A Pavement Structure That Does Not Add Up to the Required Structural Number
A new four-lane divided arterial is to be designed for 20 years with a design ESAL of W18 = 12.0 million in the design lane. Subgrade resilient modulus is 5,000 psi. The base course (Mr = 30,000 psi) drains to 50 percent saturation within one day. The subbase (Mr = 15,000 psi) is fair quality. The water table is within 3 ft of the surface for about four months each year, and the pavement structure will be near saturation more than 25 percent of the time. Design reliability is 95 percent (ZR = -1.645), overall standard deviation 0.45, and the serviceability loss is 4.2 to 2.5. Available layer coefficients are a1 = 0.44, a2 = 0.14, a3 = 0.11. The contractor has proposed 4 in of asphalt, 10 in of base, and 12 in of subbase, arguing that AASHTO is conservative anyway.
1. Solve for the required structural number on the subgrade and evaluate the contractor's proposed structure, showing the drainage adjustment.
Read the worked answer
AASHTO 1993 flexible design equation, solved for SN (W18 = 12.0e6, ZR x S0 = -1.645 x 0.45 = -0.740, delta PSI = 1.7, Mr = 5,000 psi):
log10 W18 = ZR x S0 + 9.36 log10(SN+1) - 0.20 + log10[dPSI/(4.2-1.5)] / [0.40 + 1094/(SN+1)^5.19] + 2.32 log10(Mr) - 8.07
Substituting constants: 7.079 = -0.740 + 9.36 log10(SN+1) - 0.20 - 0.201/[0.40 + 1094/(SN+1)^5.19] + 0.512
Trial SN = 6.1: 9.36 log10(7.1) = 7.968; the denominator term evaluates to 0.442 -> -0.201/0.442 = -0.455. Sum = 7.085, i.e. W18 = 1.2e7. SN required = 6.1.
Drainage coefficients. Base drains within one day = 'fair' drainage quality; exposure to near-saturation exceeds 25 percent of the time, so from the AASHTO drainage table m falls in the 1.15-1.05 range. Use m = 1.05 (the conservative end) for both base and subbase, since the water table is high for a third of the year.
Contractor's structure SN = 0.44(4) + 0.14(10)(1.05) + 0.11(12)(1.05) = 1.76 + 1.47 + 1.39 = 4.62
Result. The proposal is short by 1.48 SN, roughly 3.4 in of asphalt equivalent. Even ignoring drainage (m = 1.0) it only reaches 4.48 against a requirement of 6.1. This is not a case where 'AASHTO is conservative' bridges the gap - a 24 percent structural deficit on a 12-million ESAL facility will show up as alligator cracking and rutting well before year 20, and it will show up fastest where the water table is high.
2. The contractor's alternative is to make up the deficit with granular base rather than asphalt. Evaluate that and check the layer-by-layer requirements.
Read the worked answer
Layer requirements from the same equation at higher moduli
- SN over the base (Mr = 30,000 psi) = 3.3
- SN over the subbase (Mr = 15,000 psi) = 4.2
- SN over the subgrade = 6.1
Granular substitution. Each inch of base adds 0.14 x 1.05 = 0.147 SN. Covering 1.48 SN would require about 10 additional inches of base, taking the base to 20 in. Check that this satisfies the layered checks:
- SN over subbase = 1.76 + 0.14(20)(1.05) = 1.76 + 2.94 = 4.70 >= 4.2 OK
- SN over base = 1.76 vs required 3.3 - fails. The asphalt itself is too thin to protect a 30,000-psi base, regardless of how much base is added.
Judgment. Granular substitution is viable only if the asphalt is also thickened. Minimum asphalt = 3.3/0.44 = 7.5 in. A workable structure is 7.5 in AC + 12 in base + 10 in subbase: SN = 3.30 + 1.76 + 1.16 = 6.22 >= 6.1, with SN over base = 3.30 and SN over subbase = 5.06, both satisfied.
Why not simply pile on base. Unstabilized base loses modulus quickly when it stays wet, and this site is near saturation a third of the year. The 1.05 drainage coefficient is already discounting the base; leaning harder on it multiplies that risk. Also consider the practical limit: a 20-in granular base is a construction and right-of-way problem on an urban arterial with curbs and utilities.
3. The ESAL forecast is uncertain: truck growth could be 4 percent rather than 8 percent, which lowers W18 to 8.5 million. Should the design be revised, and how would you present this to the agency?
Read the worked answer
Recompute SN at W18 = 8.5e6 (log10 = 6.929, all other inputs unchanged): solving the same equation gives SN required = 5.87.
Sensitivity. A 29 percent reduction in design traffic buys only 0.21 SN, or about 0.5 in of asphalt (0.21/0.44). The relationship is logarithmic - traffic changes translate into small structural changes.
Should the design be revised? No, and the reasoning matters more than the number:
- 0.5 in of asphalt is within the tolerance of a single lift and is far cheaper now than a mill-and-overlay later. The cost asymmetry is roughly two orders of magnitude.
- The credible downside risk is not the truck growth rate - it is the drainage assumption. If the m = 1.05 assumption proves optimistic (poorer drainage or longer saturation), the structural penalty is comparable to the entire traffic uncertainty band, and it is not recoverable by any traffic forecast.
- The layer check on the asphalt thickness (SN over base = 3.3) is driven by base modulus, not by a small change in W18, so it does not relax meaningfully either.
How to present it. Show the agency a table of SN versus W18 (8.5, 12.0, and a high case such as 16 million) so the insensitivity is visible, keep the design at the 12-million structure, and redirect the cost discussion to the drainage details - subsurface drainage, shoulder and ditch geometry, and sealing - where the real risk sits.
Judgment. Do not switch to a mechanistic-empirical procedure to resolve this; MEPDG would add analytical sophistication without changing the conclusion that drainage and construction quality dominate. Reserve that recommendation for a facility where the layer-check or materials questions are genuinely unresolved.