Effective design for almost any kind of product, from consumer goods to industrial equipment, requires taking a large number of factors into account. By making appropriate trade-offs using simulation for digital exploration and optimization, companies can quickly develop efficient and reliable products.
For example, industrial gas turbines burn gas to turn rotors to produce electricity, with substantial amounts of hot exhaust gases as a byproduct. Instead of just warming up the surrounding air, the heat contained in exhaust gases can be put to work by capturing it and letting it flow around tubes containing water, converting the water into steam. The boiler that contains the pipes and the exhaust gases is called a heat recovery steam generator (HRSG). The steam can then flow to a steam turbine to generate more electricity.
The trick in this process is optimizing the flow rate and pressure of the gases in the ductwork at the inlet of the HRSG: The pressure drop must be minimized while maintaining a uniform flow velocity of exhaust gases into the boiler. Computational fluid dynamics (CFD) simulations are often used to design the ductwork to meet these requirements. But engineers at KeelWit Technology SL in Madrid, Spain, took the technology a step further by developing their own proprietary app called the multiobjective structured hybrid direct search (MOST-HDS) shape optimization algorithm using ANSYS ACT. With ACT, you can write APDL scripts, incorporate your company’s proprietary engineering knowledge, add third-party applications to improve your engineering simulation workflows with design exploration and manipulate simulation data.
The MOST-HDS app combined the simulation powers of ANSYS DesignModeler, ANSYS DesignXplorer and ANSYS Fluent CFD to arrive at an optimal inlet duct shape for HRSGs, reducing the pressure drop by 25 percent in the process with no loss in velocity uniformity. KeelWit engineers used DesignModeler to parameterize the wall angles and the inlet length. Next, they used DesignXplorer to create a design of experiments (DOE) with 120 design points and ran CFD simulations at each design point. The MOST-HDS app then took the data from DesignXplorer and plotted pressure drop versus velocity uniformity to generate a Pareto front diagram.
By iterating this process through the 120 design points, KeelWit engineers were able to reduce manufacturing and assembly costs by up to 95,000 euros. Employing an intensive digital exploration process with a self-developed app using ANSYS’ simulation platform was a key to this success.