Question 2 of 7: Joints in concrete pavements, pumping, and the forms of asphalt

Question 2: Joints in concrete pavements, pumping, and the forms of asphalt

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) — joints and load-transfer steel in jointed plain concrete pavement. A concrete pavement is a rigid slab that would crack of its own accord under drying shrinkage, daily and seasonal thermal movement, and the curling and warping produced by temperature and moisture gradients through its depth. The design philosophy of jointed plain (unreinforced) concrete pavement is not to prevent that cracking but to decide where it happens, so every joint in the system is a crack that has been put where the engineer wants it and detailed so that it can be sealed and can still transfer load.

Transverse joints run across the carriageway at right angles to the direction of travel and are almost always contraction joints. They are formed by sawing a groove into the fresh slab to about one quarter to one third of its depth as soon as the concrete will support the saw without ravelling, typically within four to twelve hours of placing. The groove creates a plane of weakness, and the shrinkage crack forms beneath it rather than at random. Spacing is governed by slab thickness and by the thermal coefficient of the aggregate; a common Canadian rule of thumb is a spacing in metres of about 21 to 24 times the slab thickness in metres, which gives roughly 4.0 to 5.0 m for a 200 mm slab, with a further restriction that the slab panel should not be markedly longer than it is wide. Too long a panel produces mid-panel transverse cracking; too short a panel wastes joints, and every joint is a maintenance liability.

Dowel bars are the load-transfer device at transverse joints. They are smooth, round, plain steel bars, typically 32 to 38 mm in diameter on a 200 to 250 mm slab, 450 to 500 mm long and spaced at 300 mm centres across the wheel paths. Being smooth and, on one side of the joint, coated or greased so that they are deliberately debonded, they allow the joint to open and close freely while still forcing the two slab edges to deflect together. That is the whole purpose: a wheel load arriving at an undowelled joint is carried entirely by the approach slab, whose corner deflects into the base and eventually breaks down, whereas a dowelled joint shares roughly 40 to 50 percent of the load with the leave slab. Dowels must be placed parallel to the pavement surface and to the centreline; misaligned dowels lock the joint, and a locked joint transfers the shrinkage movement to the next joint or cracks the slab around the bar.

Longitudinal joints run parallel to the direction of travel, normally on the lane lines, and control the longitudinal cracking that otherwise forms when a wide slab warps about its own axis or settles differentially across its width. They are formed either by sawing, exactly as a transverse contraction joint is, or as a keyed construction joint where adjacent lanes are paved in separate passes.

Tie bars are the steel used at longitudinal joints, and they are the opposite of dowels in both form and intent. They are deformed (ribbed) bars, typically 12 to 16 mm in diameter and 600 to 900 mm long at 600 to 900 mm centres, and they are fully bonded into the concrete on both sides. A tie bar is not a load-transfer device; it is there to hold the two lanes tightly together so that the joint faces stay in contact and transfer shear through aggregate interlock, and so that the lanes cannot separate under the outward push of a crowned cross-section on a yielding shoulder. Substituting a smooth dowel for a tie bar allows the lanes to drift apart; substituting a deformed tie bar for a dowel restrains the free thermal movement the transverse joint exists to permit. Confusing the two is the single most common error on this topic.

Expansion joints, better called isolation joints in modern practice, are full-depth gaps of 12 to 25 mm filled with a compressible board and sealed at the surface. They interrupt the slab completely, and any load transfer across them must be provided by dowels, one end of which is fitted with an expansion cap so the bar can travel. Because a contraction joint that has opened already provides room for thermal growth, routine expansion joints along a continuous pavement are no longer used — they were found to close up and let the adjacent contraction joints open excessively, spalling and losing their seals. Isolation joints are kept only where the pavement abuts something that will not move with it: bridge abutments, manholes and catch basins, buildings, and at sharp changes of alignment such as ramp gores.

Construction joints are those made where paving stops and later resumes — at the end of a day's work, at a plant breakdown, or between adjacent paving lanes. A transverse construction joint should be made to coincide with a planned contraction joint location wherever possible, and is formed as a butt joint with dowels through it so that it behaves like the contraction joint it replaces. If a stoppage falls between planned joints, the slab should be carried forward to the next joint location and the excess removed rather than a joint left in mid-panel. Longitudinal construction joints between paving lanes are usually keyed or tied, or both.

All of these joints share one further requirement: they must be sealed and the seal must be maintained. A joint reservoir is sawn to a width and depth that suit the sealant's allowable strain, backed by a compressible rod so the sealant bonds only to the two vertical faces, and filled with a hot-poured or silicone sealant. An unsealed joint admits water to the base and incompressible grit to the joint faces, and the second half of this question describes what follows.

Part (b) — pumping of joints. Pumping is the ejection of water and fine-grained material from beneath a slab through the joints, cracks and edges under the action of passing wheel loads. It requires three things at once: free water trapped under the slab, a base or subgrade containing enough fines to go into suspension, and repeated heavy axle loads. As a wheel approaches a joint the approach slab deflects and forces the trapped water forward at high velocity; as the wheel crosses, the leave slab deflects and the water is drawn back. The turbulent flow erodes the base, carries fines into suspension, and ejects a grey slurry at the joint, which dries as a fan-shaped stain on the shoulder. The consequence is not the loss of the material itself but the void it leaves: the slab corner is progressively undermined, loses its support, deflects further, and the cycle accelerates until the joint faults (a step forms between the two slabs), the corner cracks, and the ride deteriorates. Pumping is therefore controlled by attacking any of the three ingredients — by sealing joints and shoulders so surface water cannot enter, by draining the pavement structure with a permeable base and edge drains, by using a non-erodible stabilised (cement- or asphalt-treated) base rather than an untreated granular one, and by dowelling the joints so that corner deflections stay small. Widened outside lanes and tied concrete shoulders help by moving the wheel path away from the free edge.

Part (c) — the four forms of asphalt.

Cutback asphalts are asphalt cements whose viscosity has been reduced by dissolving them in a petroleum distillate, so that they can be handled at low temperatures and will coat a cold or damp aggregate. They are classified by the volatility of the solvent: rapid-curing (RC) cutbacks use a naphtha or gasoline-type distillate, medium-curing (MC) cutbacks use kerosene, and slow-curing (SC) cutbacks use a low-volatility oil. Each grade is further identified by a number indicating its kinematic viscosity, for example MC-70 or RC-250. Curing occurs when the solvent evaporates and the residual asphalt cement is left behind, so the cutback's performance ultimately depends on that residue, not on the fluid as supplied. Cutbacks were once standard for prime and tack coats, patching and cold mixes, but their use has fallen sharply because the solvent is both an air-quality problem, releasing volatile organic compounds, and an expensive fuel that is simply evaporated away. Emulsions have replaced them in most Canadian applications.

Asphalt emulsions achieve the same end — a fluid, low-temperature binder — without a solvent, by dispersing droplets of asphalt cement of the order of a few micrometres in water with the aid of an emulsifying agent. The emulsifier gives the droplets a like electrical charge so that they repel one another and stay dispersed. Anionic emulsions carry a negative charge and suit electropositive aggregates such as limestone; cationic emulsions carry a positive charge and bond well to electronegative siliceous aggregates, which is why cationic grades dominate in practice. Grades are designated by charge, setting rate and viscosity — RS, MS, SS and QS for rapid-, medium-, slow- and quick-setting, with a leading C for cationic, so CRS-2 or CSS-1h. The emulsion “breaks” when the charge is neutralised by contact with the aggregate, the droplets coalesce, and the water is displaced and evaporates, leaving the residual asphalt. They are used for tack and fog coats, chip seals and surface treatments, slurry seals and microsurfacing, and cold in-place recycling, and because the carrier is water they are far preferable environmentally to cutbacks.

Asphalt primers are low-viscosity binders applied to an untreated granular base before the first asphalt layer is laid. The primer's job is to penetrate the surface of the base for several millimetres, bind its fines, waterproof it against rain during construction, and give the overlying asphalt something to bond to. The distinction from a tack coat matters: a prime coat penetrates an absorbent unbound surface and is applied at perhaps 0.5 to 1.5 L/m², whereas a tack coat is a very light application, perhaps 0.15 to 0.4 L/m² of a diluted emulsion, applied between two asphalt layers purely to bond them. Traditional primers are medium-curing cutbacks such as MC-30 or MC-70; low-viscosity emulsion primers and specially formulated penetrating emulsions are now used to avoid the solvent.

Modified asphalts are asphalt cements whose rheology has been deliberately altered by an additive so that they perform outside the range an unmodified binder could cover. The commonest modifiers are elastomeric polymers, above all styrene-butadiene-styrene (SBS), which forms an elastic network through the binder; plastomers such as ethylene-vinyl acetate and polyethylene, which stiffen it; crumb rubber from scrap tires; and, at the cheaper end, oxidised or air-blown asphalts. The object is to widen the useful temperature interval: to raise the high-temperature stiffness so the mixture resists rutting in summer, while keeping or improving the low-temperature relaxation so it resists thermal cracking in a Canadian winter. In Superpave performance-graded terms this shows up as a wider spread between the two numbers of the grade, so that a PG 58-28 becomes, say, a PG 70-34. Modified binders also improve fatigue life, adhesion and resistance to moisture damage, and they are specified where traffic is heavy or slow-moving, at intersections and on bridge decks, and in stone-mastic and open-graded friction courses that depend on a thick, drainage-resistant binder film. The costs are a higher unit price, a narrower and higher mixing and compaction temperature window, and a greater sensitivity to construction quality.

Part (d) — advantages of adding tire rubber to an asphalt paving mixture. Crumb rubber recovered from scrap tires is used either by the wet process, in which it is blended into the hot binder and digests into it to form an asphalt-rubber binder, or by the dry process, in which coarser rubber particles replace part of the fine aggregate. The advantages claimed for it, and broadly borne out in service, are these. It increases the binder's high-temperature viscosity and elastic recovery, so the mixture resists rutting and shoving better in hot weather and under slow or channelised traffic. It lowers the stiffness and improves the relaxation of the binder at low temperature, which is directly useful in Canada because it delays thermal and reflection cracking. It permits thicker binder films — asphalt-rubber gap-graded and open-graded mixtures carry appreciably more binder without draining — and a thicker film means better resistance to age hardening, moisture damage and fatigue, so a longer service life or a thinner layer for the same life. Asphalt-rubber membranes and interlayers are among the more effective treatments for retarding reflection cracking through an overlay on a cracked pavement. Open-graded rubberised friction courses reduce tire-pavement noise by several decibels and reduce splash and spray. Finally there is the resource argument, which is not trivial: a lane-kilometre of asphalt-rubber surfacing consumes the rubber of the order of a thousand passenger-car tires that would otherwise be stockpiled or landfilled, and diverting them addresses a genuine waste management and fire-risk problem. Against these must be set a higher production temperature and more odour and fume, the need for a blending unit and continuous agitation of the wet-process binder, difficulties in recycling the resulting pavement, and a unit cost that only pays back where the improved cracking or noise performance is genuinely worth having.