In the power generation industry, axial blowers are critical equipment for boiler and desulfurization systems, operating under harsh conditions with high temperatures and corrosive atmospheres while bearing significant alternating dynamic loads. The blade, as the core component, demands exceptional external profile accuracy and internal soundness, often featuring a twisted geometry and very thin wall thickness. Traditionally, such blades are manufactured via metal mold casting or precision casting, but these methods incur high costs and prolonged lead times. Drawing on our extensive experience in producing hydraulic turbine blades using sand casting foundry techniques, we recognized that some axial blower blades share morphological similarities with turbine axial flow blades. This inspired us to explore a sand casting foundry approach for thin-wall axial blower blades. In this article, we present our systematic investigation, simulation, and successful practice of a sand casting foundry method for the JHL-583 axial blower blade.
Our research group has long been engaged in the production of hydro-generator blades using sand casting foundry, accumulating rich knowledge in gating system design, riser sizing, and defect control. Leveraging this background, we collaborated closely with a major blower manufacturer to adapt sand casting foundry for JHL-583 blades. The following sections detail the technical requirements, two alternative process designs, numerical simulations, practical implementation, and verification results.
Technical Requirements for JHL-583 Blade
The JHL-583 blade is specified as grade ZG06Cr13Ni4Mo, a low-carbon martensitic stainless steel. The chemical composition and mechanical properties must comply with standard JB/T10384—2002. The requirements are summarized in Table 1 and Table 2.
| Element | C | Si | Mn | S | P | Cr | Ni | Mo |
|---|---|---|---|---|---|---|---|---|
| Specification | 0.06 max | 1.00 max | 1.00 max | 0.030 max | 0.035 max | 11.5–13.5 | 3.5–5.0 | 0.40–1.00 |
| Property | Rp0.2 / (N·mm⁻²) | Rm / (N·mm⁻²) | A / % | Z / % | AKV / J | HBW |
|---|---|---|---|---|---|---|
| Value | ≥ 550 | ≥ 750 | ≥ 15 | ≥ 35 | ≥ 50 | 217–286 |
Non-destructive testing (NDT) was mandated per A609 standard: 100% ultrasonic testing (UT) with Class II acceptance, and 100% penetrant testing (PT) also Class II. These stringent requirements necessitated a robust sand casting foundry process to ensure internal soundness and surface quality.
Process Design
We adopted sodium silicate-bonded sand (water glass sand) as the molding material due to its good collapsibility and low cost. Machining allowances were set at +7 mm on both the pressure side and suction side of the blade, and +15 mm on the peripheral edges. Two process alternatives were investigated: vertical casting and horizontal casting.
Alternative I: Vertical Casting Process
Following conventional practice for turbine blades, we designed a vertical arrangement. The riser diameter was determined using the hot spot circle method and modulus method:
$$D_r = 260 \text{ mm}, \quad H_r = 500 \text{ mm}$$
The gating system was bottom-gated with a sprue diameter of 50 mm, a runner of 60 mm, a cross gate of 55/60 mm × 35 mm, and three ingates each 70/80 mm × 25 mm. The pouring temperature was set at 1600 °C, and the initial mold temperature at 25 °C. A three-dimensional solid model of the vertical casting layout was created using SolidEdge, and the filling and solidification were simulated. The results indicated severe distortion in the Y-direction (perpendicular to the blade face) of about 5 mm over nearly 50% of the blade surface. Moreover, the shrinkage porosity distribution was concentrated in the lower thin region, as shown in the simulation output. The long flow distance of molten steel from bottom to top caused inadequate feeding at the bottom, leading to unacceptable porosity. This confirmed that the vertical sand casting foundry approach was not feasible for such a thin-wall twisted blade.
Alternative II: Horizontal Casting Process
To shorten the flow path and improve feeding, we designed a horizontal casting orientation. The pouring temperature was reduced to 1580 °C. The riser diameter was 230 mm and height 520 mm. Two slag traps were added on the top of the blade to facilitate gas and slag removal. The bottom-gated system consisted of a sprue (50 mm), runner (60 mm), cross gate (50/55 mm × 40 mm), and six ingates (70/80 mm × 20 mm).
The modulus method was used to verify riser adequacy. For the casting, the modulus \( M_c = V_c / A_c = 13.7 \). For the riser, \( M_R = V_R / S_R = 50.2 \). The expansion factor f is:
$$ f = \frac{M_R}{M_c} = \frac{50.2}{13.7} = 3.66 $$
Since \( f \ge 1.2 \), the riser size is sufficient. The maximum feeding capacity of the riser is calculated by:
$$ G_{Cmax} = G_R \times \frac{\eta – \varepsilon}{\varepsilon} $$
where \( G_R \) is the riser weight, \( \eta = 0.14 \) (open riser efficiency), and \( \varepsilon = 0.053 \) (solidification shrinkage of ZG06Cr13Ni4Mo). Substituting values yields \( G_{Cmax} = 371.8 \text{ kg} \), whereas the blade casting weight is 275 kg. Thus, the riser provides ample feeding.
The horizontal casting process was simulated under identical boundary conditions (mold temperature 25 °C, pouring temperature 1580 °C). The results demonstrated a drastically reduced X-direction (axial direction) deformation of about 2 mm over a small area. Shrinkage porosity was confined to the riser and gate regions, with no significant defects in the blade body. This confirmed that the horizontal sand casting foundry design was viable.
| Parameter | Vertical Casting | Horizontal Casting |
|---|---|---|
| Maximum deformation (mm) | ~5 (Y-direction) | ~2 (X-direction) |
| Deformation area (%) | ~50 | Small |
| Shrinkage porosity location | Blade bottom thin region | Riser and gate only |
| Feeding adequacy | Poor | Satisfactory |
Based on these simulations, we selected the horizontal sand casting foundry process for practical trial production.
Sand Casting Foundry Practice
We proceeded to manufacture the first blade using the horizontal casting design. A high-quality Korean pine wood pattern was fabricated with a 2% shrinkage allowance. The pattern dimensions were verified using an optical three-dimensional coordinate measuring machine (SOLO type) to ensure compliance with the nominal geometry. For the mold, chromite sand was chosen as the facing sand due to its high refractoriness, which prevents sand burning and erosion. Organic ester-hardened sodium silicate was used as binder, and an alcohol-based zircon flour coating was applied. The mold was surface-dried and preheated before pouring.
Melting was performed using an electric arc furnace combined with AOD (argon oxygen decarburization) refining to achieve strict control over carbon (<0.06%) and impurity levels (P, S). The actual chemical composition of the first melt is given in Table 4.
| C | Si | Mn | S | P | Cr | Ni | Mo |
|---|---|---|---|---|---|---|---|
| 0.025 | 0.46 | 0.67 | 0.007 | 0.024 | 11.71 | 3.88 | 0.42 |
Pouring was executed at a controlled rate according to the predefined schedule. The mold was knocked out after the casting cooled below 300 °C to minimize distortion. The riser was cut off while the casting was still above 200 °C to avoid cracking. Visual inspection revealed no defects such as slag inclusions, gas holes, cold shuts, or cracks.
Heat treatment consisted of normalizing and double tempering, as shown in the following sequence (Figure omitted per instruction). The blade was placed with the hub oriented downward, supported to avoid direct contact with the furnace floor. Heating rate was limited to 70 °C/h, with a maximum temperature gradient of 5 °C across the furnace. No stacking was allowed to prevent deformation.
Verification of the First Article
After heat treatment, the blade profile was measured using a coordinate measurement machine. All surfaces had sufficient stock for subsequent machining. Ultrasonic testing and penetrant testing were performed per A609 standard on the entire blade after rough machining, and again after final machining. Both tests met Class II requirements, confirming internal soundness and surface integrity.
Mechanical properties were evaluated using a separately cast test block (Kiel block) heat-treated together with the blade. The results are summarized in Table 5.
| Property | Rm / (N·mm⁻²) | Rp0.2 / (N·mm⁻²) | A / % | Z / % | AKV2 / J | HBW 5/750 |
|---|---|---|---|---|---|---|
| Value | 789 | 631 | 21.5 | 68 | 105, 118, 124 | 253, 254, 253 |
All values well exceeded the specification minimums, demonstrating that the sand casting foundry process delivered material properties equivalent to those obtained by more expensive methods.
It is worth noting that the success of this sand casting foundry approach hinged on the careful optimization of process parameters through numerical simulation. The horizontal orientation minimized deformation and ensured directional solidification toward the riser. The use of chromite facing sand and controlled pouring temperature further enhanced the casting quality. Additionally, the integration of AOD refining guaranteed the required low carbon content and cleanliness, which are essential for the corrosion and fatigue resistance of these blades.
Conclusion
Our investigation conclusively demonstrates that sand casting foundry is a viable and cost-effective method for producing thin-wall axial blower blades made of ZG06Cr13Ni4Mo stainless steel. Through systematic simulation and practical validation, we established a robust horizontal casting process that meets all dimensional, mechanical, and NDT requirements. The key findings are:
- The horizontal casting orientation significantly reduces deformation and improves feeding compared to the vertical arrangement.
- Numerical simulation effectively predicted distortion and shrinkage, guiding the optimization of the sand casting foundry design.
- The use of chromite sand facing, ester-hardened sodium silicate binder, and AOD refining ensured defect-free castings.
- More than several hundred such blades have been successfully produced using this sand casting foundry process, resulting in substantial cost reduction and shorter lead times compared to conventional metal mold or precision casting methods.
This work enriches the application of sand casting foundry technology in the manufacturing of complex thin-walled components for the power generation industry. Future efforts will focus on further minimizing residual stresses and exploring automated process control to enhance repeatability.