NivaarExam PrepOfficial exam papers ↗

16-Civ-B7 Transportation Planning and Engineering · May 2015

Question 3 of 7: Flexible pavement design by the 1993 AASHTO method

Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)

Notes on this paper

Paper format. National Examinations, May 2015 — 98-Civ-B7 Highway Engineering. Three hours, open book, any non-communicating calculator permitted. Seven questions of equal value (20 marks each); a total of five solutions constitutes a full paper, and only the first five in the answer book are marked. All seven are solved here so the set works as a study resource. Note 1 of the paper invites a clear statement of any assumption made, and Note 2 permits any data required but not given to be assumed — both are used below and every assumption is flagged.

Reference texts. Transportation Association of Canada, Geometric Design Guide for Canadian Roads; AASHTO, Guide for Design of Pavement Structures (1993); AASHTO, A Policy on Geometric Design of Highways and Streets; Garber & Hoel, Traffic and Highway Engineering; Mamlouk & Zaniewski, Materials for Civil and Construction Engineers; Asphalt Institute, Asphalt Mix Design Methods (MS-2); TAC, Pavement Asset Design and Management Guide; Chow, Open-Channel Hydraulics; Neville, Properties of Concrete.

Question 3: Flexible pavement design by the 1993 AASHTO method

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.

Given. The traffic, serviceability, reliability and subgrade inputs of the 1993 AASHTO flexible design procedure, for a four-lane divided highway (two lanes in each direction).

Design inputs
ParameterSymbolValue
ESALs per day, per direction—900
Lane distribution factor$D_L$0.80 outside lane, 0.20 inside lane
Design period$n$20 years
Annual traffic growth rate$g$3.5 percent
Initial serviceability$p_i$4.3
Terminal serviceability$p_t$2.5
Reliability$R$90 percent
Overall standard deviation$S_o$0.40
Effective roadbed resilient modulus$M_R$30 MPa

Find. The structural number required over the design period and a layer thickness combination that provides it, for the design (outside) lane, with the inside lane checked separately.

Approach. Accumulate the daily ESALs over the design period with the compound growth factor, apportion them to the design lane, solve the 1993 AASHTO flexible equation for the structural number, then distribute that structural number across asphalt, base and subbase using standard layer coefficients.

  1. Read the growth input correctly. The paper prints a traffic growth of 3.5 percent, which is an annual rate, not the cumulative AASHTO growth factor. It cannot be the cumulative factor: that quantity is $\mathrm{GF} = [(1+g)^n - 1]/g$, whose smallest possible value over a 20-year period is $n = 20$ itself, at zero growth. Accumulating 20 years of compound growth at 3.5 percent gives $$\mathrm{GF} = \frac{(1 + g)^n - 1}{g} = \frac{(1.035)^{20} - 1}{0.035} = 28.280$$
  2. Accumulate the directional traffic. Nine hundred ESALs per day in one direction is $$W_{18,\text{dir}} = 900 \times 365 \times \mathrm{GF} = 328\,500 \times 28.280 = 9.289 \times 10^{6}\ \text{ESALs}$$ over the twenty years. Because the paper already gives the traffic per direction, no further directional distribution factor is applied — applying one as well would halve the design traffic and is the commonest error in this question.
  3. Apportion to the design lane. The outside lane carries 80 percent of the directional traffic and governs the design: $$W_{18} = D_L\,W_{18,\text{dir}} = 0.80 \times 9.289 \times 10^{6} = \boxed{7.43 \times 10^{6}\ \text{ESALs}}$$ The inside lane sees $0.20 \times 9.289 \times 10^{6} = 1.86 \times 10^{6}$ ESALs and is checked at the end.
  4. Assemble the remaining terms. At 90 percent reliability the standard normal deviate is $Z_R = -1.282$; the serviceability loss is $\Delta \mathrm{PSI} = p_i - p_t = 4.3 - 2.5 = 1.8$; and the roadbed modulus must be converted to the units in which the equation was calibrated, $$M_R = 30\ \text{MPa} \times 145.04 = 4\,351\ \text{psi}$$ Substituting 30 directly returns nonsense, because the constant 2.32 multiplies $\log_{10} M_R$ in psi.
  5. Solve the 1993 AASHTO flexible equation for the structural number. $$\log_{10} W_{18} = Z_R S_o + 9.36 \log_{10}(SN + 1) - 0.20 + \frac{\log_{10}\!\left[\dfrac{\Delta \mathrm{PSI}}{4.2 - 1.5}\right]}{0.40 + \dfrac{1094}{(SN+1)^{5.19}}} + 2.32 \log_{10} M_R - 8.07$$ The equation is implicit in $SN$ and is solved by bisection. With $\log_{10}(7.43\times10^{6}) = 6.871$ the root is $$\boxed{SN_{\text{required}} = 5.47}$$ Substituting 5.47 back reproduces $7.43 \times 10^{6}$ ESALs, which is the check that the root is the right one.
  6. Choose layer coefficients and drainage coefficients. The paper supplies none, so under its Note 2 the standard AASHTO values for good-quality Canadian materials are adopted: $a_1 = 0.44$ for dense-graded asphalt concrete, $a_2 = 0.14$ for a crushed granular base, $a_3 = 0.11$ for a granular subbase, with drainage coefficients $m_2 = m_3 = 1.00$ for a structure that drains within a day and is saturated a small fraction of the year.
  7. Distribute the structural number. The structural number is assembled as $$SN = a_1 D_1 + a_2 m_2 D_2 + a_3 m_3 D_3$$ with the thicknesses in inches. Trying 150 mm of asphalt concrete over 300 mm of base and 300 mm of subbase, $$\begin{aligned}a_1 D_1 &= 0.44 \times (150/25.4) = 2.598 \\a_2 m_2 D_2 &= 0.14 \times 1.00 \times (300/25.4) = 1.654 \\a_3 m_3 D_3 &= 0.11 \times 1.00 \times (300/25.4) = 1.299\end{aligned}$$ so that $$SN_{\text{provided}} = 2.598 + 1.654 + 1.299 = \boxed{5.55 \ge 5.47}$$ a margin of about 1.5 percent, over a total pavement thickness of 750 mm.
  8. Check the inside lane. Repeating the solution at $W_{18} = 1.86 \times 10^{6}$ gives $SN_{\text{required}} = 4.50$, about 82 percent of the outside-lane requirement even though it carries a quarter of the traffic — the equation is strongly logarithmic in load. Holding the asphalt and base constant, that structural number would need only 56 mm of subbase, which is below any practical placing and compaction thickness. The inside lane is therefore built to the same section as the outside lane, which is normal practice: the saving from thinning one lane of a two-lane carriageway is not worth the longitudinal construction joint and the difference in future performance.
asphalt concrete150 mma·D = 2.598crushed granular base300 mma·D = 1.654granular subbase300 mma·D = 1.299prepared subgradeTrial section — total 750 mmSN provided 5.55 ≥ SN required 5.47layer coefficients a1 = 0.44, a2 = 0.14, a3 = 0.11; m2 = m3 = 1.00
Adopted flexible pavement section for the outside (design) lane. The label on each layer is its contribution to the structural number; the layer thicknesses are what is actually built.

It is worth naming what the structural number is and is not. It is a dimensionless index of the total load-spreading capacity of the bound and unbound layers above the subgrade, not a thickness: the section that delivers $SN = 5.55$ here is 750 mm deep. Expressed in the millimetre convention used by several prairie agencies it would be quoted as $5.47 \times 25.4 = 139$ mm, which is the same number in different units and still not a depth. Any combination of layers reaching the required index is acceptable to the equation; it is minimum practical thicknesses, the requirement that each layer protect the one beneath it, and cost that decide which combination is built.

Check — assumptions stated under the paper's Note 1, with their consequences quantified. (i) The 3.5 percent is an annual growth rate, for the reason given in step 1. (ii) The ESALs given are already directional, so no further directional distribution factor is applied. (iii) Layer coefficients 0.44 / 0.14 / 0.11 and drainage coefficients $m = 1.00$ are assumed. The design is far more sensitive to drainage than to reliability: keeping the same 750 mm section but assuming $m_2 = m_3 = 0.90$ drops the structural number provided from 5.55 to 5.25, which no longer meets the requirement, whereas raising the overall standard deviation from 0.40 to the 0.45 more usual for flexible pavements raises the requirement only from 5.47 to 5.58, and going to 95 percent reliability raises it to 5.72. If the drainage of the granular layers cannot be assured, the subbase should be increased to 400 mm.
Final results
QuantityValue
Cumulative growth factor over 20 years28.280
Directional 20-year ESALs9.29 × 106
Design-lane (outside) ESALs, $W_{18}$7.43 × 106
Inside-lane ESALs1.86 × 106
Roadbed resilient modulus30 MPa = 4 351 psi
Structural number required, outside lane$SN = 5.47$
Structural number required, inside lane$SN = 4.50$
Asphalt concrete150 mm ($a_1 D_1 = 2.598$)
Crushed granular base300 mm ($a_2 m_2 D_2 = 1.654$)
Granular subbase300 mm ($a_3 m_3 D_3 = 1.299$)
Structural number provided$SN = 5.55 \ge 5.47$
Total pavement thickness (both lanes)750 mm