SANTOS, E. N.; http://lattes.cnpq.br/7673732001137445; SANTOS, Edson Neves dos.
Resumen:
The rock massif on which the Castanhao dam is being built is composed of
Precambrian metamorphics and intrusice rocks. This formation is represented by
three stratigraphic units and they are gneissic-migmatitic complex, Ceara Group
and quartenary sediments.
Geostatistical techniques were utilized to analyze and characterize this
rocky complex from the data obtained from the samples such as recovery, RQD,
and water loss tests. Variogram modeling of these three variables was carried out
considering the values determined through the rotating drill holes. Variographic
analysis showed the presence of a good structure with sample range varying from
35 m for recovery, 60 m for RQD and 170 m for water loss. In the cross validation
of the data it was observed that the estimated mean is very close to sample mean
for all three variables, but the difference between the observed and estimated
variances remained very high. It is worthy to making some observations regarding
the variographic analysis and the homogeneity of the rock massif, in other words
the variograms were elaborated for different directions and angle of tolerances
without any significant changes in the variogram structure. The variogram structure
tend to change when the sample pair distances was varied. This lead to the
conclusion that the rock massif is isotropic when these variables are considered or
the behavior of the variables do not depend on the direction in the study area.
An important point is worth noting about the water loss test or the variable
"vazao". In this work all the data collected were used in the analysis and in some
places, in some bore holes, water absorption was very high. This resulted in the
average water loss apparently very high, when in reality it is not so. The variogram
of this variable did not show a very good structure compared to other two variables
for the reason that the water loss in the rock is not purely a natural phenomenon. It
is possible that a considerable amount of error can be introduced during this test
resulting in unrealistic data
