AM shows its mettle in sCO2 CSP application

10 November 2020

SupportFree additive manufacturing allowed the successful build of a mission critical shrouded radial expander wheel (impeller) for use with sCO2 as the working fluid in a concentrating solar power (CSP) application – with an 80% reduction in build time and a 90% reduction in support material

Chad Robertson, senior engineer at Hanwha Power Systems, and his team are developing turbomachinery for a high-efficiency power-generation system utilising supercritical CO2 (sCO2) as the working fluid in a recompression Brayton cycle (RCBC). Heat input to the cycle will be delivered from a concentrating solar power array. The project, in-part supported by the US Department of Energy, has an end-goal of developing equipment for use in CSP plants.

The sCO2 to be used for the CSP project is a fluid state of carbon dioxide in which it is held above its critical pressure and temperature.

At these conditions the fluid is very dense, resulting in compact machinery and optimal thermodynamic cycle conditions, allowing increased thermal-to-electric energy conversion efficiency when compared to steam Rankine cycles.

The overall CSP system works by transferring heat from a large solar array into the working fluid (sCO2), which is channeled through a series of radial expanders (aka impellers/turbine wheels) to extract power. The final expander is connected to a gearbox that drives a generator and various additional compressors needed for completing the power cycle.

“The temperatures and pressures in such a system need to be very high,” says Robertson. “Our goal of optimum efficiency drove us to design a shrouded turbine wheel, or impeller, where the flow path of the working fluid is covered on both top and bottom. This eliminates any gap between the impeller and the housing that would reduce wheel efficiency.”

From traditional to additive

The Hanwha team evaluated several different potential manufacturing techniques for making the new component. “These shrouded impellers are a significant manufacturing challenge, even conventionally,” says Robertson. “The geometry is quite complex, as the enclosed, sweeping blades are a three-dimensional shape that is not easy to define, and the high-temperature nickel alloy that we use is difficult to machine.”

With these constraints in mind, the team reviewed and rejected the use of conventional techniques such as five-axis milling or precision investment casting, identifying roadblocks in terms of cost and accuracy. For example, with traditional manufacturing, the shrouded wheel would have taken multiple steps to manufacture. An open impeller and a shroud would have had to be produced separately and then brazed, with the bonding of the two having the potential for weakness or distortion in the finished piece.

Hanwha decided to explore additive manufacturing (AM, aka 3D printing) as a more direct route to dramatic simplification of the entire process. This technology offered an opportunity to iterate more quickly, refine the design, increase performance and optimise function.

Hanwha had previously worked with Stratasys Direct Manufacturing, an additive manufacturing service bureau/contract manufacturer, on prototype test builds for shrouded impellers. “We were looking for an additive vendor that could provide us with a turnkey part,” says Robertson. “We wanted to supply design specification, materials requirement—and then get back a finished part we could basically put right on our machine.”

Stratasys Direct was up to the challenge. “What enabled us to take on this shrouded turbine wheel project was what we’re calling a next-generation additive manufacturing system,” says Andrew Carter, senior process and manufacturing engineer for Stratasys Direct. “We’ve found that the new VELO3D Sapphire system dramatically improves the process and really stands alone in this next-gen category.”

SupportFree saves material and time

Prior to owning the VELO3D Sapphire, Stratasys Direct wouldn’t have bid on the Hanwha part, Carter says. “Previous projects with other AM-equipment vendors had shown us that the removal and clean-up of all the necessary support structures required for successful prints on their machines was labour-intensive, costly and, in some sections, basically impossible,” he notes.

With the Sapphire metal AM system, however, the need for supports is greatly reduced — if not entirely eliminated — due to the printer’s ability to overcome the “45-degree rule,” which dictates that angles lower than that require additional vertical supports to hold up portions of a part during printing.

By using the VELO3D system to additively manufacture the Hanwha shrouded impeller, Stratasys Direct was able to greatly reduce both the total volume of material used and the surface area for which the system needed to print supports.

The engineers compared a Hanwha component design created with conventional AM support requirements against what would be required by the Sapphire. For the conventional AM printer, they modeled supports for all surfaces less than 45 degrees from horizontal. On the VELO3D printer, they only needed to add supports on surfaces at less than 10 degrees from horizontal. (Stratasys Direct has since then continued improving their process with the Sapphire and can now print all the way down to zero degrees in certain applications without supports.) The difference in the design for Hanwha was drastic—a 90% reduction in support material.

Using less material provides significant savings in a number of ways, notes Carter. “In addition to lower material costs to our customers overall, requiring far fewer supports has eliminated a lot of post-processing work,” he says. “This, in the long run, will contribute to reduced labour time and expense on the shop floor.”

Fine-tuning the AM build

The Sapphire also enabled Stratasys Direct to print a high-temperature nickel alloy (718) for this impeller with extreme accuracy. “Due to the consistency we get from the VELO3D system, we ended up with a near-net shape part on the build plate that required correspondingly less in the way of post-processing,” says Carter.

Build-time reduction was another benefit of using the next-gen AM system. The printer features two 1 kW lasers with full build-plate coverage aligned to less than a 50 μm overlay tolerance. This means that each laser has the capability to reach anywhere on the build plate and deliver a full kilowatt of power (for bulk-metal processing).

The lasers also create a virtually invisible overlap on larger parts like the Hanwha impeller. Combined with the time savings from having much less support material to produce, the dual high-power lasers enable an 80% reduction in overall print time.

Validation of part quality

Ensuring sound mechanical properties of the Hanwha turbine wheel is extremely important for the test programme as it moves forward. The wheel will be rotating at greater than 14 000 rpm during testing and will be in a high temperature environment — so it is critical that the material properties of the turbine are well understood. As part of the project, Stratasys Direct printed test samples and heat-treated them alongside the turbine to measure tensile and stress rupture properties.

ASTM F3055-14 provided a general specification for the additive manufacturing of nickel alloy 718. The measured tensile results all exceeded the ASTM F3055 minimum requirements. The chemical composition of the test samples was also reviewed and met ASTM requirements.

The impeller was also subjected to review using digital X-Ray, CT-scanning, and FPI (fluorescent penetrant inspection). No measurable defects were detected by the scans. The turbine was then balanced and spin-tested at speeds exceeding the design conditions and rechecked for surface cracks using FPI.

“The success of the centrifugal impeller wheel prototypes Stratasys Direct made for us with the Sapphire machine from VELO3D has definitely increased our interest in additive manufacturing,” says Robertson. “It has opened up design freedoms for our team, and sparked a renewed effort to better quantify the material properties and capabilities of additively manufacturing parts. The combination of the state-of-the-art 3D printing and expert project management truly did make the impossible possible.”

The 3D-printed impeller
(Left) Cross section of the impeller. (Right) CAD image of the exterior structure
Sapphire AM machine at Stratasys Direct Manufacturing
Most current 3D printers would need a large quantity of support structures (in red) to produce the impeller
The VELO3D Sapphire printer can manufacture the same design with minimal supports (in blue)
The impeller on a rotor shaft
Tensile sample results

Linkedin Linkedin   
Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.