Asphalt pavement design: thickness by traffic and subgrade

What pavement design actually decides

Asphalt pavement design decides one thing: how thick to build each layer so the section carries the traffic over the subgrade for the design life without failing. That is it. The output is a stack of thicknesses, so many inches of asphalt surface, so many of asphalt or aggregate base, so many of subbase, sitting on a subgrade of a known strength. Everything else in the process is figuring out what those numbers need to be.

Get them too small and you under-design. The pavement fatigues and ruts early, cracks in the wheel paths within a few seasons, and the owner pays for a rebuild a decade before they should have. Get them too large and you over-design. The section carries the traffic fine, but the owner paid for inches of asphalt and stone that the traffic never needed, and on a big lot that is real money buried in the ground. The whole skill is landing between those two, sized to the actual load and the actual soil, not to habit.

Two things sit on either side of this guide. The base and subgrade construction, how you compact the soil and the aggregate and prove the density before you pave, is its own job covered in the base and subgrade compaction guide. The asphalt mat itself, how you roll it to density inside the cooling window, is the asphalt compaction guide. This guide is the part in between: deciding how thick the section has to be in the first place.

The inputs a design needs

A pavement design needs five inputs, and a thickness produced without all five is a guess wearing a number. The first is traffic, expressed not as a vehicle count but as load repetitions, because the load is what does the damage. The second is the subgrade strength, the stiffness of the soil the whole section rests on. The third is the materials, the layer strengths you have to work with. The fourth is climate, the temperature and moisture the section will live in. The fifth is reliability, how sure you need to be that it lasts the full design period.

The two that move the answer most are traffic and subgrade. A heavy-truck route over a soft clay needs a thick, strong section. A car-only lot over good granular soil needs very little. Hold the materials and climate roughly constant and the thickness swings mostly on those first two inputs, which is why the rest of this guide spends its time there.

The mistake that recurs is skipping inputs and copying a thickness instead. Somebody pulls 'six inches of asphalt on twelve of base' off the last job because it worked, drops it on a site with different soil and different trucks, and the number is either too thin for the new traffic or richer than the new soil needs. The inputs are the design. The thickness is just the answer they produce.

What is an ESAL?

An ESAL is an equivalent single axle load, the unit pavement design uses to express traffic as damage instead of as a vehicle count. One ESAL is the damage done by one pass of a standard 18,000 lb single axle on dual tires, commonly inflated near 110 psi. Every axle that rolls over the pavement gets converted into some number of these standard passes, and the design is sized to the total ESALs expected over the design life, not to the number of vehicles.

The reason you count damage and not cars is the load equivalency, and it is steeper than people expect. The damage an axle does rises roughly with the fourth power of its load, so doubling the axle weight does not double the damage, it multiplies it by about sixteen. A loaded tractor-trailer axle can do thousands of times the damage of a passenger car axle. Run the numbers and the cars essentially disappear from the design. A parking lot full of cars with one trash truck a week is a trash-truck pavement, not a car pavement.

This is the single most useful thing to carry out of pavement design and back to the field. When you look at a site, you are not counting traffic, you are counting trucks. The bus lane, the dumpster pad, the loading dock approach, the drive-through where the delivery trucks queue, those are the ESALs. The car stalls out in the field of the lot barely register. Size the truck paths for the trucks and you have done most of the design right; treat the whole lot as if the cars mattered and you have either wasted money or missed the trucks.

ESAL

Equivalent single axle load: damage equal to one pass of a standard 18,000 lb single axle on dual tires

LEF

Load equivalency factor: how many standard axles' worth of damage one axle does, rising with about the fourth power of load

Design ESALs

Total equivalent single axle loads expected over the design life, the traffic number the section is sized to

How does the subgrade affect pavement thickness?

The subgrade sets how thick everything above it has to be, because it is the foundation the whole section rests on and its strength decides how much help it needs to carry the load down to it. A stiff subgrade spreads the wheel load over a wide footprint at the bottom of the section, so the layers above can be thinner. A soft subgrade spreads it poorly and deflects under each axle, so the section has to be thicker and stronger to keep the soil from seeing more pressure than it can take. Weak soil is the most common reason a design comes out thick.

Subgrade strength gets quantified three ways in United States practice, and a design names the one it used. The resilient modulus, Mr, describes how stiffly the soil springs back under repeated wheel loading, and the AASHTO design framework is built on it. The California Bearing Ratio, CBR, rates the soil against a crushed-stone standard, where a soft clay can read below 2 to 5 and good granular soil reads much higher. The R-value is a related stability measure some agencies use. They correlate loosely. A common rule of thumb estimates the modulus as roughly 1,500 times the CBR for fine-grained soils, with newer power-law correlations such as the one in the AASHTO mechanistic guide giving a closer fit. The lab test for the modulus is the cyclic triaxial under AASHTO T 307; confirm which measure your geotech reported.

The number that matters is whatever the geotech reported for your site, because that value sized the section above it. If the field subgrade comes in softer than the report assumed, the design is no longer valid and the section is under-built before the first truck. That is a stop-and-call moment. How you build and prove that subgrade to the strength the design counted on is the base and subgrade compaction guide; this guide is what you do with the strength once you know it.

The AASHTO 93 method and the structural number

AASHTO 93 is the empirical design method most of the United States still works from, built on the results of the AASHTO Road Test of the late 1950s. It is empirical, meaning the equations come from observing how real test sections performed under measured traffic, not from modeling the physics of the layers. It takes the inputs, the design ESALs, the subgrade modulus, reliability, and the loss of serviceability you will accept, and it produces a single required strength number for the section: the structural number, SN.

The structural number is an abstract measure of the total strength the section needs. It is not a thickness. It rolls the whole stack into one figure that says, in effect, this much structural capacity has to sit over this subgrade to carry this traffic. The design then distributes that required SN across the actual layers using each layer's strength and thickness. The relationship is the heart of the method: the section's SN is the sum of each layer's thickness times its layer coefficient, with a drainage coefficient applied to the unbound base and subbase layers.

The numbers in the equation come from the AASHTO design guide and the agency that adopts it, and they shift between editions and between states. Initial serviceability for flexible pavement is commonly taken near 4.2 and the terminal value near 2.5 for major roads, with the difference being the serviceability loss the design allows. Reliability runs from around 50 percent for low-volume local roads up to 99.9 percent for major routes. Do not carry these constants between jobs. The adopted agency method sets them, and the design is only as good as the inputs you feed this equation.

SN

Structural number: the total structural strength AASHTO 93 says the section needs over the subgrade

D1, D2, D3

Thickness in inches of the surface, base, and subbase layers

a1, a2, a3

Layer coefficients: the relative strength per inch of the surface, base, and subbase materials

m2, m3

Drainage coefficients applied to the unbound base and subbase, adjusting for how well they drain

Layer coefficients: turning the structural number into thickness

Layer coefficients are how the design converts a required structural number into actual inches, and they measure the relative strength of each material per inch of thickness. Asphalt concrete carries the highest coefficient because it is the strongest layer, commonly taken around 0.40 to 0.44 per inch in many state methods. Crushed aggregate base is much lower, often near 0.11 to 0.17. Granular subbase is lower still, commonly around 0.08 to 0.11. The exact values are an agency call tied to the local materials, so pull them from the method you are designing under, not from memory.

Read the coefficients and the trade jumps out. An inch of asphalt is worth roughly two to four inches of aggregate base in structural terms. So you can hit a required SN with a thinner asphalt layer over a thick base, or a thicker asphalt layer over a thin base, and the two can satisfy the same number. Which one you build comes down to cost. Asphalt is the most expensive layer per inch, aggregate is cheap, so the economical section usually carries as much of the structure as practical in cheaper base and only as much asphalt as the surface and the fatigue demand.

The drainage coefficient is the other multiplier, and it only touches the unbound base and subbase. A base that drains well earns a coefficient near or above 1.0, which lets it count for full strength. A base that holds water gets penalized below 1.0, which means you need more of it to reach the same SN. That penalty is the design admitting what every paver knows: a wet base is a weak base. Build the drainage into the section and the math rewards you for it.

What is the MEPDG / Pavement ME?

The MEPDG, the Mechanistic-Empirical Pavement Design Guide, implemented in the AASHTOWare Pavement ME software, is the modern method that is steadily replacing the empirical AASHTO 93. The split is in the word mechanistic. Where AASHTO 93 reads a thickness off equations fit to a 1950s road test, Pavement ME models the mechanics: it calculates the stresses, strains, and deflections the layers actually see under each axle, then uses empirical models to turn that response into predicted distress, the rutting, the fatigue cracking, the roughness, accumulated over the design life.

That changes what you give it and what it gives back. The method came out of the NCHRP 1-37A research and it wants far more input, on the order of a hundred or more values covering traffic spectra, climate from a weather database, and detailed material properties, instead of a single ESAL total and one subgrade number. In return it does not hand you one thickness. It predicts how much cracking and rutting a trial section will have at a chosen reliability over the design period, and you adjust the section until the predicted distress stays under the limits. It is a check-and-iterate design, not a plug-in formula.

For a working estimator the practical line is this: many agencies have moved to Pavement ME for their highway design, while a great deal of site, parking, and local work still runs on AASHTO 93 or an agency catalog derived from it. Both are legitimate, and which one applies is set by the owner and the jurisdiction. Pavement ME is more accurate when its many inputs are real and well calibrated to local conditions, and it is only as good as that calibration. Use the method the project and the agency actually require, and do not mix the two.

Reliability and the design life

Reliability and design life are the two policy inputs that decide how much margin gets built into the section. The design life, often called the design or analysis period, is how long the pavement is meant to carry traffic before a major rehabilitation, commonly 20 years for new flexible pavement, though heavier and longer-life designs run further. Every ESAL you expect over that period gets totaled into the traffic the section is sized for, so a longer life and a higher growth rate both push the thickness up.

Reliability is the probability the section actually survives that period as designed. A higher reliability builds in more margin against the things the design cannot know precisely, the real truck loads, the as-built materials, the weather. A major interstate might be designed at 95 to 99.9 percent reliability, a local subdivision street at 50 to 75 percent. The cost of being wrong drives the choice: a freeway that fails early is a far bigger problem than a quiet street, so it is designed with more cushion.

Serviceability ties it together. The design is built around an acceptable loss of ride quality, from a good initial condition down to a terminal serviceability where the pavement is considered to need rehabilitation. That allowed drop is the serviceability loss the AASHTO method uses, and it interacts with both the life and the reliability. None of these are field judgment calls. They are policy numbers the agency sets by road class, so read them off the adopted method rather than picking them yourself.

Flexible asphalt vs rigid concrete design

Flexible and rigid pavements carry the load by opposite mechanisms, and that difference runs through the whole design. Flexible pavement, asphalt over base over subgrade, bends under each axle and springs back, spreading the wheel load down through the layers so the pressure at the subgrade is a fraction of the pressure at the tire. The strength is in the stack, and a weak subgrade forces a thicker stack. This guide is about flexible design.

Rigid pavement, a portland cement concrete slab, is stiff. It does not spread load through layers so much as bridge over the subgrade like a beam, carrying the wheel load in bending across the slab and distributing it over a wide area regardless of what is directly under each wheel. Because the slab itself is the structure, the subgrade matters less to rigid design than it does to flexible, though uniform support and good drainage still matter. Rigid design works in a different currency, slab thickness, joint spacing, load transfer at the joints, and the concrete's flexural strength, and it is its own topic covered separately by trade.

The choice between them is a life-cycle decision, not just a first-cost one. Asphalt is usually cheaper to build and faster to open, and it renews by milling and overlaying the surface. Concrete costs more up front and lasts longer with less surface work, which can win over a long life on heavy-truck routes and at slow-moving, high-load areas like dock aprons and intersections where asphalt shoves and ruts. Many heavy industrial sites end up with concrete at the dock and asphalt on the drives for exactly that reason.

The layers and what each one does

A flexible pavement is a stack of layers, each one doing a specific job, and a good design knows which layer is carrying which duty. From the top down, the names are reasonably consistent, though specs use them loosely, so read the section the project actually calls for.

The surface, or wearing course, is the layer the tires ride on. It sheds water, provides skid resistance, and takes the direct abrasion and the weather, so it is built from a finer, denser, higher-quality mix than the layers below. The binder, or intermediate, course sits under the surface on thicker sections and carries much of the structural asphalt with a coarser, cheaper mix. Below the asphalt is the base, which can be more asphalt (a full-depth or deep-strength section) or unbound crushed aggregate, and it does most of the load spreading. The subbase, when present, is a lower-cost granular layer that bridges a weak subgrade and adds drainage. At the bottom is the subgrade, the natural soil, which carries everything.

Where the structure goes is a design lever. You can carry strength cheaply in a thick aggregate base, or you can carry it in asphalt with a deep-strength or full-depth section that uses asphalt base instead of stone. Full-depth asphalt resists water better and performs well on weak subgrades where a stone base would be hard to keep dry, but it costs more per inch. Which way the section leans is the layer-coefficient and cost trade run through the structural number.

Drainage in the design

Drainage belongs in the design, not just in the construction, because water is the fastest way to turn a sound section soft. Saturate a base or a subgrade past its optimum moisture and it loses the strength it had drained, pore pressure builds under each wheel, and the layer that carried the load dry starts to deflect and pump. A section designed for a strong, dry base that ends up wet is being asked to carry traffic on a layer weaker than the one the math assumed.

AASHTO 93 puts this into the structural number directly through the drainage coefficient on the unbound layers. A base that drains quickly and stays dry most of the time earns a coefficient near or above 1.0 and counts for full strength. A base that drains slowly and stays wet gets penalized below 1.0, so you need more of it. The design is paying you to keep water out of the base, and the way you collect that payment is in the detailing: slope the base to an outlet, daylight it to a ditch or a slope so water has a path to leave, or tie it to an edge drain that pipes the water away.

The quiet failure is a permeable base that does not actually drain, too thin, too flat, or plugged with plastic fines, so it stores water against the subgrade instead of shedding it. That is worse than no drainage layer. Keeping the base draining and the subgrade sloped is the construction half of this, covered in the base and subgrade compaction guide. The design half is making sure the section, the grades, and the edge drains give the water a real exit before the first storm finds out they do not.

Perpetual pavement: design it once, resurface forever

A perpetual pavement is a thick, durable asphalt section designed so that structural distress never reaches the bottom of the asphalt, and the only maintenance it ever needs is periodic renewal of the surface. Instead of designing for a 20-year life and then a rebuild, you design for 50 years or more of structure, and you plan to mill and inlay the top course every couple of decades as it wears. The structure underneath is meant to last the life of the road.

The mechanism is the fatigue endurance limit. Asphalt that flexes below a certain tensile strain at the bottom of the layer does not accumulate fatigue damage, no matter how many times it flexes. Build the section thick enough that the strain at the bottom stays under that limit, commonly cited somewhere in the range of 70 to 125 microstrain with more recent work pushing higher, and bottom-up fatigue cracking simply does not start. The distress that does happen, surface rutting and aging, lives in the top few inches where you can mill it off and replace it. Total asphalt thicknesses for moderate to heavy truck traffic commonly run on the order of 10 to 15 in.

The case for it is life-cycle cost on a heavy-truck route. The first cost is higher because the section is thick, but you never dig out and rebuild the structure, and the surface renewals are cheap and fast compared to a full reconstruction with the traffic disruption that comes with it. On a high-volume corridor that math often wins. On a light-duty parking lot it rarely does, because the traffic never justified the thick section in the first place.

How thick should a parking lot be vs a road?

Thickness follows the traffic, so a light-duty parking lot, a normal road, and a heavy-duty truck area get three different sections even on the same soil. A passenger-car parking lot sees almost no real ESALs, so it can run a thin section, a couple of inches of asphalt over a modest aggregate base, sized as much for constructability and a sound surface as for structure. The cars are not what it is built for. The occasional delivery truck and the trash truck are.

A road or a collector street carries a steady diet of trucks and buses, so it needs a real structural section sized to the design ESALs over its life, the full AASHTO or agency design rather than a minimum thickness. Heavy-duty areas, truck aprons, bus pads, loading docks, drive-through lanes, and dumpster pads, are thicker still, because that is where the trucks concentrate and where slow, heavy, turning loads punch the hardest. A common and costly mistake is paving a whole lot to one thickness, then watching the dumpster pad and the truck entrance rut and shove while the car field sits fine.

The economical move on a mixed-use site is to map the truck paths and build to them. Pave the car field to its light section, then thicken the entrances, the truck routes, the dock approaches, and the dumpster pads, or switch those spots to concrete where the loads are slow and heavy. You spend the money where the ESALs are and you stop spending it where they are not. That targeting is most of the value engineering on a parking-lot job.

Data center pads, industrial yards, and equipment loads

Heavy industrial pavement is its own design problem because the loads dwarf anything in the ESAL tables built for highway trucks. A container terminal, a laydown yard, a crane pad, or the haul route on a data center site carries point loads from reach stackers, loaded trailers, crane outriggers, and transformer hauls that a parking-lot section was never sized for. The standard ESAL framework, tuned to the 18,000 lb axle, starts to break down when the real load is a crane outrigger putting tens of thousands of pounds on a single pad, so these jobs lean on the geotech and a specific engineered design rather than a catalog thickness.

Two things change. The load reaches deeper, so the soft layer a car would never find is exactly what a loaded stacker punches through, which pushes the design toward a thicker, stronger section and often a stabilized or deeper subgrade. And the loads are slow and concentrated, sometimes static, which is brutal on asphalt because asphalt creeps under sustained heavy load. That is why crane pads, container stacking areas, and transformer-haul routes frequently go to concrete or to a very thick, stiff section instead of conventional asphalt.

Data center and critical-facility sites add the cost of movement. A pavement or pad that settles unevenly over a poorly compacted subgrade is a nuisance under a parking lot and a serious problem next to a building full of equipment that cannot tolerate differential settlement. The geotech on these jobs usually drives tighter compaction, deeper proof rolling, and a heavier section than routine paving, and the spec earns that rigor. Design these to the engineered loads on the drawings, not to a thickness borrowed from a commercial lot.

Overlay and rehabilitation design

Overlay design is a different calculation from new design, because you are not building a section from scratch, you are adding to one that already exists and has already used up part of its life. The existing pavement still has structural value, just less than when it was new, and the design problem is figuring out how much capacity is left and how much overlay to add to carry the future traffic. Add too little and the old fatigue telegraphs straight through the new mat. Add too much and you have paid for structure the road did not need.

The methods estimate the remaining structural capacity of the existing pavement, by condition survey, by coring, or by deflection testing with a falling weight deflectometer that back-calculates the in-place layer strengths, then design the overlay to make up the difference between that remaining capacity and what the future ESALs demand. The condition of the existing surface drives the prep: cracks reflect through an overlay if you do not address them, so the design has to deal with the existing distress, not just cap it.

The honest part of overlay design is knowing when an overlay is the wrong answer. An overlay on a pavement that is failing structurally from the base up, with widespread alligator cracking and pumping, is good money chasing bad, because the problem is below the surface and a new surface does not fix it. That is a reconstruction, or at least a mill and replace deep enough to reach sound structure. The milling and overlay work itself, the depth, the prep, the bonding, is a placement topic covered separately; here the point is sizing the overlay to the structure that is actually left.

Materials and mix by layer

The design assumes a material strength in every layer, so the mix and the aggregate have to match what the layer coefficients credited. The surface course gets a finer, denser, higher-quality mix with a binder graded for the local climate, because it takes the traffic, the water, and the weather directly. The binder and base asphalt can run coarser, cheaper mixes, since their job is structural thickness rather than surface performance. Use a richer surface mix where the layer wanted a base mix and you have overspent; use a lean base mix where the surface needed quality and you have under-built the layer that wears.

The binder grade is the climate input made concrete. A binder selected for the high and low pavement temperatures of the region resists rutting in summer heat and cracking in winter cold, and the wrong grade fails at one end or the other regardless of how good the thickness design was. Polymer-modified binders buy more performance at the temperature extremes and under heavy or slow loads, which is why they show up on truck routes, intersections, and heavy-duty areas.

The base material has to match its coefficient too. The aggregate base credited at a given strength assumed a dense-graded, angular, crushed stone within a gradation band and with controlled fines, draining the way the drainage coefficient assumed. A cheaper rounded gravel or a base running heavy on plastic fines does not perform like the material the design priced, so the section is weaker than the number on the drawing. Build the layers the design assumed; the mix selection by layer is its own placement topic, but the design depends on getting it right.

The design is only as good as the build

A pavement design is a set of assumptions about what gets built, and every one of them has to come true in the field or the section is weaker than the number says. The design assumed a subgrade at a certain strength, layers at certain thicknesses, a base compacted to density and draining, and an asphalt mat rolled to its target air voids. Miss any of those on the ground and the as-built pavement does not match the as-designed pavement, no matter how clean the calculation was.

The two construction failures that quietly undo a good design are an under-built foundation and an under-compacted mat. A subgrade or base that never reached density settles and deflects under traffic, which is the base and subgrade compaction work, the proof rolling, the Proctor target, and the hold point before paving. An asphalt mat left a couple of points low on density has interconnected air voids that let water in and fatigue early, which is the asphalt compaction window work, getting density before the mat goes cold. The design counted on both being right.

So treat the design and the build as one chain. The thinnest acceptable section that is built correctly outperforms a generous section built over a soft spot or rolled cold. When a pavement fails early, the question is rarely whether the thickness was wrong on paper. It is whether the subgrade held, the base drained, and the mat hit density. Design to the inputs, then build to the design, and document both so the failure that does not happen never gets blamed on the wrong layer.

The agency and DOT method or catalog

Most pavement is not designed from the AASHTO equation on a blank sheet. It is designed off the adopted method of the agency that owns or permits the work, and that method is the one that controls. State DOTs run their own design procedures, calibrated layer coefficients, and minimum sections, often a version of AASHTO 93 or Pavement ME tuned to local materials and climate. Cities and counties publish standard pavement sections by street classification. Many private and commercial sites are designed to a local building department standard or a geotech firm's recommendation that bakes the same logic into a recommended section.

The catalog or design table is the common form on smaller jobs. It is a lookup: for this road class or this traffic level, on this subgrade category, build this section. Those tables are the AASHTO or Pavement ME design already run for the agency's typical conditions, with the inputs pre-selected and conservative. They are legitimate and they save time, but they carry the agency's assumptions, so a catalog section is right for the conditions it was built for and not automatically right for an unusual soil or an unusual load.

The practical rule is to design under the method the project requires and confirm it against the geotech for the actual site. If the agency catalog assumes a subgrade stronger than the one the geotech found, the catalog section is under-built and you owe the owner a flag, not a silent install. The agency method governs the form of the design; the project geotech governs whether the site fits the assumptions it was built on.

Value engineering: the right thickness, not the over-build

Value engineering on pavement is finding the cheapest section that still carries the traffic for the design life, and it is not the same as building thin. The cheapest pavement over a full life cycle is the one sized correctly to the inputs, because both errors cost money: under-design costs a rebuild a decade early, over-design costs inches of asphalt the traffic never used. The win is precision, not minimalism.

The biggest savings come from the trades the structural number already lets you make. Carry structure in cheap aggregate base instead of expensive asphalt where the layer coefficients allow it. Map the truck paths and thicken only where the ESALs are, instead of paving the whole lot to the truck-area thickness. Improve a weak subgrade so the section above it can be thinner, where the cost of stabilization beats the cost of the extra structure. Each of these holds the same structural capacity for less money, which is real value engineering and not a thickness cut.

What does not count is shaving the section below what the traffic demands to win a bid. That is not value engineering, it is moving the cost from the bid to the owner's rebuild, and it shows up as fatigue cracking right on schedule. The estimator who prices the section the design actually needs, and who flags where the truck areas drove it thick, is protecting both the bid and the pavement. Cut the over-build, never the design margin the traffic earned.

What to document

A pavement design that nobody can reconstruct later is a number with no defense. When the pavement performs, or fails, the record is what answers whether the section was right for the inputs and whether what got built matched what was designed. On a disputed early failure it is the difference between a defended design and an expensive guess.

Capture the design inputs and the resulting section together. The design ESALs and how they were derived, the subgrade strength and its source, the method used and its key constants, the required structural number, and then the as-designed section layer by layer: material, thickness, and the layer coefficient credited to each. Keep it next to the construction record, the density tests on the subgrade and base and the as-built mat density, so the file shows both that the design was sound and that the build matched it. The table below is the spine of that record.

Common mistakes

• Designing without the traffic, sizing a section by eye or by habit instead of to the design ESALs.
• Counting cars instead of trucks, when the trucks do nearly all the damage under the fourth-power load law.
• Ignoring a weak subgrade, designing to an assumed strength the field soil does not actually have.
• Under-designing the truck areas, paving the dock apron, dumpster pad, and entrance to the car-field thickness.
• Leaving drainage out of the design, taking no drainage credit penalty and then letting the base stay wet.
• Copying a thickness from another job without checking the new site's traffic and subgrade.
• Confusing the structural number with a thickness, or mixing layer coefficients from different agency methods.
• Building a section that does not match the design, under-compacted base or a mat rolled cold, then blaming the thickness.

Field checklist

Standards and references

The governing documents on any given job are the adopted agency design method and the project geotechnical report. The agency, a state DOT, a city, a county, or the permitting authority, sets the design procedure, the calibrated layer coefficients, the reliability and serviceability values, and the minimum sections. The geotech sets the subgrade strength the design rests on. Where the general practice in this guide differs from those documents, those documents win.

The empirical framework most United States flexible design still draws from is the AASHTO Guide for Design of Pavement Structures, commonly called AASHTO 93, which produces the structural number from the design ESALs, the subgrade resilient modulus, reliability, and serviceability loss. The mechanistic-empirical successor is the AASHTOWare Pavement ME software, out of the NCHRP 1-37A research, which models layer response and predicts distress over the design life. The Asphalt Institute thickness-design references, including MS-1, cover flexible design and full-depth and perpetual sections.

Subgrade strength testing such as the resilient modulus follows AASHTO T 307, and the CBR and R-value have their own ASTM and AASHTO procedures and loose correlations to the modulus. The constants in any of these, the layer coefficients, the serviceability and reliability values, the correlation factors, shift between editions and agencies, so confirm the current edition and the local amendments before citing any number on a submittal, and never carry a value from one agency's method onto another agency's job.

Units, terms, and conversions

Pavement design carries a few names and unit systems across the geotech report, the design method, and the drawings, so the same idea reads differently from one document to the next.

Traffic is in ESALs, equivalent single axle loads, sometimes written as the design traffic or the 18-kip ESALs, referenced to the standard 18,000 lb (80 kN) axle. Subgrade strength shows up as resilient modulus (Mr) in psi or MPa, as CBR (a percentage against crushed stone), or as R-value. Layer thicknesses are in inches in United States practice and millimeters in metric. The structural number (SN) is dimensionless. Layer coefficients are per inch of thickness. Reliability and serviceability are dimensionless figures the agency sets by road class.

ESAL

Equivalent single axle load, the standard 18,000 lb axle pass used to express traffic as damage

Structural number (SN)

The total structural strength AASHTO 93 says the section needs over the subgrade; not a thickness

Layer coefficient

The relative structural strength per inch of a given layer material

Resilient modulus (Mr)

The stiffness of the subgrade under repeated wheel loading, the main subgrade input to AASHTO design

CBR

California Bearing Ratio, a subgrade strength measure relative to crushed stone, correlated loosely to Mr

Pavement ME / MEPDG

The mechanistic-empirical design method that models layer response and predicts distress over the life

Perpetual pavement

A thick section designed so fatigue never reaches the bottom of the asphalt; renewed by resurfacing

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is proof rolling?
• What temperature do you stop rolling asphalt?
• What is Superpave?
• What causes alligator cracking?
• Is asphalt or concrete better for pavement?