Challenges Optimizing Turbocharger for Multiple Criteria

A big challenge in the mechanical design world is the need to optimize to multiple criteria. When dealing with a turbocharger, for instance, designers will need to optimize for various constraints like production expense, heat, vibration, durability, fuel efficiency, size, and more.

Simulation of the static pressure rise in the compressor stage of a turbocharger (Courtesy of Cummins Turbo Technologies).

“Many of these constraints will also work against each other,” said Brad Hutchinson, Global Industry Director at ANSYS.  “For instance, in the auto industry smaller turbochargers are needed. However, a small impeller must spin faster to produce the desired pressure. This affects the durability and aerodynamics of the design.”

Turbochargers are used in many applications and industries, so the optimization criteria for each model can vary significantly. Therefore, engineers are not able to make a one size fits all solution.

As the development times for these industries continue to shrink, engineers need tools that ensure that their designs will work well on the first prototype. This is where simulation technology like ANSYS’ suite of multiphysics simulation is beneficial.

Optimizing a Turbocharger Design with Simulation

“In the past,” he added, “you needed to over design the turbocharger for stress which would mean it was under designed for aerodynamics and performance. With simulation you can optimize the turbocharger design to meet the criteria and targets.”

The backbone of Hutchinson’s simulation solution for turbochargers is ANSYS workbench. All the modeling and simulation tools communicate through the Workbench platform.

The starting point for the design is an experience-based 1D tool such as Vista-CCD.  From there, Hutchinson suggested to use a simplified model and simulation solver to speed up the initial design space exploration and optimizations. In the preliminary stages, engineers can first use 2D simulations to optimize the blade designs to meet the aerodynamics criteria. Next, they can move onto other physics like stress analysis, vibrations, and more complex models.

“When dropping from 3D to 2D you gain a lot of speed,” said Hutchinson. “An hour long simulation can be reduced to seconds. This is great when you are exploring the design space and performing initial optimizations. The 2D solver has limitations in that it cannot provide a good estimate of losses. This will need a full 3D simulation. But for the preliminary designs you can implement a lot of improvements with a small amount of computational power with a simplified approach.”

He added, “This improved speed you get with 2D fluid simulations will also make for faster multiphysics exploration. When you hook in ANSYS Mechanical structural simulations within Workbench you will appreciate the drop from 3D fluid simulation to 2D.”

As the engineer expands the complexity of the simulations and model, they can continuously optimize the designs along the way. The goal is to develop a well advanced model (impeller-only or full assembly with impeller + volute) before 3D simulation optimizations begins. Therefore, the more time consuming and complex simulations will be used on a much reduced design space to test the design across its full operating map.

“By bringing optimizations in at every step of the design cycle, engineers can make changes earlier in the design when it is most economical to do so,” said Hutchinson. “Using quick 2D simulations gives enough fidelity to narrow down the design space and parameters.”

Turbocharger Simulation, Modeling & Optimization Software

However, Hutchinson doesn’t agree. He said, “Software empowers but it doesn’t replace a smart engineer. We need engineers to use their experiences to guide the simulation and decision processes. Simulation provides the power and efficient tools to enable designers to come up with optimal designs. But you still need to know what you are doing.”

As such, an engineer must have an idea of what makes a good design. During the optimization and design space exploration process, the software will change the geometry and assess how the changes have affected the performance. As a result, engineers must input parameters and their ranges to control the optimization process.

During the design space exploration, engineers will also be looking for which parameters will have little effect on the results. These parameters are determined near the start of the design process, in order to focus only on those parameters that are truly significant to the design.

If you are looking into using simulation and optimization to help your turbocharger designs, ANSYS provides a significant suite of software to meet your needs. Here is a short list of applications an engineer might find useful:

• Vista CCD: To create a starting-point design based on operating requirements
• BladeModeler: to create the parametric impeller geometry and control the parameters
• Turbogrid: to mesh the impeller geometry
• Vista TF: Through-flow solver for 2D fluid flow simulations for use during initial optimizations
• DesignXplorer: To set up the design space exploration and optimization algorithms
• ANSYS Meshing: To mesh all non-bladed components, such as the volute.
• ANSYS CFX: for complete CFD simulations during final optimizations and verifications
• ANSYS Mechanical: for stress, deformation, vibration and modal analysis

For a more detailed look into how this software suite can be used in a workflow to optimize turbocharger designs, watch the video below:

Ansys has sponsored this post. They have no editorial input to this post - all opinions are mine. Shawn Wasserman