Plate fin heat exchanger design optimization and mass-specific performance assessment for aerospace applications

Efficient heat exchanger design is crucial for cost, sustainability, and energy conservation. Due to its importance, the research interest in optimal HX design has increased across various industries, especially in the aerospace sector. Here, compact heat exchangers with a heat transfer area compactness β > 400 are of high interest, mainly the plate-fin ones. These topologies are characterized by vertically stacked plates that separate the two fluids, and several ducts in which many different fin geometries can be employed. Selecting the right fin type and its geometric parameters to use in each fluid side is a fundamental and complicated step. Although optimal HX selection and design have been relying on the expertise and past choices of experts, over the years, several methods have been proposed to compare the performance of different surface geometries. These performance valuation methods aim to help the designer make informed decisions about the optimal fin topology. In the aerospace sector, where the HX mass plays a key role, few design guidelines on optimal HX selection and design are present. For this reason, this work analyzes the mass-specific performance of optimally designed compact plate-fin HXs and defines general guidelines for the assessment of different optimal fin topologies. Two models of plate-fin HXs in crossflow arrangement compatible with several fin geometries have been developed. The sizing determines the dimensions and pressure drops ΔP of the HX given the inlet conditions on both sides and the heat duty. While the rating determines the heat duty and pressure drops of a given HX. The models have been validated by comparing their predictions with those of test cases from the scientific literature. In order to obtain optimized maps of the HXs, four optimization algorithms have been implemented: the Particle Swarm Optimization (PSO) and the Differential Evolutionary Algorithm (DEA) are single-objective, while the NSGA-II and the NSGA-III are multi-objective algorithms. The objective function for the single objective algorithms is the exchanger mass, which needs to be minimized, while the objective function for the other two is to minimize the exchanger mass and the pressure drop on either just the cold side, or both. The single objective optimizations have been used to generate optimized design maps in two different ways: first by varying the heat duty Q˙ and keeping constant all the constraints, and then by varying the constraint of the ΔP on the hot side at constant Q˙. An air to air balanced HX (C∗=1) is chosen as a first case study. The optimization has been performed with three fin types, in order to compare their performances. The three fin surfaces adopted are the same for both fluid sides: the offset-strip fins, the louvered fins, and the triangular wavy fins. The mass-specific power MSP=Q˙/Mhex, the ΔP, and the two Bejan numbers are chosen as performance metrics. The MSP showed no clear trend with ΔP, but a trend with Bej was found. By decreasing the heat duty, the HX mass and the Bej both decrease following a 1/x^n trend. On the other hand, increasing the constraints on the ΔPh, at constant heat duty, allows us to define some working ranges. The multi-objective optimization showed that by increasing the MSP, the ΔPh increases as well, while, the ΔPc does not follow a clear trend, and the Bejc decreases. Finally, the three-objective optimization showed that, as expected, by increasing the allowed ΔP the designer can achieve significantly lower masses.