Computational predictions allied to fast prototyping methods speed design in the fiercely competitive automotive industry. Volkswagen A. G. models unsteady, external-flows around and within its vehicles by the methods of computational fluid dynamics (CFD). Solutions of models constructed on circa 10 million fluid elements consume upwards of 1200 CPU hours. Transient flows in inlet and exhaust manifolds determine the capacity of internal combustion engines to breathe. Power, flexibility, economy and emissions are at risk. Space constraints and the complexities of turbo-charging, emission control devices etc. exercise engine designers. Perforce, they resort to compact and essentially three-dimensional form manifolds. Computationally efficient one-dimensional approaches are defeated. Furthermore, the experimentally determined loss coefficients upon which they depend are inaccessible. This research would reduce the size and complexity of CFD models and speed their computation by analysing engine manifolds into regions where one-dimensional methods suffice to capture the flow and contiguous regions where strong, three-dimensional influences prevail. New methods discovered secure the boundaries between one and three-dimensional regions, having regard to the conservation laws and differing numerical representations of the communicating flows. This hybrid computational scheme and an associated design methodology are tested and demonstrated herein.