Noise reducing asphalt pavements : a literature review on requirements, evaluating methods and recent developments

Noise reducing asphalt pavements:

A literature review on requirements,

evaluating methods and recent developments

Ehsan Ghafoori

VTI r

apport 1022A

|

Noise r

educing asphalt pav

ements: A liter atur e r eview on r equir ements, e

valuating methods and r

ecent de

VTI rapport 1022A

Published 2019

VTI rapport 1022A

Noise reducing asphalt pavements:

A literature review on requirements,

evaluating methods and recent

developments

Author: Ehsan Ghafoori, VTI, http://orcid.org/0000-0002-5526-5896 Reg. No., VTI: 2017/0573-9.2

Publication No.: VTI rapport 1022A

Abstract

The demand for using asphalt pavements with noise-reducing properties is high. Recent experience of the durability of functional performance of this type of pavement in Sweden have been mixed. Hence, there are still needs for improvements until this type of pavement before it could be widely used in Sweden. This report is a summary of the existing knowledge about asphalt pavements with noise reducing properties and is expected to be used as a basis for future studies in vision of improving the quality of noise reducing asphalt pavements exposed to extreme weather and traffic conditions. The report contains five major sections which addresses the followings: Introduction of the existing noise reducing pavement types; the current standard and non-standard test methods used for evaluating this type of pavement; the existing guidelines and standards regarding noise reducing asphalt

pavements in different countries; summary of the recent developments and studies conducted on this topic; and finally the recommendations for research in vision of improving the durability of noise reducing asphalt pavements the risks for premature failures.

Title: Noise reducing asphalt pavements: A literature review on

requirements, evaluating methods and recent developments

Author: Ehsan Ghafoori (VTI, http://orcid.org/0000-0002-5526-5896)

Publisher: Swedish National Road and Transport Research Institute (VTI)

www.vti.se

Publication No.: VTI rapport 1022A

Published: 2019

Reg. No., VTI: 2017/0573-9.2

ISSN: 0347–6030

Project: Pre-study and literature review of noise reducing pavements

Commissioned by: Swedish road administration

Keywords: Noise-reducing asphalt pavements

Language: English

Referat

Behovet av att använda asfaltbeläggningar med bullerreducerande egenskaper är stort. De senaste erfarenheterna av att använda denna typ av beläggning i Sverige har varit både positiva och negativa. Därför finns det fortfarande behov av förbättringar till den här typen av beläggning innan den används allmänt i bullerkänsliga miljöer i Sverige. Denna rapport är en sammanställning av den befintliga kunskapen om asfaltbeläggningar med bullerreducerande egenskaper och förväntas användas som underlag för framtida studier för att förbättra hållbarheten hos bullerreducerande asfaltbeläggningar för nordiska förhållanden.

Rapporten innehåller fem delar som behandlar följande: Introduktion av befintliga bullerreducerande beläggningar, testmetoderna som används för att utvärdera denna typ av beläggning, befintliga riktlinjer och specifikationer för bullerreducerande asfaltbeläggningar i olika länder, sammanfattning av den senaste utvecklingen och studier som gjorts om detta ämne och slutligen rekommendationerna för forskning som syftar till att förbättra hållbarheten hos bullerreducerande asfaltbeläggningar motverka tidiga beständighetsbrister.

Titel: Bullerreducering av asfaltbeläggningar: En litteraturstudie om krav,

utvärdering av metoder och nya utvecklingar

Författare: Ehsan Ghafoori (VTI, http://orcid.org/0000-0002-5526-5896)

Utgivare: VTI, Statens väg och transportforskningsinstitut

www.vti.se

Serie och nr: VTI rapport 1022A

Utgivningsår: 2019

VTI:s diarienr: 2017/0573-9.2

ISSN: 0347–6030

Projektnamn: Förstudie och litteraturstudie avseende funktionella egenskaper för bullerreducerande beläggningar

Uppdragsgivare: Trafikverket

Nyckelord: Bullerreducering asfaltbeläggningar

Språk: Engelska

Foreword

This study is part of a project funded by Swedish Road Administration (Trafikverket) with the title of “Literature study on specifications for quiet pavements”. For this project, a group of experts, i.e. Andreas Waldemarson, Håkan Arvidsson and Ulf Sandberg from VTI, Thorsten Nordgren, Kenneth Lind and Robert Karlsson from the Swedish Transport Administration, Hans Lundkvist and Carl Hultin from Nynas, Lars Jansson from Peab, Jerry lngelström from Svevia, Peter Lundberg from Skanska and Hassan Hakim from NCC, participated in a workshop and shared their visions on this topic.

This report has been prepared based on the directions given by the experts in the mentioned workshop and is expected to help in upcoming projects dealing with the same subject.

Linköping, May 2019

Andreas Waldemarson Project leader

Quality review

Review seminar was carried out on 9 August 2019 where Björn Kalman reviewed and commented on the report. Ehsan Ghafoori has made alterations to the final manuscript of the report. The research director Björn Kalman examined and approved the report for publication on 14 November 2019. The conclusions and recommendations expressed are the author’s and do not necessarily reflect VTI’s opinion as an authority.

Kvalitetsgranskning

Granskningsseminarium har genomförts 9 augusti 2019 där Björn Kalman var lektör. Ehsan Ghafoori har genomfört justeringar av slutligt rapportmanus. Forskningschef Björn Kalman har därefter granskat och godkänt publikationen för publicering 14 november 2019. De slutsatser och

rekommendationer som uttrycks är författarens egna och speglar inte nödvändigtvis myndigheten VTI:s uppfattning.

Table of contents

Summary ...9

Sammanfattning ...11

1. Introduction to noise reducing pavements ...13

1.1. Single layer porous asphalt (PA) ...13

1.2. Double layer porous asphalt (DLPA) ...13

1.3. Stone mastic asphalt (SMA) ...13

Noise-reducing split mastics asphalt (SMA-LA) ...13

1.4. Thin layers...14

1.5. Very- and Ultra-thin surfacing (VTAC) ...14

1.6. Open graded friction courses ...14

1.7. Asphalt rubber friction course (ARFC) ...15

1.8. Poroelastic road surface (PERS) ...15

1.9. New generation OGFC (NGOGFC) ...15

1.10. Porous concrete ...15

Next generation concrete surface ...15

2. Evaluating test methods ...16

2.1. Aggregate tests ...16

2.2. Binder and mastic tests ...16

Binder rheological properties ...16

Stiffness of the cohesive and adhesive zones ...17

2.3. Mixture tests ...17

Draindown ...17

Air void content ...18

Compactability ...18

Indirect Tensile test (IDT) ...18

Water sensitivity ...19

Direct tensile strength test (DTS) ...19

Cyclic compression test ...19

Raveling/abrasion resistance tests ...20

Semi Circular Bending (SCB) test ...23

Resistance to rutting ...24

Permeability test ...24

2- and 4-point bending ...25

Coaxial Shear Test ...25

Superpave shear tester...26

Polishing resistance ...26

Skid resistance test ...27

Rolling resistance test ...27

Laboratory aging of the mix ...28

Adhesion/bonding with base layer or substrate ...28

Additional cohesion and adhesion tests ...28

3. Noise reducing pavements in different countries ...30

3.1. Netherlands ...32

Current requirements for porous layers in Netherlands ...32

3.2. Switzerland...38

Requirements for Semi-dense asphalt (SDA) in Switzerland ...40 3.3. Ireland ...43 Mixture requirements ...43 3.4. UK ...44 Structure ...44 Binder...44 Aggregate ...44 Mixture requirements ...44 3.5. Italy ...45 Bitumen ...45 Aggregates ...45 Gradation ...45

Drainage and sound absorbing requirements ...46

Mixture requirements ...46 Construction requirement ...46 3.6. Germany ...47 3.7. France ...47 Structure requirements ...47 Aggregate ...47 Gradation ...48 Bitumen ...48 Mixture requirements ...49 3.8. Belgium ...49 3.9. Denmark ...50

Structures and mixture requirements ...50

3.10. Finland ...50

3.11. US ...51

Gradation ...51

Mix design methods ...52

3.12. Japan ...52 Structure ...52 Aggregate ...53 Bitumen ...53 3.13. Summary ...53 Structure ...53 Bitumen ...54

Evaluating test methods ...55

4. Recent studies ...57

5. Recommendations ...61

5.1. Binder and mastics ...61

5.2. Gradation ...61

Theoretical ...61

Experimental ...61

5.3. Production and handling ...62

5.4. Construction ...62

Summary

Noise reducing asphalt pavements: A literature review on requirements, evaluating methods and recent developments

by Ehsan Ghafoori (VTI)

There is a great demand for using noise-reducing pavements in Sweden. Despite of the success achieved in some of the projects, the durability issues still prevents using noise-reducing pavements in a larger scale in Sweden. This lack of performance consistency calls for a systematic approach for providing detailed requirements in both material and mixture levels. It is also important to identify and utilize performance-based evaluating test methods that are suitable for simulating the load and

environmental conditions of Sweden. As an initial step, it is important to build up knowledge about the existing practices, standards and recent studies on this topic. Hence, this report was prepared with the aim of presenting an overview of the existing standards as well as recent advances on noise reducing pavements in different regions.

This report consists of five mains chapters as described below:

• _{in the first chapter, an overview of the most commonly used noise-reducing surface layers and} their advantages and disadvantages are briefly stated

• in the second chapter, existing standard and innovative test methods used and recommended for evaluating the quality of the noise-reducing pavements are briefly presented

• in the third chapter, the existing requirements for noise-reducing pavements in different countries are presented

• in the fourth chapter, recent studies about the causes of premature failures as well as solutions for improving the quality of highly porous asphalt pavements are summarized

• in the fifth and final chapter, a recommendation for future work with the aim of improving the durability the noise reducing pavements is presented

This report is expected to be used as part of the feasibility phase for improving the Swedish standard requirements for the noise-reducing asphalt pavements.

Sammanfattning

Bullerreducerande asfaltbeläggningar: En litteraturstudie om krav, utvärderingsmetoder och nyare utveckling

av Ehsan Ghafoori (VTI)

Det finns en stor efterfrågan för att använda bullerreducerande beläggningar i Sverige. Trots den framgång som uppnåtts i några av projekten som redovisas i rapporten hindrar hållbarhetsfrågorna fortfarande att bullerreducerande beläggningar används i större skala i Sverige. Denna brist på förutsägbar prestanda kräver ett systematiskt tillvägagångssätt för att tillhandahålla mer detaljerade krav på både material- och blandningsnivåer. Det är också viktigt att identifiera och använda prestationsbaserade utvärderingsmetoder som är lämpliga för att simulera påkänningar och

miljöförhållandena i Sverige. Som ett första steg är det viktigt att bygga upp kunskaper om befintliga metoder, standarder och senaste studier om detta ämne. Därför utarbetades denna rapport i syfte att presentera en översikt över de befintliga normerna samt de senaste framstegen på högporösa beläggningar i olika regioner.

Denna rapport består av fem kapitel som beskrivs nedan:

• i det första kapitlet beskrivs kortfattat en översikt över de mest använda bullerreducerande beläggningar och deras för- och nackdelar

• _{i det andra kapitlet presenteras både befintliga standards och innovativa testmetoder som} används och rekommenderas för att utvärdera kvaliteten på de bullerreducerande

beläggningarna

• i det tredje kapitlet presenteras de befintliga specifikationerna för bullerreducerande beläggningar i olika länder

• i den fjärde kapitlet sammanfattas nyligen utförda studier om orsakerna till tidiga

misslyckanden samt lösningar för att förbättra kvaliteten på högporösa asfaltbeläggningar • i den femte och sista kapitlet presenteras en rekommendation för framtida arbete med sikte på

att förbättra hållbarhet för de bullerreducerande beläggningarna.

Denna rapport förväntas användas som en del av genomförbarhetsfasen för att förbättra de svenska kraven för bullerreducerande asfaltbeläggningar.

1.

Introduction to noise reducing pavements

The concept of porous asphalt (PA) was first proposed in mid-1950s [Nicholls, 1998]. The initial idea was to avoid accumulation of water on the pavements during the rain to prevent aquaplaning. Later, it was found out that such pavements are also useful for absorbing the noise, coming from vehicles and in particular the interaction between vehicle tires and pavements. As demands for less traffic nuisance has increased, the need for using PA or other noise-reducing pavements has been significantly

increased. However, the main issue with the existing noise reducing pavements is in their serviceability as it is largely lower than the conventional dense graded pavements. Hence,

improvements in this field seem necessary. For making durability improvements in noise reducing pavements, it is important to build up an understanding about the existing knowledge in this field. Hence, in this section, the existing types of noise reducing pavements are briefly presented and some of their pros and cons are stated.

1.1. Single layer porous asphalt (PA)

This type of flexible wearing course layer typically consists of single size coarse aggregates, which forms the resisting skeleton of the layer, and very small amounts of the fine fractions plus

conventional or modified bitumen as binder. Typically, the air void in this type of structure after compaction is between 20 to 25% of its total volume. It has been reported that the interconnected air voids in this type of road has around 2 to 3 decibels higher noise absorption than those paved with average dense asphalt mixtures [Bendtsen and Gspan, 2017]. One of the major reported problems with the single layer PA that compromises its noise reducing properties is clogging [Ahmed, 2015]. More details and standard requirements for this pavement type is presented in the third chapter of this report.

1.2. Double layer porous asphalt (DLPA)

A double layer porous asphalt consists of a thin top layer and a thicker bottom layer. Obviously, the top layer consists of smaller aggregate fraction sizes than the bottom layer. The US version of this pavement normally has air void content between 15 to 19% whereas the European and Japanese ones have 20 to 30% void contents. This type of porous asphalt absorbs more traffic noise than the single layer PA, approximately between 4 to 6 decibels the first couple of years in service. Similar to the single layer, the double PA is also most suitable for high speed roads. During the construction of a double layer PA, a proper bonding between its top and bottom layers should be provided. The drawbacks of this type of pavement are the construction costs, its vulnerability to raveling at its top layer and clogging [Hamzah et al. 2013]. The minimum thickness for single- and double-layer porous asphalt under different loading conditions are normally recommended in standards.

1.3. Stone mastic asphalt (SMA)

A conventional SMA has a gap-graded gradation with considerable binder content and low air void with the lifetime of approximately 15 years. Normally, this type of mixture is used as a reference mixture when measuring the noise reducing properties of an asphalt layer. However, a new generation of SMA with smaller nominal maximum aggregate sizes and higher air void contents than the

conventional SMA mixture is proven to be effective for reducing the tire-pavement noise. [Praticò et al. 2012]. An example of this type of noise reducing asphalt is presented in the following.

Noise-reducing split mastics asphalt (SMA-LA)

The split mastics asphalt is commonly used in Austria as one of the noise-reducing asphalt

alternatives. This type of noise reducing asphalt has relatively lower portion of smaller aggregate sizes as compared with the conventional SMA. This results in the higher air void content in SMA-LA type which is between 9 to 11% by volume. This type of structure reduces the noise for about 2.5 dB as

compared with dense graded asphalt concrete and conventional SMA. Lowering the maximum aggregate sizes would also add to the noise reducing properties of this asphalt mixture type.

Depending on its maximum aggregate size, the thickness of a SMA-LA layer is between 20 to 40 mm. Comparing to PA, SMA-LA has higher durability and less maintenance costs [Bendtsen & Gspan, 2017].

1.4. Thin layers

This type of asphalt layer normally has a 15 to 40 mm of thickness. Its noise reducing properties lies in its smaller aggregate sizes, i.e. from 0/4mm up to 0/10 mm; for obtaining better noise reducing

properties, sometime the structure of this type of mixture is adjusted to become either open or semi-dense graded [Bendtsen & Gspan, 2017]. Thin layers are more often used as a maintenance technique in urban areas where the noise is a problem. The thin layer asphalt normally has shorter lifetime than the normal pavement types and are not recommended for intersections, parking lots and roundabouts [Kragh et al., 2011].

1.5. Very- and Ultra-thin surfacing (VTAC)

This type of layer is also used in urban areas where the porous asphalt cannot be used due to clogging problems and low shear resistance. This type of surfacing has an open-graded structure with the characteristics described in EU standard (EN 13108-2) as open-graded surface class 2. The structural difference between VTAC class 2 and PA is that the first one lacks medium size aggregates whereas the later one lack both medium and small aggregate sizes. Besides, the VTAC has the maximum aggregate sizes of 6mm or 10mm and a very thin lift thickness of 20 to 25 mm. The expected air void content in VTAC is about 18 to 25% with the sand content of 17–22%. In addition to urban areas, the VTAC has shown potential to be also used as emergency repair and for bridge decks [Praticò & Anfosso-Lédée, 2012].

Table 1 shows a comparison of the durability and noise reducing properties of different pavements based on their mechanical (bearing properties) and functional (surface properties) performances.

Table 1 Comparing the durability and noise reducing properties of different pavements [Praticò & Anfosso-Lédée, 2012]

Pavement type Durability Initial noise reduction

(dBA) Final / Min noise reduction (dBA)

Dense asphalt Variable 0 -2

Single layer PA 10 – 12 4 < 3

Double layer PA 9 6 4

SMA thin layer 9.5 4.7 3

Porous thin layer 8.5 5 3

1.6. Open graded friction courses

An OGFC is a layer of asphalt with uniform aggregate size and minimum of fine fraction particles. The service life of this type of road is expected to be around 8 to 12 years. The air void content of this pavement type is less than 20%, between 15 to 22%, in the US [Alvarez et al. 2006]; the minimum thickness of OCFC is about 1.5 times the maximum aggregate size and the air void content [Cooley et al. 2009]. Raveling is a very common type of failure for this type of pavement in cold regions.

1.7. Asphalt rubber friction course (ARFC)

This type of asphalt mixture normally has a gap- or open-graded structure containing crumb rubber. It has been used as the wearing layer mainly in US, Spain and Portugal. Despite the observed variety in performances, the overall results have shown promising results in terms of noise reduction. The range of noise reduction in this type of pavement is up to 6.7dBA. However, this type of pavement is more expensive than the normal pavements [Sandberg et al., 2011].

1.8. Poroelastic road surface (PERS)

Poroelastic concept was developed many years ago in Sweden. Then, it was improved in Japan and studied more within an EU project, called “persuade”. Normally, PERS contains 20 to 40% air void. The rubber used in this mixture comes from crap tires. Despite of its high research value, it still requires more developments until it becomes commercially available. PERS contains almost 20% rubber in volume and the aggregates and the rubber are bounded by polymer modified binder or polyurethane binder. The Swedish-Japanese studies have shown that this type of pavement can reduce the noise between 5 to 15dBA compared to the conventional dense graded asphalt layers. Low

interlayer adhesion, low skid resistance were the main issues raised with this type of pavement. In addition, during wintertime in cold regions snow removal machines can destroy the PERS easily [Sandberg and Goubert, 2011].

1.9. New generation OGFC (NGOGFC)

The open graded friction course (OGFC) first was introduced in 1950s and used for many years as permeable and noise reducing pavement layer. In spite of good noise reducing properties, this type of pavement was proven to have significantly lower durability than the normal courses. Hence, in 1999, NCAT recommended a new type of the open graded asphalt so called New Generation open graded friction course (NGOGFC). In the new generation, the highly modified binder and modified gradations were introduced [Mallick et al., 2000].

1.10. Porous concrete

Similar to porous asphalt, porous cement is made of gap- or open-graded aggregate structure. A single- or double-layer porous concrete, so called ModieSlab, has been developed in Netherlands allowing 6 to 7 dB noise reduction as compared with the reference mixture [Kragh, 2009 and Gibbs et al. 2005].

Next generation concrete surface

This type of surface has been introduced for providing noise reducing properties in concrete

pavements. In this technology, the surface of a concrete layer is grinded by means of diamond blades which results in reducing the noise between 3 to 6dBA [Rasmussen et al. 2004].

In the next chapter, some of the important test methods for evaluating the quality of the porous asphalt mixtures are briefly presented.

2.

Evaluating test methods

2.1. Aggregate tests

In all types of asphalt mixtures, the quality of the aggregates as the main skeleton of the road structure is very important and prevents premature failure of the pavements. This is especially more sensitive for porous asphalt. Below there is a list of tests that is also used for evaluating the quality of porous mixture aggregates:

Crushed faces (C) [EN 933-5]

Polished Stone Value (PSV) [EN 1097-8] Stone resistance (LA) [EN 1097-2]

Water Absorption [EN 13755]

Grain size distribution [EN 933-1] Aggregate abrasion value (AAV) [EN 1097/8] Flakiness index (FI) [EN 933-3] Frost and thaw resistance (F) [ASTM C666]

Micro Deval (MDE) [EN 1097-1]

Grading coarse (Gc) [ASTM C136]

Soundness [ASTM C88]

Soft particles in coarse [JIS A 1126]

Apparent density [ASTM C-127]

Stripping [IS:6241-1971]

Specific gravity (G) [AASHTO M 132]

2.2. Binder and mastic tests

Due to high sensitivity of porous asphalt mixtures to raveling, the adhesion and cohesion properties of the binder and mastics in such mixtures are very important. Hence, in addition to normal evaluating tests such as penetration, softening point and etc. used for testing binders, more tests have been used for assessing porous asphalt binders. Two of these tests are presented below.

Binder rheological properties

Dynamics Shear Rheometer (DSR), AASHTO T315,has been frequently used for obtaining rheological properties of bitumen. This test enables measuring elastic and viscous behaviors of an asphalt binder at high and medium temperatures. Figure 1 shows a schema of the test and an example of the device.

Figure 1. Schema of the DSR test method [Majidi et al., 2016].

Stiffness of the cohesive and adhesive zones

One of the main causes for the low durability of noise reducing mixtures is raveling. It is believed that the raveling is caused by the weak cohesion within the mastics and/or low adhesion between the mastics and aggregates [Herrington et al., 2005]. Hence, DSR has been utilized for measuring these properties for improving the durability of asphalt mixtures. Figure 2 shows a schema of the mastics tested by DSR.

Figure 2. DSR setup used for testing mastics for obtaining stiffness of the cohesive and adhesive zones [Mohan, 2010].

2.3. Mixture tests

Draindown

This test is used for determining whether the amount of bitumen drainage form a given mixture is acceptable or not. The range of acceptance for draindown is mentioned in different standards, e.g. AASHTO T305-09, EN 12697-18 or ASTM D6390-11. This type of test has been used for different asphalt mixtures especially noise-reducing mixtures to assure avoiding the premature failure of such mixtures.

In this test, a portion of a loose mixture is placed in a basket and placed in an oven, Figure 3, at the temperature of 175°C for an hour and then the amount of the drained bitumen from the mixture is measured with a simple formula.

Figure 3. The draindown test basket containing an asphalt mixture placed in the oven [Brown and Mallick, 1995].

Air void content

For measuring the air void content of asphalt mixtures, the European and US Standards, EN 12697-8 and AASHTO T 269, are used. For porous asphalt, the interconnection of the voids is very important and determines the permeability of these mixtures. The old Swiss standard, SN 640 433b, has been proposed in the literature [Poulikakos, et al., 2006] for determining the interconnected voids in highly porous mixtures.

Compactability

The quality of construction is significantly influenced by the compactibility of mixtures as well as the weather conditions during construction. In case of PA and thin layers, used as noise reducing

pavements, the thickness is a very important parameter as the thinner layers are more sensitive to the cooling of the mixtures during the construction. Hence, laboratory scale researches have been focused on the compactibility of the PA mixtures and thin layers. Both Marshal and Gyratory compactors as the two well-known laboratory compactors are used for conducting compactability measurements. Normally, for compactability measurements, asphalt mixtures are compacted right after mixing and also after being aged.

Indirect Tensile test (IDT)

Indirect Tensile Test is used for determining the tensile strength (stiffness) and the resilient modulus of asphalt mixtures at different temperatures. These tests are carried out based on the European and US standard procedures, EN 12697-23 and the ASTM D 4123. Figure 4 shows an IDT the test setup.

Water sensitivity

2.3.5.1. Indirect Tensile Strength Ratio (ITSR)

The water sensitivity test is normally carried out on laboratory compacted specimens using the indirect tensile test setup at the temperature of 25ºC. This test is also part of the European standard, EN 12697-12,and is proven to be suitable for porous asphalt. Based on almost all standards, e.g. SN 640 431-7NA, the ratio between the results of the ITS on dry and wet specimens should not be lower than 70%.

2.3.5.2. Duriez

The Duriez test [Duriez and Arrambide, 1961], EN 12697-12, used mostly in France, is another test for the water sensitivity measurements; however, there is still no field validation regarding this type of test (Figure 5).

Figure 5. Duriez test setup [CEDR, 2012].

Direct tensile strength test (DTS)

The direct tensile strength test (DTS) enables to investigate and capture material’s strengths responses. Using the direct tensile test at low temperatures has been claimed to be suitable for simulating the rapture phenomena occurring at regions with high temperature falls [Populikakos et al., 2006]. Populikakos et al. in 2006 suggested adjusting the DTS for conducting measurements on porous asphalt, Figure 6. The results of their study showed that this test method was capable of capturing clear differences between different asphalt mixtures. Besides, this test has been used for measuring adhesive strength between the thin layer and its bottom layer.

Figure 6. The direct tensile test setup adjusted for PA [Populikakos et al., 2006].

Cyclic compression test

Cyclic compression test has been suggested for PA using the European standard EN 12697-25 with cyclic axial loading. With the help of this test, the E* dynamic modulus of specimens with unconfined

as well as confined conditions can be measured. Figure 7 shows different test setups for confined and unconfined conditions.

Figure 7. The test setups for cyclic compression test [Populikakos et al., 2006].

Raveling/abrasion resistance tests

As it was mentioned before, raveling is the most common type of failure in PA mixtures. Due to its high importance, there has been different suggestions regarding evaluating the resistance of such mixtures against this type of failure.

2.3.8.1. Cantabro test

The standardized cantabro test, EN-12697-17/ASTM D7064M-08, is the mostly used test method in researches for measuring the durability of porous mixtures. This test is carried out on laboratory produced specimens to evaluate the cohesive strength of the mixtures. This test consists of a rotating drum machine, Figure 8. The amount of weight loss of the specimens after the Cantabro test is used for the durability evaluation of that asphalt mixture. In spite of its high use, this test type does not seem to represent the field conditions and is only suitable for making comparison between mixtures. The cantabro test has been used for both newly prepared specimens and also for the conditioned ones, water-conditioned/oven-aged/ frost and thaw cycled [Alvarez et al., 2011].

Figure 8. Schema of Cantabro test [Tenza-Abril et al., 2014].

2.3.8.2. The Aachener Raveling Tester

This test is carried out on asphalt mixture slabs. In this test, the surface of a porous asphalt is loaded by two turning regular tires in order to cause particle losses, Figure 9. The amount of particle loss is then measured and used as an indicator of the particle loss resistance of that mixture.

Figure 9. Aachener ravelling test setup [Wang et al., 2017].

2.3.8.3. Darmstadt scuffing device (DSD)

This device is another test for simulating the raveling on different types of asphalt mixtures. For this method a slab is placed on a moveable platform and by means of a pneumatic tire rotation and translation movements are imposed to the surface of the slab (Figure 10). This method has been used for thin and very thin pavements as well as porous asphalt.

Figure 10. DSD Test setup [De Visscher & Vanelstraete, 2016].

2.3.8.4. Rotating Surface Abrasion Test (RSAT)

For measuring the mechanical stability of porous asphalt mixtures, a new test method so called the Rotating Surface Abrasion Test (RSAT) was developed [Hartjes, 2008], Figure 11.This type of test can be used for determining the rutting resistance in noise reducing mixtures [Watson et al., 2018]. This test method was developed with the aim of simulating surface damage of an asphalt layer at the end of its service life in the field. The test consists of normal and shear forces imposed by means of a massive rubber wheel to determine the durability of porous asphalt. The test is used for PA slabs at 20⁰C. The test lasts for 24 hours and then the amount of the loose materials is measured for

Figure 11. The Rotating surface abrasion test (RSAT) [Sandberg et al., 2011].

2.3.8.5. Tribometer

This test was developed in 2008 by LCPC in France [Hammoun et al., 2008] (figure12). Tribometer is used for determining the impact of the binder type on the resistance of the asphalt mixture against tangential forces. This test hasn’t been standardized in Europe yet.

Figure 12. the setup of the tribometer [Sandberg et al., 2011].

2.3.8.6. Prall

The Prall equipment, Figure 13, is often used for testing dense graded asphalt mixtures conforming to the European standard EN 12697-16:2004. In addition to the dense grade mixtures, it has also been used for other type of mixtures including PA for determining their wearing resistance.

2.3.8.7. Road simulator

The Road simulator, Figure 14, at VTI has been used to simulate the wear. The simulator consists of four-wheel axles mounted on a central rotating axle with the rotating speed up to 70 km/h (Figure 14). The diameter of the test ring is about 6 meters. This machine has a high accuracy as it enables

measuring the particle loss of the pavement and the tires separately with high precisions. Earlier studies at VTI have shown promising results when compared with the wear in the field [Jacobson, 1995]. This method is carried out according to the European standard EN 13863-4:2004. The test can be conducted within the speed range of 20 to 70km/h and at different temperatures using different tire types [Gustafsson et al., 2009] and is very suitable for both dense graded and highly porous asphalt mixtures.

Figure 14. Road simulator [Lundberg et al., 2017].

2.3.8.8. Rolling bottle test

The rolling bottle test, EN 12697-11, has been proposed for indication of the raveling resistance (Figure 15). This test allows investigating the adhesion between an aggregate and bitumen. However, it is difficult to relate its results to the performance of the mixture in the field.

Figure 15. Rolling bottle test device [Partl et al., 2018].

Semi Circular Bending (SCB) test

This test is used for determining the fracture toughness of different mixtures [Frigio et al. 2013]. The SCB is used for providing comparison among different mixtures rather than relating its results to the reality fracture in the field. The test is carried out for dry and also water-conditioned specimens, Figure 16.

Figure 16. the SCB test setup (left) dry and(right) submerged in water specimens [Frigio et al., 2013].

Resistance to rutting

The Hamburg wheel tracking test, EN 12697-22, and asphalt pavement analyzer (APA), ASSHTO TP 63, are used for investigating the potential for rutting in open graded and porous asphalt, Figure 17.

Figure 17. (left) Hamburg wheel tracking test; (right) Aasphalt Pavement Analyzer (APA) [Stuart et al., 2002, Shu et al., 2012].

Permeability test

2.3.11.1. In situ permeability test

Clogging is one of the major problems with porous asphalt layers which causes decrease of

permeability, EN 12697-40. Hence, the “permeameter”, Figure 18, is used in the field for measuring the water permeability of porous asphalt [Jacobson and Viman, 2015].

2.3.11.2. Laboratory permeability tests

In the laboratory the permeability of asphalt mixtures in horizontal and also vertical directions are measured as shown in Figure 19 [Poulikakos et al., 2006].

Figure 19. (left) vertical (right) horizontal permeability test setups [Poulikakos et al., 2006].

2- and 4-point bending

Two- and four-point bending tests, Figure 20, are normally used for measuring the complex modulus, to evaluate the structural behavior, as well as fatigue behavior, to gain insight of the pavement

resistance under repeated loads [Mackiewicz, 2012]. These tests have been used for evaluating a wide range of asphalt mixture and supports and has also been recommended for porous asphalt [Poulikakos et al., 2006].

Figure 20. Test setups for (left) two and (right) four point bending tests [Poulikakos et al., 2006 and Mackiewicz, 2012].

Coaxial Shear Test

The Co Axial Shear Test (CAST), Figure 21, was developed at EMPA and used in different researches for determining the complex modulus (E*) of asphalt pavements [Sokolov et al., 2005], Figure 21. In this test, “a shear load is applied perpendicular to the specimen’s circular surface with lateral

confinement provided by a metal ring surrounding the specimen”. This test method seems to allow loading with similar axis to the traffic. The surrounding confinement of the specimen in the test setup tries to simulate “a semi-infinite” field situation. The CAST has been used for dry and repeatedly water-conditioned specimens

.

Figure 21. (Left) Schema and (right) a cut view of a CAST specimen [Poulikakos et al., 2006].

Superpave shear tester

The Superpave shear tester, Figure 22, allows imposing both vertical and shear loads to the specimen [Bennert et al., 2004]. This test has been used for evaluating the quality of both dense and porous asphalt mixtures.

Figure 22. Superpave shear tester device [Bennert et al., 2004].

Polishing resistance

Reports declare that the polishing action due to the tires can lower the friction of porous pavements in wet conditions. Hence, it is important to choose high quality aggregates to provide reasonably high resistant PA to the polishing.

2.3.15.1. Wehner and Schulze test device

In order to measure the polishing resistance of PA the Wehner and Schulze test device (Figure 23) have been used for evaluating noise reducing asphalt pavements [Arampamoorthy & Patrick, 2011]. This test device consists of two parts enabling polishing and friction measurements.

2.3.15.2. NCAT polishing test device

This test (Figure 23) was developed in NCAT Auburn University for testing the abrasion and polishing resistance of mixtures in dry and wet conditions [Turner & Heitzman, 2013]. This test method can be used for different types of asphalt including the noise reducing ones.

Figure 23. (Left) Wehner and Schulze test device; (right) NCAT polishing device [Arampamoorthy & Patrick, 2011 and Turner & Heitzman, 2013].

Skid resistance test

As one of the standard safety evaluation parameters, skid resistance of PA is measured, EN 13036-4. The British pendulum tester is the method that is used for measuring this criterion, Figure 24.

Figure 24. British pendulum tester [Hadiwardoyo et al., 2013].

Rolling resistance test

There is no standardized method for measuring rolling resistance of tire on a pavement. For

performing rolling resistance test in the field, different methods were introduced [Bergiers & Vuye, 2012]. The most commonly used one is the one manufactured by TUG as shown in Figure 25.

Laboratory aging of the mix

Aging of the mixture depends on both the loss of cohesion in the binder and the loss of adhesion between the binder and aggregate. Aging and subsequent testing of binder alone is not a good predictor of how a mixture will behave due to the effect of the asphalt-aggregate interaction.

Therefore, instead of aging the bitumen before the mixing, different methods of aging has been applied to the mixtures after the mixture preparation [Herrington et al., 2005].

Adhesion/bonding with base layer or substrate

The testing principle and specific test methods described in the following paragraphs may be used for assessing the bond strength. For testing very thin asphalt layer, some adaptation of the techniques may have to be applied.

2.3.19.1. Torque test method

This type of test is used for evaluating the bonding between the thin layers and its adjacent layer. Using the Torque test method enables to check whether the type of the failure is cohesive or adhesive. This test can be done on both laboratory and field cores [Sandberg et al., 2011].

2.3.19.2. Shear testing

The Leutner test [Leutner, 1979] has been used for measuring the bonding between multiple layers in shear mode. This test has also been used for porous asphalt pavements [Zhang, 2017] (see Figure 26).

Figure 26. Leutner shear test [Collop et al., 2009].

Additional cohesion and adhesion tests

2.3.20.1. Vialit Pendulum test

The Vialit pendulum test (Figure 27) has been designed for measuring the cohesion properties of bitumen. This device is claimed to indicate the maximum cohesion of a tested bitumen sample and the temperature in which the bitumen showed its highest cohesion strength. It also records the temperature range for the minimum cohesion strength [Xu et al., 2016]

.

Figure 27. Cooper-vialit cohesion pendulum test setup [Xu et al., 2016].

2.3.20.2. Bitumen Bond Strength (BBS)

This test has been designed by Youtcheff, and Aurilio in 1997 for obtaining adhesive properties of asphalt binders. In this test method, the bond between a flat aggregate surface and an asphalt bitumen is measured. The components of the BBS test is shown in Figure 28.

3.

Noise reducing pavements in different countries

Low-noise road surfaces are more expensive than conventional dense asphalt concrete. They cost more than thin asphalt layers and almost half of the two-layer porous asphalt. Table 2 shows the average difference between the lifetime of different porous asphalt (PA) structures and dense graded asphalt pavements in different regions. The noise reducing asphalt mixtures are classified as [Sandberg, 2009]:

• _{Very good noise reduction x > 7.0} • Good noise reduction 5.0 < x < 7.0 • Noise reduction 3.0 <x < 5.0

Table 2. Approximate life span of serviceability between porous and dense asphalt layers in different Europe [Sandberg, 2009].

Life time (years)

Mixture Germany Switzerland Netherlands

1-layer PA 7 – 10 8 – 12 11 – 17*

2-layer PA 9 – 13*

Dense asphalt ≥ 16 15 12 – 18

* polymer modified bitumen

Based on influential parameters such as the environmental and traffic loads impacts in different regions, the problems that porous asphalt pavements are facing can vary. Table 3 summarizes the causes of failures in different regions of the Europe and US. Raveling and clogging the pores are the main problems in the Europe and in addition to these two problems ice removal and stripping are also observed in the US for porous / open graded asphalt pavements.

Table 3. The causes of PA failures in different parts of the Europe and US [Kandhal & Mallick, 1998 and Nielsen, 2006].

Continent Location Tpical problems

Europe

Austria Raveling

Germany Raveling

France Raveling

Netherlands Raveling & rapid aging Spain Raveling & clogging

UK Clogging & rapid aging

US

Alaska Ice removal

Colorado Sripping

Hawaii Raveling

Idaho Clogging

Iowa Ice removal

Kansas Ice removal

Louisiana Raveling

Maine Ice removal

Maryland Raveling

Minnesota Raveling & clogging Rhode island Raveling

South Dakota Clogging

Tennessee Sripping & ice removal

Virginia Stripping

reducing pavement surface. One of the projects was carried out in 2010 in Huskvarna and the other one in 2014 in Rotebro. The mix design for both projects was identical since the older project showed promising results. Some details about the Huskvarna project are reported in the literature [Jacobson et al., 2016] as follows:

• Length of the road 3.7 km. • Speed limit: 90 km/h.

• _{Motor way, 4 driving lanes, K1 right and K2 left side of the drive way}

• 40 000 m² double drain layers 11 + 16 mm, 20 000 m² single drain 11 mm (Figure 1) • Layer thickness double drain: top 30 mm, bottom 50 mm.

• _{Layer thickness single drain: 30 mm.}

• ADT: 20 000 – 30 000 vehicles (15 % heavy traffic).

• Distribution of traffic in the drive ways: 70 % in K1 and 30 % in K2 • _{Stone material in top layer: K1, Rhyolite, K2, diabase or Rhyolite.} • Stone material in bottom layer: diabase in both drive ways.

• PMB binder type Endura D1 (highly polymer modified binder complying with EN14023 for Pmb 75/130-65 with the properties in Table 4.) [Jacobson et al., 2016]

Table 4. Specifications of the polymer modified bitumen used in the Huskvarna project (Nynas©, 2013).

Property Method Unit Min Max

Penetration @ 25⁰C SS-EN1426 mm/10 80 120

Softening point SS-EN1427 ⁰C 75

Resistance to hardening @ 163⁰C

Weight loss SS_EN12607-1 % w/w 0.5

Softening point increase SS-EN1427 ⁰C 10

Maintained penetration SS-EN1426 % 75

Other properties

Ignition point SS-EN ISO 2592 ⁰C 220

Technical properties

Breaking point Fraass SS-EN12593 ⁰C -22

Deformation energy @ 5⁰C SS-EN13589 J/cm2 ₁

Elastic recovery @ 10⁰C SS-EN13398 % 90

After RTFOT (EN12607-1)

Elastic recovery @ 10⁰C SS-EN13398 % 80

The measured noise reductions as compared with conventional stone mastic asphalt (SMA) with nominal maximum aggregate size of 16mm were as follows.

• _{2010: noise-reduction 7 to 8dB(A)} • _{2011: noise-reduction 7 to 8dB(A)} • 2012: noise-reduction 6.5 till 7.5dB(A) • _{2013: noise-reduction 6 till 7dB(A)}

Figure 29. The structure of the PA used in Huskvarna, Sweden [Ahmed, 2015].

After 4 years of construction very minor particle loss on the surface was observed on the Huskvarna road. Although, most of the parameters in the Rotebro project were the same as the ones used in the Huskvarna, the recent inspection of the Rotebro road has revealed that the porous layer is majorly damaged due to high wearing and is planned to be removed and repaved after only 4 years from its construction. The main reason for the low durability of the Rotebro road lies in its higher traffic volume as compared with the Huskvarna road. Therefore, it seems necessary to improve the standard requirements and material specifications for constructing more durable porous asphalt mixtures in Sweden.

In the following, the existing standard requirements for porous asphalt from European and US Standard documents are briefly presented.

3.1. Netherlands

Netherlands have a long history of broadly using porous asphalt as compared with other countries. Almost 90% of the motorways in Netherlands are open graded asphalt mixtures.

Asphalt roads in Netherlands have undergone a development and have 4 generations in it. Dense asphalt concrete is the first generation. The open graded asphalt concrete, so called “zoab” forms the second generation of the roads in Netherlands. Zoab roads are about 3 dB(A) quieter than dense coatings [Van Vilsteren, 2017a].

On the Dutch highways, open asphalt concrete (zoab) is the most widely used. This type of pavement reduces traffic noise and increases the capacity of the road through less splashing and spraying water during rain. Zoab is a pavement with a high percentage of hollow space (approximately 20%) [Hofman, 2017]. The porosity of the zoab is achieved because the mixture contains a relatively large amount of coarse granulate (crushed stone) and relatively few finer components. The zoab has a good resistance to rutting and that it is comfortable to drive over. Since the beginning of 2007, a new improved variant has been developed, called sustainable zoab. This road surface is stronger and more durable than standard zoab. On the other hand, due to the open structure of this mixture type, the bitumen ages earlier than with dense asphalt concrete. Besides, with newly constructed zoab, there is the risk of insufficient skid resistance because the stones are still coated with bitumen on the surface of new zoab. [Van Vilsteren, 2017a].

In the following, details of the requirements regarding the porous asphalt in Netherlands is mentioned.

Current requirements for porous layers in Netherlands

3.1.1.1. Structures

o Top layer Zoab 4/8 with minimum thickness of 25mm

o Bottom layer Zoab 11/16 with the minimum thickness of 45mm [Van Vilsteren, 2017a]

Figure 30. Typical porous asphalt structures in Netherlands; (left): one-layer, (right): 2-layer.

3.1.1.2. Bitumen properties

• _{For one-layer ZOAB normally bitumen type of 70/100pen is used.} • For two-layer ZOAB polymer modified bitumen is demanded.

It is required for the bitumen to have at least 40% lower penetration than its grade after construction [Van Vilsteren, 2017a].

3.1.1.3. Bitumen content

• _{≤ 0.6% lower than the mix design after the construction}

• ≤ 0.2% lower for the average bitumen content in all specimens [Van Vilsteren, 2017a]

3.1.1.4. Stone type

The aggregates must at least fulfill requirement of the level 3 as listed below. • C100/0

• PSV ≥ 58 • _LA15 • WA241

• F2 [Van Vilsteren, 2017a]

3.1.1.5. Mixture requirements

The compositions of the porous mixtures with different NMAS are shown in Table 5 [Van Vilsteren, 2017a].

Table 5. Boundaries for preparing PA in Netherlands [Van Vilsteren, 2017a].

Passing sieve ZOAB 11 ZOAB 16 2L**--ZOAB 16 2L-ZOAB 8

22.4 mm - 100 100 16 mm 100 93-100 90-100 11.2 mm 91-100 70-85 100 8 mm 15-40 - - 90-100 5.6 mm - - - 4 mm - - - 2 mm 15-25 15-25 5-25 5-25 0.5 mm 0.063 mm (2-10) * (2-10) * (2-10) * (2-10) *

Bitumen 4.2% _(70/100pen) 4.2% _(70/100pen) 4.2% _(PMB) 5.4% _(PMB)

Vmin 25% 20%

* The filler smaller than 0.063 should contain 25% hydrated lime ** 2L in the table (Two-layer)

3.1.1.6. Air void content

• _{For the mix design the minimum requirements is 20% of air void content} • The constructed layer in the field

o ≤ 5% lower than the mix design void content

o ≤ 2% lower for the average void contents in all specimens [Van Vilsteren, 2017a]

3.1.1.7. Density

• _{After construction the mixture should reach 100% of the design density}

• _{After construction the maximum density difference between the design and reached densities} must lower than 0.3% [Van Vilsteren, 2017a].

3.1.1.8. Water sensitivity

• Minimum Indirect Tensile Strength Ratio (ITSR) should be at least 80 [Van Vilsteren, 2017a].

3.1.1.9. Noise reduction

• _{No noise requirements will be met through the composition [Van Vilsteren, 2017a].}

3.1.1.10. Skid resistance

Two test methods for skid resistance:

• _{Initial dry friction, brake deceleration test (100% slip) ≥ 6.5m/s}2 • Initial wet friction test (86% slip) ≥ 0.44 [Van Vilsteren, 2017a]

3.1.1.11. Evenness

• _{The maximum longitudinal evenness should be around 2% [Van Vilsteren, 2017a].}

3.1.1.12. Water drainage

3.1.1.13. Construction considerations

Dynamic roller compactor is forbidden due to possible crushing the stone skeleton and raveling initiation. The degree of compaction must be at least 97% [Van Vilsteren, 2017a].

3.1.1.14. Two-layer porous asphalt

• _{Faster laying process}

• Better bonding between layers • _{Less sensitive for bad weather} • More homogeneous temperature Disadvantages:

Due to lower voids content in interface of two-layer PA: • Lower noise reduction

• _{Danger for clogging} • Poor longitudinal evenness

3.1.1.15. Using shuttle Buggy

• Homogeneous mix (no segregation, no dripping of bitumen) • Homogeneous asphalt temperature

• _{Continuous laying process (no stops)} • No bumping of asphalt trucks • _{Faster laying process}

• _{Better quality} Disadvantages:

• _{Not suitable for small sites (city) [Van Vilsteren, 2017a]}

3.1.1.16. Early life skid resistance

Characteristic values for braking deceleration are shown in Table 6.

Table 6. Recorded deceleration of reference vehicles on different pavement surfaces [Van Vilsteren, 2017a].

Braking Open graded asphalt _{New Old} Dense graded asphalt _{New Old} Without ABS 5.4 m/s2 7.0 m/s2 7.0 m/s2 8.0 m/s2 With ABS 9.0 – 9.5 m/s2 9.5 – 10.0 m/s2

As shown in Table 5, in some cases the braking deceleration results do not fulfill the requirements. In order to solve this problem, the newly laid zoab is scattered with a small amount of fine, sharp material, i.e. > 6.5 m/s², gritting with 100 to 200 g/m² crushed sand, just before the first roller passage is recommended to improve the deceleration with no influence on other properties (Figure 31). The sand may mainly consist of crushed sand with maximum size of 3mm. On the basis of intensive research, this method has no negative consequences for noise reduction, water storage capacity and

sustainability. Hence, gritting/ sanding the PA is recommended in the Dutch standard since 2009. The other method to improve the skid resistance is to reduce the speed limit for the lane when it is newly laid and then increase it [Van Vilsteren, 2017a].

Figure 31. Gritting for improving the skid resistance properties [Van Vilsteren, 2018].

3.1.1.17. Quality control

For evaluating the quality of the newly built pavement the following actions are made. • Taking field cores from porous layer

o One from every 2000 m² o Two at every location • _Analysis o Layer thickness o Density o Voids o Bitumen percentage o Mix composition

o Skid resistance [Van Vilsteren, 2017a]

3.1.1.18. Climate considerations

In the following conditions the ZOAB shouldn’t be used:

• _{Extremely cold climates with studded tires or aggressive snow removal} • Locations with high shear stresses, e.g.

• _{Narrow curves} • _Roundabouts • Urban areas

• _{Junctions [Van Vilsteren, 2017a]}

3.1.1.19. Damage distribution in Netherlands

Based on the road inspections raveling is the most common damage on open graded asphalt layers in Netherlands, as shown in Figure 32.

Ravelling Cracking Evvennes s & Rutting Skid resistanc e

Figure 32. The distribution of the damages observed on PA in Netherlands [Bouman, 2017].

3.1.1.20. Maintenance

For the maintaining the surfaces subjected to particle loss, the holes and voids are filled in-situ with an open graded emulsion sand asphalt mixture, Figure 33. The recipe of the mixture is shown in Table 7.

Figure 33. An example of the maintained PA in Netherlands.

Table 7. Recipe of the in-situ mixture used for maintaining the damaged porous asphalt [Van Vilsteren, 2017b].

Sieve sizes Open emulsion sand asphalt mixture 1/3 _{Minimum Maximum} 5.6 mm 4 mm 97 2 mm 60 80 0.063 mm 2 10 Bitumen 5 8 Void ≥ 25%

3.1.1.21. Laying speed of maintenance

• _{Inlay Porous Asphalt 200 m/h} • Spraying rejuvenator 3000 m/hv

Serviceability of the Open graded asphalt layers after construction and maintenance in Netherlands: • After construction

o Fast lane 17 years • After maintenance

o Spraying rejuvenator 3 years

o Open Emulsion Sand Asphalt 3 years o Thin inlay 6 years [Van Vilsteren, 2017b]

Note: The costs of construction and maintenance for a double layer porous asphalt with the noise

reduction of 6dB in Netherlands are respectively 10% and 75% higher when compared with a single layer porous asphalt with noise reduction of 4dB [Van Vilsteren, 2017b].

Dutch road administration, “Rijkswaterstaat”, is investigating the possibility for an Ultra Quiet Road Surface (USW). The bar is set high here. The aim is to develop a road surface with a noise reduction of 10 dB compared to a dense coating and a service life of at least 7 years. According to an

international state-of-the-art study, this objective can be achieved with a so-called porous-elastic road surface containing a non-bituminous binder, such as a polyurethane resin, and rubber granules. This new coating is still under development and will only become available after 2020.

3.2. Switzerland

There are two different recipes available for highly porous mixtures in Switzerland, i.e. open graded and semi-dense asphalt (SDA) mixtures. Below, first, the standard requirements for open graded (porous) and then the SDA mixtures used in Switzerland are briefly presented.

Requirements for open graded mixtures

3.2.1.1. Binder type

In the Swiss standard for porous asphalt layer, both wearing and binder courses, using polymer modified bitumen is mandatory. The requirements for the bitumen and polymer bitumen must be according to the European standard EN 12591 and EN 14023. Table 8 contains recommendations for the choice of binders depending on the type of mixed material [SN 640 431-7NA].

Table 8. Recommended bitumen types for PA in Switzerland [Ongel et al., 2007].

Bitumen sort & type Road layers _{Wearing course Binder course Drainage layer} Road construction bitumen

50/70pen o

70/100pen +

Polymer modified & special bitumen

PMB 25/55-65 (CH-E) o o

PMB 45/80-65 (CH-E) + + o

PMB 65/105-60 (CH-E) + + o

Special bitumen o o o

+ Varieties that are usually used

o Varieties to be used depending on the demands of traffic and climate

Additives such as organic and mineral fibers, polymers, rubber additives, etc. may be used provided their suitability has been proven [Ongel et al., 2007].

3.2.1.2. Aggregates

The types of mix are based on the combined grain group 0/4 (possibly the grain groups 0/2, 2/4) and the grain groups 4/8, 8/11, 11/16, 16/22 and 22/32 according to EN 13043. The aggregates used for

requirements of EN 13043 (e.g. percentage of fractured surfaces of coarse aggregates) are to be distinguished according to layers and types of mix.

3.2.1.3. Gradations

The grain size distribution shall be determined on the basis of the initial test in accordance with EN 13108-20. The nominal values of the aggregate size distribution must be within the ranges given in Table 9.

Table 9. Recommended gradation curve limits for PA in Switzerland [Ongel et al., 2007].

Sieve analysis Road layers Wearing course Binder course Drainage layer

PA 8 PA 11 PA B 16 PA B 22 PA S 16 PA S 22 PA S 32 45 mm 100 31.5 mm 100 100 90–100 22.4 mm 100 90–100 100 90–100 16.0 mm 100 90–100 90–100 11.2 mm 100 90–100 15–35 15–65 15–60 8.0 mm 90–100 20–40 15–35 15–60 5.6 mm 4.0 mm 15–35 2.0 mm 10–17 8–15 7–14 6–13 7–20 6–20 5–20 0.5 mm 4–10 4–10 4–10 4–10 4–10 4–10 4–10 0.063 mm 3–5 3–5 3–5 3–5 3–5 3–5 3–5

3.2.1.4. Binder content

The binder content shall be determined based on the initial test in accordance with EN 13108-20. The binder content for the nominal composition should not be less than the minimum values Bmin indicated in Table 10. The binder content is based on the weighted average apparent grain density of the total grain fraction pa of 2.65 Mg/m3 _{[Ongel et al., 2007].}

Table 10. Minimum required binder content for PA in Switzerland [Ongel et al., 2007].

Road layers Gradation Minimum permitted binder content (%) Wearing course PA 8 _{PA 11} ≥ 6.0 _{≥ 5.5}

Binder course PA B 16 _{PA B 22} ≥ 4.0 _{≥ 3.5} Drainage layer PA S 16 PA S 22 ≥ 3.5 ≥ 3.0

PA S 32 ≥ 3.0

3.2.1.5. Air void content

Based on the gradation and the layer type, for each layer there is a minimum required porosity that is shown in Table 11.

Table 11. Minimum requirements for the air void content for open graded asphalt layers in Switzerland [Ongel et al., 2007].

Road layers Gradation Permitted air void content (%) Wearing course PA 8 _{PA 11} ≥ 16.0 _{≥ 18.0}

Binder course PA B ≥ 22.0

Drainage layer PA S ≥ 18.0

3.2.1.6. Water sensitivity

Table 12. Minimum required ITSR for open graded asphalt layers in Switzerland [Ongel et al., 2007].

Road layers Gradation ITSR (%)

Wearing course PA ≥ 70.0

Binder course PA B ≥ 70.0

Drainage layer PA S ≥ 80.0

3.2.1.7. Temperature range for production and construction

When using road construction bitumen, the mixture temperatures in all phases of the construction must be within the ranges listed in Table 13. The highest temperature applies for any points in the mixer. The minimum temperature of the bituminous mixture must be complied with upon delivery. The use of polymer bitumen or special bitumen may be subject to different temperatures.

Table 13. Recommended temperature range for using the conventional bitumen types in blending and construction [Ongel et al., 2007].

Penetration (1/10 mm) Allowable temperature ranges (⁰C)

50/70 140-175

70/100 140-170

Requirements for Semi-dense asphalt (SDA) in Switzerland

The second type of pavement used as noise reducing asphalt layer in Switzerland, i.e. semi-dense asphalt, has shown more promising performance and durability as compared with the open graded one. Hence, the contractors in Switzerland currently use this type of asphalt when drainage and sound absorption is required for a road. In the following, standard details about the semi-dense asphalt used in Switzerland is presented.

3.2.2.1. Binder

Only polymer bitumen should be used. The use of PMB (CH-E) is recommended [SNR 640436].

3.2.2.2. Air void content in Marshall test specimens

The required air voids contents of the Marshall test specimens and their limits are shown in Table 14.

Table 14. Characteristic void content and limit values for the voids content of the Marshall test specimens [SNR 640436].

Class -12 _{Volume (%)} -16 -20

SDA 4 12 16 20

SDA 8 12 16

Limit values of the voids content of Marshall specimens

SDA 4 10 – 14 14 – 18 18 – 22

SDA 8 10 – 14 14 – 18

3.2.2.3. Water sensitivity

The requirements for the ratio of the indirect tensile strengths ITSR are shown in Table 15.

Table 15. The ITSR requirements [SNR 640436].

Class -12 -16 -20

Volume (%)

SDA 4 ≥ 70 ≥ 70 ≥ 70

SDA 8 ≥ 70 ≥ 70

Table 16. Grain size distribution of the SDA mixtures [SNR 640436].

Sieve analysis

(mm) Passing mass (%) SDA 4 SDA 8

11.2 100 8.0 90 – 100 5.6 100 50 – 70 4.0 90 – 100 15 – 52 2.0 12 – 50 10 – 35 1.0 7 – 29 7 – 26 0.5 4 – 24 4 – 21 0.063 3 – 12 3 – 12

3.2.2.5. Minimum Bitumen content

The lower limits required for the bitumen content in the SDA mixtures are shown in Table 17.

Table 17. Guide values for the metered binder content Bmin [SNR 640436].

Class -12 -16 -20

Mass (%)

SDA 4 ≥ 6.0 ≥ 6.0 ≥ 6.0 SDA 8 ≥ 5.8 ≥ 5.8

3.2.2.6. Resistance to permanent deformation

The maximum proportional rut shall meet the requirements specified in Table 18.

Table18. Proof conditions and requirements for the proportional rut depth SN EN 12697-22.

Wearing type SDA*

Class -12, -16, -20

Conditions of the test

Thickness of the specimen (mm) 50 Temperature of the test (⁰C) 60

Number of cycles 30000

Requirements

Percentage of the rutting depth (%) ≥ 7.5 * No requirements for SDA 4

3.2.2.7. Production control

Table 19 shows the required evaluating tests for the Semi dense asphalt mixtures in Switzerland.

Table 19. Type and number of tests on semi dense mixed material SDA [SNR 640436].

Property Test method

Grain size distribution SN EN 12697-2

Binder content SN EN 12697-1

Air void content SN EN 12697-8 SN EN 12697–6, Method D SN EN 12697–5, Method A

Binder drainage SN EN 12697-18

Water sensitivity SN EN 12697-12

Resistance to permanent deformation SN EN 12697-22

3.2.2.8. Recommended varieties and classes

Table 20 classifies the types of mixed material (SDA4, SDA8) and their classes (12, 16, 20) on the basis of past experience.

Table 20. Recommended varieties and classes [SNR 640436]. Class -12 -16 -20 SDA 4 + + o SDA 8 + o + Recommended O Conditionally accepted

3.2.2.9. Recommended layer thickness

The recommended target value ranges of the layer thicknesses are contained in Table 21.

Table 21. Recommended nominal value ranges of the layer thicknesses depending on the type and the class of mixtures [SNR 640436].

Class -12 -16 -20

[mm]

SDA 4 20 – 35 20 – 35 20 – 35 SDA 8 25 – 40 25 – 40

3.2.2.10. Air void content (field compaction)

Based on the class and the gradation of the SDA mixtures compacted in the field, the Swiss standard requires different air void content limits as shown in Table 22.

Table 22. Limit values of the air void contents of the built-in layers [SNR 640436].

Class -12 -16 -20 Volume (%) Individual values SDA 4 10 – 20 14 – 24 18 – 28 SDA 8 9 – 17 13 – 23 Average values SDA 4 10 – 18 14 – 22 18 – 26 SDA 8 10 – 16 14 – 20

3.2.2.11. The compaction degree on the lower layer

The degree of compaction is defined as the quotient of the bulk density of a core or section of a layer and the bulk density of Marshall specimens of associated mix. The values in Table 23 show the requirements for the degree of compaction regarding different SDA mixtures.

Table 23. Requirements for the degrees of compaction of the layers [SNR 640436].

Class -12 -16 -20 (%) Individual values SDA 4 ≥ 97 ≥ 97 ≥ 97 SDA 8 ≥ 97 ≥ 97 Average values SDA 4 ≥ 98 ≥ 98 ≥ 98 SDA 8 ≥ 98 ≥ 98

3.2.2.12. Layer thicknesses

The allowable deviations of the layer thickness from the target thicknesses are as follows. • The average layer thickness calculated from the consumption of the mixture should not

deviate more than ± 10% from the target nominal thickness.

• _{The average density of the Marshall samples should be used for the calculation.}