The Vacuum Sealed Molding (V-Process) is a distinctive casting method characterized by its use of dry, unbonded silica sand and a thin plastic film to form molds. The mold strength is derived entirely from the differential pressure created by evacuating air from the sealed flask. This process offers significant advantages, including superior surface finish, high dimensional accuracy of the final casting part, reduced production costs, and minimal environmental impact, earning it the designation of “green casting.” This study focuses on the application of V-Process for manufacturing large, sparse-body casting parts. These parts are typically characterized by substantial planar dimensions, relatively low weight, and a high ratio of overall size to wall thickness. While initial process design is often successful for simpler geometries, the mass production of complex sparse-body structures can reveal persistent defects. This paper details the investigation and resolution of a recurring sand washing defect encountered during the high-volume production of such a casting part, employing numerical simulation as a primary diagnostic tool and subsequently validating the optimized process through actual production.
The casting part under investigation serves as a structural component in heavy engineering vehicles. Its core function necessitates a design comprising large, flat panels for strength and mounting platforms for assembly. A detailed structural and thickness analysis, as shown below, reveals key characteristics critical for process design.
The primary specifications and structural analysis of the casting part are summarized in Table 1. The part is produced in gray iron, grade HT200.
| Parameter | Value / Description |
|---|---|
| Overall Dimensions | 1479 mm × 1022 mm × 349 mm |
| Weight | 330 kg |
| Material | Gray Iron HT200 |
| Average Wall Thickness | 24.19 mm |
| Minimum Wall Thickness | 10 mm |
| Classification | Medium-sized, thick-walled, sparse-body casting |
| Key Features | Large planar areas, multiple mounting bosses and threaded holes. |
The significant planar expanse relative to its volume classifies this as a sparse-body casting part. This classification immediately informs several aspects of the gating system design, primarily the need to ensure adequate fluidity to fill the extensive, thin sections without creating mist runs, cold shuts, or incomplete filling.
The initial casting process was designed specifically for V-Process production. A summary of the key process parameters is provided in Table 2.
| Process Parameter | Specification |
|---|---|
| Molding Method | Vacuum Sealed Molding (V-Process) |
| Molding Material | Dry, unbonded silica sand |
| Parting Line | At the maximum contour in the thickness direction. |
| Flask Size | 2300 mm × 1600 mm × 600/850 mm (One part per flask) |
| Dimensional Tolerance | DCTG11 Grade |
| Weight Tolerance | MT11 Grade (330 kg ± 15 kg) |
| Machining Allowance | RMAG 8 Grade (8 mm on upper surfaces) |
| Shrinkage Allowance | 0.7% linear in all directions |
| Draft Angle | 0°35′ (increased wall thickness method) |
| Pattern Parting Allowance | 1 mm on the cope pattern |
The gating system was designed as a pressurized-open, top-pour, slow-fill type. This design aims to reduce the velocity of the molten metal as it enters the mold cavity, thereby minimizing turbulence and erosion of the fragile mold face. The choke area was placed at the junction of the sprue well and runner. Ceramic filters were incorporated in the runner for metal cleanliness. Six side risers were placed around the casting part for venting gases. Furthermore, the mold was tilted by 7°, elevating the side farthest from the sprue to aid metal flow and gas venting. Key gating calculations were performed to determine the pouring time and choke area.
The theoretical pouring time \( t \) (s) was estimated based on the casting weight \( G_L \) (kg) and an empirical coefficient \( S \) (s/kg\(^{1/2}\)) related to the section thickness:
$$ t = S \sqrt{G_L} $$
For a thick-walled casting part of this size, an \( S \) value of approximately 2.2 was selected, yielding a calculated pouring time of about 40 seconds.
The minimum choke area \( A_{choke} \) (cm²) was calculated using the basic fluid flow equation:
$$ A_{choke} = \frac{G_L}{\rho_L \cdot \mu \cdot t \sqrt{2g H_p}} $$
Where:
- \( \rho_L \) is the density of molten iron (~7000 kg/m³).
- \( \mu \) is the discharge coefficient (taken as 0.55 for this system).
- \( g \) is the acceleration due to gravity (9.81 m/s²).
- \( H_p \) is the effective metallostatic pressure head (m).
Based on the mold geometry and tilt, \( H_p \) was calculated to be approximately 0.35 m. Substituting the values yielded a choke area \( A_{choke} \approx 11.75 \text{ cm}^2 \).
The gating ratio was established as \( A_{sprue} : A_{runner} : A_{choke} : A_{ingate} = 2 : 1.5 : 1 : 2 \). This ensures the system is initially pressurized (sprue > runner > choke) to prevent aspiration, while the larger ingates relative to the choke help reduce flow velocity into the cavity.
Initial trial production of four prototype casting parts was successful. All parts met mechanical property specifications via test bars, dimensional accuracy via CMM scanning, and visual quality standards with no apparent defects. This validated the fundamental soundness of the process design for the casting part.
However, upon scaling to batch production of 50 units, significant quality issues emerged, as detailed in Table 3. A sand washing defect, predominantly located near the ingate farthest from the sprue, was the primary cause of rejection. This defect manifested as rough, irregular metal protrusions on the surface of the casting part, indicative of loose sand grains being incorporated into the solidifying metal.
| Defect Type | Quantity Rejected | Rejection Rate (%) | Primary Location |
|---|---|---|---|
| Sand Washing | 10 | 20% | Near farthest ingate from sprue |
| Cold Shut | 2 | 4% | Lower corners of the casting part |
| Gas Porosity | 1 | 2% | Scattered in upper sections |
| Total Rejects | 13 | 26% |
Initial corrective actions focused on mold strength. Mold hardness measurements consistently exceeded 90, indicating adequate compaction and vacuum-held strength according to V-Process standards. Reinforcement with embedded steel rods in the defect-prone area also failed to eliminate the problem, ruling out insufficient mold strength as the root cause for this specific casting part. This necessitated a deeper investigation into the fluid dynamics during mold filling.
A comprehensive numerical simulation of the filling process was conducted. The model consisted of approximately 20 million finite element cells. The boundary conditions and material properties used are listed in Table 4.
| Parameter | Setting |
|---|---|
| Mesh Elements | ~20 Million |
| Mold Material | Silica Sand (Initial Temp: 25 °C) |
| Alloy | HT200 (Actual Composition) |
| Pouring Temperature | 1430 °C |
| Metal-Mold HTC | 500 W/(m²·K) |
| Mold-Ambient Cooling | Air Cooling |
The simulation results provided a clear, time-resolved visualization of the filling sequence. The analysis confirmed that the gating system functioned as intended initially: metal flow was calm, with no severe turbulence, splashing, or air entrainment observed. The liquid front progressed smoothly and uniformly. However, a critical finding emerged from the analysis of metal flow velocity and flux at the ingates. Monitoring points were placed at the center of each of the four ingates (labeled Ingate 1 nearest the sprue to Ingate 4 farthest from the sprue). The metal mass flow rate \( \dot{m} \) (g/s) over time \( t \) (s) for each ingate revealed a non-uniform distribution, as plotted in the simulation output and quantified in Table 5 for the peak flow period.
| Ingate Number (Location) | Peak Mass Flow Rate, \( \dot{m}_{max} \) (g/s) | Time of Peak Flow (s) | Relative Flow Intensity |
|---|---|---|---|
| Ingate 1 (Near Sprue) | 1.19 | ~25 | High |
| Ingate 2 | 0.89 | ~26 | Medium |
| Ingate 3 | 0.86 | ~26 | Medium |
| Ingate 4 (Farthest from Sprue) | 1.23 | ~27 | Highest |
The data shows that Ingate 4, the one most distant from the sprue, experienced the highest peak flow rate \( \dot{m}_{max} \approx 1.23 \text{ g/s} \). This was approximately 43% higher than the flow rate in Ingates 2 and 3. Furthermore, this peak occurred later in the pouring sequence (~27 seconds), coinciding with a stage where the metal stream had stabilized and was imparting sustained kinetic energy onto a specific area of the mold wall. This perfectly correlated with the exclusive appearance of the sand washing defect on the casting part near Ingate 4. The simulation thus identified a localized over-velocity condition as a key contributing factor.
The fundamental mechanism of sand washing in V-Process is distinct from that in bonded sand molds. In V-Process, the mold surface integrity during pouring relies on a fragile, carbonaceous “skin” or char layer formed from the pyrolysis of the plastic film. This skin temporarily maintains the vacuum seal and contains the sand. The integrity of this skin is vulnerable to two main forces during the filling of a large, sparse-body casting part: 1) Direct hydrodynamic shear stress from high-velocity metal impingement, and 2) Buildup of back-pressure from gases generated by the film pyrolysis that cannot vent quickly enough. The numerical simulation confirmed the first factor: a localized high-velocity stream was persistently scouring the mold wall at the Ingate 4 entry point. The second factor is inherent to the location; being at the far end of a large cavity, it is a natural dead-end for gas flow, especially if venting (via side risers) is not optimally placed. The combined effect of sustained fluid shear and localized gas pressure spikes likely breached the fragile charred film, allowing sand grains to be dislodged and entrapped in the solidifying metal, resulting in the defective surface on the casting part.
The cold shut defects observed in two units were also explained by the simulation, which showed the meeting of two metal fronts in the lower corners of the casting part. This, coupled with the lower temperature of metal from the bottom of the ladle during the final stage of pouring, created the condition for cold shuts.
Based on the root cause analysis, the gating and venting system was redesigned with the following key modifications to protect the integrity of the mold skin for this specific sparse-body casting part:
- Relocation of Gating System: The entire gating system was moved to a central location on the casting part, adjacent to a large window structure. This fundamentally changed the flow pattern, eliminating long, unobstructed flow paths and the associated end-of-flow high-velocity condition. The incoming metal now enters a more confined central region, dissipating energy more evenly.
- Ingate Design Modification: The connection points between the ingates and the casting part were redesigned to present a larger target area and a less direct impingement angle, further reducing shear stress on the mold wall.
- Strategic Vent Placement: Two of the side risers were repositioned to the areas previously prone to defects (the far corners from the original sprue location). This provided dedicated, low-resistance escape paths for pyrolysis gases accumulating in those regions, preventing pressure buildup that could rupture the film char layer.
The modified design ensures a more balanced fill, lower localized velocities, and superior venting for the gases produced during the casting of this part.
The optimized process was implemented in a subsequent production batch of 55 casting parts. The results were definitive, as shown in Table 6. The sand washing defect was completely eliminated. Furthermore, the more balanced fill and improved thermal profile likely contributed to the disappearance of cold shuts. The overall rejection rate dropped to 0%, confirming the effectiveness of the changes.
| Production Batch | Quantity Produced | Quantity Rejected | Primary Defect (Sand Wash) | Rejection Rate |
|---|---|---|---|---|
| Initial Batch (Old Design) | 50 | 13 | 10 | 26% |
| Validation Batch (Optimized Design) | 55 | 0 | 0 | 0% |
This case study underscores the critical importance of understanding the unique failure mechanisms in V-Process casting, especially for challenging geometries like sparse-body casting parts. The defect was not due to conventional “weak mold” issues but to a combination of unfavorable fluid dynamics and inadequate venting for pyrolysis gases, both of which compromised the integrity of the essential film-derived skin. Numerical simulation proved to be an indispensable tool for diagnosing this problem, moving beyond trial-and-error. It pinpointed the precise location and cause of the high-velocity flow responsible for eroding the mold surface of the casting part. The successful optimization, which involved recentering the gating and reorienting the vents, demonstrates that effective solutions often require a holistic re-evaluation of the feeding and venting architecture relative to the casting part’s geometry. This approach significantly improved the yield and operational efficiency for producing this complex sparse-body casting part via V-Process.