High-tech machinery increasing demands results in more and more heat output by it’s components. To lower the temperature of these components cooling channels are used. The performance of these cooling channels can be increased by adding flow disrupting structures inside the channel. This study explores the use of density-based topology optimization to optimize the geometry of these structures. A Darcy-Forchheimer penalization method is used combined with a vorticity-based objective to avoid the use of the heat transfer model during optimization. The resulting designs show increased heat transfer as the amount of vorticity increases. However, post-processing results show that overall thermal performance largely related to the pressure drop in the channel rather than detailed geometry. Under these very specific conditions increased flow velocity by narrowing the channel has more effect on thermal performance than disrupting the flow. However, more research is needed making use of a turbulence flow model or different restrictions to the design.