In the production of wedge-type rigid-seal gate valve bodies, I have successfully implemented a lost foam casting (EPC) method that embeds a stainless steel valve seat directly into the foam pattern, enabling integral casting. The valve body has an oval shape with dimensions approximately φ220 mm × 229 mm, a mass of 17 kg, and is made of ductile iron QT450-10. The working pressure is 1.6 MPa. The valve seat material is 2Cr13 or 1Cr18Ni9Ti stainless steel. This paper describes the entire process of lost foam castings for this component, including pattern making, coating, molding, pouring, and resulting quality. My aim is to share practical insights into achieving metallurgical bonding between the stainless steel seat and the ductile iron body, which is critical for leak-proof performance in hard-seal gate valves.
The key innovation in these lost foam castings lies in the pre-placement of a machined stainless steel ring into the foam pattern before assembly. During pouring, the high-temperature iron melt surrounds the stainless steel insert, causing partial melting at the interface and forming a strong metallurgical bond. This eliminates the need for separate machining or welding of the valve seat, reduces production time, and improves sealing reliability. The following sections detail each step of the lost foam castings process.
Pattern Fabrication for Lost Foam Castings
I used STMMA (styrene-methyl methacrylate) beads with a pre-expanded density of 23–26 kg/m³. The foam pattern for the valve body was formed in a single-piece mold with pneumatic core pulling. After pattern molding, the assembly was dried for 120 hours. The stainless steel valve seat (pre-treated by degreasing and preheating) was then embedded into the dried foam pattern. To ensure proper feeding, I designed the gating system with an ingate cross-section of 40 mm × 13 mm and a runner cross-section of 50 mm × 40 mm. Each cluster consisted of eight patterns, assembled as shown in a schematic representation during the lost foam castings setup. A zircon-based coating (ZL-1 type, composed of mullite and quartz powder) was mixed to a density of 60 ± 2 Bé and applied in two layers, each dried thoroughly before the next application. The coating thickness was controlled to 1.5–2.0 mm to prevent metal penetration and ensure smooth surface finish of the lost foam castings.
Molding and Pouring of Lost Foam Castings
The molding was carried out on an automated lost foam casting production line. Each flask contained two clusters, with flask dimensions of 1500 mm × 1200 mm × 1100 mm. Vacuum was applied from five sides (bottom and four corners) to maintain uniform negative pressure gradient within the flask. A two-stage adjustable rain-type sand filling device ensured even sand compaction. The vibrating table, equipped with hydraulic clamping, operated at high frequency and low amplitude to avoid pattern deformation, with the vibration frequency tuned to avoid resonance between the flask and the table. The sand used was dry silica sand with AFS grain fineness number 40–50.
For melting, I used a 1.5-ton medium-frequency induction furnace. The iron composition before treatment and after treatment is summarized in Table 1. The melt was superheated to 1550–1580°C, then treated using a single-step nodularization process with MgFeSi alloy and inoculant. The pouring temperature was maintained at 1500–1520°C at the point of pouring. The vacuum level during pouring was set to 0.03–0.05 MPa, and the vacuum was maintained for 5 minutes after completion of pouring to ensure complete decomposition of the foam pattern and removal of gaseous products. These parameters are critical for successful lost foam castings of ductile iron with embedded inserts.
| Element | C | Si | Mn | P | S | Mg | RE | Ti | Fe |
|---|---|---|---|---|---|---|---|---|---|
| Base iron | 3.74 | 1.58 | 0.246 | 0.0391 | 0.0147 | 0.00113 | 0.003 | 0.0406 | Balance |
| Final casting | 3.57 | 2.67 | 0.252 | 0.0398 | 0.0136 | 0.0325 | 0.0127 | 0.0427 | Balance |
Quality Evaluation of Lost Foam Castings
Visual Inspection
After shakeout, cleaning, and shot blasting, the valve body castings exhibited a smooth surface with sharp edges and clear contours. No defects such as cold shuts, sand adhesion, or gas porosity were observed. The small bolt holes (φ19 mm on the flanges and φ14 mm on the center flange) were cast cleanly with good dimensional accuracy. Figure 1 shows a sectioned valve body after wire-cut machining, revealing the integrity of the embedded stainless steel seat.
The above photograph illustrates a typical cross-section of the lost foam castings after cutting, showing the intimate contact between the stainless steel ring and the ductile iron matrix.
Hydrostatic Testing and Mechanical Properties
All castings were subjected to hydrostatic testing using a ZZB valve-specific pressure tester. A water pressure of 2.4 MPa was applied, and every valve body passed without any leakage. The tensile test results from specimens cut from the castings (and separately cast test bars) are given in Table 2. The average tensile strength reached 469.3 MPa, and elongation was 18.86%, exceeding the QT450-10 specification. These excellent mechanical properties are attributed to the optimized chemical composition and controlled cooling rate achieved during lost foam castings.
| Property | Value | Standard (QT450-10) |
|---|---|---|
| Tensile strength (MPa) | 469.3 | ≥450 |
| Elongation (%) | 18.86 | ≥10 |
| Hardness (HB) | 170–210 | – |
Microstructural Examination
Metallographic samples were prepared from the interface region between the stainless steel seat and the ductile iron body. Using a GX40 optical microscope and SRMAS quantitative image analysis system, I observed the microstructure at various magnifications. The stainless steel seat showed a clear melting zone at its edge, indicating strong metallurgical bonding. The microstructure gradually changed from the interface outward: a narrow transition zone (combined zone), followed by a layer of pearlite (approximately 76.5% pearlite, 16.0% cementite, 7.5% ferrite), and then the typical ferrite-pearlite matrix of ductile iron. At the interface itself, a thin layer of gray iron (pearlite 65.26%, ferrite 32.39%) was detected, which is a result of localized chilling by the stainless steel insert and subsequent recalescence. This is a common phenomenon in lost foam castings when massive inserts are used. The nodularity of graphite in the valve body was rated as grade 3 (about 87–89%), with an average graphite nodule diameter of grade 7. This confirms that the thermal conditions during lost foam castings were favorable for proper graphite spheroidization even near the insert.
Discussion: Advantages of Embedded Seat Design in Lost Foam Castings
The integration of a stainless steel valve seat directly into the lost foam castings eliminated the need for separate machining and assembly of the seat, reducing processing time and cost. The metallurgical bond achieved ensured that no leakage paths exist at the seat-body interface, which is a common failure mode in conventional mechanically fastened seats. Moreover, the design is flexible: different seat materials (e.g., 2Cr13, 1Cr18Ni9Ti) can be embedded depending on service requirements. The lost foam castings process provided excellent dimensional accuracy for the small bolt holes and flanges, further minimizing post-casting machining.
The key process parameters for these lost foam castings are summarized in Table 3. The combination of vacuum level, pouring temperature, and sand compaction is crucial for avoiding defects such as misruns or shrinkage porosity. The use of STMMA beads with controlled pre-expanded density ensured a low ash residue and smooth foam degradation.
| Parameter | Value |
|---|---|
| Pattern material | STMMA beads, pre-expanded density 23–26 kg/m³ |
| Coating | Mullite + quartz powder, density 60±2 Bé, two layers |
| Pouring temperature | 1500–1520°C (from melt at 1550–1580°C) |
| Vacuum level | 0.03–0.05 MPa (maintained 5 min after pour) |
| Sand type | Dry silica sand, AFS 40–50 |
| Vibration frequency | High frequency (50–60 Hz), low amplitude (0.3–0.5 mm) |
| Pouring time per cluster | 8–12 seconds |
Conclusions
Through systematic development, I have demonstrated that lost foam castings can successfully produce DN100 wedge-type rigid-seal gate valve bodies with embedded stainless steel seats. The metallurgical bond formed during pouring ensures leak-free operation under hydrostatic test pressure of 2.4 MPa. The mechanical properties exceed QT450-10 requirements. The process is economical, repeatable, and applicable to a wide range of valve sizes and seat materials. By careful control of pattern density, coating, vacuum, and pouring parameters, defect-free lost foam castings are achieved consistently. This technology significantly advances the manufacturing of hard-seal gate valves by reducing machining steps and improving sealing integrity. Future work may explore optimization of the transition zone microstructure to further enhance bond strength in large-diameter lost foam castings.