Scanning Path Planning for Laser Cladding

In the maintenance and remanufacturing of key industrial components, the restoration of dimensional accuracy and surface integrity is paramount. One prominent challenge arises during the overhaul of electric traction motor housings for high-speed trains, where components like node sleeves, often manufactured from spheroidal graphite cast iron, suffer from wear, scoring, or dimensional deviations. Conventional subtractive repair methods are inadequate for restoring both geometry and material properties. Laser cladding, an additive manufacturing technique, presents a powerful solution by depositing a new, metallurgically bonded layer onto the worn substrate. The success of this process, however, is not solely dependent on material selection and basic process parameters. For complex geometries, such as the internal cylindrical surface of a node sleeve, the strategy employed to move the laser beam—the scanning path—profoundly influences the thermal history, stress distribution, and ultimately, the quality of the clad layer. This work, therefore, investigates the critical role of scanning path planning in laser cladding Ni-Cu alloy onto a spheroidal graphite cast iron substrate, aiming to identify the optimal strategy for defect-free, high-integrity repairs.

1. Material System and Experimental Rationale

The substrate material selected for this study is a low-temperature grade ductile iron, GJ-400-18LT, a type of spheroidal graphite cast iron. Its microstructure, characterized by graphite nodules embedded in a ferritic matrix, provides good castability and mechanical properties but poses challenges for welding and cladding due to its high carbon content and tendency for martensite and carbide formation (leading to “white etching” structures) in the heat-affected zone (HAZ). The nominal chemical composition of this spheroidal graphite cast iron is detailed in Table 1.

Table 1: Nominal Chemical Composition of the Substrate (GJ-400-18LT Spheroidal Graphite Cast Iron)
Element C Si Mn P S Ni Mg Fe
wt.% 3.70 2.08 0.13 0.035 0.011 0.45 0.037 Balance

The clad material is a gas-atomized Ni-Cu alloy powder. Nickel-based alloys are favored for cladding on cast iron due to their good wettability, compatibility with the high-carbon substrate, and ability to suppress brittle phase formation. The addition of Copper enhances fluidity and corrosion resistance. The powder composition is given in Table 2.

Table 2: Chemical Composition of the Ni-Cu Alloy Cladding Powder
Element Ni Cu Si B Fe C O
wt.% Balance 21.93 2.02 1.03 0.250 0.015 0.028

The core mechanism of laser cladding involves the formation of a small molten pool on the substrate surface via a high-power laser beam. Powder is synchronously fed into this pool, where it melts and mixes with a thin layer of the melted substrate. Upon solidification, a dense, metallurgically bonded layer is formed. The degree of substrate melting, or dilution, must be controlled. Excessive dilution can import undesirable elements (like carbon from the spheroidal graphite cast iron) into the clad, degrading its properties, while insufficient dilution leads to poor bonding. This dilution ratio $\\eta$ can be conceptually described by the ratio of the cross-sectional area of substrate melt ($A_s$) to the total melt area ($A_c + A_s$):

$$ \eta = \frac{A_s}{A_c + A_s} $$

where $A_c$ is the cross-sectional area of the melted clad powder. The process is governed by key parameters: laser power ($P$), scanning velocity ($v_s$), powder feed rate ($\\dot{m}$), and beam diameter ($d$). Their interrelationship influences the energy density $E_d$ delivered per unit length:

$$ E_d = \frac{P}{v_s \cdot d} $$

A higher $E_d$ generally increases melt pool depth and dilution. After optimization for this specific material system, the fixed process parameters used for all path planning experiments are summarized in Table 3.

To repair a typical damage depth of 0.5 mm, a two-layer clad strategy was employed, as the optimized single-layer clad thickness was found to be 0.6-0.7 mm. The challenge lies in the cylindrical geometry. Planning the sequence of individual clad tracks (or “beads”) to cover the entire internal surface area is non-trivial. The scanning path dictates the localized heating and cooling cycles, which in turn control thermal stress accumulation, distortion, and defect formation.

Table 3: Optimized Laser Cladding Process Parameters
Parameter Laser Power Scan Speed Powder Feed Rate Beam Spot Size Overlap Ratio
Value 1.5 kW 8 mm/s 9.69 g/min 3 mm 50%

2. Scanning Path Strategies and Thermal Physics

Based on the geometry of the node sleeve, we designed and implemented three distinct scanning path strategies. The thermal history for any point in the clad layer is a superposition of the heat input from successive laser passes. This history $T(x,y,z,t)$ can be modeled by the transient heat conduction equation with a moving heat source $Q$:

$$ \rho c_p \frac{\partial T}{\partial t} = \nabla \cdot (k \nabla T) + Q(x,y,z,t) $$

where $\\rho$ is density, $c_p$ is specific heat, $k$ is thermal conductivity, and $Q$ represents the laser energy input. Different scanning paths lead to vastly different $Q(x,y,z,t)$ functions, resulting in different temperature gradients $\\nabla T$ and cooling rates $\\frac{\partial T}{\partial t}$.

Path 1: Bi-Directional Circumferential Lap (Width-wise A/B Partitioning). The internal cylindrical surface was conceptually divided into two longitudinal halves, A and B. The clad head first covered half A by rotating the workpiece continuously while the head translated along the axis for each circumferential track. The process was then paused, the workpiece was re-positioned (rotated 180°), and half B was clad. The final step involved a very short circumferential track to join the end of zone A with the start of zone B. This strategy induces a significant thermal discontinuity. The temperature field for the first clad track in zone B, $T_B^1$, starts from a pre-heated state from zone A’s residual heat, but the long pause leads to substantial cooling of the interface region. The thermal gradient at the A-B seam can be extreme:

$$ \nabla T_{seam} \approx \frac{T_{A,hot} – T_{B,cold}}{\delta x} $$

where $\\delta x$ is a small distance across the seam. This promotes defects.

Path 2: Unidirectional Circumferential Lap (Axial Progression). This path involves a continuous, helical-like progression. The clad head deposits a single full circumferential track. Then, the head steps axially by a distance equal to the track width multiplied by (1 – overlap ratio), and the next circumferential track is deposited in the same rotational direction. This creates a continuous, unidirectional thermal front moving along the axis. The pre-heating effect is optimal because the previous track ($n$) acts as a direct thermal source for the subsequent track ($n+1$) being deposited alongside it. The cooling rate is more uniform.

Path 3: Axial “Zigzag” Lap (Oscillating). Here, the laser head moves linearly back-and-forth along the axis of the cylinder for a short segment. After each axial pass, the workpiece is indexed rotationally by a small angle to achieve the circumferential overlap, and the head reverses direction. This creates very short, discrete clad segments. The primary issue is intense localized heat accumulation. The time interval between the end of one axial pass and the start of the next pass on an adjacent area is minimal. The heat input $Q$ is nearly continuous in a small volume, leading to a high average temperature rise $\\Delta \\bar{T}$ which can be approximated by:

$$ \Delta \bar{T} \propto \frac{P \cdot t_{dwell}}{v_s \cdot d \cdot \rho c_p \cdot V_{zone}} $$

where $t_{dwell}$ is the effective time the heat source spends near a specific zone and $V_{zone}$ is the volume of that zone. This path maximizes $t_{dwell}/V_{zone}$, causing potential overheating and element burn-off.

3. Results: Macroscopic, Microscopic, and Mechanical Analysis

The macroscopic appearance of the clad layers provided the first clear indicator of path-dependent quality.

  • Path 1 (A/B Partition): The surface was relatively flat but exhibited visible pits and lack-of-fusion defects precisely at the seam between zones A and B, correlating with the predicted thermal discontinuity.
  • Path 2 (Unidirectional Circumferential): The clad surface was uniform, smooth, and free from visible defects like spatter or unmelted powder. The track ripple pattern was consistent, indicating stable melt pool dynamics.
  • Path 3 (Axial Zigzag): The surface quality was poorest, showing high roughness, uneven track geometry, and signs of excessive melting/burn-off, consistent with the model of severe heat accumulation.

Post-cladding machining to the final service dimensions was performed, followed by liquid penetrant testing (PT), a critical non-destructive evaluation. The results starkly contrasted the paths.

Table 4: Defect Analysis via Penetrant Testing for Different Scanning Paths
Scanning Path Defect Type Observed Defect Location & Characteristics Root Cause (Thermal History)
Path 1 (A/B Partition) Porosity, Minor Lack-of-Fusion Concentrated at the longitudinal seam between zones A and B. High thermal gradient and incomplete melting at the cold seam interface during the second-stage cladding.
Path 2 (Unidirectional Circumferential) None No indications over the entire service surface. Uniform thermal progression, effective pre-heating, and gradual cooling minimized thermal stress and promoted complete fusion.
Path 3 (Axial Zigzag) Axial Cracks, Cluster Porosity Cracks propagating axially along track boundaries; porosity clusters within tracks. Extreme heat accumulation led to high residual tensile stress upon cooling; gas entrapment due to turbulent pool dynamics from rapid re-melting.

Metallographic examination of cross-sections perpendicular to the clad direction provided microstructural evidence. For the spheroidal graphite cast iron substrate, the critical region is the interface and HAZ.

  • In Path 1 samples, the seam area revealed oxide inclusions and pores up to 300 µm in size, confirming the PT findings.
  • In Path 2 samples, the interface was wavy, indicating a metallurgical bond with controlled dilution. The graphite nodules near the interface showed limited carbon dissolution, effectively suppressing the formation of a continuous brittle white iron layer (ledeburite). The clad layer itself exhibited a fine, homogeneous dendritic/cellular structure.
  • In Path 3 samples, large gas pores (140-350 µm) were prevalent within the clad layer. Micro-cracks were observed initiating from these pores, following the principle that pores act as stress concentrators. The stress intensity factor $K_I$ near a pore of radius $a$ under residual stress $\\sigma$ is:

$$ K_I \approx \sigma \sqrt{\pi a} $$

When $K_I$ exceeds the material’s fracture toughness, crack propagation occurs, explaining the axial cracking observed.

Microhardness (HV0.1) traverses from the top of the clad layer into the substrate for the optimal Path 2 sample revealed a mechanically graded structure, which is beneficial for service performance. The results are summarized in Table 5 and depicted graphically.

Table 5: Microhardness Profile for Path 2 (Unidirectional Circumferential Clad)
Region Average Microhardness (HV0.1) Standard Deviation Remarks
Clad Layer (Top) 288.2 ± 15.3 Highest hardness due to fine microstructure and solid solution strengthening from Cu/Si/B.
Clad Layer (Mid-Depth) 265.7 ± 18.1 Slight decrease due to thermal cycling from the second layer.
Fusion Interface Variable (220-350) – Fluctuation due to mixing zone and possible minor carbide formation.
Heat-Affected Zone (HAZ) 179.5 – 245.0 – Hardness spike indicates possible transformation hardening; depends on proximity to interface.
Spheroidal Graphite Cast Iron Substrate 159.2 ± 12.7 Base ferritic-pearlitic matrix hardness.

The gradual decrease in hardness from the clad surface to the substrate helps mitigate abrupt property transitions, reducing the risk of delamination under load. The hardness in the HAZ, while elevated, did not reach critically high levels that would indicate extensive brittle martensite, thanks to the controlled heat input and the inherent graphitic structure of the spheroidal graphite cast iron which buffers carbon availability.

4. Discussion: Synthesis of Thermal Management and Path Efficacy

The experimental outcomes can be synthesized through the lens of thermal management. The primary objective in laser cladding a sensitive material like spheroidal graphite cast iron is to manage the thermal cycle to avoid:
1. Excessive quenching (high $\\frac{dT}{dt}$) causing martensite.
2. Excessive heat accumulation (high $\\int Q \\, dt$) causing burning, distortion, and high tensile stress.
3. Large thermal gradients ($\\nabla T$) causing cracking and lack-of-fusion.

Path 2 (Unidirectional Circumferential) excelled because it created a quasi-steady-state thermal condition. The heat source moved continuously, allowing the previously deposited material to act as a distributed pre-heater. This minimized the temperature difference $\\Delta T$ between the new melt pool and its neighboring solid, reducing thermal shock and promoting good fusion. The cooling rate was sufficiently fast to produce a fine clad microstructure but not so fast as to cause cracking. The residual stress state, primarily governed by the temperature gradient mechanism, is more balanced in this configuration. The stress $\\sigma_{res}$ can be related to the constrained thermal strain:

$$ \sigma_{res} \approx E \cdot \alpha \cdot \Delta T_{constraint} $$

where $E$ is Young’s modulus, $\\alpha$ is the coefficient of thermal expansion, and $\\Delta T_{constraint}$ is the effective temperature drop under constraint. Path 2 minimizes localized $\\Delta T_{constraint}$.

Path 1 (A/B Partition) failed at the seam due to a breakdown of this steady state. The thermal discontinuity introduced a cold barrier, creating a large $\\nabla T$ at the seam during the second cladding stage, leading to incomplete melting and porosity.

Path 3 (Axial Zigzag) failed due to the opposite problem: excessive heat concentration. The short, rapid reciprocating motion did not allow sufficient time for heat dissipation ($t_{cool} \\ll t_{heat}$), leading to a constantly elevated local temperature. This caused element vaporization, increased melt pool turbulence (promoting porosity), and, upon final cooling, generated very high tensile stresses as the large, overheated volume contracted against the cooler surrounding spheroidal graphite cast iron substrate, resulting in cracking.

5. Conclusion

This investigation conclusively demonstrates that scanning path planning is a decisive factor in the quality of laser-clad repairs on spheroidal graphite cast iron components, equally critical as the optimization of power, speed, and feed rate. For the specific case of an internal cylindrical surface, such as a motor node sleeve:

  • The Unidirectional Circumferential Lap (Axial Progression) strategy is optimal. It ensures a continuous, controlled thermal front that promotes uniform pre-heating, minimizes thermal gradients, and prevents excessive heat accumulation.
  • This path yielded a defect-free clad layer with excellent surface finish, a sound metallurgical bond with controlled dilution, a favorable graded hardness profile, and most importantly, a service surface that passed rigorous non-destructive inspection.
  • Alternative strategies like partitioned cladding or short-segment oscillating paths introduce significant thermal management challenges—discontinuities or accumulation—that manifest as porosity, lack-of-fusion, and cracking defects, rendering the repaired component unfit for service.

The principles derived here—favoring continuous, unidirectional thermal progression to manage heat input and stress—are broadly applicable to the laser additive repair of other complex geometries on challenging substrates like spheroidal graphite cast iron. This work underscores that in laser-based remanufacturing, the “toolpath” is not merely a geometric coverage instruction but a fundamental process parameter dictating the thermomechanical history and ultimate integrity of the restored component.

Scroll to Top