc construction, delivery trucks were waiting in line to feed the AC paver. The
project was much shorter than the Lompoc project ( about 245 m on the first day) and the paver
needed to maneuver around corners and backup to the start point after it reached the end.
Delivery temperatures of the asphalt mix were more variable and generally lower than the
Lompoc construction. The average initial temperature of the hot mix asphalt was 144° C. Initial
temperature measurements were also a bit lower for this project because they were taken behind
the paver rather than from the truck or the windrow, as was done in Lompoc. Recording of
asphalt cooling temperatures were performed in a similar manner to the Lompoc construction.
49
Measurements shown for the San Leandro project were the average of the three locations ( edge,
surface, mid- depth).
3.3.3 Comparison of CalCool and Field Measurements
One of the goals of recording cooling temperatures of field construction of asphalt
concrete is to validate and calibrate CalCool. The two construction projects used for calibration
were selected to include different values for most of the variables included in CalCool. The two
projects included day and night construction, extremes in cloud cover ( clear and dry to overcast),
different existing surface materials ( except subgrade), wet and dry conditions in the granular
base, and single and multi- lift construction.
As shown in Figures 16 and 17, the field data correlated very closely with CalCool for
single and double lift construction. With three lifts, CalCool overestimated how fast the lift
would cool down and underestimated how much the lift heats back up when a new lift is placed
on top of it, as shown in Figure 18. As shown in Figure 19, CalCool underestimated the time
required to reach the stop temperature for AC placed over a PCC surface.
50
Lompoc H- Street AC Construction
Point 4, Lift 2, One Lift over Rich Bottom AC
0
20
40
60
80
100
120
140
160
0: 00 0: 28 0: 57 1: 26 1: 55 2: 24 2: 52 3: 21
Time ( hours)
Temperature (° C)
In- depth
Surface
Edge
CalCool
Date: October 7, 1999
Time: 7: 55 a. m.
Avgerage Air Temp: 11° C
Average Wind: 2.5 kph
Existing Surface: AC
Existing Surface Temp: 14.0° C
Cloud Cover: Clear and dry
Mix Specification: DGAC
Lift Thickness: 90 mm
Figure 16. Cooling curve for a single lift of rich bottom AC placed on granular base
( Lompoc project).
Lompoc H- Street AC Construction
Point 11, Two Lifts over Rich Bottom AC
0
20
40
60
80
100
120
140
160
0: 00 1: 12 2: 24 3: 36 4: 48 6: 00 7: 12 8: 24 9: 36 10: 48
Time ( hour)
Temperature (° C)
In- depth ( lift 2) Surface ( lift 2)
Edge ( lift 2) In- depth ( lift 3)
Surface ( lift 3) Edge ( lift 3)
CalCool Lift 2 CalCool Lift 3
Date: October 8, 1999
Times: 8: 50 and 11: 22 a. m.
Average Air Temp: 23, 32° C
Average Wind Speed: 5.0, 0.0 kph
Existing Surface: AC
Existing Surface Temp: 23, 58° C
Cloud Cover: Clear and dry, Clear and dry
Mix Specification: DGAC
Lift Thickness: 80 mm
Figure 17. Cooling curve for a double lift of AC placed on rich bottom AC layer ( Lompoc
project).
51
San Leandro Marina Blvd. AC Construction
Location 2
0
20
40
60
80
100
120
140
160
0: 00 1: 12 2: 24 3: 36 4: 48 6: 00 7: 12 8: 24 9: 36 10: 48
Time ( hours)
Temperature (° C)
Measured Lift 1 Measured Lift 2
Measured Lift 3 CalCool Lift 1
CalCool Lift 2 CalCool Lift 3
Date: July 17, 2000
Times: 1: 17, 2: 55, 4: 20 a. m.
Average Air Temp: 14.0, 16.4, 16.3 ° C
Average Wind Speed: 7.1, 5.8, 16.3 kph
Existing Surface: Granular Base ( wet), AC, AC
Existing Surface Temp: 19.4, 64.6, 77.6° C
Cloud Cover: Overcast
Mix Specification: DGAC
Lift Thickness: 100, 100, 91 mm
Figure 18. Cooling curves for a three lift AC layer placed on granular base ( San Leandro
project).
Marina Blvd San Leandro AC Construction
Location 3
0
20
40
60
80
100
120
140
160
0: 00 0: 28 0: 57 1: 26 1: 55 2: 24 2: 52 3: 21 3: 50 4: 19
Time ( hour)
Temperature (° C)
Measured Lift 1 Measured Lift 2
Measured Lift 3 CalCool Lift 1
CalCool Lift 2 CalCool Lift 3
Date: July 17, 2000
Times: 2: 00, 3: 37, 4: 45 AM
Average Air Temp: 15.9, 15.8, 16.0 ° C
Average Wind Speed: 5.8, 4.0, 4.0 kph
Existing Surface: PCC, AC, AC
Existing Surface Temp: 20.0, 81.1, 86.6° C
Cloud Cover: Overcast
Mix Specification: DGAC
Lift Thickness: 61, 76, 76 mm
Figure 19. Cooling curve for a three lift AC layer placed on existing PCC ( San Leandro
project).
52
53
4.0 RESULTS OF THE AC CONSTRUCTABILITY ANALYSIS
Two types of calculation were implemented for the asphalt constructability analysis as
follows:
• Deterministic analysis, in which major input parameters such as resource availability,
scheduling factors, and delay for AC cooling time were treated as constants without
variations, and
• Stochastic analysis, in which these parameters were treated as random variables with
defined probability distributions. The stochastic engine used was called Crystal
Ball from Decisioneering( 18) along with the UCB prototype analysis spreadsheet
for deterministic analysis.
The rehabilitation production capability analysis results are expressed in two different
ways: centerline- meters and lane- meters. Lane- meters is the product of the number of
rehabilitated lanes and centerline- meters.
4.1 CSOL Production Capability
4.1.1 Deterministic Analysis
The initial comparison between rehabilitation options was based on the deterministic
analysis. The purpose of the deterministic analysis was to measure the sensitivity of the freeway
rehabilitation production capability to all input parameters.
54
4.1.1.1 CSOL Production Capability in Centerline- meters
The result of the deterministic analysis of CSOL production capability ( centerline-meters)
for a 55- hour weekend closure is summarized in Table 4. For the partial lane closure
options, the total productivity required for two weekends was determined and then divided by
two to come up with the production capability for one weekend in order to facilitate easy
comparison to the other rehabilitation options. The Layer Profile “ A” option for the CSOL Half
Closure Partial Completion strategy was found to be similar to the CSOL Full Closure Full
Completion option ( Profile “ A”) and therefore was not included in Table 4.
Table 4 Deterministic Analysis Results for CSOL Production per 55- Hour Weekend
Closure, Four- Lane Rehabilitation.
Production per Weekend Closure ( Centerline- meters)
Semi Bottom Dump Full Closure Half Closure
Truck Cycles Full
Completion
Full
Completion
Partial
Completion*
Cycle Time
( min.)
Trucks
per Hour
Profile
“ A”
Profile
“ B”
Profile
“ A”
Profile
“ B”
Profile
“ B”
7 9 859 988 708 806 930
6 10 1,002 1,153 825 940 1,085
5 12 1,202 1,384 991 1,128 1,302
4 15 1,503 1,729 1,238 1,410 1,628
3 20 1,552 1,750 1,253 1,427 1,647
* Total productivity required for two weekends was determined and then divided by two to come
up with the production capability for one weekend in order to facilitate easy comparison to the
other rehabilitation options
The CSOL production table was converted into production graphs for better visual
understanding and comparison between the rehabilitation options, as shown in Figures 20 and
21. In Figure 20, the rehabilitation production was presented as a function of the cycle time of
the asphalt delivery trucks for each rehabilitation option ( because the number of semi bottom
55
CSOL Production ( centerline- meters)
500
1,000
1,500
2,000
2 3 4 5 6 7 8
Cycle Time of Semi Bottom Dump Trucks ( minutes)
Production ( centerline- meters)
Full Closure Full Completion Layer Profile " B"
Full Closure Full Completion Layer Profile " A"
Half Closure Full Completion Layer Profile " B"
Half Closure Full Completion Layer Profile " A"
Half Closure Partial Completion Layer Profile " B"
Figure 20. Deterministic analysis of CSOL production in centerline- meters as a function of
semi bottom dump truck cycle time.
CSOL Production ( centerline- meters), Deterministic Analysis
500
1,000
1,500
2,000
Full Closure Full
Completion Layer
Profile " B"
Full Closure Full
Completion Layer
Profile " A"
Half Closure Full
Completion Layer
Profile " B"
Half Closure Full
Completion Layer
Profile " A"
Half Closure Partial
Completion Layer
Profile " B"
Rehabilitation Option
Production ( centerline- meters)
9 SBT/ hour 10 SBT/ hour 12 SBT/ hour
15 SBT/ hour 20 SBT/ hour
Number of Semi Bottom Dump Trucks per Hour
Figure 21. Deterministic analysis of CSOL production in centerline- meters as a function of
rehabilitation option and number of semi bottom dump trucks/ hour.
56
dump trucks is a primary constraint). In Figure 21, the production was plotted in comparison
with various rehabilitation options with a range of delivery trucks per hour.
4.1.1.2 CSOL Production Capability in Lane- meters for Four- Lane Rehabilitation
Similarly, the result of the CSOL production capability in terms of total lane- meters for
four- lanes rehabilitation is summarized in Table 5 for the various options. Figures 22 and 23
show a graphic display of the production capability results presented in Table 5 with respect to
delivery truck cycle time and number of delivery trucks per hour, respectively.
Table 5 Deterministic Analysis Results for CSOL Production, Four- Lane
Rehabilitation
Production per Weekend Closure ( Lane- meters)
Semi Bottom Dump Full Closure Half Closure
Truck Cycles Full
Completion
Full
Completion
Partial
Completion*
Cycle Time
( min.)
Trucks
per Hour
Profile
“ A”
Profile
“ B”
Profile
“ A”
Profile
“ B”
Profile
“ B”
7 9 3,435 3,953 2,830 3,222 3,720
6 10 4,007 4,612 3,302 3,759 4,340
5 12 4,808 5,534 3,962 4,511 5,208
4 15 6,010 6,918 4,953 5,639 6,510
3 20 6,088 7,001 5,014 5,707 6,589
* Total productivity required for two weekends was determined and then divided by two to come
up with the production capability for one weekend in order to facilitate easy comparison to the
other rehabilitation options
The Layer Profile “ B” with a full lane closure and full completion of the rehabilitation on
all four lanes is the most productive strategy in terms of centerline- meters. The productivity of
the rehabilitation increases for all options with an increase in AC delivery trucks per hour. The
least productive option was the Layer Profile “ A” with the CSOL Half Closure Full Completion
strategy.
57
CSOL Production ( lane- meters)
2,000
4,000
6,000
8,000
2 3 4 5 6 7 8
Cycle Time of Semi Bottom Dump Trucks ( minutes)
Production ( lane- meters)
Full Closure Full Completion Layer Profile " B"
Full Closure Full Completion Layer Profile " A"
Half Closure Full Completion Layer Profile " B"
Half Closure Full Completion Layer Profile " A"
Half Closure Partial Completion Layer Profile " B"
Figure 22. Deterministic analysis of CSOL production in lane- meters as a function of semi
bottom dump truck cycle time.
CSOL Production ( lane- meters), Deterministic Analysis
2,000
4,000
6,000
8,000
Full Closure Full
Completion Layer
Profile " B"
Full Closure Full
Completion Layer
Profile " A"
Half Closure Full
Completion Layer
Profile " B"
Half Closure Full
Completion Layer
Profile " A"
Half Closure Partial
Completion Layer
Profile " B"
Rehabilitation Option
Production ( lane- meters)
9 SBT/ hour 10 SBT/ hour 12 SBT/ hour
15 SBT/ hour 20 SBT/ hour
Number of Semi Bottom Dump Trucks per Hour
Figure 23. Deterministic analysis of CSOL production in lane- meters as a function of
rehabilitation option and semi bottom dump trucks/ hour.
58
4.1.2 Stochastic Analysis
In order to calculate a realistic range of production capabilities for the various
rehabilitation options, a stochastic analysis was conducted by treating the parameters affecting
production as random variables.
4.1.2.1 Random Variable Parameters for Stochastic Analysis.
Table 6 summarizes how major input parameters for the stochastic analysis were treated
as random variables. The CSOL Half Closure Full Completion Layer Profile “ A” option is used
as an example.
Table 6 Example of Random Variables for the CSOL Half Closure Full Completion
Layer Profile “ A” Option, Stochastic Analysis
Variable ( parameter) Unit Distribution
Type Probability Distribution Function
Mobilization time hours Triangular min = 1, mean = 2, max = 3
De- mobilization time hours Triangular min = 2, mean = 3, max = 4
Mix plant capacity m3/ hour Normal mean = 150, standard deviation = 15
Cycle time of SBT minutes Normal mean = 5, standard deviation = 0.5
Efficiency of SBT n/ a Triangular min = 0.85, mean = 0.95, max = 1.0
Traffic switch time hours Triangular min = 0, mean = 1, max = 2
Delay for AC cooling hours Triangular min = 3.5, mean = 5.5, max = 6.5
The type of distribution was assumed realistically with resource reference information
from AC field data and the I- 10 project concrete case study. The mean of the distribution is the
same as the typical value for the deterministic analysis.( 6)
The parameters were randomly generated and combined to complete 1,000 runs in the
constructability analysis spreadsheet. The prediction of the production capability is shown in
Figure 24 along with a “ one- standard deviation” of confidence interval around the mean. As the
sum of the independent input parameters of random variables, the production capability has an
59
Certainty is 68% from 895 to 1,118 lane- meters
Mean = 990
.000
.008
.016
.023
.031
0
7.75
15.5
23.25
31
726 862 998 1,135 1,271
1,000 Trials 10 Outliers
Forecast of Production ( CSOL Half Closure
Full Completion Layer Profile " A")
Probability
Frequency
Lane- meters
Figure 24. Forecast of production for CSOL from stochastic analysis ( CSOL Half Closure
Full Completion Layer Profile “ A”).
approximate normal distribution, based on the “ Central Limit Theorem”( 19), as shown in Figure
24. In Figure 24, one standard deviation for the confidence interval means there is a 68 percent
likelihood the production capability of the rehabilitation will fall within the interval given the
resource inputs and productions.
Another advantage of the stochastic analysis is to indicate the sensitivity of the results to
the input parameters. Figure 25 shows that the cycle time of the asphalt delivery trucks ( SBT) is
the most influential variable in the rehabilitation production capability.
4.1.2.2 Result of the CSOL Stochastic Analysis
Table 7 summarizes the result of the CSOL stochastic analysis in terms of centerline-meters
categorized into different intervals of likelihood, ( i. e., lower bound, mean, and upper
bound). The same results are plotted into a centerline- meters production graph ( Figure 26)
60
CSOL Half Closure Full Completion Layer Profile " A"
Cycle of Semi Bottom Dump Trucks ( min.) -. 92
Semi Bottom Dump Truck Efficiency .28
First AC Cooling Delay -. 15
Second Traffic Switch -. 11
Mobilization Duration -. 10
Paver Travel Speed ( kph) .09
Third AC Cooling Delay -. 07
Batch Plant Capacity ( m3/ hour) .06
Second AC Cooling Delay -. 03
Cooling Delay ( hours) .03
First Traffic Switch -. 01
- 1 - 0.5 0 0.5 1
Measured by Rank Correlation
Resource Sensitivity
Figure 25. Resource sensitivity for CSOL stochastic analysis ( CSOL Half Closure Full
Completion Layer Profile “ A”).
Table 7 Stochastic Analysis Results for CSOL Production, Four- Lane Rehabilitation
Production per Weekend Closure for Given
Rehabilitation Option ( Centerline- meters)
Closure Option Full Closure Half Closure
Completion
Option
Full
Completion
Full
Completion
Partial
Completion2
Layer Profile Profile
“ A”
Profile
“ B”
Profile
“ A”
Profile
“ B”
Profile
“ B”
Lower Bound 1,080 1,231 894 1,003 1,193
Mean ( average) 1,190 1,358 990 1,106 1,316
Deterministic1 1,202 1,384 991 1,128 1,302
Upper Bound 1,322 1,515 1,116 1,245 1,456
112 semi bottom dump trucks per hour
2Total productivity required for two weekends was determined and then divided by two to come
up with the production capability for one weekend in order to facilitate easy comparison to the
other rehabilitation options
61
CSOL Production ( centerline- meters), Stochastic Analysis
500
1,000
1,500
2,000
Full Closure Full
Completion Layer
Profile " B"
Full Closure Full
Completion Layer
Profile " A"
Half Closure Full
Completion Layer
Profile " B"
Half Closure Full
Completion Layer
Profile " A"
Half Closure Partial
Completion Layer
Profile " B"
Rehabilitation Option
Production ( centerline- meters)
Lower bound Mean Deterministic Upper bound
Figure 26. Stochastic analysis of CSOL production in centerline- meters as a function of
rehabilitation option.
showing the likely production interval for the various analysis options. Similar to the centerline-meter
production, the results of the stochastic analysis for CSOL in terms of total lane- meters for
four lanes rehabilitated are summarized in Table 8 and Figure 27.
The mean production capability from the stochastic analysis is very close to the
deterministic analysis when using an average of 12 asphalt delivery trucks ( SBT cycle time of 5
minutes, as used for the stochastic analysis).
62
Table 8 Stochastic Analysis Results for CSOL Production, Four- Lane Rehabilitation
Production per Weekend Closure for Given
Rehabilitation Option ( Lane- meters)
Closure Option Full Closure Half Closure
Completion
Option
Full
Completion
Full
Completion
Partial
Completion2
Layer Profile Profile
“ A”
Profile
“ B”
Profile
“ A”
Profile
“ B”
Profile
“ B”
Lower Bound 4,321 4,925 3,575 4,010 4,773
Mean ( average) 4,758 5,431 3,956 4,422 5,264
Deterministic1 4,808 5,534 3,962 4,511 5,208
Upper Bound 5,289 6,060 4,465 4,979 5,826
112 semi bottom dump trucks per hour
2Total productivity required for two weekends was determined and then divided by two to come
up with the production capability for one weekend in order to facilitate easy comparison to the
other rehabilitation options
CSOL Production ( lane- meters), Stochastic Analysis
2,000
4,000
6,000
8,000
Full Closure Full
Completion Layer
Profile " B"
Full Closure Full
Completion Layer
Profile " A"
Half Closure Full
Completion Layer
Profile " B"
Half Closure Full
Completion Layer
Profile " A"
Half Closure Partial
Completion Layer
Profile " B"
Rehabilitation Option
Production ( lane- meters)
Lower bound Mean Deterministic Upper bound
Figure 27. Stochastic analysis of CSOL production in lane- meters as a function of
rehabilitation option.
63
4.1.3 Production Comparison of the Rehabilitation Options for CSOL
Table 9 compares the relative average production capability from the CSOL stochastic
analysis between each rehabilitation option and the fastest option ( i. e., CSOL Full Closure Full
Completion Layer Profile “ B”). Table 9 also includes the number of hours of delay due to
waiting for hot AC to cool and switching of traffic between lanes. The results show that the
amount of delay greatly affects the overall productivity of the rehabilitation.
Table 9 Production Comparison for CSOL Rehabilitation
Comparison of Production per Weekend Closure for Given
Rehabilitation Option ( Lane- meters)
Closure Full Closure Half Closure
Completion Option Full
Completion
Full
Completion
Partial
Completion3
Layer Profile Profile
“ A”
Profile
“ B”
Profile
“ A”
Profile
“ B”
Profile
“ B”
Average Production1 4,758 5,431 3,956 4,422 5,264
Comparison2 88 % 100% 72% 81% 97%
Delay ( hours) 0 0 9 8.5 3
AC
Cooling
Traffic
Switching 0 0 0 0 2 7 2 6.5 0 3
1Stochastic analysis in terms of total lane- meters for four- lane rehabilitation
2Compared with CSOL Full Closure Full Completion Layer Profile “ B”
3Total productivity required for two weekends was determined and then divided by two to come
up with the production capability for one weekend in order to facilitate easy comparison to the
other rehabilitation options
The Layer Profile “ B” ( 200- mm overlay) has approximately 12 percent more production
capability than the Layer Profile “ A” ( 230- mm overlay) for full- and half- lane closure strategies.
This production capability ratio is almost the same as the ratio of the overlay thicknesses of the
two pavement profiles, ( i. e., 88 percent = Profile “ B”/ Profile “ A” thickness = 200 mm/ 230 mm).
On average, the Half Closure Full Completion case is approximately 20 percent less
productive than the Full Closure Full Completion option for both pavement profiles. The 20
64
percent decrease in productivity must be compared with the reduced traffic delay caused by
leaving two lanes open to traffic in the Half Closure option instead of having all four lanes
closed as in the Full Closure option.
The Half Closure Full Completion option is less productive than the Full Closure Full
Completion option because of delays for AC cooling and traffic switches. However, the road
user is less inconvenienced with the Half Closure Full Completion option relative to the Full
Closure Full Completion option.
In the Half Closure Partial Completion option, the delay caused by AC cooling is
negligible and therefore the production capability was found to be almost the same as the Full
Closure Full Completion case. With the Half Closure Partial Completion option, two out of four
lanes are always open to traffic with only a 3 percent loss in productivity compared to the Full
Closure Full Completion option. The only issue to resolve is the impact on pavement life of
opening two out of the four lifts of AC for one week to normal urban freeway traffic.
4.2 Full- Depth AC Replacement Production Capability
The results of deterministic and stochastic analyses for Full- Depth AC Replacement with
Single- and Double- Lane Rehabilitation are described in this section.
4.2.1 Deterministic Analysis
4.2.1.1 Production Capability of Full- Depth Single- Lane Rehabilitation
The linear scheduling technique descried in detail for the concrete constructability
analysis in Reference ( 4) was used again in the analysis for Full- Depth AC Replacement. This
technique was used to determine the optimum time allocation between the demolition and paving
65
activities for a given set of resource constraints. For example, in a 55- hour weekend closure
there were 24 hours for paving ( including 3 hours AC cooling) and 28 hours for demolition
assuming 12 demolition trucks and 10 asphalt delivery trucks per hour ( Full- Depth Single- Lane
Layer Profile “ B” case).
Table 10 shows the constructability analysis results for the Single Lane Rehabilitation
using the Full- Depth AC Replacement strategy. The constraints on production capability were
the pavement profile ( Profile “ A” or “ B”) and the number of demolition teams. The number of
demolition and asphalt delivery trucks also plays a key role in the production of this strategy. In
the case of two demolition teams, more than one construction access lane needs to be provided
during the demolition work. If the shoulder width is more than 3 meters, then it can be used as
one of the access lanes. If only one access lane is available for two demolition teams, then the
resultant productivity will be equivalent to 1.5 demolition crews.( 6) The poor productivity of
two teams with one access lane is caused by construction traffic congestion.
Table 10 Deterministic Analysis Results for Production of Full- Depth AC
Replacement, Single- Lane Rehabilitation
Production ( Lane- meters)
Trucks per hour Profile
“ A”
Profile
“ B”
Profile
“ A”
Profile
“ B”
Profile
“ A”
Profile
“ B”
Semi Bottom
Dump Truck
Dump
Truck
1 Demolition
Team
1 Demolition
Team
1.5 Demolition
Teams
1.5 Demolition
Teams
2 Demolition
Teams
2 Demolition
Teams
10 10 1,544 1,216 2,028 1,600 2,356 1,879
10 12 1,723 1,357 2,222 1,753 2,548 2,032
12 10 1,648 1,298 2,203 1,738 2,593 2,068
12 12 1,853 1,460 2,433 1,920 2,827 2,255
15 10 1,766 1,391 2,411 1,902 2,883 2,299
15 12 1,943 1,530 2,597 2,049 3,057 2,438
66
4.2.1.2 Production Capability of Double- Lane Rehabilitation
The productivity results of Double- Lane Rehabilitation using Full- Depth AC
Replacement are summarized in Table 11 for both pavement profiles and as a function of the
number of demolition and asphalt delivery trucks per hour. Two demolition teams work
simultaneously in the model, but because of the availability of only a single access lane, the
calculation assumed 1.5 demolition teams ( the net effect of 2 demolition teams with a single
access lane).
The production capability of the Single- Lane Rehabilitation option in Table 10 and
Double- Lane option in Table 11 were combined and the results shown in Figure 28 and Table
12. The Single- Lane Rehabilitation strategy was more productive than the Double- Lane
Rehabilitation strategy because fewer working hours were spent waiting for AC cooling
compared with the double- lane option.
Table 11 Deterministic Analysis Results for Production of Full- Depth AC
Replacement, Double- Lane Rehabilitation
Production
Trucks per hour
Centerline- meters Lane- meters
Semi Bottom
Dump Trucks Dump Trucks Profile
“ A”
Profile
“ B”
Profile
“ A”
Profile
“ B”
10 10 890 714 1,781 1,427
10 12 976 782 1,951 1,564
12 10 967 775 1,935 1,551
12 12 1,069 856 2,137 1,713
15 10 1,059 848 2,117 1,697
15 12 1,181 947 2,362 1,893
67
Table 12 Deterministic Analysis Results for Full- Depth AC Replacement Production,
Single- versus Double- Lane Rehabilitation
Production ( Lane- meters)
Trucks per hour
Single Lane Double Lane
Semi Bottom
Dump Truck Dump Truck Profile
“ A”
Profile
“ B”
Profile
“ A”
Profile
“ B”
10 12 2,222 1,753 1,951 1,564
12 10 2,203 1,738 1,935 1,551
12 12 2,433 1,920 2,137 1,713
11.5 demolition teams for both Single- and Double- Lane Rehabilitation
Full- Depth AC Replacement Production, Deterministic Analysis
1,000
1,500
2,000
2,500
3,000
Single- Lane Layer
Profile " A"
Single- Lane Layer
Profile " B"
Double- Lane Layer
Profile " A"
Double- Lane Layer
Profile " B"
Rehabilitation Option
Production ( lane- meters)
SBT= 10, DT= 10 SBT= 10, DT= 12 SBT= 12, DT= 10
SBT= 12, DT= 12 SBT= 15, DT= 10 SBT= 15, DT= 12
Number of Semi Bottom Dump Trucks and Dump Trucks Per Hour
Figure 28. Deterministic analysis of Full- Depth AC Replacement production as a function
of Single- or Double- Lane Rehabilitation, and type and number of trucks per hour.
68
4.2.2 Stochastic Analysis
For the Full- Depth AC Replacement strategy, a stochastic analysis was completed and
the results were compared with the results of the deterministic analysis. Table 13 shows an
example of the random variables used for the Full- Depth Double- Lane Layer Profile “ B” case
and their corresponding distribution types and probability distribution functions ( PDF). Similar
to the stochastic analysis for the CSOL case, the distribution types were realistically assumed
using reference information from AC field data and the concrete case study with the I- 10
project.( 6) The typical value of the deterministic analysis was used as the mean of the
distribution.
Table 13 Example of Random Variables for the Full- Depth AC Replacement, Double-
Lane, Layer Profile “ B,” Stochastic Analysis
Variable ( Parameter) Unit Distribution
Type Probability Distribution Function
Mobilization time hours Triangular min = 0.5, mean = 1, max = 1.5
Demobilization time hours Triangular min = 2, mean = 3, max = 4
Mixing plant capacity m3/ hour Normal mean = 150, standard deviation = 15
Demolition team* number Discrete min = 1, mean = 1.5, max = 2
Dump truck frequency* trucks/ hour Normal mean = 10, standard deviation = 0.1
Dump truck efficiency* - Triangular min = 0.45, mean = 0.6, max = 0.75
Semi bottom dump
truck frequency trucks/ hour Normal mean = 12, standard deviation = 1.2
Efficiency of semi
bottom dump truck - Triangular min = 0.85, mean = 0.95, max = 1.0
Delay for AC cooling hour Triangular min = 4, mean = 7, max = 9
* New variables in addition to the CSOL stochastic analysis ( refer to Table 6)
An example of the Full- Depth AC Replacement stochastic analysis is shown in Figure 29
for the Full- Depth Double- Lane Layer Profile “ B” case. For this rehabilitation case, the
stochastic analysis forecasted an AC production capability with a range of 1.2 to 1.8 lane- km
with a mean of 1.5 lane- km during a 55- hour weekend closure. As shown in Figure 30, the
overall production of the AC Rehabilitation was most sensitive to the number of demolition
69
Certainty is 68% from 1,211.45 to 1,784.22 lane- meters
Mean = 1,523.85
.000
.007
.015
.022
.029
0
7.25
14.5
21.7
5
29
850 1,193 1,536 1,879 2,222
1,000 Trials 2 Outliers
Forecast of Production ( Full- Depth AC Replacement
Double- Lane Layer Profile “ B”)
Probability
Frequency
Lane- meters
Figure 29. Forecast of production for Full- Depth AC Replacement from stochastic analysis
( Full- Depth Double- Lane Layer Profile “ B”).
Full- Depth Replacement Double- Lane Layer Profile “ B”
Number of Demolition Teams .76
Dump Trucks per hour .34
Dump Truck Efficiency .31
Semi Bottom Dump Trucks per hour .27
Standby from cooling time -. 22
Semi Bottom Dump Truck Efficiency .10
Overlap of demolition and paving .08
Paver return speed ( kph) .03
Batch Plant Capacity ( m3/ hour) -. 03
Mobilization duration -. 01
- 1 - 0.5 0 0.5 1
Measured by Rank Correlation
Resource Sensitivity
Figure 30. Resource sensitivity for Full- Depth AC Replacement stochastic analysis ( Full-
Depth Double- Lane Layer Profile “ B”).
70
teams, the number of dump trucks per hour, the efficiency of the dump trucks, the number of
asphalt delivery trucks per hour, and the efficiency of the AC delivery trucks.
The results of the stochastic analysis for Full- Depth AC Replacement are summarized in
Table 14 for the Single- and Double- Lane cases for each layer profile. The data from Table 14
was plotted to show the potential range of rehabilitation productivity, ( i. e., lower and upper
bounds with mean), as shown in Figure 31. Using stochastic analysis, the Single- Lane
Rehabilitation case was found to be more productive than the Double- Lane Rehabilitation case.
The mean productivity for each strategy was close to what was calculated using deterministic
analysis because the mean of random variable distributions is same as the typical value of the
deterministic analysis.
Table 14 Stochastic Analysis Results for Full- Depth AC Replacement Production.
Production ( Lane- meters)
Lanes Rebuilt Single Lane Double Lane
Layer Profile Profile “ A” Profile “ B” Profile “ A” Profile “ B”
Lower bound 1,647 1,330 1,512 1,211
Mean 2,103 1,694 1,914 1,524
Deterministic* 2,203 1,738 1,935 1,551
Upper bound 2,429 1,958 2,232 1,784
* Semi bottom dump trucks: 12/ hr.; dump trucks: 10/ hr.
4.2.3 Productivity Comparison of Full- Depth AC Replacement
Table 15 compares the production capability of Single- and Double- Lane Rehabilitation
strategies for both pavement profiles (“ A” and “ B”) along with the number of hours the paving
operation was delayed due to AC cooling. The production for each option is compared to the
most productive option ( Single- Lane Layer Profile “ A”).
71
Full- Depth AC Replacement Production, Stochastic Analysis
1,000
1,500
2,000
2,500
3,000
Single- Lane Layer
Profile " A"
Single- Lane Layer
Profile " B"
Double- Lane Layer
Profile " A"
Double- Lane Layer
Profile " B"
Rehabilitation Option
Production ( lane- meter)
Lower bound Mean Deterministic Upper bound
Figure 31. Stochastic analysis for Full- Depth AC Replacement production, Single- versus
Double- Lane Rehabilitation.
Table 15 Production Comparison for Full- Depth AC Replacement, Four- Lane
Rehabilitation
Lanes Rebuilt Single Lane Double Lane
Layer Profile Profile
“ A”
Profile
“ B”
Profile
“ A”
Profile
“ B”
Avg. production1 2,103 lane- meters 1,694 lane- meters 1,914 lane- meters 1,524 lane- meters
Comparison2 100% 80% 91% 72%
Suspension ( hours) 3 1 hrs. 3 hrs. 6 hrs. 7 hrs.
1Stochastic analysis results
2Compared with Full- Depth AC Replacement Single- Lane Layer Profile “ A”
3Delay for AC cooling
72
The production capability for the Layer Profile “ B” ( 406- mm thickness) was about 80
percent of the Layer Profile “ A” ( 330- mm thickness) case. This reduction is similar to the extra
amount of asphalt thickness that is required for the Layer Profile “ B” ( 81 percent = Profile
“ A”/ Profile “ B” = 330 mm/ 406 mm). This suggests that the production difference was mainly
the result of the amount of existing pavement to be removed and the quantity of asphalt material
to be delivered.
Double- Lane Rehabilitation was about 10 percent less productive than Single- Lane
Rehabilitation for both Layer Profile “ A” and “ B.” In the concrete constructability analysis,
Double- Lane paving was more productive than Single- Lane paving because both lanes were
paved simultaneously and the constraints for Single- and Double- Lane paving were different.( 4)
The AC cooling time for Full- Depth AC Replacement for Double- Lane rehabilitation is much
longer than Singe- Lane rehabilitation ( See Table 15). For Double- Lane construction during the
55- hour weekend closure, the paving time required for each lift is much shorter than the Single-
Lane case, which results in more hours waiting for the previous AC lift to cool.
73
5.0 VALIDATION OF THE AC CONSTRUCTABILITY ANALYSIS MODEL ( I- 710
PROJECT)
5.1 Background of the I- 710 Project
Caltrans is in the process of constructing a demonstration project for the Long Life
Asphalt Concrete Pavement Rehabilitation Strategy ( LLACPRS) on Interstate 710 ( Long Beach
Freeway). The project construction was started in spring of 2001. The I- 710 project will be a
good case study for the validation and calibration of the asphalt constructability analysis model
described in this report, similar to the role the I- 10 project played for the concrete
constructability analysis, as described in References ( 6) and ( 7).
Given that the main reconstruction has not started yet, the asphalt constructability
analysis model will be used to predict the most probable production capability of the I- 710
project based on the preliminary design and planning information developed by Caltrans
engineers. The predicted production capability from the analysis model will be compared with
the production estimate developed by a committee of AC construction engineers from the
Southern California Asphalt Pavement Association ( SCAPA) and Caltrans.
As shown in Figure 32, the objective of the I- 710 project is to rebuild about 4.8 km ( 3
miles) of existing PCC pavement with asphalt concrete during a series of 55- hour weekend
closures ( approximately 12 consecutive weekends). The site, located on I- 710 between the
Pacific Coast Highway ( State Route 1) and Interstate 405, the freeway has three lanes in each
direction.
Crack Seat and Overlay ( CSOL) is the main rehabilitation strategy to be employed. The
site also includes four bridge structure underpasses under which AC ( Full- Depth AC
74
Figure 32. Site layout of the LLACPRS demonstration project on I- 710.( 11)
Replacement) will be placed to provide adequate clearance. Figure 33 shows the design profile
of the CSOL and Full- Depth AC Replacement portions of the project. The CSOL portion will
use the CSOL Layer Profile “ A” ( total AC thickness of 230 mm in four lifts), and the Full- Depth
AC section will excavate 430 mm of the existing pavement and replace it with 330 mm of AC in
five lifts ( Layer Profile “ A”) with 100 mm additional clearance for the new pavement system
under the bridge underpasses. The 4.8- km project length consists of a total of 2.8 km of CSOL
and 2.0 km of Full- Depth AC. Most of the rehabilitation work is planned to be completed during
3 months of weekend closures; the project schedule is shown in Figure 34.
Existing Profile
Removed for Full- Depth AC Under Bridges Retained AC
OR
101 mm ( 4 in.)
additional clearance
Existing Surface
Fabric
Portland Cement Concrete ( PCC)
203 mm ( 8 in.)
Subgrade
Cracked and Seated PCC
203 mm ( 8 in.)
Cement Treated Base ( CTB)
102 mm ( 4 in.)
Aggregate Base ( AB)
305 mm ( 12 in.)
1st Lift 55 mm 2 hours
Layer Thickness Cooling
Subgrade
2nd Lift 75 mm 4 hours
3rd Lift 75 mm 4 hours
Final Lift 25 mm 0.5 hours
Total Thickness = 230 mm ( 9 in.)
Typical CSOL
Subgrade
Aggregate Base ( AB)
178 mm ( 7 in.)
1st Lift 76 mm 1 hour
2nd Lift 76 mm 2 hours
3rd Lift 76 mm 6.5 hours
4th Lift 76 mm 1.5 hours
Final Lift 25 mm 0.5 hours
Layer Thickness Cooling
Total Thickness = 330 mm ( 13 in.)
Demolition = 431 mm ( 17 in.)
Under Bridges
Cement Treated Base ( CTB)
102 mm ( 4 in.)
Aggregate Base ( AB)
305 mm ( 12 in.)
Note: Longer cooling time may be required for some lifts due to paving at around noon. The difference in layer cooling times between
Figure 33 and Figure 10 ( showing typical cooling times) is due to the scheduling unique to the I- 710 project.
Figure 33. Proposed pavement profiles for I- 710 project.
75
76
Figure 34. I- 710 rehabilitation stage construction schedule.( 11)
All three lanes in one direction of the freeway will be closed and traffic will be switched
to the other side ( counter- flow traffic). Shoulder and median work for the traffic switch will
occur during a series of nighttime closures over the first 6 months of the project, as the project
schedule shows in Figure 34. Shoulders on both sides of the CSOL segments will be overlaid
simultaneously with the main traffic lanes. Caltrans will use a “ stage construction” concept for
the pavement rehabilitation; the 4.8- km project will be split into two equally divided segments in
each direction for a total of four segments, as shown in Figures 32 and 35. According to the
initial Caltrans plan ( Figure 35), two or three 55- hour weekend closures will be assigned for each
segment. During each closure, the entire segment being rehabilitated will receive the 230- mm
CSOL pavement and the 330- mm Full- Depth AC pavement underneath the two bridge structures
contained therein.
77
Figure 35. Schematic of the stage construction for the I- 710 project.( 11)
78
5.2 Predicted Production Capability for the 710 Project
The results of the stochastic analysis to predict the production capability on the I- 710
project for both the CSOL and the Full- Depth AC Replacement sections are summarized in
Table 16 and plotted in Figure 36. The predicted production capability for the CSOL portion
( 6.8 lane- km, 3 lanes overlaid) is similar to the typical production for the CSOL Full Closure
Layer Profile “ A” ( 6.8 lane- km, 4 lanes overlaid) shown in Table 8. For three lanes of CSOL
rehabilitation, there is negligible time lost to AC cooling delay.
Table 16 Stochastic Analysis for Proposed I- 710 Case Study
Design Profile CSOL Production( 1) Full- Depth AC Production( 2)
Mileage Centerline-meters
3 Lanes
( lane- meters)
Centerline-meters
3 Lanes
( lane- meters)
Lower bound 1,408 4,230 390 1,180
Mean ( Average) 1,544 4,638 500 1,490
Deterministic3 1,537 4,624 510 1,520
Upper bound 1,720 5,202 590 1,780
11 paving team
21.5 demolition teams
312 semi bottom dump trucks per hour; 10 dump trucks per hour
The predicted production capability of the Full- Depth AC Replacement option for the I-
710 project ( 1.5 lane- km, 3 lanes rehabilitated, 430 mm demolition depth, and 330 mm AC) is
less than the Full- Depth AC Replacement Layer Profile “ A” option shown in Table 14 ( 1.9 lane-km,
2 lanes rehabilitated, 330 mm demolition, and 330 mm AC). The main reasons for the
reduced production capability were 1) more material had to be demolished on the I- 710 project
( 430 versus 330 mm) to obtain additional bridge clearance, and 2) because of the short paving
distance, there was more delay due to AC cooling.
79
I- 710 ( CSOL Stochastic Analysis )
0
1,000
2,000
3,000
4,000
5,000
6,000
Centerline- meters Lane- meters
( three lanes)
Rehabilitation Option
Production
Lower bound
Mean
Deterministic
Upper bound
I- 710 ( Full- Depth AC Replacement
Stochastic Analysis )
0
500
1,000
1,500
2,000
Centerline- meters Lane- meters
( three lanes)
Rehabilitation Option
Production
Lower bound
Mean
Deterministic
Upper bound
Figure 36. Stochastic analysis for the I- 710 project.
The results of the predicted production capability from the asphalt analysis model for the
CSOL and Full- Depth AC Replacement sections were compared with the production
performance plan developed by the SCAPA/ Caltrans committee. The comparison between the
predicted production capability and the Caltrans planned production capability indicates that the
current performance target of the I- 710 project looks reasonable, but is somewhat “ tight” or
optimistic and doesn’t h