⚡ electrical engineering

Study of Thermal and Hydrodynamic Behavior of Incone-625 Additively Manufactured Surfaces with Varying Scan Orientations: Surface Topography Analysis, Experimental Investigation with Reproducibility Tests and CFD Verification

This study investigates the impact of laser powder bed fusion scan orientations (0°, 45°, and 90°) on the thermal and hydrodynamic performance of Inconel-625 cooling channels by combining experimental measurements with CFD simulations, revealing that the 90° build orientation yields superior heat transfer and pressure drop characteristics compared to other orientations and a smooth baseline.

Original authors: Kuldeep Mandloi, Harish Cherukuri, Angela Allen, Jimmie Miller

Published 2026-09-01

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Kuldeep Mandloi, Harish Cherukuri, Angela Allen, Jimmie Miller

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the world of modern manufacturing, engineers have long sought a way to build parts with intricate internal cooling channels, much like the complex blood vessels inside a living body. These channels are vital for removing heat from high-performance machinery, such as the engines in jetliners or the components inside powerful computers. Traditionally, making these channels required drilling or milling, processes that struggle with sharp turns and complex shapes. Additive manufacturing, often called 3D printing, offers a solution by building parts layer by layer, allowing for designs that were previously impossible. However, this method leaves the inside of these channels with a unique texture. Unlike the smooth walls of a machine-milled pipe, a 3D-printed channel is lined with microscopic ridges, bumps, and valleys left behind by the laser that melts the metal powder. This roughness is not just a cosmetic issue; it fundamentally changes how water flows through the channel and how effectively it carries heat away. Understanding the relationship between this built-in texture and the cooling performance is the key to unlocking the full potential of 3D-printed cooling systems.

A team of researchers at the University of North Carolina at Charlotte set out to map this relationship, focusing on a specific metal alloy known as Inconel-625, which is prized for its strength in extreme heat. They wanted to know if the direction in which the 3D printer lays down its layers—the scan orientation—could be used as a dial to tune the cooling performance. To find the answer, they did not rely on theory alone. Instead, they built a physical experiment that could be taken apart and reassembled with different metal inserts, allowing them to test multiple surface textures under identical conditions. They also created detailed computer simulations to model the flow of water and the transfer of heat, using real measurements of the metal surfaces to make the virtual models as accurate as possible.

The researchers fabricated small metal blocks using a laser powder bed fusion process, a technique where a high-powered laser melts thin layers of metal powder to fuse them together. They produced samples with three distinct layering patterns: one where the laser tracks ran parallel to the direction of the water flow, one where they were angled at 45 degrees, and one where they ran perpendicular, or across, the flow. To ensure their results were reliable, they printed three copies of each orientation and subjected them to a rigorous heat treatment to remove internal stresses, carefully checking that the surface texture did not change during this process. They also created a control sample with a perfectly smooth surface, machined from a solid bar of the same metal, to serve as a baseline for comparison.

To test these surfaces, the team designed a specialized test rig. Half of the rig was made from a high-temperature resin printed with a stereolithography process, acting as an insulating shell to prevent heat from escaping into the air. The other half was the interchangeable metal section containing the channel. This design allowed them to swap out the different metal inserts without rebuilding the entire machine. Inside the channel, they placed heaters to warm the water and sensors to measure the temperature at various points along the path. They ran water through the channels at different speeds, ranging from a slow, steady stream to a faster, more turbulent rush, while applying a constant amount of heat. By measuring how much the water temperature rose and how much pressure was lost as it pushed through the channel, they could calculate how efficiently each surface removed heat and how much resistance it created.

The results revealed a clear and surprising pattern. The surface with the laser tracks running perpendicular to the water flow, the 90-degree orientation, proved to be the most effective at transferring heat. The water flowing over this rough, cross-hatched texture picked up heat faster than it did over any other surface, including the smooth control sample. However, this superior cooling came with a cost. The same texture that helped grab heat also created the most resistance to the water, requiring significantly more pumping power to move the fluid through the channel. The 45-degree angled surface offered a middle ground, providing better heat transfer than the smooth or parallel surfaces but with less resistance than the perpendicular one. The surface with tracks running parallel to the flow performed the worst, sometimes offering even less cooling efficiency than the smooth metal.

These findings were not just observed in the physical lab; they were mirrored in the computer simulations. The researchers used a modeling technique that treated the complex, jagged 3D-printed surface as if it were covered in a uniform layer of tiny sand grains, a standard way to represent roughness in fluid dynamics. By adjusting the size of these virtual sand grains to match the actual measurements of the metal surfaces, the computer models predicted the same trends seen in the experiments. The simulations confirmed that the orientation of the laser tracks was the dominant factor, outweighing other minor surface features. The study demonstrated that while 3D printing can create surfaces that enhance cooling, it is not a simple matter of "rougher is better." The specific direction of the print matters immensely. A surface that excels at cooling might be too difficult to pump fluid through for a given application, while another might offer a balanced performance.

Ultimately, the work provides a practical guide for engineers designing the next generation of cooled machinery. It shows that by simply changing the angle at which a part is printed, one can tailor the thermal and fluid behavior of the internal channels. The 90-degree orientation delivers maximum heat removal for applications where pumping power is not a constraint, while the 45-degree angle offers a more efficient compromise. The study confirms that the as-built finish of a 3D-printed part is a critical design parameter, not just a byproduct of the manufacturing process. By understanding and controlling these surface textures, engineers can move beyond trial and error, designing cooling systems that are precisely tuned to the needs of high-performance aerospace, medical, and industrial tools.

Technical Summary: Thermal and Hydrodynamic Behavior of Inconel-625 Additively Manufactured Surfaces

Problem Statement
Additive manufacturing (AM), specifically Laser Powder Bed Fusion (LPBF), offers the capability to produce parts with complex internal geometries, such as cooling channels, which are difficult to achieve with conventional methods. However, the as-built surface finish of AM parts is inherently rough and varies significantly based on build parameters, particularly scan orientation. While previous research has established links between AM surface topography and fluid flow or heat transfer, there remains a gap in understanding the specific relationship between as-built surface characteristics (including weld tracks, spatter, and partially melted particles) and thermal-hydrodynamic performance across different scan orientations. Furthermore, many existing studies are limited to specific channel geometries or lack a comprehensive correlation between Computational Fluid Dynamics (CFD) modeling using equivalent roughness and experimental validation. This study aims to investigate the effects of AM roughness characteristics on heat transfer (Nusselt number) and pressure drop across cooling channels for varying scan orientations, validating numerical models with reproducible experimental data.

Methodology
The research employed a dual approach combining CFD simulations and experimental investigation.

Fabrication and Sample Preparation: Inconel-625 surfaces were fabricated using LPBF on an EOS M290 machine with three critical scan orientations: 0°, 45°, and 90° relative to the flow direction. Three samples per orientation were printed to ensure reproducibility. A smooth Inconel-625 part, machined from a forged bar, served as the baseline control. All AM samples underwent heat treatment (900°C for 4 hours, followed by annealing) to relieve residual stresses, with surface topography verified before and after treatment to ensure no significant changes occurred.
Surface Characterization: Areal surface roughness was measured using Coherence Scanning Interferometry (CSI) with a 20X objective. A low-pass Fourier filter (100 µm) was applied to remove form and waviness. The study utilized the Root Mean Square roughness ( $S_q$ ) to define the equivalent sand-grain roughness ( $k_s$ ) for CFD modeling, acknowledging that $S_q$ better captures areal topography features than the arithmetic mean ( $R_a$ ).
Experimental Setup: An innovative, exchangeable experimental setup was designed and fabricated. It consisted of two halves: a stationary section made of high-temperature resin (printed via SLA) acting as insulation, and an exchangeable section made of Inconel-625 containing the cooling channel. This design allowed for the rapid swapping of AM samples with different scan orientations. The setup included converging-diverging sections to minimize flow disturbances, embedded cylindrical heaters for constant heat input, and thermistors for temperature profiling.
CFD Modeling: Simulations were conducted using StarCCM+ with a steady-state conjugate heat transfer model. The k- $\omega$ SST turbulence model was employed to resolve near-wall effects. An equivalent sand-grain roughness model was implemented to correlate the CFD results with the 90° build orientation experimental data.
Validation and Uncertainty: Experiments were conducted across Reynolds numbers ( $Re$ ) ranging from 1,000 to 4,000. Reproducibility tests were performed by dismantling and reassembling the setup with new components. Type-A uncertainty analysis was conducted based on standard deviations from these repeated tests.

Key Contributions

Exchangeable Experimental Platform: The development of a modular experimental setup that facilitates the testing of multiple AM surface orientations under identical flow and thermal conditions, addressing the difficulty of isolating surface effects in complex AM geometries.
CFD-Experimental Correlation: The successful correlation of CFD results using a sand-grain roughness model with experimental data, specifically validating the model for the 90° scan orientation.
Comprehensive Orientation Analysis: A systematic investigation of three critical scan orientations (0°, 45°, 90°) alongside a smooth baseline, providing a comparative analysis of how build direction influences thermal and hydraulic performance.
Roughness Modeling Approach: The application of $S_q$ (RMS roughness) rather than $R_a$ to define equivalent sand-grain roughness, aiming to better represent the complex areal topography of AM surfaces in CFD simulations.

Results

Heat Transfer (Nusselt Number): Significant differences were observed across the different AM surfaces. The surface with a 90° build orientation (transverse tracks relative to flow) demonstrated the highest Nusselt numbers, indicating superior heat transfer performance. The 45° orientation provided intermediate values, while the 0° orientation (parallel tracks) performed poorly, sometimes yielding lower heat transfer coefficients than the smooth baseline.
Pressure Drop and Friction Factor: The surface with the highest heat transfer (90°) also incurred the highest pressure drop and friction factor. The 45° orientation offered a balance, providing improved heat transfer over the 0° and smooth surfaces with a moderate increase in pressure drop compared to the 90° case.
Flow Regime Dependence: Under turbulent flow conditions, the transverse track alignment (90°) showed the highest efficiency. In laminar flow, the differences in Nusselt numbers between the various surfaces were less pronounced.
Surface Topography Impact: The study concluded that the orientation of weld tracks and spatter deposits are the dominant features influencing heat transfer and fluid flow, while the total surface area impact was found to be minimal for both laminar and turbulent conditions.
Validation: Experimental results for Nusselt numbers and pressure drops closely matched the CFD predictions, validating the use of the sand-grain roughness model for this specific application.

Significance and Claims
The paper claims that not all additively manufactured surfaces inherently provide high rates of heat transfer; performance is critically dependent on the scan orientation and flow conditions. The study demonstrates that while a 90° scan orientation maximizes heat transfer, it does so at the cost of significantly higher pumping power requirements due to increased pressure drop. Conversely, a 45° orientation presents an optimal compromise, offering enhanced heat transfer relative to smooth or 0° surfaces without the extreme pressure penalties of the 90° orientation.

The research underscores the necessity of understanding the relationship between as-built surface finish and thermal-hydrodynamic performance for the successful industrial adoption of AM in applications requiring complex cooling channels, such as aerospace and electronic cooling. By providing validated experimental data and a robust CFD correlation method, the study offers a framework for optimizing AM build parameters to achieve desired thermal performance without incurring excessive hydraulic losses.

More like this