<p>  Use of casting CAE is effective for preventing and countermeasuring casting defects. However, this method is ineffective for predicting defects caused by sand mold characteristics, because casting CAE treats the actual mixture of sand particles and air in the sand mold as one continuum material, thus the correspondence between the essential sand mold characteristics and thermophysical properties of the mold in unclear in numerical analysis. </p><p>  In this study, we investigated a method for predicting penetration defects by mold filling, heat transfer, and solidification analyses using a casting CAE software. First, the differences in sand mold characteristics were identified based on the thermophysical properties of the mold material in casting CAE. Next, the relationship between some cooling curves obtained by numerical analysis and the presence/absence of penetration defects observed in casting experiments was investigated, and the conditions for the occurrence of penetration defects were determined. In the present study, penetration defects occurred at a mold surface temperature of 1310 °C or higher for 5s or longer after mold filling was completed. This condition was verified with eight types of molds and the results showed that the proposed method is effective for determining the existence of penetration defects..</p>