In my years of working with the vacuum lost foam casting process, I have faced many casting defects, but none has been as persistent and challenging as the elliptical deformation of ductile iron pipe fittings. When producing small and medium sized fittings, especially in high volume runs, the problem becomes increasingly severe. Through careful observation, systematic experimentation, and process refinement, I have developed and validated an effective countermeasure known as the wet sleeve sand ring technique. In this article, I will share my practical experience, the root cause analysis, the development of the improvement plan, and the quantitative results that ultimately brought the ovality ratio down to a stable level of less than three percent. My goal is to provide engineering insights that can be applied to similar thin-walled circular castings produced by lost foam casting.
The vacuum lost foam casting (also known as expendable pattern casting) process is widely used in the production of ductile iron pipe fittings due to its near-net-shape capability and excellent surface finish. However, one recurring problem is the loss of circularity at the socket and spigot ends. The deformation appears as an elliptical cross-section, which not only violates dimensional specifications but also creates difficulties in subsequent machining and assembly. The larger the batch size, the more frequently the defect appears. From the making of the EPS (expandable polystyrene) pattern, through coating dipping and drying, to molding and compaction, every stage has the potential to induce ovality. Based on my practical production records and a series of comparative experiments, I have established a process improvement that effectively controls the deformation of the socket and spigot ends.
In our production facility, we employ vacuum lost foam casting to manufacture ductile iron pipe fittings ranging from DN400 to DN2600. Among these, the DN400 to DN1200 specifications are produced on an automated continuous line. To understand the severity and distribution of the ovality defect, I tracked the proportion of non-circular parts for each specification over a period of continuous production. The results are shown in Table 1.
| Nominal Diameter (DN) | Number of Parts Inspected | Oval Parts Count | Percentage of Ovality (%) |
|---|---|---|---|
| 400 | 1250 | 18 | 1.44 |
| 600 | 1120 | 31 | 2.77 |
| 800 | 978 | 46 | 4.70 |
| 1000 | 854 | 55 | 6.44 |
| 1200 | 720 | 62 | 8.61 |
The trend in Table 1 clearly shows that the ovality percentage increases with nominal diameter. This suggests that larger pipe fittings have lower structural stiffness relative to their size, making them more vulnerable to deformation. At this stage, I suspected that the main influencing factors were the initial roundness of the foam pattern, the rigidity of the model during handling, and the mechanical effects of sand filling and vibration during molding. To isolate these factors, I conducted a series of measurements and observations on the pattern condition and the molding behavior.
Analysis of Ovality Characteristics and Causes
Pre-Molding Pattern Roundness
In the original process, we used a dry method of installing resin sand rings to maintain roundness. After the coating layer on the pipe fitting pattern was dried, but before assembling the pattern group, we manually fitted pre-cured resin sand rings onto the socket and spigot ends. The pattern group was then assembled outside the flask and lifted into the molding box for sand filling and compaction. This dry sleeve method was supposed to provide external support during the molding operation. However, when I examined the pattern groups before entering the flask, I found two major categories of problems.
First, the coating layer on the pattern surface was not perfectly smooth. Due to natural leveling effects, there were local protrusions, pinholes, or coarse lumps. When the sand ring was slipped over the coated surface, these irregularities created local gaps or uneven contact. As a result, the support provided by the sand ring was non-uniform, and the pattern could easily deform under the pressure of the molding sand. Second, during the drying process, a small number of patterns had already lost their roundness. The dried coating layer became hard and brittle. When the dry sand ring was forced over the pattern, the coating was sometimes cracked or scratched by the edge of the sand ring. Figure 3 in the original study shows such coating cracking, but I will not reproduce that figure here; instead, I will describe the phenomenon. The presence of cracks weakened the mechanical integrity of the coating and caused local stress concentrations, which worsened the deformation.
To quantify the initial roundness of the pattern, I measured the socket end of several patterns after the dry sand ring had been installed. The results, shown in Table 2, confirmed that the yellow foam pattern itself already exhibited a certain degree of ovality. More importantly, the ovality ratio of the pattern was strongly correlated with the final ovality ratio of the cast product. Therefore, controlling ovality must start from the pattern itself.
| Specification | Number of Patterns Measured | Average Ovality of Pattern (%) | Maximum Ovality of Pattern (%) |
|---|---|---|---|
| DN400 | 40 | 0.92 | 1.35 |
| DN600 | 38 | 1.54 | 2.10 |
| DN800 | 36 | 2.31 | 3.02 |
| DN1000 | 35 | 2.95 | 3.87 |
| DN1200 | 32 | 3.68 | 4.55 |
Deformation Observed During Molding and Compaction
When I observed the molding process closely, I noticed several events that contributed to the ovality. Because the entire pattern group was assembled before being lifted into the flask, the group experienced significant shaking during transport. Some sand rings, which were attached with dry contact only, had a tendency to become loose or tilted. If a worker did not notice this, the pattern would enter the molding process in an imperfectly supported state. Furthermore, during the initial low-frequency, low-amplitude vibration that accompanies sand filling, I could visually detect that some sand rings rotated slightly around the pattern. This relative motion prevented the sand ring from staying firmly in contact with the socket end, and thus the ring no longer provided reliable support. The horizontal sand rings, in particular, tended to rotate under the vibration, as illustrated in Figure 5 of the original report. It became evident that the dry sleeve sand ring method could not guarantee the roundness of the pattern at the time of compaction.
After casting and shakeout, I measured every oval part and marked the direction of the elliptical major and minor axes on the pipe body. I then reconstructed the position of each pipe fitting inside the flask. The results were astonishing and highly systematic. More than eighty percent of the deformed fittings exhibited the same orientation relative to the molding sand box. The minor axis of the ellipse was aligned with the direction connecting the two vibrators mounted on the side walls of the flask, while the major axis was perpendicular to that direction. This pattern strongly suggested that the vibration energy from the two side-mounted motors creates a sand flow that compresses the pattern along the vibrator axis. In other words, the sand in the central region between the two vibrators is dynamically compacted more vigorously, producing a local pressure concentration on the pattern surface. Additionally, I found that the socket ends located in the upper part of the mold showed a higher degree of ovality than those in the lower part. This is because the upper areas are less confined by the surrounding sand and therefore more prone to movement during vibration.
Based on all these observations, I concluded that in order to control ovality without changing the existing molding equipment and process, we needed to solve three critical problems. First, we had to ensure the roundness of the foam pattern itself and of the sand ring. Second, we had to achieve a firm and intimate bond between the sand ring and the pattern so that the assembly could withstand transport, lifting, and sand filling without separation. Third, during the vibration compaction step, the sand ring must remain locked to the pattern, maintaining continuous support against the external pressure of the sand. This realization led me to abandon the dry sleeve method and develop a wet sleeve process in which the sand ring is applied while the coating is still wet.
Development of the Wet Sleeve Sand Ring Technique
The wet sleeve sand ring technique is conceptually simple but requires careful execution. In the original dry method, the sand ring was fitted after the pattern had been coated and dried. In the new wet method, I reversed the order. After the final dip-coating operation, while the coating was still wet and before the surface began to form a dry skin, I placed the pre-cured resin sand ring over the socket and spigot ends. The gap between the sand ring and the pattern was filled by the wet coating, which acted as an adhesive and as a gap-filling medium. The entire assembly was then placed in the drying oven. As the coating dried, it shrank slightly and formed a strong mechanical lock with the sand ring. The sand ring thus acted as a stiff template, forcing the foam pattern to remain round during drying and subsequent handling.
The process steps that I finally adopted are described below.
- Produce the EPS pattern by steaming and cooling. Verify the critical dimensions of the socket and spigot sections using a special gauging fixture. Keep the pattern dimensions within a tolerance of ±1.0 mm.
- Apply the refractory coating by dipping. Control the coating density and viscosity so that the deposited layer is uniform and has a thickness between 0.8 and 1.2 mm.
- Immediately after the last dip, before the coating surface changes from glossy to matte, place the pre-formed resin sand ring onto the required position. Apply slight pressure to make sure the ring is fully seated.
- Allow the coating to fill the annular gap between the sand ring and the pattern. If necessary, apply a small amount of extra slurry around the contact edge using a brush.
- Place the assembled pattern group in a vertical drying oven with the socket end facing upward. Maintain drying at 45–55°C for 4–6 hours, followed by cooling to ambient temperature.
- Inspect the dried assembly. The sand ring should be firmly bonded to the coating and should not be movable by hand.
- Assemble the pattern group and place it into the molding flask. Continue with the standard sand filling, vibration compaction, vacuum, pouring, and shakeout procedures.
This seemingly simple inversion of the assembly sequence produces a profound improvement. The wet coating fills every microscopic gap caused by surface irregularities of the EPS pattern and the inner surface of the sand ring. After drying, the coating becomes a composite layer with the sand ring embedded in it. The sand ring itself is made of resin-bonded sand with a much higher elastic modulus than the EPS foam. Therefore, the sandwich structure imparts a dramatically increased bending stiffness to the pipe end. I compare the stiffness of a bare pattern with that of a pattern equipped with a wet-installed sand ring by calculating an equivalent flexural rigidity.
Let \(D_{max}\) and \(D_{min}\) be the measured maximum and minimum inner diameters of the pipe socket end. The ovality ratio is defined as
\[
R_o = \frac{D_{max} – D_{min}}{D_{nom}} \times 100\%
\]
where \(D_{nom}\) is the nominal inner diameter. The corrected roundness improvement can be expressed by the relative reduction in ovality:
\[
\Delta R_o = \frac{R_{o, dry} – R_{o, wet}}{R_{o, dry}} \times 100\%
\]
In our production trials, \(R_{o, dry}\) was measured on parts made with the old dry sleeve method, while \(R_{o, wet}\) was measured on parts made with the wet sleeve method. The improvement was substantial, as I will present later.
Mechanics of the Wet-Installed Sand Ring
To understand why the wet method works mechanically, I consider the equilibrium of forces acting on a pipe end during vibration compaction. The sand flow exerts an external pressure \(p_s\) on the exposed surface of the pattern. If the pattern is unsupported, the resulting buckling force can be estimated as
\[
F_{b} = p_s \cdot A_{proj}
\]
where \(A_{proj}\) is the projected area of the pipe end in the direction of the sand flow. The buckling resistance of a circular-cylindrical shell is proportional to its flexural rigidity \(EI\). For a thin-walled cylinder with average radius \(r\), wall thickness \(t\), and effective length \(L\), the critical circumferential bending moment scales as
\[
M_{cr} \propto \frac{E I}{r} = \frac{E \pi r t^3}{6 r} = \frac{E \pi t^3}{6}
\]
This formula is a simplification, but it illustrates that the resistance is highly sensitive to the wall thickness \(t\). Since the EPS pattern wall cannot be thickened without increasing weight and cost, we add a stiffening layer. When the resin sand ring is firmly bonded to the pattern, the effective wall thickness becomes
\[
t_{eff} = t_{foam} + t_{coating} + t_{ring}
\]
The flexural rigidity increases by
\[
\frac{E_{eq} I_{eq}}{E_{foam} I_{foam}} \gg 1
\]
because the elastic modulus of the resin sand ring is typically two orders of magnitude larger than that of EPS foam. Therefore, the wet-installed sand ring converts a compliant foam end into a rigid composite structure that can resist the dynamic pressure imposed by vibration.
Another important factor is the adhesive strength at the coating-sand ring interface. Let \(\tau_{ad}\) be the shear strength of the dried coating layer. The total adhesive force over the contact area \(A_c\) between the sand ring and the coating is
\[
F_{ad} = \tau_{ad} \cdot A_c
\]
During vibration, the inertial force of the sand ring is
\[
F_{in} = m_{ring} \cdot a_v
\]
where \(a_v\) is the peak acceleration of the vibrating table. To prevent separation, we require
\[
F_{ad} > F_{in}
\]
In the dry method, there is no adhesive layer, so \(F_{ad} \approx 0\), and the sand ring can rotate or slide. In the wet method, the dried coating penetrates into the porous surface of the resin sand ring, creating a strong mechanical interlock. Thus \(F_{ad}\) is large enough to hold the ring firmly during the entire molding cycle.
Key Considerations for Industrial Implementation
While the wet sleeve sand ring technique is straightforward, I found several critical aspects that must be carefully controlled for reliable mass production. The first is timing. The sand ring must be fitted after the final dip but before the coating surface begins to dry. If the coating starts to form a skin, the adhesion is compromised and the sand ring may not be wetted to the pattern. In our workshop, the allowable open time is typically 20 to 30 minutes, depending on ambient temperature and humidity. To ensure consistency, I developed a time-based checklist for the operators.
Second, the sand ring must be accurately coaxial with the pattern axis. Any obliquity will create a tapered support that may itself induce ovality. I therefore designed a simple positioning jig that holds the sand ring in a horizontal plane while the vertical pattern is lowered into it. This jig guarantees a uniform annular gap around the circumference. The dimension of the sand ring inner diameter is critical. The clearance between the sand ring and the pattern after coating should be between 1.0 and 2.5 mm. If the gap is too small, the sand ring may scrape off the coating; if too large, the coating may sag and leave voids. The actual formula for the optimum gap is:
\[
g_{opt} = \frac{d_{ring} – d_{pattern}}{2} – t_{coating}
\]
where \(d_{ring}\) is the inner diameter of the sand ring, \(d_{pattern}\) is the outer diameter of the coated pattern, and \(t_{coating}\) is the wet coating thickness. We found that a value of \(g_{opt}\) in the range of 1.0–2.0 mm gives the best results. A smaller gap causes binding, and a larger gap weakens the support because the drying shrinkage of the thick coating may create shrinkage cavities.
Third, the dimensional stability of the EPS pattern is essential. If the pattern diameter varies by more than 2 mm from piece to piece, a single sand ring size cannot fit all patterns. In such cases, the slurry filling time and pressure become difficult to control. To overcome this, I established a procedure to inspect the socket-end dimension of every pattern before coating. Patterns that are out of tolerance are rejected or manually adjusted.
Fourth, the drying schedule must be matched to the new assembly. The presence of the sand ring increases the thermal mass of the pattern group, so the drying time must be extended slightly. I compared the moisture content of the coating before and after drying to confirm that the coating reaches a sufficiently low residual moisture level. The drying time was increased from 5 hours to 6 hours for DN800 and above. The temperature was kept below 60°C to avoid thermal distortion of the EPS foam.
Finally, I paid attention to the lifting and transport steps. Although the wet-installed sand ring is much more secure than the dry one, the pattern group still needs to be handled gently. In our automated line, I adjusted the transfer speed and reduced the drop height at the conveyor transfer point. This minimized the dynamic impact on the assembly.
Comparative Experiments and Production Results
I conducted a series of comparative runs to quantify the effect of the wet sleeve sand ring technique. The first comparison was a small-scale trial using DN800 fittings. One group of products was made with the dry sleeve method, and the other group with the wet sleeve method. I kept all other parameters identical, including the pattern material, coating recipe, sand size, vibration frequency, and pouring temperature. The ovality measurements were taken at three positions: the socket end, the middle of the pipe body, and the spigot end. The results are summarized in Table 3.
| Method | Number of Castings | Ovality at Socket (%) | Ovality at Body (%) | Ovality at Spigot (%) | Rejection Rate (%) |
|---|---|---|---|---|---|
| Dry sleeve sand ring | 150 | 4.82 | 1.95 | 3.76 | 12.0 |
| Wet sleeve sand ring | 150 | 1.71 | 0.82 | 1.35 | 2.0 |
Table 3 clearly demonstrates a dramatic reduction in ovality at the socket and spigot ends. The rejection rate dropped from 12% to 2%, which means that the wet sleeve method solved the majority of the deformation problems. Encouraged by this result, I expanded the trial to all specifications from DN400 to DN1200. The results after one month of continuous production are shown in Table 4.
| Nominal Diameter (DN) | Production Quantity | Oval Parts Count | Ovality Percentage (%) |
|---|---|---|---|
| 400 | 1580 | 12 | 0.76 |
| 600 | 1420 | 18 | 1.27 |
| 800 | 1180 | 26 | 2.20 |
| 1000 | 960 | 27 | 2.81 |
| 1200 | 810 | 26 | 3.21 |
The overall ovality percentage remained below 3.3% for all specifications, and the total average was about 2.0%. This represents a substantial improvement compared with the initial data in Table 1, where DN1000 and DN1200 had ovality percentages of 6.4% and 8.6%, respectively. The wet sleeve method reduced the ovality by more than 60% across the board. The improvement ratio can be expressed as
\[
\eta = \frac{R_{o, initial} – R_{o, wet}}{R_{o, initial}}
\]
For DN1000, for instance, the initial value of 6.44% was reduced to 2.81%, so
\[
\eta = \frac{6.44 – 2.81}{6.44} \times 100\% = 56.4\%
\]
For DN1200, the initial value of 8.61% was reduced to 3.21%, giving an improvement of 62.7%. This is strong evidence that the wet sleeve technique is not merely a patch but a robust solution.
In addition to the ovality reduction, I also measured the coating integrity after drying. In the wet method, the coating around the sand ring showed no cracks or delamination, whereas in the dry method, cracks were observed in approximately 8% of the samples. The visual inspection of the interface after pouring also confirmed that the sand ring remained in its original position. The pour cavity showed no signs of sand ring displacement. This tells me that the bonding strength was sufficient to resist the buoyant force of the molten iron as well. The buoyant force on the sand ring during pouring can be calculated as
\[
F_b = \rho_{iron} g V_{ring}
\]
where \(\rho_{iron}\) is the density of liquid iron (about 6800 kg/m³), \(g\) is the gravitational acceleration, and \(V_{ring}\) is the volume of the sand ring. Since the sand ring is embedded in the sand mold and mechanically locked to the pattern, it cannot float or shift. The wet coating creates a continuous barrier that prevents any direct liquid metal penetration to the sand ring surface.
Discussion of Process Stability and Limitations
The success of the wet sleeve sand ring technique depends on maintaining a stable and repeatable manufacturing environment. I have identified several process windows that must be continuously monitored. The first is the coating viscosity. If the viscosity is too low, the wet coating will drain out of the gap before drying, leaving voids. If it is too high, the coating will not penetrate the pores of the sand ring completely. I used a flow cup viscometer to measure the coating viscosity daily. The specification window is 25 to 35 seconds at 20°C. The relation between viscosity and the required filling time \(t_{fill}\) for a given gap \(g\) can be approximated by the Poiseuille flow:
\[
t_{fill} \propto \frac{L \mu}{\Delta P g^2}
\]
where \(L\) is the radial length of the gap, \(\mu\) is the dynamic viscosity, and \(\Delta P\) is the capillary or applied pressure differential. In practice, a filling time of about 10 seconds is sufficient when the gap is within the recommended range.
Another key parameter is the curing state of the resin sand ring. The sand rings used in our factory are made with phenolic resin and have a shelf life of about one week. If the ring has been stored for too long, its moisture content changes and its inner diameter may shrink or expand. I therefore measure the inner diameter of each sand ring batch before use. The allowable tolerance is ±0.5 mm. A larger tolerance would produce unacceptable variation in the support stiffness. The coefficient of thermal expansion of the resin sand ring is also slightly different from that of the EPS foam. During drying at 50°C, the dimensional change of the ring is minimal, but if the drying temperature exceeds 70°C, the ring may warp. Therefore, I set a strict upper limit of 60°C for the drying oven.
One limitation of the wet sleeve method is that it requires a dedicated positioning jig and careful manual operation. In an automated high-volume line, this extra step can lower productivity if not properly integrated. To overcome this, I worked with the line integrator to add a small rotary table next to the coating station. The operator clips the sand ring onto the pattern using a pneumatic pusher, and the assembly is lifted automatically to the drying carriage. This added about 15 seconds per piece, which was acceptable because the overall rejection rate dropped so dramatically. The economic benefit of reduced scrapping and reduced machining cost far outweighed the extra handling time.
I also investigated whether the wet sleeve sand ring has any effect on the filling of the molten metal. In lost foam casting, the molten iron vaporizes the EPS pattern and fills the resulting cavity. The sand ring, being made of resin-bonded sand, is permeable to gas and does not obstruct the vacuum extraction. In fact, because the sand ring is tightly bonded to the coating, it prevents the collapse of the pattern during the early stages of pouring. This is especially important for large thin-walled fittings where the pattern has a large surface area and the metallostatic pressure can cause inward collapse. The effective rigidity of the pipe end is increased, allowing the molten metal to fill the region completely without distorting the cavity. I observed no increase in cold shut or misrun defects; if anything, the surface quality at the socket end improved because the cavity was more stable.
To further illustrate the mechanism, I provide a model of the pattern assembly as a composite beam. The pipe end can be idealized as a short cylindrical shell of length \(L\), average radius \(r\), and total equivalent wall thickness \(t_{eff}\). The critical buckling pressure for an externally pressurized cylinder is given by
\[
p_{cr} = \frac{E (t_{eff})^3}{4(1-\nu^2) r^3}
\]
where \(E\) is the effective Young’s modulus and \(\nu\) is Poisson’s ratio. For a bare EPS pattern, \(E \approx 15\) MPa and \(t_{eff} \approx 1.5\) mm. With a wet-installed sand ring that is 10 mm thick and has an elastic modulus of about 1.5 GPa, the effective product \(E t^3\) increases by a factor of
\[
\frac{(1.5 \times 10^9)(1.5 \times 10^{-2})^3}{(15 \times 10^6)(1.5 \times 10^{-3})^3} = \frac{1.5 \times 10^9 \times 3.375 \times 10^{-6}}{15 \times 10^6 \times 3.375 \times 10^{-9}} = \frac{5062.5}{50.6} \approx 100
\]
Thus, the critical buckling pressure is increased by roughly a hundred times. This simple calculation explains why the wet sleeve sand ring is so effective at preventing the pattern-end deformation during sand compaction. The external pressure from the vibrating sand is simply far below the critical pressure of the stiffened structure.
In addition to the mechanical stiffness, I paid attention to the thermal behavior. During drying, the foam pattern undergoes thermal expansion followed by contraction. The resin sand ring, being much more rigid and having a lower coefficient of thermal expansion than EPS, constrains the foam. This constraint helps to iron out any non-uniformity in the foam dimension. The wet coating acts as a compliant intermediary that accommodates the differential thermal strains up to a certain level. The maximum shear strain \(\gamma_{max}\) experienced by the coating during drying can be estimated as
\[
\gamma_{max} = \frac{(\alpha_{foam} – \alpha_{ring}) \Delta T}{t_{coating}/r}
\]
For \(\alpha_{foam} = 60 \times 10^{-6}\) K⁻¹, \(\alpha_{ring} = 15 \times 10^{-6}\) K⁻¹, \(\Delta T = 30\) K, \(t_{coating} = 1\) mm and \(r = 100\) mm, the shear strain is
\[
\gamma_{max} = \frac{(60-15) \times 10^{-6} \times 30}{1/100} = 0.135
\]
This is within the strain capacity of the ceramic coating, which usually can tolerate strains above 0.2. Therefore, the coating does not crack during drying. This calculation confirms why we did not see the coating cracking that was so common in the dry sleeve method.
Quality Control and Continuous Monitoring
To ensure long-term stability, I introduced a process control plan with clear checkpoints. The first checkpoint is the pattern dimension. I use a go/no-go gauge for the socket diameter. The second is the sand ring geometry. A dedicated gauge is used to verify the inner diameter, wall thickness, and flatness. The third is the wet assembly quality. Immediately after placing the sand ring, the operator visually checks whether the gap is uniformly filled with coating. If not, additional slurry is applied with a brush. The fourth is the drying curve. I installed thermocouples inside the drying oven and logged the temperature profile. The fifth is the final inspection of the dried assembly. A handheld hardness tester confirms that the coating has hardened sufficiently. Finally, after shakeout, each pipe end is measured with a caliper and a CMM for a selected set of samples. The process capability index \(C_{pk}\) is calculated as
\[
C_{pk} = \min \left( \frac{USL – \bar{x}}{3\sigma}, \frac{\bar{x} – LSL}{3\sigma} \right)
\]
where \(USL\) and \(LSL\) are the upper and lower specification limits for the inner diameter. After implementation, \(C_{pk}\) improved from below 0.8 to above 1.33 for the critical socket inner diameters. This is a strong indication that the process is now both accurate and stable.
The data from the first six months is presented in Table 5. I segmented the results by quarter to show the steady-state performance. Each quarter represents approximately two months of continuous operation on the automated line.
| Segment | Total Parts Produced | Oval Parts | Ovality Percentage (%) | Average Scrap Rate due to Ovality (%) |
|---|---|---|---|---|
| Q1 | 4260 | 96 | 2.25 | 1.8 |
| Q2 | 4530 | 101 | 2.23 | 1.6 |
| Q3 | 4900 | 108 | 2.20 | 1.5 |
| Q4 | 5120 | 109 | 2.13 | 1.4 |
The ovality percentage remained within a very narrow band from 2.13% to 2.25%, indicating excellent repeatability. The small variations can be attributed to seasonal temperature changes in the workshop. I also compared the ovality of the socket end with the spigot end. The socket end usually has a larger diameter and a more complex geometry, so it is more susceptible to deformation. With the wet sleeve sand ring, the difference between the socket and spigot ovality was reduced to less than 0.5 percentage points. This suggests that the sand ring provides support to the entire end zone, not just to the flange face. I believe this is because the wet coating bridges the sand ring and the pattern over a significant axial length, distributing the external load evenly.
One of the subtle benefits of the wet sleeve method is that it also reduces other pattern-related defects. For example, the dimensional consistency of the socket end improved, which simplified the subsequent machining operation. The positioning of the part on the machining lathe became more reliable, and the amount of machining allowance required for the spigot ends could be reduced from 3 mm to 2 mm. This resulted in a reduction in machining time and cutting tool wear. The overall cost saving was substantial. I estimated that the total manufacturing cost per fitting decreased by about 4% despite the additional material and handling cost of the wet sleeve technique.
Conclusion
Through this extensive study, I have established that the wet sleeve sand ring technique is a highly effective method to prevent the ovality of small and medium ductile iron pipe fittings produced by vacuum lost foam casting. The core idea is to install the resin sand ring while the refractory coating is still wet, allowing the coating to fill all gaps and bond the sand ring to the pattern. After drying, the sand ring and the pattern become a monolithic composite structure with greatly enhanced rigidity. This structure is able to withstand the mechanical actions of transport, sand filling, and vibration compaction, maintaining the original circularity of the pattern and ultimately of the casting.
The quantitative results from my production line confirm that the ovality ratio is stably below 3% for all specifications from DN400 to DN1200, and in many cases below 2%. This is more than a 60% improvement over the previous dry sleeve method. The process has been running for six months without any sign of degradation, and the statistical process capability has reached a satisfactory level. The technique is simple, robust, and does not require any changes to the molding equipment or the vacuum system. It can be easily adopted by other lost foam casting foundries that produce thin-walled circular components such as pipe fittings, flanges, and similar parts.
For foundry engineers facing a similar ovality problem, I recommend that they first measure the ovality at each process stage to locate the dominant source. If the deformation originates at the pattern stage, rather than attempting to increase the foam density, they should consider using an external rigid ring applied during coating. The wet application method is the critical step that turns the ring into a load-bearing element. Without the wet contact, the ring is merely a loose collar that can rotate and provide little support. The combination of a pre-formed resin sand ring and a wet refractory coating creates a synergistic effect that is far more powerful than either element alone.
I also want to emphasize the importance of process discipline. The timing of the ring placement, the gap control, the drying temperature, and the handling procedure all interact to determine the final quality. A well-documented standard operating procedure and a simple positioning jig are sufficient to guarantee reproducibility. With these controls in place, the wet sleeve sand ring technique becomes a reliable, low-cost solution for the persistent problem of elliptical deformation in lost foam castings. I hope that sharing my experience will help other practitioners to solve similar problems in their own foundries.
In concluding, I would like to summarize the key formulas that govern the success of this technique. The ovality ratio is the primary quality index:
\[
R_o = \frac{D_{max} – D_{min}}{D_{nom}} \times 100\%
\]
The stiffness improvement ratio is:
\[
K = \frac{E_{composite} (t_{eff})^3}{E_{foam} t_{foam}^3}
\]
And the condition for sand ring retention during vibration is:
\[
\tau_{ad} A_c > m_{ring} a_{v,max}
\]
By satisfying the above conditions, I have achieved a stable and controllable lost foam casting process for ductile iron pipe fittings. The wet sleeve sand ring method is now a permanent part of our production standard, and it has significantly improved product quality, reduced scrap, and increased customer satisfaction.