The production of high-quality, heavy-section castings for machine tools, such as beds and columns, presents significant technical challenges. The integrity of these large machine tool castings is paramount, as any defect can lead to substantial economic loss. In foundries utilizing split-box molding techniques for such components, the bottom-gated shower gating system is widely adopted. Its primary advantages lie in promoting a tranquil mold filling process and ensuring a relatively uniform temperature distribution within the casting cavity. However, the design principles for this system have often been based on empirical knowledge rather than a thorough understanding of the underlying fluid dynamics. This work details our comprehensive experimental and theoretical investigation aimed at demystifying the flow characteristics of the bottom-gated shower system, establishing a scientific foundation for its design, and ultimately enhancing the quality and reliability of heavy-duty machine tool castings.
Our research methodology centered on hydraulic simulation experiments, a powerful tool for visualizing and quantifying flow phenomena that are otherwise impossible to observe during actual metal pouring. The foundational theory for this approach is the mechanical similarity of flows. For the filling process governed primarily by gravity and resistance forces, we ensured our model satisfied both the Froude and Reynolds similarity criteria. To most accurately replicate the conditions in a sand mold during iron pouring, we selected a model scale of 1:1 (Lr = 1). This allowed the physical parameters measured in the water model to be directly applied to the design of the actual gating system for machine tool castings.
The experimental setup was constructed from transparent acrylic to allow full visual access. The system consisted of a pouring basin, sprue, main runner, multiple branching sprues (one for each shower line), branching runners, and the final bottom-gated ingates (shower gates). A representative structure with two rows of shower gates was studied in depth. The key dimensions of the gating system model are summarized in the table below:
| Model Component | Dimensions (mm) | Cross-Sectional Area (mm²) |
|---|---|---|
| Model Cavity (L×W×H) | 1200 × 400 × 400 | – |
| Pouring Basin Orifice (Ød0) | Varied: 20, 25, 30, 35, 40 | – |
| Sprue (Ø) | 40 | 1256 |
| Main Runner | 50 × 60 | 3000 |
| Branching Sprue (Ø) | 30 | 706 |
| Branching Runner | 30 × 40 | 1200 |
| Each Ingate (Ø) | 20 | 314 |
To capture the dynamic behavior of the system, we developed and employed several measurement techniques. A multi-point flow rate measurement device was constructed. Electrical contacts were placed at different heights within each casting cavity module. As the water level rose and touched each contact, a signal was sent to a light-beam oscillograph, recording the height-time curve for each module. Since the cross-sectional area of each module was constant, this data directly yielded the flow rate versus time relationship for each individual ingate.
Pressure distribution at the base of each ingate was measured using a similar principle. Thin pressure tubes were connected to each ingate location, with contacts placed inside these tubes. The oscillograph recorded the rise of the water column in each tube, corresponding to the dynamic pressure head at each ingate during filling.
Perhaps the most insightful measurements came from Laser Doppler Velocimetry (LDV). The LDV system was used to measure the instantaneous flow velocity within the ingates. Based on the optical Doppler effect, it accurately provided data on mean velocity, turbulence intensity, and velocity distribution statistics, which were processed and printed by a connected microcomputer.
Analysis of Experimental Results on Flow Characteristics
The experimental campaign yielded rich data, allowing us to analyze the system’s behavior in detail.
Ingate Flow Rate Distribution: A critical finding was the characteristic flow distribution among the multiple shower gates. Contrary to top-gating systems which often show a “first-in, first-served” uneven distribution, our bottom-gated shower system demonstrated a remarkably uniform flow allocation under stable pouring conditions. The distribution became even more stable as filling progressed. This uniformity is a significant advantage for producing dimensionally consistent machine tool castings. The phenomenon is attributed to the “superimposed” flow within the branching runner. The kinetic energy of the flow traveling from the branching sprue towards the end of the branching runner gradually diminishes. When the pouring basin head is sufficiently high, this energy does not fully dissipate, leading to a stable, filled condition in the branching runner that feeds all ingates quasi-equally from a common pressurized manifold.
Ingate Pressure Distribution: The pressure at the base of each ingate was found to be directly correlated with its flow rate. During the initial instant when the branching runner fills and flow begins through the ingates, a pressure spike or “jetting” phenomenon was observed. Immediately after this transient, the pressure stabilized and remained constant throughout the filling process for each ingate, mirroring the stability of the flow rate. The pressure head at each ingate was consistently about 100 mm of water column higher than the metal level in its corresponding cavity module, confirming the pressurized nature of the feeding system.
Ingate Velocity Distribution: The LDV measurements and cavity fill-time analysis confirmed a fundamental characteristic of this system: the velocity of metal through each ingate remains constant throughout the filling cycle. This is a direct consequence of the near-constant effective pressure head maintained by the connected, pressurized gating network as the cavity fills. The linear relationship between cavity fill height and time, with correlation coefficients above 0.99, empirically validates this constant velocity assumption, which is crucial for modeling. The velocity was influenced by the pouring basin head; a higher head generally increased ingate velocity. The effect of the pouring basin orifice diameter (d0) was non-linear, with velocity decreasing for larger orifices at very low heads due to increased wall-flow and friction, but increasing with orifice size at higher, more practical pouring heads.
Observation of Flow Phenomena and Defect Formation Mechanisms
The transparency of the acrylic model provided unparalleled insight into flow structures that contribute to defects in actual machine tool castings.
Formation of the Impact Turbulence Zone and Gas Entrainment: As the fluid falls from the sprue into the main runner and changes direction into the branching sprue, high turbulence is generated. This turbulence entrains air, creating a frothy mixture. When this stream impinges on the free surface in the branching sprue well, it creates a churning “Impact Turbulence Zone.” While some bubbles escape, others are carried down into the branching runner and can ultimately enter the mold cavity through the first few ingates, potentially creating gas holes in the final casting if the metal temperature is low.
Formation of the ‘V’-Shaped Zone and Gas Ingestion: A more persistent source of gas was identified at the junction between the branching sprue and the branching runner. The sudden change in flow direction at the sharp corner created a “dead zone” of relatively stagnant fluid. Adjacent to this, an unfilled, V-shaped air pocket formed along the top of the branching runner’s initial section. Air bubbles carried by the flow readily rose into this pocket and traveled along the top of the runner, inevitably being drawn into the very first ingate. Simply adding a well at the base of the branching sprue did not eliminate this V-zone. We successfully mitigated this by designing a “gas collection pocket” with a vent at this junction. This pocket trapped the ascending bubbles and vented them out, dramatically reducing gas entry into the casting.
Jetting Phenomenon and Its Prevention: The initial pressure spike when an ingate starts flowing can cause a high-velocity jet into the mold cavity. This can lead to defects like sand erosion, dross entrainment, and cold shuts. We found that increasing the size of vents on slag traps at the end of branching runners helped dampen this jet. Furthermore, implementing a conical-shaped slag collector with a vent at the runner end effectively buffered the flow and almost completely eliminated the jetting effect.
Slag/Dross Entrainment and Filtration: Using polystyrene foam particles to simulate non-metallic inclusions, we traced their path. Most slag was carried with the initial flow front and floated on the surface in the branching runner. While a simple trapezoidal slag trap at the runner end caught some, a significant amount could still be drawn into ingates. The conical centrifugal slag collector with a contracted inlet proved highly effective. The contraction increased velocity, inducing a rotational flow in the collector that efficiently separated and trapped the slag particles, preventing them from entering the machine tool casting cavity.
Theoretical Modeling and Numerical Analysis
Building upon the experimental data, we developed a mathematical model to facilitate the rational design of the bottom-gated shower system for machine tool casting.
Optimization of Pouring Time: Selecting the optimal pouring time (T) is critical. It involves balancing thermal and hydrodynamic factors to achieve sound casting. We formulated this as a constrained nonlinear optimization problem:
$$\text{Minimize: } F = T$$
$$\text{Subject to: } T_{min} \le T \le T_{max} \quad \text{and} \quad v_{min} \le \frac{P}{T} \le v_{max}$$
where \(P\) is the casting height, and \(v\) is the average rise velocity of the metal in the cavity. The limits \(T_{min}, T_{max}, v_{min}, v_{max}\) depend on the metal’s properties and casting geometry. This one-dimensional problem was solved using the Complex Method algorithm to find the optimal \(T^*\).
Mathematical Model of the Gating System: The model is based on hydraulic principles. Applying Bernoulli’s equation between the pouring basin surface (point 0) and the exit of the i-th ingate (point i), and considering the back-pressure from the rising metal in the cavity, we derive the mean flow velocity \(v_i\) for each ingate:
$$v_i = \mu_i \sqrt{2gH_{pi}}$$
Here, \(\mu_i\) is the effective flow coefficient for the path to the i-th ingate, accounting for all friction and local losses, and \(H_{pi}\) is the effective metallostatic pressure head for that ingate, adjusted for back-pressure. For a bottom-gated system, the average effective head \(H_p\) is given by:
$$H_p = H_0 – \frac{P}{2}$$
where \(H_0\) is the initial sprue height and \(P\) is the total cavity height.
The total loss coefficient for each flow path is the sum of friction losses in each segment (sprue, runner, etc.) and local losses at every junction (bends, contractions, etc.):
$$\sum \xi_i = \sum (\lambda_j \frac{l_j}{d_j} + \zeta_k)$$
The friction factors (\(\lambda\)) and local loss coefficients (\(\zeta\)) were determined empirically from our velocity and pressure measurements. A sample of calculated loss coefficients (\(\sum \xi_i\)) for different ingates under specific pouring conditions is shown below:
| Ingate Number (i) | 1 | 2 | 3 | 4 | 5 | 6 |
|---|---|---|---|---|---|---|
| \(\sum \xi_i\) (H0=400mm, d0=30mm) | 6.24 | 6.31 | 6.35 | 6.40 | 6.52 | 6.72 |
Determination of Gating System Dimensions: With the optimal pouring time \(T^*\) and the calculated ingate velocity \(v_i\) from the model, the required total cross-sectional area of the ingates (\(\sum A_i\)) can be determined from the mass balance:
$$ \sum A_i = \frac{W}{\rho \cdot T^* \cdot \bar{v}} $$
where \(W\) is the casting weight, \(\rho\) is the metal density, and \(\bar{v}\) is the average ingate velocity. For a casting with two symmetrical rows of shower gates, assuming uniform wall thickness, the area for each of the \(n\) ingates per row is:
$$A_i = \frac{W}{2 \cdot n \cdot \rho \cdot T^* \cdot \bar{v}}$$
In practice, for manufacturing convenience, ingate areas are often made equal: \(A_1 = A_2 = … = A_n\). Once the total ingate area is fixed, the areas of other gating components (runners, sprues) are determined based on established gating ratios to create a non-pressurized (open) system, commonly used for this application to minimize turbulence. A typical ratio is:
$$\sum A_{sprue} : \sum A_{main-runner} : \sum A_{branch-runner} : \sum A_{ingates} = 1.0 : 1.5 : 2.0 : 4.0$$
This entire design process—from data input and pouring time optimization to hydraulic calculation and drafting—was integrated into a computer program, enabling the Computer-Aided Design (CAD) of the bottom-gated shower system for bed-type machine tool castings.
Conclusion
This first-person experimental investigation provides a comprehensive hydrodynamic analysis of the bottom-gated shower gating system. We have demonstrated that hydraulic simulation is an invaluable tool for visualizing flow phenomena and optimizing system geometry to prevent defects in large machine tool castings. Key findings include the inherent stability of the system, evidenced by uniform flow distribution, constant ingate velocity, and stable pressure during filling—attributes highly beneficial for producing sound, dimensionally accurate castings.
Furthermore, we identified and proposed solutions for critical defect-forming mechanisms: the ‘V’-shaped gas pocket (mitigated by a vented collection pocket) and slag entrainment (effectively controlled by a centrifugal slag collector). These design modifications directly address common quality issues in heavy-section foundry practice.
The developed mathematical model, grounded in experimental data for loss coefficients, provides a rational method for calculating gating dimensions based on an optimized pouring time. The implementation of this methodology into a CAD system marks a significant step from empirical rule-of-thumb towards a science-based design process for the production of high-integrity machine tool castings. This work underscores that a deep understanding of fluid flow within the gating system is fundamental to achieving consistent quality and economic efficiency in the casting of heavy machinery components.