Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Computational Science, Engineering & Technology Series
ISSN 1759-3158
Edited by: Y. Tsompanakis, P. Iványi and B.H.V. Topping
Chapter 10

Modelling Deterioration and Maintenance of Australian Low Volume Traffic Roads

K.J. McManus

Centre for Sustainable Infrastructure, Swinburne University of Technology, Melbourne, Australia

Full Bibliographic Reference for this chapter
K.J. McManus, "Modelling Deterioration and Maintenance of Australian Low Volume Traffic Roads", in Y. Tsompanakis, P. Iványi and B.H.V. Topping, (Editors), "Civil and Structural Engineering Computational Methods", Saxe-Coburg Publications, Stirlingshire, UK, Chapter 10, pp 185-203, 2013. doi:10.4203/csets.32.10
Keywords: pavement deterioration models, maintenance model, low volume traffic roads, pavement roughness, surface profile, long wavelength roughness, neural network analysis, Markov decision process.

The modelling of pavement performance assists asset managers to effectively plan for the maintenance and timely replacement of deteriorated pavements. The development of deterioration and maintenance models for roads have been based on data from well-built and well-maintained heavy volume traffic (HVT) pavements of the major road systems subjected to heavy traffic loads as measured by traffic volume and axle loads. The models generally require substantial performance data obtained over the long term. In Australia, in addition to the heavily loaded pavements, there is a substantial network of thin sealed flexible pavements providing essential access to regional areas. The models for the major pavement set do not necessarily apply to the low volume traffic (LVT) roads in the regions of Australia away from the major cities. The quality of the performance data for the LVT roads is generally not as good as that for the major road system. Hence, the need to move towards the development of deterioration and maintenance models for LVT roads. The LVT roads selected for the project were of single carriageway, two-way, two-lane roads with unbound sealed granular pavements and unsealed shoulders. LVT roads are defined as those carrying average annual daily traffic (AADT) below 2000 with a minimum of 10% heavy vehicles (HV).

The relatively recent availability of objective surface profilometer measurements for some LVT roads has provided a new database supplementing the previous incomplete road roughness measurements supported by rutting measurements, and the mainly subjective assessments of surface cracking. These recent profile data allow for an analysis of the wavelengths making up the surface supporting defect analysis of the pavement. The examination of various wavebands gives guidance to the deterioration of the different pavement elements and their rehabilitation. Significantly, one outcome from this research is to support diagnosis of the most likely cause of pavement deterioration using the now readily available roughness data.

Another outcome is the detection of the presence of long wavelength roughness (LWR), not previously detected by roughness measurements. LWR has the effect of increasing dynamic axle loads in the rough section and downstream. The result is increased failure rates of pavements in these regions. Early detection allows preventative maintenance to alleviate the problem.

In addition to the surface analysis, the outcomes of the research project include an improved deterioration model for LVT road pavements. Initially, in the absence of profilometer data, LVT road deterioration models were developed using roughness data from a sample of local government authority (LGA) roads. An approach using a neural network analysis (NNA) relating age and total pavement thickness to roughness was examined and found to be useful but in need of improvement. Another model was based on regression analysis linking rutting depth and adjusted structural number (SNP), for different climatic zones and levels of traffic loading.

A maintenance model (MM) was developed by combining a probabilistic approach with a Markov chain procedure and linking into forecasting the need for a maintenance budget. A probabilistic model is used to address the change of pavement condition, as it is seen to better handle the unknowns, such as changes in traffic load and deterioration modes. The Markov decision process is employed, which can also cope with different levels of maintenance effort. It is also possible to optimise maintenance activity under budget constraints. This model relates maintenance expenditure with road performance.

This study has concentrated on LVT roads in Australia that constitute a substantial proportion of the total road network and provide important freight and communication links within regional Australia. Its purpose was to seek improved deterioration models for the thin seal flexible pavements that make up the bulk of the LVT roads. Furthermore, guidelines concerning maintenance type, effectiveness and optimisation of maintenance activity were proposed.

purchase the full-text of this chapter (price £20)

go to the previous chapter
go to the next chapter
return to the table of contents
return to the book description
purchase this book (price £75 +P&P)