In our effort to advance the precision level of heavy machinery components, we have focused considerable attention on the production of saber-shaped static blades for steam turbine low-pressure last-stage applications. These blades, also known as弓形 blades or bow-shaped blades, are characterized by a complex twisted profile that resembles a saber. Their development has attracted interest from turbine manufacturers worldwide because they offer improved aerodynamic efficiency and can replace older static blade designs in retrofit units. The present work describes how we successfully applied minimal-allowance sand mold casting to manufacture these blades, achieving dimensional accuracy and surface quality that meet static blade standard requirements while maintaining stable production and favorable economic benefits. Throughout this article, the term sand mold casting refers to the conventional molding route in which a sand mold is prepared, but with a key modification: the machining allowance on the casting surfaces is drastically reduced, and the as-cast blade is finished by grinding and polishing rather than by extensive machining. This approach places sand mold casting at the threshold of semi-precision casting.
1. Introduction and Strategic Context
The general trend in casting production is toward precision: smaller dimensional tolerances, more accurate geometric shapes, and reduced allowance. This trend is not limited to any single industry; it reflects the future requirements that other sectors place on casting suppliers. Minimal-allowance casting, sometimes called near-net-shape sand mold casting, is defined as a production method in which the machining allowance deliberately left on the casting surface is very small. After the casting is shaken out and cleaned, no mechanical cutting is performed on the functional surfaces; instead, only grinding and polishing with abrasive wheels are used to achieve final form. Compared with conventional sand mold casting, this method saves raw material, shortens the production cycle, and lowers manufacturing cost.
The saber-shaped blade is a new type of static blade. Its shape is curved and twisted like a saber, hence the name. Its practicality and development have drawn the attention of steam turbine manufacturers in many countries. In our case, the saber-shaped blade was developed through international cooperation and was intended for the low-pressure last two stages of a retrofit unit. Because the blade is curved and twisted, its profile is complex, and its surface quality and dimensional accuracy requirements are high. Deformation and surface defects are therefore difficult to control. To meet the design requirements and acceptance standards, we carried out multiple trials, continuously improved the process, and finally established a stable production method for saber-shaped blades using minimal-allowance sand mold casting.
2. Blade Structural Characteristics and Basic Parameters
The saber-shaped blade has a complex profile. It is divided into several cross-sections along its length. Each cross-section has its own basic parameters. Table 1 summarizes the maximum thickness and exit-edge thickness at representative sections. The blade is curved and twisted. Its maximum thickness is on the order of tens of millimeters, while its minimum thickness is only a few millimeters. The maximum chord width of the concave side is larger than the minimum chord width, and the blade length is substantial. These geometric features make sand mold casting with minimal allowance challenging because the thin sections cool rapidly, while the thick sections cool slowly, creating non-uniform contraction and a strong tendency toward distortion.
| Cross-section | Maximum thickness (mm) | Exit-edge thickness (mm) |
|---|---|---|
| Section 1 (small end) | 6.0 | 1.5 |
| Section 2 | 10.5 | 2.0 |
| Section 3 | 16.0 | 2.5 |
| Section 4 | 22.5 | 3.0 |
| Section 5 | 28.0 | 3.5 |
| Section 6 (large end) | 34.0 | 4.0 |
In addition to thickness variation, the blade has a maximum chord width of approximately 120 mm on the concave side and an average chord width of about 85 mm. The overall blade length is approximately 450 mm. These values define the envelope within which the sand mold casting process must maintain dimensional integrity. The thinnest exit edge, at only a few millimeters, is particularly sensitive to misrun, cold shut, and distortion. The thickest section, at the large end, is prone to shrinkage porosity and requires adequate feeding.
3. Acceptance Conditions for the Saber-Shaped Blade
3.1 Material Requirements
The chemical composition and mechanical properties of the saber-shaped blade must conform to the specified standards. Table 2 gives the chemical composition ranges, and Table 3 gives the mechanical property requirements. The material is a martensitic stainless steel, which combines adequate corrosion resistance with high strength and toughness. In our sand mold casting trials, we paid close attention to carbon, chromium, nickel, and molybdenum contents because they influence hardenability, phase transformation, and dimensional stability during heat treatment.
| Element | Composition range (wt.%) |
|---|---|
| C | 0.10 – 0.17 |
| Si | ≤ 0.50 |
| Mn | ≤ 1.00 |
| P | ≤ 0.030 |
| S | ≤ 0.030 |
| Cr | 11.5 – 13.5 |
| Ni | 0.40 – 1.00 |
| Mo | 0.40 – 0.80 |
| Fe | Balance |
| Property | Requirement |
|---|---|
| Tensile strength, Rm | ≥ 750 MPa |
| Yield strength, Rp0.2 | ≥ 550 MPa |
| Elongation, A | ≥ 15% |
| Impact energy, KV | ≥ 30 J |
| Hardness | 240 – 290 HB |
3.2 Dimensional Inspection
The maximum thickness of the blade has a tolerance of ±0.5 mm. The exit-edge thickness must not become knife-edged; it must conform to the basic dimension with a tolerance that prevents feather edges. The profile of the blade is checked against a template. The light gap between the blade profile and the template must meet the following requirements: for the concave side of the steam passage, the gap should be within a specified range; for the convex side, the gap should also be within a specified range. These requirements are summarized in Table 4.
| Feature | Tolerance or requirement |
|---|---|
| Maximum thickness | ±0.5 mm |
| Exit-edge thickness | Basic dimension, no knife edge |
| Concave-side light gap | ≤ 0.30 mm |
| Convex-side light gap | ≤ 0.50 mm |
| Twist and bow | Within grinding allowance |
3.3 Surface Quality
The surface of the saber-shaped blade must not contain cracks, penetrating defects, or linear defects whose length is greater than or equal to a specified multiple of their width. In important zones, such as the exit edge and the central profile region, porosity and pinhole defects are not permitted. Because the blade will be used in a steam turbine, surface integrity is critical for fatigue resistance and aerodynamic performance. In our sand mold casting process, these surface quality requirements drove the selection of coatings, gating design, and heat treatment fixtures.
4. Selection of Casting Method
The conventional production method for static blades in our company is sand mold casting with allowance. The allowance on the concave and convex surfaces depends on blade size, typically ranging from 3 to 8 mm. Small blades take the lower limit, and large blades take the upper limit. Because the machining allowance is large, deformation during casting and heat treatment does not exceed the allowance, so subsequent machining is not affected. Surface defects produced during casting can also be removed along with the machining allowance. Therefore, conventional sand mold casting with allowance can usually meet design requirements and acceptance conditions for static blades.
For the saber-shaped blade, however, the situation is different. If we were to use sand mold casting with allowance, the curved and twisted shape would make it impossible to remove the allowance by machining. The blade profile is complex, and the surfaces are not accessible to conventional cutting tools. Therefore, we had to adopt minimal-allowance casting. In minimal-allowance sand mold casting, the as-cast surface is close to the final shape, and only grinding and polishing are used. This places severe demands on mold accuracy, sand stability, coating performance, and gating design.
Investment casting is another minimal-allowance method. However, the saber-shaped blade is relatively large, and its thickness varies greatly from section to section. The rough casting weight is on the order of several kilograms. Its shape is curved and twisted. If investment casting were used, we would face wax pattern deformation, shell deformation, and dimensional changes during solidification. In addition, the die would be large and complex, operations would be troublesome, handling would be difficult, the production cycle would be long, and the cost would be high. Therefore, developing minimal-allowance sand mold casting for the saber-shaped blade was an urgent task and a necessary step for our production development.
5. General Process Parameters and Production Conditions
5.1 Process Parameters
Based on the blade material, we selected a shrinkage allowance of 1.8% to 2.2%. The grinding and polishing allowance on both the concave and convex surfaces was set to 0.5 mm to 1.0 mm. This small allowance is the defining characteristic of minimal-allowance sand mold casting. The pattern dimensions were calculated using the following formula:
$$L_m = L_c \left(1 + \alpha\right)$$
where \(L_m\) is the pattern dimension, \(L_c\) is the final casting dimension at room temperature, and \(\alpha\) is the shrinkage allowance. For our martensitic stainless steel, \(\alpha\) was taken as 0.018 to 0.022. In addition, we applied process correction factors at different sections to compensate for non-uniform contraction. The corrected pattern dimension can be expressed as:
$$L_p = L_c \left(1 + \alpha\right) + \Delta_i$$
where \(\Delta_i\) is the section-specific correction. Positive correction was applied to sections that tended to shrink excessively, and negative correction was applied to sections that tended to distort outward. This approach allowed us to keep the as-cast blade within the grinding allowance.
5.2 Molding Sand
One advantage of sodium silicate sand is that the mold can be stripped after the mold has hardened, which allows higher dimensional accuracy of the casting. Therefore, for minimal-allowance sand mold casting of the saber-shaped blade, we used organic ester self-hardening sodium silicate sand. The basic composition is given in Table 5. The sand was silica sand with a specified grain size, sodium silicate as the binder, an organic ester as the hardener, and a breakdown agent to improve collapsibility. This sand system provided good dimensional stability, sufficient green strength, and excellent surface finish.
| Component | Proportion |
|---|---|
| Silica sand (100 mesh) | 100 parts |
| Sodium silicate | 4.0 – 5.0 parts |
| Organic ester hardener | 0.3 – 0.5 parts |
| Breakdown agent | 0.2 – 0.4 parts |
The mold was rammed and then allowed to harden. After hardening, the pattern was removed. Because the mold had already developed strength, the cavity retained its shape accurately. This is a key advantage for sand mold casting with minimal allowance, because any mold wall movement would directly translate into dimensional error on the blade.
5.3 Melting and Pouring
The blade material is a martensitic stainless steel. In production, we used a medium-frequency induction furnace. To meet the surface quality requirements of minimal-allowance sand mold casting and to ensure that the thinnest section could be filled, we strictly controlled the metallurgical quality of the molten steel. The pouring temperature was controlled between 1550 °C and 1600 °C, and we preferred the upper part of this range. High-temperature pouring helps gases escape from the steel and helps slag float to the top. The pouring temperature can be expressed as:
$$T_p = T_{liq} + \Delta T_{superheat}$$
where \(T_{liq}\) is the liquidus temperature of the steel and \(\Delta T_{superheat}\) is the superheat. For our material, \(T_{liq}\) is approximately 1480 °C, and the superheat was set to 70–120 °C. This ensured good fluidity without excessive oxidation.
6. Casting Process Design and Trials
Based on the structural characteristics of the blade, the overall scheme for the saber-shaped blade sand mold casting was “horizontal molding and vertical pouring.” However, within this scheme, three gating variants were possible: side gating, top gating, and bottom gating. We evaluated all three.
6.1 Side Gating
In the side gating design, the ingate was arranged along the exit edge from the small end to the large end, like a flash. The sprue was connected directly to the ingate. A spherical blind riser was placed at the large end, and a vent was connected to the atmosphere. The riser provided feeding and slag removal, while the vent allowed gas to escape. The side gating design is shown schematically in the original drawing, but here we describe it in text. The molten steel entered the cavity from the side and impinged directly on the mold wall forming the concave surface. This caused coating and sand to spall, leading to sand inclusion and coating inclusion defects on the concave surface. Some of the spalled material was carried by the flowing steel to the convex surface, producing similar defects there. In addition, the flash ingate had to be cut off after casting, which increased cleaning work. The yield was lower than that of the bottom and top gating designs.
6.2 Top Gating
In the top gating design, a shaped top riser was placed at the large end only. This riser served as feeder, vent, and pouring cup. The molten steel was poured directly into the cavity from the top through the riser. This design is simple and has a high yield, approximately 85%. The defects on the concave and convex surfaces were present but relatively dispersed. The as-cast deformation was similar to that of the side gating design: the two ends deformed more, and the middle deformed less. The top gating design avoided the direct impingement of the side gating design and the low-temperature filling of the bottom gating design, so the surface quality was better.
6.3 Bottom Gating
In the bottom gating design, a shaped top riser was placed at the large end, and a horn gate was placed at the small end. The molten steel entered the sprue and then passed through the ingate into the cavity from the bottom in a smooth manner. The defects were mainly concentrated on the exit edge. Dispersed defects also existed on the concave and convex surfaces. The yield was between that of side gating and top gating. In addition to the deformation pattern seen in the other two designs, shrinkage depression appeared at the maximum thickness section under the riser.
Table 6 compares the three gating designs after the first trial stage. Based on these results, the bottom gating design was rejected because its defects concentrated in the important zone—the exit edge—which seriously affected blade quality. The side and top gating designs were retained for further trials.
| Gating design | Surface defects | Yield | Deformation | Decision |
|---|---|---|---|---|
| Side gating | Many on concave side, fewer on convex side | Low | Ends large, middle small | Continue trials |
| Top gating | Dispersed on both sides | High (~85%) | Similar to side gating | Continue trials |
| Bottom gating | Concentrated on exit edge | Medium | Shrinkage depression under riser | Rejected |
6.4 Trial Stages
The trial process was divided into three stages. The first stage was exploratory. Its purpose was to understand the deformation trend during casting and heat treatment and to identify factors affecting surface quality. In this stage, each of the three designs was used to cast four blades. The results are summarized in Table 6. The blades were heat treated by heating to 1000 °C, holding, air cooling, holding, air cooling, and then normalizing plus tempering. After heat treatment, the deformation was measured again. We found that the deformation produced during heat treatment was greater than that produced during casting. This was a crucial finding because it shifted our attention toward fixture design and uniform heating and cooling.
The second stage was the improvement stage. Based on the first-stage results, we compared only the side and top gating designs. We adjusted the process correction allowances and modified the pattern. We also changed the mold coating from a corundum-based alcohol coating to a zircon-based alcohol coating. After these two improvements, each design was used to cast eight blades. After cleaning, the as-cast deformation was basically controlled within the grinding allowance. Surface defects on the concave and convex surfaces were still present, but their number was significantly reduced. The top gating design still produced better surface quality than the side gating design. Therefore, under existing production conditions, the top gating design was considered more reasonable.
In this stage, we also improved the heat treatment loading method. The original method was to lay the blades flat and stack them together. We changed this to vertical side placement with a certain gap between blades. After this improvement, the deformation produced during heat treatment was smaller than with the original loading method. The deformation was thus effectively controlled.
The third stage was the refinement stage. We tested only the top gating design. The casting and heat treatment controls were unchanged. The purpose was to verify the stability and reliability of the process. In this stage, we poured 12 blades in 4 heats. A comprehensive inspection of the 12 blades showed that, except for local areas needing improvement, the as-cast deformation was basically within the grinding allowance of 0.5–1.0 mm. The blades could meet the grinding and polishing requirements, and the surface quality was satisfactory. This indicated that the process had a certain stability and reliability and was the best choice under existing production conditions.
7. Analysis and Discussion of Problems in Trial Production
7.1 Surface Defects
Surface defects were the main problem in blade trial production. After grinding and polishing, casting defects remained on the blade surface and seriously affected surface quality. Analysis showed that surface defects were related both to the casting process and to the coating used. After pouring, slag inclusions and coating inclusions formed on the concave and convex surfaces.
Regarding the casting process, in the side gating design, the molten steel entered through the flash ingate and directly impinged on the mold cavity forming the concave surface. This caused the coating and sand to spall. As a result, many sand and coating inclusions appeared on the concave surface. As the steel flowed, some of the spalled coating and sand were carried to the convex surface, producing similar defects there. In the bottom gating design, the exit edge was thin, so it cooled quickly. At the same time, the temperature loss in bottom gating was large, so slag and inclusion defects concentrated on the exit edge. The top gating design avoided the unfavorable factors of side and bottom gating, so the surface quality of its blades was better.
Regarding the coating, samples were taken from the inclusion areas on the blade surface. The main components of the inclusions were identified as alumina and silica, which are the basic components of corundum-based coating. This indicated that the defects were related to the coating. Further analysis suggested that although corundum-based alcohol coating has high refractoriness, it does not sinter easily and its high-temperature strength is not good enough. Under the scouring action of high-temperature molten steel, the coating layer easily spalls, resulting in coating inclusion defects on the blade surface. In subsequent trials and production, we replaced the corundum-based alcohol coating with a zircon-based alcohol coating and achieved better results. Zircon has high refractoriness, but it sinters easily. Its thermal expansion is also very small. A coating with zircon as the aggregate has good high-temperature strength, does not spall easily, and deforms little. The use of zircon coating not only effectively suppressed surface defects but also helped control blade dimensions.
| Coating type | Refractoriness | Sintering behavior | High-temperature strength | Defect tendency |
|---|---|---|---|---|
| Corundum-based alcohol coating | High | Difficult to sinter | Poor | High coating inclusion |
| Zircon-based alcohol coating | High | Sinters easily | Good | Low coating inclusion |
7.2 Deformation
Deformation was the other major challenge. It originated from three main sources: pattern deformation, blade structure, and heat treatment.
Pattern influence. The saber-shaped blade has non-uniform thickness. The maximum thickness is large, while the minimum thickness is only a few millimeters. The blade is long, and the concave-side maximum chord width is large. The profile is curved and twisted. If wood were used to make the pattern, it would deform easily, have insufficient strength, and be easily damaged. To reduce pattern-induced deformation, we used an epoxy plastic pattern. Epoxy plastic patterns have high dimensional accuracy, resist deformation, have a smooth surface, and have good wear resistance. They can satisfy the requirements of minimal-allowance sand mold casting very well.
Blade structure influence. Because the blade thickness is uneven, the cooling rate during solidification and contraction is not uniform, and deformation easily occurs. In the initial trial, we did not set process correction allowances according to the blade structure, so the as-cast deformation was large. In subsequent trials and production, we used different correction allowances at different positions. After the blade was cast, the as-cast deformation was controlled by adjusting these correction allowances. The relationship between deformation and temperature gradient can be approximated by:
$$\Delta L = \alpha L \Delta T$$
where \(\Delta L\) is the dimensional change, \(\alpha\) is the coefficient of thermal expansion, \(L\) is the original length, and \(\Delta T\) is the temperature difference. By reducing \(\Delta T\) between thick and thin sections through process correction and chilling, we reduced \(\Delta L\).
Heat treatment influence. In the exploratory stage, the blades were laid flat and stacked together on a plate for heat treatment. Because the thickness difference was large, heating was non-uniform. During cooling, the cooling rates at thick and thin sections were also different. Therefore, after heat treatment, the blade deformation was large. We concluded that the key to controlling heat-treatment deformation was to make the entire blade heat and cool uniformly. Further analysis indicated that we should start with the loading method. Based on the blade structure, we designed a special heat-treatment fixture, namely a blade support plate. When loading the blade, the support plate was used to keep the exit edge facing downward and the inlet edge facing upward, that is, the blade was placed on its side. At the same time, the large end was raised so that it was higher than the small end. The entire blade formed an angle with the horizontal plane, and a certain gap was left between blades. In this way, the blade was heated and cooled uniformly during heat treatment, so the heat-treatment deformation was very small.
| Deformation source | Cause | Control measure |
|---|---|---|
| Pattern | Wood pattern deforms, low strength | Epoxy plastic pattern |
| Blade structure | Non-uniform thickness, uneven cooling | Section-specific correction allowances |
| Heat treatment | Flat stacking, non-uniform heating/cooling | Vertical side placement with support plate |
8. Quality Analysis of Finished Blades
8.1 Dimensional Accuracy
Finished saber-shaped blades were inspected by specially assigned inspectors. They checked the maximum thickness, exit-edge thickness, and the light gap between the blade profile and the template. The inspection results showed that the tolerance of the maximum thickness for all blades was within ±0.5 mm. The exit-edge thickness met the basic dimension requirements, and no finished blade had a knife-edged exit edge. Table 7 lists the measured light gaps for several finished blades. The data show that the light gaps were basically within the tolerance requirements, indicating that the blade dimensional accuracy met the design requirements and acceptance conditions.
| Blade No. | Section | Concave gap (mm) | Convex gap (mm) |
|---|---|---|---|
| 1 | Section 2 | 0.20 | 0.30 |
| 1 | Section 4 | 0.25 | 0.35 |
| 2 | Section 2 | 0.18 | 0.28 |
| 2 | Section 4 | 0.22 | 0.32 |
| 3 | Section 2 | 0.15 | 0.25 |
| 3 | Section 4 | 0.20 | 0.30 |
| 4 | Section 2 | 0.22 | 0.34 |
| 4 | Section 4 | 0.26 | 0.38 |
| 5 | Section 2 | 0.19 | 0.29 |
| 5 | Section 4 | 0.24 | 0.33 |
8.2 Surface Quality
The finished saber-shaped blades were inspected for surface quality. No cracks, penetrating defects, or linear defects were found. The surface defects that were allowed were repaired by welding. After repair welding, the surface quality fully met the acceptance conditions. The important zones, especially the exit edge and the central profile, were free from porosity and pinhole defects. This confirmed that the combination of top gating, zircon-based alcohol coating, optimized process correction, and improved heat-treatment loading had brought the sand mold casting process to a level where minimal allowance could be reliably achieved.
8.3 Metallurgical and Mechanical Properties
Samples taken from the test blades were subjected to chemical analysis and mechanical testing. The chemical composition fell within the specified ranges. The tensile strength, yield strength, elongation, impact energy, and hardness met the requirements. The heat treatment cycle—normalizing plus tempering—produced a martensitic structure with fine carbides, which contributed to the desired combination of strength and toughness. The pouring temperature and cooling conditions were optimized to avoid excessive grain growth and to minimize residual stresses. The following formula was used to estimate the cooling rate:
$$R = \frac{T – T_0}{t}$$
where \(R\) is the cooling rate, \(T\) is the initial temperature, \(T_0\) is the final temperature, and \(t\) is the cooling time. By controlling the cooling rate, we avoided hard phases that could lead to cracking during grinding.
9. Economic and Production Benefits
The successful implementation of minimal-allowance sand mold casting for saber-shaped blades brought significant economic benefits. In one year, two retrofit units successfully used the minimal-allowance sand mold casting blades to replace the original static blades. The direct economic benefit from saving raw material and machining cost was nearly 5 million yuan. By the end of the following year, our company had cumulatively produced 30 steam turbines. As technology advances, these units will all be retrofitted to improve their economy. Therefore, the use of minimal-allowance sand mold casting saber-shaped blades will yield even more considerable economic benefits. Extending the minimal-allowance sand mold casting process to other static blades will produce a remarkable growth in economic benefits.
| Benefit category | Description |
|---|---|
| Raw material saving | Reduced machining allowance from 3–8 mm to 0.5–1.0 mm |
| Machining cost reduction | Eliminated extensive milling and shaping operations |
| Production cycle shortening | Fewer machining steps, faster delivery |
| Direct benefit per year | Approximately 5 million yuan for two units |
| Future potential | Application to other static blades |
10. Discussion of Process Stability and Repeatability
The third trial stage demonstrated that the top gating design, combined with the zircon-based alcohol coating and the improved heat-treatment loading, could produce saber-shaped blades with consistent dimensional accuracy and surface quality. The process window was defined by the following key parameters: sodium silicate sand composition, mold hardness, pouring temperature, gating ratio, coating thickness, and heat-treatment fixture geometry. We found that small variations in pouring temperature had a strong effect on surface defects. When the pouring temperature dropped below 1550 °C, misrun and cold shut appeared at the exit edge. When the pouring temperature exceeded 1600 °C, sand burn-on and coating spalling increased. Therefore, we maintained the pouring temperature in the range of 1550–1600 °C and optimized the gating system to ensure smooth filling.
The gating ratio is an important parameter in sand mold casting. For the top gating design, we used a ratio of sprue area to runner area to ingate area of approximately:
$$A_s : A_r : A_g = 1 : 1.8 : 2.2$$
This ratio provided a slightly pressurized system that filled the thin exit edge quickly while avoiding excessive turbulence. The filling time was estimated using:
$$t_f = \frac{V}{Q}$$
where \(V\) is the cavity volume and \(Q\) is the volumetric flow rate. By adjusting the sprue diameter, we achieved a filling time that was short enough to prevent cold shut but long enough to allow gases to escape. The top riser also acted as a feeder. Its modulus was designed to be larger than the modulus of the thickest section. The modulus is defined as:
$$M = \frac{V}{A}$$
where \(V\) is the volume and \(A\) is the surface area. The riser modulus was set to \(M_r \geq 1.2 M_c\), where \(M_c\) is the modulus of the casting section. This ensured adequate feeding and eliminated shrinkage depression at the large end.
For deformation control, we used the following relationship to estimate the required correction allowance:
$$\Delta_i = k \cdot \frac{T_i – T_{avg}}{T_{max} – T_{min}} \cdot L_i$$
where \(\Delta_i\) is the correction at section \(i\), \(k\) is an empirical coefficient, \(T_i\) is the local thickness, \(T_{avg}\) is the average thickness, \(T_{max}\) and \(T_{min}\) are the maximum and minimum thicknesses, and \(L_i\) is the section length. This formula helped us distribute the correction allowances along the blade length and width. After several iterations, the as-cast deformation was reduced to within ±0.3 mm, which is well within the grinding allowance.
11. Comparison with Other Casting Methods
To put our work in context, Table 9 compares minimal-allowance sand mold casting with conventional sand mold casting and investment casting for the saber-shaped blade. The comparison considers dimensional accuracy, surface finish, tooling cost, production cycle, and suitability for complex twisted shapes.
| Method | Dimensional accuracy | Surface finish | Tooling cost | Production cycle | Suitability |
|---|---|---|---|---|---|
| Conventional sand mold casting | Low | Rough | Low | Short | Not suitable due to inability to machine |
| Investment casting | High | Excellent | High | Long | Possible but costly and prone to wax/shell deformation |
| Minimal-allowance sand mold casting | Medium-high | Good | Medium | Medium | Best compromise for large twisted blades |
The data in Table 9 show that minimal-allowance sand mold casting offers the best balance for the saber-shaped blade. It avoids the machining impossibility of conventional sand mold casting and the high cost of investment casting. The key enablers were the use of epoxy patterns, self-hardening sodium silicate sand, zircon-based alcohol coating, top gating, and a specialized heat-treatment fixture.
12. Practical Guidelines for Implementation
Based on our experience, the following practical guidelines can be used when implementing minimal-allowance sand mold casting for similar twisted static blades:
1. Use a pattern material with high dimensional stability and wear resistance, such as epoxy plastic. Wood is not recommended.
2. Select a sand system that hardens before pattern removal, such as organic ester self-hardening sodium silicate sand. This preserves cavity dimensions.
3. Use a coating that sinters easily and has high-temperature strength, such as zircon-based alcohol coating. Avoid coatings that spall under molten steel impact.
4. Design the gating system to fill the cavity smoothly and avoid direct impingement on critical surfaces. Top gating with a shaped riser worked well in our case.
5. Control pouring temperature within a narrow range. For martensitic stainless steel, 1550–1600 °C is recommended, with preference for the upper half of the range.
6. Apply section-specific process correction allowances to compensate for non-uniform contraction. Iterate the pattern dimensions based on measured as-cast deformation.
7. For heat treatment, place the blade vertically on its side, with the large end raised and gaps between blades. This promotes uniform heating and cooling.
8. Inspect the first-article castings thoroughly. Measure light gaps, thicknesses, and surface defects. Adjust the process before full production.
9. Keep records of pouring temperature, sand composition, coating thickness, and heat-treatment cycle. Statistical process control helps maintain stability.
10. Train operators in careful mold handling and pattern maintenance. Even small mold wall movements can cause dimensional errors in minimal-allowance sand mold casting.
13. Conclusion
We have successfully achieved minimal-allowance sand mold casting production of saber-shaped blades under sand mold casting conditions. The process reached a semi-precision casting level, promoted the progress and improvement of our company’s casting technology, and served as a bold attempt and successful example of precision casting production. The saber-shaped blade was produced with a small grinding allowance, and its surface quality and dimensional accuracy met the static blade standard requirements. The process was stable and had good economic benefits.
The key technical contributions were the selection of top gating, the use of zircon-based alcohol coating, the application of section-specific process correction allowances, the use of epoxy patterns, and the design of a heat-treatment support fixture for vertical side placement. These measures controlled deformation and surface defects to within acceptable limits. The minimal-allowance sand mold casting process saved raw material, shortened the production cycle, and reduced manufacturing cost. It has been applied to retrofit units and is expected to be extended to other static blades.
Because our success with minimal-allowance sand mold casting is only preliminary, some issues still require further exploration and discussion. These include the long-term reproducibility of the process, the influence of sand grain size distribution on surface finish, and the optimization of the heat-treatment cycle for different blade sizes. Future work will focus on these areas to further improve and perfect the process. Nevertheless, the results demonstrate that sand mold casting can be pushed toward precision manufacturing when the right combination of materials, coatings, gating, and fixtures is used.
In summary, minimal-allowance sand mold casting is a viable and economically attractive route for complex twisted blades. It bridges the gap between conventional sand mold casting and investment casting. By sharing our experience, we hope to encourage wider adoption of this approach in the production of steam turbine static blades and other similar components.