Thompson - September 2008

Page 1 of 20

ROAD REFLECTIVE CRACKING SYSTEM UNDER COLD CLIMATIC CONDITIONS, EVALUATION AND STUDY OF THE FIBER-REINFORCED EMULSION MEMBRANE Martin Thompson Midland Asphalt Materials Inc. International Symposium on Asphalt Emulsion Technology, ISAET September 24-26th 2008

Abstract
The paper will present the application and effects of a road maintenance technique combining polymer modified asphalt emulsion and glass fibers. This process entered the North American market in New York State, 2003. New York State suffers from severe winter climatic conditions with abundant winter snow plow activities affecting in particular surface treatments such as chip seals/ surface dressings. Background data on these climatic conditions will be mentioned. The fiber-reinforced membrane can be used as a stand alone enhanced chip seal system or as a Stress Absorbing Membrane Interlayer, SAMI. When used as a SAMI system, the process provides an alternative cost effective treatment to some traditional techniques designed to delay reflective cracking. A new piece of equipment that applies the fiber reinforced membrane was developed in order to increase production and effectiveness of the technique and this will be discussed. Two studies at Texas Transportation Institute and Pennsylvania Transportation Institute will be presented showing to date the benefits of this process.

Thompson - September 2008

Page 2 of 20

1. Introduction In the United States the majority of pavements are composed of bituminous or asphalt concrete (AC). A large number of Portland cement concrete (PCC) also exist, especially on high-traffic volume highways. Two of the three major pavement distresses found in this region are fatigue and reflective cracking. Both these cracking mechanisms are considered low to moderate temperature phenomena attributed to both load and thermal stresses in pavements. As is well known and documented, in the later periods of their life, both types of pavements exhibit crack distresses: fatigue and thermal cracking in AC pavements, joint faulting and mid-slab cracking in the case of PCC pavements. Following an overlay application on an existing pavement, physical deterioration of this overlay takes place as a result of movement at the joints and cracks of the underlying pavement layer. When an asphalt overlay is placed over an existing pavement surface the former should be fully bonded by the tack/bond coat to the latter. Any movement taking place in the underlying pavement at a joint/crack will produce stresses in the overlay, which can promote reflective crack propagation if the stresses in the overlay exceed its fracture resistance. Reflective cracking occurs in nearly all type of overlays. Temperature induced horizontal movements concentrated at the underlying joints and cracks in the existing pavement lead to tensile stresses and is an important contributor to reflective cracking. Load or traffic induced vertical movements lead to shear stresses in the overlay that also contribute to reflective cracking. This study will show that a fiber-reinforced product, readily available in the North American market, performs under severe winter conditions not only in terms of seasonal and diurnal temperature ranges but also under aggression due to the effects of snowplows in New York State as compared with a normal CRS-2p based Chip Seal. In addition this study will show the product when used as an interlayer, extends the life of the overlay subsequently applied, several fold as compared with no treatment. Technical studies undertaken on this product will underpin the actual reality of using this product in New York and other similar States.

Thompson - September 2008

Page 3 of 20

2. Phenomena of Reflective Cracking Asphalt concrete overlays and surface treatments are some of the most economical forms of maintenance for a distressed pavement. However a problem often encountered is when cracks start to appear due to traffic and temperature induced fatigue through and on the surface of the overlay. While asphalt overlays and bituminous surface treatments are relatively cheap maintenance techniques, they are however, considered a quick fix because the existing cracks in the underlying layers eventually propagate upward through the overlay and reappears at the surface and hence forms what are referred to as reflective cracks. Similarly surface shrinkage cracks may also initiate top down cracking and lead to problems such as raveling and alligator cracking or crazing. Membranes, as bituminous surface courses and/or as an interlayer as described herein, have high shear and tensile strength in addition to high ductility and can act as crack relieving layers when placed on/or in-between the old and new surface. Because of their strength, the crack propagation through that interlayer requires higher energy and stress concentrations, ultimately leading to a delay in the formation of the reflective cracks. The interlayers ductility allows it to absorb some of the strain energy developed at the bottom of the new overlay as the wheel loads are applied cyclically on the top of the pavement diagram 1. Diagram 1
AC Overlay Interlayer

AC Overlay

Joint or crack

Cracked Layer

Cracked Layer

Cracking also occurs through asphalt concrete pavements due to cold temperatures and/or temperature cycling especially in the more northern states and provinces of North America. Cracking that occurs from cold temperatures is referred to as low temperature cracking whereas cracking due to thermal cyclic changes is referred to as thermal fatigue cracking. Both forms can propagate through new asphalt overlays or bituminous surface treatments as reflective cracks. These thermal induced cracks allow the ingress of water into the layers below, deteriorating these layers through freeze thaw cycles and/or by freezing and expansion of ice focal points that may produce an upward force on the pavement overlay. The result from both traffic and thermal induced fatigue is a deterioration of pavement life and a reduction in ride quality for the end-user.

Thompson - September 2008

Page 4 of 20

3. New York State and Climate

Diagram 2 Map of NY State courtesy of MAGELLAN Geographix

3.1 Background Eastern New York is dominated by the Great Appalachian Valley. Lake Champlain is the chief northern feature of the valley, which also includes the Hudson River. West of the lakes are the rugged Adirondack Mts. The rest of NE New York is hilly, sloping gradually to the valleys of the St. Lawrence and Lake Ontario, both of which separate it from Ontario. The Mohawk River, which flows from Rome into the Hudson north of Albany, is part of the New York State Canal System's Erie Canal, once a major route to the Great Lakes and the Midwestern United States as well as the only complete natural route through the Appalachian Mts. Most of the southern part of the state is on the Allegheny plateau, which rises in the SE to the Catskill Mts. New York City, in turn, attracts tourists from all over the world. On the extreme SE, the state extends into the Atlantic Ocean to form Long Island. The western extension of the state to Lakes Ontario and Erie contains many bodies of water, notably Oneida Lake and the Finger Lakes. In the northwest the Niagara River, with scenic Niagara Falls, forms the border with Ontario between Lake Ontario and Lake Erie. 3.2 The Physical & Climate of New York State New York State contains 49,576 square miles, inclusive of 1,637 square miles of inland water, but exclusive of the boundary-water areas of Long Island Sound, New York Harbor, Lake Ontario, and Lake Erie. The climate of New York State is broadly representative of the humid continental type, which prevails in the northeastern United States, but its diversity is not usually encountered within an area of comparable size. The character of the topography, and proximity to large bodies of water have pronounced effects on the climate.

Thompson - September 2008

Page 5 of 20

Cold winter temperatures prevail over New York whenever Arctic air masses flow southward from central Canada or from Hudson Bay. High-pressure systems often move just off the Atlantic coast, become more or less stagnant for several days, and then a persistent airflow from the southwest or south affects the state. This circulation brings the very warm, often humid weather of the summer season and the mild, more pleasant temperatures during the fall, winter, and spring seasons. The climate of New York State is marked by abundant snowfall. With the exception of the coast, the state receives an average seasonal amount of 100cm (40 inches) or more. The average snowfall is greater than 178cm (70 inches) over some 60 percent of New York's area. Seasonal snowfall, averaging more than 4.5m (175 inches), occurs on the western and southwestern slopes of the Adirondacks. A secondary maximum of 3.8m to 4.5m (150 to 180) inches prevails in the southwestern highlands, some 16 to 50 kilometers (10 to 30 miles) inland from Lake Erie. Heavy snow and lake effect squalls frequently occur, generating from 30 to 60cm (1 to 2 feet) of snow and occasionally 1.2m (4 feet) or more in a single day.

4. The fiber reinforced membrane process The product at the center of this study is a made in-place membrane that waterproofs and produces a fiber impregnated and reinforcing layer that allows a dissipation of some of the stresses generated in the pavement. The fiber-reinforced membrane produced is all applied on one unit, whereby two layers of polymer modified asphalt emulsion sandwich fiberglass strands that are chopped in-place by a special chopping and distribution system. Columns of fiber are held in a storage area on a trailer unit. Strands of fiberglass are then taken and fed pneumatically through lines to a fiber chopping unit. In advance of the chopping unit a layer of polymer asphalt emulsion is applied through a traditional slotted jet distributor spray bar arrangement. The strands of fiberglass are then chopped in place and randomly orientated by air in a chamber and are lightly blown down onto the surface of the polymer modified emulsion. A second layer of polymer modified emulsion then seals in the fibers and completes the membrane component Diagram 3 over. An aggregate layer is then applied.

Thompson - September 2008

Page 6 of 20

Diagram 3 Process layout

Storage of columns of fiber

Strands of fiber cut in place.

2nd layer of polymer modified emulsion applied

Fibers blown down onto 1st layer

1st layer of polymer modified emulsion

Fiber reinforced membrane

The fiber reinforced membrane process, FibreDec, was developed back in the late 1980s in the UK and has been used in a number of countries world wide from the UK to Australia and now North America. Initially the unit was limited to 2.4m (8ft) wide for a single lane pass and was also limited to the production application rate per day, typically 12,500-17,000m2 (15-20,000Yd2). It was also difficult to work on and hence for the North American market had to be revamped making it more appropriate for use. The outcome was the trailer unit previously described, that contains fiber in the storage area of the trailer unit capable of 38,000 to 64,000m2 (45-75,000 Yd2) of fiber reinforced membrane before the need for re-loading. This means a daily production rate of up to 85,000m2 (100,000Yd2). In addition the application width of the unit is now capable of making a 4 metre (13ft5) pass. This ensures in the majority of instances a full lane pass for the roads in North America. We also addressed the problem of working on and maintaining the unit by making the spray bar and fiber configuration unit fold vertically in the middle for easier access.

Thompson - September 2008

Page 7 of 20

The fiber-reinforced membrane has been used as the wearing course in traditional chip sealing markets adding value by extending the life of the seal and as a Stress Absorbing Membrane Interlayer, SAMI to compete against such products as Geotextiles and the like. This whole process was designed to delay the process of reflective cracking and seal alligator cracks within and on a pavement structure. It acts not only as a wearing surface but can be used at various levels within the pavement structure from surfacing pre-primed gravel bases, to interlayers before HMA, Ultra-Thin mixes such as NovaChip; Microsurfacing or Slurry Seal; giving in the interim before the final overlay, a resilient wearing surface.

4.1 The fiber reinforced membrane as a wearing surface The first type of application for this product is a bituminous wearing surface course treatment akin to a chip seal. This application combines a special polymer modified asphalt emulsion at typically 1.8 to 3.2m2 (0.4-0.8Gal/Yd2), chopped glass fibers of nominally 60mm (23/8) length at a rate of 30g/m2 to 120g/m2 (0.06lbs to 0.22lbs/Yd2). This mixture produces a membrane that acts as a highly resilient waterproofing layer that effectively bridges cracks with fiber and seals the cracks with a residual polymer asphalt membrane. The membrane is superior to traditional bituminous or chip sealing surface treatments as greater tensile strength properties resist stresses placed upon it. This is shown in the Photo Log below and over the next page where a side-by-side evaluation was performed with a regular chip sealing emulsion in New York State back in 2003. This was subjected to progressive winters where extensive snowplow operations have pronounced effects on all types of pavements because of their carbide blades. A 5-year evaluation study of the fiber reinforced membrane process was performed in upstate New York. One side of the chosen road was treated with a conventional chip seal overlay placed while the other with the FiberMat Type A process as the fiber-reinforced membrane is called when used as an enhanced chip seal. This road surface had many reflective cracks likely caused by years of stress due to seasonal shifting and snow plow damage. After one year of wear and tear on this test road, longitudinal reflective crack had started to reappear on the regular chip seal side while no cracks had appeared on the fiber-reinforced membrane side of the road. Note that cracking was present on both sides of the road prior to application.

Thompson - September 2008

Page 8 of 20

The following year, longitudinal reflective cracking and snowplow damage were evident on the chip seal side while the FiberMat Type A side was still in very good condition.

On the third year of evaluation, further evidence of snowplow damage was noted on the chip seal side while only slight damage started to occur on the fiberreinforced membrane side; all the while the FiberMat Type A membrane itself was still intact.

By year four, the chip seal side showed significant deterioration. It was now becoming a candidate for total rehabilitation. The FiberMat Type A side, after 4 years of use, only just started to show some minor damage.

Year five showed that the original chip seal side had fully deteriorated, with aggregate loss and raveling along with all afore mentioned distresses; a candidate now for full depth reclamation. However, the condition of the FiberMat Type A

Thompson - September 2008

Page 9 of 20

side altered minimally from the previous year. It was still years away from requiring any repairs or rehabilitation.

Work utilizing the Mini-Fretting test on this process was carried out back in the 90s, and showed the benefit of using the polymer modified emulsions for this process. It also showed that the aggregate layer applied firmly penetrated into the membrane and the fibers were able to lock-in the aggregate thus ensuring greater aggregate retention. Further work was carried out, this time showing in terms of the Vialit Adhesion Test at 5oC a dramatic improvement when using a polymer modified versus non-modified asphalt emulsions This resistance to general wear and tear on the surface is also linked to the experiences in New York, that of, under normal practices & conditions a greater resistance to aggression from such things as snow plows. In addition the field study from Australia, (Lysenko & Scott, 1998), on various bases showed that crack retardation was effective at a 3 fold improvement over nonfibered reinforced membrane treatments. One of the aspects here was the fact that the membrane produced was held in tact after traffic loading and the aggregate retained was firmly set in place. If we consider then a simple cost analysis, Equivalent Annual Costs, EAC, then for typically $ 2.25/m2 spent on the fiber process as a wearing surface and $1.80/m2 for a modified chip seal. The EAC is calculated as = Price per unit or SY of material/ expected life of treatment. For the Fiber Reinforced Membrane For a regular polymer modified chip seal = $2.25/ 6 years = $0.38/yr = $1.80/ 3 years = $0.60/yr

This is just assuming a two fold improvement when using the fiber reinforced membrane versus a regular polymer modified chip seal. In reality the fiber process is lasting three and more times longer than a regular chip seal surface; as in New York. Therefore although the initial outlay is higher for the fiber process the actual return on that investment is short.

Thompson - September 2008

Page 10 of 20

4.2 The fiber reinforced membrane as an interlayer

When the fiber reinforced membrane is used as an interlayer treatment, known as FiberMat Type B, then again there are benefits to be had. But before consideration of this aspect the requirements of a successful interlayer system should possess must be considered.

Diagram 4 - SAMI

The ability to provide a strong, waterproof membrane. The ability to absorb some stresses generated in the pavement and give enhanced tensile properties. To be installed easily, quickly without excessive preparation. Cost effective in whole life costs of the pavement. Easily recycled.

As with all systems that act as SAMIs there is an imperative need for a tack or bond coat to waterproof and seal the existing surface and provide an anchor of adhesion for the subsequent layer(s). With the fiber reinforced membrane this provides the polymer modified asphalt emulsions applied through a split bar giving the waterproofing and sealing characteristics. It is not suggested that all the stresses will be absorbed but just some stresses as there are no SAMI systems in the market that can proclaim to absorb all the stresses generated in a pavement. The aim is to reduce the effect of the stresses within the pavement layer and interlayer by absorption. Tensile properties are also important to give enhanced integral strength to the layers, especially when used as the enhanced chip seal wearing surface allowing greater retention of the aggregate. With the fiber glass cut and introduced between the spray bars this gives the ability to withstand some of the stresses and give enhanced tensile properties. Cost effective as shown from the previous example is more so when comparing this membrane against traditional interlayer treatments such as geotextiles and polymer rich mixes that may be as high as $12/m2 ($10/Yd2) or more installed in comparison. In addition when used as an interlayer with an ultra-thin wearing course

Thompson - September 2008

Page 11 of 20

the need for mill and fill with HMA is eliminated and the cost savings are considerable- Diagram 5.

Diagram 5 Cost Comparison

Unlike many SAMI systems the layer ideally should be easily placed and shaped without the need for excessive preparation such as truing and leveling courses or extra milling. The fiber reinforced membrane can be quickly installed without the need for excessive preparation. It is made on the move, cut to size and tailored to shape. In subsequent years it needs to be recycled, easily milled and reprocessed back at a HMA plant without having to constantly remove wads of textile from a conveyor as with some geotextiles for example. The fiber reinforced membrane can be easily recycled and used more usually in base RAP courses. The fiber reinforced membrane truly fits into the family of what are known as Stress Absorbing Membrane Interlayers, SAMIs and we would suggest is the only true SAMI that encompasses what the customer be they a Contractor or a Governmental DOT/County or Town really need. We have applied nearly 4.25 million square meters (5 million square yards) of fiber reinforced membrane in New York alone since 2003. About 60% of the projects to date throughout North America have been as an interlayer application. Often this has been overlaid with an ultra-thin hot mix overlay as shown in the photo of the core sample below. The rule of thumb, at least in New York State is that a pre-existing crack will propagate through a new overlay at a rate of about 25.4cm (1 inch) per year. Numerous projects with ultra-thin surfaces as the overlay, Photo 6, have been inplace since 2003 and have no cracks re-appearing whereas nearby regular projects of HMA with non fiber reinforced membrane treatment have cracks re-appearing within the 1 year period.

Thompson - September 2008

Page 12 of 20

Ultra-thin overlay NOVACHIP FiberMat Type B

Photo 6 FiberMat Type B Interlayer with NovaChip Ultra-thin overlay

5. US based Research The benefit of using the fiber reinforced membrane as the interlayer has not only been proven in the field but also through laboratory studies here in the US. 5.1 The Texas Transportation Institute, TTI, study mimics the thermal cracking phenomena and is aimed at generating a cracking number which can be compared against other systems. At TTI use is made of the Overlay Tester that is a displacement controlled repetitive loading machine to initially produce a small crack (due to tension) at the base of the test specimen and then continues to induce repetitive horizontal displacements which causes the crack to propagate upward through the specimen. This process is intended to simulate the cyclic tensile stresses of pavements due to periodic thermal variations. R.L Lytton has conducted studies on reflective cracking and fracture mechanics since the early seventies. The TTI overlay tester (simulating the cyclic horizontal displacement related to thermal shrinkage) has been developed in the early eighties. Thermally induced fatigue is the primary cause of reflective cracking initiating in pavements. Secondly Traffic Induced Fatigue

Thompson - September 2008

Page 13 of 20

comes into play followed by Bending/curling fatigue of the pavement. The testing protocol developed is summarized below:

1. TTI Overlay Tester Data output:

2.

Fracture mechanics modelization; predicted cracking life

3.

Test section / Model validation.

- Linear horizontal Load P vs. time - Horizontal displacement (u) - Crack length (c) vs. cycles (N) - Elastic/ relaxation Modulus of the Complex (overlay + interface) @ initial loading.

Cycles at failure: N = h _ 1__ de

0

A(Ki)n

Test section with TXDOT Verification of material fracture mechanics properties on core samples. - Section monitoring: Predicted Cracking Life verification.

Where: - A, n fracture mechanics properties deducted from Overlay tester data. - Ki Stress intensity factor deducted from measured modulus thru finite elements analysis. - Sample depth: h. Reinforcing factor & Crack speed Index

However the Finite Element Analysis model used by TTI assumes that the crack propagated will do so vertically this is not to date the case however with the fiber reinforced membrane. The findings from the study concluded that - There was horizontal crack propagation along the FiberMat Type B interface, Photo 8, rather than by cracking vertically above as in control samples, Photo 7. This is consistent with previous studies at Nottingham University in the UK.

Photo 7 Control Photo 8 FiberMat Type B SAMI - Generally, specimens containing FiberMat Type B improved cracking resistance in the small overlay testers 3 to 4 times more than control samples. The large overlay FiberMat Type B samples survived 14 times more compared to the control. Diagram 6 - Small Overlay Tester

Thompson - September 2008

Page 14 of 20

5.2 Pennsylvania Transportation Institute, PTI, Work was also carried out at Penn State Test Track where they evaluate commercial buses for use in the US. This is part of the Pennsylvania DOT approval process for Municipal Services with Pennsylvania Transportation Institute, PTI.

HMA Sections

PCC Sections

FiberMat Type B
HMA Overlay

FiberMat Type A

Cracked HMA
Base

Cracked HMA
Base

PCC
Strong Base

PCC
Weak Base

Several areas were applied with fiber reinforced membrane of the types A & B and to date, as the sections were laid last year, are performing very well. In addition three live sites where chosen throughout Pennsylvania and again these are performing well. A preconstruction distress survey was also done for all the sections. The sections were divided into a grid of 6x6 ft parcels. Monthly distress surveys were and are conducted during this period for the sections constructed at the PTI test track. Distress observations and pictures are routinely taken as part of the visual survey process. The observed cracks are identified, measured, and matched with the original cracks that existed before the overlay. Some of the techniques employed as part of the approval process included Accelerated Pavement Testing is performed using the Mobile Model Load Simulator Scale 3, (MMLS3), unit that determine loss of aggregate on the test sections. Chipping or aggregate loss was inspected for the FiberMat Type A & control and experimental sections using the MMLS3 as seen in Figure 7. Evaluation is done by weighing the mass of aggregates chipped from the pavement after a certain number of wheel cycles are applied.

Thompson - September 2008

Page 15 of 20

(a) (b) Photo 9 a) MMLS3 testing over a designated area for aggregate loss; b) MMLS3 covered with a tarp to prevent any loss of aggregate from leaving the area. For the fiber-reinforced chip seal section versus the control no major differences were observed initially but are continuing to be monitored periodically. However field cores were taken from the fiber-reinforced SAMI experimental and control sections. A thin ring corresponding to the polymer modified emulsion can be seen in Photo 11(b)

Photo 10(a) Cored control

Photo 10(b) Cored FiberMat SAMI

Fiber-bitumen emulsion

11(a) 11(b) Photo 11(a) Core from FiberMat Type B section; and b) FiberMat Type B section (black ring consisting of the fiber-bitumen emulsion) It can be seen in the top two photos 10 that there is a substantial difference when the fiber-reinforced system is employed as a SAMI, Photo 10(b). Crack propagation is delayed for the fiber-reinforced samples after 18months, while cracks propagate through the control section Photo 10(a).

Thompson - September 2008

Page 16 of 20

In addition and not part of the formal evaluation, X-ray Tomography was carried out on some of the core specimens. These clearly showed the layer of fiberreinforced membrane containing the fibers; aggregate fracture of the old underlying layer; and distinguish between voids within the mix. This was more from an academic standpoint than anything else but shows the uniformity of the fiber-reinforced membrane produced.

Photo 12 (a) - Cross section of core

Photo 12(b) Top view

Bending Beam Rheometer, (BBR), analysis was carried out on the fibermembrane albeit with a great deal problems associated with sample preparation. The results showed a definite increase in stiffness of the residual binder due to the inclusion of the fibers. The fibers act with the residual binder and reduce the deflection caused due to the load. Furthermore, it was seen that the stiffness of the specimens with fibers is greater than the case of samples without fibers. The creep rate (m value) is lower in case of samples with fibers (0.192 and 0.173) and higher in case of samples with without fibers (0.462 and 0.431), suggesting an increase in elastic behavior when the glass fibers are added.

Photo 13 BBR Samples

Thompson - September 2008

Page 17 of 20

After applying one million cycles on top of pre-existing cracks on both FiberMat Type B and the control sections, no visible reflective cracks appeared initially. However, after taking the cores on top of the pre-existing pavement cracks, a hairline crack was visible on the control section. Therefore, it was decided to further investigate that area using the Portable Seismic Pavement Analyzer (PSPA). The PSPA is an instrument designed to determine the variation in modulus with depth of exposed layer be it concrete or asphalt. The operating principle of the PSPA is based on generating and detecting stress waves in a medium. The device consists of two transducers and a source packaged into a hand portable package, photo 14(a), which performs the Ultrasonic Surface Wave (USW), and Impact Echo (IE) tests. The USW method was used on the project to determine the modulus of FiberMat Type B and the control section. A number of areas were chosen to calculate the modulus of the top 2 using the PSPA.

Photo 14(a) Portable PSPA

Photo 14(b) Transducer and source over crack

Moduli from the four sections were determined at -7oC and at 7oC (Table 5). At -7oC, it was observed that for locations without pre-existing cracks, the modulus of the FiberMat Type B section was similar to that of the control section, photo 14(b). However for locations with pre-existing cracks, the modulus at -7oC of the Control section is higher than that of fiber-reinforced. This intimates that the preexisting crack reflected to the upper layers of the control section while it did not in the fiber-reinforced. The higher modulus of the control stems from the high modulus of ice entrapped; frozen water entrapped; in the reflected crack in the top 2 of the control section which implies more voids or the reflective crack is more prominent. Comparing crack and no crack sections, both crack sections are higher to the ones without cracks, because of this high modulus of ice.. For the higher temperatures on top of the pre-existing crack, the fiberreinforced section has a higher modulus than the control due to absence of the reflective crack. For the section on top of no pre-existing cracks, the fiber-reinforced section has a lower modulus, due to the inclusion of the interlayer which has a lower modulus than regular HMA overlay. This fact makes the treated section less stiff and more ductile and self-healing.

Thompson - September 2008

Page 18 of 20

6. Conclusions Pavement preservation is one of the key tools for road engineers around the world to prevent the complete fatigue of their network and to optimize their road budget. In this field of pavement preservation, the fiber reinforced membrane chip seal (surface dressing in the UK) or as a SAMI will bring one of the most important techniques to the fore. The use of fiber reinforced systems with polymer modified asphalt emulsion will enhance this process and will give another appropriate tool in their range of maintenance products. The new applicator unit developed is more ideal and applicable for the North American market than previous versions. This allows coverage of single lane widths, higher production per day and easier to maintain. The long term resilient characteristics of the fiber reinforced membrane, as shown in New York State during five winters, gave local road engineers an excellent technique to help them maintain their pavements. At low temperatures, the results obtained in New York State after three winters showed a combination of advantages: 1. Economical in terms of Equivalent Annual Costs when compared with traditional systems. 2. Resilience under winter conditions and maintenance (snow plough effects and low temperature). SAMI (Stress Absorbing Membrane Interlayer) systems are also a domain road engineers can gain by using the fiber reinforced membrane system with hot mix asphalt. The action of the polymer modified asphalt emulsion will bond both materials and therefore give good reflective crack resistance when used with the fibers. The polymer modified asphalt emulsion is the best bond component for this application. The use of this system with an ultra-thin overlay can give the road engineer substantial savings over traditional mill and HMA fill practices and effectively reduce the environmental impact of road construction. In cold climates, reflective cracks can reveal the status of the pavement. Fatigue cracks indicate the end of the life of the wearing course if there is no deformation, these techniques will help the road engineer; by keeping water out and insuring a good protection of the pavement to crack propagation. Thermal cracks also can be dealt with. Those cracks especially during the thaw period will be abated as the fiber reinforced membrane gives great imperviousness and flexibility to and therefore helps prevent water penetrating into the pavement. The higher flexibility given by the combination of the polymer modified asphalt emulsion and fibers will insure good performance for longer than usual. Research done in the USA showed the great advantage of this technique to delay crack propagation coming either from fatigue or thermal effects. The findings from the TTI study concluded that There was horizontal crack propagation along the FiberMat Type B interface rather than by cracking vertically above as in control samples. This is consistent with previous studies.

Thompson - September 2008

Page 19 of 20

Generally, specimens containing FiberMat Type B improved cracking resistance in the small overlay testers 3 to 4 times more than control samples. The large overlay FiberMat Type B samples survived 14 times more compared to the control.

From the PTI study BBR results suggest a higher ductility at lower temperature for the fiberreinforced membrane interlayer which would delay the onset of cracks reflecting upwards. Cores removed from both sections having pre-existing cracks show that reflective cracks occur in the control section; whereas, reflective cracks are absent in the fiber-reinforced sections. X-ray Tomography also showed evidence of fiber-reinforcement homogeneity and characteristics of existing substrates not normally seen. An interesting tool for evaluation of mixes in general. One million cycles of MMLS3 on top of pre-existing cracks did not induce any visible cracks neither on the treated nor the control sections. Results from Portable Seismic Pavement Analyzer show that more voids and micro/macro cracks are present in the overlay surfaces in the control section as compared to the fiber-reinforced sections. Visual observations of both studied sections suggest that cracks are more likely to reflect in the control section than in the fiber-reinforced section.

Research undertaken in the USA will continue on in order to understand better, through American tests, the behavior of both systems. Finally with the future use of mechanistically driven pavement structural design in the US market, we should be able to give data in order to introduce this SAMI system in the design.

Thompson - September 2008

Page 20 of 20

References

I.L.Al-Qadi, M.M.Elseifi, (2003), A Simplified Overlay Design Model against Reflective Cracking Utilizing Service Life Prediction, TRB 82nd Annual Meeting January 12-16 2003, Washington D.C. W.G.Buttlar, D.Bozkurt, B.J.Dempsey, (2000), Cost-Effectiveness of Paving Fabrics to Control Reflective Cracking, Journal of the Transportation Research Board, No.1730, National Research Council, National Academy Press, Washington D.C. G.S.Cleveland, R.L.Lytton, J.W.Button, (2003), Reinforcing Benefits of Geosynthetic Materials in Asphalt Concrete Overlays using Pseudo Strain Damage Theory, TRB 82nd Annual Meeting January 12-16 2003, Washington D.C. K.E. Cooper, S.L. Held, (1987), The evaluation of fibre reinforcement techniques to inhibit reflection cracking in overlays, Nottingham University Dept of Civil Engineering. P.Jayawickrama, R.L.Lytton, Methodology for Predicting Asphalt Concrete Overlay Life against Reflection Cracking. J. Lysenko, G Scott (1998), Are your roads getting enough fibre?, AAPA Pavements Industry Conference. C.Yeates, ( 1994) Evalaution of fibre-reinforced membranes., Colas Limited. F.Chaignon & M.Thompson, (2006), Road Reflective Cracking System Under Cold Climatic Conditions, World Congress On Emulsions October 2006, Lyons, France