As a casting process engineer specializing in large-scale industrial components, I have been deeply involved in the development of steel casting techniques for renewable energy applications. In recent years, the rapid advancement of offshore wind power technology has driven a significant increase in the unit capacity of wind turbines, with megawatt-scale systems becoming the norm. Among these, the secondary planetary frame—a critical load-bearing part in wind turbine gearboxes—plays a pivotal role in ensuring operational reliability and longevity. The steel casting of this component, made from materials like ZG35CrMo, presents formidable challenges due to its complex frame structure, substantial wall thicknesses ranging from 80 mm to 199 mm, and stringent quality requirements, particularly in internal soundness and crack resistance. In this article, I will share my firsthand experience in addressing these challenges through process optimization, leveraging numerical simulation, and implementing rigorous quality control measures, with a focus on enhancing steel casting integrity for such demanding applications.
The secondary planetary frame casting, with an outline dimension of Φ1,668 mm × 939 mm and a rough weight of approximately 5,500 kg, features a framework design that includes upper and lower annular plates connected by triangular columns. This geometry inherently leads to high thermal stresses and solidification issues, as the vertical contraction is constrained by the horizontal sections, creating stress concentrations at junctions. Initially, our foundry employed a traditional casting process using cold-set furan resin molds with zircon-based coatings. The original scheme, as illustrated in early designs, incorporated a single large circular insulated riser above the planetary axis, with un-machined holes in the annular plates and a riser feed channel between them. The gating system was designed with three bottom-gated in-gates on the lower annular plate to promote directional solidification through low-temperature, fast-pouring principles. However, this approach resulted in a low yield rate of 45% and several critical defects that compromised the steel casting quality.
During production, we encountered three primary issues that necessitated a thorough reevaluation. First, ultrasonic testing of the internal spline region (the long-shaft end inner hole) revealed rejectable defects. Customer specifications mandated that single discontinuities must not exceed Φ3 mm in equivalent size, with no allowance for clustered, extended, or linear defects, and prohibit any weld repairs. Despite the absence of obvious shrinkage cavities in simulations, the area showed coarse grain structures likely due to excessive riser heat influence, leading to unsatisfactory ultrasonic inspection results. Second, shrinkage porosity at the fillets between the triangular columns and annular plates acted as stress raisers, initiating cracks that jeopardized structural integrity. Third, the large riser and its feed channel were difficult to cut off post-casting, prolonging machining cycles and increasing costs. These problems underscored the need for a optimized steel casting process that could ensure internal quality while improving efficiency.
To tackle these challenges, I led a comprehensive process optimization initiative centered on redesigning the riser system and incorporating stress-relief features. The key modifications included: (1) Implementing separate risers for the upper and lower annular plates to localize feeding and reduce thermal impact on the internal spline zone. For the upper plate’s hot spot (designated as area A), an open riser was placed above the shaft head with a riser pad in the cavity to form a feed channel. For the lower plate’s hot spot (area B), a blind riser with lateral channels was used to enhance feeding. This not only refined the grain structure in critical regions but also reduced riser size and weight, facilitating easier cutting and boosting the yield rate to 58%. (2) Adding process ribs, specifically 120 mm × 120 mm × 20 mm割筋 (stress-relief ribs), at the roots of the fillets where triangular columns meet the annular plates. These ribs help dissipate contraction stresses by providing additional cooling surfaces and mechanical support, thereby mitigating crack formation. The optimized layout significantly improved the steel casting’s soundness and manufacturability.
To validate these changes, we utilized InteCAST CAE numerical simulation software to analyze the solidification behavior and defect formation. The simulation involved solving the heat transfer and fluid dynamics equations governing the steel casting process. For instance, the temperature field during solidification can be described by the transient heat conduction equation: $$ \frac{\partial T}{\partial t} = \alpha \nabla^2 T + \frac{Q}{\rho c_p} $$ where \( T \) is temperature, \( t \) is time, \( \alpha \) is thermal diffusivity, \( Q \) is latent heat release, \( \rho \) is density, and \( c_p \) is specific heat. By applying this to the 3D model, we predicted shrinkage porosity using the Niyama criterion, which relates thermal gradients to defect susceptibility: $$ G / \sqrt{R} \leq C $$ where \( G \) is temperature gradient, \( R \) is cooling rate, and \( C \) is a material constant. The results, as shown in simulation outputs, confirmed that the optimized riser arrangement effectively eliminated macroscopic shrinkage, with sound metal throughout the casting. A comparative table summarizes the key parameters before and after optimization:
| Parameter | Original Process | Optimized Process |
|---|---|---|
| Number of Riser | 1 main riser | 2 separate risers |
| Riser Size (approx.) | Large circular, high profile | Smaller, reduced height |
| Yield Rate | 45% | 58% |
| Ultrasonic Test Result | Unacceptable in spline area | Grade 2 in spline area |
| Crack Incidence | High at fillets | Minimal due to ribs |
| Cutting Difficulty | High, time-consuming | Reduced, faster processing |
The numerical simulation also allowed us to quantify the thermal stresses responsible for cracking. Using elastic-plastic constitutive models, we estimated the von Mises stress distribution: $$ \sigma_{v} = \sqrt{ \frac{(\sigma_1 – \sigma_2)^2 + (\sigma_2 – \sigma_3)^2 + (\sigma_3 – \sigma_1)^2}{2} } $$ where \( \sigma_1, \sigma_2, \sigma_3 \) are principal stresses. The analysis indicated stress concentrations exceeding the material’s yield strength at fillet junctions in the original design, which were alleviated by the added ribs. This underscored the importance of integrated thermal-mechanical simulation in steel casting optimization. Furthermore, we derived a feeding distance formula to ensure adequate riser coverage: $$ L_f = k \cdot \sqrt{V/A} $$ where \( L_f \) is feeding distance, \( k \) is a constant dependent on alloy and mold properties, \( V \) is volume, and \( A \) is surface area. By tailoring riser placements based on this, we achieved uniform solidification patterns.
Upon implementing the optimized process in production, we conducted extensive non-destructive testing to verify the steel casting quality. All trial castings underwent 100% ultrasonic inspection per GB7233.1 standards and 100% magnetic particle testing. The internal spline region now met Grade 2 requirements, with single defects below Φ3 mm equivalent and no unacceptable defect types. Other areas conformed to Grade 3, demonstrating a significant improvement in internal soundness. Magnetic particle inspection revealed nearly no crack indications at the column-annular plate junctions, validating the effectiveness of the stress-relief ribs. The finished castings exhibited excellent dimensional stability and surface quality, ready for subsequent machining and assembly. This success not only ensured compliance with customer specifications but also reduced rework and scrap rates, enhancing overall productivity.
In conclusion, the optimization of the steel casting process for large wind turbine secondary planetary frames has proven highly effective in overcoming quality challenges. By redesigning the riser system to minimize thermal influence and incorporating strategic stress-relief features, we eliminated critical defects such as shrinkage-related cracks and ultrasonic test failures. The use of numerical simulation provided a scientific basis for these modifications, enabling precise control over solidification and stress dynamics. The resulting increase in yield rate from 45% to 58% translates to substantial cost savings and shorter lead times, bolstering the competitiveness of steel casting in renewable energy sectors. Future work will focus on extending these principles to other complex steel casting components, leveraging advanced materials and simulation tools to further push the boundaries of quality and efficiency. Through continuous innovation, steel casting remains a cornerstone of durable and reliable wind power infrastructure.