Civil Engineering July 2021 | Vol 29 No 6

Civil Engineering July 2021 31 Q Q Scenario 2: using the ELFs given in the contract documents and increased annual traffic growth of 10% per year from 2020 to 2034 Q Q Scenario 3: using the proposed in- creased ELFs based on maximum legal axle loads and annual traffic growth of 5% per year from 2020 to 2034. A comparative plot of the cumulative ESAL for Gbargna and Ganta based on the actual traffic counts and the original predictions for Scenario 3 is given in Figure 10. Assessment of the cumulative traffic indicated that the revised design traffic based on actual traffic count data varies significantly from the original forecasted values. The traffic at Gbargna is domi- nating as opposed to the original forecast which dictated that the traffic at Ganta is to be used in the pavement design. This demonstrates the uncertainties that exist in determining the design traffic, and a conservative approach was therefore considered prudent. OVERLAY DESIGN Methodology As described in AASHTO, 1993 (Part III Design of Rehabilitation of Existing Pavement), the effective pavement life and the structural number (SN eff ) of the existing pavement can be deter- mined using three different estimation methods: Q Q Method 1 – Visual condition: adjust the material coefficients for the various pavement layers, based on the percentage and type of distress observed Q Q Method 2 – Deflections: determine the roadbed resilient modulus based on measured pavement deflections for estimating pavement life . Q Q Method 3 – Fatigue damage: de- termine the remaining structural capacity based on past traffic and the use of a condition factor. This method is considered particularly appropriate when there is little visible distress on the pavement surface. Layer coefficients The original design was essentially based on CBR-correlated layer coefficient values. However, the layer coefficient can vary significantly from the CBR- correlated value, based on in-situ condi- tions such as stress state, moisture state and layer thickness. The various layer coefficient values depending on laboratory CBR, resilient modulus and visual condition are pro- vided in Table 2. Subgrade resilient modulus As part of the original pavement design verification, the subgrade or roadbed resilient modulus was determined as 15 151 psi, based on CBR values. In terms of Method 2, M r is to be determined from the mean pavement deflections under a single load, as measured in the FWD tests. Using this method resulted in a back-calculated value of 10 428 psi. Asphalt overlay thickness A sensitivity analysis was carried out based on varying subgrade resilient moduli, design traffic and SN eff . A sum- mary of the results is given in Table 3. CONCLUSIONS AND RECOMMENDATIONS It was recommended that the contractor proceed with placing the 50 mm asphalt overlay, as originally planned, based on the following: Q Q Although the actual traffic volumes are lower than originally forecasted, it is recognised that uncertainties with regards to future traffic growth and actual load equivalency factors necessitates a conservative approach to determining the 20-year design traffic. Q Q The actual layer coefficient values (and associated pavement capacity) may be less than what the design was originally based on due to a variation in moisture state, stress state and defects developing (pavement visual condition). Q Q The subgrade resilient modulus value, back-calculated from the average FWD deflection, is significantly less than the value assumed for the original design, resulting in lower pavement capacity. Q Q Some individual high pavement deflec- tions are already present, exceeding the maximum 95 th percentile value of 0.7 mm. If an overlay is not placed now, as originally planned, there is a risk that the contractor may not achieve the specified pavement condition at the end of the contract.  Table 2 Variation in layer coefficient values Layer coefficient CBR Value E (psi) – damp Value E (psi) – wet Value Visual condition Value Asphalt, a 1 - - 400 000* 0.42 - - No or little cracking 0.35 to 0.40 Base, a 2 100 0.14 20 000 0.09 15 000 0.06 No pumping 0.10 to 0.14 Subbase, a 3 30 0.11 14 000 0.10 12 000 0.09 No pumping 0.10 to 0.14 *All moisture states Table 3 Variation in designed asphalt overlay thickness Traffic scenario 3 3 3 1 1 1 Design traffic (W18) 20 years, mill ESALs 23.4 23.4 23.4 17.1 17.1 17.1 SN F required 4.344 4.344 4.53 4.35 4.35 4.35 Method/approach to determine SN eff Stress state Visual condition Fatigue damage Stress state Visual condition Fatigue damage Mr (psi) 12 000 12 000 10 428 10 428 10 428 10 428 SN eff 3.38 3.44 3.50 3.38 3.44 3.60 Overlay thickness (mm) 56 52 62 56 53 45