A hydraulic collapse mechanism was incorporated into a recirculating pulsatile flow system to simulate the physiologic problem cf coronary artery vasospasm. A dimensional analysis of the hemodynamic problem provided the basis for i) the specifications for elastic test sections (both straight and branching), ii) the determination of the flow modelling parameters, Reynolds number and unsteady Reynolds number, and iii) the determination of the dimensionless collapse parameters. The models were collapsed in a controlled manner while changes in volumetric flow rate into and out of the models as well as axial pressure drop were monitored. It was found that the driven collapse of the vessel acts as a pump, the effectiveness of which is dependent on upstream and downstream resistance. There was noted a difference in the volumetric flow curves representing fluid leaving the pre- and the post-collapse models under the same inflow conditions. This was due to both the elastic properties of the models and to the post-collapse shape of the models (curved walls and non-circular cross-section). Time-exposed photographs of tracer particle displacements within the model indicate increased volumetric flow in each branch during the initial phase of the collapse process. Moreover, it was seen that the radial gradient of the axial velocity at each wall surface varied in magnitude (and possibly in sign) during the collapse. The in vitro results do not substantiate the coronary spasm/myocardial ischemia connection, but do further implicate vasospasm as a factor in atherogenesis. / Ph. D.