Documents

Laboratory versus plant production impact of material properties and performance for RAP and RAS mixtures.pdf

Description
International Journal of Pavement Engineering ISSN: 1029-8436 (Print) 1477-268X (Online) Journal homepage: http://www.tandfonline.com/loi/gpav20 Laboratory versus plant production: impact of material properties and performance for RAP and RAS mixtures Reyhaneh Rahbar-Rastegar & Jo Sias Daniel To cite this article: Reyhaneh Rahbar-Rastegar & Jo Sias Daniel
Categories
Published
of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=gpav20 Download by:  [Birla Institute of Technology] Date:  07 July 2017, At: 00:09 International Journal of Pavement Engineering ISSN: 1029-8436 (Print) 1477-268X (Online) Journal homepage: http://www.tandfonline.com/loi/gpav20 Laboratory versus plant production: impact of material properties and performance for RAP andRAS mixtures Reyhaneh Rahbar-Rastegar & Jo Sias Daniel To cite this article:  Reyhaneh Rahbar-Rastegar & Jo Sias Daniel (2016): Laboratory versus plantproduction: impact of material properties and performance for RAP and RAS mixtures, InternationalJournal of Pavement Engineering, DOI: 10.1080/10298436.2016.1258243 To link to this article: http://dx.doi.org/10.1080/10298436.2016.1258243 Published online: 23 Nov 2016.Submit your article to this journal Article views: 123View related articles View Crossmark data  INTERNATIONAL JOURNAL OF PAVEMENT ENGINEERING, 2016http://dx.doi.org/10.1080/10298436.2016.1258243 Laboratory versus plant production: impact of material properties and performance for RAP and RAS mixtures Reyhaneh Rahbar-Rastegar and Jo Sias Daniel Department of Civil and Environmental Engineering, University of New Hampshire, Durham, NH, USA ABSTRACT Agencies are moving towards performance-based design methodologies for asphalt pavements, and different methods to evaluate the asphalt performance in the laboratory have been developed.  The laboratory performance can be evaluated at the mix design and/or production stages. A good understanding of differences in the behaviour of mixtures produced in the laboratory and plant is required to assess anticipated field performance at the mix design stage. The objectives of this paper are to compare the measured properties of plant-produced and laboratory-produced mixtures, to evaluate the effect of mixture variables on the differences observed, and to translate these to anticipated differences in fatigue performance through pavement evaluation using a linear viscoelastic layered analysis. In this study, 11 plant mixed, plant compacted, and their corresponding laboratory-mixed, laboratory-compacted mixtures are evaluated through binder and mixture testing. Mixture variables include aggregate gradation, binder grade and source, and recycled materials’ type and content. Performance grading on extracted and recovered binders, and complex modulus and SVECD fatigue testing on mixtures were conducted, and fatigue life was predicted using layered viscoelastic pavement design for critical distresses software. Most of the results show the laboratory mixtures are generally stiffer than the plant mixtures, but there is no constant shift for all mixtures. Larger differences are observed for the 19 mm and PG 58-28 mixtures and binder source appears to influence the differences as well. Different plants result in different effects on the properties of plant and lab-produced mixtures. This study provides a unique set of data that expands understanding of differences between laboratory and plant production of asphalt mixtures. © 2016 Informa UK Limited, trading as Taylor & Francis Group KEYWORDS HMA; laboratory mixtures; plant-produced mixtures; RAP; RAS; dynamic modulus; SVECD fatigue; pavement evaluation ARTICLE HISTORY Received 4 May 2016 Accepted 28 October 2016 CONTACT Jo Sias Daniel  jo.daniel@unh.edu 1. Introduction Te asphalt industry has been moving towards performance-based design, reinforced by federal legislation under the Moving Ahead Progress or the twenty-first Century Act. Many different meth-ods and approaches have been developed over the last several decades to evaluate the perormance o asphalt mixtures in the laboratory. Originally, most laboratory testing was performed on laboratory-abricated specimens; more recently, the differences between laboratory and plant production methods have been rec- ognised. An understanding of differences between the properties and perormance measured on specimens abricated in differ- ent ways is important for implementation of performance-based approaches. Tere are different methods to fabricate asphalt mix- ture test specimens; the most common ones are:ã Laboratory mixed, laboratory compacted (LMLC): the specimens are mixed and compacted in the laboratory using conditioning methods that are intended to simulate what happens in the plant and are generally used or mix design purposes (Kim et al.  2003).ã Plant mixed, laboratory compacted (PMLC): the spec-imens are abricated in the laboratory by reheating and compacting the loose mix produced at the plant.ã Plant mixed, plant compacted (PMPC): the specimens are compacted in a laboratory at the plant immediately ollowing production without reheating o the loose mixture.ã Field cores: the specimens are taken rom the asphalt pavement and are the best representation o in-place mix-ture conditions.Tese specimen abrication methods use different handling, mixing and compaction methods that can potentially impact the properties measured from the resulting specimens. Te handling and storage condition o asphalt binder in plant and lab are di- ferent. Te asphalt binders in the lab are kept in small containers at room temperature, while the asphalt is handled in enclosed systems at the plant and may result in differences in stiffness (  A Manual for Design of Hot Mix Asphalt with Commentary   2012). Another potential source o difference is the breakdown of aggregates that occurs during plant production and differences in mineral filler amounts that are added in the plant (  A Manual  for Design of Hot Mix Asphalt with Commentary   2012). Te tem- peratures to which the asphalt and aggregate are subjected are different in the plant vs. the laboratory and the method o com-paction is different rom lab to field as well.  2 R. RAHBAR󰀭RASTEGAR AND J. S. DANIEL ã Evaluate the impact o differences in measured properties on predicted atigue perormance. 2. Materials and test methods  2.1. Materials Tis study includes testing on 11 plant-produced (PMPC) and 11 lab-produced (LMLC) mixtures. Te PMPC specimens were abricated at two different drum plants in Lebanon, NH and Hooksett, NH. Te raw materials were collected or ab-rication o the LMLC specimens. Te Lebanon mixtures were placed in the field along New Hampshire (NH) State Route 12 near Westmoreland during the 2013 construction season. Te mixtures are varied in binder PG grade (PG 52-34, PG 58-28, PG 64-28), binder source, nominal maximum size o aggregate (NMAS) (9.5, 12.5 and 19 mm), recycled material type and binder replacement (16–32% RAP or RAP/RAS). able 1 shows the combinations evaluated, mix design volumetric information and actual binder replacement values. able 2 shows the aggre-gate gradation o different mixtures. Te RAP binder used in Lebanon plant had a continuous grade o 81.3–19.3 °C; this is a typical value or RAP materials in NH. Te stiffness o the Hooksett RAP was not measured, but is likely similar to the Lebanon RAP. Te RAS material is primarily tear-off shingles and could not be graded in the laboratory.  2.2. Binder testing Te asphalt binder rom each o the mixtures (both plant and laboratory mixed) was extracted in accordance with AASHO  164 (American Association of State Highway and ransportation Officials 2008-a) procedure 12 using a centriuge extractor and toluene solvent and recovered based on ASMD 7906-14 (American Society or esting and Materials 2014-b) using a rotary evaporator. Performance grading of the virgin binders and the extracted and recovered binders were conducted ollowing AASHO MP1-93. (American Association of State Highway and ransportation Officials 1993).  2.3. Mixture testing Complex modulus testing is a way to determine two important mixture properties: dynamic modulus and phase angle. Complex modulus testing was perormed on three cylindrical specimens o each mix ollowing AASHO P-79 (American Association Most studies conducted on plant and lab-produced mixtures show the lab-produced specimens are stiffer than plant-produced specimens. Johnson et al.  (2010), evaluated asphalt mixtures containing reclaimed asphalt pavement (RAP) and recycled asphalt shingle (RAS) and showed that the dynamic modulus (  E  ∗  ) o plant-produced specimens are lower than those of lab-produced mixtures. However, the research did not evaluate the impact o RAP/RAS on the difference between plant- and lab-produced mixtures. A pooled fund study on high RAP mixtures (Mogawer 2012) indicated that the reheating effect in PMLC mixtures causes them to be significantly stiffer than PMPC ones. Research performed on plant-foamed asphalt mixtures containing RAP (Xiao et al.  2014) showed the rut depth o PMLC specimens are lower than PMPC, indicating higher stiffness o the PMLC specimens. Researchers believe the reason is the effect o reheating in the lab, resulting in aged binder and stiffer materials. Also, the binder testing on recovered asphalt binder indicates the failure temperature of lab- produced specimens is higher than that of plant-produced ones (Xiao et al.  2014). On the other hand, the results o requency sweep testing (AASHO P7-94) on Michigan, Missouri, and Indiana mixtures (Mc Daniel et al.  2002) show that the stiffness of plant- and lab-produced mixtures of Michigan and Missouri is similar, while a higher stiffness ( G *) is observed or plant-produced mixtures o Indiana.Tere are many variables that affect the perormance o asphalt mixtures such as binder grade, binder source, gradation,  volumetric properties and recycled content and it is important to understand how these may potentially impact differences in measured properties o laboratory- and plant-produced speci-mens. Te scope o this paper is to evaluate PMPC vs. LMLC mixtures to compare the material properties that would be measured during the mix design phase and during production. Tis study provides new inormation on properties measured on binders and mixtures that has been rarely discussed in the literature. It is anticipated that the findings o this research will lead to a better understanding o differences between labora-tory- and plant-produced mixtures and would be a basic work or uture investigations.Te main objectives o this study are to:ã Compare the measured properties o plant-produced and lab-produced specimens.ã Evaluate the impact o mixture variables on the differ-ences between properties measured on PMPC and PMLC specimens. Table 1. Mixtures properties and volumetric information. PlantVirgin binder PG gradeBinder sourceNMAS (mm)%Total asphalt (P be )VMA (%)(VFA) (%)%Total binder replacement (% RAP/% RAS)Average % air PMPC specimensAverage % air LMLC specimens Lebanon58–28112.55.3 (4.7)15.574.918.9 (18.9/0)7.76.4112.55.3 (5.0)16.275.828.3 (28.3/ 0)7.46.52194.8 (4.4)15.071.320.4 (8.2/ 12.2)6.35.72194.7 (4.4)14.175.931.3 (31.3/ 0)6.05.652–34112.55.3 (4.7)15.574.918.9 (18.9/0)6.36.8112.55.3 (5.0)16.275.828.3 (28.3/ 0)6.96.43194.8 (4.4)15.071.320.4 (8.2/ 12.2)5.75.63194.7 (4.4)14.175.931.3 (31.3/ 0)6.15.7Hooksett58–28212.55.8 (5.5)15.979.522.4 (22.4/0)5.35.629.56.1 (5.7)16.578.921.3 (21.3/0)5.76.064–2829.56.1 (5.7)16.578.916.4 (16.4/0)5.96.0  INTERNATIONAL JOURNAL OF PAVEMENT ENGINEERING 3 o State Highway and ransportation Officials 2015). Te test-ing was conducted using asphalt mixture perormance tester (AMP) equipment in unconfined compression using our LVDs with a 70 mm gauge length to measure deormations. Te complex modulus data were analysed using Abatech RHEA® sofware, and the results are presented in the orm o dynamic modulus master curves represented using a generalised sigmoid ormat (Equation (1)) and phase angle diagrams to evaluate the relaxation capability o the mixtures. where  E    . is dynamic modulus; α ,  β , δ  ,   and γ  are fit coefficients; and ω  is reduced requency. Reduced requency is equal to re-quency used in the test multiplied by shif actor, T  , obtained rom Equation (2).where 1 , a 2  and a 3  are shif actor coefficients, and T   is temper-ature (Rowe 2009). Uniaxial tensile fatigue testing and analysis using the S-VECD approach was conducted on our specimens o each mixture ollowing AASHO P-107, (American Association o State Highway ransportation Officials (AASHO) 2014). Damage analysis for each mixture was performed and damage characteristic curves (DCC) were obtained using models avail-able within ALPHA-Fatigue sofware. Also, the atigue cracking resistance was assessed by fatigue failure criterion of asphalt mix-tures vs. number of cycles ( G R  − N   f  ). G R  is the rate of change in the averaged released pseudo strain energy (per cycle) throughout the test, and is calculated rom the Equation (3). where W  C   is released pseudo strain energy, and N   f   is the number o cycles beore ailure (Sabouri and Kim 2014).  2.4. Pavement evaluation Layered viscoelastic pavement design or critical distresses (LVECD) is a programme developed by North Carolina State (1) log   E  ∗   =    +  1  +   e   +    log (  )  1  (2) og a T   =  a 1 T   + a 2 T   + a 3 (3) G R =    f  0  W  RC  N  2  f  University to calculate responses and predict the atigue and rutting behaviour o asphalt pavements (Eslaminia 2012). o assess the atigue behaviour, this 3D finite element-based sof-ware employs SVECD approach. A DCC rom SVECD is used in this model. Since DCC is developed by removing the bulk  viscoelastic response o material rom the constitutive response, it can be used to evaluate the mixture’s response to any uniaxial loading history and temperature combination (Daniel and Kim 2002, Chehab 2003). 3. Results and discussion 3.1. Binder testing and analysis Te continuous high and low PG temperatures or the differ-ent virgin and extracted and recovered binders are shown in Figures 1 and 2, respectively. Te high PG temperatures rom the lab-produced mixtures are greater than those rom the plant-produced mixtures and there are slight differences with the different binder sources. Te two PG 52-34 virgin binders did not quite meet the required perormance grade on the low side. Te binders extracted and recovered rom Hooksett plant and lab mixtures show higher difference in high-temperature PG grade than Lebanon mixtures. Te difference between PMPC and LMLC mixtures is less pronounced on the low temperature side and all o the low grades are controlled by the m -value. In most cases for both 12.5 and 19 mm Lebanon-extracted binders, PMPC mixtures show colder temperatures, while two PG 58-28 binders extracted rom Hooksett LMLC mixtures have colder temperatures than PMPC mixtures.Te mixtures containing RAS have warmer temperatures than 31.3% RAP mixtures. Te stiffer, more highly aged material in RAS is what contributes to this difference observed. Also, the binders extracted rom the 19 mm mixtures have warmer temperatures than those extracted from the 12.5 mm mixtures for the same recy- cled material content. Te different binder sources or 12.5 and 19 mm may cause the difference in high- and low-temperature PG grade o extracted and recovered binders, so that warmer high temperatures of virgin binders from sources 2 and 3 used in 19 mm mixtures result in warmer high temperature of extracted and recov- ered binders from 19 mm than 12.5 mm. Te slightly higher actual binder replacement or the 19 mm mixture (31.3% vs. 28.3% or 12.5 mm) may contribute to the warmer temperatures as well. Table 2. Gradation information for different mixtures. Sieve size (mm)12.5 mm 28.3% RAP12.5 mm 18.9% RAP19 mm 31.3% RAP19 mm 20.4% RAP RAS12.5 mm Hooksett9.5 mm Hooksett% Passing 37.51001001001001001002510010010010010010019100100999910010012.598.698.683.483.498.91009.586.986.37070.386.5984.756059.247.246.357.9782.3641.741.532.43244.0621.1830.730.723.523.334.3490.621.121.316.21625.2350.311.411.49.3915.9220.156.15.95.24.78.0120.0753.93.93.33.14.688.5
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks