Problems with interpretation of sand state from cone penetration test

Abstract

Considerable attention has been given to the inference of sand density from indirect tests, particularly the cone penetration test (CPT). Data from tests performed in large-scale chambers have demonstrated that for a given sand there is an approximately unique relationship between CPT tip resistance, density (or void ratio) and effective stress level. Data have been presented to support the supposition that CPT tip resistance, if normalized by division by the mean effective stress, is uniquely related to the difference between the current void ratio and the void ratio at the steady state at the same mean stress level. The Paper presents a study of chamber test data for the Ticino sand and demonstrates that there is no such unique relationship for this sand. Use of published correlations that fail to recognize this and other potentially important factors could result in interpretations of sand state that, if carried through to design, could in some circumstances be catastrophic. A series of flow slides that occurred during the construction of an hydraulically placed subsea sand berm at Nerlerk, in the Canadian Beaufort Sea, highlights these difficulties and provides an opportunity to compare field performance with laboratory data. On a beaucoup étudié la détermination de la densité du sable à l'aide d'essais indirects, particulièrement a partir de l'essai de pénétration au cône. Des données obtenues au cours d'essais effectés dans des chambres de grande capacité ont démontré qu'il existe une relation à peu près unique entre la résistance à la pointe du cône, la densité (ou l'indice des vides) et le niveau effectif des contraintes. On a présenté des données d l'appui de l'hypothèse que si la résistance à la pointe est normalisée en la divisant par la contrainte effective moyenne elle est reliée de façon unique à la différence entre l'indice instantané des vides et cet indice pour l'état stationnaire et pour le même niveau moyen des contraintes. Cet article étudie les résultats d'essais effectués dans de telles chambres sur du sable du Tessin et démontre qu'il n'existe aucune telle relation dans le cas de ce sable. L'emploi de corrélations publiées qui ignorent ce facteur entre d'autres aspects d'importance potentielle pourrait conduire à des déterminations de l'état du sable qui seraient catastrophiques pour la construction dans de certaines circonstances. Ces difficultés sont soulignées par une série de glissements par liquéfaction qui eurent lieu lors de la construction d'une risberme mise en place sous la mer par moyen hydraulique dans la Mer de Beaufort (Canada) et qui donnent l'occasion de comparer les performances sur place avec les données de laboratoire.

Evaluation of State Parameter of Crushable Jeju Sand Using Cone Resistance

“…The Q p -w relation appears to be linear for each level of stresses and this observation supports the dependency of the Q p -w relation on the stress level. The slopes (m) of each relationship are observed to be similar, within the range of 6.9-8.2, which is consistent with the result obtained by Sladen (1989).…”

“…This observation conflicts with Sladen (1989), in which the j value of Ticino sand decreases with an increasing stress level. In other words, the Q p -w relation of Jeju sand moves to the right side in the plot when increasing the stress level.…”

“…This induces less influence of the stress variation on the cone resistance of Jeju sand compared with the state parameter and, therefore, an extremely low m value of Jeju sand is observed in Figure 12. Sladen (1989) raised the possibility of the influence of stress level on the Q p -w relation from the interpretation of cone penetration test results of Ticino sand. In addition, his investigation was supported by a numerical examination using the cavity expansion analysis and NorSand model by Shuttle and Jefferies (1998).…”

Drained cavity expansion in sands exhibiting particle crushing

“…Therefore, when using spherical cavity expansion theory to interpret CPT results, the assumed value for G must closely represent the elastic behaviour of the sand. However, the sensitivity of s [13,37]. This dependency is also observed when crushing is included, as shown in Figure 7, where a series of results were generated for different values of p 0 0 in which G/G max was held constant.…”

Interpretation of sand state from cone penetration resistance

“…Collins et al (1992) provided the first realistic model for cavity expansion in soil using a state parameter based numerical analysis. Their results showed that the relation between normalised CPT tip resistance and ł 0 was still affected by the stress level (as suggested by Sladen, 1989) and depended on material properties of the sand. Shuttle & Jefferies (1998) used a general work hardening/softening critical state model to evaluate changes in CPT calibration in terms of critical state properties M, º that can be determined in routine triaxial testing of reconstituted samples.…”