In this work, we present a novel method for automated floor plan design in architectural engineering, focusing on the placement of rooms with known dimensions within a rectangular building envelope while adhering to a diverse set of constraints. These constraints include ensuring non-overlapping room placement, providing external openings for rooms, establishing door connections to corridors, and ensuring that corridors lead to external exits. Additionally, compliance with building codes, which restrict the placement of openings on certain sides of the building, is enforced. Special constraints tailored for 3D printing, such as limiting the size of building components, are also considered, along with the flexibility to implement arbitrary design rules. In contrast to previous methods, our method models the floor plan as a discrete grid (raster), with each room converted into a raster representation and placed on this grid. Corridors are treated as unoccupied cells within the grid, connecting the rooms and ensuring access to external exits. Violations of constraints are assigned penalties based on both the nature of the constraint and the extent of the violation. The overall objective is to minimize the total penalty through an optimization approach based on Simulated Annealing. The method demonstrates robust performance, yielding efficient floor plans that comply with a wide range of practical design requirements. This approach opens new possibilities for further exploration in the realm of automated architectural design, particularly in integrating constraints for emerging technologies like 3D printing.