Rheocasting Fins under the Microscope

UBE Machinery analyzed thin-walled heat-sink fin sections produced by Comptech’s Rheocasting process. The geometry was aggressive by die-casting standards, ~105 mm fin height, ~0.8 – 1.0 mm tip width, with a 0.5 – 1° draft angle, precisely the kind of features that challenge conventional liquid die casting.

The alloy listed was EN42000 (A356-equivalent) made from 100% recycled feedstock (approx. Si 6.95 %, Fe 0.3 %) and, notably, no Sr or other grain-refining modifiers were added.

The microstructure UBE observed

At low magnification (~×100), the fin cross-sections, from the ~3 mm thick base out to the ~1 mm-thick tip, showed a uniform, well-packed structure composed of:

Primary, near-spherical α-Al particles (~50 µm)
Extremely fine α-Al crystallized between these spheroids
Eutectic Si occupies the remaining interstices

Minor air entrapment and occasional shrinkage cavities were present, but the metal matrix was evenly distributed without segregation bands, leading the analysts to expect stable, repeatable quality across parts.

At high magnification (up to ×5000), UBE reported that the “extremely fine α-phase” appearing between the spheroids is in fact secondary-crystallized, petal-like dendrites only a few microns across, while the eutectic Si adopts an exceptionally fine granular/fibrous morphology both signatures of a rapidly formed semi-solid feed that solidified with suppressed coarsening and limited segregation. Some Al-Fe-Si intermetallics were visible, as expected for this chemistry, but not in networks that would mechanically or thermally dominate the section.

Why that matters (in UBE’s words)

In many semi-solid processes, the low-melting liquid tends to rush ahead, and eutectic Si segregates into thin sections, exactly where designers can least afford brittleness or porosity. In the studied Rheocasting samples, segregation is suppressed. UBE hypothesizes a high density of crystal nuclei in the ladle/slurry that continue nucleating in the sleeve and during mold filling, yielding the observed fine, uniform microstructure along the fin height. Their practical call-out, even as-cast, such a structure could significantly improve tensile elongation versus typical die-cast material, encouraging for thin fins, leak-tight bases, and even large structural die castings, like Gigacastings.

The report’s summary bullets translate into a clear manufacturing thesis:

Homogeneous “tissue” (matrix) filling lowers part-to-part variability and supports higher production efficiency
A large number of nuclei in the slurry helps inhibit grain growth and sustain the micro-fineness across long flow paths
Low-Si casting is feasible without resorting to post-cast heat treatments that can distort thin features, which is important for leak-tightness and dimensional control
The process was demonstrated with 100% recycled feedstock, pointing to CO₂ and cost benefits
Long flow-length filling is attainable, and, as the UBE team notes, applying the method to very large Gigacasting body/chassis components would benefit from sufficient injection energy/locking force, or conversely allows smaller projected areas for a given press

Linking the microstructure to heat-sink performance

Every foreign atom, every grain boundary, and every second phase scatters electrons and phonons, cutting thermal conductivity (k); porosity and leaks are catastrophic in liquid-cooled applications. Standard HPDC Al-Si alloys (≈7–12 % Si) are castable but typically plateau around ~100–140 W/m·K in real parts, while low-Si Rheocasting alloys have demonstrated ~170–195 W/m·K in practice. This step-change directly reduces base-plate conduction resistance in cold plates.

Where this points next

For designers chasing higher heat flux, think direct-to-chip cold plates, immersed fin cores, or high-aspect-ratio air fins. The UBE findings support three practical moves:

Lean on Rheocasting for thin-wall reliability: the documented suppression of eutectic-Si segregation in thin sections is a credible path to as-cast elongation and reduced crack/leak risk in long, narrow passages.
Use lower-Si chemistries where thermal conductivity matters: with the semi-solid route keeping porosity in check, you don’t need to trade castability for conductivity; you gain both. This shift lifts the thermal conductivity into the 180–195 W/m·K range that many high-power modules require.
Exploit recycled feedstock: the UBE samples show that 100% recycled input can still deliver fine, uniform microstructure, opening cost, and sustainability wins without sacrificing performance.

Disclosure and citation

Again, the microstructural observations and conclusions summarized above are drawn from UBE Machinery’s report “2025.06.04 Microstructural Analysis — UBE-Vitto-Comptech,” based on Rheocasting samples provided by Comptech via Vitto.

This article presents those third-party findings and offers design/manufacturing implications. Any errors of interpretation are mine.