Question 5 of 7: Marshall optimum asphalt content, and the forms of asphalt

Question 5: Marshall optimum asphalt content, and the forms of asphalt (10 + 10 marks)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

Part (a) — Optimum asphalt content by the Marshall method

Given. Five Marshall property curves, each plotted against asphalt cement content by weight of total mix. The curves are supplied graphically only, so the first task is to read them. The values scaled from the printed plots at the test asphalt contents are:

Find. The optimum asphalt content by the classical Marshall averaging procedure, and a verdict on whether the mixture at that asphalt content satisfies the Asphalt Institute mix design criteria.

Approach. Locate the asphalt content at the peak of the unit-weight curve, the asphalt content at the peak of the stability curve, and the asphalt content at the median design air void content of 4 percent; take the optimum as the average of the three; then read every property at that optimum and test it against the Asphalt Institute criteria for the appropriate traffic level.

1. Part (a) — locate the peak of the unit-weight curve. Fitting a parabola through the three points that bracket the maximum, at 5.5, 6.0 and 6.5 percent, and differentiating gives the turning point at $$AC_{\gamma} = 6.19 \approx \boxed{6.2\ \text{percent}}$$ This is the asphalt content at which the compacted mixture is densest — beyond it, the additional binder begins to push the aggregate particles apart faster than it fills voids.
2. Locate the peak of the stability curve. The stability data are supplied only to 6.5 percent, but the maximum is comfortably inside that range. Fitting a parabola through the points at 5.5, 6.0 and 6.5 percent gives $$AC_{S} = 5.86 \approx \boxed{5.9\ \text{percent}}$$ Peak stability occurs slightly leaner than peak density, which is the usual ordering: the mixture loses internal friction as soon as the binder film becomes thick enough to lubricate the aggregate contacts, and that happens before the mixture reaches its densest state.
3. Locate the asphalt content at 4 percent air voids. Four percent is the midpoint of the 3 to 5 percent design range for a dense-graded surface mixture. Interpolating linearly on the descending limb of the air-void curve between 5.5 percent asphalt (4.7 percent voids) and 6.0 percent asphalt (2.7 percent voids), $$AC_{V} = 5.5 + 0.5\times\frac{4.7-4.0}{4.7-2.7} = 5.5 + 0.175 = \boxed{5.68\ \text{percent}}$$
4. Average the three to obtain the optimum asphalt content. The classical Marshall procedure takes the design binder content as the mean of the three asphalt contents identified above: $$AC_{\text{opt}} = \frac{6.19 + 5.86 + 5.68}{3} = 5.91 \Rightarrow \boxed{AC_{\text{opt}} = 5.9\ \text{percent by weight of mix}}$$
5. Read every property at the optimum. Interpolating each curve at 5.9 percent gives a unit weight of 156.9 pcf, air voids of 3.1 percent, a Marshall stability of 1790 lb (7960 N), a flow of 12.8 hundredths of an inch, and voids filled with asphalt of 80.8 percent. The voids in the mineral aggregate follow from the identity relating the three void measures, $$\mathrm{VMA} = \frac{V_a}{1 - \mathrm{VFA}/100} = \frac{3.10}{1-0.808} = \boxed{16.1\ \text{percent}}$$ which is comfortably above the 14 percent minimum that applies to a 12.5 mm nominal maximum size mixture, so the aggregate skeleton itself is adequate.
6. Test the mixture against the design criteria. Comparing each property with the Asphalt Institute criteria for heavy traffic, that is more than $10^{6}$ ESALs:

The mixture therefore satisfies flow, air voids and VMA, sits exactly on the stability minimum, and clearly fails the voids-filled criterion. The 1790 lb stability is a 0.6 percent shortfall against the 1800 lb minimum, which is well inside the precision with which a printed curve can be scaled, so stability should be regarded as marginally satisfied rather than failed. The VFA result is a different matter: at 80.8 percent it is six points above the upper limit, which is far outside any reading error and is a genuine finding.

Interpretation and recommendation. A high voids-filled value with acceptable VMA and low air voids is the signature of a mixture carrying rather more binder than its aggregate skeleton can safely hold. Under heavy traffic such a mixture is at risk of further densification in service, followed by flushing of binder to the surface and rutting. Two remedies are available. The simpler is to reduce the design asphalt content: at 5.7 percent the air voids rise to 3.9 percent, VFA falls to 76.4 percent and stability is essentially unchanged at 1770 lb, and at 5.5 percent the mixture returns air voids of 4.7 percent, VFA of 72.0 percent, flow of 12.0 and stability of 1750 lb, which satisfies every Asphalt Institute criterion for medium traffic and satisfies the heavy-traffic voids and flow criteria as well. The more thorough remedy, if the mixture must serve heavy traffic with full margin, is to revise the aggregate gradation to raise the VMA by a point or two, which permits the same film thickness at a lower voids-filled value. On the evidence supplied, the mixture is a sound medium-traffic dense-graded surface mixture at about 5.5 to 5.7 percent asphalt, and should not be used at 5.9 percent under heavy traffic without a gradation change.

Part (b) — The forms of asphalt

Asphalt cement at ambient temperature is a stiff, nearly solid material, and almost everything one wants to do with it other than hot-mix paving requires it to be made temporarily fluid. The four families the question names are four answers to that requirement, plus one that changes the binder’s properties rather than its handling.

Cutback asphalts are asphalt cements blended with a petroleum distillate so that they can be applied cold or warm and then stiffen as the solvent evaporates. They are classified by the volatility of the solvent, which determines how fast they cure. Rapid-curing cutbacks, designated RC, use a naphtha or gasoline-type distillate and are used for tack coats and surface treatments where the binder must set quickly. Medium-curing cutbacks, MC, use kerosene and cure more slowly, which makes them suitable for prime coats and for cold-mix patching material that must remain workable in the stockpile. Slow-curing cutbacks, SC, use a heavy oil, cure very slowly and are used for dust control and light stabilisation. Within each class a grade number, such as RC-250 or MC-3000, gives the minimum kinematic viscosity and therefore the consistency. Cutbacks are simple to make and to use, they wet aggregate well, and they tolerate a wide range of application temperatures. Their decisive disadvantage is environmental and economic: the solvent is a valuable fuel that is deliberately evaporated into the atmosphere as a volatile organic compound. Emissions regulation in Canada and elsewhere has consequently pushed cutbacks out of most applications, and they survive mainly where an emulsion will not work, such as prime coats on dense bases and cold-weather patching.

Asphalt emulsions achieve the same fluidity without a solvent, by dispersing droplets of asphalt cement, typically 1 to 10 micrometres across, in water with the aid of an emulsifying agent. The emulsifier gives the droplets a surface charge that keeps them apart, and that charge classifies the product: anionic emulsions carry a negative charge and work best with electropositive aggregates such as limestone, while cationic emulsions carry a positive charge and adhere well to electronegative siliceous aggregates such as granite and quartzite, which is why cationic grades dominate in most of Canada. Grades are further described by setting speed, rapid, medium, slow and quick setting, as in CRS-2, CMS-2 or CSS-1h, where the C denotes cationic and a trailing h denotes a harder base asphalt. Emulsions break when the water separates and the droplets coalesce, leaving a continuous asphalt film, and cure fully as the remaining water evaporates. Their advantages are substantial: no solvent is lost, so they are far cleaner and cheaper; they can be applied to damp aggregate and in cool weather; and they are the basis of chip seals, slurry seals, microsurfacing, tack coats, fog seals and cold in-place recycling. Their limitations are that they cannot be applied in the rain or in freezing conditions, that they have a limited storage life and can settle or break in the tank if agitated wrongly or frozen, and that the choice of emulsion must be matched to the aggregate mineralogy and the weather, so they are less forgiving of poor site judgement than a cutback.

Asphalt primers are low-viscosity binders applied to an untreated granular base before the first asphalt layer is placed. Their function is not to bind but to penetrate: the primer soaks into the top few millimetres of the base, coats the fines, waterproofs the surface, binds the loose material so that construction traffic does not tear it up, and promotes adhesion between the base and the first bituminous layer. Traditionally the medium-curing cutbacks MC-30 and MC-70 have been the standard prime materials, precisely because a solvent carries the asphalt into the pores in a way water does not; specially formulated penetrating emulsions have been developed to replace them for environmental reasons and are now common, although they are less effective on tightly bound dense-graded bases. A prime coat should be distinguished carefully from a tack coat: a prime is applied to a granular surface and penetrates it, whereas a tack coat is a light application of emulsion between two bituminous layers to bond them together and does not penetrate at all.

Modified asphalts address a different problem. Conventional asphalt cement has a narrow useful temperature range, being too soft in summer and too brittle in a prairie winter, and modifiers widen it. Elastomeric modifiers, principally styrene-butadiene-styrene block copolymer and styrene-butadiene rubber latex, form an elastic network through the binder that raises the high-temperature stiffness and improves elastic recovery, so the mixture resists rutting and reflective cracking and recovers from deformation. Plastomeric modifiers such as ethylene-vinyl acetate and polyethylene stiffen the binder at high temperature but add little elasticity. Crumb rubber from scrap tyres, blended either wet into the binder or dry into the mixture, improves both high- and low-temperature behaviour and consumes a waste stream. Other modifiers include polyphosphoric acid, sulphur, natural asphalts such as Trinidad Lake asphalt or gilsonite, cellulose and mineral fibres for stone mastic asphalt, and anti-stripping amines for moisture resistance. The advantages are longer service life, greater rut and fatigue resistance, improved low-temperature cracking performance, which matters a great deal in Canada, and the ability to meet wide performance grades such as PG 58-34 that a neat binder cannot reach. The disadvantages are cost, typically 30 to 80 percent above a neat binder, higher mixing and compaction temperatures with the associated energy use and emissions, potential storage instability requiring agitation, and greater sensitivity in both plant and laydown operations. Their use is therefore justified where the traffic or the climate is severe, and is difficult to justify on a lightly trafficked road.