There are different types of protective clothing available to protect human body from different external conditions such as rain, dust, direct sun radiation, insect access and their bites. The problem of overheating of the body may arise when such proactive clothing is required to wear in warm environment or heavy workload conditions. This is because the outer layer of cloth lacks sufficient air permeability, which causes the accumulation of warm and moist air at the body and causes discomfort. To enhance air exchange, various closable vents and open spaces of clothing have been developed. However, this only results in a partial improvement in air exchange, and reducing mechanical strength of the clothing. The mechanical strength of the clothing can be increased by attaching appropriate ventilation elements at the inner side of ventilation holes, which may permit proper air exchange as well as restrict direct access of insects to the body. The design of elements involving fluid flows usually is based on time-consuming Computational Fluid Dynamics (CFD) simulations. In this paper metamodeling approach different order polynomial local and global as well as kriging approximations are compared for shape optimization purposes of ventilation elements. The main goal is to identify the geometrical shape of the element that causes the least amount of flow energy losses along the cell flow channel, which can also be known from pressure difference. For this a multistep procedure was realized to achieve the best results. 1) Planning the position of control points of Non-Uniform Rational B-Splines (NURBS) for obtaining elements with a smooth shape. 2) Building geometrical models using Computer Aided Design (CAD) software SolidWorks in conformity with the design of the experiment. 3) Calculation of responses for a complete model using Computer Aided Engineering (CAE) software SolidWorks Flow Simulation. 4) Building metamodels for responses based on the computer experiment. 5) Using metamodels for shape optimization. 6) Validating the optimal design using CAE software for the complete model.