Interpretation of Structural Evaluations
General information regarding principles of pavement design and structural evaluation is given in the AUSTROADS Pavement Design Guide. More theoretical background is given in Ref.1 for structural asphaltic pavements while unbound pavements are discussed in Ref. 2.
The layer 1 moduli are normally shown on a scale of 0-1000 MPa (typical range for unbound basecourse). (Other typical moduli are: asphaltic concrete 1000-5000 MPa; cement stabilised basecourse 2000-20,000 MPa; concrete roads >50,000 MPa.) The files may be imported to any word processor or spreadsheet to obtain a printout table of results. For project level surveys, the lane for each testpoint can be identified from the last digit of the stationing (0 for Increasing Route Position (IRP or left side), 5 for Decreasing RP, or other conventions for severely distressed points or multiple lanes).
The Austroads strain criteria used assume 'typical' soils, which may be overly conservative for strong yet very compressible soils such as unweathered ash. If any section shows excessive overlays from local experience, then precedent strain analyses should be considered. In cases where precedent strain analysis has been carried out, note that appropriate conditions (see Appendix) must apply before overlay results from GMP may be adopted.
The Precedent design method is given in the Transit Supplement to the AUSTROADS Pavement Design Guide. This contains two methods of pavement overlay design. The first is given in the above reference as Equation 10.3 and is equivalent to the NRB State Highway Pavement Design and Rehabilitation Manual (SHPDRM). The disadvantage of this method is that it provides similar thicknesses in weak and strong subsections of the road, ie overlays may be excessive in places but marginal elsewhere. These limitations are overcome by using the full Precedent Strain Method (Equations 10.3 and 10.4), which establishes an appropriate strain criteria for the subgrade soil throughout the section, and adjusts overlay thicknesses to accommodate the weaker and stronger subsections. The results should be useful in volcanic soils, where the prerequisite conditions apply. If the Precedent results are less conservative those for the GMP analysis then particular care is required to check that input parameters and the assumptions of the Precedent method are valid. The methods do not apply to network level surveys, but relate only to unbound granular pavements which have reached a terminal condition because of excessive subgrade strain. Separate checks (Austroads Supplement, Section 10.3) are required for distress in the pavement layers, especially degraded unbound basecourses (Section 10.3.2) or cement stabilised materials (Section 6.3.3). For conventional unbound granular basecourses, the Precedent methods use as input, the ratio of future to past traffic (ESA), and the percentage of the section which has reached a terminal serviceability condition. If these values have not been provided for this evaluation, results will not be valid but respective defaults of 3 and 50% will have been adopted for provisional analysis. Verification of these parameters and the total design ESA will be necessary. The numerical values for the overlays can be found in the xls spreadsheet file.
For unsealed roads programmed for seal extension, the in situ CBR's determined should be conservative for the season of testing. However overlay requirements are a LOWER BOUND result only. These thicknesses can only be used if the old surface layer (due to become the subbase) is free-draining. Consideration of possible seasonal moisture changes in all layers after sealing, is essential for reliable design.
Some tables give 2 values of CBR. For the existing pavement, a lower quartile percentage (LQP) CBR is given, using conservative correlations between insitu CBR and subgrade modulus. If reconstruction is considered, ie the designer is using AUSTROADS Fig. 8.4, then because this chart assumes anisotropic moduli and a correction is also required for modulus non-linearity, then an appropriate CBR (labelled RCN for reconstruction) is listed separately.
Preliminary graphs are intended to let the designer give due consideration to the findings of our initial modelling. It is important that the ESA traffic loading used in the analysis is confirmed prior to the adoption of any rehabilitation options presented. We request feedback regarding any results which appear anomalous so that we can study the initial models more closely to see whether refinements are appropriate. The techniques used are generally robust but there are occasional instances where changing the back-analysis parameters will cause significant variations in predictions. Feedback from the practical site experience of the local design team and their knowledge of regional precedence is therefore essential as a check on the standard mechanistic analyses used in the production of this pavement evaluation.
Please feel free to contact us for support on any aspect.
Austroads Publications can be purchased from here
The latest NZTA Supplement can be downloaded for free from here
Ullidtz, Per (1998), Pavement Analysis, Elsevier
Tonkin & Taylor Ltd (1998) Pavement Deflection Measurement and Interpretation for the Design of Rehabilitation Treatments,
Transfund Research Report No. 117
Diagnosis of Pavement Distress
In some cases the cause of premature distress in a pavement can be difficult to confirm from visual survey. However using visual condition data in conjuction with structural analysis provides a methodical tool for definitively establishing the mechanism of any distress.
The best technique in this instance is to have an observer working on site with the FWD operator, to ensure the geophones are effectively located on intervals of distress where the severity is essentially the same, ie avoid positions where the spread of geophones is partly on intact pavement and partly on distressed. The spread is usually 1.5 to 1.8 m, therefore constant distress severity over at least a 2 m length of wheelpath is necessary. The observer than moves along with the FWD, stopping the driver at meaningful positions and recording the nature and severity of distress for each test. Categories are usually rutting, cracking and shoving. Roughness may also be recorded if relevant lengths of constant severity are present. Severity must also be recorded. We recommend a simple ranking of 0-4. (None, Initial, Advanced, Terminal). Diagnosis software is available that allows the severity of distress to be related to any parameter, eg cracking severity may be plotted against the modulus of the surfacing to see if poor quality surface materials are implicated. Rutting severity may be plotted against modulus of each layer in turn, to determine if any one layer is the likely source of excessive deformation.
Correlation of severity against subgrade modulus exponent is often a guide to poor compaction at the top of the subgrade, or the need for drainage improvement.