The prediction of moisture content in untreated pavement layers and an application to design in southern Africa

Methods have been developed to predict pavement moisture content in South Africa so that it may be used in pavement design. The data used were taken from existing surfaced roads in the Transvaal, Cape Province, and Natal. The effects of materials and climate on moisture content were found. Models were developed to predict equilibrium moisture content and the ratio of equilibrium to optimum moisture content for all pavement layers and climates of South Africa. The seasonal and spatial variation of moisture content was studied. The minimum width of sealed shoulder to substantially reduce the probability of trafficking in the edge zone of seasonal moisture variation was defined. A new suction/climate relationship was developed to predict equilibrium suction on a regional basis, and a new climate map of Thornthwaite's moisture index was produced to complement this. Laboratory and field data were used to quantify the variation of strength with moisture for pavement materials. A new unsoaked CBR criterion was proposed for subgrade classification and design. The costs of pavements designed using the proposed unsoaked criteria were compared to the costs of those designed using the present soaked criterion. A Bayesian probability analysis approach was used to find the total discounted costs over the analysis period, and unsoaked design was shown to be less expensive than the soaked design for all but very high predicted moisture contents.


  • English

Media Info

  • Pagination: 82p
  • Serial:
    • Issue Number: 20

Subject/Index Terms

Filing Info

  • Accession Number: 01388042
  • Record Type: Publication
  • Source Agency: ARRB
  • Files: ATRI
  • Created Date: Aug 23 2012 12:35AM