Blower coupler optimization through flow simulations

The optimization of a blower coupler (connecting piece) was carried out using CFD simulations in order to minimize the pressure drop. Through a combination of generative design and adjoint-based optimization, the flow control was significantly improved. The final component was additively manufactured and experimentally validated.

This article deals with the following topics:

• Problem: High pressure drop between the connections of a blower coupler.
• Solution: Iterative optimization using CFD simulations in ANSYS Fluent.
• Methods: Generative design and adjoint-based optimization.
• Result: Reduced pressure drop, flow-optimized geometry and successful experimental validation.

The blower coupler is a connecting piece for a lung simulator. With this additively manufactured component with three connections, optimum flow control is crucial in order to achieve the desired minimum pressure loss. In this analysis, the focus was on reducing the pressure loss between connections 2 and 3 in order to improve the efficiency of the system. The other connections, such as between connection 1 and connection 2, are not relevant because there will be no flow in this connection due to the design of the overall system. CFD simulations were used to understand, evaluate and optimize different designs.

In the first iteration, an initial design (“Baseline”) was created classically in parametric CAD and evaluated with ANSYS Fluent. In addition, a generative design (“Generative”) was created for comparison. The simulation setup can be seen in Figure 2. A stationary solution is calculated.

Even in this iteration, significant improvements were achieved compared to the initial design. The generatively designed component resulted in better flow control, more even pressure distribution and a reduction in turbulence.

Based on the simulation results of the first iteration and inspired by the generatively created shape, another geometry was created in CAD (“Engineered”). This geometry served as the starting point for a flow optimization using the ANSYS Adjoint Solver (“Optimized”).

Background Adjoint Solver

The ANSYS Adjoint Solver is a powerful optimization tool used in CFD simulations to efficiently identify design improvements. Instead of running multiple simulations with different geometries, the Adjoint Solver uses a mathematical sensitivity analysis to suggest specific changes that improve the desired goal. In this case, the reduction of pressure loss.

The advantages of this method lie in its efficiency and accuracy. By automatically identifying areas with high optimization potential, targeted adjustments can be made without having to perform numerous iterations manually. Using this method, further geometry improvements (“Optimized”) could be achieved and the pressure loss further reduced.

Figure 6 shows the significantly more uniform velocity distribution of the flow. In addition, the maximum velocities are significantly lower in the optimized design. A closer look at the streamlines also shows the even and smooth distribution. The simulated pressure drop (static pressure) could be reduced by approx. 0.22 mbar (see Figure 7), which is more than 50% less pressure loss compared to the “baseline”.

After the flow-optimized adjustment, the geometry had to be reconstructed in order to be able to manufacture a physical component. The mesh generated by the adjoint solver had to be converted back into a CAD geometry. The flow-optimized features were extracted and transferred to the design. ANSYS Discovery provided the necessary tools for this. This step is required in order to smooth the mesh and obtain a less computationally intensive solid. The design of the connections to the rest of the system was carried out in SolidWorks.

A major advantage of additive manufacturing in this context is the ability to realize complex internal geometries that would be difficult or impossible to produce using conventional manufacturing processes. This made it possible to create flow-optimized channels without sharp edges or abrupt changes in cross-section, which enabled a reduction in pressure loss.

The optimized component was manufactured in the internal SLS production and tested experimentally in a test setup. The test setup can be seen in Figure 10. In contrast to the simulation, the pressure loss of the blower is also measured. The measurements confirmed the predicted improvements, although there were slight deviations from the simulation. The reduction in pressure drop between connection 2 and 3 with the “Optimized” design is approx. 0.26 mbar (Figure 11 left). Reasons for deviations could be the different measurement setup compared to the simulation, in which the two blowers are also measured. However, the flow simulations predicted the trend of the experimental results precisely enough. In order to explain the same measured values between “Blower” and “Optimized”, a measurement was taken in the opposite direction between connection 3 and 2. This additional measurement confirms the improvement in pressure drop achieved in the “Optimized” design. In addition, the measured pressure drop for the “Blower” is now measurably lower, as expected.

By combining various tools, including simulations, experimental measurements, generative design and optimization, it was possible to develop a high-performance design that meets the requirements. The iterative approach with CFD simulation and subsequent validation proved to be an effective approach to improving the blower coupler. Additive manufacturing in particular played a decisive role, as it enabled a high degree of design freedom and made complex, flow-optimized geometries feasible without additional manufacturing restrictions.

In the future, further optimization measures could be implemented by using multi-objective optimization methods in order to minimize not only the minimum pressure loss between connections 2 and 3, but also other targets such as the pressure loss in the opposite direction between connections 3 and 2.