Compared to conventional intuition-based design, topology optimization (TO) provides considerable mass savings by clearing excess material from lightly loaded regions of a structural part. The remaining material may be distributed in a purely truss-like fashion, or in the form of a closed-walled design consisting of flat plates or curved shells with variable thickness. Unless buckling is of critical concern, closed-walled designs are in general more efficient than trusses which makes them particularly interesting for challenging applications in lightweight design. However, closed-walled designs obtained by topology optimization are still the exception rather than the rule. This paper investigates the applicability of the recently developed selective penalization approach to the design of a motor bracket for an unmanned aerial vehicle (UAV) to deliver defibrillators which is currently being developed by the HORYZN student initiative at the Technical University of Munich. The optimization results are closed-walled designs as desired. A comparison to a truss-like design as well as to a conventional off-the-shelf motor bracket reveals that the closed-walled design even outperforms the topology optimized truss-like design by additional 3% in terms of stiffness-to-weight ratio. Moreover, it provides a streamlined housing protecting the motor cables and contributing to the reduction of aerodynamic drag at cruise speed. Another key finding of this case study is: Depending on the specific optimization problem, and a suitable build orientation provided, closed-walled designs may lower the amount of necessary sacrificial support structures or may even be almost self-supporting. For the closed-walled motor bracket design we found a reduction by more than 25% compared to the truss-like design. This did not require limiting the freedom of design by imposing any additional constraints. The motor bracket was successfully manufactured from aluminium alloy using laser powder bed fusion (LPBF) followed by removal of support structures and CNC machining of functional surfaces.