Contribution à la théorie macrostructurale de la résistance d'un béton à la rupture par compression simple

Résumé

L'auteur présente une hypothèse de résistance à la rupture des bétons, fondée sur une forme possible de rupture par compression simple avec élimination du frettage aux appuis et sur la mise en évidence de trois paramètres principaux: résistance au fendage du granulat, résistance à la traction du mortier ou de la pâte de ciment durcie, adhérence entre granulat et pâte de ciment ou mortier.

L'auteur tient compte des contraintes structurales, dues au retrait d'un béton à base de ciment portland et de la relaxation de ces contraintes dans le temps.

L'hypothèse avancée est confirmée expérimentalement par des résultats d'essais faits sur des bétons à base de concassés obtenus à partir de plusieurs variétés de roches.

Summary

A macrostructural hypothesis as to the rupture strength of concretes subjected to a simple compression test is presented, based on a possible form of rupture at the limit state, obtained by eliminating the surface friction between the plates of the press and the test piece bases. The macrostructural hypothesis has recourse to three fundamental parameters: the splitting strength of the aggregate (R ^g_f ), the tensile strength of the mortar or the hardened cement paste (R_t) and the adhesion between the aggregate and the hardened cement paste or the mortar (R ^m_coh ). Account is taken of the shrinkage stresses of the Portland cement concrete (R_str). The hypothesis is expressed in the form of a linear relation between the fundamental parameters to which certain coefficients A, B and C are given; these coefficients express the possible number of rupture planes at the limit state which appear when there is a splitting of the aggregates (A), decohesion (loss of adhesion) (B), and tension of the hardened cement paste or the mortar (C); this relation is written: R_c=A R ^g_f +B R ^m_coh +C R_t+R_str. Taking experimental data relating to two values of the water/cement ratio as a starting point, an estimation of the degree of relaxation of the structural stresses in relation to time was obtained; a relation expressing the relaxation of the stresses in terms of time is proposed.

Experimental results are given showing the part played by the friction of the plates of the testing machine in the estimation of the strength of a concrete at rupture by pure compression. A linear relation is established between the real strength of a concrete subjected to a simple compression and the dimensions of the test piece used in the test, leaving the binding of the bases out of consideration. With equal dimensions of the aggregates the strength of the concrete increases with the dimension of the test piece, and not the reverse as is generally assumed.

The hypothesis is verified experimentally by results of compression tests carried out on concretes having a base of crushed aggregates obtained from several varieties of rocks, the friction on bearing faces being eliminated.