Gas Turbine Swirlers: Co-Optimizing Despite the Complexity

The Penn State Center for Gas Turbine Research, Education, and Outreach (GTREO) is part of the same university that also hosts the Center for Innovative Materials Processing through Direct Digital Deposition (CIMP-3D). And the simple fact of this co-location on the same campus might prove to have a notable influence on the power efficiency of gas turbines in the future. CIMP-3D is The Pennsylvania State University’s engineering research organization and facility devoted to additive manufacturing, particularly in metal. GTREO director Dr. Jacqueline O’Connor turned to this group to apply laser powder bed fusion toward the redesign of a specific turbine component: the swirler.

The swirler is arguably an underappreciated part of a gas turbine. For those at least somewhat familiar with gas turbine operation, the picture generally goes this far: Hot gas in the turbine generates power as its pressure on blades causes a shaft to spin. But the swirler is a vital component that comes just ahead of this action, and it does a lot. This one piece of hardware mixes air and fuel, introduces turbulence and swirl to the mix, and in this way holds the flame in place that keeps the turbine gas hot.

Problem: The swirler does so much at once that geometry optimization is a difficult problem. To explore the effect of geometry changes, there would need to be a way to make and test many different designs to see what effects design changes have on various parameters of swirler performance. Until now, there has not been an easy way to manufacturer many variations on the design of this complex and expensive component. Until, that is, the arrival of additive manufacturing. The intricate components, conventionally made through precision assembly, can now be 3D printed in one piece — and to geometric forms the assembled versions could not realize. Dr. O’Connor recently turned to Penn State additive manufacturing instructor Dr. Guha Manogharan, and students he teaches in Penn State’s master’s program in additive manufacturing, instructing these students in the basics of swirler function and design. She invited them to engineer their own swirler geometries so that GTREO researchers could begin to develop a record of the performance variations associated with different sets of swirler characteristics.

“In swirler design, you are trying to co-optimize a whole bunch of stuff at the same time,” says Dr. O’Connor in a recent episode of The Cool Parts Show devoted to this work. “You're trying to co-optimize your fuel and air mixing, and your flame stabilization, but you're also trying to optimize things like emissions performance. You're trying to optimize for acoustics, too … and then, there’s an entire structure around this part, so you're also trying to optimize for heat transfer and durability. That’s really where additive manufacturing can make a huge impact: It opens up the design space for this one component that’s trying to satisfy a huge number of constraints.”

Turbulence in the fuel-air flow is particularly meaningful for turbine performance, she says, and she had students focus on this outcome. A turbine burning a lot of fuel will generate a lot of power, while a turbine starved for fuel will generate low emissions. Both outcomes are good — is there a compromise? Potentially yes: Turbulence that optimizes combustion can strike a balance between high power output and low environmental impact.

Student designs were 3D printed in Inconel 718 on a ProX 320 powder bed fusion system from 3D Systems, with a baseline swirler printed on a powder bed fusion machine from Xact Metal. There were constraints bounding the project: “I needed certain things to be the same,” Dr. O’Connor says. “The swirlers had to be the same size because I needed to be able to bolt them into the same [test rig]. And I needed the same number of blades on each because there are certain patterns in the flow that are generated by the blades and I wanted to make sure I wasn’t losing those. And then we had some constraints on the open area. I wanted to make sure that we weren’t closing things down too much. That would change the pressure drop through the system, and it could also dramatically change the acoustics.” Given all that, however, the surface detail remained unconstrained. In addition to changing the shapes of airfoil blades, the students worked with the airfoil surface in their pursuit of turbulence. They turned to biomimicry.

“We asked, what if we were to add additional features on the outside?” says Paula Clares, one of the Penn State AM students involved in the swirler work. “We found this research from Cambridge University in which they talk about the hydrodynamics of adding shark-fin structures to parts.” The surface detail of shark scales enhances turbulence. The students used shark scale models obtained via scanning electron microscope and added these to the surfaces of swirler blades, in search of the same beneficial effect.

Are shark scales the performance-enhancing vital feature that has long been missing from gas turbines? Perhaps yes. More likely, of course, it’s not that simple. At this writing, the physical evaluation of the 3D printed swirlers using the flame test stand within the GTREO Reacting Flow Dynamics Laboratory has not yet been performed. The work is just beginning. Yet just the creation of these novel swirler forms, using surface features not previously practical to attain, represents an important advance — and a hint at what is coming for turbine engineering. Swirlers have been a component essentially too complex to optimize. Interrelated features contribute to interrelated performance outcomes, and exploration of new design has been limited to what designs could be produced. Now, with AM potentially providing a relatively low-cost method to iterate swirler designs and produce them rapidly, the way is clear to discovering just how much the swirler geometry can contribute to what a gas turbine can do.