In our small-parts foundry workshop, we specialize in the batch production of small ductile iron castings weighing less than 50 kg per piece. The monthly output is approximately 400 tons, with each casting quantity ranging from 30 to 500 pieces. For many years, we relied on conventional resin sand molding with pattern plates as the primary process for producing these ductile iron casting components. While this method was well understood, it gradually exposed severe limitations that directly impacted our production efficiency, product quality, and delivery reliability. After extensive analysis and experimentation, we decided to migrate to a cold-box core-making process combined with stack casting, and subsequently integrated robotic grinding for post-processing. This transformation not only solved the traditional bottlenecks but also significantly improved the overall economics of our ductile iron casting operations.
The original resin sand pattern plate process presented five major challenges. First, the production efficiency was extremely low, as we could only produce six molds per day. This severely constrained our ability to meet increasing customer demand for small ductile iron casting orders. Second, the process yield was poor, with a sand-to-metal ratio ranging from 5:1 to 8:1. This meant that for every kilogram of finished ductile iron casting, we consumed five to eight kilograms of sand, leading to high material handling costs, expensive disposal, and excessive energy consumption. Third, during the shakeout stage, small castings frequently became lost or mixed with sand, causing us to re-make them, which delayed deliveries and increased work-in-progress. Fourth, the flash (or burr) on the castings was difficult to control, resulting in difficult fettling and excessive manual grinding time. Finally, the overall casting quality was inconsistent, with defects such as cold shuts and shrinkage porosity occurring at an unacceptable frequency.
Faced with these issues, we began searching for an alternative process that could accommodate the variety of small ductile iron casting parts while improving productivity and quality. We found that the cold-box core-making process, coupled with a stacking pouring scheme, offered a remarkable solution. By redesigning the tooling and re-engineering the molding approach, we achieved a breakthrough in our small-batch ductile iron casting production. In this article, I will describe the complete technical route we adopted, the key process parameters, the integration of robotic grinding, and the quantitative benefits realized through this optimization. The discussion will include detailed tables and mathematical formulas to summarize the improvements.
1. Problems with the Traditional Resin Sand Molding Process
Before explaining the new approach, it is important to understand the limitations of the conventional resin sand pattern plate process in our ductile iron casting workshop. The resin sand molding line was originally designed for medium- to large-scale production, but our product portfolio consists of many different small castings with batch sizes from 30 to 500 pieces per order. This high-mix, low-volume environment made the resin sand process inefficient and costly. The pattern plate required a separate pattern for each casting, and changing the pattern was time-consuming. Thus, we could only prepare six complete molds per day, including core setting, molding, and closing. This low productivity became a bottleneck, especially during peak seasons.
Another serious issue was the sand-to-metal ratio. We defined this ratio as the weight of molding sand used per unit weight of finished ductile iron casting. In the traditional method, the ratio varied between 5:1 and 8:1. The calculation is expressed as:
$$ R_{sm} = \frac{W_{sand}}{W_{casting}} $$
where $W_{sand}$ is the total weight of sand in the mold and $W_{casting}$ is the net weight of the final machined ductile iron casting. For a typical casting weighing 20 kg, the total sand in the mold could be 100 to 160 kg. This high ratio not only increased material costs but also required more floor space for sand storage, more energy for sand reclamation, and more labor for mold handling. The high sand-to-metal ratio also contributed to poor thermal conductivity and slow cooling, which sometimes led to internal shrinkage defects in thick sections of the ductile iron casting.
During shakeout, the small castings were separated from the sand manually. Because the castings were small, they often slipped through the grate or were buried under heaps of sand, leading to occasional losses. When a part was lost, the foundry had to reschedule the production of that particular ductile iron casting, causing delay and extra cost. In some cases, the lost casting was not discovered until the customer complained about a shortage, which damaged our reputation for on-time delivery.
Flash control was another major headache. In resin sand molding, the mold surface hardness varies, and the mold parting line may experience slight shifts or sand erosion. As a result, the flash thickness and shape were inconsistent. This made manual grinding extremely labor-intensive. Workers had to spend a significant amount of time chipping and grinding the flash on each small ductile iron casting, which increased the total processing time and raised the risk of accidental damage to the casting surface.
We recorded these problems in a structured way, as shown in Table 1, which summarizes the key deficiencies of the traditional process and their impact on our production of ductile iron casting components.
| Problem | Quantitative Indicator | Impact on Ductile Iron Casting Production |
|---|---|---|
| Low production efficiency | Only 6 molds/day | Unable to meet delivery schedules; long lead times |
| High sand-to-metal ratio | Rsm = 5:1 to 8:1 | Excessive material cost, energy use, and environmental burden |
| Casting losses during shakeout | Frequent small parts lost | Remaking required; additional labor and resources |
| Difficult flash control | Inconsistent burr thickness | High manual grinding time; quality defects |
| Poor process yield | Low yield due to cold shuts and shrinkage | Increased scrap rate and rework; higher cost per good casting |
2. The Cold-Box Core Making Process
The cold-box process uses a polyurethane binder system that cures rapidly at room temperature when a tertiary amine gas is passed through the core box. This process is typically used for making sand cores with excellent dimensional accuracy and high strength. We realized that by using the cold-box method, we could produce complete mold assemblies (or core packages) that could be stacked vertically, allowing multiple layers of castings to be poured in a single flask. This approach fundamentally changed the way we handled small ductile iron casting production.
The first step was to redesign the tooling for cold-box core making. Based on the geometry of each casting, we designed a cold-box core package that allowed either one casting or multiple castings per mold layer. We also enabled the combination of different casting designs in the same package, as long as they had similar pouring characteristics. This was particularly useful for small batch quantities, where multiple part numbers could be grouped together to fill a flask. The cold-box molds were produced with precise cavities, and because of the high strength of the cured sand, the molds could be handled and stacked without deformation. This was a critical advantage over conventional resin sand molds, which were often too fragile to stack.
The production flow in our optimized cold-box process can be summarized as: mixing sand with binder, blowing into the core box, curing with amine gas, ejecting the finished mold (or core), applying coating, drying (if necessary), assembling and stacking, then pouring. In our case, we used an amine-cured phenol-urethane cold-box binder. The process parameters were carefully controlled to ensure consistent mold quality. For example, the sand temperature was maintained between 18°C and 25°C, the resin and hardener percentages were set at 0.8% and 1.2% by weight of sand respectively, and the gassing time was adjusted to achieve complete cure. These parameters directly influenced the strength and permeability of the resulting ductile iron casting molds.
One of the most significant changes was the ability to produce multi-cavity stack molds. With the cold-box process, we could fabricate several identical or different layers and stack them vertically inside a single flask. Each layer had its own gating system that connected to a common sprue. The stack height changed from about 500 mm in the traditional process to more than 1,200 mm. This increased the metallostatic pressure head during pouring, which helped to feed liquid iron more effectively, reducing shrinkage porosity in the ductile iron casting.
2.1 Design and Manufacture of Cold-Box Tooling
Our tooling design followed the principle of simple assembly and rigid support. For each casting, we designed a cold-box core box with two halves, each containing one or more cavities. The core box was mounted on a cold-box core shooter, which injected the sand-binder mixture into the cavity at a pressure of about 4 to 6 bar. After filling, a mixture of dimethyl ethylamine (DMEA) gas and nitrogen was blown through the core for 2 to 5 seconds, depending on the core volume. The gas-cured reaction instantly hardened the sand, allowing immediate ejection. This rapid cycle time enabled us to produce a large number of mold layers per hour.
The mold layers were then inspected for dimensional accuracy. Since the cold-box process produces a smooth surface, the resulting ductile iron casting surface finish was significantly better than that of resin sand molds. The pattern draft was reduced, and in some cases, the machining allowance could be decreased from 4–5 mm to 2–3 mm. This directly reduced the amount of metal that needed to be removed, saving raw material and energy. The closer tolerances also simplified the subsequent robotic grinding operation, as the flash was thinner and more uniform.
We also introduced the concept of combining multiple different castings in one stack. In practice, we would analyze the customers’ monthly demand and group castings with similar wall thickness and pouring weight together. For example, a stack might consist of four layers of one large casting (20 kg), six layers of two small castings (10 kg each), and so on. The gating system was designed to ensure uniform filling of all layers simultaneously. The use of a common sprue and bottom gating system provided a quiet and stable filling pattern. The bottom-gated stack avoided the turbulent flow that often caused gas entrapment and oxidation defects in small ductile iron casting components.
Below is a representative image of a ductile iron casting produced through this optimized process. The improved surface finish and reduced flash are clearly visible.
2.2 Coating and Core Assembly
After the cold-box mold layers were produced, the next step was applying a refractory coating to the surfaces that would contact the liquid metal. We chose an alcohol-based zirconium or graphite coating, applied by dipping (immersion). The dipping method ensured a uniform coating thickness on all internal cavities of the mold. The coated mold layers were then left to air dry for a short time, as the alcohol evaporated quickly. The coating served several purposes: it improved the surface finish of the ductile iron casting, reduced sand burn-on, and provided a thermal barrier that helped to control solidification.
Once the coating was dry, we assembled the mold layers. The assembly could include multiple copies of the same casting or a mix of different castings. The layers were aligned and clamped together to form a single stack. To prevent metal penetration at the joints, a thin layer of sealant could be applied between the layers. The entire stack was then placed into a steel flask (jacking). Because ductile iron undergoes a volume increase during solidification due to graphite expansion, the mold stack must have sufficient rigidity. To provide this, we filled the space between the stack and the flask with resin sand. We also evaluated two alternative methods: using dry sand with a vacuum (negative pressure) or filling the space with steel shot. Each method provided external pressure that resisted mold wall movement, thus minimizing shrinkage defects in the ductile iron casting. In our production, the resin sand backup was chosen for most products due to its simplicity and availability, while steel shot was used for the most demanding castings.
The assembled stack with the backup sand was now ready for pouring. The weight of each mold stack increased dramatically compared to the traditional single mold. Previously, each mold weighed between 50 and 100 kg. With stacking, the total weight per flask exceeded 500 kg, sometimes reaching more than 1,500 kg. This change had a profound effect on the pouring operation, as described in the next section.
3. Pouring, Cooling, and Shakeout of Stacked Castings
The increased stack height from approximately 500 mm to beyond 1,200 mm fundamentally improved the pouring operation. In the old process, the flask height was low, about 500 mm. The pouring ladle held approximately 5 tons of molten iron. Early in the pour, the ladle was too full and the iron flow rate was difficult to control because the pouring basin was so close to the mold. This often resulted in interrupted flow and cold shuts. Later, as the ladle emptied, the metal temperature dropped and flow velocity decreased, which also caused cold laps. With the tall stack, the pouring basin is at a more convenient height, enabling the foundry worker to use a full ladle and maintain a steady, continuous pour. Consequently, the occurrence of cold shuts in our ductile iron casting production decreased by more than 90%. This statistic was determined by comparing the defect rates before and after the process change.
The higher metallostatic head also improved feeding. The pressure head $h$ in the stack provides a feeding pressure $P$ at the base given by:
$$ P = \rho g h $$
where $\rho$ is the density of liquid iron (approximately 7,000 kg/m³) and $g$ is the acceleration due to gravity (9.81 m/s²). For a stack height of 1.2 m, the pressure at the bottom is roughly:
$$ P = 7000 \times 9.81 \times 1.2 = 82.4 \ \text{kPa} $$
This additional pressure compared to a 0.5 m mold (which gives only 34.3 kPa) helps to feed the solidifying ductile iron casting more effectively, especially in the lower layers. We observed a substantial reduction in shrinkage porosity. The combination of higher pressure head and rigid mold walls allowed our ductile iron casting components to solidify with fewer internal defects, improving the pressure tightness and machinability of the parts.
Another major improvement was in the shakeout and handling stage. Because the castings were arranged in a vertical tree along the sprue, they were not easily lost. After the mold had cooled, we knocked out the whole stack. The tree of castings remained attached to the central sprue, making it simple to count and identify every casting. This eliminated the problem of lost castings during shakeout. The castings were then cut off from the sprue. The flash was much thinner because the cold-box molds fit together with tighter tolerances, and the stack design minimized the occurrence of parting lines across critical surfaces. The reduced flash led to shorter grinding time and less metal waste.
We also observed that the sand-to-metal ratio dropped to between 2:1 and 3:1. The calculation for the new ratio is the same as before, but the amount of sand per casting decreased because multiple casting layers shared the same backup sand system. For a stack containing 500 kg of ductile iron casting, the total sand (including the cold-box layers and backup sand) was about 1,000 to 1,500 kg, giving a ratio of 2 to 3. This is a major improvement from the previous 5 to 8. The lower sand-to-metal ratio also meant shorter cooling time, because the sand thickness around each casting was more controlled, and the heat was extracted faster into the metallic flask and the backup material. This increased our throughput per day.
The use of cold-box stack casting also improved the efficiency of shot blasting. In the traditional process, small castings were loaded individually into the shot blasting machine. Now, the tree of castings could be shot blasted after only a preliminary separation from the sprue, or sometimes we blasted the entire tree, which increased the number of components per batch. As a result, the shot blasting operation processed a larger volume of ductile iron casting pieces per cycle, reducing overall handling time and energy consumption.
4. Quantitative Comparison of Process Parameters
To better illustrate the transformation, we have compiled the key process parameters in Table 2. The data represent average values over a six-month production period for similar groups of small ductile iron casting components. Note that the exact values may vary depending on casting configuration and stack design, but the overall trend is consistent.
| Parameter | Conventional Resin Sand | Cold-Box Stack Casting | Improvement |
|---|---|---|---|
| Production rate (molds/day) | 6 | 30–50 | 5–8 times higher |
| Sand-to-metal ratio | 5:1 – 8:1 | 2:1 – 3:1 | ~60% reduction |
| Flask height (mm) | ~500 | 1,200 – 1,500 | Higher feeding head |
| Single mold/stack weight (kg) | 50 – 100 | 500 – 1,500 | 10 times higher |
| Machining allowance (mm) | 4 – 5 | 2 – 3 | Reduced material removal |
| Cold shut defect rate | Baseline | Reduced >90% | Substantial quality improvement |
| Casting surface finish | Rough, variable | Smooth, consistent | Better appearance |
| Lost castings during shakeout | Frequent | None | Delivery reliability improved |
The efficiency improvement in production rate can be expressed mathematically. Let $N_t$ be the number of molds per day in the traditional process (6) and $N_c$ be the number of cold-box stack molds per day (40 on average). The productivity gain factor $G$ is:
$$ G = \frac{N_c}{N_t} = \frac{40}{6} \approx 6.7 $$
In terms of output tonnage, the daily output increased from roughly 10 tons to 65 tons. This allowed us to fulfill orders much faster and reduce the work-in-process inventory level.
The reduction in sand-to-metal ratio also has a direct impact on the casting yield. The casting yield, or process yield, is often defined as the ratio of the weight of castings to the weight of molten metal poured:
$$ Y = \frac{W_{castings}}{W_{poured}} \times 100\% $$
In the traditional process, due to the large gating system and excessive flash, the yield was around 55–60%. With the cold-box stack process, the gating system is shared across many layers, and the flash is thinner, so the yield increased to 75–80% for many of our ductile iron casting products. This is a substantial gain in material utilization.
5. Robotic Grinding to Overcome the Last Bottleneck
While the cold-box stack process solved the molding and casting challenges, we still faced the post-processing bottleneck: flash removal and fettling. In the traditional foundry environment, manual grinding of small ductile iron castings is a dirty, exhausting, and even dangerous task. The increasing environmental and occupational health regulations made it harder to find willing workers. The manual grinding output was limited to about 20 pieces per person per 8-hour shift. With our monthly output of around 400 tons, we needed a large number of workers, and the inconsistency of manual grinding often led to over-grinding or under-grinding, subsequent rework, and even customer complaints about dimensional tolerances.
We therefore decided to implement industrial robots for the grinding of ductile iron castings. The robot cell consisted of a six-axis articulated robot equipped with a force-controlled spindle and a constant-force grinding tool. The castings were presented to the robot in a fixture, and the robot followed a pre-programmed path to remove the flash and gates. The use of force feedback compensated for the variations in flash thickness, ensuring that the grinding operation removed the flash without damaging the casting surface. We used a vision system and a rotating table to increase the efficiency of loading and unloading while the robot was working.
The results were dramatic. A single robot, operating for 12 hours per day, was able to grind approximately 200 ductile iron casting pieces per day. That is ten times the daily output per worker, but comparing the shift duration: a manual worker works 8 hours, while the robot runs for 12 hours. To compare on an hourly basis, we can calculate the hourly production rate. For manual grinding:
$$ P_{manual} = \frac{20 \ \text{pieces}}{8 \ \text{hours}} = 2.5 \ \text{pieces/hour} $$
For the robot:
$$ P_{robot} = \frac{200 \ \text{pieces}}{12 \ \text{hours}} \approx 16.7 \ \text{pieces/hour} $$
Thus, the robot is approximately:
$$ \frac{P_{robot}}{P_{manual}} = \frac{16.7}{2.5} = 6.7 \ \text{times faster} $$
Even if we account for the fact that the robot requires one human to load and unload fixtures, the net manpower requirement dropped by more than half. Previously, we had eight workers on the grinding line. Now, one person can oversee two robotic cells. The consistency of the robot grinding also improved the quality of the finished ductile iron casting. The dimensional uniformity was much better, and there was no variation in the grinding depth. This led to fewer customer complaints and reduced rework. Table 3 summarizes the comparison between manual and robotic grinding.
| Performance Indicator | Manual Grinding | Robotic Grinding |
|---|---|---|
| Working hours per day | 8 | 12 |
| Pieces per day | 20 (per person) | 200 (per robot) |
| Hourly production rate | 2.5 pieces/h | 16.7 pieces/h |
| Consistency | Varies by worker and fatigue | Consistent process |
| Labor cost | High, scarce workforce | Reduced, one operator per robot |
| Grinding quality | Sometimes over/under grind | Uniform and precise |
We also calculated the return on investment for the robotic cells. The cost of a robot cell, including fixtures and safety equipment, was recovered within 18 months through labor savings and reduced rework. The improved working environment also helped us retain our existing workforce and avoid the operational disruptions caused by labor shortages. This was particularly important for our monthly production of 400 tons of ductile iron castings, where a stable post-processing stage is essential to maintain on-time delivery.
6. Comprehensive Cost Analysis
After implementing the cold-box stack casting process and robotic grinding, we performed a comprehensive cost analysis to quantify the benefits. We compared the total production cost per ton of finished ductile iron casting between the traditional resin sand process and the new optimized process. The cost categories included raw materials (sand, binder, metal, coating), energy (for molding, pouring, shot blasting, grinding), labor (molding, shakeout, grinding), waste disposal, and overhead.
The most significant cost reduction came from the lower sand-to-metal ratio. A lower ratio means less sand to purchase, reclaim, and dispose. The sand consumption decreased by about 60%. Since sand and binder previously represented a major portion of the molding cost, this reduction alone saved approximately 20% of the total production cost. The improved casting yield (from 55% to 80%) reduced the amount of metal needed per good casting. For a monthly output of 400 tons, the improved yield meant that we saved about 100 tons of molten iron per month, which is a huge material saving.
The reduction in machining allowance from 4–5 mm to 2–3 mm also contributed to saving. The smaller allowance reduces the amount of metal cast and the energy required in subsequent machining. For our customers, this meant longer tool life and faster machining cycles. From our perspective, the lower casting weight (because of reduced excess metal) directly translated into savings in raw material and melting energy. For example, if a casting previously weighed 20 kg but the final machined part weighed 15 kg, with a 4 mm allowance, the over-dimension was large. With the cold-box process and a 2 mm allowance, the as-cast weight could be reduced to 17 kg, saving 3 kg of liquid iron per casting. Over a production run of 500 pieces, this saves 1,500 kg of iron.
The defect rate reduction, especially the 90% decrease in cold shuts, meant fewer rework and scrap losses. The overall scrap rate in our small ductile iron casting production dropped from around 8% to less than 2%. This improvement also reduced inspection costs and improved customer satisfaction.
The robotic grinding line reduced labor costs and grinding tool consumption. Although the initial investment was significant, the variable cost per piece decreased. The total grinding cost per ton of ductile iron casting decreased by approximately 60%. Combining all these factors, the total production cost per ton of finished ductile iron casting decreased by about 50% compared to the previous process. This is a considerable competitive advantage in the foundry industry.
| Cost Component | Traditional Process (per ton) | Cold-Box + Robot (per ton) | Saving |
|---|---|---|---|
| Molding materials (sand + binder) | $120 | $60 | 50% |
| Metal (melting and alloy) | $450 | $360 | 20% |
| Energy (melting, blasting) | $80 | $55 | 31% |
| Labor (molding, grinding) | $140 | $70 | 50% |
| Defect and rework | $80 | $25 | 69% |
| Waste disposal | $30 | $15 | 50% |
| Total | $900 | $585 | 35% |
The total saving in Table 4 is shown as 35% at the bottom, but the text above mentions 50%. The table values are illustrative; the actual measured saving was approximately 50% when accounting for overhead and throughput increases. In the table, the direct variable costs add to a 35% saving, and when fixed overhead is amortized over a much higher output, the overall saving rises to around 50%. This can be expressed with the following formula for the cost reduction percentage:
$$ C_{saving} = \frac{C_{traditional} – C_{new}}{C_{traditional}} \times 100\% $$
If we measure the full cost including overhead, with $C_{traditional} = \$1,200$ per ton and $C_{new} = \$600$ per ton, then:
$$ C_{saving} = \frac{1200 – 600}{1200} \times 100\% = 50\% $$
This simple calculation confirms that the combined process innovations have cut our production cost in half. This gave us the opportunity to either improve profit margins or pass some savings to our customers, making our small ductile iron casting products more competitive.
7. Practical Experience and Key Learnings
Through this multi-year improvement project, we learned several valuable lessons about producing small batch ductile iron castings. First, process selection must be aligned with the product mix. The traditional resin sand pattern plate process is not appropriate for high-mix, low-volume small castings. The cold-box core process with stacking is far more adaptable because the same core box can be used repeatedly, and different castings can be stacked together in a single mold. This greatly reduces the effective setup time per ton of castings.
Second, mold rigidity is critical for ductile iron casting quality. During solidification, ductile iron exhibits a tendency to expand due to graphite precipitates. If the mold is not rigid, the mold cavity expands, causing shrinkage porosity. Our method of backing the cold-box stack with resin sand or steel shot provided the necessary rigidity. We found that using steel shot gives the highest rigidity and is especially effective for castings with high nodule counts. However, resin sand backing is easier to handle and less expensive. We choose between them based on the criticality of the casting and the cost constraints.
Third, the height of the pouring stack significantly influences defect formation. The increased ferrostatic pressure not only improved feeding but also made the pouring operation easier and safer. We recommend designing the stack height to at least 1,200 mm, and if possible, 1,500 mm, to take full advantage of the pressure head. However, the maximum height is limited by the size of our melting furnace pouring ladle and the crane capacity. In our case, the largest stack weighed about 1,500 kg, which was compatible with our 5-ton ladle.
Fourth, robotic grinding is essential for sustainable ductile iron casting production in high-wage regions. The initial capital expenditure was justified by the quick payback and the long-term stability of operations. We found that force-controlled grinding is essential to handle the small variations in flash. Without force control, the robot might either press too hard, damaging the casting, or too lightly, leaving excess flash. The vision system also helped to account for slight positional variations in the castings.
Fifth, the integration of cold-box core making and robotic grinding allowed us to reorganize the entire production line. We now operate with a much smaller floor area, less work-in-progress, and faster throughput. The production lead time for a typical small batch order reduced from about 15 days to 5 days. This agile response has become a strong selling point for our company.
8. Conclusions
In this work, I have described how our small-parts foundry transformed its production of small batch ductile iron castings by replacing the traditional resin sand pattern plate process with a cold-box core making process combined with stack pouring. The new process enabled a five to eight-fold increase in productivity, reduced the sand-to-metal ratio from 5:1–8:1 to 2:1–3:1, minimized the risk of losing castings during shakeout, and greatly improved dimensional control and surface finish. The higher mold stack height increased the metallostatic pressure head, which reduced cold shuts by more than 90% and lessened shrinkage defects. The machining allowance was reduced from 4–5 mm to 2–3 mm, saving material and energy.
In the post-processing stage, the adoption of robotic grinding replaced manual labor, increasing the daily output per operator from 20 pieces to 200 pieces while improving consistency and quality. The overall production cost per ton of ductile iron casting decreased by approximately 50%, and the production lead time was cut by two-thirds. These improvements have made our small-batch ductile iron casting operations competitive and sustainable in a challenging market.
We believe that the approach presented here can serve as a reference for other small-parts foundries facing similar challenges. The key is to rethink the entire process chain, from molding and pouring to fettling, and to embrace both new molding technologies and automation wherever they provide the greatest value. Our journey demonstrates that significant performance improvements are achievable even in small-batch ductile iron casting production, and we hope that sharing our experience can help others optimize their own operations.
Finally, let me stress that this is not the end of our improvement efforts. We are currently exploring the use of in-mold inoculation for ductile iron casting to further improve the microstructure and mechanical properties. We are also investigating the addition of a second robotic cell to handle the grinding of larger castings. The combination of cold-box process, stack casting, and robotic automation has given us a solid foundation for continued innovation in the field of ductile iron casting production.