## LSU Historical Dissertations and Theses

1994

Dissertation

#### Degree Name

Doctor of Philosophy (PhD)

#### Department

Civil and Environmental Engineering

Experimental investigations and procedures for the determination of damage are presented for the macro- and micro-analysis of a SiC-Titanium Aluminide metal matrix composite (MMC). Uniaxial tension tests are performed on laminate specimens of two different layups. The layups are balanced symmetric and given as (0/90)$\sb{s}$ and ($\pm45)\sb{s}$, each containing four plies. Dogbone shaped flat plate specimens are fabricated from each of the layups. Specimens for the different layups are loaded to various load levels ranging from rupture load down to 70% of the rupture load at room temperature. By loading specimens to various load levels, damage evolution is experimentally evaluated through a quantitative micro-analysis technique. Micro-analysis is performed using scanning electron microscopy (SEM) on three mutually perpendicular representative cross sections of all specimens for the qualitative and quantitative determination of damage. Together these representative cross sections form a representative volume element (RVE) defined for the theoretical development of damage evolution. Results from the micro-analysis are used in evaluating the damage parameters defined in a previously developed damage theory (Voyiadjis and Kattan, 1993c). This theory uses a second order damage tensor of which it is proposed will have non-zero off diagonal terms only if the load applied to the laminate is in a direction other than parallel to the fibers. Damage parameters are evaluated for each of the load levels below the rupture load, since damage features present at this load level are beyond the range of valid damage mechanics. Two types of damage evaluations are performed for the RVE, one for the overall quantification of damage and the other quantifying damage in the matrix and fibers separately. Damage as a result of delamination is excluded in this investigation for both types of damage evaluation, but may be included if needed. The damage curves presented show graphically a consistent evolution of damage, which validates the methods used in evaluating the damage parameter. Curves for the different laminate layups show that damage profile curves are similar for different laminate layups and orientations. The off diagonal terms of the damage tensor as evaluated by the proposed scheme are found to be significant and should be included in any constitutive damage relation. Numerical modeling of these observations is performed using a finite element program that incorporates the proposed damage model.