aggregates. The size distribution of the aggregates follows a Fuller curve. The material parameters attributed to the aggregates and the two different matrixes under investigation are all given in Table 3. The composite structure to be analysed is shown in Fig. 4 (a) and (b). TABLE 3: MATERIAL PARAMETERS USED IN THE NUMERICAL SIMULATION Materials Aggregates Matrix in NC Matrix in HSC E (GPa) 60.0 20.0 30.0 υ(-) 0.2 0.2 0.2 ft (MPa) 10.0 3.0 8.0 w1 (mm) 0.002 0.02 0.01 σ1 (Mpa) 1.0 0.5 1.0 w2 (mm) 0.008 0.03 0.02 GF (N/m) 14.0 37.5 50.0 (a) (b) Figure 4: Damage (fictitious cracks) and crack formation in numerically simulated concrete: (a) normal concrete and (b) high strength concrete If a moderate force F is applied the system reacts linearly elastic. But after a characteristic load is reached a fictitious crack (damaged zone) will be formed in the weaker matrix. The crack propagation in normal concrete is shown in Fig. 4. As the aggregates are stronger than the matrix cracks will run around the aggregates. Finally a damaged band with a width of about twice the maximum diameter of the aggregates is being formed. It is quite clear that due to the mechanical interaction between aggregates and matrix the fracture energy of the composite material becomes considerably bigger than the fracture energy of both the aggregates and the matrix. If crack propagation in high strength concrete is simulated cracks can run through aggregates as the matrix and the aggregates are mechanically very similar. A typical example is shown in Fig. 4 (b). A comparatively narrow crack band is being formed in high strength concrete. The results shown in Fig. 4 agree well with the photos of crack surfaces as shown in Fig. 2. 4. SHRINKAGE CRACKING Measured shrinkage is in fact the response of a concrete specimen to the strains and stresses generated by the drying process. If the water loss of a specimen with known geometry is measured the hygral diffusion
RkJQdWJsaXNoZXIy MjM0NDE=