A Critical Review Of Bituminous Mixes Used In India

A CRITICAL REVIEW OF BITUMIOUS PAVIG MIXES USED I IDIA By PRITHVI SIGH KADHAL*, V. K. SIHA** & A. VEERARAGAVA*** [This paper was published as Paper o. 541 in the Journal of the Indian Roads Congress, Volume 69-2, July-September 2008. Response to comments received on this paper are also given at the end.]

ABSTRACT A critical review of bituminous paving mixes used in India in accordance with the current MORTH Specifications (2001) has been made keeping in perspective the fundamentals of mix selection based on their intended functions in different courses within the flexible pavement. There is a proliferation of bituminous paving mixes in India. MORTH Specifications broadly provides 4 mixes for base courses, 6 mixes for binder courses, and 4 mixes for wearing courses. Further two grading, each of BM, DBM, SDBC and BC are specified in the MORTH specifications. Too many options for a specific bituminous course have created confusion in mix selection and are mainly responsible for the poor performance of flexible pavements in India. A case has been made on technical grounds to have only 5 dense graded mixes of different nominal maximum aggregate size (NMAS) in the specifications, as is the case in most developed countries of the world. The following 5 dense graded mixes have been accordingly proposed along with their recommended rut resistant gradations: • • • • •

37.5 mm NMAS DBM Base Course Grading 1 25 mm NMAS DBM Base Course Grading 2 19 mm NMAS BC Binder Course 12.5 mm NMAS BC Wearing Course Grading 1 (for heavy traffic) 9.5 mm NMAS BC Wearing Course Grading 2 (for light to medium traffic, urban areas, and thin application) 1. ITRODUCTIO

The bituminous paving mixes as specified in MORTH “Specifications for Road and Bridge Works”, Fourth Revision, 200128 are commonly used in India. Some of these mixes have evolved since 1960s, an era when the present day hot mix asphalt plants were not common and mixes were produced with small portable mixing plants with limited _______________________________________________________________________ * (Prof.) Associate Director Emeritus, National Center for Asphalt Technology (NCAT), Auburn University, USA (currently Jaipur 302 006) e- mail: [email protected] ** Secretary General, Indian Roads Congress, New Delhi –110 011 *** Professor of Civil Engineering, Indian Institute of Technology, Madras Chennai – 600 036

1

aggregate heating, blending and mixing capabilities. The proliferation of bituminous paving mixes as specified in the MORTH publication basically manifest the constraints of non-availability of modern hot-mix plant besides cost reduction of bituminous mixes. Today the scenario has substantially changed. There is a growing concern to construct long lasting pavements with minimum maintenance. It is accordingly, felt that only those specifications for bituminous mixes be allowed, which are durable, irrespective of their location (NH or SH or Rural Road) because durability and long term performance is central to all of them.

A critical review of commonly used bituminous paving mixes in India has been accordingly attempted keeping in perspective the following: (a) fundamentals of mix selection based on their intended function and location within the flexible pavement structure, and (b) capabilities of the present day hot mix asphalt plants. The gradations of some of the mixes also need to be updated based on proven field performance of similar mixes in the test tracks and in regions of developed countries like USA with climate reasonably close to that of India. This review is also expected to be helpful to the contractors who are currently designing their own projects (including flexible mix selection) under the Public-Private-Partnership (PPP) projects. The future trend of mix design improvements should finally aim to achieve long-lasting perpetual pavements. 2. FUDAMETALS OF MIX SELECTIO BASED O THEIR ITEDED FUCTIO AD LOCATIO WITHI PAVEMET STRUCTURE Bituminous mixes are used in a flexible pavement to serve the following three important functions: • • •

Provide structural strength Facilitate subsurface drainage Provide surface friction especially when wet

2.1 Provide Structural Strength Fig. 1(a) presents a typical cross-section of flexible pavement in a developed country like USA. The structural bituminous courses can consist of bituminous binder course and bituminous surface or wearing course as shown in Fig. 1(a). Providing structural strength is the primary purpose of most bituminous mixes except those used in very thin surfacing. The objective is to disperse appropriately the dynamic and static effects of

2

traffic wheel loads to the underlying pavement layer such as bituminous/crushed stone base course. For low-volume roads only a granular base and a bituminous wearing course may suffice based on structural requirements. Normally, lower layers of base courses including those of bituminous base courses (as provided in developed countries) should have desired stiffness characteristics to act as good foundation, which should be effective in dispersing the traffic loads to the lower layers. Upper layers of bituminous binder course has the requirement of being effective in re-bounding against the dynamic effect of traffic load. In other words, the top layers of bituminous binder courses should have adequate stiffness to resist rutting coupled with the flexibility to be effective in rebounding. The flexibility characteristics should, therefore, increase when going from bottom to upwards layer.

Fig. 1(a) Typical cross-section of flexible pavement in USA

From the perusal of the current literature and practices, it is observed that the preceding requirements can be fulfilled by using continuously dense-graded bituminous mixes with nominal maximum aggregate size (NMAS) decreasing from base course through binder course to surface or wearing course30. The nominal maximum aggregate size is defined as one sieve larger than the first sieve to retain more than ten percent of combined aggregate31.

Base course mixes, which use relatively larger size aggregate, are not only stiff/stable but also are economical because they use relatively lower bitumen contents. Surface or wearing course mixes with smaller aggregate on the other hand have relatively higher bitumen contents, which not only impart high flexibility but also increase their durability. The binder (intermediate) course mix serves as a transition between the base course and

3

wearing course. Several studies3,30,34 have shown that permanent deformation (rutting) within flexible pavement is usually confined to the top 100 to 150 mm of the pavement. This means both the binder and wearing course mixes should be designed to be resistant to rutting. That is why in extreme cases of heavy traffic loads and high tyre pressures, it is considered prudent to use Stone Matrix Asphalt (SMA) mix in which due to stone-onstone contact the load is carried directly by the coarse aggregate skeleton. An idealized cross-section for a heavy-duty flexible pavement is shown in Fig.1(b). Recently, the Indian Roads Congress (IRC) has adopted a tentative SMA specification, which could be used under such circumstances. Kandhal13 has published a manual on design and construction of SMA mixes, which is widely used in the USA.

Fig. 1(b) Idealized cross-section of flexible pavement using SMA

2.2 Facilitate Subsurface Drainage Typically, granular sub-base in a flexible pavement is intended to provide subsurface drainage. However, in many situations where granular sub-bases contain high percentages of fines (less than 75 micron size material) such layers are found to be not very effective. In developed countries like USA, Permeable Asphalt Treated Base (PATB) has been used extensively on major highways to provide positive subsurface drainage. PATB basically does not constitute a conventional base course. It is considered as a separate course exclusively for subsurface drainage. From cost consideration, PATB is not recommended for most highways in India where GSB alone should suffice. However, it is felt that a specification similar to those of PATB should be available for

4

use in exceptional drainage problem situations. The lift or layer thickness of the PATB generally ranges from 75 to 100 mm.

PATB is provided between the granular sub-base (GSB) and the bituminous base course. Figures 2(a) and 2(b) shows two typical provisions of PATB. The concept is to provide a two-layer drainage system. It is on the presumption that water accumulated in the subbase always seeks least path of flow and some part of it flows into the PATB rather than traveling altogether a long distance to the edge of the sub-base. Water collected in the PATB is then drained out in two ways. The PATB can either be connected to a subsurface pavement edge drain as shown in Fig. 2(a) or it can be extended all the way to the edge of the embankment or “daylighted” as shown in Fig. 2(b). This two-layer subsurface drainage system is very effective in quickly removing water, which may enter the pavement by any manner.

Fig. 2(a) Permeable Asphalt Treated Base (PATB) connected to pavement edge drain

Fig. 2(b) Permeable Asphalt Treated Base (PATB) daylighted

5

It is of paramount importance to ensure that the PATB layer is not blocked in any way otherwise it will be saturated with water and thus create a “bath tub” scenario in the pavement. This phenomenon will not only induce stripping21,22,24 within the PATB (due to pore water pressure buildup under traffic) but also in the bituminous courses overlying it. Premature failures of pavements from this phenomenon have been experienced in California. In India, open-graded permeable Bituminous Macadam (BM) is used as binder/base course in pavements without any positive outlet for water, which can result in such failures. This will be discussed later when the BM mix is reviewed.

2.3 Provide Surface Friction Mixes like bituminous concrete, semi-bituminous concrete, besides premix carpet, mixseal surfacing is commonly provided as per MORTH Specification as wearing course. Some bituminous wearing course mixes are designed to provide high surface friction especially when the pavement is wet. Open-Graded Friction Courses (OGFC) usually 2025 mm thick (Figure 1b) are designed as an open graded mix with interconnected voids that provide drainage during heavy rainfall. OGFC reportedly provides the following advantages12: •

Reduce splash and spray

•

Reduce tyre-pavement noise14

•

Enhance visibility of pavement markings, and

•

Reduce night time surface glare in wet weather

It is important that the bituminous layer underneath the OGFC is very dense, impermeable, and highly resistant to stripping. The design, construction, and maintenance of OGFC is fully developed and well documented by Kandhal in manual and papers12,18,27. The OGFC has been known to induce stripping in the underlying bituminous layer15,22. OGFC basically is a specification for cold country having skid problem. It involves additional cost and hence may be used only on highways, which are accident-prone during rains. This course should be constructed with proper anti stripping agents like hydrated lime.

If an existing pavement with OGFC needs to be overlaid, it is necessary to first remove (mill off) the OGFC and also examine the underlying layer for potential moisture damage

6

(stripping). Open graded wearing courses like premix carpet, surface dressing and to some extent semi-dense bituminous concrete falls under the above category. In case of overlay these wearing course layer should be totally removed, which is often not done in India. Experience in the US has shown that if an existing OGFC layer is not removed it will trap water and cause premature moisture damage by stripping, particularly when the underlying layer is a dense mix. Photo 1 and 2 exhibits such failures in Australia and Oklahoma (USA)15.

Photo 1. Premature OGFC related distress on a highway west of Sydney, Australia (Ref. 15)

Photo 2. Premature OGFC related distress on a highway in Oklahoma, US (Ref. 15)

7

3. PERPETUAL PAVEMETS – A EMERGIG COCEPT The concept of Perpetual Pavement was first launched by the Asphalt Pavement Alliance (APA) in a joint promotional effort with Asphalt Institute, National Asphalt Pavement Association, and the State Asphalt Pavement Associations of USA in 2003-04. Fig. 3 depicts the perpetual design concept.

Fig. 3 Perpetual pavement design concept (HMA = hot-mix asphalt) (Source : US Department of Transportation FHWA) From the perusal of the Fig. 3 it will be observed that the bituminous portion is divided into three zones and bitumen mixes to be adopted for these three zones are characteristically different. The first zone is wearing course 40-75 mm thick. It could be high quality HMA or OGFC. The succeeding zone/layer is high compression zone with high modulus rut resistant mix 100-200 mm thick. The third layer is to cater maximum tensile strain and should be able to resist flexible fatigue. The design of the bituminous layers is done on mechanistic principles by keeping the strain within each layer less than endurance limit. Thus no damage accumulation takes place in any layer and the pavement layer constructed, normally do not need any replacement/rehabilitation. It is only the top layer, which is required to be replaced in case of any renewal/strengthening.

Figure 4 shows how high asphalt content improves fatigue resistance. This proves why international specifications are opting for dense graded mixes with more bitumen content rather than open graded mixes with less bitumen. This emerging concept has

8

already been followed, at least in principle, at many locations in USA. For example, on I-710 (the Long Beach Freeway) in Los-Angeles County a mix comprising of 25 mm OGFC + 75 mm dense-graded HMA + 150 mm asphalt has been used on the design concept of perpetual pavement (Monismith and Long, 1999). The perpetual pavements last long, provided they are built on a solid foundation. It is observed that rutting on such roads built on sub-grade with the CBR greater than 5 % originates mostly in the HMA layer. This suggests that a sub-grade with a CBR greater than 5% (M

R

greater than 50

MPa) is considered adequate. Proper construction techniques and quality control are essential for perpetual pavements like any other pavement).

Fig.4 Improve fatigue resistance with high asphalt content mixes (Source: Idaho Project Development Conference 2007)

Figure 5 shows how the design concepts of the perpetual pavements minimizes the tensile strain with pavement thickness. Figure 6. shows how the design concept of perpetual pavements deals with high temperature encountered during summer in case of most of our pavements in DBM layers.

9

Fig. 5 Minimize Tensile Strain with pavement thickness (Source: Idaho Project Development Conference 2007)

Figure 6. Impact of Temperature Gradient on Asphalt Grade. (Source: Idaho Project Development Conference 2007) The concept of perpetual pavements is briefly introduced to emphasize the current trend of development in respect of specifications of bituminous mixes. This also highlights the required change in mindset of Indian engineers to review the existing specifications of bituminous mixes from the point of view of long-lasting pavements rather than on the consideration of cost and conveniences. The subsequent chapter describes the current specifications followed in many developed countries. It is felt that even these specifications in those countries might go further changes consistent with the perpetual pavement concepts or similar concepts emerging in future. The point of concern is that long-lasting pavements going to be future concern in India. The advantages of such specifications lie not only in long life but also in the reduced cost of travel with better serviceability conditions. Figure 7 shows one such comparison.

10

Years Fig. 7 Pavement Serviceability Comparison (Source: Idaho Project Development Conference 2007)

4. EXPERIECE I DEVELOPED COUTRIES (USA)

Most specifications for dense graded bituminous mixes in the USA and other countries are intended for four pavement courses, namely, base course, binder course, and two surface (wearing) courses (one coarse and one fine). Table 1 shows four such Marshall mixes used by the Georgia Department of Transportation (GDOT) in the recent past, before changing to Superpave designed mixes6. The table shows nominal maximum aggregate size (NMAS), minimum and maximum layer (lift) thickness, and gradations of the four mixes. Unlike other highway agencies, a narrow range of material passing the 2.36 mm sieve is specified by GDOT for developing the job mix formula. The second sieve size has been changed from 85-100 to 90-100 percent in this table to suit the specified NMAS of the mix.

GDOT has held the general reputation of constructing one of the best and most durable flexible pavements in the US for several years. This time period encompasses the use of both Marshall and Superpave mixes in Georgia. Eleven projects consisting of Marshall mixes were evaluated in Georgia recently35. The average rut depth after about 6 years of heavy traffic was determined to be 1.5 mm only. Georgia’s experience should be of interest to India because Georgia being in the southeastern US has climate similar to north India. Air temperature up to 44.50 C has been recorded in Georgia. That is why in the past, Georgia used AC-30 viscosity grade bitumen, which is equivalent to VG-30

11

TABLE 1. GRADATIO OF GEORGIA DOT MARSHALL MIXES (Ref. 6) Mix Type Lift Thickness

25 mm NMAS Base Course 75-125 mm

19 mm NMAS Binder Course 45-75 mm

Sieve Size, mm 37.5 25 19 12.5 9.5 4.75 2.36 1.18 0.600 0.300 0.150 0.075 Bitumen Content

100 90-100 60-80 32-40 11-19 4-7 4.0 – 5.5

100 90-100 55-75 30-36 11-19 4-7 4.0 - 5.5

12.5 mm NMAS Wearing Course 35-60 mm

Percent Passing 100 90-100 70-85 44-48 10-25 4-7 5.0 - 7.0

9.5 mm NMAS Wearing Course 25-50 mm

100 90-100 55-75 44-50 14-25 4-7 5.2 - 7.5

grade (50-70 penetration) used in India. Georgia’s experience is, therefore, quite relevant in reviewing and revising Indian dense graded mix specifications in terms of NMAS, gradation and layer (lift) thickness.

As shown in Table 1, the base course consists of a 25 mm NMAS mix; the binder course consists of a 19 mm NMAS mix; and the surface (or wearing) course consists of 12.5 mm or 9.5 mm NMAS mix. The 12.5 mm wearing course mix is used for heavy traffic roads, whereas the 9.5 mm wearing course mix is used for low to medium traffic roads, in urban areas, and in thin (25 mm) applications. Any of the four dense graded mixes can be used for leveling or profile corrective course (PCC) depending upon the required thickness. Only dense graded mixes are used in the US in PCC30.

The Georgia DOT and some other state DOTs in the US attempted to use 37.5 mm NMAS mix for base course, which was only marginally more stable than the 25 mm mix, but had the following disadvantages:

•

The 37.5 mm NMAS mix was found very prone to segregation resulting in honeycombing (Photo 3). 12

•

The 37.5 mm NMAS mix becomes permeable at relatively lower air void content levels as compared to the 25 mm NMAS mix, thereby increasing the potential for water/moisture related damage. Fig. 8 clearly shows the effect of NMAS on field permeability

2,26

. As the NMAS increases, the permeability also increases

multifold at a given void level. For example, at an in-place air void content of 6 percent, the following permeability values were measured for each NMAS.

9.5 mm NMAS 12.5 mm NMAS 19.0 mm NMAS 25.0 mm NMAS

6×10-5 cm/sec 40×10-5 cm/sec 140×10-5 cm/sec 1200×10-5 cm/sec

The preceding data26 clearly shows larger NMAS 37.5 mm will be highly permeable since the permeability increases multifold from one NMAS to the next higher one.

•

Modified Marshall method developed by Kandhal

20,23,25

and referred to in

Asphalt Institute MS-21, which uses 6-inch diameter mould needs to be used for designing and testing 37.5 mm mix.

Photo 3. Segregation of 37.5 mm MAS mix resulting in honeycombing

13

Fig. 8 Effect of nominal maximum aggregate size (MAS) on permeability of inplace pavement (Ref. 26) Some engineers in India including 2nd author believe that the use of 37.5 mm mix should be considered for base course in India because of overloading problem. It is felt that this gets well validated by the observation made before that in case of heavy loads traffic stone mastic asphalt is a correct specification as per international practice. The axle loads in India is quite heavy and further the speed is low with many stop/start condition. Rutting of bituminous mixes like DBM in India is quite common

32

. It is, therefore, felt

that issue of providing DBM with NMAS of 37.5 mm be, therefore, kept under consideration pending further research.

5. REVIEW OF FLEXIBLE PAVIG MIXES USED I IDIA For the purpose of this paper a detailed review of the following bituminous paving mixes specified in the MORTH Specifications (2001) is undertaken. a. Bituminous Macadam (BM) b. Dense Bituminous Macadam (DBM) c. Semi-Dense Bituminous Concrete (SDBC) d. Bituminous Concrete (BC) 5.1 Bituminous Macadam (BM) Bituminous Macadam (BM) is an open graded, permeable, and recipe type mix produced without any quality control on its volumetrics or strength (stability). The primary problem with the BM mix is that being very open graded, it is highly permeable and therefore will trap moisture or water. BM and SDBC were developed several years ago, when conventional hot mix plants were not common. At that time, hot mixing was done in small portable plants or concrete mixers in which much fine aggregate could not be

14

used due to limitations of the available heating and mixing equipment. Now, good hot mix plants are normally available.

Fig. 9, 10 & 11 shows a typical cross-section of flexible pavement as being used in India. Fig. 9 does not have a BM layer and DBM is resting directly on WMM. However, Figs. 10 and 11 show cross-sections where BM has been used as a base, binder or profile corrective course (PCC) with no outlet for water thus creating a “bath tub” situation within the pavement.

Fig. 9 Typical cross-section of flexible pavement in India

Fig. 10 Flexible pavement with BM as a base course or PCC

Fig. 11 Flexible pavement with BM as a base/binder course 15

The fundamental question thus boils down to BM versus DBM. Should BM be deleted and DBM used instead in all cases? To answer that question BM and DBM should be compared both from the engineering aspect (primary) and economical aspect (secondary). This has been done considering the following factors:

5.1.1 Permeability: It has been acknowledged in many IRC and MORTH publications that BM is a much more open mix compared to the DBM. The MORTH Manual for Construction and Supervision of Bituminous Works29 states on page 52, “Because of the open-graded aggregate matrix, the voids content (in the BM) can be as high as 20-25 percent.” Some researchers have reported air voids of about 10% in lab compacted BM specimens. This range of 10-25% air voids can occur because BM has two gradings and within each grading BM can be relatively coarse or fine considering the combination of lower and upper values for each sieve. Table 2 gives air voids and permeability data obtained recently on the BM mix8. In this case four BM gradations were used: Grading 1 (both coarse and fine) and Grading 2 (both coarse and fine). The test data was obtained on 150-mm diameter specimens compacted with 75 blows (equivalent of 50 blows on 100-mm diameter specimens). The air void content ranges from 8.3 to 15.4 percent. The test data on Gradings 1 and 2 are comparable because both gradings have about the same amount of material passing 4.75 mm sieve. Photos 4 and 5 show the open texture of BM specimens Grading 1 and Grading 2, respectively. When these specimens were placed under a water tap, the water readily passed through indicating very high permeability.

TABLE 2. AIR VOIDS AD PERMEABILITY TEST DATA FOR FLEXIBLEMACADAM (Ref. 8) Mix Type Bitumen Content, Air Voids, % Permeability, % cm/sec BM Grading 1 (Coarse) 3.25 13.6 3.4 BM Grading 1 (Fine) 3.25 8.9 0.4 BM Grading 2 (Coarse) 3.4 15.4 3.6 BM Grading 2 (Fine) 3.4 8.3 0.6 Even if the scenario of about 10% air voids in the BM in the lab is considered, the voids in the field can be as much as 15% (at least 95% compaction of the lab density is usually required). According to numerous studies all over the world, dense graded bituminous

16

mixes become permeable when air voids are more than 7-8%. BM type open graded mix, which has a large number of interconnected voids, becomes permeable at relatively lower air voids, i.e. for air voids more than 5-7% . So there cannot be any argument about the fact that the BM is a highly permeable mix compared to the DBM. It has been said, three

Photo 4. Open surface texture of BM Grading 1

Photo 5. Open surface texture of BM Grading 2 things are important in highway construction – drainage, drainage, and drainage. No permeable asphalt layer is desirable within the pavement structure (unless it is specifically for drainage with proper outlets) whether it is a PCC, base course, binder course or whatever. If this fundamental requirement is disregarded, the potential for premature pavement distress is increased. A permeable layer always attracts and traps water, moisture or moisture vapour. Water can come from the top, from the sides, or from the non-flexible courses underneath22. If there is a premix carpet (which is highly 17

permeable) right over the BM, rainwater will have direct access to the BM and can cause havoc. The first author has investigated and reported15,24 many real-life field case histories of premature pavement failures from across the world. In a majority of cases, bituminous layers, which trapped water, were the real culprits. There was stripping of bitumen in the permeable layer as well as in the adjacent layers overlying or underlying it due to traffic action. Photos 6 and 7 show a failure in Oklahoma, US, which was investigated by the first author15. On this project an open type binder course was used. It was saturated with water since there were no subsurface edge drains at the edge of the pavement. This led to stripping in the binder course under traffic. Note that majority of the potholes appeared near the pavement edge where water accumulated and did not have any positive outlet. Potholes are often found more in number in the right lane, which carried heavy truck traffic. It is, therefore, felt that BM layer without an outlet should not be used for long term pavement performance. 5.1.2 Structural Strength: Many highway agencies across the world give structural value to a BM type mix (used for drainage) of 50% of dense graded DBM type mix. IRC Publications 37 and 81 on flexible pavement design state that 7 mm of DBM is equal to 10 mm of BM. In either case, the DBM is far superior to the BM in terms of structural strength and fatigue life. Some engineers are suggesting using polymer-modified bitumen (PMB) in the BM to increase its structural strength. If that is the objective, why not simply use the stiffer DBM in the first place. First of all, hardly any agency in the world uses PMB in a base course mix. Moreover, using PMB in an un-designed, recipe type BM mix, which unlike the DBM has hardly any quality control criteria at the design or mixing stage, is not justified. Therefore, DBM is by far superior to the BM in terms of structural strength, mix design criteria, and mix production control.

18

Photo 6. Road failure in Oklahoma, US resulting from saturated binder course (Ref. 15)

Photo 7. Close up of stripped binder course in Oklahoma, US (Ref. 15) 5.1.3 Use as a PCC: It has been surmised that BM is a good material for profile corrective course (PCC) because it resists reflection cracking. No other country is using a permeable, water-trapping type mix for PCC. Only dense graded mixes such as DBM or BC are used for transverse or longitudinal profile correction in other countries30 in courses called leveling courses or wedge courses, which are same as India’s PCC. The reasons for using dense mixes are: to stay away from water-trapping permeable mixes and also to facilitate easy feathering of the mix from a specified depth to almost zero in a wedge type PCC. It is normally argued/believed that BM has a better resistance to reflection cracking and accordingly, many pavement designers introduced a layer of BM between DBM and WMM in India. It does not appear that this conclusion is based on any research.

19

Assuming that BM has a better resistance to reflection cracking, it is still a water-trapping permeable mix and as discussed earlier, the potential for failure by far exceeds the perceived advantage of resisting reflection cracking. It is better to seal the cracks or to remove the top cracks by milling/scraping or prevent their upward movement by use of geo-synthetics, rather than over trust the BM to prevent reflection cracking at the cost of stripping. 5.1.4 Cost Considerations: The use of BM is quite often made on the premise that BM is cheaper than DBM and, therefore, it is suitable for use in developing country like India. That is not correct. The following cost analysis of BM versus DBM was conducted at IIT, Madras in response to the revised draft BM specification published in Indian Highways for comments5.

According to the current prevailing schedule of rates in his area, the cost of BM layer is Rs. 3,465/cu.m and the cost of DBM is Rs. 4,193/cu.m. Consider that 100 mm of BM is required for a pavement and it is equal to 50 mm of DBM as per equivalency used in some other countries. The cost of a 2-lane highway (10 m wide including shoulders) with BM will cost Rs. 34.65 lakhs/2-lane-km and the cost of the same highway with DBM will cost Rs. 20.97 lakhs/2-lane-km. Now, that is a saving of Rs. 13.7 lakhs per km or 39.5%.

Then, assume that 100 mm of BM is equal to 70 mm of DBM as per IRC guidelines. In that case, the cost of highway with BM is Rs. 34.65 lakhs/2-lane-km and the cost of highway with DBM is Rs. 29.35 lakhs/2-lane-km. Thus using DBM in lieu of BM will reduce the cost by Rs. 5.3 lakhs per km, which amounts to a saving of about 15 percent.

Comparative cost analysis of BM and DBM has also been done based on the current 2007 schedule of rates of BM and DBM obtained from the Rajasthan PWD Circle in Jaipur. The cost of BM in place is Rs 1,404 per ton and the cost of DBM in place is Rs. 1,588 per ton. Considering that 100 mm of BM is equal to 70 mm of DBM as per IRC guidelines, the actual cost of DBM in place comes out to be Rs. 1,112 per ton. That is a saving of Rs. 292 per ton or 21%, when DBM is used in lieu of BM.

20

The preceding cost analyses have clearly established that DBM is much cheaper than the BM on equivalency cost basis.

5.1.5 Traffic Conditions: According to some highway engineers, BM is intended for low-traffic roads only, although it is being used extensively on national highways and state highways. However, the fact remains that any layer, which traps water, should not be used whether it is a low-volume or high volume road. The concept of perpetual pavement or long term performing pavements is relevant even for less traffic roads. The need today is to construct pavements needing less maintenance avoiding frequent overlays, besides providing a good riding surface for many years. The design concept should, therefore, be same for both heavy traffic and low traffic. In any case, the thickness of different bituminous layers will be different depending upon the traffic intensity. As discussed earlier, water is the enemy of the road. The only tangible argument against the use of DBM could be non availability of proper batch mix plants in required number gradually to produce DBM on rural roads/state highways. It is felt that this should be enforced at least to tone up the quality level of our pavements.

5.1.6 General Statements: General statements are sometimes made like (a) BM is a “popular” mix or (b) BM has been widely used with “success” throughout the country. The concept of success is myopic and it does not envisage in concept of long-term performing pavements. The normal life of pavement in India is between 2 to 4 years compared to 8 to 10 years in other countries. The developed countries are talking of perpetual/long-term pavements capable of performing for 50 years or more. This may look strange but our vision should accordingly extend to give precedence to durability over deceptive cost saving. The concept of sound economics/engineering suggest that we should accept changing the permeable mixes by dense and relatively less permeable mixes to give long life to our bituminous pavements.

5.2 Dense Bituminous Macadam (DBM) At the present time the dense flexible macadam (DBM) is specified for use as a base course and/or binder course. Two gradations of the DBM are specified in Section 507 of

21

MORTH specifications: Grading 1 has a NMAS of 37.5 mm and Grading 2 has a NMAS of 25 mm.

Table 3 gives the existing MORTH composition of DBM Gradings 1 and 2. The specified percentage of fine aggregate is the same in both gradings (28-42 percent), the main difference is just some large size aggregate particles (25-45 mm size) are contained in Grading 1. It was discussed earlier that the use of large stone mix (NMAS of 37.5 mm or larger) has several disadvantages such as segregation (Photo 3) and high permeability (Fig. 8). Although these disadvantages outweigh the marginal gain in stability over a 25 mm NMAS mix, some engineers would like to use it in India due to overloading problem. Therefore, DBM Grading 1 has been retained. Since Grading 1 is highly permeable, it should be sealed before rainy season otherwise water will penetrate and TABLE 3. EXISTIG MORTH GRADATIOS FOR DESE BITUME MACADAM (DBM) (Ref. 28) Grading Nominal Aggregate Size Lift Thickness Sieve, mm 45 37.5 26.5 19 13.2 9.5 4.75 2.36 1.18 0.6 0.3 0.15 0.075 Bitumen Content, %

1 40 mm 80-100 mm

2 25 mm 50-75 mm Percent Passing

100 95-100 63-93 55-75 38-54 28-42 7-21 2-8 Min. 4.0

100 90-100 71-95 56-80 38-54 28-42 7-21 2-8 Min. 4.5

damage the underlying WMM course. This neglect is commonly observed during construction of our roads. Overall, DBM Grading 2 with 25 mm NMAS is best suited for a base course similar to GDOT base course specification in Table 1.

22

The MORTH specification mentions a lift thickness of 50-75 mm for Grading 2. It needs to be debated whether it should be 50-100 mm as practiced in many countries including the US30.

Table 4 makes a comparison of existing DBM Grading 2 and GDOT gradation for base course (extracted from Table 1) and proposes a new gradation for 25 mm NMAS DBM Base Course to be used in India. As mentioned earlier, Georgia has hot climate similar to north India and has one of the best flexible roads in the US. It is encouraging to note that the existing DBM Grading 2 is reasonably similar to GDOT Base Course. Therefore, TABLE 4. COMPARISO OF GRADATIOS: EXISTIG MORTH DBM GRADIG 2, GDOT BASE COURSE AD PROPOSED DBM BASE COURSE Grading

Existing DBM Grading 2

GDOT Base Course

Nominal Aggregate Size Lift Thickness

25 mm

25 mm

50-75 mm

37.5 26.5 19 13.2 9.5 4.75 2.36 1.18 0.6 0.3 0.15 0.075 Bitumen Content, %

100 90-100 71-95 56-80 38-54 28-42 7-21 2-8 Min. 4.5

75-125 mm Percent Passing 100 90-100 60-80 32-40 11-19 4-7 4.0-5.5

Proposed DBM Base Course Grading 2 25 mm 75-100 mm 100 90-100 71-95 56-80 38-54 28-42 7-21 4-7 4.0-5.5

the proposed DBM Base Course Grading 2 has been kept the same as existing DBM Grading 2 except for the percentage of fines (material passing the 0.075 mm sieve). At least 4 percent fines are needed in the job mix formula (JMF) to impart some stiffness to the bitumen-fines mortar. Eight percent fines are considered too excessive. It should be noted that the range of 4 to 7 percent is intended for the JMF design gradation. Normal variation during production will be allowed. It is also recommended to specify and use 25

23

mm sieve in lieu of 26.5 mm sieve and 12.5 mm sieve in lieu of 13.2 mm sieve so that Indian standards generally conform to world standards. It applies to all gradations. The lift thickness for DBM Base Course with NMAS of 25 mm has been revised to 75-100 mm. Similar changes have been made to proposed DBM Grading 1 with NMAS of 37.5 mm as reported later in Table 8. 5.3 Semi-Dense Bituminous Concrete (SDBC) There is no engineering logic in using a “semi-dense” mix when only dense, continuously graded mixes are technically desirable. In most developed countries30 either dense mixes (HMA) are provided or OGFC is provided as wearing course. Semi-dense mixes which are neither dense graded nor open graded contain the so-called “pessimum” voids when constructed. Terrel and Shute32 advanced the concept of “pessimum” void concept for stripping. Fig. 12 shows the general relationship between air voids and relative strength of bituminous mixes following water conditioning. The amount of strength loss depends upon the amount and nature of voids. As shown in the figure, at less than 4 percent air voids, the mix is virtually impermeable to water, so it is essentially unaffected. Unfortunately, region B to C of Fig. 12 is where mix is semi-dense. As the voids increase to D and beyond, the mix strength becomes less affected by water because the mix is now free draining like an ATPB. The region B to C can be called “pessimum” void content

Fig. 12 Pessimum voids in semi-dense mixes (Ref. 32)

24

because it represents opposite of optimum. The objective is to stay out of the “pessimum” void range. A “semi-dense” mix, which has a potential for having “pessimum” voids in it, is likely to trap water and causing stripping. It is felt that in a tropical country like India thicker wearing courses of 40 mm should be provided on all roads to ensure against the penetration of water from top and thereby to prevent crack initiation from top. Unfortunately, the use of SDBC has been advertently promoted to some extent because only SDBC Grading 2 has been specified by MORTH for a thin layer of 25 mm. If this is the consideration then there is no reason as to why a dense graded 9.5 mm NMAS BC cannot be used like in other countries in lieu of SDBC Grading 2 to provide a thin layer of 25 mm. It should also be noted that BC is only 3 percent more expensive than the SDBC as is evident from the following prices obtained from the 2007 Schedule of Rates of Rajasthan PWD, Jaipur Circle:

Item 16.7.6 Semi-Dense Bituminous Concrete (SDBC) Grading 2

Rs. 1,756 per ton

Item 16.8.6 Bituminous Concrete (BC) Grading 2

Rs. 1,812 per ton

5.4 Bituminous Concrete (BC) Two gradings of the Bituminous Concrete (BC) have been specified in Section 509 of the MORTH Specifications (2001). According to MORTH, the BC can be used for wearing and profile corrective courses. Grading 1 has a NMAS of 19 mm and Grading 2 has a NMAS of 13 mm.

As discussed earlier, a DBM base course was already selected. Now, there is a need to select a binder course and two wearing (surface) courses (one coarse for heavy traffic and one fine for light to medium traffic, urban areas, and thin application). BC Grading 1 with a NMAS of 19 mm is suitable for a binder course. Its gradation needs to be revised to make it more rut resistant. BC Grading 2 with a NMAS of 13 mm is suitable for a wearing course on heavy-traffic roads. Its gradation also needs to be revised to make it more rut resistant.

There is a need to add a third BC gradation with a NMAS of 9.5 mm, which can be used for light to medium traffic, urban areas, and in thin (25 mm) applications. Right now, BC 25

Grading 2 has been specified for a layer thickness of 30-45 mm. As mentioned earlier, this has forced pavement engineers to select SDBC grading 2 if they wanted a thin (25 mm) application. Now that the SDBC has been eliminated, the new BC Grading with 9.5 mm NMAS should replace it. This is suggested on the premise that 25 mm wearing course is required to be provided for less trafficked roads for cost saving. Sinha et.al32 has studied the problem of rutting on a heavy trafficked NH 32 and have found deep rutting within a year or so of construction. DBM mix was found to have been affected with high inside pavement temperature (DBM layer) during summer, exceeding the softening point of bitumen used. The pavement thus gets a kneading effect by heavy trucks moving with slow speed/stop-start condition. The quality of bitumen needs to be toned up to take care of such problem of rutting. The rutting was observed almost the entire depth of 150 mm thick DBM. Rut depths between 75 to 100 mm within 2 to 3 years of opening to traffic is quite common in India. This needs to be researched and improved. A high modulus rut resistant mix is needed to avoid rutting problem. The concept of perpetual pavement briefly mentioned before suggest that there is also a need to improve the three gradations of the BC to make them more rut resistant. This can be achieved by taking advantage of successful field experience in an area or on a test track, which is located in a region anywhere in the world with hot climatic conditions similar to India in the absence of any significant research in India.

One can emulate the gradations of the base course, binder course and surface course used in Georgia. It is generally argued that experience from other countries cannot be emulated. Indigenous research is always welcome but in the absence of same we cannot continue with poor specifications. Granite is granite or limestone is limestone, whether it is in Georgia or India. Similarly, AC-30 bitumen used in Georgia in the past is similar to VG-30 bitumen (50-70 penetration) used in India. Traffic is also computed in ESALs both in Georgia and India. Therefore, there is no reason as to why experience with bituminous mixes cannot generally be used in India with some adjustments at least to start with.

26

Another source of excellent experience with performance of flexible mixes in hot climate similar to north India is the 2.7 km long oval test track operated by the National Center for Asphalt Technology (NCAT), Auburn University, Alabama since year 2000. This test track has 46 asphalt test sections, which comprise of different binder course and wearing course mixes. Variables include type of mix (such as Marshall, Superpave, SMA and OGFC), type of aggregate (such as granite, limestone, quartzite, gravel and slag), type of gradation (such as coarse graded and fine graded)11,16, and type of bitumen (such as unmodified and modified).

Real trucks with loaded trailers are used to apply traffic loading of 10 million ESALs to the track within a 2-year cycle. This amount of traffic is equivalent to 10-year traffic on a typical interstate (national) highway in the US. Performance of the asphalt test sections in terms of rut resistance and cracking is monitored periodically. Photos 8 and 9 show the NCAT Test Track and trucks with loaded trailers. The test track is located in Alabama (latitude of 32.6 degrees) in southeastern US with climate similar to north India. Average maximum pavement surface temperature of 61.40 C has been recorded on the track. The test track has primarily used a 19 mm NMAS binder course and a 12.5 mm NMAS wearing course. Most of the test sections have performed extremely well with average rut depth of 3 mm and maximum rut depth of 6 mm and no significant cracking. It should be mentioned that maximum acceptable rut depth of a pavement is considered to be about12.5 mm during its service life. Therefore, guidance can be taken from gradations of binder course and surface (wearing) course actually used on the test track with excellent performance under hot climatic conditions.

Photo 8. NCAT Test Track in Auburn University, Alabama

27

Photo 9. Trucks with loaded trailers applying traffic load to NCAT Test Track Table 6 has been prepared to compare the existing MORTH BC Grading 1 of 19 mm NMAS mix (intended now for use as a binder course) with the gradation specified by GDOT and the actual gradation range used on NCAT test track with excellent performance results. The last column in this table has been formulated from the comparison as a recommended 19 mm NMAS BC binder course gradation for India. Only the values for 19 mm, 13.2 mm, 9.5 mm and 0.075 mm have been revised based on excellent performance experience in Georgia and on NCAT test track. Remaining values in the existing MORTH specifications for other sieves are reasonably close to Georgia TABLE 6. COMPARISO OF GRADATIOS: EXISTIG MORTH BC GRADIG 1, GDOT BIDER COURSE, CAT TRACK BIDER COURSE AD PROPOSED BC BIDER COURSE Grading

Nominal Aggregate Size Lift Thickness Sieve Size, mm 26.5 19 13.2 9.5 4.75 2.36 1.18 0.6 0.3 0.15 0.075

Existing MORTH BC Grading 1 19 mm

GDOT Binder Course 19 mm

NCAT Test Track Binder Course 19 mm

Proposed BC Binder Course 19 mm

50-65 mm

45-75 mm

-

50-75 mm

100 79-100 59-79 52-72 35-55 28-44 20-34 15-27 10-20 5-13 2-8

100 90-100 60-89 55-75 30-36 11-19 4-7

100 97-100 66-86 48-80 32-53 24-38 20-30 16-24 11-15 7-12 4-8

100 90-100 66-86 55-75 35-55 28-44 20-34 15-27 10-20 5-13 4-8

28

Bitumen Content

5.0-6.0

4.0-5.5

4.0-5.0

4.0-5.5

and test track and, therefore, have been kept the same. The reasons for changing the values for 0.075 mm sieve have been discussed earlier in the paper. The second largest sieve size (19 mm in this case) of a dense graded mix should be 90-100 percent material passing rather than 79-100 percent. The reasons for the suggested changes are as follows: •

A few large size, isolated aggregate particles hardly increase the mix stability and can unnecessarily cause segregation problems. Photo 10 shows the presence of such scattered particles in a compacted mat on a national highway in India.

•

With 79-100 percent passing the second sieve, the same mix can have different NMAS values, which will create confusion in mix selection and will also have different minimum VMA requirements19. That is why; the Superpave mix design has control points of 90 and 100 percent for the second largest sieve size.

Photo 10. Large size, isolated aggregate particles in compacted mat. Table 7 has been prepared to compare the existing 13 mm NMAS MORTH BC Grading 2 (intended now for use as a coarse surface or wearing course) with the gradation specified by GDOT and the gradation range used on NCAT test track with excellent performance results. The gradation range for the wearing course used on the test track encompasses gradations below the Superpave restricted zone (BRZ), through the

29

restricted zone (TRZ), and above the restricted zone (ARZ). All these gradations performed really well on the track. The restricted zone specified earlier in Superpave through which no gradation was permitted to pass, was already deleted from the Superpave mix design based on a half-million dollar research study conducted by Kandhal and Cooley10,14,17.

Again, the proposed gradation for India in the last column of Table 7 has been formulated from the comparison. Most of the values except for 9.5 mm and 4.75 mm sieves have been revised. Reasons for revising the values for the second largest sieve 13.2 mm (or 12.5 mm) and the last 0.075 mm sieve were given earlier. The values for sieves 2.36 mm down to 0.15 mm have been generally lowered because it is believed that the present MORTH gradation has excessive amounts of fine aggregate (compared to GDOT and NCAT test track), which can increase the potential for rutting.

TABLE 7. COMPARISO OF GRADATIOS: EXISTIG MORTH BC GRADIG 2, GDOT SURFACE COURSE, CAT TRACK SURFACE COURSE AD PROPOSED 12.5 mm BC SURFACE COURSE Grading BC Grading 2 GDOT Wearing NCAT Test Proposed BC Course Track Wearing Wearing Course Course Nominal 13 mm 12.5 mm 12.5 mm 12.5 mm Aggregate Size Lift Thickness 30-45 mm 35-60 mm 30-50 mm Sieve Size, mm Percent Passing 19 100 100 100 100 13.2 79-100 90-100 94-99 90-100 9.5 70-88 70-85 73-92 70-88 4.75 53-71 51-73 53-71 2.36 42-58 44-48 34-54 38-54 1.18 34-48 22-38 24-38 0.6 26-38 17-29 17-29 0.3 18-28 10-25 12-19 12-22 0.15 12-20 7-11 7-15 0.075 4-10 4-7 4-8 4-7 Bitumen Content 5.0-7.0 5.0-7.0 4.3-7.8 5.0-7.0 Table 8 shows the recommended final gradations of 5 dense graded bituminous mixes for India: 37.5 mm NMAS DBM base course, 25 mm NMAS DBM base course, 19 mm NMAS BC binder course, 12.5 mm NMAS BC wearing course Grading 1, and a new 9.5 30

mm NMAS BC wearing course Grading 2. The finer 9.5 mm NMAS BC wearing course Grading 2 has been adapted from the GDOT specification for Marshall mixes (Table 1), which has an excellent performance history for application on low to medium traffic roads, in urban areas, and in thin (25 mm) applications. Values for three sieve sizes 1.18 mm, 0.6 mm and 0.15 mm have been interpolated. It is recommended to use Table 8 to revise the gradations of MORTH DBM mix and BC mixes. All these 5 mixes are adequate for constructing conventional dense graded bituminous pavements in India, both for new construction and overlays. They are also suitable for PCC based on the required thickness. As mentioned earlier, it is not necessary to use all three courses (base, binder, and wearing) in a new flexible pavement unless the traffic is very high. For example, the following combinations can be used depending upon the total thickness of the bituminous course(s) required as per structural design based on IRC:37.

TABLE 8. PROPOSED FIAL FIVE DESE GRADED BITUMIOUS MIXES FOR IDIA Grading Proposed Proposed Proposed Proposed Proposed BC DBM Base DBM Base BC Binder BC Wearing Wearing Course Course Course Course Course Grading 1 Grading 2 Grading 1 Grading 2 Nominal 37.5 mm 25 mm 19 mm 12.5 mm 9.5 mm Aggregate Size Lift Thickness 75-100 75-100 mm 50-75 mm 30-50 mm 25-40 mm Sieve Size, Percent Passing mm 45 100 37.5 90-100 100 26.5 63-93 90-100 100 19 71-95 90-100 100 13.2 55-75 56-80 66-86 90-100 100 9.5 55-75 70-88 90-100 4.75 38-54 38-54 35-55 53-71 55-75 2.36 28-42 28-42 28-44 38-54 40-55 1.18 20-34 24-38 29-44 0.6 15-27 17-29 21-33 0.3 7-21 7-21 10-20 12-22 14-25 0.15 5-13 7-15 7-15 0.075 4-7 4-7 4-7 4-7 4-7 Bitumen 4.0-5.5 4.0-5.5 4.0-5.5 5.0-7.0 5.2-7.5 Content

31

ote: Wearing course grading 1 is recommended for heavy traffic roads. Wearing course grading 2 is recommended for light to medium traffic roads, in urban areas, and for thin (25 mm) applications. • • •

WMM + DBM Base Course + BC Binder course + BC Wearing Course WMM + BC Binder Course + BC Wearing Course WMM + BC Wearing Course only

It has been surmised by some that a BC wearing course is too stiff and will crack if placed directly over WMM. This is not correct because the BC wearing course has relatively lower stiffness due to its lower NMAS (12.5 or 9.5 mm) and high bitumen content. This combination is being used in other countries including Australia and South Africa. Similarly, a bituminous overlay required for strengthening flexible pavement can consist of the following depending upon the required thickness as per IRC:81:

• •

BC Binder Course + BC Wearing Course BC Wearing Course only

Unlike most developed countries, overloading is a major concern in India. On very heavily trafficked road with severe overloading problem, it is recommended to modify the BC wearing course and BC binder course (that is, the top 100 mm of the pavement only, which is likely to rut) as follows:

•

Ensure to use viscosity graded VG-30 grade bitumen as per latest IS73:20064, which is significantly more rut resistant than the old 60/70 penetration bitumen.

•

Use polymer modified bitumen (PMB)

•

Use Stone Matrix Asphalt (SMA) as per IRC Specifications approved recently

6. COCLUSIOS

1. For ensuring long-term performing pavements focus should shift to dense graded bituminous mixes rather than open graded lean bituminous mixes. The future trend is towards introducing high modulus rut resistant mix material to take care of problems of rutting in bituminous layers. Rutting is broadly due to inadequacies in our mix design rather than granular bases or sub-grades.

32

2. Presently there is a proliferation of bituminous paving mixes in India. Many of them are open graded lean bituminous mixes. Too many options for a specific bituminous course have created confusion in mix selection. Many pavement design engineers do not understand the individual characteristics of these mixes and, therefore, select them based on their specified layer thickness and/or cost. 3. Most countries of the world generally do not have more than four dense graded mixes. Notwithstanding other lean mixes like BUSG, BPM there are 8 mixes like BM (2 gradings), DBM (2 gradings), SDBC (2 gradings), and BC (2 gradings). The study recommends in total five numbers of bituminous mixes in lieu of current eight specified in MORT&H specifications as per details below. • • • • •

37.5 mm NMAS DBM Base Course Grading 1 25 mm NMAS DBM Base Course Grading 2 19 mm NMAS BC Binder Course 12.5 mm NMAS BC Wearing Course Grading 1 9.5 mm NMAS BC Wearing Course Grading 2

4. Paper recommends phasing out of lean mixes as they are mainly responsible for poor performance of bituminous pavements. It is recommended not to use them on primary network like NH or even State Highways. Specifications should restrict the mixes to ensure better quality roads.

5. Some new specifications like those similar to open graded mixes such as permeable asphalt treated base (PATB) and OGFC may be introduced to complete the specifications. Both PATB and OGFC need to be used in India only in exceptional circumstances due to their additional costs. These may be used when specifically required.

6. Bituminous Macadam is a highly permeable mix and promotes rutting. Use of Bituminous Macadam a very popular mix at present may be deleted and substituted with DBM because it is finally cost effective and better performing. Similarly, use of Semi-dense Bituminous Concrete is also not considered to be allowed in the specifications. It suffers from “pessimum” voids, which have potential to trap water resulting in moisture damage. It should be substituted by Bituminous Concrete as it is better performing and cost effective.

33

7. Use of thick wearing/surface course of 40 mm is recommended in a tropical country like India to seal the pavement top from the ingress of water and moisture. To take care of possible use of wearing course of 25 mm on cost consideration as was the case in SDBC a 9.5 mm NMAS BC wearing course may be introduced in the specifications after removing SDBC.

8.

The Profile Corrective Course (PCC) should be only in DBM or BC. The current use of BM as PCC should be discouraged and not allowed.

7. RECOMMEDATIOS

A. A suitable research study to study the performance of specifications of bituminous mixes suggested above be carried out on the pattern of National Center for Asphalt Technology (NCAT), Auburn University, Alabama since year 2000 by constructing a test track. Pending the above research output the suggested change in the specifications may be incorporated and their performance in field be evaluated with respect to older specifications. B. The research scheme suggested above should attempt new mixes which comprises of high modulus rut resistant material to tackle the problem of rutting on the lines of design concept of perpetual pavements. C. Ongoing debate for the use of larger Nominal Maximum Aggregate Size (NMAS) in DBM etc. should be settled by conducting indigenous research and the existing grading should be modified suitably. Pending the above research output suggested grading be incorporated in the specifications. 7. ACKOWLEDGEMET The opinions expressed in this paper are those of the authors only. 8. REFERECES 1. Asphalt Institute. Mix Design Methods for Asphalt Concrete. MS-2, Sixth Edition, 1997.

34

2. Brown, E. R. et al. Relationship of Air Voids, Lift Thicknesses, and Permeability of Hot Mix Asphalt Pavements, TRB, NCHRP Report 531, 2004. 3. Brown, E. R. and S. A. Cross. A Study of In-Place Rutting of Asphalt Pavements. Asphalt Paving Technology, Vol. 58, 1989. 4. Bureau of Indian Standards. Paving Bitumen – Specification (Third Revision) IS 73:2006, July 2006. 5. Comments on IRC Draft FlexibleMacadam Specifications published in Indian Highways, Vol. 35, No. 8, August 2007. 6. Communication with Don Watson, Formerly with Georgia Department of Transportation, Now at NCAT, February 2, 2008. 7. Communications with the Border Roads Organization, August 26, 2007. 8. Communication from IJM (India), BM Test Data Report. October 18, 2007. 9. Kandhal, P.S. Quiet Pavements: Asphalt Pavements Mitigate Tire/Pavement Noise. Hot Mix Asphalt Technology Journal, Vol. 9, No. 2, 2004 10. Kandhal, P.S. and L.A. Cooley Jr. Investigation of the Restricted Zone in the Superpave Aggregate Gradation Specifications. Asphalt Paving Technology, Volume 71, 2002. 11. Kandhal, P.S. and L.A. Cooley Jr. Coarse Versus Fine-Graded Superpave Mixtures: Comparative Evaluation of Resistance to Rutting. Transportation Research Board, Transportation Research Record 1789, 2002. 12. Kandhal, P.S. Design, Construction, and Maintenance of Open-Graded Asphalt Friction Courses. National Asphalt Pavement Association Information Series 115, May 2002. 13. Kandhal, P.S. Designing and Constructing Stone Matrix Asphalt Mixtures—State-ofthe-Practice. National Asphalt Pavement Association Quality Improvement Publication QIP-122 (Revised Edition), March 2002. 14. Kandhal, P.S. and L.A. Cooley Jr. The Restricted Zone in the Superpave Gradation Specification. Transportation Research Board, National Cooperative Highway Research Program Report 464, 2001. 15. Kandhal, P.S. and I. Rickards. Premature Failure of Asphalt Overlays from Stripping: Case Histories. Asphalt Paving Technology, Volume 70, 2001. 16. Kandhal, P.S. and R.B. Mallick. Effect of Mix Gradation on Rutting Potential of Dense-Graded Asphalt Mixtures. Transportation Research Board, Transportation

35

Research Record 1767, 2001. 17. Kandhal, P.S. and L.A. Cooley Jr. Effect of Restricted Zone on Permanent Deformation of Dense-Graded Superpave Mixtures. ASTM Special Technical Publication 1412, 2001. 18. Kandhal, P.S. and R.B. Mallick. Open-Graded Friction Course: State of the Practice. Transportation Research Board, Transportation Research Circular Number E-C005, December 1998. 19. Kandhal, P.S., K.Y. Foo, and R.B. Mallick. Critical Review of VMA Requirements in Superpave. Transportation Research Board, Transportation Research Record 1609, 1998. 20. Kandhal, P.S. and T. Wu. Precision of Marshall Stability and Flow Test Using 6-inch (152.4-mm) Diameter Specimens. ASTM. Journal of Testing and Evaluation, Vol. 24, No. 1, January, 1996. 21. Kandhal, P.S. Field and Laboratory Evaluation of Stripping in Asphalt Pavements: State of the Art Report. Transportation Research Board, Transportation Research Record 1454, 1994. 22. Kandhal, P.S. Moisture Susceptibility of HMA Mixes: Identification of Problem and Recommended Solutions. National Asphalt Pavement Association, Quality Improvement Publication (QIP) No. 119, December 1992. 23. Kandhal, P.S. Large Stone Asphalt Mixes: Design and Construction. Proceedings, Association of Asphalt Paving Technologists, Vol. 59, 1990. 24. Kandhal, P.S., C.W. Lubold, and F.L. Roberts. Water Damage to Asphalt Overlays: Case Histories. Proceedings, Association of Asphalt Paving Technologists, Vol. 58, l989. 25. Kandhal, P.S. Changes in Mix Design to Improve Performance: Selected State Experiences. Proceedings, Association of Asphalt Paving Technologists,.Vol. 57, l988. 26. Mallick, R. B. et al. Evaluation of Factors Affecting Permeability of Superpave Designed Pavements. National Center for Asphalt Technology, Report 03-02, June 2003. 27. Mallick, R.B., P.S. Kandhal, L.A. Cooley Jr., and D. Watson. Design, Construction, and Performance of New-Generation Open-Graded Friction Courses. Asphalt Paving Technology, Volume 69, 2000. 28. Ministry of Road Transport & Highways. Specifications for Road and Bridge Works, Section 500, Fourth Revision, 2001, Indian Roads Congress, New Delhi. 29. Ministry of Road Transport & Highways. Manual for Construction and Supervision of FlexibleWorks. Indian Roads Congress, New Delhi, November 2001.

36

30. National Asphalt Pavement Association. HMA Pavement Mix Type Selection Guide. Information Series 128, 2001. 31. Roberts, F.L., P.S. Kandhal, E.R. Brown, D.Y. Lee, and T.W. Kennedy. Hot Mix Asphalt Materials, Mixture Design and Construction. NAPA Education Foundation, Lanham, Maryland, Second Edition, 1996. 32. Sinha V.K., H.N. Singh and Saurav Shekhar “Rutting in Flexible Pavements - A Case Study” Journal of the Indian Roads Congress Volume 68-3 October-November 2007 33. Sinha V.K. “Perpetual Pavements – An emerging concept” Editorial, Indian Highways, Indian Roads Congress, August 2008 issue 34. Terrel, R. L. and J. W. Shute. Summary Report on Water Sensitivity. SHRP Report SHRP-A/IR-89-003, November 1989. 35. Watson, D. E. et al. Verification of Superpave Ndesign Compaction Levels for Georgia. Paper presented at the 2008 Annual Meeting of the Transportation Research Board , Washington, DC, January 2008.

Response of Authors to Comments on Discussion Paper o. 541, “A Critical Review of Bituminous Paving Mixes Used in India” [The following response of authors to comments is under publication in the Journal of the Indian Roads Congress, Volume 69-4, 2009.] The authors appreciate the fact that their paper has solicited numerous valuable comments from the readers. Since there are many comments pertaining to recommended deletion of the Bituminous Macadam (BM) from the MORTH specification, it is prudent to make a general response on this topic at the beginning in order to avoid repetitions later in individual responses. General Response on BM The authors had made a detailed technical and economical comparison of BM with DBM in Section 5.1 of the paper in terms of permeability, structural strength, use as PCC, cost considerations, traffic conditions, and general statements. It was concluded with technical justifications that dense graded DBM should be used in lieu of open graded, undrained BM especially to obtain long lasting pavements. However, despite many fundamental, technical flaws associated with BM as mentioned in the paper, some commentators have advocated to retain it in the specifications. This is probably due to the following misconceptions: •

Dense graded DBM is not flexible enough to be placed directly on WMM and therefore a “flexible” BM course is necessary between the WMM and DBM. If the 37

•

DBM was not flexible it would not rut at all. But that is not the case. If there is uneven settlement /consolidation of WMM, the DBM is flexible enough to deform and adjust similar to BM if that is what is desired. It is a common practice in most countries of the world to place DBM type bituminous base course directly on crushed stone base course (we call it WMM). That practice has resulted in durable long lasting pavements without any problems. BM is cheaper than DBM and that is why it is good for a developing country like India. A detailed, comparative cost analysis given in the paper shows that the DBM is cheaper than the BM by 15 to 21% if the relative structural strengths are considered. Only when the BM is used as PCC to correct camber/super elevation it is cheaper than the DBM. But the problem still remains that the undrained BM PCC would trap moisture/water and thus is potentially detrimental to the pavement.

Some commentators have suggested retaining the BM but providing outlet for the water trapped in the open graded BM. To do this, the BM has to be extended all the way to the edge of the embankment (that is, day lighted) or pavement edge drains have to be constructed to drain the BM. Both of these configurations which are shown in Figs. 2(a) and 2(b) of the paper to drain a permeable asphalt treated base (PATB) are very expensive propositions. Even developed countries use the PATB as a drainage layer only on selected heavy duty roads. It should also be mentioned here that if we must use a drainage layer in India in exceptional circumstances, we should use PATB rather than the BM because the former with 2-3% bitumen content and coarser gradation is not only cheaper but also more permeable. If BM is used as PCC for correcting camber, it may not be possible to drain the BM wedge (triangle) especially if it is towards a raised median. Moreover, rainwater falling in the raised median may also enter sideways into the BM wedge and cause stripping and potholes. The first author has seen such a case on a national highway in India. As mentioned in the paper, it is time to move on from open graded “cheaper” mixes to dense graded, durable mixes if our objective is to have long lasting pavements both for low-volume and high-volume roads.

Responses to Individual Comments Shri Arlikar has asked about the effect of using low penetration 10/20 or 20/30 paving bitumen in bituminous mixes. It is not desirable to use such hard bitumens in India. Although these hard bitumens would increase mix resistance to rutting but they would decrease the mix resistance to fatigue cracking. Viscosity graded VG-30 bitumen is most suited for Indian climatic and traffic conditions and is recommended for all bituminous courses. Concerning Shri Bhattacharya’s comment (a), the authors agree with him that our drainage layers (assuming he means GSB) hardly drain even when day lighted. This is primarily due to too many fines in the GSB material. To make it reasonably

38

permeable, the material passing 0.075 mm sieve should not exceed 5 percent. The authors also agree with his comment (b) that without a geotextile between drainage layer and soft shoulder soil, the former will get choked with time rendering it almost ineffective. Concerning his comment (c), Shri Bhattacharya is referred to authors’ general response on BM given earlier. In response to Shri Chyrmang’s comment, it is stated in the paper that PMC should not be sandwiched between two dense graded bituminous layers (not BM courses as he mentioned). This is because PMC being open graded with no drainage outlet will trap water causing stripping in underlying and overlying bituminous layers when subjected to traffic. This would result in pavement disintegration and potholing. The paper gives some field case histories on this potential problem. Shri Garg should please note that the authors have recommended in Section 2.2 the use of stone matrix asphalt (SMA) in extreme cases of heavy traffic loads and high tyre pressures. SMA is practically impermeable (it is designed at 4% air voids) although its surface appears to be open graded. Shri Ghosh has mentioned that the paper “presents new and interesting ideas, which require thorough examination at different places and different soil conditions before adoption”. In fact, there is nothing new or interesting about using the recommended four dense graded mixes in India because these are the mixes being used successfully in various countries of the world with different climatic and soil conditions. What is so special about India? There is absolutely no need to do any further examination and waste time. The paper is simply recommending deleting the outdated, open graded mixes as explained in the beginning. Shri Joshi has mentioned that the paper has given a very good “non-conventional approach”. Actually, the current use of BM and SDBC in India is non-conventional considering that most countries of the world are conventionally using dense graded mixes only. The paper does not propose any crust thickness for each mix type, only the construction lift thicknesses are given; the former is determined from proper pavement thickness design as per IRC guidelines. The authors agree with Shri Jain that training of engineers is badly needed in India especially in view of our ambitious highway building programme. Shri Jain is also referred to the general response on BM given at the beginning. It is reiterated that the DBM is also a flexible course. On most national highways, there is a provision for placing two lifts of DBM and one lift of BC over the WMM. Sometimes, while under construction the highway is opened to heavy traffic after placing only one lift of the DBM. Under such circumstances the DBM has a potential for cracking because the highway does not have adequate pavement thickness to sustain heavy traffic loads. A crack relief layer is hardly used on crushed stone base course (we call it WMM) in other countries when a DBM type mix is placed on it. By definition, crack relief layer is used if the existing course (WMM in this case) already has cracks and there is a need to minimize reflection cracks in the overlay. In any case, if we do provide such an open graded layer, it must be provided with proper drainage as mentioned in the beginning. But that would be very costly.

39

Shri Mandal is requested to read the general response to BM in the beginning and also the paper, which clearly states that DBM is more economical than BM by 15 to 21%. It is also stated that dense graded mixes are likely to result in long lasting pavements and therefore cost effective both for low- and high-traffic volume roads. Premix carpet placed on WBM should suffice for PMGSY type rural roads. Not every project with BM will fail because conditions are different on every project. However, the potential for failure increases when an open graded mix is used within the pavement system without any outlet for water thus creating a “bath tub”. The authors agree with Shri Nayak that during every monsoon thousands of kilometres of bituminous roads are damaged due to insufficient drainage and pervious layers. The authors also agree with him that the existence of too many bituminous mixes in the specifications is confusing. The authors agree with Shri Ramakrishnan that IRC codes and MORTH specifications should be amended in view of this paper. It should be noted that the small size, portable plants can also produce dense graded bituminous mixes. Concerning Shri Ramakrishnan’s coments on Para 2.1 of the paper he is referred to the general response on BM given at the beginning. His comments on Para 2.2 imply that BM need not be day lighted but a GSB type material be used in shoulder against for its drainage. Since GSB type low permeability hardly drains, BM will still act like a “bath tub”. For his comments on the use of BM as PCC, he is referred to the general response on the BM. Shri Ramakrishnan has further commented about bitumen contents in Tables 5 and 6. Optimum bitumen content in a mix is indirectly related to the amount of aggregate passing the 2.36 mm sieve (that is, the amount of fine aggregate). Most of the aggregate surface area in a bituminous mix is contributed by the fine aggregate. As the proportion of the fine aggregate is increased in a mix, a larger surface area of the aggregate has to be coated thus increasing the bitumen demand. It should be noted that DBM Grading 2 (base course) and BC Grading 1 (binder course) have almost equal amounts (28-42%) passing the 2.36 mm sieve. Therefore, the same recommended range of 4.0 to 5.5% for bitumen content for both is reasonable. Obviously, this is just a range; the optimum bitumen content will have to be determined by conducting a Marshall mix design. The minimum bitumen content for these courses (base and binder courses) cannot be arbitrarily increased from 4 to 5% as suggested by Shri Ramakrishnan. If a proper Marshall mix design which satisfies all void parameters (such as VMA, air voids and VFA), stability and flow, gives 4.5% bitumen content for the base or binder course it cannot be increased to 5% without causing bleeding and/or rutting problems. It should be noted that the minimum VMA requirements ensures there is sufficient bitumen in the mix for durability. Some of our national highways are already exhibiting such problems. Shri Ramakrishnan should also note that the lift thickness of 75-100 mm as given in Table 7 for DBM Grading 2 with a NMAS of 25 mm has been used successfully in

40

other countries. In fact, the Georgia DOT is using a lift thickness of 75-125 mm for this mix as shown in Table 4. Dr. Raju has reported satisfactory experience with thin BC layers placed directly on WMM. This is practiced in other countries such as Australia and South Africa without any problem. This shows, as mentioned earlier, that the BC is flexible (contrary to the misconception that it is not flexible) and there is no need for a BM layer between WMM and DBM. According to Shri Rao, GSB under BM will function well if the GSB gradation is proper. Even if the GSB gradation is proper its permeability will be much less than the permeability of the BM. Therefore, according to the law of hydraulics, excess water in the GSB will seek the path of least resistance and will flow into the BM rather than travelling long distances within the GSB as explained in Section 2.2 of the paper. The authors concur with Shri Sahay that the concept of perpetual pavement (which requires the use of dense graded mixes only) should be encouraged in India. In response to Shri Rajiv Sharma, any dense graded mix given in Table 7 of the paper can be used for PCC on hill roads as well. Mix with an appropriate NMAS should be selected to suit the thickness of the PCC. Shri R. S. Sharma has asked about the opening of traffic on the DBM prior to placing the BC; the MORTH specification requires placement of BC immediately after laying DBM. Deterioration of DBM when exposed to traffic has been observed on many projects in India. This occurs especially when heavy traffic is opened on the first lift of the DBM because the pavement thickness is grossly inadequate at that time to sustain heavy loads. The deterioration also takes place from the ingress of rainwater because the DBM mix (especially Grading 1) is quite open. The first author has observed stripping in DBM Grading 1 on a national highway in India. That is why, the engineer would like to seal the DBM if BC placement is delayed. According to the first author, the following construction practice (also practiced in other countries like US) should be followed. Do not use DBM Grading 1 for base course, it is too open. Use DBM Grading 2 as a base course. As recommended in Table 7 of the paper, use BC Grading 1 as a binder course (this is in lieu of the second lift of the DBM. Since the finer BC Grading 1 will not allow any ingress of rainwater into the pavement and reasonable pavement thickness would be obtained (only 40 mm BC wearing course is lacking), traffic can be allowed on the road until the BC wearing course is placed. However, the time period between the binder course and the wearing course should be held as minimal as possible. Contrary to what the specification may require, sometimes it is not practical to keep the traffic away from the pavement during construction. Why spend money on sealing DBM Grading 1 or 2 when relatively impermeable BC Grading 1 is used as a binder course as recommended in Table 7 of the paper. Shri R. S. Sharma has also asked whether surface dressing can be used in place of PCC. This is not advisable because it will introduce a weak, permeable layer within

41

the pavement system. Only dense graded bituminous mixes should be used for PCC as recommended in the paper and discussed earlier. In response to Shri V. K. Sharma, again only dense graded bituminous mixes as given in Table 7 of the paper are suitable for PCC for Indian climatic conditions. Under no circumstances BM should be used as PCC. In response to Shri Gopal Singh, only dense graded mixes can be used for structural courses such as base, binder, and wearing course. Porous mix like PATB mentioned in Section 2.2 of the paper can be used only as a drainage layer below the structural courses. However, such bituminous drainage layers should be used only in case of exceptionally poor drainage conditions because of their prohibitive costs. In response to Shri Balbir Singh, it is not necessary to use a different mix for a different traffic volume. The mix type remains the same, only its thickness in the pavement is changed to suit the design traffic as per IRC guidelines. Shri A. V. Sinha has expressed concern about skid resistance when small size aggregate is used in the wearing course. Numerous research studies have shown that skid resistance of a pavement surface is dominated by its micro texture (contributed by the aggregate particle surface) rather than its macro texture (contributed by the aggregate size); although both contribute to skid resistance. That is, use of an aggregate with good micro texture and adequate resistance to polishing is more important than the NMAS of the mix which provides macro texture. It should be noted that Table 7 recommends BC Wearing Course Grading 1 with a NMAS of 12.5 mm for heavy traffic roads in India. In the US, BC Wearing Course Grading 2 with a NMAS of 9.5 mm is also used on many interstate highways with acceptable skid resistance. So the wearing course mixes recommended for India are actually on conservative side in terms of their NMAS and therefore there is absolutely no cause for concern. Shri Vakharia concurs with the recommendations made in the paper. The authors agree with him that the IRC Flexible Pavement Committee (H-2) should deliberate and bring out necessary amendments in relevant codes. However, it will be a difficult task to change the mindset of some engineers who continue to insist on using open graded mixes within the pavement. In response to Shri Bagalkot, retaining all existing mixes and just tabulating them on one sheet with their individual details is not the solution. It is necessary to delete the mixes, which are outdated and technically flawed as explained in detail in the paper. The authors agree with his suggestion that using proper GSB and WMM is a better option in lieu of PATB for India. Shri Bagalkot states that the use of BM and SDBC should be discoursed and he recommends experimental studies before arriving at a final decision. As emphasized earlier, the use of dense graded mixes in bituminous pavements is a well established and highly successful practice in the world for obtaining durable pavements and, therefore, there is no need to do any more discourses or experimental studies. That would simply delay the implementation unnecessarily to the detriment of our roads. [END]

42