This work is situated within the scope of the emerging complexity sciences and tries to bridge scientific and artistic intentions for understanding complexity by doing experiments. To this end, we propose the use of simple, "minimal" models coupled with expression functions that translate model results to visual elements or properties. Number series appear to be good candidates for this type of experiment, because they are simple, inflexible and infinite. Expression functions on the other hand may be arbitrary and subjective. Our case study is based on the well-known Fibonacci series and shows that, by constraining some aspect of the visual form, an expression function may translate the number series to a complex visual form. Various expression functions are investigated and their results exemplified by selected images. Several dimensions of future work are also briefly outlined.