Development of gravel loss and other condition performance models for unsealed roads

Transcription

1

2 Development of gravel loss and other condition performance models for unsealed roads George Giummarra ARRB Group Theuns Henning University of Auckland Janice Brass LTNZ

3 Outline of presentation Importance of performance models Developments in Australia Developments in New Zealand Practices to reduce gravel loss

4 Why update existing models? move from an art to a science better able to assess future needs achieving greater value justifying funding requirements

5 Making it happen in Australia

6 Nominated monitoring sites on local roads (as at end March 2006 State/Territory Number councils or organisations Asphalt* road sites Sealed road sites Unsealed road sites Total sites nominated New South Wales Victoria Queensland Western Australia South Australia Tasmania Australian Capital Territory Northern Territory Total

7 Performance parameters selected Sealed Roads Unsealed Roads Cracking Gravel loss Binder hardness Loss of shape Stone loss Roughness (ride quality) Rutting Loose stone Roughness (ride quality) Material Properties Structural strength

8 Typical unsealed roads in Australia

9 Measuring gravel loss

10 Gravel loss model for Australia GL D * (F1 *ADT + F2 * MMP + F3 * PF) Where: GL D ADT MMP PF P075 PI F1, F2, F3 gravel thickness loss (mm) time period in hundreds of days (days/100) average daily vehicular traffic in both directions, in veh/day mean monthly precipitation, in mm/month plasticity factor (PI P075) amount of material passing the mm sieve, in per cent by mass plasticity index model coefficients

11 Roughness model for unsealed roads IRITG2 IRImax exp{[ * (F0 + F4 * ADT * MMP/1000)] * (TG2-TG1)} * (IRImax IRITG1) Where: IRITG1 IRITG2 IRImax TG1, TG2 ADT MMP F0, F4 roughness at time TG1, in m/km IRI roughness at time TG2, in m/km IRI maximum allowable roughness for specified material, in m/km IRI time elapsed since latest grading, in days average daily vehicular traffic in both directions, in veh/day mean monthly precipitation, in mm/month model coefficients

12 Shape loss model SL F0 + F1 * ADT + F2 * P075 Where: SL ADT P075 F0, F1, F2 percentage (%) change in cross-fall per year average daily vehicular traffic in both directions, in veh/day amount of material passing the mm sieve, in per cent by mass model coefficients

13 NZ-Gravel loss background During 2002, Land Transport New Zealand established a gravel loss experiment in cooperation with 10 local authorities. Data collection on 51 sections for a period of five years 2007, Land Transport New Zealand Research Development of Gravel Loss Models (MWH NZ, ARRB, University of Auckland)

14 Study objectives Developing condition deterioration models such as gravel loss Develop a framework for adopting the deterioration models into a decision framework for the GRMS Propose some strategic level best practice guidelines for managing gravel roads from a performance perspective

15 Survey methodology Timber peg Concrete Benchmark Reference 1 0.5m Offset 1 ##.#m A1 A2 A3 A4 A5 A6 Iron Tube Cross Section 10m A B1 B2 B3 B4 B5 B6 Cross Section B 60m Fence Concrete Benchmark 3 Etc F1 F2 F3 F4 F5 F6 Cross Section F Timber peg Concrete Benchmark 2 Offset 2 ##.#m G1 G2 G3 G4 G5 G6 Iron Tube Cross Section G

16 Average Daily Traffic Average Rainfall (mm/yr) Statistical analysis results Number of Bladings Number of Bladings 9 10

17 Gravel data offers a challenge to understand Number of Average Site readings 20 Histogram Slope Loss 0 Number of Average Site readings Histogram of Gravel Loss Gravel Loss mm/yr Slope Loss m/m

18 A new performance indicator was developed Slope a. Datum Shape Loss Slope (%) Width ShapeIndex 10 ( Slopefactor + StdDevFactor ) Slope b

19 Model development results Three models were developed Gravel Loss GL ADT, P75, P265, Blading Freq, Surface width Slope Loss Sl plasticl, cbr, ADT, Days Since Last Blade, grade, p265, width.f Shape Loss Shape L plasticl, cbr, ADT, Days Since Last grade, p265, width.f Blade,

20 Impact on gravel road management A simple data collection process is recommended Have to start collecting & managing the data Include models to decision process Discontinue current monitoring more detail monitoring for specific cases

21 Factors contribution to gravel loss poor geometric standards drainage provisions wearing course materials construction and maintenance practices

22 Geometric design standards

23 Drainage, drainage, drainage

24 Selection of a wearing course

25 Construction and maintenance practices

26 Questions??

27 Research offers great promise but there are risks Lessons learnt Obtaining information from councils on site locations, can take a long time Council s records at selected sites on maintenance history, traffic counts and age of seals was often very limited While sites nominated were well posted many councils undertook maintenance works without prior advice Historic pavement performance records to compliment monitoring undertaken were of little value due to a lack of adequate maintenance records Many variables to content with in modelling unsealed roads