Category Stone Matrix Asphalt. Theory and Practice

In countries where paving grade bitumens (unmodified) are used in SMA, it is usu­ally assumed that the application of hard binders to ensure improvement in rut resis­tance is not necessary. It is generally accepted that, ensuring rut resistance should be accomplished by creating the correct mineral skeleton. So medium-grade binders are justified for use in SMAs, such as the popular 50/70 binder used in Germany or the 70/100 used in the Netherlands. A very soft binder like a 160/220 may be seen in Europe (e. g., in Finland); in very cold climate conditions, use of such a soft binder in not surprising.

The selection of a binder for an SMA wearing course is determined, on the one hand, by the temperature range over which the pavement is expected to perform and, on the other, by the expected traffic loads. The type of binder selected is limited by the local temperature conditions, which in most cases disallow for the application of excessively hard binders to prevent low-temperature cracking. In cold climate coun­tries, softer PMBs are applied due to their high elastic recovery at low temperature. Some examples of these binders are presented in Chapter 5. A tendency to change the type of binder in consideration of the increase in traffic load—from paving grade, through multigrade, and up to a low penetration PMB—is clearly evident.

Despite particular emphasis being placed on securing rutting resistance mainly by the SMA skeleton, there is no doubt that the proper selection of a binder is an extra element supporting the stone skeleton performance. German research studies (Graf, 2006; Kreide, 2000) show that in most cases modified binder significantly increases SMA rutting resistance.

3.3 SUMMARY

• SMA mastic consists of fine (or passive) aggregate, filler, stabilizer, and bituminous binder. Binder and filler together create mortar.

• One reason for using fine aggregate is to fill the voids among the coarse (active) grains and participate in their interlocking.

• The term filler denotes all the aggregate that passes through the limit sieve (0.063 mm in Europe or 0.075 mm in the United States). It is suggested that the properties of the whole filler fraction passing through limit sieve (i. e., added filler plus aggregate fines) be tested. Test results for the entire filler fraction may differ from the results of testing only the added fillers.

• An excess of filler leads to the stiffening of the mortar or mastic and an increase in its susceptibility to cracking.

• The correct filler-binder (F:B) ratios and mortar properties have an impact on the workability, compaction, and permeability of a given mixture. In

fact, formulating a universal F:B index is a kind of approximation; such an index should be defined for each filler individually.

• Fillers with high contents of voids (analyzed according to any Rigden’s method) should be used cautiously due to the insufficient content of free binder in a mortar, which results in stiffening the mortar.

• Adding hydrated lime at a rate of about 1.0-1.5% (m/m) is a positive solu­tion, improving water resistance and the adhesion of the binder-filler to the aggregate.

• SMA binders can be divided into paving grade (unmodified) binder, PMB and special (most often multigrade) binder. Each of these types can be success­fully used, providing the mix is well designed with properties appropriate for the traffic loading and local climate.

• The application of hard (low-penetration) paving grade binder (e. g., 35/50) for wearing courses is not recommended due to the high risk of low-tem­perature cracking.

In this section, we will examine the different types of bituminous binders used in SMAs and methods for selecting one.

3.3.1 Types of Applied Binders

Various SMA binders may be seen in the highway engineering practices of many countries. These binders can be divided into paving grade bitumens (unmodified), polymer-modified bitumens (PMB), and special binders (multigrade and others).[16]

Paving grade bitumens are frequently used. In Europe the most commonly used binder is the penetration graded 50/70 type, and to a lesser extent the 70/100 type. Performance graded binders are routinely used in the United States and may or may not be polymer modified, depending on the base asphalt, the desired range of temperatures at which the binder is expected to perform, and the anticipated traffic level.

Polymer-modified bitumens are increasingly being used. They are found mostly in mixtures laid on roads with high traffic loadings, in special conditions (e. g., road crossings, slow traffic lanes) or on special pavements, although one should remember that PMBs require a suitable technological regime. U. S. experts recom­mend that highly modified PMBs with polymer contents over 5% (m/m) should not be used (Asphalt Review, December 2004). High polymer contents create problems with fast stiffening and increased difficulties during compaction. Use of highly- modified PMBs combined with other stiffness enhancers should be especially avoided.

Properly chosen and tested modified binders could increase rutting resistance and decrease the risk of low-temperature cracking of an SMA pavement.

Among the special binders, multigrade ones are sometimes used in SMA mix­tures; for example, in Australia (NAS AAPA, 2004) they are used for heavily traf­ficked pavements.

Summing up:

• When designing an SMA mixture, one should not decrease the content of filler below a minimum value (defined by using a 0.063-mm sieve in

Europe or a 0.075-mm sieve in the United States). The lack of filler will be reflected in a decrease in the durability of the mix and the void content and an increase in the risk of the appearance of fat spots.

• An increased amount of filler causes higher viscosity of the mortar, which promotes resistance to deformation. But one should not use an excessive amount of filler because it is easy to overstiffen the mortar, which can lead to cracking.

• Too fine a filler can cause problems since it absorbs much of the binder and may plasticize the mix.

• It is worthwhile to take note of Rigden’s test results, particularly when changing filler in the same mixture during continuous production. A new filler with a decidedly different void content than the former one may result in unwanted surprises.

• Adding hydrated lime at a rate of about 1.0-1.5% is a good move as doing so improves the binder-aggregate adhesion and boosts resistance to water damage.

In some countries a substantial amount of research has been dedicated to baghouse fines, considering them as potential material for use in mixtures. Their use has an economical aspect since the high efficiency of dust collectors in up-to-date asphalt plants collects considerable amounts of dust that are then available for use essentially free of charge. Using some of the dust from dust removal is an everyday practice in many countries. It has been applied to SMA, along with other types of fillers.

The results of studies on the practicality of using baghouse fines are quite diver­gent. One might conclude that the appropriateness of their use cannot be gener­alized. Properties of extracted dusts may differ widely, depending on their origin (i. e., source rock type). It is common knowledge that large amounts of collected dust added to mixtures may substantially stiffen them, making them susceptible to cracking and water damage, and asphalt mixes that contain them are not easy to place and compact. Despite that, baghouse fines can be a very good filler (Asphalt Review, December 2004).

Finally it is worth noting one more application aspect of collected baghouse dust. If we want to use baghouse fines as filler, they have to satisfy the standard requirements for fillers, including those concerning the repeatability and uniformity of obtained results. Frequent changes of aggregate types (i. e., rock types) may result in fluctua­tions in the mineralogical composition of the collected dust and its properties.

The European standard EN 13043 accepts baghouse fines as filler aggregate if they meet the standard requirements for fillers.

Many accessible publications (Iwahski, 2003; Judycki and Jaskula, 1999, Little and Epps, 2001) on the use of hydrated lime as an additive to asphalt mixtures have pointed out its positive effect. Apart from a substantial increase in water and frost resistance arising from an improvement in the binder adhesion to the aggregate, an increase in resistance to permanent deformation may also be noticed.

Hydrated lime is distinguished by its strong mixture-stiffening properties at high temperatures. It should be kept in mind that, according to Rigden, voids in hydrated lime can be very large. In connection with that, one should not exceed the standard con­tent of lime in an asphalt mixture, usually accepted as 1.0-1.5% of the aggregate mass. Such an addition of lime will benefit the mixture without the risk of overstiffening.

According to EN 13034, ready-made filler containing hydrated lime is called a mixed filler.

3.2.4.1 Commercially Produced Fillers (Added Fillers)

According to European terminology,! added fillers are made by crushing stone to produce fillers aimed at use in highway engineering. For a long time, the most popu­lar of these has been, and is likely to remain, limestone filler. Limestone filler is distinguished by its affinity with binder, which is one of its strong points. Therefore in Europe limestone filler is most often used for SMA.

The other important feature of industrially manufactured fillers is their repeat­ability and uniformity of parameters. Finally it is worth observing their constant and repeatable mineralogical composition.

3.2.4.2 Fly ashes

The use of fly ashes as fillers for SMA is uncommon. Their disadvantages include large specific area (fly ashes are very fine) and the spherical shape of the grains. So fly ashes have only been used to a limited extent and always need an engineering assessment before use.

The density of fly ashes is lower than that of crushed rocks or baghouse fines and fluctuates between 2.0 and 2.6 g/cm3. To obtain a similar volume share in a mineral
blend, ashes are metered in at a lower weight than a standard filler. The modified Rigden void content is usually less than 50% (Report FHWA-IF-03-019, 2003).

One popular approach used during design practice in many countries is to indicate the recommended range of the filler-bitumen ratio (or F:B index) by weight or vol­ume. Researchers in the United States have said that this factor better describes the maximum content of filler in the mix than does setting specific limits on the filler content. It is also worth adding that those studies have defined the maximum F:B index for asphalt concrete at the level of 1.2-1.5 (by weight) (Anderson, 1987). The F:B index was later altered to 0.6-1.6 in the Superpave method (Superpave Mixture Design Guide. WesTrack Forensic Team Consensus Report, 2001). Finally the sug­gested F:B ratio for SMA mixes is at 1.5 by weight, taking the total amount of dust on aggregates and added filler as the filler content (Harris and Stuart, 1995). But the F:B index has been criticized for some time, and there are suggestions regarding its replacement by other factors based on the free binder concept. As an example, Australian research studies (Bryant, 2006) suggest the application of an additional filler fixing factor (FFF) apart from the F:B index. Tests have proved that FFF may be also used to estimate the workability of a mixture.

It is necessary to remember that fillers differ markedly in terms of gradation, den­sity, and void content, therefore formulating a universal F:B index is only an approxi­mation. In fact, such an index should be defined for each filler individually. The goal is clear for each case—to produce a mastic that is neither too dry nor too soft.

Figure 3.4 illustrates the relationship between the content of voids in fillers accord­ing to Rigden’s method and the amount of filler needed to fix binder completely. The higher the F:B index, the more filler is needed to fix the binder. This relationship is illustrated by the graph; the estimated line of completely filled voids in the filler represents the zero-amount of free binder—all the binder is fixed.

As we can see, to fix the binder we need roughly two times more filler that has approximately 30% free voids (the lowest content according to Rigden) than filler that has 50% voids. In the latter case, less filler is sufficient to accommodate all the binder in the free voids (there is a lot of free space for binder within the high void content). It would be difficult for us to use such dependencies unless the filler manufacturer supplies data on the content of free voids according to Rigden or we conduct suitable tests ourselves. All in all, it is better to perform the tests in our own laboratories since eventually a voids parameter may be applied to the entire filler fraction of a mixture (i. e., including the filler fraction that may be coating the coarse and fine aggregate particles).

Another way to evaluate the properties of mortars and the F:B ratio is the applica­tion of the softening point (SP) method. According to a publication from Germany (Schroeder and Kluge, 1992) mortars with SPs between 85°C and 100°C perform

well in SMA. Also the Superpave binder test methods could be used for mortar test­ing (DSR[15] and BBR). In Chapter 8, one can find a short description of the methods and some additional remarks.

Let us imagine a set of grains that are going to be dry-compacted by tamping.[13] The result will be a mixture with its volume consisting of grains and some free spaces among them. In a regular binder mortar (blend of filler and binder), these free spaces in a compacted filler would be occupied by binder. The rest of binder would remain as excess filler. Thus binder contained in a mortar can be divided into the following two types (Figure 3.2):

Bulk volume of compacted mineral filler

FIGURE 3.2 Free and fixed binder concept. (From Harris, B. M. and Stuart, K. D., Journal of the Association of Asphalt Paving Technologists, 64, 54-95, 1995; Kandhal, P. S., Journal of the Association of Asphalt Paving Technologists, 50, 150-210, 1981. With permission.)

• Fixed binder—binder inside the voids (filling the voids among compacted filler grains)

• Free binder—excess binder remaining after the voids have been filled

As in the comparison of two fillers with differing gradations, here we may dem­onstrate much the same tendencies—the same quantity of two fillers but with dif­ferent contents of voids may bond differently to the amount of binder. Actually what really matters is the quantity of free binder, because the properties of the mas­tic are dependent on it. The lower the content of free binder in a mortar, the faster the growth of its stiffness (Harris and Stuart, 1995). The minimum amount of free binder has been defined in U. S. research, based on the modified Rigden method, as 30% (v/v) of an asphalt mortar (Anderson, 1987; Chen and Pen, 1998). With that level of free binder, filler grains are suspended in the binder and they do not touch each other. In addition, the rest of the mineral mixture (the coarse aggregate) will be coated by only the free binder, so it is important that a sufficient quantity is available.

Figure 3.3 illustrates the process of gradually filling the voids in a compacted filler; the binder essentially plays two roles—that of a lubricant making the reloca­tion of grains easier and that of a liquid in which they can be suspended.[14]

With a constant content of binder in an asphalt mixture, the quantity of free binder depends on the voids in the compacted filler. With fixed proportions of components in an asphalt mix, the quantity of free binder can be increased by changing to a filler

FIGURE 3.3 Gradually filling the voids in a compacted filler. (From Anderson, D. A., Guidelines on the use of baghouse fines. National Asphalt Pavement Association. Information Series 101-111, 1987b; Harris, B. M. and Stuart, K. D. Journal of the Association of Asphalt Paving Technologists, 64, 54-95, 1995. With permission.)
with a lower void content. Obviously the reverse is also true. The final requirements for the void content in a compacted filler may be defined as follows:

• They cannot be too high—so as to prevent fixing the whole binder, or too much of it, and to leave enough binder for the rest of the asphalt mix— otherwise the mortar will be too dry, stiff, and susceptible to cracking and water damage

• They cannot be too low, because too much unbonded, excess binder will create a greater risk of mix instability, excessive bleeding and binder drain – down, and the deformability (rutting susceptibility) of an asphalt mixture

The following are the recommended contents of voids in a dry-compacted filler:

• When using results of Rigden’s method (test according to EN 1097-4), the mini­mum content of voids in a dry-compacted filler should amount to 28% (v/v) and the maximum content should not exceed 45% (v/v) (Schellenberger, 2002).

• When using results of Rigden’s method modified by Anderson (as in the United States) (test according to Anderson [1987]), the maximum content of voids in a dry-compacted filler should not exceed 50% (v/v) (Brown and Cooley, 1999).

The main factors that influence the content of voids in a dry-compacted filler are as follows (Kandhal, 1981):

• Particle size

• Particle shape

• Particle surface structure

• Particle size distribution

Examples of void contents in various dry-compacted fillers tested according to Rigden’s original method and specific areas according to Blaine’s method [Schellenberger, 2002] are as follows:

Limestone (added filler) 27.7-31.6%

Diabase (baghouse fines) 30.4-34.2%

Limestone (baghouse fines) 28.3-32.1%

Dolomite (added filler) 27.1-28.1%

Dolerite-microdiabase (baghouse fines) 32.4-36.4% Greywacke (baghouse fines) 27.6-31.8%

The following are some sample results of void contents in various dry-compacted fillers according to Rigden’s method modified by Anderson (Schroer, 2006):

Mineral filler 39-47%

Baghouse fines 30-60%

Hydrated lime 66-71%

Fly-ash 37-57%

Generally, the results obtained from Rigden’s original method (as used in Europe) are slightly lower than the results from the modified Rigden method (as used in the United States). The reason is the difference in compactive effort. A similar test, which is an indirect usage of the Rigden voids concept, is the German Filler Test; this test shows good correlation to the modified Rigden method results (see Chapter 8).

Let us start with the definition of an aggregate’s specific area. Specific area is defined as the grains’ surface area related to a unit mass, usually given in terms of square centimeter per gram (cm2/g). Fillers are tested in Europe with Blaine’s method* according to EN 196-6 (see Chapter 8). The measured specific area depends on how much the parent material was reduced in size (broken up or crushed). The more the material was milled (crushed), the finer the filler, and the larger its specific area.

An example will serve to illustrate the influence of gradation on specific area size. Let us take two fillers and call them A and B. Let them both pass completely (100%) through a 0.063-mm sieve. Laser analyzer tests reveal significant differences in the material smaller than 0.063 mm. There are 20% and 90% of grains smaller than 0.005 mm in fillers A and B, respectively. This means the filler B has a higher content of very fine grains. What difference does this make? A higher content of finer grains means a substantial increase in their specific area. Because of these large differences in the gradation of the fillers that are less than 0.063 mm, the specific areas of fillers A and B may differ widely—even by two or three times—though, at first sight, a simple screening analysis through the 0.063-mm sieve may not suggest such a distinction.

If a filler makes up 10% of a typical SMA mix (100 kg per 1 metric ton), then the specific area of this ingredient alone amounts to 350 million cm2 or 3500 m2, which is equivalent to the area of a sports field that is 100 m long and 35 m wide. That is quite a lot, isn’t it?

In an SMA mixture, the grains’ surface area should be evenly covered with a binder film (a couple of microns thick or a bit more). Different specific areas of filler require different quantities of binder to coat them, which means a changeable demand for binder in a mortar. If the true sieve analysis of a filler passes unnoticed and only a simplified screening through the limit sieve is observed (as with fillers A and B), one should not wonder why the same SMA mixture behaves utterly differ­ently when fillers are changed. The specific area that needs to be coated with binder may be substantially different in one case versus another, and a correction of the added binder content may be necessary.

An increase in the specific area of filler requires an increase in the binder quan­tity just to preserve a suitable mortar consistency. However, an increase in the filler content with a constant binder content causes a reduction of the film-thickness coat­ing the filler grains, making the mix appear to be drier. A substantial increase in the filler-binder (F:B) index increases the risk of cracking. [12]

We can see that the concept of binder content in a mixture based on the specific area size is pretty vivid and appealing to the imagination. There are some analytic formulae approximating the required binder content related to the specific area of a mineral mix; one can find an example in The Asphalt Handbook (MS-4).

Before we leave the issue of filler gradation and specific area, it should be noted that much research has raised questions about a direct relationship existing between the gradation and the stiffening properties of a filler (e. g., Anderson, 1987). Also, the significance of the amount of material passing the 0.02-mm sieve and its specific area are questioned as they do not appear to influence mortar properties (Brown and Cooley, 1999). Thus it is time to proceed to another concept, that of voids in a dry – compacted filler.

The term filler means an aggregate that mostly passes through a specified sieve (0.063 mm in Europe, 0.075 mm in the United States). It should be emphasized that the material just discussed, which is generally called filler, denotes all the grains—that is, both those coming from the added filler and those occurring on fine and coarse aggregate grains in the form of dust. Thus if we want to know the behavior of a filler fraction in a given mixture, then all the grains below a specified sieve size in a final mineral blend should be separated, regardless of their source. All that material should be tested. If we only test the added filler, the results do not show the influence of the entire filler fraction on the properties of the mix.

The significant influence of filler on asphalt mixtures may be defined in the fol­lowing way (Anderson et al., 1982; Kandhal et al., 1998; Druschner, 2006): [11]

One could say that filler is the most underestimated component of SMA. After all, it constitutes from 8% to 12% of a mixture, which actually is a significant amount.

Next we will turn our attention to two concepts used in explaining the behavior of filler—the specific area of a filler and the content of voids in a compacted filler. These concepts can help describe various phenomena occurring in mixes.