The semiconductor field has witnessed a dramatic increase in demand for denser computing resources. This requirement will definitely keep getting larger due to the emergence of wide applications for artificial intelligence and machine learning. To satisfy this trend, transistor density must keep getting larger, which requires the technic to manufacture extremely small-scale features precisely and economically. As the feature size gets tiny, the traditional multiple lithography patterning method has reached its physical limit. Introduction of EUV lithography can be a resort however it brings many new issues like the high cost, low part lifetime, special photoresist requirements and stochastic effect, etc. On the other hand, block copolymer (BCP) directed self-assembly (DSA) can be a promising way of generating tiny patterns (tens of nanometer or less feature size) at a low cost due to its multiplication property and not constrained by the physical properties like wavelength. This method can also serve as a crucial process in the whole flow when manufacturing 3D structures on top of some existing layers. The chemical property difference of the bottom layer can serve as a guide to different types of polymers so the alignment process can be neglected. For those reasons, the BCP DSA process was thoroughly studied with the help of coarse-grained molecular dynamic (CGMD) simulation. Having the ability to analyze the location and movement of every molecule, the influence of the chemoepitaxial guiding substrate pinned area (cross-linked PS) and unpinned area (PS/PMMA short brush chains) interfacial properties on the formed pattern both qualitatively and quantitatively has been figured out and the reasons for those effects were explained at the mechanism level. LiNe flow was used for these substrate influence studies. Later, we studied another type of substrate that is only composed of polymer brush short chains. DSA on this type of substrate is worth studying because of its huge potential in the 3D patterning field. This method has proven to be practical and has room to be optimized.