Wednesday the 20th - Geosolve Ltd. -
Inventory of pavement case histories where distress is considered due to shear instability of unbound granular basecourse or subbase.

Inventory of pavement case histories where distress is considered due to shear instability of unbound granular basecourse or subbase.

The intention of this project is to establish, promote and expand a database of aggregate parameters so that a prominent distress mode: shear instability of basecourses that comply with the current M/4 :2006 specification, can be progressively reviewed to eliminate this recurring problem.


1 Source data are not yet fully verified and also a wider range of samples must be obtained before adopting the interim recommendation below. The findings here may be coincidental only as there are insufficient examples here of post-compaction gradings from well-performing basecourses.
2 A common feature of a large proportion of cases of basecourse or subbase instability appears to be excessive fines (passing 0.15 mm) plus gap-grading within the sand fraction. The consequence of a sand deficit is the tendency to either have the gravel particles floating in a silt matrix, or if there is no silt to have sparse point to point contacts between gravel particles that will inevitably encourage degradation. Ensuring sufficient sand is therefore essential for shear stability.
3 The data have been analysed using the structural mechanistic-empirical approach, i.e. identifying the functional form and potential primary variables, then applying regression analysis. This indicates (i) the percentage passing 0.15 mm and (ii) gap grading are the key parameters. (iii) The sand equivalent is significant with 35 being the critical value. (iv) After these parameters there is little influence from clay index or plasticity but that may be changed as more case histories become available.
4 The principal measure adopted to quantify gap-grading for this study, is the slope of the grading curve when plotted (log-log), i.e. Talbot's grading exponent "n".
5 Grading exponents of about 0.4 and 0.6 correspond to the fine and coarse limits of M/4 basecourse.
6 The "incremental n" value (ni) has been calculated over each individual sieve interval, i.e. ni=log10(P1/P2)/log10(d1/d2) where P is percentage passing and d is the sieve size. M/4:2006 grading shape control implicitly allows incremental n values as low as 0.26, a concernedly low value.
7 A simple criterion that may be a predictor of virtually all cases of shear instability is where the grading exponent is less than 0.40 over the sand range, i.e. the sieve range of 0.150 to 4.75 mm.
8 Four "governing" sieve ranges for calculation of ni, have been determined by regression. If any two of these are less than 0.4, or are likely to become so after compaction, particular caution appears to be indicated by the lessons of history (1965-2009).
9 A logical regression equation has been explored to see if expected design life ESA for a basecourse can be predicted as a function of grading (governing exponents) and sand equivalent.(Ref: ) This equation must not be used to reject any source at this stage, but shows promise for future use once more case histories are added to the inventory. The equation is based on samples recovered from pavements in a terminally distressed condition. When using stockpile gradings, an additional allowance must be made for degradation during compaction and trafficking. The preliminary equation is presented for use by producers who may wish to consider the likely practical outcome of refining their screening and crushing practices, rather than simply doing the minimum to comply with the pass/fail criteria of the M/4 2006 specification. Ultimately an ESA based rating could lead to more efficient use of resources allowing more marginal basecourses including those that fail sand equivalent, to be used on low volume roads and encouraging the preservation of higher quality product for use on heavily trafficked pavements. If the ESA determined by the regression equation meets the design value, then sand equivalents less than 40 may be justified, if there is a record of good precedent performance.
10 Where gap grading is marginal, additional representative sampling to assess the production variation and range is important, as a relatively small shift in gradation can result in a major change to shear stability performance.

Calculator for shear stabilty criterion

Sieve Size (mm) Percentage Passing (%) Critical Range (mm) Grading Exponent (%) M4-40 Limits M4-40 Shape

Note there is a very fine line between well performing and poorly performing gradings.

Uncertain aggregates may be satisfactory in suitable conditions.

GeoSolve Ltd.       Contact: 03 445 0061       Email: