Interpretation of results: soil classification charts
As has already been mentioned, the CPT does not allow sampling, so that the soil type must be determined, either directly, by means of boreholes carried out in parallel, or indirectly, by associating certain trends of the measured quantities with soil characteristics, naturally based on the experience gained in cases where the first-mentioned option was adopted.
When the original version of the apparatus was used (without measuring pore water pressure), charts were developed as shown in Figure 1.18. As can be seen, typically fine soils
Figure 1.18 Simplified soil classification chart from CPT results (Robertson and Campanella, 1983).
tend to exhibit lower values of qc and higher values of the friction ratio. The limitation of this type of chart arises from the fact that it is based on test results that have not reached very great depths, typically less than 30 m. In a very deep hole, for example, in a normally consolidated clay, qc (or q,) will tend to reach relatively high values, for which the application of this chart would be misleading, resulting in an incorrect classification of the soil.
The continuous measurement of the pore pressures during driving, which was made possible by the CPTu, represented, in this regard, remarkable progress. In fact:
- (i) crossing layers of soft or medium clayey soils occurs in undrained conditions, i.e., generating excess pore pressures (positive, in the present case), so that values of pore water pressure generally greater than u0 (the hydrostatic at-rest pore pressure) are measured;
- (ii) on the contrary, crossing sandy layers (with less than 10% fines) typically occurs under drained conditions, with the piezocone recording pore pressure values typically very close to u0;
- (iii) in some overconsolidated and fissured clayey soils or very dense sandy soils, i.e., dilat- ant soils, pore pressures lower than un may be recorded (Tamiolkowski et al., 1985; Mayne et al., 2001).
Figures 1.19 to 1.21 illustrate examples of CPTu results in recent (Holocene) alluvial deposits, involving different soils. As can be seen, the equipment provides nearly continuous logging in depth (generally, at 2-cm intervals) of the four quantities defined above. In each of the figures, the stratigraphic profile of the site was added to the right side, deduced from conventional boreholes.
Figure 1.19 CPTu results in the University of Aveiro Campus, Portugal - soft clays scenario.
Figure 1.19 corresponds to a scenario of about 9 m of soft clayey soils over dense sandy soils and overconsolidated clayey soils. The CPTu results corroborate the above-mentioned trends, in which the very low cone resistance in the soft clays is particularly evident.
On the other hand, Figure 1.20 represents a scenario of essentially sandy soils over 30 m thick. As would be expected, the cone resistance is clearly greater than that shown in the previous figure for the soft clays and tends to increase with depth; below a depth of 20 m, the occurrence of a denser layer is clear. In this sandy deposit, soft clay (mud) intercalations occur, particularly between approximately 11m and 15 m in depth. As can be seen, these interspersed clayey layers imply: i) low values of cone resistance; (ii) positive excess pore pressures; and iii) increases in the friction ratio.
Finally, Figure 1.21 shows a scenario essentially characterized by soft clays, extending from 5 m to about 25 m depth. What is essentially different in this test is the appearance of several (fine) interbedded sands in this thick layer, namely at about 13 m and at 23 m depth. These intercalations are characterized by: i) higher values of cone resistance; (ii) pore pressures close to hydrostatic pressures; and iii) decreases in the friction ratio.
From the previous examples, it can be recognized that the CPTu, by its ability to identify very thin layers within other, larger layers, gains exceptional utility in the context of embankment works on soft, clayey soils, in the correct definition of the hydraulic boundary conditions, fundamental to a realistic approximate prediction of the consolidation time.
The experience with the CPTu allowed the development of much more reliable and complete charts for identification of the soil type and soil behavior than the one in Figure 1.18. Among those, the chart proposed by Robertson (1990) shown in Figure 1.22, which is already considered to be a classic example, is based on the comparison of the normalized
Figure 1.20 CPTu results in Aveiro Harbour, Portugal - loose sandy soils scenario, with interbedded clays and muds.
Figure 1.21 CPTu results in the River Tagus Valley, Portugal - soft clayey soils scenario, with interbedded sandy layers.
Note: OCR corresponds to the overconsolidation ratio of the soil, the ratio of the pre-consolidation effective stress to (ftll
Figure 1.22 Normalised CPT or CPTu soil behavior type chart (Robertson, 1990).
cone penetration resistance, Qt, the normalized friction ratio, FK, and the pore pressure ratio, Bq, which are expressed as follows, respectively:
where и is the instantaneous pore pressure measured by the piezocone.
The left-hand side of the chart represents CPT results while the right-hand side represents CPTu results. As can be seen, on the diagonal area of the left-hand figure are the normally consolidated soils, with the coarser soils (intersected by the cone in drained conditions) above and the finer soils (in undrained conditions) below. Progressive upward displacement of the diagonal corresponds to older and overconsolidated soils, while downward displacement corresponds to increasingly sensitive soils.
Since then, the right-hand chart has been scarcely used, whereas, on the contrary, the left- hand chart has been extensively used, motivating the introduction of a new parameter. In fact, Jefferies and Davies (1993) proposed a curious complement to Robertson’s Q(-F, chart. They noticed that the boundaries between the classes could be fitted by concentric circles, the radius of which was designated the soil behavior type index, Ic.
Robertson and Wride (1998) modified the definition of lc, as defined by:
The boundaries based on lc provide a good approximation to the soil behavior type descriptions, especially in the center of the chart, where normally to lightly overconsolidated soils are located, as shown by Figure 1.23.
Robertson (2009) pointed out that lc = 2.6 may be considered a good proposal for a boundary between sandy (lc<2.6) and clayey soils (Ic > 2.6), based on cyclic liquefaction case histories that essentially involved normally consolidated soils.
Figure 1.23 Contours of soil behavior type index, lc (thick lines), on soil behavior type Q,-Fr Robertson (1990) chart (Robertson, 2009)
-  For this purpose, it is crucial that the filter element is fully pre-saturated so that the pore pressures can actuallybe transmitted to the pore pressure cell.
-  The sensitivity of a clay, S„ is the ratio between the undrained shear strength of the undisturbed soil and that ofthe reconstituted or remolded soil (see equation 1.30).