Neste trabalho abordamos a fragmenta¸c˜ao de um disco fino e quebradi¸co (quase bidimensional) ocorrida pelo impacto pontual de um proj´etil. Este problema ´e tratado via estat´ıstica geom´etrica, onde linhas e c´ırculos desempenhar˜ao o papel das rachaduras e ser˜ao distribu´ıdas aleatoriamente sobre a superf´ıcie. A escolha das rachaduras ser´a feita com a distribui¸c˜ao de probabilidade respectiva a cada ente geom´etrico: a linha ou o c´ırculo. Estas distribui¸c˜oes foram obtidas a partir de uma an´alise da propaga¸c˜ao e, principalmente, da dissipa¸c˜ao da energia cedida pelo proj´etil. Ao construirmos o disco fragmentado, calculamos a ´area de cada fragmento e, ent˜ao, chegamos `a distribui¸c˜ao cumulativa das massas dos fragmentos maiores ou iguais a m, P(m). Observamos que esta ´ultima est´a em grande concordˆancia com os dados experimentais obtidos por T. Kadono et al, ou seja: P(m)∼ m−β, onde 0.1 < β < 0.3, no regime de pequenas massas.CNPqIn this dissertation we address the fragmentation of a flat (two-dimensional) brittle disk occurred due to the punctual impact of a projectile. This problem is addressed via geometric statistics, where lines and circles represent the disk’s cracks, randomly distributed over its surface. The distribution of cracks will be made through each geometric entity’s respective probability distribution: the line or circle. These distributions were obtained from a propagation and dissipation analysis of energy transferred by the projectile. To construct the fragmented disk, calculations were made to obtain the area of each fragment and to plot the cumulative distribution of the fragments masses (greater than or equal to m). It was noted that the latter are in accordance with the experimental data obtained by T. Kadono et al, ie. P(m)∼ m−β, where 0.1 < β < 0.3, in the small mass regime.