Docstoc

Economic Impact of Refining Dynamic Internal Angle of Superpave

Document Sample
Economic Impact of Refining Dynamic Internal Angle of Superpave Powered By Docstoc
					Economic Impact of Refining The Dynamic Internal Angle Of The Superpave® Gyratory Compactor
By Dr. George K. Chang, P.E. Project Manager The Transtec Group, Inc., 1012 East 38 ½ Street, Austin, Texas 78751 Phone: (512) 451-6233, Fax: (512) 451-6234 gkchang@thetranstecgroup.com Dr. Robert Otto Rasmussen, P.E. Vice President and Chief Engineer The Transtec Group, Inc., 1012 East 38 ½ Street, Austin, Texas 78751 Phone: (512) 451-6233, Fax: (512) 451-6234 robotto@thetranstecgroup.com Mr. Thomas Harman, P.E. Asphalt Pavement Team Leader Federal Highway Administration Turner-Fairbank Highway Research Center 6300 Georgetown Pike, HDRI-11, McLean, Virginia 22101-2296 Phone: (202) 493-3072 Tom.Harman@fhwa.dot.gov

2600 words, 4 Figures and 5 Tables = 4850 words

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

1

Economic Impact of Refining The Dynamic Internal Angle Of The Superpave® Gyratory Compactor
By Dr. George K. Chang, P.E., Dr. Robert Otto Rasmussen, P.E., and Mr. Thomas Harman, P.E. Abstract The purpose of the paper is to present qualitative as well quantitative comparisons of hot mix asphalt pavement performance designed on Superpave® gyratory compactors with different dynamic internal angles. The Superpave® gyratory compactor became the standard compaction instrument for HMA design in the mid-1990’s with the national adoption of the Superpave® system. Differences in compacted mixtures using Superpave® gyratory compactors from various manufacturers have become a concern of State agencies and suppliers in the asphalt industry. Through the analysis in this paper, the economic impact due to changes in the dynamic internal angles of Superpave® gyratory compactors has been shown to be significant. It has been determined that an increase in the DIA of 0.06°, from the target dynamic internal angle, results in a national increase in the life-cycle cost by as much as $2 billion annually. This economic impact justifies the need for a robust procedure to adjust and maintain dynamic internal angles of Superpave® gyratory compactors using the Federal Highway Administration Dynamic Angle Validator.

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

2

INTRODUCTION The purpose of the report is to present qualitative as well quantitative comparisons of hot mix asphalt (HMA) pavement performance designed on Superpave® gyratory compactors (SGC) with different dynamic internal angles (DIA). The Superpave® gyratory compactor became the standard compaction instrument for HMA design in the mid-1990’s with the national adoption of the Superpave® system. Differences in compacted mixtures using SGCs from various manufacturers have become a concern of State agencies and suppliers in the asphalt industry. Currently AASHTO defines SGC compactive effort based on the external angle of gyration (see Figure 1), where AASHTO T-312 (1) requires the external angle to be 1.25° ± 0.02°. Under this standard, HMA mixtures compacted by different SGCs may have significantly different bulk specific gravities (Gmb). Investigation of this problem has revealed that the DIA of the SGC is an important factor in defined compactive effort. The Federal Highway Administration (FHWA) has developed a Dynamic Angle Validator (DAV) to measure the DIA of SGCs. The FHWA Asphalt Team, working with the TRB Superpave® Mix/Aggregate Expert Task Group, has developed a draft AASHTO procedure for the incorporation of DAV into the Superpave® system. The ultimate goal is to use the DAV to adjust the dynamic internal angle of all SGCs to comply with a standard target value and a specified tolerance. In this report, the potential economic impacts due to changes to the DIA are assessed via mix volumetric design and pavement performance. INTERNAL ANGLE OF SUPERPAVE® GYRATORY COMPACTOR (SGC) Gyratory compaction employs a kneading effort to fabricate laboratory HMA specimens. Various versions/types of gyratory compactors have been invented and modified since the 1930’s (2, 3). The development of Superpave® gyratory compactor occurred during the Strategic Highway Research Program (SHRP) and is based on the concepts of the Texas gyratory shear compactor, US Corps of Engineers gyratory test machine, and the Laboratoire Central des Ponts et Chaussées (LCPC) Gyratory compactors (2). Through the standardization process, the static external angle of the SGC (a in Figure 1) was established at 1.25° ± 0.02 (1). Although the external angle is fixed in the specification, it was later determined that differences in the DIA among SGCs could be attributed to differences in compactive effort. Greater than allowable precision differences in bulk specific gravities (Gmb) between specimens compacted in different SGCs have been reported in the field as AASHTO T166-00 (4) precision states that duplicate specific gravity results by the same operator should not be considered suspect unless they differ by more than 0.02. The precision stated in AASHTO T166-00 (4) of 0.02 equates to a difference in calculated air voids of approximately 0.8 percent. This could be the difference between passing and failing test results.

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

3

Research has established an appropriate target and tolerance for the DIA of 1.16° ± 0.02° (5). Table 1 summarizes the adjustment range of external angles and the typical DIA when the external angle is fixed at 1.25° for the prevalent SGCs used in the United States. METHODOLOGY FOR THIS INVESTIGATION The approach adopted in this study is to use simple methods to demonstrate the economic impacts due to differences in the SGC DIA. Various scenarios were developed based on a range of SGC DIA values, which centered on a target value of 1.16°. For each scenario, Gmb at a given Ndesign revolutions was computed using a ratio of (difference in Gmb – to – difference in DIA). Assuming that a target 4% air voids was achieved by adjusting the asphalt binder content, the maximum theoretical specific gravity (Gmm) of mix was calculated based on Gmb at Ndesign to be 96% Gmm. Then, a volumetric calculation for each scenario was performed. A material model was then selected to predict the dynamic modulus of an HMA mix. Using a linear-elastic based structural response model, the tensile strain under the HMA layer, and the compressive vertical strain on top of subgrade, were calculated based on an assumed pavement structure. Pavement lives based on fatigue as well as rutting failures were then calculated using selected distress transfer functions. Based on additional assumptions, the life-cycle costs for each scenario were estimated and compared to one another. Although the above approach requires a number of key assumptions, the analysis was employed sound engineering fundamentals to provide a reasonable assessment of the potential economic impact due to differences in the SGC DIA. IDENTIFICATION OF MODELS NEEDED FOR THIS INVESTIGATION The models and assumptions used in this analysis are listed as follows: Conversion Between The SGC Dynamic Internal Angle (DIA) To The Bulk Specific Gravity Of HMA (Gmb) The relation between Gmb and DIA, adapted from (5), is expressed in equation 1.

DGmb » 0.15 DDIA
Volumetric and Gravimetric Conversion

(1)

For this analysis, the Gmm was assumed to be 2.500 for the case where DIA equals 1.16° (the “Base Case”). Changes in the volumetric binder content (Pb), by mass, were assumed to result in changes in air content (Pa), by volume, with the relationship in equation 2.

DPb = 0.1% ® DVa = -0.25%
where: Pb = Percent asphalt binder by weight of mixture

(2)

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

4

Va

= Percent air voids by volume of mixture

Gmb is assumed at Ndesign revolution for each scenario using the relationship in equation 3. Gmb = 0.96 × Gmm where: Gmm = Maximum theoretical specific gravity (Rice gravity) (3)

Conversion from Binder Content (Pb) to HMA Stiffness (E) Equation 4 presents the current form of the Witczak predictive equation (6). It is based on a database of 2750 dynamic modulus measurements from 205 different asphalt mixtures tested over the last 30 years in the laboratories of the Asphalt Institute, the University of Maryland, Arizona State University, and the FHWA.

log E = -1.249937 + 0.02932 r 200 - 0.001767( r 200 ) 2 - 0.002841r 4 - 0.058097V a æ Vbeff - 0.802208ç çV +V a è beff
where: E h f Va Vbeff r34 r38 r4 r200 = Dynamic modulus, 105 psi = Bitumen viscosity, 106 Poise = Loading frequency, Hz = Air void content, % = Effective bitumen content, % by volume = Cumulative % retained on the 19-mm sieve = Cumulative % retained on the 9.5-mm sieve = Cumulative % retained on the 4.76-mm sieve = % passing the 0.075-mm sieve

ö 3.871977 - 0.0021r 4 + 0.003958 r 38 - 0.000017( r 38 ) 2 + 0.005470 r 34 ÷+ ÷ 1 + e ( -0.603313- 0.313351 log( f ) - 0.393532 log(h )) ø

(4)

Assumptions for above parameters are as follows: h f r34 r38 r4 r200 = 19.6 ´ 106 Poise (assuming 70°F pavement temperature, A = 10.67 and VTS = -3.564) = 10 Hz = 22 % = 40 % = 58 % = 5%

An aged HMA modulus value at depth of 2 inches was used for subsequent analysis.

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

5

Structural Response Model WinLayer is a multi-layer linear elastic response software package developed by The Transtec Group, and based on ELSYM5. It was used to compute the stress and strain responses of an assumed HMA pavement structure. The pavement layer properties were assumed to be those in Table 2. A load level of 9,000 lb and tire pressure of 90 psi were used. Distress Transfer Models for Fatigue Cracking and Permanent Deformation Equation 5 from Asphalt Institute was selected to compute fatigue lives (7):

N f ( fat ) = 0.00432 ´ e t-3.291 ´ E AC
where: Nf(fat) C M et EAC

-0.854

(5)

= Number of load to failure (20% fatigue cracking over the entire pavement area, which relates to about 37% of the wheel path area.); = 10M ; =

ö æ Vbe 4.84 ´ ç ÷ ç V + V - 0.69 ÷ ; be ø è a

= tensile strain at bottom of HMA layer; = dynamic modulus of HMA, psi.

Equation 6 from Asphalt Institute (8), also adopted by NCHRP 1-37A and WesTrack for initial calibration, was selected to compute rut depth and a critical rut depth of 0.5 inch was used.

N f ( soil ) = 1.365 ´ 10 -9 ´ e v ( soil )
where: Nf(soil) ev(soil) Life-cycle Cost Model

-4.477

(6)

= Number of load to limit subgrade distortion to 0.5 inches; = Vertical compressive strain on top of the soil foundation.

For this analysis, the annual HMA production is estimated to be 500 million tons. The unit cost of HMA for the Base Case was assumed to be $40 per ton. The unit cost of asphalt binder was assumed to be $180 per ton. Thus, the unit costs of HMA for other scenarios can be adjusted based on differences in their binder contents as they deviate from the Base Case. The design life for Base Case was assumed to be 18 years. Relative design lives for other scenarios were calculated based on the design traffic for the Base Case. An analysis period of 40 years was selected, with the last construction for each scenario including a

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

6

partial reconstruction for the degraded portion estimated by the remaining life for that specific life cycle (after year 40). An annual discount rate was calculated based on an annual interest rate of 6% and an inflation rate of 3% using equation 7.

Rdis =
where: Rdis Ri Re

Ri - Re 1 + Re

(7)

= annual discount rate (0.0291); = annual interest rate (0.06); = annual inflation rate (0.03).

The Net Present Value (NPV) was calculated using equation 8.

NPV = å (1 + Rdis )
i =1

m

[

-Yr ( i )

´ F (i )

]

(8)

where: m i Yr(i) F(i) = number of construction activities (including full construction and partial construction); = construction activity count from 1 to m; = years from the original construction for construction activity i; = non-discounted construction cost for construction activity i.

ECONOMIC IMPACT ANALYSIS In this section, an analysis was performed that results in a relationship between changes in DIA of the SGC and the predicted life-cycle costs of pavements using HMA mixture design based on various SGCs. Based on assumptions of the total annual production of HMA in the United States, an estimate of the life-cycle cost differential was calculated as a function of the change in SGC DIA. Analysis Results The range of SGC dynamic internal angles considered was from 1.10° to 1.22° with a step of 0.03°, a tolerance specified by (5). The analysis results are presented in Table 3, Table 4, and Table 5. The economic impacts of changes in DIA are also shown in Figure 2, Figure 3, and Figure 4, in terms of NPV, differences in NPV, and percent differences in NPV, respectively. From Table 3, changes in Gmb from -0.009 to +0.009 result from deviations in DIA from -0.06° to +0.06°. Consequently, the Gmm changes range from -0.009 to 0.009 due to adjustments in asphalt binder content to maintain the target 4% air voids. The differences in calculated asphalt content range from +0.24

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

7

to -0.24. As a result, mix designers would need to reduce binder content for a given Ndesign using SGC with larger than standard DIA to maintain the 4% air voids. The HMA dynamic modulus (for an assumed aggregate gradation and field condition) is affected because of changes in the binder content. The tensile strain at the bottom of HMA layer, and vertical compressive strain at the top of the subgrade, is also shown to deviate slightly from the Base Case. Using the selected distress transfer functions, the load repetitions to failure for each scenario are presented in Table 4. Due to the assumed structure, the fatigue-based life is shown to be more critical than that based on rutting. However, normalized pavement lives, normalized to be multiplications of the Base Case, for both fatigue and rut-based analyses are presented for comparison purposes. predictions from fatigue-based and rut-based analyses should be considered separately. As shown in Table 5, the total number of constructions (including the initial construction) is presented for both fatigue-based and rut-based analyses. As the binder content decreases as DIA increases, the decreasing fatigue lives demands additional reconstruction events. The reverse is true for the increasing life (from rutting) due to decreasing binder contents – thus, less reconstruction is necessary. The changes in binder content also affect the unit cost of the HMA. Though not significant at first glance, these changes can be significant when summed. The annual HMA cost for the base year may differ by as much as $220 million for Scenario 1 (DIA = 1.1°), and the differences in life-cycle cost can be as much as $2 billion (or a 4.8% increase) for Scenario 5 (DIA = 1.22°) (for the fatigue-based analysis). SUMMARY OF FINDINGS An analysis of the potential economic impact due to differences in dynamic internal angles was performed. To complete the analysis, a number of assumptions were made. Employing a multi-layer linear elastic analysis model, the pavement responses were calculated for mixes “designed” with varying DIAs, and distress models were then used to predict pavement lives. According to the predicted number of loads to failure, fatigue was observed to be a dominant factor (as opposed to rutting) for the assumed cases. Assuming a base-case pavement life of 18 years, relative pavement lives were calculated for each scenario, both by fatigue and by rutting criteria. Life-cycle cost techniques were then used, based on a 40-year analysis period, in order to assess the relative life-cycle costs based on an assumed annual HMA production and unit costs of both the HMA and asphalt binder. From this analysis, it has been determined that an increase in the DIA of 0.06°, from the target DIA, results in a national increase in the life-cycle cost by as much as $2 billion annually. Although many assumptions were made for this analysis, the economic impact due to changes in the SGC DIA has been shown to be significant. This economic impact justifies the need for a robust procedure to adjust and maintain SGC DIAs using the Dynamic Angle Validator. Therefore,

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

8

REFERENCES 1. AASHTO T312 (formerly TP4-00), “Standard Practice for the Evaluation of Different Superpave® Gyratory Compactors (SGCs) Used in the Design and Field Management of Superpave® Mixtures,” Standard Specifications for Transportation Materials and Methods of Sampling and Testing, Part II – tests, 21st Edition, 2001. 2. Harman. T., Budowski, J.R., Moutier, F., Huber, G, McGennis, R., “The History And Future Challenges of Gyratory Compaction – 1939 to 2001”, The Transportation Research Board Meeting, 2001. 3. Huber, A. G., “Development of the Superpave® Gyratory Compactor” web document, http://www.utexas.edu/research/superpave/articles/gyr_hist.html, 1996. 4. AASHTO T166-00, “Bulk Specific Gravity of Compacted Bituminous Mixtures Using Saturated Surface-Dry Specimen” Standard Specifications for Transportation Materials and Methods of Sampling and Testing, Part II – Tests, Twentieth Edition, 2000. 5. Al-Khateeb, G., Paugh, C., Stuart, K., Harman, T., and D’Angelo, J., “Target and Tolerance Study for the Angle of Gyration Used in the Superpave® Gyratory Compactor (SGC)”, The Transportation Research Board Meeting, 2002. 6. Andrei, D., M.W. Witczak, and M.W. Mirza. “Development of a Revised Predictive Model for the Dynamic (Complex) Modulus of Asphalt Mixtures.” NCHRP 1-37A Inter Team Report, University of Maryland, March 1999. 7. May, R.W., and M.W. Witczak, “An Automated Asphalt Concrete Mix Analysis System”, Proceedings of the Association of Asphalt Paving Technologists, Volume 61, Charleston, S.C., 1992. 8. Asphalt Institute, Thickness Design – Asphalt Pavements for Highways and Streets, Manual Series No. 1, 1991.

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

9

Table 1: Allowable External Angles and DIA when External Angle is Fixed at 1.25° SGC Manufacturer/ Models Pine Instrument Company (AFGB1A, AFG1A) Pine Instrument Company (AFGC125X) Troxler (4140, 4140-B)* Troxler (4141) Rainhart (144) ELE International – Soiltest (ELE Gyrotest, Servopac) * ** *** Troxler supplies a True Mold Angle Device (TMA) that can be used to verify and calibrate the external angle of gyration of SGC. N/A: Not available. (reference 5) 0.50 to 2.0° set at 1.25° N/A** 0.0 – 3.0° (± 0.02°) 1.14°*** 1.1° 1.2° N/A 0.50 to 2.0° 1.18° *** Adjustment Range of External Angles set at 1.25° Typical DIA when the External Angle Fixed at 1.25° 1.2°

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

10

Table 2: Assumed Pavement Structure for the Analysis Layer Type HMA Granular Subbase Subgrade Thickness (inches) 6 12 (assumed to infinite) Poisson’s Ratio 0.4 0.35 0.45 Modulus (psi) predicted by equation 4 70,000 15,000

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

11

Table 3: HMA Volumetric, Mechanical Properties, and Responses Scenarios SGC Internal Angle, DIA (degrees) HMA Bulk Specific Gravity Gmb@Ndesign Maximum Theoretical Specific Gravity Gmm Percent Volume of Air Void Percent Mass of Binder Effective Percent Mass of Binder Deviation from Pbe @ DIA=1.16 Effective Percent Binder by Volume HMA Dynamic Modulus Tensile Strain @ bottom of HMA Vertical Strain @ top of subgrade Va (%) Pb (%) Pbe (%) DPbe(%) Vbe (%) EAC (psi) (microstrain) (microstrain) 1 1.10 2.391 2.491 4.00 5.71 5.31 0.24 12.51 61.8 162 2 1.13 2.396 2.495 4.00 5.59 5.19 0.12 12.24 61.6 161.7 3 1.16 2.400 2.500 4.00 5.47 5.07 0.00 11.98 61.4 161.4 4 1.19 2.405 2.505 4.00 5.34 4.94 -0.12 11.71 61.3 161.2 5 1.22 2.409 2.509 4.00 5.22 4.82 -0.24 11.45 61.1 160.9

2,063,000 2,071,000 2,079,000 2,088,000 2,096,000

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

12

Table 4: Predicted Pavement Lives Scenarios SGC Internal Angle, DIA Loads to fail - Fatigue Fatigue-based Loads to fail - Rut (SG) Rut-based Normalized Fatigue Life Fatigue-based Normalized Rut Life Rut-based 1 1.10 2,643,631 1.08 0.98 2 1.13 2,548,773 1.04 0.99 3 1.16 2,453,873 1.00 1.00 4 1.19 2,345,419 0.96 1.01 5 1.22 2,251,221 0.92 1.01

127,389,587 128,451,119 129,523,492 130,244,496 131,335,231

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

13

Table 5: Projected Life Cycle Cost Scenarios SGC Internal Angle, DIA Numbers of Constructions in 40 years Fatigue-based Numbers of Constructions in 40 years Rut-based Unit Cost of HMA (base year) Annual HMA Cost - base year Life-Cycle Cost in NPV (fatigue) Life-Cycle Cost in NPV (Rut) Diff Life-Cycle Cost in NPV (Rut) Diff Life-Cycle Cost in NPV (Rut) $/ton (million $) (million $) (million $) (million $) % 1 1.10 3.06 3.26 2 1.13 3.14 3.24 3 1.16 3.22 3.22 4 1.19 3.32 3.21 5 1.22 3.42 3.19

$ 40.44 $ 40.22 $ 40.00 $ 39.78 $ 39.57 $ 20,219 $ 20,109 $ 20,000 $ 19,891 $ 19,783 $ 38,848 $ 39,629 $ 40,453 $ 41,486 $ 42,410 $ 41,361 $ 40,905 $ 40,453 $ 40,081 $ 39,635 $ (1,605) $ (824) $ 908 -3.97 2.24 $ 452 -2.04 1.12 $ $ 0.00 0.00 $ 1,033 $ 1,957 $ (372) $ (818) 2.55 -0.92 4.84 -2.02

Diff Life-Cycle Cost in NPV (fatigue) (million $) Diff Life-Cycle Cost in NPV (fatigue) %

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

14

Figure 1: Illustrations of External and Dynamic Internal Angle of SGC

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

15

$43,000

$42,000

Life-Cycle Cost (million $)

$41,000

$40,000

$39,000

$38,000

$37,000 1.10 1.13 1.16 1.19 1.22

SGC Dynamic Internal Angles (degrees)
Fatigue Rut

Figure 2: Life-Cycle Cost Prediction

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

16

$2,500

$2,000

Difference in Life-Cycle Cost (millions $)

$1,500

$1,000

$500

$1.10 $(500) 1.13 1.16 1.19 1.22

$(1,000)

$(1,500)

$(2,000)

SGC Dynamic Internal Angles (degrees)
Fatigue Rut

Figure 3: Difference in Life-Cycle Cost Prediction

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.

Chang, Rasmussen, and Harman

17

6.00 5.00 4.00

Difference in Life-Cycle Cost (%)

3.00 2.00 1.00 0.00 1.10 -1.00 -2.00 -3.00 -4.00 -5.00 1.13 1.16 1.19 1.22

SGC Dynamic Internal Angles (degree)
Fatigue Rut

Figure 4: Percentage Difference in Life-Cycle Cost Prediction

TRB 2003 Annual Meeting CD-ROM

Original paper submittal – not revised by author.


				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:19
posted:7/31/2009
language:English
pages:18