In our manufacturing facility, we embarked on the trial production of adjustable-speed hydraulic couplers several years ago. The pump and turbine components are critical elements in this system, and at the time, there were no domestic precedents for producing these parts in cast steel. Facing a lack of reference materials and established expertise, the challenges were substantial. However, the successful fabrication of qualified pumps and turbines was essential for the overall success of the hydraulic coupler project. After thorough analysis and research, we concluded that employing the lost wax casting method was a viable approach to achieve the required quality for these components. This report details our journey, methodologies, and outcomes in utilizing lost wax casting for these complex parts.
The pump and turbine are both disc-shaped and fabricated from cast steel. Their working cavities are composed of equally spaced blades, with the pump featuring a specific number of blades and the turbine another. The blades have a height of approximately 20 mm, a root thickness of around 5 mm, and a tip thickness of about 2 mm. The rough casting diameter for the pump is roughly 300 mm, and for the turbine, it is slightly smaller. The rough weights are approximately 50 kg and 40 kg, respectively, while the net weights are about 30 kg and 25 kg after machining. The structural complexity and high technical requirements made traditional sand casting unsuitable, reinforcing the necessity of lost wax casting.
The technical specifications for the pump and turbine are stringent, as summarized in the table below:
| Parameter | Requirement |
|---|---|
| Surface Roughness of Blade Flow Passage | Not lower than Grade 6.3 (approximately Ra 1.6 μm) |
| Heat Treatment | Quenched and tempered to hardness HB ≥ 170 |
| Tensile Strength (σ_b) | ≥ 60 kg/mm² (approximately 588 MPa) |
| Yield Strength (σ_s) | ≥ 35 kg/mm² (approximately 343 MPa) |
| Elongation (δ) | ≥ 14% |
| Reduction of Area (ψ) | ≥ 30% |
| Impact Toughness (α_k) | ≥ 4 kg·m/cm² (approximately 39 J/cm²) |
| Chemical Composition | C ≤ 0.32%, Si ≤ 0.8%, Mn 0.5-0.8%, P ≤ 0.04%, S ≤ 0.04%, Cr ≤ 0.3%, Ni ≤ 0.3% |
| Radiographic Inspection | No cracks or shrinkage defects allowed; circular defects ≤ 3 mm diameter permitted, but no linear defects |
| Blade Thickness and Spacing Tolerance | ≤ 0.5 mm |
| Static Balance Test | Unbalance weight tolerance ≤ 10 g |
| Dynamic Balance and Overspeed Test | Test speed 6000 rpm, working speed 5500 rpm, test duration 5 minutes |
The mechanical properties can be related to the material’s behavior under stress. For instance, the yield strength and tensile strength follow a typical relationship for steel, which can be expressed as: $$ \sigma_s = k \cdot \sigma_b $$ where \( k \) is a material constant typically around 0.5-0.7 for cast steels. Additionally, the impact toughness \( \alpha_k \) is critical for dynamic applications and is measured using standardized specimens. The hardness requirement ensures wear resistance, and it correlates with strength through empirical formulas such as: $$ \text{HB} \approx \frac{\sigma_b}{3} $$ for many steels, though this is approximate and varies with composition.
The lost wax casting process began with the design of the wax pattern mold. To achieve high dimensional accuracy, we used a one-piece integral mold for the wax patterns. The feeding risers and gating system were fabricated separately and then welded onto the wax pattern. This approach minimized joints and potential defects. During casting, we adhered to the principle of directional solidification by orienting the blade ends upward and the thicker disc sections downward. This configuration promotes effective feeding and reduces shrinkage porosity. The wax pattern, as illustrated below, was critical in defining the final casting geometry.
The shell molding process involved multiple layers to build a robust mold. The surface layer coating consisted of sodium silicate (water glass) and alumina powder, while the reinforcement layers used sodium silicate and bauxite. Ammonium chloride served as the hardening agent. Typically, we applied six layers, with the sixth layer reinforced with iron wire to enhance shell strength. Dewaxing was conducted swiftly to prevent shell deformation, and immediately after, we inserted three wedge-shaped refractory bricks into the cavity to support the shell during handling and firing, preventing sagging.
For shell firing, we placed the shell in a sand-filled box and used a box-type resistance furnace. The firing temperature ranged from 850°C to 900°C, with a holding time of 2 hours to ensure complete burnout of residues and achieve thermal stability. Melting was performed in a medium-frequency induction furnace with a capacity of about 500 kg. The melting temperature reached approximately 1600°C, and the pouring temperature was maintained between 1560°C and 1580°C, as measured by an optical pyrometer. The shell temperature during pouring was kept at 200-300°C to avoid thermal shock. The pouring speed was critical and set to 20 seconds per casting to ensure complete filling of the thin-walled blades. The relationship between pouring temperature \( T_p \) and solidification time \( t_s \) can be approximated by: $$ t_s = C \cdot (T_p – T_s)^{-n} $$ where \( C \) is a constant, \( T_s \) is the solidus temperature, and \( n \) is an exponent typically around 2 for steel. High-temperature, rapid pouring helped achieve full filling but required careful control to avoid defects like porosity.
Since the inception of trial castings, we have accumulated over three years of experience. Initially, the scrap rate exceeded 50%, primarily due to issues such as short pours, mold lifting, shell expansion, and shrinkage porosity. To tackle these challenges, we implemented quality improvement cycles, similar to the Plan-Do-Check-Act (PDCA) methodology, conducting three full cycles. This systematic approach involved analyzing defect root causes, implementing corrective actions, and monitoring results. As a result, the scrap rate was reduced to approximately 10-15%. The table below summarizes the mechanical properties obtained from test specimens cast in the same heat as the components:
| Mechanical Property | Average Value | Standard Deviation |
|---|---|---|
| Tensile Strength (σ_b) | 62 kg/mm² | ±2 kg/mm² |
| Yield Strength (σ_s) | 38 kg/mm² | ±1.5 kg/mm² |
| Elongation (δ) | 16% | ±1% |
| Reduction of Area (ψ) | 32% | ±2% |
| Impact Toughness (α_k) | 4.5 kg·m/cm² | ±0.3 kg·m/cm² |
Chemical composition analysis confirmed compliance with specifications, as shown in the following table:
| Element | Composition Range (%) | Measured Average (%) |
|---|---|---|
| Carbon (C) | ≤ 0.32 | 0.28 |
| Silicon (Si) | ≤ 0.8 | 0.65 |
| Manganese (Mn) | 0.5-0.8 | 0.70 |
| Phosphorus (P) | ≤ 0.04 | 0.025 |
| Sulfur (S) | ≤ 0.04 | 0.022 |
| Chromium (Cr) | ≤ 0.3 | 0.25 |
| Nickel (Ni) | ≤ 0.3 | 0.20 |
Radiographic inspection after rough machining involved multiple films per casting, with a tentative standard allowing circular defects up to 3 mm in diameter but prohibiting any linear defects. The blade thickness and spacing tolerances were within the specified 0.5 mm, and surface roughness consistently met or exceeded Grade 6.3. These quality metrics demonstrated that the lost wax casting process could produce components meeting design requirements. In practical operation, the trial-produced adjustable-speed hydraulic couplers were installed in a power plant and have been running reliably for over two years, confirming the effectiveness of our approach.
Despite the success, several issues persist in the lost wax casting process. Minor blade swelling, small blowholes, slag inclusions, and dimensional deformation occasionally occur. These defects are often attributed to variables in the wax pattern, shell molding, and pouring stages. For example, gas entrapment during pouring can lead to blowholes, which may be modeled by the ideal gas law: $$ PV = nRT $$ where pressure \( P \) and temperature \( T \) interactions during solidification can cause porosity. To address these, we are exploring optimized gating designs and stricter process controls. The continuous improvement in lost wax casting is essential to further enhance quality and yield rates.
In conclusion, the lost wax casting method has proven instrumental in manufacturing high-quality pump and turbine components for hydraulic couplers. Through iterative refinements, we have overcome initial challenges and achieved significant reductions in scrap rates. The integration of detailed process controls, material science principles, and quality assurance measures has enabled us to meet stringent technical specifications. Moving forward, we aim to resolve remaining defects through advanced simulation and process optimization, ensuring that lost wax casting remains a cornerstone of our production strategy for complex castings.