Optimization of Sand Mould Casting for a Tractor Saddle

In this study, I examined the replacement of forged steel with ductile iron for a tractor saddle produced by sand mould casting. The tractor saddle is a critical connecting component between a tractor and a semitrailer. During service, it is subjected to impact loads from multiple directions, including longitudinal, vertical, lateral, and torsional loading. These loads are severe, and therefore the saddle must satisfy strict requirements for strength, toughness, dimensional accuracy, and internal soundness. The conventional saddle is made from forged steel, which offers good impact resistance but has a higher cost than cast iron. My objective was to achieve iron-for-steel substitution through sand mould casting while maintaining the required mechanical properties and reducing production losses caused by defects such as blow, cold shut, shrinkage porosity, shrinkage cavity, sand inclusion, and slag inclusion.

I used a numerical simulation approach based on AnyCasting together with orthogonal experimental optimization. The work included structural process analysis, material selection, gating system design, filtration design, venting design, filling and solidification simulation, defect probability prediction, orthogonal optimization, and final experimental verification. The results showed that a six-ingate closed gating system with a bottom-pouring and stratified-pouring configuration, combined with an optimized set of process parameters, greatly reduced the probability of shrinkage porosity and shrinkage cavity. The pass rate increased from approximately 60% in the original process to approximately 95% in the optimized sand mould casting process.

1. Background and Significance

Sand mould casting remains one of the most important manufacturing methods for mechanical components because it is flexible, cost-effective, and suitable for both small and large production batches. In sand mould casting, the mould cavity is formed by compacted sand, and the molten metal is poured into the cavity, where it cools and solidifies into the desired shape. The process can be divided into sand mould casting and special casting methods. Special casting includes centrifugal casting, differential pressure casting, squeeze casting, permanent mould casting, and lost foam casting. These special methods can provide better surface finish, dimensional accuracy, and mechanical properties than ordinary sand mould casting. However, sand mould casting is still widely used because of its low moulding material cost, simple operation, and excellent adaptability to large and complex parts.

The traditional sand mould casting route includes sand preparation, moulding, core making, melting, sand handling, cleaning, and inspection. Among these steps, core making and moulding are especially important. The quality of moulding sand and core sand directly affects the final casting quality. Poor sand quality can cause gas holes, sand burn-on, slag inclusion, misrun, and other defects. For the tractor saddle, the manufacturing challenge is increased by its large size, thin walls, complex geometry, and high service load. A small internal defect can become a fatigue crack initiation site under repeated impact, so internal soundness is essential.

The original forged steel saddle has good impact toughness, but its cost is high. Casting the saddle from ductile iron can reduce material and machining costs if the casting process is properly controlled. In actual production, however, iron-for-steel substitution often leads to blow, cold shut, shrinkage porosity, shrinkage cavity, sand inclusion, and slag inclusion, with the risk probability decreasing in that order. These defects cause significant economic loss. Therefore, I focused on the sand mould casting process itself, especially the gating system, exhaust system, and filtration system, to improve filling behavior and solidification behavior.

2. Research Status of Casting Numerical Simulation

Numerical simulation of casting has developed over several decades. Early work focused on heat conduction and solidification calculations. Later, researchers introduced finite element and finite difference methods to predict temperature fields. With the rapid development of computers, casting simulation expanded from simple heat transfer to coupled fluid flow, heat transfer, solidification, stress, and defect prediction. Many commercial software packages have been developed and applied in aerospace, automotive, shipbuilding, and general machinery industries.

In the sand mould casting of large thin-wall parts, simulation can reveal filling sequences, velocity fields, temperature fields, isolated liquid regions, and probability defect maps. It helps engineers determine whether the gating system is reasonable, whether the venting system is sufficient, and whether shrinkage defects will occur. The main numerical methods used in casting simulation include finite difference, finite volume, finite element, and cellular automaton methods. For mold filling, the solution of mass, momentum, and energy conservation equations is essential. For solidification, latent heat release and phase change must be considered. For defect prediction, criteria such as temperature gradient, cooling rate, and solid fraction are commonly used.

In my work, I used AnyCasting because it provides strong pre-processing, solving, and post-processing capabilities. The software allowed me to import three-dimensional models, generate meshes, set thermal physical properties, define initial and boundary conditions, and visualize filling, solidification, and defect probability. The simulation results were then used to compare different gating systems and to perform orthogonal optimization.

3. Structural Process Analysis of the Tractor Saddle

The tractor saddle is a large, thin-wall casting. Its overall dimensions are approximately 755 mm × 580 mm × 322 mm, and its blank mass is approximately 63 kg. The average wall thickness is about 10 mm, while the maximum local thickness is about 30 mm. Because the part is large and thin-walled, it is difficult to fill completely and easy to form cold shut, misrun, and shrinkage defects. The structure is highly symmetrical, which is favorable for gating design and solidification control. However, the saddle contains several thick sections and complex connecting regions, which can act as hot spots during solidification.

I analyzed the structural processability of the saddle with the following considerations:

  • The wall thickness should not be excessively large, and wall thickness variation should be minimized to reduce hot spots and cold shut risk.
  • Large flat surfaces should be avoided where possible to reduce sand inclusion and sand burn-on.
  • Holes that are required in the final part should be cast directly whenever possible to reduce machining.
  • Sharp corners should be replaced by fillets to reduce stress concentration and to improve filling.

3.1 Material Selection and Chemical Composition

I selected QT700-6 ductile iron for the saddle. This material is a pearlitic ductile iron with high strength and good wear resistance. It is suitable for parts such as crankshafts and heavy-duty machine spindles. The casting is used in the as-cast condition without additional heat treatment. The required mechanical properties are tensile strength not less than 700 MPa, elongation not less than 6%, and hardness between 240 and 290 HBW. According to DIN EN ISO 945, the graphite should be type VI, with a graphite size of 5 to 8, and flake graphite is not permitted.

Element C Si Mn S P Mg Ce Cr Cu Mo
Content, wt% 3.5–3.9 1.8–2.1 0.35–0.5 0.006–0.02 ≤0.05 0.035–0.055 0.006–0.03 ≤0.1 0.4–0.6 ≤0.15

The density of QT700-6 increases with temperature up to about 800°C and then tends to stabilize. Its thermal conductivity decreases as temperature increases. These thermophysical properties strongly affect the temperature field and solidification behavior during sand mould casting. Because the saddle is large and thin-walled, the cooling rate is relatively high in thin sections, while thick sections remain liquid longer and can form isolated liquid regions. The carbon equivalent is normally controlled between 4.3 and 4.4. If the carbon and silicon contents are too low, shrinkage porosity and shrinkage cavity become more likely. If they are too high, other defects such as graphite flotation may appear.

Property Requirement
Tensile strength ≥700 MPa
Elongation ≥6%
Hardness 240–290 HBW
Graphite type VI, no flake graphite
Graphite size 5–8
Matrix Pearlite-type ductile iron

3.2 Raw Material Composition and Melting Control

The raw materials for ductile iron usually include recycled iron, clean ductile iron chips, and scrap steel. The main auxiliary materials include carburant, ferrosilicon, and composite agents. The charge composition has a significant influence on shrinkage behavior. In my production design, the returned material accounted for about 50%–60% of the charge, and scrap steel accounted for about 40%–50%. The carbon and silicon contents were controlled near the middle of the specification range. Alloying elements were added after the first chemical analysis of the melted charge. The carburant and composite agent were added at a ratio of about 7:3. After the first sampling, if carbon or silicon was below the design target, the furnace was tapped partially, the corrective additions were made, and the remaining molten iron was returned after complete melting. After chemical composition was confirmed, spheroidization treatment was performed.

3.3 Moulding and Parting Method

I evaluated three-box moulding, whole-mould moulding, and two-box moulding. The saddle has a complex shape and large size, so two-box moulding is the most practical. In two-box moulding, the pattern is divided into upper and lower halves, and the upper and lower flasks are moulded separately. I also compared vertical and horizontal parting. Vertical parting would require a very complex core, and the vertical height difference of the casting would be too large, approximately 794.32 mm, which is unfavorable for production line operation. Horizontal parting gives a more uniform height distribution, simplifies the core, and facilitates moulding and pattern removal. Therefore, I selected horizontal parting and horizontal pouring.

The important assembly surface of the saddle, which connects to the vehicle body, must have high quality. During filling, sand and slag tend to float upward. Therefore, I placed this important assembly surface in the lower mould to improve its quality. This is a key decision in the sand mould casting process design.

3.4 Process Parameters

The linear shrinkage of ductile iron depends on the matrix and restraint conditions. For pearlitic ductile iron, free shrinkage is about 0.9%–1.1%, while restrained shrinkage is about 0.6%–0.8%. I selected a shrinkage rate of 1.1% for the pattern design. The dimensional tolerance grade was selected as CT12, the machining allowance grade as RMAG F–H, and the weight tolerance grade as MT11–MT14. The draft angle was selected as 3° to ensure smooth pattern removal and to protect the mould and core.

Parameter Selected Value
Linear shrinkage 1.1%
Dimensional tolerance grade CT12
Machining allowance grade RMAG F–H
Weight tolerance grade MT11–MT14
Draft angle 3°
Moulding method Two-box, horizontal parting
Pouring position Horizontal

3.5 Sand Core Design

To reduce core making cost and improve core rigidity, I used a single sand core for the entire internal cavity. The core is a cold-box sand core. The working surface of the core must be free from flashing, deformation, looseness, cracks, severe lack of material, and excess material. Core print defects larger than 15 mm × 15 mm require rejection or repair. The core is fixed by both vertical and horizontal core prints. The vertical upper core prints have dimensions of 108 mm, 75 mm, and 42 mm; the lower core prints have dimensions of 108 mm, 98 mm, and 42 mm. Horizontal core prints are symmetric left and right. The rear large core print is 64 mm, and the front small core print is 35 mm. The core print draft is 3°. This design provides accurate location and prevents displacement or rotation during pouring.

4. Theoretical Basis of Liquid Metal Filling and Solidification

In sand mould casting, the ability of the liquid metal to fill the mould cavity is called filling capacity. It is influenced by metal properties, pouring conditions, and casting structure. Higher pouring temperature generally increases filling capacity, but excessively high temperature can cause sand burn-on, gas defects, and coarser microstructure. The pouring pressure head also affects filling. A higher sprue produces a larger static pressure head, but excessive pressure can cause turbulence and mould erosion. The gating system geometry controls flow resistance and filling sequence.

To predict filling and solidification, I used the fundamental conservation equations. The mass conservation equation is:

$$
\frac{\partial \rho}{\partial t}
+\frac{\partial(\rho u)}{\partial x}
+\frac{\partial(\rho v)}{\partial y}
+\frac{\partial(\rho w)}{\partial z}
=0
$$

For steady incompressible flow, this becomes:

$$
\frac{\partial u}{\partial x}
+\frac{\partial v}{\partial y}
+\frac{\partial w}{\partial z}
=0
$$

The momentum conservation equations in three directions are:

$$
\rho\left(\frac{\partial u}{\partial t}
+u\frac{\partial u}{\partial x}
+v\frac{\partial u}{\partial y}
+w\frac{\partial u}{\partial z}\right)
=
-\frac{\partial p}{\partial x}
+\mu\nabla^2u
+\rho g_x
$$

$$
\rho\left(\frac{\partial v}{\partial t}
+u\frac{\partial v}{\partial x}
+v\frac{\partial v}{\partial y}
+w\frac{\partial v}{\partial z}\right)
=
-\frac{\partial p}{\partial y}
+\mu\nabla^2v
+\rho g_y
$$

$$
\rho\left(\frac{\partial w}{\partial t}
+u\frac{\partial w}{\partial x}
+v\frac{\partial w}{\partial y}
+w\frac{\partial w}{\partial z}\right)
=
-\frac{\partial p}{\partial z}
+\mu\nabla^2w
+\rho g_z
$$

The energy conservation equation is:

$$
\rho c_p\left(\frac{\partial T}{\partial t}
+u\frac{\partial T}{\partial x}
+v\frac{\partial T}{\partial y}
+w\frac{\partial T}{\partial z}\right)
=
\frac{\partial}{\partial x}\left(\lambda\frac{\partial T}{\partial x}\right)
+\frac{\partial}{\partial y}\left(\lambda\frac{\partial T}{\partial y}\right)
+\frac{\partial}{\partial z}\left(\lambda\frac{\partial T}{\partial z}\right)
+Q
$$

During solidification, latent heat is released. The source term can be written as:

$$
Q=\rho L\frac{\partial f_s}{\partial t}
$$

The relationship between temperature and solid fraction can be approximated by linear or quadratic functions. For a linear relationship:

$$
f_s=\frac{T_L-T}{T_L-T_S}
$$

For a quadratic relationship:

$$
f_s=\left(\frac{T_L-T}{T_L-T_S}\right)^2
$$

Shrinkage porosity and shrinkage cavity prediction can be based on temperature gradient and cooling rate. The temperature gradient is:

$$
G=
\sqrt{
\left(\frac{\partial T}{\partial x}\right)^2
+\left(\frac{\partial T}{\partial y}\right)^2
+\left(\frac{\partial T}{\partial z}\right)^2
}
$$

The cooling rate is often expressed as:

$$
R=\frac{T_{\text{upper}}-T_{\text{lower}}}{T_{\text{upper}}-T_{\text{lower}}}
$$

In the G/R criterion, a value below approximately 0.8 indicates a high probability of shrinkage porosity. I used this criterion together with the probability defect parameter in AnyCasting to evaluate the three gating designs.

5. Design of Gating, Filtration, and Venting Systems

The gating system is the channel through which molten metal enters the mould cavity. It usually consists of a pouring cup, sprue, runner, and ingates. A well-designed gating system should fill the mould smoothly, avoid turbulence and gas entrapment, trap slag and sand, provide proper temperature distribution, and facilitate cleaning. For the tractor saddle, the gating system must also satisfy the requirements of large thin-wall sand mould casting.

I compared top gating, bottom gating, middle gating, and stratified gating. Top gating is easy to fill but causes severe impact and turbulence. Bottom gating provides smooth filling and good venting but is less favorable for directional solidification and riser feeding. Stratified gating combines smooth filling with better temperature distribution and feeding capacity. Because the saddle is large and thin-walled, I selected a bottom-pouring system as the main configuration, supplemented by stratified pouring. I also selected a closed gating system. A closed gating system is beneficial for slag entrapment and is suitable for production lines without skimming. It helps reduce sand inclusion and slag inclusion.

5.1 Average Pressure Head and Minimum Remaining Pressure Head

The average pressure head for bottom gating is calculated as:

$$
H_P=H_0-\frac{h_0}{2}
$$

For the saddle, the upper flask height was 30 cm, and the casting height was 32 cm. Using the selected pouring position, I obtained an average pressure head of approximately 14 cm.

The minimum remaining pressure head at the farthest and highest point of the casting is:

$$
H_M \ge L\tan\alpha
$$

where L is the horizontal distance from the farthest highest point to the sprue center, and α is the pressure angle. I selected L = 195 mm and α = 8°, giving H_M ≥ 27 mm. Therefore, the sprue length was required to be at least 347 mm.

5.2 Pouring Time

For iron castings weighing less than 450 kg, the pouring time can be estimated by:

$$
t=S\sqrt{G}
$$

where t is the pouring time in seconds, G is the total casting mass in kg, and S is a coefficient related to wall thickness. For a wall thickness between 8.0 and 15 mm, S is about 2.2; for 3.5–8.0 mm, S is about 1.85. I selected S = 1.85 and G = 63 kg, giving t ≈ 14 s. Therefore, the pouring time was set to 14 s.

5.3 Choke Area Calculation

The total ingate area was calculated by the hydraulic method:

$$
\sum A_{\text{in}}
=
\frac{G}{0.3\mu t\sqrt{H_P}}
$$

Using G = 157 kg, t = 14 s, μ = 0.64, and H_P = 14 cm, I obtained:

$$
\sum A_{\text{in}} \approx 15.6\ \text{cm}^2
$$

For the closed gating system, the area ratio was selected as:

$$
\sum A_{\text{in}}:\sum A_{\text{run}}:\sum A_{\text{sprue}}
=
1:1.65:1.9
$$

Based on this ratio, I designed three different gating systems with four, six, and eight ingates. The sprue and runner areas were kept the same in all three designs so that the effect of ingate number could be compared directly.

Design Number of Ingates Single Ingate Area, cm² Total Ingate Area, cm² Runner Area, cm² Sprue Area, cm²
Scheme 1 4 3.9 15.6 25.74 29.64
Scheme 2 6 2.6 15.6 25.74 29.64
Scheme 3 8 1.95–2.0 15.6 25.74 29.64

For Scheme 1, the single ingate cross-section was square with a side length of about 2.0 cm. For Scheme 2, the side length was about 1.6 cm. For Scheme 3, the side length was about 1.4 cm. The sprue was cylindrical with a diameter of about 6.0 cm. The runner was square with a side length of about 5.0 cm. All three schemes used one sprue and one runner. Only the number of ingates changed.

5.4 Filtration System

The filtration system removes suspended impurities from the molten metal. Common filtration methods include flow stabilization, choking, slag trapping, and filter screens. I used a 90 mm × 90 mm × 22 mm ceramic foam filter with 10 ppi. In addition, I installed a choking plate at the sprue outlet, a filter screen at the front of the runner, and a slag trap at the end of the runner. This combined filtration system improved slag removal and reduced sand inclusion and slag inclusion in the sand mould casting process.

5.5 Venting System

A proper venting system is essential for large thin-wall sand mould casting. Gas can come from the moulding sand, core sand, binder decomposition, air in the cavity, and gas dissolved in the molten metal. If gas cannot escape, blowholes, gas holes, and reaction pores may form. I designed vents for both the mould cavity and the sand core.

The venting design principles I used were:

  • Vents should be located at high points, far from feeding paths, and away from thick sections that require shrinkage feeding.
  • Vents should be placed at the end of the gating system and at locations where gas is difficult to remove.
  • For complex thin-wall castings, the total vent area should be about 1.5–2.5 times the ingate or choke area.
  • The vent bottom thickness should be less than half of the local casting thickness.
  • The core should have low gas evolution and high permeability.

The vent area can be estimated by:

$$
S=(1.5\sim4)\frac{22.6G}{\rho t h_p \mu}
$$

Using G = 157 kg, ρ = 6.984 kg/cm³, t = 14 s, h_p = 140 mm, and μ = 0.35, I obtained a total vent area of approximately 11 cm². I placed one large vent with an area of 8 cm² and two small vents with an area of 1.5 cm² each. The large vent was placed at the end of the gating system, and the small vents were placed at the highest pouring position. For the sand core, I used upward venting along the parting plane. This design improved gas evacuation and reduced blow and cold shut defects.

6. Numerical Simulation Setup

I built the three-dimensional model of the saddle and gating system and exported it in STL format. The model was imported into AnyCasting. I used different mesh densities for different regions. The runner region was meshed with approximately 160,069 cells. The mould cavity region was meshed with approximately 382,641 cells. The core region, which contains hot spots and stress concentration areas, was meshed with approximately 789,085 cells. The average element volumes were about 2.2 mm³ for the core, 4.5 mm³ for the runner, and 9.1 mm³ for the mould cavity. This mesh strategy provided a balance between accuracy and computational cost.

The initial and boundary conditions were set as follows:

Parameter Value
Pouring temperature 1395°C
Gravity 9.8 m/s², +x direction
Pouring time 14 s
Cooling condition Natural cooling in sand mould
Material QT700-6 ductile iron
Gating ratio 1 : 1.65 : 1.9

I used the temperature-dependent density, thermal expansion coefficient, thermal conductivity, and specific heat of FCD 700 in the simulation. These properties affect the filling and solidification results, especially the formation of isolated liquid regions.

7. Simulation Results for Different Ingate Numbers

7.1 Filling Behavior with Four Ingates

For the four-ingate design, the molten metal filled the runner and then entered the cavity through four ingates. At about 25% filling, the metal had just entered the cavity. At about 35% filling, the flow became unstable, with a maximum velocity of about 157 mm/s and a minimum velocity of about 0.699 mm/s. At about 55% filling, the maximum velocity reached about 179 mm/s. The large velocity difference caused turbulence and gas entrapment. The flow was not smooth, and the cavity was not filled in a progressive manner. This increased the probability of blowholes, gas holes, and sand defects. The four-ingate design was therefore not considered suitable for the saddle.

7.2 Filling Behavior with Six Ingates

For the six-ingate design, the molten metal entered the cavity more evenly. At about 24% filling, the metal was flowing into the ingates. At about 36% filling, the ingates were full and the metal had entered the cavity. At about 56% filling, the flow was stable with no large velocity difference. At about 94% filling, the cavity was almost completely filled, and the flow remained smooth. There was no significant turbulence, and the cavity was filled completely. The six-ingate design showed the best filling behavior among the three designs.

7.3 Filling Behavior with Eight Ingates

For the eight-ingate design, the filling behavior was also relatively smooth in the early and middle stages. At about 25% filling, part of the metal had entered the cavity. At about 35% filling, the flow velocity difference was small. At about 55% filling, the cavity was mostly filled with stable flow. At about 95% filling, the flow was slightly faster than in the six-ingate design, but still acceptable. The eight-ingate design was better than the four-ingate design but not as good as the six-ingate design in terms of overall stability.

7.4 Solidification Behavior

The solidification behavior was evaluated using the temperature field and the formation of isolated liquid regions. For the four-ingate design, at t = 65 s the solid fraction was 5.7%; at t = 97 s it was 15.4%; at t = 294 s it was 66.2%; and at t = 516 s it was 85.9%. The isolated liquid regions were small and dispersed, which was favorable for soundness. However, the filling behavior was poor, so this design was not optimal overall.

For the six-ingate design, at t = 50 s the solid fraction was 2.4%; at t = 117 s it was 21.2%; at t = 302 s it was 65%; and at t = 734 s it was 91.9%. The isolated liquid regions were very small. The solidification sequence was favorable, and the probability of shrinkage porosity and shrinkage cavity was low. This was the best solidification result.

For the eight-ingate design, at t = 67 s the solid fraction was 5.4%; at t = 127 s it was 25.3%; at t = 301 s it was 66.5%; and at t = 506 s it was 85.4%. Elongated isolated liquid regions appeared at the edge of the cavity. Several dispersed isolated liquid regions remained in the later stage of solidification. These regions can become internal voids because liquid metal cannot feed them. Therefore, the eight-ingate design had a higher risk of shrinkage porosity and shrinkage cavity than the six-ingate design.

Design Filling Behavior Isolated Liquid Region Defect Probability Overall Assessment
4 ingates Unstable, high velocity difference, turbulence Small and dispersed High, with line defects on top and bottom Not suitable
6 ingates Smooth, complete filling Very small Low, only minor defect near lock hole Best
8 ingates Relatively smooth, faster at late stage Elongated and dispersed Moderate to high, edge and cavity defects Acceptable but not best

Based on the filling behavior, solidification behavior, and probability defect results, I selected the six-ingate gating system as the optimal initial design. The six-ingate scheme had one sprue, one runner, and six ingates. The single ingate area was 2.6 cm², the total ingate area was 15.6 cm², the runner area was 25.74 cm², and the sprue area was 29.64 cm². The pouring temperature was 1385°C, and the pouring time was 14 s. This was a no-riser and no-chill sand mould casting process.

8. Orthogonal Experimental Optimization

After selecting the six-ingate gating system, I further optimized the process parameters using an orthogonal experiment. The objective was to minimize the total pore volume and the solidification time. A smaller pore volume means fewer shrinkage porosity and shrinkage cavity defects. A shorter solidification time means higher production efficiency, provided that the casting remains sound.

I selected four factors and three levels. The factors were pouring temperature, filling pressure, pouring speed, and molten metal mass. The levels are shown in the following table.

Factor Level 1 Level 2 Level 3
A: Pouring temperature, °C 1395 1385 1405
B: Filling pressure, Pa 1.0 0.9 1.1
C: Pouring speed, m/s 0.34 0.30 0.40
D: Molten metal mass, kg 157 155 159

I used an L9(3⁴) orthogonal array. The simulation results for the nine trials are given below.

Trial A B C D Pore Volume, cm³ Solidification Time, h
1 1395 1.0 0.34 157 0.986 0.5575
2 1395 0.9 0.30 155 0.654 0.5581
3 1395 1.1 0.40 159 0.473 0.5542
4 1385 1.0 0.30 159 0.813 0.5458
5 1385 0.9 0.40 157 0.634 0.5411
6 1385 1.1 0.34 155 0.746 0.5417
7 1405 1.0 0.40 155 0.682 0.5708
8 1405 0.9 0.34 159 0.914 0.5728
9 1405 1.1 0.30 157 0.753 0.5739

8.1 Range Analysis

For the pore volume, the smaller the value, the better the result. The range analysis is summarized below.

Index A B C D
K1 0.704 0.827 0.882 0.791
K2 0.731 0.734 0.740 0.694
K3 0.783 0.657 0.596 0.733
Range R 0.079 0.170 0.286 0.097
Optimal level A1 B3 C3 D2

The influence order on pore volume was C > B > D > A. That is, pouring speed had the greatest effect, followed by filling pressure, molten metal mass, and pouring temperature. The optimal combination for minimum pore volume was A1B3C3D2: pouring temperature 1395°C, filling pressure 1.1 Pa, pouring speed 0.40 m/s, and molten metal mass 155 kg.

For solidification time, the range analysis is summarized below.

Index A B C D
K1 0.5566 0.5580 0.5573 0.5575
K2 0.5429 0.5573 0.5593 0.5569
K3 0.5725 0.5566 0.5554 0.5576
Range R 0.0296 0.0014 0.0039 0.0007
Optimal level A2 B3 C3 D2

The influence order on solidification time was A > C > B > D. Pouring temperature had the greatest effect, followed by pouring speed, filling pressure, and molten metal mass. The optimal combination for minimum solidification time was A2B3C3D2: pouring temperature 1385°C, filling pressure 1.1 Pa, pouring speed 0.40 m/s, and molten metal mass 155 kg.

The trend analysis showed that increasing pouring speed reduced the pore volume and shortened the solidification time. Increasing pouring temperature increased both pore volume and solidification time. Increasing filling pressure first increased and then decreased pore volume; the solidification time decreased when filling pressure exceeded 1.0 Pa. Increasing molten metal mass first increased and then slightly decreased pore volume, while solidification time increased and then leveled off above 157 kg.

8.2 Matrix Analysis

Because the orthogonal experiment involved two indicators, I also used matrix analysis to determine the optimal combination. The indicator layer matrix M is constructed as:

$$
M_{ij}=
\begin{cases}
K_{ij}, & \text{if larger is better}\\
\dfrac{1}{K_{ij}}, & \text{if smaller is better}
\end{cases}
$$

The factor layer matrix T is:

$$
T_i=\frac{1}{\sum_{j=1}^{n}K_{ij}}
$$

The level layer matrix E is:

$$
E_j=\frac{R_j}{\sum_{i=1}^{m}R_i}
$$

The weight matrix is:

$$
\omega_i=M_iT_iE_i
$$

For the pore volume indicator, the calculated weights for the factors were approximately A = 0.0437, B = 0.0794, C = 0.1232, and D = 0.0477. For the solidification time indicator, the calculated weights were approximately A = 0.2774, B = 0.0131, C = 0.0365, and D = 0.0066. After combining the two indicators, the final weights were approximately A = 0.1605, B = 0.0463, C = 0.0799, and D = 0.0271 for the three levels. The maximum values were obtained at A2, B3, C3, and D2. Therefore, the matrix analysis also gave the optimal combination A2B3C3D2: pouring temperature 1385°C, filling pressure 1.1 Pa, pouring speed 0.40 m/s, and molten metal mass 155 kg.

Factor Optimal Level Selected Value
A: Pouring temperature A2 1385°C
B: Filling pressure B3 1.1 Pa
C: Pouring speed C3 0.40 m/s
D: Molten metal mass D2 155 kg

9. Verification of the Optimized Scheme

I simulated the optimized scheme to verify its performance. In the filling stage, the molten metal flowed smoothly through the gating system and filled the cavity completely. At about 22% filling, the metal had entered the cavity; at about 38% filling, the flow was stable; at about 58% filling, the cavity was mostly filled; and at about 96% filling, the cavity was almost full. There was no significant turbulence or gas entrapment.

In the solidification stage, the thinner walls solidified first, and the main cavity solidified later. This sequence is favorable for feeding. At about 4.9% solidification, the temperature field began to develop. At about 24.7% solidification, the temperature distribution was uniform. At about 68.2% solidification, isolated liquid regions appeared but were small. At about 92.2% solidification, the isolated liquid regions were further reduced. The optimized scheme showed good solidification behavior.

The defect probability results after optimization showed that the line defects observed in the original six-ingate scheme were eliminated. The overall probability defect map showed no obvious shrinkage porosity or shrinkage cavity. The lock hole area, which previously showed a small line defect, was improved. The optimized parameter set reduced the pore volume to approximately 0.401 cm³.

Stage Observation in Optimized Scheme
Filling Smooth flow, complete filling, no obvious turbulence
Solidification Progressive solidification, small isolated liquid regions
Defect probability No obvious shrinkage porosity or shrinkage cavity
Pore volume Reduced to about 0.401 cm³

10. Experimental Validation

The original production process used five vents, a pouring temperature of 1420°C, and an ingate area of about 14 cm². This process produced shrinkage cavity, blow, and other defects. The defects reduced the service life of the saddle and increased the burden on machining and metallographic inspection. The pass rate was only about 60%. The original process problems were mainly caused by insufficient venting, excessive pouring temperature, and insufficient ingate area for feeding during solidification.

For the optimized scheme, I produced the saddle using the six-ingate closed gating system and the optimized parameters. The surface quality was good, with no visible shrinkage porosity or shrinkage cavity. I selected two batches of fourteen products for inspection. Ultrasonic testing was used, and no unacceptable defects were detected. I then performed metallographic examination after etching with nitric acid alcohol to reveal the matrix structure. The pearlite content was greater than or equal to 80%, and the ferrite content was less than or equal to 20%. The graphite size was grade 6, and the spheroidization level was VI. No flake graphite was observed. These results met the technical requirements.

Test Item Requirement Result
Elongation ≥6% 10.0–10.7%
Yield strength — 410–436 MPa
Tensile strength ≥700 MPa 725–749 MPa
Hardness 240–290 HBW 240–249 HBW
Graphite size 5–8 6
Pearlite content ≥80% 75–83%
Spheroidization VI, no flake graphite VI
Ultrasonic inspection No unacceptable defects Passed
Pass rate Improvement required About 95%

The optimized sand mould casting process significantly improved the internal quality of the tractor saddle. The combination of a six-ingate closed gating system, bottom and stratified pouring, combined filtration, effective venting, and orthogonal parameter optimization reduced shrinkage porosity, shrinkage cavity, sand inclusion, and slag inclusion. The pass rate increased from about 60% to about 95%. This confirmed that iron-for-steel substitution is feasible for the tractor saddle when the sand mould casting process is properly designed and optimized.

11. Conclusions

I completed a systematic study of the sand mould casting process for a large thin-wall tractor saddle. The main conclusions are as follows:

  1. The saddle is a large thin-wall ductile iron casting with overall dimensions of approximately 755 mm × 580 mm × 322 mm and a mass of about 63 kg. QT700-6 ductile iron was selected because it can satisfy the required tensile strength, elongation, hardness, and graphite morphology.
  2. Horizontal two-box moulding was selected because it simplifies core design, improves pattern removal, and provides uniform height distribution. The linear shrinkage was set to 1.1%, the draft angle to 3°, and the dimensional tolerance to CT12.
  3. A closed gating system with bottom pouring and stratified pouring was designed. The gating ratio was ∑A_in : ∑A_run : ∑A_sprue = 1 : 1.65 : 1.9. Three designs with four, six, and eight ingates were compared.
  4. Numerical simulation showed that the four-ingate design caused unstable filling and high defect probability. The eight-ingate design produced elongated and dispersed isolated liquid regions. The six-ingate design provided smooth filling, complete cavity filling, small isolated liquid regions, and low defect probability.
  5. Orthogonal optimization further improved the process. The optimal parameters were pouring temperature 1385°C, filling pressure 1.1 Pa, pouring speed 0.40 m/s, and molten metal mass 155 kg. The pore volume was reduced to about 0.401 cm³.
  6. Experimental verification showed that the optimized sand mould casting process produced sound saddles with no unacceptable ultrasonic defects. The microstructure and mechanical properties met the requirements. The pass rate increased from about 60% to about 95%.

12. Future Work

Although the optimized sand mould casting process achieved good results, further work can improve the study. First, different ductile iron grades and heat treatment conditions should be tested to determine whether other materials can provide better performance after iron-for-steel substitution. Second, more accurate thermophysical property data for the specific ductile iron should be measured and used in the simulation. Third, regression analysis can be applied to the orthogonal experiment results to further reduce experimental error and improve prediction accuracy. Fourth, the relationship between local cooling rate, graphite morphology, and mechanical properties should be studied in more detail. Finally, the optimized gating and venting design can be integrated with intelligent process monitoring to improve stability in large-scale sand mould casting production.

The results of this study provide a practical reference for the sand mould casting of large thin-wall tractor saddles and similar components. The combination of numerical simulation, gating system design, filtration, venting, and orthogonal optimization offers a reliable route for reducing defects, improving pass rate, and achieving cost-effective iron-for-steel substitution.

Scroll to Top