In my master’s research, I focused on the sustainability challenges of traditional sand casting, particularly the high energy consumption and carbon emissions inherent in manufacturing sand casting parts. The foundry industry is under increasing pressure to adopt cleaner production technologies. Patternless sand casting methods, such as 3D sand printing and CNC sand milling, offer disruptive advantages in flexibility and precision for producing complex sand casting parts. However, a comprehensive sustainability evaluation framework for these emerging processes was lacking. To address this gap, I developed a low-carbon modeling approach for the patternless sand casting process and proposed an optimization design method for composite molds. This article details my research methodology, the mathematical models I established, and the application to a real motor end cover casting. The results demonstrate that composite molds produced by combining traditional pattern molding, 3D sand printing, and CNC sand milling can significantly reduce resource consumption and carbon emissions while improving production efficiency for sand casting parts.
Keywords: patternless sand casting, sand casting parts, composite mold, low-carbon modeling, particle swarm optimization, process parameters
1 Introduction
The manufacturing sector is responsible for nearly one-third of global energy demand and CO2 emissions. Traditional manufacturing, especially the foundry industry, faces severe environmental challenges. In China, the energy consumption of the manufacturing sector accounts for over 83% of industrial energy use. Sand casting, being one of the most versatile metal forming processes, is widely used for producing sand casting parts ranging from automotive components to aerospace structures. However, the conventional sand casting process is characterized by low material utilization, high energy intensity, and significant pollutant generation. The average energy utilization rate in Chinese foundries is less than 30%, and the investment in pollution control is only about 5-8% of total equipment investment, far below developed countries.
To mitigate these issues, researchers have explored various sustainable manufacturing strategies. For sand casting, process parameter optimization and mold design improvements have been shown to reduce defects and energy consumption. Meanwhile, additive and subtractive manufacturing technologies, such as 3D sand printing and CNC sand milling, have been introduced to the foundry sector. These patternless techniques eliminate the need for physical patterns, enabling rapid production of complex sand casting parts with higher accuracy. Nevertheless, the sustainability performance of these novel methods was not thoroughly quantified. Previous studies focused on mold strength, surface finish, and material properties, but a holistic low-carbon evaluation model was missing. Therefore, my work aimed to develop a comprehensive low-carbon modeling framework for patternless sand casting and to use it for optimizing composite mold designs.
The specific objectives of my research were:
- To analyze the process characteristics of patternless sand casting and identify the key factors influencing resource demand.
- To establish mathematical models for resource consumption, energy consumption, carbon emissions, and carbon efficiency for sand casting parts production.
- To propose a composite mold design that combines multiple forming methods (traditional pattern, 3D printing, CNC milling) and to optimize its coupled process parameters using particle swarm optimization (PSO).
- To validate the proposed models and optimization method through a case study of a motor end cover, comparing the low-carbon indicators of single-technology and composite-mold approaches.
2 Process Characteristics and Resource Demand Analysis
2.1 Process Characteristics of Patternless Sand Casting
Patternless sand casting refers to mold-making processes that do not require a physical pattern. The main technologies include 3D sand printing (binder jetting) and CNC machining of sand blocks. These methods differ significantly from conventional pattern-based molding. In my analysis, I compared the process flow for three approaches: traditional pattern molding, 3D sand printing, and CNC sand milling. Figure 1 illustrates the manufacturing route and material flows for the composite molding approach I adopted in this research.

The traditional method involves pattern shop activities, sand filling, ramming, and pattern withdrawal, which are labor-intensive and time-consuming. In contrast, 3D sand printing builds molds layer-by-layer by selectively depositing binder into sand, offering high geometric freedom and eliminating pattern costs. CNC sand milling starts from a pre-solidified sand block and removes material to form the cavity. Both methods produce near-net-shape molds with excellent dimensional accuracy, which is beneficial for sand casting parts with thin walls and intricate features.
One of the key advantages of patternless technologies is the possibility to design complex runner systems and hollow or truss-like mold structures, which are impossible with conventional pattern-equipped molding. These designs can reduce sand consumption and improve casting quality. However, the metal casting process still requires molds with adequate strength and thermal properties. Therefore, a combination of methods—called composite molds—can be used to exploit the strengths of each technology. For example, 3D printing can handle intricate cores, CNC milling can efficiently machine large flat surfaces, and traditional pattern molding can economically produce simple runner parts.
2.2 Analysis of Mold Characteristics
In contrast to conventional solid molds, patternless sand casting allows for modular mold designs. The mold can be segmented into several parts, each manufactured by a different method. In my work, I followed the following rules for mold segmentation:
- The mold units should have regular parting surfaces, and the number of modules should be minimized to maintain assembly accuracy.
- High-precision regions of the casting should be assigned to higher-accuracy forming methods (e.g., 3D printing) and should not be cut further.
- Avoid thin or fragile modules that could break during handling.
This modular approach increases flexibility and allows each module to be optimized for its specific function, ultimately benefiting the overall sustainability of producing sand casting parts.
2.3 Resource Demand Influencing Factors
To develop a low-carbon model, I first identified the factors that affect resource demand in patternless sand casting. These factors were classified into four categories:
| Category | Description | Examples |
|---|---|---|
| Mold forming method | The chosen technology influences material consumption, equipment energy use, and labor requirements. | 3D printing vs. CNC milling vs. traditional pattern molding |
| Mold intrinsic properties | Properties such as sand type, binder composition, mold thermal conductivity, and compactness directly affect the process. | Sand particle size, resin content, mold strength |
| Mold-casting coupling | These are parameters that link the mold design to the casting design, such as machining allowance, casting shrinkage, and draft angle. | Mold wall thickness, core size, surface finish requirements |
| Batch production conditions | The scale of production influences the amortization of fixtures and equipment, as well as the overall efficiency. | Number of patterns, production line automation, batch size |
These factors were then used to construct the resource consumption models.
3 Low-Carbon Modeling of the Patternless Sand Casting Process
3.1 Resource Consumption Model
Resource consumption in sand casting parts production mainly includes material resources (MR), energy resources (ER), and labor resources (LR). The total resource consumption (RC) can be expressed as:
$$RC = MR + ER + LR$$
Material resources consist of metallic materials, molding materials, auxiliary materials, and energy materials. For iron castings, the metallic materials include pig iron, return scrap, and steel scrap. Molding materials are the sand systems (e.g., furan resin sand, green sand). Auxiliary materials include coatings, chillers, and cooling water. Energy materials include coke, natural gas, and alcohol. The material resource term is calculated as:
$$MR = \sum_{a=1}^{n} Em_{a}^{\Delta1} + \sum_{b=1}^{n} Em_{b}^{\Delta2} + \sum_{c=1}^{n} Em_{c}^{\Delta3} + \sum_{d=1}^{n} Em_{d}^{\Delta4}$$
where $Em^{\Delta1}$ denotes metallic materials, $Em^{\Delta2}$ molding materials, $Em^{\Delta3}$ auxiliary materials, and $Em^{\Delta4}$ energy materials.
Energy resources are quantified as the electricity or fuel consumed by equipment and processes. The details are provided in the energy consumption model below. Labor resources are evaluated by the number of workers in each process step:
$$LR = \sum_{f=1}^{n} N_{f} \quad (f \in \text{process steps})$$
where $N_f$ is the number of workers involved in the $f$-th process.
3.2 Energy Consumption and Energy Efficiency Ratio
Energy consumption in patternless sand casting includes melting energy, auxiliary equipment energy, and recycling energy. The total energy consumption $E_{Total}$ per casting can be written as:
$$E_{Total} = E_{N\alpha} + E_{N\beta} + E_{N\gamma}$$
where $E_{N\alpha}$ is the energy for melting the molten metal, $E_{N\beta}$ is the electrical energy for all auxiliary equipment, and $E_{N\gamma}$ is the energy for material recycling (e.g., sand reclamation, scrap handling).
The melting energy per casting depends on the furnace capacity and the fraction of melt allocated to the specific casting. If the furnace produces $G_1$ tons of iron in one heat, the number of castings poured from one heat is:
$$N_{i} = \frac{G_{1} \cdot \eta \cdot (1-\delta)}{G_{cast}}$$
where $\eta$ is the production plan ratio, $\delta$ is the loss coefficient during transfer and pouring, and $G_{cast}$ is the mass of the final casting (including runner system). Then, the melting energy per casting is:
$$E_{N\alpha} = \frac{P_{i} \cdot G_{1}}{N_{i}}$$
where $P_i$ is the specific energy consumption (kWh/ton).
The auxiliary and recycling energy are obtained by summing the product of the average power of each equipment type and its operating time per casting:
$$E_{N\beta} + E_{N\gamma} = \sum_{\beta} \overline{En}_{\beta} t_{\beta} + \sum_{\gamma} \overline{En}_{\gamma} t_{\gamma}$$
Thus, the total energy consumption per sand casting part is:
$$E_{Total} = \frac{P_{i} G_{1}}{N_{i}} + \sum_{\beta} \overline{En}_{\beta} t_{\beta} + \sum_{\gamma} \overline{En}_{\gamma} t_{\gamma}$$
To compare the efficiency of different molding technologies, I define the Energy Efficiency Ratio (EER) as the ratio of the standard production time per qualified casting to the actual labor-time adjusted energy consumption:
$$EER = \frac{t_{st}}{\sum_{f} (t_{w}^{f} – t_{los}^{f}) \cdot N_{f} \cdot E_{Total}} \times 100\%$$
Here, $t_{st}$ is the standard time, $t_{w}^{f}$ is the working time in process $f$, $t_{los}^{f}$ is the lost time due to delays, and $N_f$ is the number of workers in that process.
3.3 Carbon Emission and Carbon Efficiency
Carbon emissions from the sand casting process are divided into material-related emissions (CM), energy-related emissions (CE), and non-expected emissions (CU). The total carbon emission $C$ per casting is:
$$C = CM + CE + CU$$
Material-related emissions account for the carbon footprint of all input materials (including their upstream production) and are adjusted for recyclable materials:
$$CM = \left[ M_{1}, \ldots, M_{n-1}(1-\lambda_{i}), M_{n} \right] \cdot \left[ f_{1}, \ldots, f_{n} \right]^{T}$$
where $M_i$ are the consumed quantities, $\lambda_i$ is the recycling rate for recyclable materials, and $f_i$ are the carbon emission factors.
Energy-related emissions include emissions from fuel combustion and from electricity generation. The generic formulation is:
$$CE = \left[ E_{1}, \ldots, E_{n}, F_{1}, \ldots, F_{n} \right] \cdot \left[ g_{1}, \ldots, g_{n} \right]^{T}$$
where $E_i$ are electricity consumptions, $F_i$ are fossil fuel consumptions, and $g_i$ are the corresponding emission factors.
Non-expected emissions arise from chemical reactions (e.g., gas evolution from sand binders) and from the treatment of wastes. They are expressed as:
$$CU = \left[ \varepsilon_{1} U_{1}, \ldots, \varepsilon_{n} U_{n} \right] \cdot \left[ h_{1}, \ldots, h_{n} \right]^{T}$$
where $\varepsilon_i$ are generation functions for unexpected sources and $h_i$ are their emission factors.
To evaluate the environmental performance relative to productivity, I defined the Carbon Efficiency (CER) as:
$$CER = \frac{t_{st}}{\sum_{f} (t_{w}^{f} – t_{los}^{f}) \cdot N_{f} \cdot C} \times 100\%$$
Both EER and CER provide useful benchmarks when comparing alternative manufacturing processes for sand casting parts.
4 Low-Carbon Process Design and Optimization of Composite Molds
4.1 Selection of Forming Methods and Materials
Composite molds combine different forming techniques to leverage their advantages. In my study, I considered three primary methods: traditional pattern molding (suitable for simple runner structures), 3D sand printing (ideal for complex core and cavity geometries), and CNC sand milling (efficient for large, relatively simple surfaces). The material selection depends on the required strength, surface finish, and thermal properties. For instance, fine sand (70-140 mesh) is preferred for 3D printing to achieve higher resolution and strength. CNC milling can handle various sand systems, including furan resin sand and sodium silicate sand.
The choice of forming methods for each mold module was based on the following criteria:
- High-precision zones of the sand casting parts (e.g., critical seal faces) → 3D printing with fine sand.
- Large flat areas and simple contours → CNC milling for speed.
- Standard runner and riser systems → traditional pattern molding to lower cost.
4.2 Classification of Process Parameters
To optimize the mold design, I categorized the process parameters of composite molds into three types:
| Parameter type | Definition | Examples |
|---|---|---|
| Independent process parameters | Parameters inherent to the mold material and mold making process. | Sand thermal conductivity, binder content, compaction density |
| Compensation process parameters | Parameters used to correct mold or core dimensions due to deformation or shrinkage. | Parting allowance, core allowance |
| Coupled process parameters | Parameters that link the mold geometry to the casting geometry. | Mold dimensions, mold wall thickness, machining allowance, number of mold modules |
In the following optimization, I focused on the coupled parameters because they directly affect the amount of sand and metal used, and consequently the overall sustainability of producing sand casting parts.
4.3 Particle Swarm Optimization Model
I employed Particle Swarm Optimization (PSO) to obtain the optimal coupled process parameters. The objective was to minimize a weighted sum of resource consumption, energy consumption, carbon emissions, and production time. The multi-objective optimization problem was converted into a single-objective problem using the weighted sum method and normalization based on the traditional pattern molding baseline.
The decision variables included the number of mold modules ($MN$), mold size ($MS$), minimum mold wall thickness ($MT$), and machining allowance ($MR$). The constraints were:
- The number of modules must not exceed a maximum allowed value: $g_1(X) = MN(X) – mn_{max} \le 0$.
- The mold wall thickness must be greater than the minimum allowable thickness: $g_2(X) = MT(X) – f_{min} \ge 0$.
- The total mold forming time must be less than a maximum allowed time: $g_3(X) = ST(X) – st_{max} \le 0$.
The PSO algorithm updates velocity and position for each particle as follows:
$$V_{id}^{t+1} = \omega V_{id}^{t} + C_{1} \cdot rand(0,1) \cdot (P_{id} – X_{id}) + C_{2} \cdot rand(0,1) \cdot (P_{gd} – X_{id})$$
$$X_{id}^{t+1} = X_{id}^{t} + V_{id}^{t+1}$$
where $\omega$ is the inertia weight (0.8), $C_1 = C_2 = 1.49445$, and $rand(0,1)$ is a random number.
The overall optimization function is:
$$Optimal = \theta_1 a_{R}(X) + \theta_2 a_{E}(X) + \theta_3 a_{C}(X) + \theta_4 a_{T}(X)$$
with weights $\theta_4 = 0.4$ (time), $\theta_3 = 0.3$ (carbon), and $\theta_1 = \theta_2 = 0.25$ (resource and energy). The individual objective functions are normalized as:
$$a_R(X) = \frac{EM_{max} – EM(X)}{EM_{max}}, \quad a_E(X) = \frac{EN_{max} – EN(X)}{EN_{max}}$$
$$a_C(X) = \frac{C_{max} – C(X)}{C_{max}}, \quad a_T(X) = \frac{T_{max} – T(X)}{T_{max}}$$
where the subscript $max$ refers to the baseline traditional pattern molding results.
The optimization procedure is shown in the flow diagram below (not reproduced here due to formatting), and the algorithm was implemented in MATLAB.
5 Case Study: Motor End Cover
5.1 Casting Process Conditions
To validate my models and optimization method, I selected a motor end cover made of HT200 gray cast iron. The casting mass was 32.86 kg, with maximum dimensions of 420 mm × 420 mm × 111 mm. The wall thickness varied from 6 mm to 36 mm, classifying it as a thin-walled iron casting. The required surface roughness was Ra ≤ 50 μm, and no shrinkage defects were allowed.
I designed three casting process variants: (a) traditional pattern molding, (b) CNC sand milling, and (c) 3D sand printing. For the composite mold approach, I combined traditional pattern molding for the runner and riser modules, 3D printing for the upper mold (complex geometry), and CNC milling for the lower mold. Table 1 lists the key casting parameters for the three single technologies.
| Parameter | Traditional | CNC milling | 3D printing |
|---|---|---|---|
| Draft angle (°) | 3 | 0 | 0 |
| Machining allowance (mm) | 5 | 3.5 | 2 |
| Sand box size (mm) | 700×700×300 | 700×700×300 | – |
| Pouring time (s) | 12.0 | 11.12 | 9.99 |
| Pouring temperature (°C) | 1380 | 1380 | 1380 |
The gating system used was a side-pouring annular stepped design, with one open riser and three trapezoidal ingates. For the composite mold design, the mold was divided into several modules. The original design and the optimized design (after PSO) are compared in Table 2.
| Parameter | Original (before optimization) | Optimized |
|---|---|---|
| Sand box size (mm) | 650×650×300 | 650×650×300 |
| Machining allowance (mm) | 3 (approx.) | 1.8 |
| Number of mold modules | 4 | 6 |
| Mold wall thickness (mm) | ~70 (typical) | 65 |
| Casting yield (%) | 84.83 | 85.52 |
| Sand filling volume in box (m³) | 0.0219 | 0.1538 |
5.2 Simulation Verification
I used ProCAST simulation software to verify the casting quality of the optimized composite mold. The finite element model included all mold modules with their respective material properties and interface heat transfer coefficients. The pouring temperature was 1380°C, and the initial mold temperature was 25°C. The simulation results showed that the contraction defects appeared only in the runner/riser system, which would be trimmed later, and no harmful shrinkage porosity was detected in the sand casting part itself. Thus, the optimized composite mold design was validated for production.
5.3 Results of Single-Technology Approaches
For the single-technology approaches, I calculated the material consumption, energy consumption, and carbon emissions per unit casting using the models in Section 3. Table 3 summarizes the key energy and emission results for two production conditions: (a1) single-unit small batch, and (A1) mass production of 3000 pieces.
| Method | Condition | Total energy (J) | Carbon emission (kg CO2) | Production time (h) | EER (%) | CER (1/kg CO2) |
|---|---|---|---|---|---|---|
| Traditional | a1 | 2.590E+08 | 200.6 | 33.99 | 0.49 | 0.00177 |
| Traditional | A1 | 2.453E+08 | 197.0 | 4.28 | 0.56 | 0.00191 |
| CNC milling | a1 | 2.781E+08 | 204.5 | 22.68 | 1.15 | 0.00435 |
| CNC milling | A1 | 2.729E+08 | 203.2 | 11.09 | 1.18 | 0.00442 |
| 3D printing | a1 | 1.909E+08 | 163.3 | 18.27 | 1.70 | 0.00551 |
| 3D printing | A1 | 1.857E+08 | 162.0 | 9.72 | 1.80 | 0.00573 |
These results show that 3D sand printing performs best in terms of energy consumption and carbon emissions due to its simplified process and reduced material usage. CNC milling has higher emissions because of the additional machining process and lower sand recovery rate. Traditional pattern molding is less efficient in small batches but improves significantly in mass production due to amortization of pattern costs.
5.4 Results of Composite Mold Approach
For the composite mold approach, I evaluated both the original design and the optimized design. Table 4 presents the main energy and emission results.
| Parameter | Original composite | Optimized composite |
|---|---|---|
| Total energy (a1B1) (J) | 2.490E+08 | 2.359E+08 |
| Total energy (A1B1) (J) | 2.424E+08 | 2.281E+08 |
| Total carbon (a1B1) (kg CO2) | 188.6 | 177.2 |
| Total carbon (A1B1) (kg CO2) | 186.9 | 175.2 |
| Production time (a1B1) (h) | 13.55 | 10.12 |
| Production time (A1B1) (h) | 7.18 | 5.57 |
| Efficiency (a1B1) (%) | 90.1 | 93.2 |
| EER (a1B1) (%) | 1.30 | 1.42 |
| CER (a1B1) (1/kg CO2) | 0.00478 | 0.00526 |
Compared with the traditional pattern molding approach, the optimized composite mold reduced total carbon emissions by 11.63% in small-batch production and by 11.06% in mass production. The production time per sand casting part decreased from about 34 h to about 10 h in the small-batch scenario, a 70% reduction. This demonstrates the substantial sustainability benefits of combining patternless methods.
5.5 Comparative Analysis of Resource Consumption
Figure 2 (omitted here) shows the material consumption comparison. The optimized composite mold reduced metallic material consumption by about 5% compared with the original composite design, and by about 10% compared with traditional pattern molding. The total sand consumption (including new sand and recycled sand) was also significantly lower because the mold modules were designed with reduced wall thickness and because a large portion of the mold consisted of green sand with a higher recycling rate.
Labor resource analysis revealed that traditional pattern molding required at least 10 workers for the pre-pouring stages, whereas the 3D printing method required only 2 workers, and the composite method needed 6 workers but with less working hours per part. In mass production, the labor time per part was far lower than in small-batch production.
5.6 Energy and Carbon Efficiency Discussion
The EER and CER indicators show that 3D sand printing has the highest values among single methods, but the composite mold approach achieves a good balance between efficiency and environmental impact. The optimized composite mold reached an EER of 1.42% and a CER of 0.00526, which are close to 3D printing values while maintaining a lower production time. This indicates that the composite mold approach combines the strength of each technology to achieve an overall sustainable production route for sand casting parts.
6 Conclusion and Outlook
In this research, I developed a comprehensive low-carbon modeling framework for patternless sand casting processes and applied it to optimize composite molds for a motor end cover. The main conclusions are:
- Patternless technologies, especially 3D sand printing, can reduce resource consumption and carbon emissions significantly compared to traditional pattern molding for small-batch production of sand casting parts.
- The composite mold approach, which combines traditional pattern molding for runner systems, 3D printing for complex features, and CNC milling for simpler sections, provides a flexible and sustainable solution that shortens production time while maintaining low emissions.
- The PSO-based optimization successfully improved the original composite mold design, further reducing resource use and carbon emissions by optimizing mold module count, wall thickness, and machining allowance.
Future work could extend this model to consider post-casting operations such as heat treatment and machining, and to explore intelligent mold designs with embedded sensors for adaptive cooling. The application of topology optimization and lattice structures to sand casting parts is also a promising direction to increase material efficiency.
I believe that the proposed low-carbon modeling and optimization methodology provides a practical tool for foundries aiming to transition toward cleaner production of sand casting parts.
