In my experience as an engineer specializing in foundry processes, addressing internal defects in critical components like steam turbine cylinders is paramount. These steel casting components must meet stringent quality standards, particularly for the split flanges where ultrasonic testing is required per specifications such as JB/T9630.2. The goal is to achieve a Grade 2 quality level, ensuring no defects like porosity or shrinkage that could compromise performance. This article details my first-person approach to resolving these issues through process optimization, numerical simulation, and practical adjustments in steel casting production.
The challenge centered on the split flanges of a steam turbine cylinder上半, made from ZG20CrMo steel, with a casting weight of 6,400 kg and a yield of 55%. The structure involved varying wall thicknesses: a 45 mm cylinder wall with horizontal and vertical flanges, where the small flange was 160 mm thick and the large flange 95 mm thick. This configuration created thermal junctions prone to defects. Initial ultrasonic inspection revealed porosity and shrinkage in areas between risers on the split face, leading to high repair rates and production delays. My task was to refine the steel casting process to eliminate these defects.
To begin, I analyzed the original steel casting process, which used cold-set furan resin sand molding, electric arc furnace melting, and a bottom-pouring gating system. The key was to ensure directional solidification through proper riser design. I calculated the modulus for the flange sections to determine riser sizes. For any complex steel casting, the modulus \( M \) is defined as the volume-to-surface area ratio, \( M = \frac{V}{A} \), which guides riser design. For the large flange, with a hot spot diameter derived from inscribed circles, the modulus was computed as follows:
$$ M_{\text{large}} = \frac{V_{\text{large}}}{A_{\text{large}}} = 6.4 \, \text{cm} $$
Similarly, for the small flange:
$$ M_{\text{small}} = \frac{V_{\text{small}}}{A_{\text{small}}} = 6.95 \, \text{cm} $$
Applying a safety factor of 1.2 for riser design, the adjusted moduli were \( M_{\text{large}} = 7.68 \, \text{cm} \) and \( M_{\text{small}} = 8.34 \, \text{cm} \). Using a feeding efficiency \( \eta = 12\% \) for elliptical risers, I designed risers as summarized in Table 1.
| Flange Section | Calculated Modulus (cm) | Riser Dimensions (mm) | Riser Weight (kg) |
|---|---|---|---|
| Large Flange | 7.68 | Y475 × 575 × 400 | 700 |
| Small Flange | 8.34 | Y475 × 675 × 400 | 850 |
The riser layout covered thermal junctions, with distances between risers less than three times the minimum hot spot size, theoretically ensuring adequate feeding. Numerical simulation was employed to predict defect formation. Using simulation software, I modeled the solidification process, and the results initially indicated no defects between risers, suggesting the steel casting process was feasible. However, actual ultrasonic testing told a different story: defects like porosity were detected in the mid-layer of the flange between risers, spanning approximately 150 mm × 100 mm. This discrepancy highlighted that the simulation might not have captured localized thermal conditions leading to mushy zone solidification and shrinkage.
Upon reflection, I realized that the riser heat influence created a flat temperature gradient in the flange center, promoting equiaxed solidification and porosity. To address this, I first experimented with chills placed between risers. Chills are external cooling devices used in steel casting to enhance directional solidification. I tested various chill thicknesses, but ultrasonic inspection still showed defects, albeit shifted upward. This indicated that chills provided only temporary cooling, quickly reaching thermal saturation and failing to sustain a steep temperature gradient. Table 2 summarizes the chill trial results.
| Chill Thickness (mm) | Defect Location After UT | Defect Severity |
|---|---|---|
| 20 | Upper flange region | Moderate porosity |
| 30 | Mid-flange shifted up | Minor shrinkage |
| 40 | Near riser interface | Significant defects |
The persistence of defects led me to reconsider the riser configuration. In steel casting, risers not only feed molten metal but also allow gas and inclusion escape. I proposed connecting the risers with a feed path or using a continuous riser across the split face. This approach, known as riser interconnection, ensures the entire flange is under riser coverage, promoting better feeding and reducing isolated mushy zones. The height of this connection was set equal to the hot spot size, based on the modulus calculation. Implementing this, I modified the steel casting pattern to include a unified riser system, as illustrated in the process schematic.
To validate this, I ran numerical simulations again, focusing on thermal analysis. The solidification time \( t_s \) for a steel casting can be estimated using Chvorinov’s rule: \( t_s = k \cdot \left( \frac{V}{A} \right)^2 \), where \( k \) is a constant dependent on material and mold properties. For the interconnected riser design, I calculated the modulus for the combined flange region:
$$ M_{\text{combined}} = \frac{V_{\text{flange}} + V_{\text{riser\_connector}}}{A_{\text{flange}} + A_{\text{riser\_connector}}} $$
Assuming a connector height of 160 mm (matching the small flange hot spot), the modulus increased, enhancing feeding. Simulation results showed a reduced porosity index, with the Niyama criterion \( N_y \) used to predict shrinkage: \( N_y = \frac{G}{\sqrt{T}} \), where \( G \) is the temperature gradient and \( T \) is the cooling rate. Values above a threshold indicate sound steel casting. After modification, \( N_y \) improved significantly in the flange mid-region.
Practical trials were conducted with the new riser interconnection design. The steel casting process involved meticulous control of pouring temperature and speed to ensure optimal feeding. Post-casting, the split face was machined and subjected to ultrasonic testing per JB/T9630.2. The results were promising: no超标 defects were detected, and the steel casting met Grade 2 quality requirements. This confirmed that the interconnected riser system effectively eliminated porosity by maintaining a directional solidification front.
Further analysis involved comparing the original and modified processes through key parameters. Table 3 provides a summary of the improvements.
| Parameter | Original Process | Modified Process (Riser Interconnection) |
|---|---|---|
| Riser Coverage | Isolated risers | Continuous riser network |
| Temperature Gradient (G) | Low in flange center | High, sustained gradient |
| Solidification Mode | Mushy zone formation | Directional solidification |
| Ultrasonic Test Results | Porosity defects present | No defects, Grade 2 achieved |
| Production Efficiency | Low due to repairs | High, reduced rework |
The success of this approach can be generalized to other steel casting components with similar geometry. The modulus method remains a cornerstone for riser design, but it must be complemented with thermal analysis. For future steel casting projects, I recommend integrating numerical simulation early in process planning. The governing heat transfer equation during solidification is:
$$ \rho c_p \frac{\partial T}{\partial t} = \nabla \cdot (k \nabla T) + L \frac{\partial f_s}{\partial t} $$
where \( \rho \) is density, \( c_p \) is specific heat, \( k \) is thermal conductivity, \( L \) is latent heat, and \( f_s \) is solid fraction. Solving this with boundary conditions for risers and chills helps optimize steel casting parameters.
In conclusion, by shifting from isolated risers to an interconnected riser system, I effectively solved internal defects in steam turbine cylinder split flanges. This steel casting improvement not only enhanced quality but also boosted productivity by minimizing post-cast repairs. The key takeaway is that in complex steel casting, ensuring continuous feeding paths through riser connectivity can mitigate shrinkage and porosity, aligning with industry standards for high-integrity components. Moving forward, continuous innovation in steel casting processes, supported by simulation and empirical data, will drive further advancements in defect-free manufacturing.