This paper was produced for the 2019 NAFEMS World Congress in Quebec Canada
In the pump industry, a segment of customers is requesting a high-pressure multistage pump that can operate with elevated fluid temperatures. At the pump and pump solutions provider Grundfos the solution to this request is called an air cooled top (ACT). The working principle of the ACT is a cavity of air that insulates and provide cooling to the fluid around the shaft seal of the multistage pump.
To increase the value offering in this customer segment, it is investigated how to improve the cooling rate of the current ACT solution. The immediate solution is to increase the volume of air to increase the insulation effect. However, this would increase the footprint or height of the pump, which is not considered a valid solution in the industry. The alternative approach is to develop a redesign within the same box of dimensions as the current solution based on heat transfer engineering.
Initially, a numerical model of the multistage pump is developed to simulate the heat transfer during operation. The heat transfer simulations are conducted in commercial code. Subsequently, the numerical model is calibrated with test results, hence, a sequence of experimental testing is conducted. The temperature is measured in the fluid, in the air-filled cavity, at contact surfaces, and it is measured ambiently in the test rig. The purpose is to calibrate the parameters of the numerical model, and the goal is to accurately predict the steady-state temperatures without coupling the simulations with CFD to reduce the computational cost.
A response surface is generated, and parametric optimization is performed to complete the calibration of the numerical model. This is followed by a number of redesign iterations evaluated with the numerical model. The final iteration is a topology optimization with thermal and structural compliance objectives. Since the ACT is also a load carrying component, that is an integral part of the high-pressure pump, multiple objectives are included in the topology optimization. Different challenges have been encountered when conducting this coupling because the problem is contradicting, thus the redesign is dependent on the relative weighting of the two objectives.
Additive manufacturing constraints have been applied in the topology optimization setup, since the intention is to produce the final redesign by metal additive manufacturing - specifically laser powder bed fusion. Once the topology optimization is done, the AM Lab at Grundfos has made a functional prototype of the redesigned ACT. Subsequent experimental testing with the prototype has confirmed a superior performance compared with the original design. This is a validation of the approach of design for additive manufacturing with simulation driven development of heat transfer critical applications at Grundfos.