Asphalt mill and overlay: cold milling, tack coat, and resurfacing

Mill and overlay, and what it actually fixes

Mill and overlay is resurfacing in two moves: grind off the worn top of the existing asphalt with a cold-milling machine, then pave a new lift over what is left. The milling takes off the oxidized, cracked, rutted surface and brings the grade back down. The overlay puts a fresh wearing course back on. Done right, the road or lot looks and rides new, and the new surface ties back into the curbs, gutters, and door thresholds it has to meet.

The reason you mill first instead of just paving over the top is grade. Every overlay raises the surface by the thickness of the lift. Pave over an existing road without milling and you bury the curb reveal, you push water the wrong way at the gutter, and the new surface stands proud of every driveway and threshold it meets. Milling off roughly what you put back keeps the finished grade where it started, so the ties-in work and the drainage still runs the way it was built to.

What this guide is about is the chain of small decisions between those two moves, because that is where mill and overlay is won or lost. The milling depth and the cleanup, the tack bond, the leveling course, the lift over the stone size, the joints, the structures that have to come up to the new grade, and the old cracks that want to come back through. Get any one of them wrong and the surface that looked new on opening day tells on you within a year or two.

When is mill and overlay the right fix?

Mill and overlay is the right fix when the pavement is failing on the surface but the structure underneath is still sound. Surface raveling, weathering, oxidation, block cracking, rutting confined to the top mix, and worn ride all live in the upper layer, and grinding that layer off and replacing it restores the surface without touching a base that is still carrying the load. That is the sweet spot, and it is the case the PCI survey is built to find.

It is the wrong fix when the failure is structural. Alligator cracking and rutting that runs down into the base or subgrade are load failures, and a mill and overlay laid over them flexes on the same bad base and fails again within a season or two. The cause check from the pavement condition assessment guide is the gate. A load failure needs full-depth patching or reconstruction first, not a fresh surface over a sound-looking skin. Treat a base failure with a surface fix and you have paved over the problem and paid to do it twice.

Three options sit on the table and the cause picks between them. An overlay-only, with no milling, fits where grade is not a constraint and the existing surface is sound enough to bond to, often an open lot with room to raise the surface. Mill and overlay fits where grade matters and the top layer is shot but the base is good. Reconstruction, full-depth removal and rebuild, is the honest call when the base and subgrade are gone. The pavement condition assessment and the PCI band tell you which one you are looking at, and the structural check under the surface confirms it before the milling machine shows up.

How does cold milling work?

Cold milling, also called cold planing or profiling, grinds off the existing asphalt with a rotating drum studded with carbide cutting teeth. The machine runs the drum down into the mat, the teeth fracture and lift the material, and a conveyor throws the millings into a truck running ahead of it. No heat, hence cold. The depth is set on the machine and held by a grade-and-slope control system, and modern machines with sonic or laser references hold the cut tight, commonly within a few millimeters of target, so the milled surface comes out to a planned grade and cross slope instead of just following the old, deformed one.

That last part is the point most people miss. Milling does not just take material off. It corrects the profile. Where an overlay or a leveling course fills the low spots, milling removes the high spots, so a machine running on a reference string or an averaging ski grinds out the ruts and the humps and leaves a surface truer than the one it cut. Set the cross slope on the controller and the milled surface drains, even if the old one had gone flat. That is why milling and a leveling course often run together on the same job: milling fixes the high side, the leveling course fills the low side, and the wearing course goes on a corrected base.

Depth depends on what you are after. A light cut just knocks off surface irregularities and texture. A deeper, uniform cut removes a set thickness, usually matched to the overlay going back on, so the finished grade comes out even. Deeper still gets you a cross-slope correction or a cut to the base for a full reconstruction. The milled surface itself comes out with a coarse, grooved texture from the teeth, and that texture is a feature, not a flaw: it gives the new lift mechanical grip and more surface for the tack to bond to. A drum with worn or missing teeth leaves a ragged, torn surface and ridges, which is why the teeth get checked and changed, not run to failure.

Why mill instead of just paving over the top?

Two reasons, and grade is the first. Every lift you lay raises the surface. On a curbed road or a lot, that raise has nowhere to go: it shrinks the curb reveal, it floods the gutter line, and it leaves the new surface standing above every driveway apron, door threshold, and adjacent lane it has to meet. Mill off roughly the thickness you intend to put back and the finished grade lands where it started, so the ties-in are flush and the water still runs to the inlets the way the site was graded to drain.

The second reason is that you are removing the failed material, not burying it. The top layer is where the oxidation, the raveling, the surface cracking, and most of the rutting live. Grind it off and you take the bad asphalt away. Pave over it and you have laid a fresh surface on a layer that is already cracked and debonding, which is a head start on reflective cracking and a delamination waiting to happen. Milling gives you a clean, sound, textured surface to build on instead of a tired one to hide.

There is a structural angle too. Milling and replacing keeps the total pavement thickness about where it was, which matters on bridge approaches and under overpasses where vertical clearance is fixed, and on a road where raising the crown would dump water into the shoulders or the adjacent properties. When clearance or drainage will not let you raise the surface, milling is not optional. It is the only way to resurface and keep the grade.

Milling cleanup and the swept surface

The milled surface is full of loose grindings and fine dust, and none of it can stay there. The cutting teeth leave the grooves packed with fractured fines, and a power broom and often a vacuum sweeper have to pull that material off before anything else happens. A milled surface that looks clean to the eye is usually still holding dust down in the grooves, so the sweeping is more thorough than it looks like it needs to be.

The reason is the tack. Tack coat bonds the new lift to the old surface, and it can only bond to the surface, not to a layer of dust sitting on top of it. Spray tack over a dirty milled surface and the emulsion glues itself to loose fines that shear away under the first truckload, and now you have a debonded lift that slips, shoves, and delaminates under traffic. The dust is a bond breaker. Sweep it off or you have spent the tack money for nothing.

On a job that mills today and paves tomorrow, the surface also catches whatever blows in overnight and whatever traffic tracks across it, so a re-sweep right ahead of the tack truck is normal practice, not a redo. Clean, dry, and dust-free is the condition the tack needs, and the foreman who walks the milled surface before the distributor truck rolls is the one whose overlay stays stuck down.

What is a tack coat, and why is it the most skipped step?

A tack coat is a light spray of asphalt emulsion applied between the old surface and the new lift to bond them into one structure instead of two layers sitting loose on each other. It is the single most skipped and most botched step in resurfacing, and it is the one that quietly decides whether the overlay lasts. A bonded overlay carries load as a unit. A debonded one slips under braking and turning, and you get shoving, slippage cracks, and the new lift peeling off in sheets at the intersections where the shear is highest.

The material is usually an asphalt emulsion: a slow-set or rapid-set grade like SS-1h, CSS-1h, or a proprietary trackless tack, depending on the agency and the schedule. The number that matters is the residual rate, the asphalt left on the surface after the water in the emulsion breaks off and evaporates, measured in gallons per square yard. A milled surface needs more tack than a smooth one because the grooves multiply the surface area, so a residual rate commonly in the range of about 0.05 to 0.08 gal/sy on a milled surface is typical, against something closer to 0.03 to 0.05 gal/sy on a smooth existing pavement. If the emulsion is diluted, the applied gallons run higher than the residual to put the same asphalt down. Those are common ranges, not a spec. The agency or project specification sets the rate, and the distributor truck is calibrated to hit it, not eyeballed.

Coverage and curing are where it goes wrong on the ground. The tack has to be uniform and complete, no streaks, no skips, no bare strips between spray bar nozzles, because every gap is an unbonded patch the overlay will find. And the emulsion has to break and cure before the mat goes over it: the brown emulsion has to turn black, the water has to leave, and the surface has to get tacky rather than wet. Break times run roughly 5 to 15 minutes for a rapid-set, longer for a slow-set, and a trackless tack is built to set up fast, often within 10 to 30 minutes, so it does not track off on the haul truck tires. Pave over tack that has not broken and you trap water at the interface and the bond fails. Pave over tack that trucks have tracked away and you have no bond where the tires lifted it. Tack thin, tack complete, let it break, then pave.

Tack coat

A light asphalt-emulsion spray that bonds the new overlay to the existing surface so the layers act as one

Residual rate

The asphalt left on the surface after the emulsion breaks, in gal/sy; the number the spec controls

Break / cure

The emulsion turning from brown to black as water leaves; paving before it breaks traps water and kills the bond

Trackless tack

A proprietary tack that sets fast and does not pick up on truck tires, so coverage survives to the paver

The leveling or scratch course

A leveling course, also called a scratch course or a wedge course, is a thin lift laid before the wearing course to fill ruts, depressions, and elevation differences so the surface lift goes down on a true, smooth base. Where milling takes off the high spots, the leveling course fills the low spots, and on a badly deformed pavement you use both: mill the humps and ruts down, then scratch in the remaining lows before the final surface.

The leveling course is laid by the paver or, on variable ruts, with a spreader box, and its job is to bury the profile defects, not to be the finished surface. Its thickness varies across the mat by design, feathering to nothing where the existing surface is already true and building up where it dips. Because it varies, you cannot hold a fixed compacted thickness on it, and you cannot expect it to reach the same density as a uniform lift in the feathered-out zones, which is exactly why it goes under the wearing course and not on top.

Skip the leveling course where the surface needs it and the rut telegraphs straight up through the overlay. A uniform-thickness wearing course laid over a rutted surface just paves a thinner mat in the ruts and a thicker one on the high side, the compaction comes out uneven, and the rut shadow is back in a season because the deformation underneath was never corrected. The leveling course is how you reset the profile before you commit the surface.

The overlay lift thickness and the stone size

The overlay lift has a floor set by the stone in the mix, not just by the structure you want. The compacted lift has to be thick enough that the largest aggregate can shift and seat without bridging on itself, expressed as the ratio of lift thickness to nominal maximum aggregate size, t/NMAS. The common rule is a minimum lift of about 3 times the NMAS for fine-graded mixes and about 4 times for coarse-graded mixes, so a 1/2 in NMAS surface mix wants a compacted lift on the order of 1.5 to 2 in, not 3/4 in. Confirm the ratio against the spec, because agencies vary and some push it higher.

Go thinner than the ratio allows and the lift will not reach density, it turns permeable, and it ravels early, which is the same trap covered in the compaction window guide for any thin lift. It also cools too fast to compact, because a thin mat sheds heat in a hurry and the roller runs out of working time before it reaches target. So the surface lift is sized to the stone and to the cooling, and the leveling course underneath carries the variable thickness that the surface lift cannot.

On the structural side, a thicker overlay or a second lift adds capacity, and that decision belongs to the thickness design, not to habit. Match the mix to the lift and the lift to the mix: a coarse mix wants a thicker lift to seat, a thin surface lift wants a finer mix. Spec the surface course, the lift, and the NMAS together so the mat you spread can actually reach density at the thickness you spread it.

NMAS

Nominal maximum aggregate size, a measure of the largest stone in the mix, read from the gradation

t/NMAS

Lift-thickness-to-NMAS ratio; a minimum near 3 (fine) to 4 (coarse) sets the thinnest lift that will compact

Laying and compacting the overlay

Once the surface is milled, swept, tacked, and the leveling course is in, the overlay is paved and rolled like any other mat, and the temperature and rolling rules from the compaction window guide carry straight over. The mat comes off the screed hot, the breakdown roller stays tight behind the paver in the hot zone, and the density gets built while the binder is still soft enough to let the stones move. Lose the window on an overlay the same way you lose it on a new mat: a thin lift on a cool, milled surface cools fast, so the rollers cannot wander.

The milled surface adds one wrinkle. It pulls heat out of the bottom of the mat faster than a smooth, sound pavement would, because the grooved face has more area and the milled layer underneath may run cooler. On a thin surface lift over a cold milled base on a cool day, the working time is short, so run a cooling estimate and keep the roller train tight. The tack underneath has to be broken before the mat goes over it, and a hot mat helps the bond set, but it does not rescue tack that was paved over wet.

Everything the compaction window guide says about the roller train, the test strip, the tender zone, and density acceptance applies to the overlay mat. The difference with resurfacing is everything underneath it: the milled profile, the bond, and the leveling, all of which were set before the paver ever rolled.

Longitudinal and transverse joints

Joints are where an overlay starts to come apart, and resurfacing has two kinds. The longitudinal joint is the seam between adjacent paving passes, running with traffic. The transverse joint is the seam across the mat where a day's paving stops and the next starts, or where the overlay meets existing pavement at the project limits. Both are low-density, water-entry points by nature, and both ravel and crack from the joint out before the field of the mat shows any age.

The longitudinal joint fails from the unconfined edge, the same mechanism as on a new mat: the first pass spreads sideways under the roller with nothing beside it to push against, so that edge runs lean. The fixes are the same too, edge restraint on the breakdown roller, rolling the edge with the drum hanging over, and not leaving the joint cold. On a resurfacing job you also have the chance to plan the joint location so it does not land in a wheelpath, where the worst loading hits the weakest seam. Echelon paving, two pavers running staggered so the longitudinal joint is made hot against hot, builds a far denser seam than a hot lane laid against a cold, day-old one, and on wide mats it is the difference between a joint that lasts and one that opens in two winters.

The transverse joint at the project limits is where the new overlay meets the old road, and it has to be a clean, compacted butt joint, not a feathered ramp of thin mix. The transverse joints between paving runs get cut back to a vertical face and tacked before the next mat butts against them, and a joint heater can be run on the cold transverse joint to soften it so the new mat bonds rather than sitting cold against a stiff edge. A bad transverse joint is the bump every driver feels at the same spot, and it is usually a joint that was feathered instead of cut and butted.

Why do you raise the manholes and valves?

You raise the manholes, valve boxes, and inlets because the overlay raises the surface around them, and a casting that does not come up with it ends up buried under the mat or sitting in a dish below the new grade. Both are failures. A buried casting is lost until someone digs for it, and a low casting is a hole in the road: a jarring bump, a wheel trap, a place water ponds and freezes, and a spot the mat ravels around from the impact loading. The fix is to bring every structure up to the new finished grade so the surface runs flush across it.

The sequence matters. Castings are commonly milled around or lowered before paving so the milling machine and the paver are not fighting them, then adjusted up to the new grade after the overlay, often with adjustable riser rings that set the casting to the finished surface without excavation, or by resetting the frame on grade rings and collaring it. On a road job the structures get located and referenced before milling so nobody loses one under the millings, because a manhole found by a milling drum is a bad day.

The bump from a missed structure is the complaint that outlives the whole project. Drivers feel a low casting at the same spot every pass, and it becomes the thing the owner remembers about an otherwise good overlay. Locate the structures, protect them through milling and paving, and adjust every one of them to grade before the job is called done. A finished overlay with a sunken manhole in the wheelpath is not finished.

Transitions, tie-ins, and the wedge at the gutter

At the edges of the work the new overlay has to tie back into pavement, curb, and structures that are staying, and those transitions are where the geometry gets fussy. At the project limits, where the resurfacing stops, you mill a notch into the existing pavement and butt the new mat into it, a header or butt joint, so the surface steps cleanly from old to new instead of ramping up over a feathered tail that breaks off under traffic. The milled header gives the transverse joint a vertical face to compact against and keeps the grade matched across the limit line.

Along the curb and gutter, the overlay has to meet the existing concrete without burying the gutter or losing the drainage. If the resurfacing is not milled tight to the curb, the surface near the gutter gets a tapered wedge so the new mat feathers down to meet the existing gutter line and the water still runs to the inlets. Bury the gutter under a full-thickness overlay and you have created a dam that ponds water against the curb and floods the lane in the first rain. The grade at the gutter is checked, not assumed.

Driveways, aprons, ramps, and door thresholds are the other tie-ins, and they are why milling earns its keep. Hold the finished grade at the original elevation through milling and these transitions stay flush, the accessible route keeps its slope, and nothing trips or scrapes. Raise the whole surface without milling and every one of these meeting points becomes a lip you have to chase with a transition that should not have been necessary.

What causes cracks to come back through an overlay?

Cracks come back through an overlay because the crack in the old pavement underneath is still moving, and that movement concentrates stress in the new mat directly above it until the new mat tears along the same line. This is reflective cracking, the old crack pattern reappearing on the fresh surface, and it is the classic disappointment of resurfacing: a smooth new overlay with the old joints and cracks tracing back through it within a few seasons. The driver is the underlying crack opening and closing with traffic load and with daily and seasonal temperature swings, which the overlay rides like a tendon over a knuckle until it splits.

Milling helps because it removes the most cracked, oxidized material and reduces how much movement the overlay has to bridge, but milling alone does not stop a working crack that runs deeper than the cut. Where reflective cracking is the known risk, the trade puts an interlayer between the old surface and the new overlay to absorb or interrupt that stress. The common ones are a stress-absorbing membrane interlayer, a SAMI, a heavy asphalt-and-aggregate or polymer-modified layer that flexes with the crack instead of transmitting the stress; a paving fabric or geotextile saturated with binder; and a stiff geogrid or geocomposite interlayer that spreads the strain. Each one buys time against reflection rather than guaranteeing the crack never returns, and agencies assign them rough equivalences, treating an interlayer as worth some additional thickness of overlay for crack relief.

None of these saves an overlay laid over a structural failure. A working thermal or load crack over a sound base is the case an interlayer addresses. Alligator cracking over a dead base is a different problem, and no interlayer or membrane stops a base that is moving under the wheel. The honest order is to fix what is moving first, full-depth patch the load failures, then choose milling depth and an interlayer for the working cracks that remain. Lay a fabric over a failing base and you have buried a problem you will pay to dig back up.

Pavement markings: removal and restriping

Pavement markings come off before milling and go back on after the overlay, and the timing is part of the traffic plan, not an afterthought. On a mill and overlay the milling itself takes the old striping off with the surface, so removal is usually automatic, but where a section is overlaid without milling, the old markings get ground or blasted off first so they do not bleed through or interfere with the bond. Either way, the old lines are gone before the new surface is down.

After paving, the surface gets restriped, and a fresh mat needs a temporary plan in the meantime. A newly overlaid road or lot with no lines is a hazard, so crews lay temporary tabs or short-term markings to carry traffic until the permanent striping goes down, which often waits a short period for the mat to cure and for the layout to be set. The permanent layout is the chance to fix what was wrong before: faded geometry, non-compliant accessible stalls and aisles, and stall counts, all of which ride on the same surface the resurfacing just renewed.

On a commercial lot the restripe is what the owner actually sees, so it is worth getting the layout, the accessible route, and the stall geometry right while the surface is blank. The detailed striping and ADA layout work is its own scope, but the overlay schedule has to leave room for it and the traffic plan has to cover the gap between paving and permanent lines.

What smoothness does the finished overlay have to meet?

The finished overlay has to meet a ride or smoothness spec, and on most agency work that is an International Roughness Index, IRI, target measured by an inertial profiler over the finished surface, plus a straightedge tolerance for localized bumps and dips. Resurfacing is partly bought to improve ride, so the smoothness is often an acceptance item with a pay adjustment: hit the IRI target and you get paid, miss it and you get docked or have to grind and correct the rough spots. The exact IRI number, the segment length, and the pay schedule come from the agency spec.

The straightedge is the field check anyone can run. A common requirement is no deviation greater than about 1/4 in under a 10 ft straightedge, which catches the localized defects a profiler averages over: the bump at a transverse joint, the dip at a structure, the hump from a paver that surged. Where the straightedge finds a high spot, it gets diamond-ground out; a low spot usually has to be addressed in the next lift or lived with against the tolerance.

Smoothness on an overlay is mostly built underneath the surface, which is the part people miss. A surface lift laid over an uncorrected rutted profile telegraphs the ruts and reads rough no matter how well the paver runs, so the milling and the leveling course are where the ride is actually won. The automatic grade control on the paver, running off a string line, an averaging ski, or a sonic reference, holds the new surface to a true profile instead of copying the old one, and that is what turns a corrected base into a smooth finished ride.

Thickness design and the milling-to-overlay balance

Thickness design sets how deep you mill and how thick you overlay, and for resurfacing it is mostly a balance, not a from-scratch structural calculation. On a true mill and overlay where the goal is to hold grade, you mill off about what you pave back on, so the finished elevation lands where it started. Mill 2 in, lay 2 in, and the grade is preserved while the surface is renewed. When the design needs to add structural capacity, the overlay goes thicker than the mill, and the surface rises by the difference, which then has to be reconciled with curbs, clearances, and ties-in.

The structural side uses the language of pavement design: the existing pavement carries some remaining capacity, and the overlay adds to it, expressed in the structural-number framework of the AASHTO pavement-design method or a mechanistic-empirical equivalent, where each layer contributes according to its thickness and a layer coefficient. Milling complicates that arithmetic because grinding off material removes some existing capacity along with the bad surface, so a deep mill that takes sound material has to be paid back in overlay thickness if the structure is to hold. That balance is an engineering call on a designed road, and it belongs to the project documents, not to a field rule of thumb.

On a parking lot or a light-duty resurfacing the math is usually simpler, driven by grade and surface condition more than by traffic loading, and a typical mill-and-inlay holds the existing thickness while renewing the surface. The line to hold is that the design, the project specification, and the agency govern the milling depth, the overlay thickness, and how the two balance. Confirm them before the milling machine sets its drum, because the depth you cut is hard to put back.

Quality control: cores, density, thickness, and the tack

Quality control on a mill and overlay checks four things the surface will not tell you by eye: the bond, the density, the thickness, and the milled grade. Density is the headline number, measured as a percent of Gmm the same way as any mat, tracked with a correlated gauge during rolling and confirmed with cores for acceptance, with the joint carrying its own lower target. The compaction window guide covers the density methods and the gauge-versus-core reconciliation in full; on an overlay the same acceptance applies to the new lift.

Cores do double duty on resurfacing. The same plug that gives you the in-place density also shows the lift thickness and, just as important, whether the new lift is bonded to the milled surface underneath. A core that separates cleanly at the interface is a debonded overlay, and it tells you the tack failed, whether from a dirty milled surface, a missed coverage, or paving over emulsion that had not broken. That is the QC check that catches the most expensive mistake before traffic does, because a debonded lift looks perfect until it shovels up at the first hard-braking intersection.

The milling itself gets checked before any of that: the cut depth and the cross slope against the plan, the surface texture and uniformity, and the cleanup, because the tack and the bond ride on a clean, true milled surface. Tack coverage and rate get verified at application, by the distributor calibration and a visual walk for skips and streaks, since once the mat is over it the only way to check the bond is to cut a core and try to pull it apart. Check the milled surface, check the tack, check the density, and core the result. The surface that passes all four is the one that lasts.

The commercial site, night work, and traffic

Resurfacing happens under traffic far more often than new construction does, because you are rebuilding a surface people are still using. On roads that means lane closures, staged work, and frequently night paving, since many agencies will not allow daytime closures on busy routes. Night work brings the cooling problem from the compaction window guide to the front: falling temperatures in the early-morning hours shrink the compaction window, so the rolling has to be planned around a clock that is working against you, and the mill-pave-open cycle has to fit inside the closure.

The phasing is the planning. You mill a section, sweep it, tack it, pave it, and reopen it to traffic, often within the same shift, which means the milled surface may carry traffic before it is overlaid. A milled surface open to traffic is rideable but coarse and loud, and it sheds loose stone, so the gap between milling and paving gets kept short and the millings get swept again before the tack. On a commercial lot the same logic applies in smaller pieces: you keep half the lot open while you resurface the other half, you stage around the anchor tenants and the access drives, and you sequence so the business keeps running.

The handoff to traffic is its own decision. A fresh overlay needs to cool before it carries traffic or it shoves and marks under tires and turning loads, and the temporary striping and the traffic control have to carry the public through until the surface is cool, marked, and the structures are adjusted to grade. Open it too early and you damage the mat you just paid to lay.

What to document

The record is what defends a resurfacing job when a core comes back light or a section debonds months later, and on a mill and overlay there is more to capture than on a single mat because the result depends on what happened underneath. The crew that writes down the milling, the tack, the leveling, the lift, and the density by station is the crew that gets paid for marginal lots instead of arguing from memory.

Capture it by station or area: the milling depth and whether the cross slope was corrected, the cleanup, the tack type and the residual rate actually applied, the leveling course where it was used, the overlay mix and compacted lift, the density results and method, and the structures adjusted to grade. Note the joints and any interlayer used for reflective-crack relief, and flag any section where the cause check pushed the work from overlay to full-depth patch. A note made at the time about a short milling section, a tack that ran thin, or a structure left for a follow-up crew is worth more than a memory at the dispute meeting. Holding the milling depths, the tack rate, the lift, the density, and the structure adjustments together in a record like FieldOS keeps the as-built tied to the lot it covers.

Common mistakes

• Skipping or under-applying the tack coat, or paving over emulsion that has not broken, so the overlay debonds and shoves at the intersections.
• Leaving milling debris and dust in the grooves, so the tack bonds to loose fines that shear away under the first truckload.
• Not raising the manholes, valves, and inlets to the new grade, leaving a sunken casting and a bump the owner feels every pass.
• Spreading the surface lift thinner than the t/NMAS minimum, so it never reaches density, turns permeable, and ravels early.
• Overlaying a section with alligator cracking or a wet, failing base instead of patching full-depth first, so it reflects and fails in a season.
• Ignoring reflective cracking on working cracks, skipping the interlayer where the spec or the condition called for one.
• Feathering the transverse joint at the limits into a thin ramp instead of milling a header and butting against a vertical face.
• Burying the gutter line under a full-thickness overlay with no wedge, damming water against the curb.
• Holding a uniform surface lift over an uncorrected rutted profile instead of milling the highs and leveling the lows first.

Field checklist

Standards and references

What governs a mill and overlay is the project specification and the state DOT or owner agency, layered on top of the trade standards. The milling depth and tolerance, the tack type and residual rate, the leveling course, the overlay mix and lift, the joint requirements, the structure-adjustment items, and the smoothness acceptance all live in the agency spec, and the numbers in this guide are the commonly cited ranges, not the contract values. Confirm the rate, the lift, and the tolerance against the project documents and the adopted specification before you build to any single figure.

The Asphalt Institute is the trade source for the practical side. MS-22, Construction of Quality Asphalt Pavements, covers placement, tack, and compaction, and the Institute's overlay and rehabilitation references in the MS series cover the mix and overlay design. For mix design and the volumetrics behind the density target, the Superpave system and the gyratory procedure under AASHTO T312 set the framework.

The test methods behind acceptance come from AASHTO and ASTM. Theoretical maximum specific gravity, Gmm or the Rice value, is AASHTO T209 and ASTM D2041; bulk specific gravity of cores is AASHTO T166, with T331 vacuum sealing for mixes that drain; in-place nuclear density follows ASTM D2950, and non-nuclear gauges still require correlation to cores. Smoothness is reported as IRI from an inertial profiler under the standardized profiling and computation methods, with a straightedge tolerance commonly near 1/4 in under a 10 ft straightedge for localized defects. Structural overlay thickness draws on the AASHTO pavement-design method and the structural-number framework or a mechanistic-empirical equivalent. The exact designations and editions shift on a cycle, so verify them against the adopted versions before citing a clause on a submittal.

Units and terms

Resurfacing gets described across a few unit systems and a stack of terms that mean specific things, so the same job reads differently across a plan set, a mix sheet, and a spec. Milling and lift thickness are in inches, with aggregate size in inches or millimeters, so a 1/2 in NMAS surface mix is 12.5 mm. Tack is a residual rate in gallons per square yard, gal/sy, which in metric is liters per square meter. Density is a percent of Gmm, and smoothness is an IRI value, commonly in inches per mile or meters per kilometer, with a straightedge tolerance in inches.

The terms carry the meaning. Cold milling is grinding the surface off with a toothed drum and no heat. A leveling or scratch or wedge course is the thin lift that fills lows before the surface lift. The tack coat is the bond between old and new, set by its residual rate after the emulsion breaks. Reflective cracking is the old crack coming back through the new mat, and an interlayer or SAMI is the stress-relieving layer that slows it. A mill-and-inlay holds the existing grade by milling off about what you pave back, and the t/NMAS ratio is the lift-to-stone check that sets the thinnest lift that will compact.

Mill and overlay

Resurfacing by milling off the worn top layer with a cold-milling machine, then paving a new lift over it

Cold milling / cold planing

Grinding off the existing asphalt with a rotating toothed drum, no heat, to a set depth and slope

Leveling / scratch / wedge course

A thin, variable-thickness lift that fills ruts and lows before the wearing course

Tack coat / residual rate

The emulsion bond between old and new, measured by the asphalt left after it breaks, in gal/sy

Reflective cracking / interlayer

The old crack reappearing through the overlay, and the stress-relief layer (SAMI, fabric, geogrid) that slows it

t/NMAS

Lift-thickness-to-nominal-maximum-aggregate-size ratio that sets the thinnest lift that will compact

IRI

International Roughness Index, the profiler-based smoothness number used for ride acceptance

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What temperature do you stop rolling asphalt?
• What is the pavement condition index (PCI)?
• What is asphalt milling?
• What is the difference between tack coat and prime coat?
• What is the minimum temperature to pave asphalt?