Finite-element reconstruction of molten-zone dynamics and chemical segregation in optical floating-zone grown Ni--Mn--Ga
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Condensed Matter. The day in condensed matter: materials, phases and the measurements under them.
Kai: I'm Kai.
Mira: And I'm Mira, here with Lev.
Kai: We're looking at the paper: "Finite-element reconstruction of molten-zone dynamics and chemical segregation in optical floating-zone grown Ni--Mn--Ga".
Mira: Here's the gist of the paper. Homogeneous single-crystal growth of Ni–Mn–Ga magnetic shape-memory alloys is limited by incongruent melting, Mn reactivity, and longitudinal composition segregation The coupled finite element method and mass-balance modelling reproduces the molten-zone evolution and composition segregation The calculated solidificationfront geometry is independently supported by EBSD analysis of a quenched growth interface The local solidification velocity differs substantially from the imposed translation rate.
Kai: Growth Methodology and Stabilization The study reports a modified optical floating-zone method where growth is performed inside a transparent sapphire crucible, which stabilizes the molten zone while preserving optical heating and visual monitoring This modification addresses the issue of oxide-crust formation that typically destabilizes conventional crucible-free OFZ growth.
Mira: The experimental setup utilized a four-mirror optical floating-zone furnace equipped with 1000 W halogen lamps and was modified for high-vacuum operation.
Kai: A stable molten zone of approximately 10 mm height was visually established, typically at 42 % of the furnace power.
Title and authors: Kai: So we're looking at this paper today, "Finite-element reconstruction of molten-zone dynamics and chemical segregation in optical floating-zone grown Ni--Mn--Ga". This whole study is about how you grow these magnetic shape memory alloys without messing up the crystal structure.
Mira: Yeah, it tackles the fundamental problem that limits homogeneous single crystal growth: incongruent melting and all that Mn reactivity stuff. It focuses on how those things lead to composition segregation along the length of the ingot.
Lev: And what they do is use a coupled finite element method and mass balance modeling to track exactly how the molten zone evolves and how the chemical makeup segregates as it happens, which sounds way more detailed than just looking at a final sample.
Kai: Exactly. The paper uses this modeling approach to reproduce the evolution of that molten zone geometry, which is usually really hard to see directly during growth. It even independently supports what they calculated by showing that the solidus profile has a convex curvature, matching what EBSD analysis saw on a quenched sample interface.
Mira: That’s key. They show that deviations from the ideal Scheil–Pfann behavior happen because of how the molten zone geometry changes and how solidification conditions evolve throughout the process. It’s not just one fixed condition, right?
Lev: Right, it's dynamic. The paper shows that the local solidification velocity isn't set by the translation speed you impose on your furnace; instead, it changes substantially as that molten zone geometry evolves during growth.
Kai: That means if you keep pushing the furnace at a constant rate, the actual speed of freezing in different spots is fluctuating wildly because of this evolving shape. That’s something we have to account for when we think about any real hardware.
Mira: It suggests that the segregation process isn't happening under one set of fixed conditions; it’s a continuous process driven by those changing local conditions, even if your external setup is constant.
Lev: And to make this accessible, they simplified the finite element model for laboratory scale optical floating-zone growth because a fully coupled fluid flow and species transport model would be way too much overhead right now.
Title and authors: Kai: So what’s the practical implication here? They used this reconstruction to provide a practical means of reconstructing otherwise inaccessible growth conditions, which is super useful when you're trying to optimize something.
Mira: It gives researchers a way to test parameters before they commit to time-consuming experiments, like figuring out the right lamp power or translation speed for better material quality.
Lev: And that reconstruction method can actually be used diagnostically to identify exactly which mechanisms are causing those composition gradients, linking the geometry evolution directly to the segregation physics.
Kai: It’s also a bit of a double-edged sword, though. They mention that their effective partition coefficients are constrained experimentally and only describe the segregation under these specific growth conditions, not equilibrium partitioning on its own.
Mira: Which is fair; they admit those partition coefficients are an intermediate description between simple analytical models and a fully coupled simulation of fluid flow and species transport. They aren't claiming equilibrium partitioning here.
Lev: And they flag a limitation right in the text: this present finite element model is intentionally simplified to be computationally accessible for laboratory scale optical floating-zone growth because explicitly coupling it to full fluid flow calculations would be too much overhead.
Kai: So, while this paper is great for reconstructing what happened in a lab setting, it doesn't fully capture all the physics of the segregation mechanism itself.
Mira: The future work they suggest is to couple this heat transfer model with explicit fluid flow and solute transport calculations so they can treat liquid mixing and interface destabilization directly instead of just absorbing those effects into effective partition coefficients.
Lev: That’s where things get interesting for real hardware; running that full coupled simulation would be the next step, allowing you to see how liquid mixing actually affects instability at higher translation speeds.
Kai: So, looking at the results, they showed that a constant crucible translation rate doesn't correspond to a constant solidification velocity; both the molten zone height and that local solidification velocity vary substantially during growth as the thermal balance evolves.
Title and authors: Mira: That variation is what explains those major departures from ideal Scheil–Pfann behavior we expect in simpler models. The continuous evolution of the solid liquid interface dictates everything about the segregation process.
Lev: And when you look at their comparison plot, Figure six you see how the composition segregation profiles track differently depending on whether you use a weighted average of molten zone height versus velocity compared to just the imposed crucible translation speed.
Kai: It shows that relying only on one parameter—like the fixed translation speed—gives a misleading picture of what's actually happening inside the melt.
Mira: And they confirmed this by showing that their calculated solidus profile has a convex curvature, which matches the interface geometry seen by EBSD, validating that reconstruction.
Lev: That independent experimental check with EBSD is crucial because it confirms the physical validity of the thermal field they reconstructed using COMSOL Multiphysics.
Kai: So, to wrap up this discussion on "Finite-element reconstruction of molten-zone dynamics and chemical segregation in optical floating-zone grown Ni--Mn--Ga", we see that by coupling heat transfer with mass balance, they can reconstruct the evolution of the molten zone geometry during growth.
Mira: This reconstruction is practical because it allows us to understand those complex composition gradients by linking them to the actual geometry changes and local solidification conditions.
Lev: The paper shows that this method provides a basis for reducing trial-and-error optimization, which helps in improving the efficiency of crystal growth experiments without needing endless material consumption.
Kai: And they’ve set up a clear path forward by suggesting future work that couples this heat transfer model with explicit fluid flow and solute transport calculations to capture the full physics of mixing.
Mira: It's a solid piece of modeling for now, and it gives us tools to diagnose the mechanisms causing segregation, even if it doesn't solve the entire underlying transport problem yet.
Lev: For me, from an error-correction perspective, this model shows how much overhead we might have when trying to simulate these coupled systems on real hardware versus what’s actually required for a good laboratory-scale description.
Kai: It’s a valuable tool for understanding the dynamics of growth in sapphire confined optical floating zone setups, even with its current limitations regarding the full fluid physics.
The paper's summary: Kai: It’s about taking what you measure experimentally—the segregation profiles—and using a coupled finite element method and mass balance calculation to reconstruct what was actually happening inside that melt zone while it was growing.
Mira: Right, so they’re not just looking at the final ingot; they are tracking the evolution of the molten zone geometry over time and linking that directly to how much material is melting or solidifying at any given moment.
Lev: I mean, what’s interesting here for hardware simulation is that this reconstruction shows that the local solidification velocity doesn't follow the translation speed you set on your furnace; it changes a lot because the shape of that molten zone keeps evolving.
Kai: That means even if you keep pushing the heater at a constant rate, the actual freezing speed in different spots is fluctuating constantly as things change. It’s not a steady process.
Mira: And they show this dynamic behavior is what causes those big deviations from the ideal Scheil–Pfann behavior we usually expect in these growth processes. It’s all about the changing local conditions of solidification, not just a single fixed set of parameters.
Lev: From an error correction standpoint, this helps us see exactly where the physics breaks down when we simplify things too much. This model provides a bridge between simple analytical segregation models and a full fluid flow simulation.
Kai: It’s practical because they validated their reconstruction against something independent, using EBSD analysis on a quenched sample to confirm the calculated solidus profile has that convex curvature they predicted.
Mira: That's important because it means the geometry they modeled actually matches what was physically observed in the material itself, which gives us confidence in their reconstruction of those inaccessible growth conditions.
Lev: This method lets researchers test parameters—like lamp power or translation speed—before committing to a long experiment, which is a huge efficiency gain when you're dealing with expensive materials.
Kai: So they’re essentially giving us a tool to diagnose those composition gradients by showing that they arise from the combined effects of how the molten zone evolves and how the interface moves.
Mira: It’s not just about getting a final number; it’s about understanding why that number looks the way it does based on the underlying physics of growth.
Lev: What this suggests for future work is to take this heat transfer model and couple it with explicit fluid flow and solute transport calculations so they can actually treat liquid mixing directly, instead of just using some effective partition coefficients.
Kai: So they’ve laid out a path forward where the next step is making the model more computationally intensive but physically complete by including those direct fluid interactions.
Mira: It sets up a clear benchmark for optimizing sapphire confined floating-zone growth because you now have a way to reconstruct what’s happening without needing to perform every single experiment yourself.
Lev: The generic heat-input formulation they used is also handy, meaning this model template can be adapted for other crucible confined zone melting configurations, which expands its utility beyond just Ni--Mn--Ga.
The paper's improvements: Tom: So we're looking at how they suggest improving this model for future work, focusing on what’s missing or where it needs to go next, right?
Kai: They say the current finite element model is intentionally simplified for lab scale optical floating-zone growth because coupling it with explicit fluid flow and solute transport calculations would be too much overhead right now.
Mira: That means they acknowledge that their current version doesn't fully capture liquid mixing or the onset of interface destabilization at higher translation speeds, so that’s where the next step needs to go.
Lev: Exactly; if you want real hardware simulations, you need to move past just using effective partition coefficients and start treating those fluid dynamics directly.
Kai: And they propose that future development should couple this heat transfer model with explicit fluid flow and solute transport calculations, allowing liquid mixing and interface destabilization to be treated without being absorbed into those effective partition coefficients.
Mira: That's a big conceptual leap; moving from an intermediate description to a fully coupled simulation that handles the actual movement of material around the melt front.
Lev: It’s necessary if you want to see how things actually behave when you push the translation speed higher, because those high speeds introduce mixing effects that this current setup just smears into a single partition coefficient.
Kai: So they're aiming for a more complete description where we can study the actual physical mechanisms of segregation, not just an approximation of them.
Mira: And they also mention that the generic heat-input formulation they used makes this model adaptable to other crucible confined zone melting configurations, which opens up its use in other intermetallic systems too.
Lev: It’s a good thing for general research; it means you can use this template to investigate composition segregation from incongruent melting or impurity redistribution in other materials.
Kai: So the implication is that this paper isn't just about Ni--Mn--Ga anymore; it’s about creating a versatile modeling framework for optimizing any zone melting process confined within a sapphire crucible.
Mira: It gives researchers a method for testing parameters before they start time-consuming material-consuming growth experiments, which is huge for efficiency in materials science.
Lev: The model also serves as a diagnostic tool to identify the specific mechanisms responsible for those composition gradients you see in experimental data, linking geometry evolution directly to the segregation physics.
Kai: It’s about turning hard experimental observations into a predictive tool that helps us design better growth conditions without having to try every single combination ourselves.
Conclusion: Kai: So to wrap up, this paper on "Finite-element reconstruction of molten-zone dynamics and chemical segregation in optical floating-zone grown Ni--Mn--Ga" shows how coupling heat transfer with mass balance allows them to reconstruct the evolving molten zone geometry during growth.
Mira: It’s a practical way to understand those complex composition gradients by linking them directly to the actual geometry changes and local solidification conditions, which is really useful for materials science.
Lev: The paper confirms that this reconstruction is physically sound because they validated it against independent EBSD analysis, showing the calculated solidus profile matches what they saw in the quenched sample interface.
Kai: It means we have a method to test parameters before doing time-consuming material-consuming experiments, which is a huge step toward optimizing crystal growth efficiency.
Mira: The implication is that even with simplified modeling, you can get a practical basis for reducing trial and error when working with these complex magnetic alloys.
Lev: For the hardware side, it confirms that this approach is viable for laboratory-scale optical floating-zone growth setups without needing the full overhead of explicit fluid flow calculations right away.
Kai: It’s a solid tool for diagnosing those mechanisms responsible for segregation, showing how they arise from the combined effects of molten zone evolution and interface motion.
Mira: We have to remember that while this is a great reconstruction method, it doesn't fully capture all the physics yet—it stops short of modeling full liquid mixing or solute transport explicitly.
Lev: That’s true; future work needs to couple this heat transfer model with explicit fluid flow and solute transport calculations so they can treat those liquid mixing effects directly instead of just using effective partition coefficients.
Kai: So, while the current version is a powerful diagnostic tool, the goal is to move toward a more physically complete simulation that addresses those remaining transport physics.
Mira: It sets up a clear path forward for modeling growth optimization in these confined systems by enabling testing of parameters prior to long experiments.
Lev: The generic heat-input formulation they used is also handy because it means this model template can be adapted for other crucible confined zone melting configurations, expanding its utility beyond just Ni--Mn--Ga.
Denys Musiienko, Taras Kovaliuk, Petr Harcuba, Oleg Heczko, Ross H. Colman
Department of Magnetic Measurements and Materials, FZU - Institute of Physics of the Czech Academy of Sciences · Department of Condensed Matter Physics, Faculty of Mathematics and Physics, Charles University · Department of Physics of Materials, Faculty of Mathematics and Physics, Charles University
cond-mat.mtrl-sci
Submitted: 2026-11-05
Updated: 2026-11-05
License: http://creativecommons.org/licenses/by/4.0/
Importance score: 79/100
The gist: Homogeneous single-crystal growth of Ni–Mn–Ga magnetic shape-memory alloys is limited by incongruent melting, Mn reactivity…
Key concepts
- Coupled Finite Element and Mass-Balance Modelling
- This involves linking a thermal heat-transfer simulation (finite element) with a mass-balance calculation. This combination allows the model to track how heat moves, how the molten zone changes shape over time, and how elements move within that changing zone. It helps predict chemical segregation based on evolving geometry.
- Molten-Zone Evolution
- This describes how the liquid pool of metal changes size and shape while it is being fed into the solid crystal during growth. The model shows this evolution is complex; the zone doesn't just shrink uniformly. It shifts towards areas with less remaining thermal mass, meaning its shape changes dynamically throughout the process.
- Solidification Front Geometry
- This refers to the precise shape of the boundary where liquid metal turns into solid crystal. The model calculates this geometry and shows it has a convex curvature, which matches what was observed experimentally using EBSD on a quenched sample, confirming the physical validity of the simulation.
- Constant Translation Rate vs. Solidification Velocity
- The machine moves the heating zone at a fixed speed (translation rate), but the actual speed at which metal solidifies locally is not constant. This difference arises because the molten-zone's geometry is constantly changing, meaning local solidification conditions are continuously evolving, which deviates from simple theoretical models.
Terminology
Summary
Homogeneous single-crystal growth of Ni–Mn–Ga magnetic shape-memory alloys is limited by incongruent melting, Mn reactivity, and longitudinal composition segregation The coupled finite element method and mass-balance modelling reproduces the molten-zone evolution and composition segregation The calculated solidificationfront geometry is independently supported by EBSD analysis of a quenched growth interface The local solidification velocity differs substantially from the imposed translation rate. p. 2 Colmanb,∗
aDepartment of Magnetic Measurements and Materials, FZU - Institute of Physics of the Czech Academy of Sciences, Na Slovance
1999/2, Prague, 182 00, Czech Republic
bDepartment of Condensed Matter Physics, Faculty of Mathematics and Physics, Charles University, Ke Karlovu 5, Prague, 121 16, Czech
Republic
cDepartment of Physics of Materials, Faculty of Mathematics and Physics, Charles University, Ke Karlovu 5, Prague, 121 16, Czech Republic
A R T I C L E I N F O
Keywords:
Crystal Growth
Floating zone
Incongruent melting
Finite element modelling
Chemical segregation
Magnetic Shape Memory
A B S T R A C T
Homogeneous single-crystal growth of Ni–Mn–Ga magnetic shape-memory alloys is limited by
incongruent melting, Mn reactivity, and longitudinal composition segregation. We report a modified optical floating-zone method in which growth is performed inside a transparent sapphire
crucible, stabilising the molten zone while preserving optical heating and visual monitoring.Where the paper says it
Growth Methodology and Stabilization
The study reports a modified optical floating-zone method where growth is performed inside a transparent sapphire crucible, which stabilizes the molten zone while preserving optical heating and visual monitoring This modification addresses the issue of oxide-crust formation that typically destabilizes conventional crucible-free OFZ growth. The experimental setup utilized a four-mirror optical floating-zone furnace equipped with 1000 W halogen lamps and was modified for high-vacuum operation. p. 3 Crystal growth was
performed using a four-mirror optical floating-zone furnace (Crystal Systems Corp. FZ-T-4000-VI-VPM-PC) equipped
with 1000 W halogen lamps and modified for high-vacuum operation. elliptical mirrors with focal points F1 and F2. approximately 3 h until a stable molten zone of ≈ 10 mm height was visually established, typically at 42 % of the
furnace power.Where the paper says it
Where the paper says it
Modeling and Reconstruction
The interpretation of measured segregation profiles involved coupling a transient finite-element heat-transfer model in COMSOL Multiphysics with an effective molten-zone mass-balance calculation The model reveals that the local solidification velocity is not fixed by the imposed translation speed but varies strongly as the molten-zone geometry evolves. The coupled model reproduces the dominant features of the measured composition profiles. p. 2 The model reveals that the local solidification velocity is not fixed by the imposed translation
speed but varies strongly as the molten-zone geometry evolves. The calculated solidificationfront geometry is independently supported by EBSD analysis of a quenched growth interface,
while the coupled mass-balance calculation reproduces the dominant features of the measured
composition profiles.Where the paper says it
Key Findings on Dynamics
The analysis demonstrates that deviations from ideal Scheil–Pfann behaviour arise from the evolving molten-zone geometry and solidification conditions. The thermal model predicts significant variation in molten-zone size throughout the process, as illustrated by selected time snapshots in Fig. p. 7 In contrast to the classical Scheil–Pfann description, which assumes a constant molten-zone volume, the present model
predicts significant variation in molten-zone size throughout the process, as illustrated by the selected time snapshots
in Fig. This variation arises naturally from the evolving thermal balance in the finite-length rod geometry. 3, the molten zone
does not remain centred on the optical focal position, 𝑧𝑓
, but becomes increasingly shifted towards the side with the
lower remaining thermal mass. Thus, the computed temperature field reveals not only changes in molten-zone volume,
but also a progressive offset between the molten zone and the nominal heating focus.Where the paper says it
Where the paper says it
Validation and Implications
The model's predictions are validated by independent EBSD analysis of a quenched sample, which confirms that the calculated solidus profile exhibits a convex curvature consistent with the experimentally preserved interface. This reconstruction provides a practical means of reconstructing otherwise inaccessible growth conditions and serves as a basis for optimisation of sapphire-confined floating-zone growth. p. 2 These results show that deviations from ideal Scheil–Pfann behaviour
arise from the evolving molten-zone geometry and solidification conditions. The experimentally
constrained model therefore provides a practical means of reconstructing otherwise inaccessible
growth conditions and a basis for optimisation of sapphire-confined floating-zone growth of
incongruently melting alloys. We complement
this modified growth method with a time-dependent finite element heat-transfer model, coupled with an effective
molten-zone mass-balance calculation. Using the measured post-growth composition profile together with the initial
compositions of the seed crystal and feed rod, the model reconstructs the evolving molten-zone shape and local
solidification velocity during growth, rather than assuming a constant molten-zone length and equating solidification
velocity with the imposed crucible translation rate, as in classic Scheil–Pfann descriptions.Where the paper says it
Where the paper says it
Future Directions
The present FEM model is intentionally simplified to provide a computationally accessible description of the coupled thermal and segregation behaviour during laboratory-scale optical floating-zone growth. Future development should couple the present heat-transfer model to explicit fluid-flow and solute-transport calculations, allowing liquid mixing and the onset of interface destabilisation at higher translation speeds to be treated directly rather than absorbed into effective partition coefficients. p. 9 The present FEM model is intentionally simplified to provide a computationally accessible description of the
coupled thermal and segregation behaviour during laboratory-scale optical floating-zone growth. Most importantly,
fluid flow within the molten zone is not explicitly included, although buoyancy-driven convection, Marangoni flow,
and flow associated with motion of the solid–liquid interfaces may all affect solute redistribution and the degree
of liquid mixing. Sapphire-confined optical floating-zone growth
Stable crucible-free OFZ growth of Ni–Mn–Ga was difficult to maintain because of oxide-crust formation at the
melt surface. This behaviour is attributed to the high reactivity and vapour pressure of Mn near the melting temperature,
producing a Mn-rich vapour that reacts with trace residual oxides released from furnace surfaces. The
resulting crystals show improved growth reproducibility and controlled macroscopic geometry. To interpret the measured segregation profiles, a transient finite-element heat-transfer model
in COMSOL Multiphysics was coupled to an effective molten-zone mass-balance calculation. The
resulting crystals show improved growth reproducibility and controlled macroscopic geometry. To interpret the measured segregation profiles, a transient finite-element heat-transfer model
in COMSOL Multiphysics was coupled to an effective molten-zone mass-balance calculation. FEM Reconstruction of Incongruent Ni–Mn–Ga Single-Crystal Growth
Table 1
Physical parameters used in the Heat Transfer model. Parameter Value Description
𝑃integral 65 (W) Deposited heating power
𝑣vert 5 (mm ⋅ h
−1) Vertical translation speed
𝑇𝑚 1393 (K) Melting point
Δ𝑇𝑚
20 (K) Temperature interval of phase transformation
𝐿 223 (kJ ⋅ kg−1) Latent heat of melting/solidification
𝜌 6900 (kg ⋅ m−3) Material density near phase transformation
ℎconv 20 (W ⋅ m−2
⋅ K−1) Effective convective heat-transfer coefficient
𝜖 0. In the present work, we present the results of growth experiments in a sapphire-confined modified optical
floating-zone geometry, overcoming several limitations of the standard optical floating-zone growth. We complement
this modified growth method with a time-dependent finite element heat-transfer model, coupled with an effective
molten-zone mass-balance calculation. Sapphire-confined optical floating-zone growth
Stable crucible-free OFZ growth of Ni–Mn–Ga was difficult to maintain because of oxide-crust formation at the
melt surface. This behaviour is attributed to the high reactivity and vapour pressure of Mn near the melting temperature,
producing a Mn-rich vapour that reacts with trace residual oxides released from furnace surfaces. Sapphire-confined optical floating-zone growth
Stable crucible-free OFZ growth of Ni–Mn–Ga was difficult to maintain because of oxide-crust formation at the
melt surface. This behaviour is attributed to the high reactivity and vapour pressure of Mn near the melting temperature,
producing a Mn-rich vapour that reacts with trace residual oxides released from furnace surfaces. The final chemical profile therefore
provides one of the few experimentally accessible records of the intrinsic growth history, allowing the model to serve
as a physical reconstruction of otherwise inaccessible thermal and solidification conditions throughout the growth
process. In the present work this model is used diagnostically: it identifies the mechanisms responsible for the observed
composition gradients and shows that they can be explained by the combined effects of molten-zone evolution, interface
motion, and elemental partitioning. An independent experimental check of the reconstructed thermal field, provided
by EBSD analysis of the quenched growth shown in Fig. 5, delineates a convex solidification front between the grown
crystal and quenched melt, demonstrating good agreement with the solidus geometry predicted by the thermal model
and providing additional support that the model captures the principal radial and axial heat-flow conditions governing
the re-solidification.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
REFERENCE
Denys Musiienko, Taras Kovaliuk, Petr Harcuba, Oleg Heczko, Ross H. Colman Finite-element reconstruction of molten-zone dynamics and chemical segregation in optical floating-zone grown Ni–Mn–Ga Denys Musiienkoa,∗, Taras Kovaliukb, Petr Harcubac, Oleg Heczkoa and Ross H. p. 2 Finite-element reconstruction of molten-zone dynamics and
chemical segregation in optical floating-zone grown Ni–Mn–Ga
Denys Musiienkoa,∗
, Taras Kovaliukb
, Petr Harcubac
, Oleg Heczkoa
and Ross H. Colmanb,∗
aDepartment of Magnetic Measurements and Materials, FZU - Institute of Physics of the Czech Academy of Sciences, Na Slovance
1999/2, Prague, 182 00, Czech Republic
bDepartment of Condensed Matter Physics, Faculty of Mathematics and Physics, Charles University, Ke Karlovu 5, Prague, 121 16, Czech
Republic
cDepartment of Physics of Materials, Faculty of Mathematics and Physics, Charles University, Ke Karlovu 5, Prague, 121 16, Czech Republic
A R T I C L E I N F O
Keywords:
Crystal Growth
Floating zone
Incongruent melting
Finite element modelling
Chemical segregation
Magnetic Shape Memory
A B S T R A C T
Homogeneous single-crystal growth of Ni–Mn–Ga magnetic shape-memory alloys is limited by
incongruent melting, Mn reactivity, and longitudinal composition segregation.Where the paper says it
Experimental Setup Details
The alloy preparation involved pre-alloying Ni50Mn28.5Ga21.5 by inductive heating in an alumina crucible within an RF furnace operating at 100 kHz under argon atmosphere. p. 3 98%, etched prior to use by dilute HNO3(aq)
) all
purchased from Alfa Aesar, were weighed to give a nominal alloy composition of Ni50Mn28. Pre-alloying
was performed by inductive heating in an alumina crucible within an RF furnace operating at 100 kHz under argon
atmosphere. Sapphire-confined optical floating-zone growth
Stable crucible-free OFZ growth of Ni–Mn–Ga was difficult to maintain because of oxide-crust formation at the
melt surface. This behaviour is attributed to the high reactivity and vapour pressure of Mn near the melting temperature,
producing a Mn-rich vapour that reacts with trace residual oxides released from furnace surfaces. Crystal growth was
performed using a four-mirror optical floating-zone furnace (Crystal Systems Corp. FZ-T-4000-VI-VPM-PC) equipped
with 1000 W halogen lamps and modified for high-vacuum operation.Where the paper says it
Where the paper says it
Where the paper says it
Model Implementation
A transient heat-transfer model was developed in COMSOL Multiphysics using an axisymmetric geometry reproducing the experimental configuration. Heat transport was described by conduction, with radiative and convective heat losses applied at the external boundaries. p. 6 2 using an axisymmetric geometry reproducing the experimental configuration. Heat transport within the Ni–Mn–
Ga rod was described by conduction, with radiative and convective heat losses applied at the external boundaries. Figure 3: Simulation snapshots at selected time steps, showing the ingot cross-sectional temperature profile, phase change
interfaces, and the focal position of the lamps, 𝑧𝑓
, dot-terminated black line. Melting and solidification were treated using an apparent-heat-capacity formulation incorporating the latent heat
over a finite temperature interval around the melting point.Where the paper says it
Where the paper says it
Compositional Tracking
Mn and Ga were tracked explicitly, while Ni was calculated as the remaining fraction. The effective partition coefficients were constrained experimentally from the sharp composition change observed near the end of the re-solidified bar. p. 8 FEM Reconstruction of Incongruent Ni–Mn–Ga Single-Crystal Growth
The effective partition coefficients were constrained experimentally from the sharp composition change observed
near the end of the re-solidified bar, using the mean compositions measured immediately before and after this transition,
as indicated in Fig. They therefore describe the effective segregation under the present growth conditions rather
than equilibrium partitioning alone. The complete mass-balance formulation, state-variable definitions, time-step
bookkeeping, and numerical implementation are given in Appendix A. The resulting experimentally constrained HT–mass-balance model reproduces the dominant features of the
measured longitudinal segregation profiles in Fig.Where the paper says it
Where the paper says it
Conclusion on Segregation Physics
The effective partition coefficients obtained here are close to those previously derived from Scheil–Gulliver fits to macroscopic segregation in Bridgman-grown Ni–Mn–Ga ingots. These parameters describe the effective segregation under the present growth conditions rather than equilibrium partitioning alone. p. 8 FEM Reconstruction of Incongruent Ni–Mn–Ga Single-Crystal Growth
The effective partition coefficients were constrained experimentally from the sharp composition change observed
near the end of the re-solidified bar, using the mean compositions measured immediately before and after this transition,
as indicated in Fig. They therefore describe the effective segregation under the present growth conditions rather
than equilibrium partitioning alone. Nevertheless, 𝐾∗
Mn and 𝐾∗
Ga should be regarded
as phenomenological parameters incorporating equilibrium partitioning together with deviations arising from finite
liquid mixing, local solidification conditions, and the simplified treatment of solute transport. The coupled model
therefore provides an intermediate description between conventional analytical segregation models and a fully coupled
thermo-fluid and species-transport simulation.Where the paper says it
Where the paper says it
REFERENCE
Denys Musiienko et al. FEM Reconstruction of Incongruent Ni–Mn–Ga Single-Crystal Growth Table 1 Physical parameters used in the Heat Transfer model Parameter Value Description Pintegral 65 (W) Deposited heating power vvert 5 (mm ⋅ h −1 Vertical translation speed Tm 1393 (K) Melting point ΔTm 20 (K) Temperature interval of phase transformation L 223 (kJ ⋅ kg−1 Latent heat of melting/solidification ρ 6900 (kg ⋅ m−3) Material density near phase transformation hconv 20 (W ⋅ m −2 K −1 Effective convective heat-transfer coefficient ε 0. p. 6 FEM Reconstruction of Incongruent Ni–Mn–Ga Single-Crystal Growth
Table 1
Physical parameters used in the Heat Transfer model. Parameter Value Description
𝑃integral 65 (W) Deposited heating power
𝑣vert 5 (mm ⋅ h
−1) Vertical translation speed
𝑇𝑚 1393 (K) Melting point
Δ𝑇𝑚
20 (K) Temperature interval of phase transformation
𝐿 223 (kJ ⋅ kg−1) Latent heat of melting/solidification
𝜌 6900 (kg ⋅ m−3) Material density near phase transformation
ℎconv 20 (W ⋅ m−2
⋅ K−1) Effective convective heat-transfer coefficient
𝜖 0.Where the paper says it
Model Geometry and Dynamics
The transient heat transfer model was developed using an axisymmetric geometry reproducing the experimental configuration. The width parameter was fixed at ω = 0. p. 6 To account for translation of the heating zone,
the profile was evaluated in a coordinate system moving with the focal point, following the conditional boundary
implementation described by Sönnerlind [43]. Its width parameter was fixed at 𝜔 = 0. Its width parameter was fixed at 𝜔 = 0. 016 m (𝜔 = 2𝜎), with
FWHM = √
2 ln 2 𝜔 ≈ 0. 016 m (𝜔 = 2𝜎), with
FWHM = √
2 ln 2 𝜔 ≈ 0. 0188 m corresponding to the effective filament length perpendicular to the growth direction. Full governing equations, boundary conditions, numerical implementation,
and parametrisation are provided in Appendix A. The calculated temperature field provides direct access to the evolution of the molten-zone geometry during growth.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Coupled Mass Balance
The coupled model additionally allows the measured composition profile to be related to the reconstructed moltenzone geometry and local solidification conditions. At each time step, the phase fractions obtained from the heat-transfer calculation were used to determine the volumes of material entering and leaving the molten zone through melting and solidification. p. 7 The effective partition coefficients used in the model are denoted by 𝐾∗
Mn and 𝐾∗
Ga. At each time step, the phase fractions obtained from the heat-transfer calculation were used to determine the
volumes of material entering and leaving the molten zone through melting and solidification. The complete mass-balance formulation, state-variable definitions, time-step
bookkeeping, and numerical implementation are given in Appendix A. The resulting experimentally constrained HT–mass-balance model reproduces the dominant features of the
measured longitudinal segregation profiles in Fig.Where the paper says it
Where the paper says it
Interface Geometry
The spatial transition between the re-solidified crystal and the remaining unremelted feed rod delineates a preserved solidification front, which exhibits a convex profile across the ingot radius. The calculated solidus profile shown in Fig. p. 9 For this particular growth, the deposited heating power and initial focal position were
adjusted to reproduce the measured longitudinal composition profile and experimental growth conditions, while the
interface shape itself was not fitted. The calculated solidus profile shown in Fig. The calculated solidus profile shown in Fig. 5c exhibits a convex curvature
consistent with the experimentally preserved interface. 200 mm, respectively. The corresponding dimensionless convexities,
ℎ∕𝑎, where 𝑎 is the crystal radius, are approximately 0.Where the paper says it
Where the paper says it
Where the paper says it
Growth Rate Comparison
Although the imposed translation speed remained constant at 5 mm/h, both the molten-zone height and the actual solidification velocity varied substantially during growth. Changes in vs reflect the continuous evolution of the solid–liquid interface rather than simply the imposed translation of the heating zone. p. 9 Although the imposed translation speed remained constant at 5 mm/h, both the molten-zone height and the actual
solidification velocity varied substantially during growth. In particular, changes in 𝑣s
reflect the continuous evolution
of the solid–liquid interface rather than simply the imposed translation of the heating zone. In particular, changes in 𝑣s
reflect the continuous evolution
of the solid–liquid interface rather than simply the imposed translation of the heating zone. The segregation process
therefore takes place under continuously changing local solidification conditions, even though the externally imposed
growth parameters remain constant.Where the paper says it
Where the paper says it
REFERENCE
Denys Musiienko et al. FEM Reconstruction of Incongruent Ni–Mn–Ga Single-Crystal Growth Figure 6 Comparison of calculated elemental composition segregation plotted against the crystal length, z, and weighted averages of molten-zone height, hw, and solidification-interface velocity, vs Crucible translation speed is denoted by dashed horizontal line Vertical dot-dashed lines indicate left-to-right: starting point of re-crystallisation solidus position when at the middle-point of the growth solidus coordinate at the point of full melting of the feed rod position are accounted for explicitly by the coupled thermal and mass-balance model
Parameter Identification Process
To identify a physically reasonable parameter set reproducing the experimental composition profile, a rectangular mesh with an element size of 5 ⋅ 10−4 m was used. The manual adjustment process involved adjusting parameters such as lbar to approximately reproduce the position of feature (3), total power Pintegral and lbar to fit the position of kink (2) and the magnitude of the bulge (3), and zinit to fit the coordinate of composition jump (1). p. 9 FEM Reconstruction of Incongruent Ni–Mn–Ga Single-Crystal Growth
The corresponding thermal model independently provides the instantaneous molten-zone geometry and position
of the solidification interface. For this particular growth, the deposited heating power and initial focal position were
adjusted to reproduce the measured longitudinal composition profile and experimental growth conditions, while the
interface shape itself was not fitted. measured composition variation along the growth direction; and e. calculated macroscopic
segregation expected in floating zone (Scheil–Pfann) and Bridgman (Scheil–Gulliver) growth types, discussed in the text.Where the paper says it
Where the paper says it
Limitations and Future Work
The present FEM model is intentionally simplified to provide a computationally accessible description of the coupled thermal and segregation behaviour during laboratory-scale optical floating-zone growth. The effective partition coefficients necessarily represent the remaining unresolved segregation physics, including equilibrium partitioning, finite solute redistribution, incomplete melt mixing, and possible dependence on the local solidification rate. p. 8 The effective partition coefficients obtained here are close to those previously derived from Scheil–Gulliver fits to
macroscopic segregation in Bridgman-grown Ni–Mn–Ga ingots [29]. Nevertheless, 𝐾∗
Mn and 𝐾∗
Ga should be regarded
as phenomenological parameters incorporating equilibrium partitioning together with deviations arising from finite
liquid mixing, local solidification conditions, and the simplified treatment of solute transport. Nevertheless, 𝐾∗
Mn and 𝐾∗
Ga should be regarded
as phenomenological parameters incorporating equilibrium partitioning together with deviations arising from finite
liquid mixing, local solidification conditions, and the simplified treatment of solute transport. The coupled model
therefore provides an intermediate description between conventional analytical segregation models and a fully coupled
thermo-fluid and species-transport simulation. In the present work, we present the results of growth experiments in a sapphire-confined modified optical
floating-zone geometry, overcoming several limitations of the standard optical floating-zone growth. We complement
this modified growth method with a time-dependent finite element heat-transfer model, coupled with an effective
molten-zone mass-balance calculation.Where the paper says it
Where the paper says it
Where the paper says it
REFERENCE
Denys Musiienko et al. FEM Reconstruction of Incongruent Ni–Mn–Ga Single-Crystal Growth 5 Conclusions Sapphire-confined optical floating-zone growth of Ni–Mn–Ga single crystals was demonstrated and combined with a transient finite-element heat-transfer model coupled to an effective molten-zone mass-balance calculation The experimentally constrained model reproduces the principal longitudinal segregation trends and reconstructs the evolution of the molten-zone geometry and solidification conditions during growth Independent comparison with a quenched sample further shows that the calculated solidification-front geometry is consistent with that observed by EBSD, supporting the physical validity of the reconstructed thermal field The EBSD also confirms adoption of the seed crystal orientation utilising this modified FZ growth technique, despite contact of the melt with the single crystalline sapphire crucible We show that a constant crucible translation rate does not correspond to a constant solidification velocity Both the molten-zone height and local solidification velocity vary substantially during growth as the thermal balance evolves, explaining the major departures from ideal Scheil–Pfann behaviour By reconstructing these otherwise inaccessible growth conditions, the model provides a practical basis for reducing trial-and-error optimisation, limiting material losses, and improving the efficiency and reproducibility of crystal growth
REFERENCE
Denys Musiienko et al. FEM Reconstruction of Incongruent Ni–Mn–Ga Single-Crystal Growth Data availability Data and COMSOL model template files supporting the findings of this study will be made publicly available on Zenodo on the date of publication at https://doi. p. 6 FEM Reconstruction of Incongruent Ni–Mn–Ga Single-Crystal Growth
Table 1
Physical parameters used in the Heat Transfer model. Parameter Value Description
𝑃integral 65 (W) Deposited heating power
𝑣vert 5 (mm ⋅ h
−1) Vertical translation speed
𝑇𝑚 1393 (K) Melting point
Δ𝑇𝑚
20 (K) Temperature interval of phase transformation
𝐿 223 (kJ ⋅ kg−1) Latent heat of melting/solidification
𝜌 6900 (kg ⋅ m−3) Material density near phase transformation
ℎconv 20 (W ⋅ m−2
⋅ K−1) Effective convective heat-transfer coefficient
𝜖 0. Growth was then carried out without rotation at a translation rate of 5 mm/h over the full length of the
ingot. Composition and microstructural characterisation
Samples for all characterisation methods were prepared by mechanical grinding and polishing, followed by
chemical etching using a copper-sulfate based nickel-alloy etchant, known in the laboratory as blue poison. Finite-element reconstruction of molten-zone dynamics and
chemical segregation in optical floating-zone grown Ni–Mn–Ga
Denys Musiienkoa,∗
, Taras Kovaliukb
, Petr Harcubac
, Oleg Heczkoa
and Ross H. Colmanb,∗
aDepartment of Magnetic Measurements and Materials, FZU - Institute of Physics of the Czech Academy of Sciences, Na Slovance
1999/2, Prague, 182 00, Czech Republic
bDepartment of Condensed Matter Physics, Faculty of Mathematics and Physics, Charles University, Ke Karlovu 5, Prague, 121 16, Czech
Republic
cDepartment of Physics of Materials, Faculty of Mathematics and Physics, Charles University, Ke Karlovu 5, Prague, 121 16, Czech Republic
A R T I C L E I N F O
Keywords:
Crystal Growth
Floating zone
Incongruent melting
Finite element modelling
Chemical segregation
Magnetic Shape Memory
A B S T R A C T
Homogeneous single-crystal growth of Ni–Mn–Ga magnetic shape-memory alloys is limited by
incongruent melting, Mn reactivity, and longitudinal composition segregation.Where the paper says it
Where the paper says it
Where the paper says it
REFERENCE
Denys Musiienko et al. Appendix A Finite-element model and composition profile fitting The commercially available COMSOL Multiphysics 6. p. 2 The
resulting crystals show improved growth reproducibility and controlled macroscopic geometry. To interpret the measured segregation profiles, a transient finite-element heat-transfer model
in COMSOL Multiphysics was coupled to an effective molten-zone mass-balance calculation.Where the paper says it
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Coupled finite element method and mass-balance modelling can reproduce molten-zone evolution and composition segregation, allowing AI systems to predict chemical profiles during growth based on evolving geometry rather than fixed assumptions.
-
The system can accurately reconstruct otherwise inaccessible growth conditions by using the experimentally constrained model, which provides a
practical means of reconstructing otherwise inaccessible growth conditions.
-
The AI system can perform predictive modeling for material optimization by testing parameters such as
relative lamp power, translation speed, crystal diameter, seed crystal and feed-rod lengths
before conducting time-consuming experiments. -
The system can identify the mechanisms responsible for observed composition gradients by showing that they
can be explained by the combined effects of molten-zone evolution, interface motion, and elemental partitioning.
-
The AI can be used to validate physical models through independent experimental checks; for example, an
independent experimental check of the reconstructed thermal field
using EBSD analysis can confirm the calculated solidification-front geometry.
Abstract
Homogeneous single-crystal growth of Ni--Mn--Ga magnetic shape-memory alloys is limited by incongruent melting, Mn reactivity, and longitudinal composition segregation. We report a modified optical floating-zone method in which growth is performed inside a transparent sapphire crucible, stabilising the molten zone while preserving optical heating and visual monitoring. The resulting crystals show improved growth reproducibility and controlled macroscopic geometry. To interpret the measured segregation profiles, a transient finite-element heat-transfer model in COMSOL Multiphysics was coupled to an effective molten-zone mass-balance calculation. The model reveals that the local solidification velocity is not fixed by the imposed translation speed but varies strongly as the molten-zone geometry evolves. The calculated solidification-front geometry is independently supported by EBSD analysis of a quenched growth interface, while the coupled mass-balance calculation reproduces the dominant features of the measured composition profiles. These results show that deviations from ideal Scheil--Pfann behaviour arise from the evolving molten-zone geometry and solidification conditions. The experimentally constrained model therefore provides a practical means of reconstructing otherwise inaccessible growth conditions and a basis for optimisation of sapphire-confined floating-zone growth of incongruently melting alloys.
Related papers
- Structural contribution to light-induced gap suppression in Ta 2 NiSe 5
- Anatomy of fast current-induced skyrmion motion in synthetic antiferromagnets
- Magnetic control of an exciton-polariton condensate in a van der Waals magnet
- High-pressure elastic properties of GeO2 polymorphs up to 120 GPa
- Decoupling Strain-Rate Sensitivity and Deformation Length Scale Effects in Neutron-Irradiated Tungsten: A Coupled Nano-Indentation, HR-EBSD and Crystal Plasticity Study
- Magnetism modulation in Co3Sn2S2 by current-assisted domain wall motion