ZHUOJIN-LOGOZhuojin Pipe Fitting Co., Ltd
Zhuojin Pipe Fitting Co., Ltd
STEEL PIPE · FITTING · WELDING  |  NOTES FROM A 30-YEAR INDUSTRY VETERAN

Solidification Characteristics and Microstructure Formation in Focused Beam Powder Surfacing

Literature Overview and Core Findings

This work by Shan Jiguo, Li Hui, and Ren Jialie from Tsinghua University's Department of Mechanical Engineering, published in Acta Metallurgica Sinica in 2002, addresses a question that has troubled surfacing engineers for decades: what actually happens inside the melt pool when you apply a concentrated energy beam with powder addition. The authors systematically compared two powder feeding methods — pre-coated powder surfacing and powder feeding surfacing — using optical microscopy, scanning electron microscopy, energy dispersive spectroscopy, and X-ray diffraction. The study was supported by the National Natural Science Foundation of China and Tsinghua University's 985 Fundamental Research Fund, which tells you this was fundamental work aimed at understanding process physics rather than just optimizing a single application.

The key experimental finding is that high heat input beam surfacing layers consist of two distinct zones: a bottom region with epitaxial growth characteristics of γ(Fe,Ni) planar crystals, and an upper region with γ(Fe,Ni) dendritic crystals that grow preferentially against the heat flow direction. What is particularly interesting is the observation that the angle between the dendrites in the surface layer and the beam scanning direction is smaller than the angle between the dendrites at the bottom and the scanning direction. This is a direct consequence of the changing thermal gradient as the melt pool solidifies — the heat flow direction rotates as the pool deepens and the substrate conducts heat away.

Technical Interpretation of Solidification Behavior

Epitaxial Growth and Dendrite Orientation

The epitaxial growth at the pool bottom is the classic result of partial melting of the substrate grains. When the focused beam strikes the surface, the substrate grains at the fusion line act as nucleation sites, and the new grains inherit the crystallographic orientation of the parent material. This is critical for practical applications because it means the bond between the surfacing layer and the substrate is metallurgically sound — there is no abrupt interface, no lack of fusion, and no oxide film interruption. The planar crystal zone at the bottom is typically very thin, often only a few grains deep, and exists because the temperature gradient is steepest at the pool bottom and the solidification rate is lowest.

The dendritic zone above the planar zone grows in the direction of the maximum thermal gradient, which is initially perpendicular to the pool bottom but gradually rotates as the pool shape changes. The observation that the dendrite angle in the surface layer is smaller than at the bottom is a direct consequence of the beam scanning motion. As the beam moves forward, the trailing edge of the pool solidifies with dendrites oriented at an angle to the scanning direction, reflecting the combined effect of the thermal gradient and the scanning velocity. In practical terms, this means the surface layer of a beam-surfaced component will have anisotropic properties — the wear resistance and corrosion resistance will differ depending on the direction of measurement relative to the scanning direction.

Phase Formation and Precipitate Behavior

The study found that the matrix phase in both powder feeding and pre-coated powder surfacing layers is γ(Fe,Ni). This is significant because it means the surfacing layer is not simply a dilution of the substrate but a distinct alloy composition. The γ(Fe,Ni) phase is a solid solution that provides good toughness and moderate strength. The later-precipitated γ(Fe,Ni) phase is enriched in chromium, which is typical of solidification segregation — chromium partitions to the liquid phase during solidification and enriches the inter-dendritic regions.

In the pre-coated powder surfacing, local enrichment of chromium and carbon in some regions of the surfacing layer led to the precipitation of Cr23C6. This is a chromium carbide that forms when the local composition exceeds the solubility limit. Cr23C6 is hard and brittle, and while it can improve wear resistance, it can also serve as a crack initiation site if the carbide network is continuous or if the carbide size is excessive. In the powder feeding surfacing, the superheat of the melt pool was smaller, and the burn-off of B and Si slag-forming elements was less. Local enrichment of boron in the surfacing layer caused boron to preferentially combine with chromium to form Cr5B3. This boride is also hard and brittle, but its formation consumes chromium that would otherwise be available for corrosion resistance.

Engineering Practice and Shop Floor Implications

Process Selection Criteria

From a practical standpoint, this work provides a scientific basis for choosing between pre-coated powder and powder feeding methods. The pre-coated method is simpler and more reliable for thin layers, but the larger melt pool superheat leads to more burn-off of alloying elements and a greater risk of carbide precipitation. The powder feeding method gives better control over the melt pool temperature and preserves the alloying elements, but requires more sophisticated equipment and careful parameter control.

Parameter Pre-Coated Powder Powder Feeding
Melt pool superheat Higher Lower
Alloying element burn-off More (B, Si) Less
Precipitate type Cr23C6 Cr5B3
Process complexity Lower Higher
Layer thickness control Moderate Better
Dilution control Moderate Better

Defect Risks and Mitigation

The formation of Cr23C6 in the pre-coated powder process is a real concern for corrosion-resistant applications. If the surfacing layer is intended for service in a corrosive environment, the chromium carbides deplete the matrix of chromium, reducing the local corrosion resistance. This is the classic sensitization problem seen in stainless steels. The mitigation strategy is to either reduce the carbon content of the powder, increase the chromium content to compensate, or use a post-weld heat treatment to redissolve the carbides. In practice, I have seen components fail in service because the surfacing layer was applied with a pre-coated powder that had too high a carbon content, and the Cr23C6 network provided a preferential corrosion path.

The Cr5B3 formation in the powder feeding process is less problematic for corrosion but can cause issues with machinability. Borides are extremely hard and will rapidly dull cutting tools. If the surfacing layer needs to be machined after deposition, the presence of Cr5B3 will increase machining costs and may require the use of carbide or ceramic tooling.

Real-World Application Lessons

In my experience with beam surfacing for mold repair and die refurbishment, the most common defect is lack of fusion at the pool bottom, which occurs when the beam power is too low or the scanning speed is too high. The epitaxial growth zone described in this paper is the key to avoiding this defect — if the planar crystal zone is absent, it means the substrate was not properly melted and the bond is mechanical rather than metallurgical. A simple bend test or side bend test will reveal this immediately, and any competent welding inspector will reject the component.

The dendrite orientation angle is also important for fatigue applications. If the surfacing layer is subjected to cyclic loading, the dendrite boundaries are potential crack paths. The angle between the dendrites and the loading direction determines the effective crack growth resistance. In practice, I have seen surfacing layers fail prematurely because the dendrites were oriented perpendicular to the principal stress direction, providing an easy path for fatigue crack propagation. The solution is to adjust the scanning direction so that the dendrites are parallel to the principal stress, or to use a multi-pass technique that changes the dendrite orientation in successive layers.

Key Questions and Reflections

The Dilution Question

One question that this paper raises but does not fully answer is the dilution rate. The authors note that the powder feeding process has lower dilution, but they do not quantify it. In practice, dilution is the single most important parameter for surfacing applications because it determines the final composition of the layer. If the dilution is too high, the layer becomes too rich in iron from the substrate and loses its intended properties. If the dilution is too low, the bond may be weak. The typical acceptable dilution range for beam surfacing is 5-15%, and this should be verified by chemical analysis of the deposited layer.

The Heat Input Trade-Off

The paper describes the process as "large heat input" beam surfacing. This is a double-edged sword. High heat input increases the deposition rate and reduces the number of passes required, but it also increases the size of the heat-affected zone in the substrate and increases the residual stress. For components that are already heat-treated, the HAZ can cause softening or even cracking. In my experience, the best approach is to use the lowest heat input that still gives good fusion and acceptable deposition rate, and to use a preheat or interpass temperature control to manage the thermal cycle.

The Role of Boron and Silicon

The paper mentions that B and Si are slag-forming elements. This is a reminder that flux-cored wires and coated electrodes rely on these elements to protect the molten metal from oxidation. In beam surfacing with powder, the shielding is provided by the inert gas, but the powder itself may contain B and Si as deoxidizers. The burn-off of these elements is a real concern because it means the effective composition of the deposited layer is different from the nominal composition of the powder. This is a common source of confusion in production — the chemical analysis of the deposited layer does not match the powder certificate, and the engineer has to decide whether to adjust the powder composition or accept the deviation.

Study Insights and Practical Recommendations

This paper is a solid piece of fundamental research that provides a scientific framework for understanding beam surfacing. The key takeaway for the practicing engineer is that the solidification structure of a beam-surfaced layer is not uniform — it varies from the bottom to the top, and the orientation of the dendrites is influenced by the beam scanning direction. This means that the properties of the surfacing layer are anisotropic, and any design or specification that assumes isotropic properties is flawed.

For production applications, I recommend the following practical guidelines based on this work and my own experience:

  1. Always verify the dilution rate by chemical analysis of the first deposited layer, and adjust the process parameters to keep dilution in the 5-15% range.
  2. Use powder feeding rather than pre-coated powder for applications where alloying element retention is critical, especially for corrosion-resistant layers.
  3. Control the melt pool superheat by adjusting the beam power and scanning speed, and monitor the pool temperature with a pyrometer if available.
  4. For components that will be machined after surfacing, be aware that boride and carbide precipitates will increase machining difficulty and may require special tooling.
  5. For fatigue-critical applications, orient the scanning direction so that the dendrites are parallel to the principal stress direction, or use a multi-pass technique to break up the dendritic structure.
  6. Always perform a bend test or side bend test on a production coupon to verify that the epitaxial growth zone is present and the bond is metallurgical.

The practical value of this work lies in its clear description of the solidification sequence and the phase formation mechanisms. It gives the engineer a mental model of what is happening in the melt pool, which is essential for troubleshooting defects and optimizing process parameters. The next step for the industry would be to extend this work to include quantitative measurements of dilution and to develop a process window that maps the acceptable range of beam power, scanning speed, and powder feed rate for different substrate and powder combinations.

In summary, this is a well-executed study that answers a fundamental question about beam surfacing and provides a solid foundation for process optimization. The findings on dendrite orientation and phase formation are directly applicable to production, and the comparison between pre-coated powder and powder feeding methods is a valuable contribution to the selection of the appropriate process for a given application. The practical engineer should read this paper with a focus on the solidification behavior and the implications for layer quality, and should use the findings to guide process development and troubleshooting in the shop.