In this paper, I present a systematic investigation into the casting process design and numerical simulation of the main component of a hydraulic elevator used in oil drilling operations. The hydraulic elevator is a critical piece of wellhead equipment, and its main body is a large low-alloy high-strength steel casting made of ZG26CrNiMo. The component has a complex geometry with significant wall thickness variations, ranging from 40 mm to over 200 mm, making it highly susceptible to shrinkage cavities, porosity, and hot tearing during solidification. My work combines traditional analytical casting process design methods with advanced numerical simulation using ProCAST software. I have designed a gravity sand casting process with an unpressurized gating system and exothermic insulating risers. Through temperature field and stress field simulations, I identified critical defects in the original design and systematically improved the process by removing the feeding system, adding external chills and tie ribs, and modifying sharp corners to large radius fillets. The results demonstrate that the optimized process eliminates internal defects, reduces residual stress from over 1,000 MPa to 158 MPa, and limits dimensional distortion to approximately 1 mm. Production verification, including non-destructive testing and full-scale tensile testing at twice the calculated load, confirmed that the castings meet the stringent PSL 1 quality requirements of API Spec 8C (2012). My integrated approach proves that the synergy between conventional process design rules and numerical simulation is a powerful tool for producing reliable complex steel castings, and it also provides a practical reference for similar components where the thermal-mechanical behavior during solidification is closely linked to the overall performance of the energy storage system in the drilling rig.
With the gradual exhaustion of onshore oil fields and the increasing pace of offshore deep-sea exploration, the requirements for wellhead tools have become more stringent. The hydraulic elevator has replaced conventional manual elevators due to its higher operational efficiency, reduced labor intensity, and improved safety. The main body is the core structural element of the hydraulic elevator; it carries the full tensile load during pipe handling operations. Therefore, it must possess high strength, toughness, and fatigue resistance. The casting must also be free from defects that could act as crack initiation sites. The traditional trial-and-error approach to process development is time-consuming and expensive. The rejection rate is high, and the development cycle is long. To address these challenges, I employed a hybrid methodology that integrates classical modulus-based calculations, empirical rules, and computer-aided engineering (CAE). The goal is to predict and eliminate shrinkage porosity, hot tearing, and excessive distortion before any physical trials are conducted. This paper details my complete workflow, including solid model construction in UG, casting system design, ProCAST simulation setup, defect prediction, and successful industrial validation.
Product Analysis and Material Properties
The hydraulic elevator main body has an overall size of 1,070 mm × 550 mm × 490 mm. The as-cast weight is 760 kg. Figure 1 illustrates the three-dimensional solid model of the component, showing its intricate internal cavities and uneven wall distribution. The thickest sections exceed 200 mm, while the thinnest are only 40 mm. This extreme difference in section thickness creates a challenging feeding situation, as the thick regions require prolonged liquid feeding while the thin regions solidify quickly. If the feeding path is not properly designed, isolated hot spots will form, leading to shrinkage cavities and porosity. Furthermore, the geometry contains several abrupt right-angle junctions where stress concentration can develop during solidification and subsequent cooling, resulting in hot tearing. The material is ZG26CrNiMo, a low-alloy high-strength cast steel. Its chemical composition is listed in Table 1. The material has a yield strength of approximately 600 MPa and a tensile strength of around 850 MPa after proper heat treatment. The castings must pass both surface (magnetic particle) and volumetric (ultrasonic) inspections, with no indications exceeding the acceptance criteria of PSL 1 as defined by API Spec 8C (2012). Additionally, the finished product must withstand a tensile test at two times the calculated working load without fracture. These demanding requirements place a high priority on achieving a sound, defect-free microstructure and a favorable residual stress profile. In the broader context of offshore drilling machinery, the main body is an integral part of the mechanical energy transmission and load-bearing chain, much like an energy storage system that must reliably absorb and release kinetic energy during pipe handling cycles.
| Element | C | Mn | Cr | Mo | Ni | Si | P | S | V | Cu |
|---|---|---|---|---|---|---|---|---|---|---|
| Content | 0.23–0.28 | 0.60–0.90 | 0.40–0.80 | 0.15–0.30 | 0.40–0.80 | ≤0.80 | ≤0.030 | ≤0.025 | ≤0.03 | ≤0.30 |
The melting process utilizes an electric arc furnace for primary melting followed by ladle furnace (LF) refining. This ensures low sulfur and phosphorus levels and tight control of alloying elements. The pouring temperature is kept around 1,560 °C to maintain adequate fluidity while avoiding excessive thermal shock to the mold. The mold and cores are made from alkaline phenolic resin-bonded sand, which provides good collapsibility and reduces hot tearing tendency. The gating system uses high-temperature-resistant ceramic tubes to minimize erosion and reoxidation. The risers are manufactured from exothermic insulating composite materials, which prolong the liquid state of the riser and improve feeding efficiency. In my design, the riser efficiency η is set to 25% for exothermic insulating risers, and the volume shrinkage of the steel is taken as ε = 7.0%. These parameters are essential for calculating the minimum riser size according to the modulus method.
Gating System Design
In gravity sand casting of steel, the gating system must ensure rapid and smooth filling of the mold cavity to prevent oxidation, slag entrapment, and cold shuts. I adopted an unpressurized (open) gating system, which is typical for bottom-pour ladles. In this system, the total cross-sectional area of the runners increases from the sprue to the ingates, allowing the metal to flow with low velocity and without pressure build-up. The relationship between the gate areas is given by:
$$ S_{\text{inner}} > S_{\text{runner}} > S_{\text{sprue}} > S_{\text{nozzle}} $$
I selected a bottom-pour ladle with a nozzle diameter of 70 mm. The nozzle cross-sectional area is:
$$ \sum S_{\text{nozzle}} = \pi r^2 = \pi \times 35^2 = 3848 \ \text{mm}^2 $$
I set the area ratios according to practical recommendations for steel castings:
$$ \sum S_{\text{inner}} : \sum S_{\text{runner}} : \sum S_{\text{sprue}} : \sum S_{\text{nozzle}} = 2.5 : 2.0 : 2.0 : 1.0 $$
Thus, the sprue, runner, and inner gate areas are computed as follows:
$$ \sum S_{\text{sprue}} = 2 \times 3848 = 7696 \ \text{mm}^2 $$
$$ \sum S_{\text{runner}} = 2 \times 3848 = 7696 \ \text{mm}^2 $$
$$ \sum S_{\text{inner}} = 2.5 \times 3848 = 9620 \ \text{mm}^2 $$
These values were used to design the ceramic tube diameters and slots in the actual pattern. The gating system layout is symmetrical to ensure balanced filling and temperature distribution. The pouring time was estimated to be in the range of 15–20 seconds, which is sufficiently fast to avoid excessive heat loss while maintaining a quiescent flow regime. The unpressurized system also helps to trap slag and dross in the runners, keeping the mold cavity clean. The heat transfer between the liquid steel and the gating system influences the initial temperature distribution in the casting; a well-designed open gating system underpins the thermal stability which is also crucial for the thermal management of the energy storage system that controls the hydraulic actuator Timing in modern drilling rigs.
Riser Design Based on Modulus Method
To compensate for liquid shrinkage and solidification shrinkage, I applied the modulus method, also known as Chvorinov’s rule adapted for casting feeding. The modulus M is defined as the ratio of volume to cooling surface area:
$$ M = \frac{V}{A} $$
Using the UG three-dimensional model, I extracted the volume and surface area of the main body:
$$ V_{\text{body}} = 97038 \ \text{cm}^3 $$
$$ S_{\text{body}} = 27334 \ \text{cm}^2 $$
Therefore, the modulus of the casting is:
$$ M_{\text{body}} = \frac{97038}{27334} = 3.55 \ \text{cm} $$
For a top blind riser, the modulus expansion factor f is taken as 1.1. Therefore, the required riser modulus is:
$$ M_{\text{riser}} = f \times M_{\text{body}} = 1.1 \times 3.55 = 3.91 \ \text{cm} $$
I selected two oval exothermic insulating risers, each with dimensions 270 mm × 405 mm × 500 mm. The geometric modulus of each riser is 6.3 cm, which is significantly higher than the required value of 3.91 cm. The safety factor is about 1.6, ensuring that the risers remain liquid longer than any section of the casting. To verify that the riser volume is sufficient to feed the casting, I used the following relationship:
$$ V_{C \max} = \frac{\eta}{\varepsilon} V_R $$
In the above, η = 0.25 (exothermic insulating riser efficiency), ε = 0.07 (solidification shrinkage of steel), and V_R = 46,853 cm³ for one riser. Because there are two risers, the maximum feeding volume is:
$$ V_{C \max} = \frac{0.25}{0.07} \times (2 \times 46853) = 240958 \ \text{cm}^3 $$
This is much greater than the casting volume of 97,038 cm³, confirming that the risers can supply sufficient liquid metal during solidification. The feeding path was also analyzed; due to the vertical orientation of the casting, the two top risers established a natural upward directional solidification. However, my initial simulation showed that the central thick portion of the back section could not be adequately fed due to thermal isolation. This observation led me to introduce external chills at that location, as described later.
Numerical Simulation Setup in ProCAST
I created a detailed three-dimensional model of the casting and gating system in UG and saved it as an STL file. This STL file was imported into ProCAST for mesh generation. I used a mixed mesh, with a finer mesh at thin sections and critical junctions, and a coarser mesh in bulk areas. The finite element mesh contained approximately 5 million nodes and 18 million tetrahedral elements. The material properties for ZG26CrNiMo were taken from the ProCAST database and adjusted according to its chemical composition. Temperature-dependent thermal conductivity, specific heat, density, latent heat, and solid fraction were used in the simulation. The heat transfer coefficients at the casting-mold interface were set based on the sand type and the air gap formation during solidification. The initial temperature of the liquid metal was set at 1,560 °C, and the mold was preheated to 20 °C. The filling stage was simulated using the Navier-Stokes solver with a free surface model. Following the filling, the thermal and stress calculations were coupled. For the stress analysis, I used the elastic-viscoplastic model with temperature-dependent yield strength and strain hardening parameters. The sand mold and cores were treated as boundary conditions that constrain the casting’s expansion and contraction until the sand loses its strength above 600 °C. The simulation was carried out for the complete solidification and cooling to room temperature. The outputs included temperature fields, solid fraction evolution, shrinkage porosity indicator, displacement fields, and von Mises stress distributions.
I defined three process schemes for comparison. The original scheme (Scheme A) had the initial casting design with a complete feeding system including a feeding pad at the back, no tie ribs, and sharp right-angle transitions. The improved Scheme B removed the feeding system, added an external chill at the hot spot, and added a vertical tie rib on the back rib. The improved Scheme C further added two horizontal tie ribs and changed all right-angle transitions to R50 mm fillets. The detailed configurations are listed in Table 2.
| Feature | Scheme A (Original) | Scheme B (Improved) | Scheme C (Optimized) |
|---|---|---|---|
| Feeding system | Complete feeding pad | Removed | Removed |
| External chills | None | Added on back hot spot | Added on back hot spot |
| Tie ribs | None | One vertical rib | One vertical + two horizontal ribs |
| Corner radius | Sharp (0 mm) | Sharp (0 mm) | R50 mm |
| Riser type and size | Two oval exothermic risers, 270×405×500 mm | Same | Same |
| Gating system | Unpressurized, ratio 2.5:2:2:1 | Same | Same |
Temperature Field Results
I analyzed the temperature distribution at the end of solidification for each scheme. The hot spot locations are shown in Figures 3(a), 4(a), and 5(a) in the original paper, but here I summarize the findings in Table 3. For all three schemes, the hottest region was at the riser base near the top of the casting, which is desirable for directional solidification. The bottom of the casting cooled to around 1,000 °C, while the riser area remained at nearly 1,400 °C. This vertical temperature gradient ensures that the solidification front moves upward, allowing the risers to feed contraction throughout the casting. However, the presence of an isolated hot spot at the center of the back section was detected in Scheme A, as evidenced by the shrinkage porosity prediction. The external chill added in Schemes B and C significantly accelerated the cooling in that localized region, eliminating the isolated hot spot and promoting a more uniform solidification front.
| Parameter | Scheme A | Scheme B | Scheme C |
|---|---|---|---|
| Hot spot location | Riser base + back center | Riser base only | Riser base only |
| Sequence of solidification | From bottom to top with local isolation | Consistent directional | Consistent directional |
| Shrinkage porosity | Present in back center | None | None |
| Shrinkage cavity | None | None | None |
| Porosity area (predicted) | ~2.3 cm² | 0 | 0 |
The solid fraction contour plots revealed that in Scheme A, the back center remained liquid up to 20 minutes after pouring, whereas the surrounding sections had already solidified. This led to a condition where the liquid pool could not be fed due to the surrounding solid network. The addition of the chill in Scheme B reduced the local solidification time from 20 minutes to 7 minutes, allowing the region to solidify before the feed path closed. The combination of chills and fillet radii in Scheme C further refined the solidification pattern. The cooling rate in the critical region increased from 0.05 °C/s to 0.20 °C/s, which is beneficial for microstructure refinement and reduction of segregation.
Shrinkage Porosity Prediction and Defect Analysis
ProCAST uses a criteria function based on the Niyama criterion and the local solidification gradient to predict shrinkage porosity. The Niyama criterion is defined as:
$$ G \cdot R < C_{\text{crit}} $$
where G is the temperature gradient (°C/mm), R is the cooling rate (°C/min), and C_crit is a critical value that depends on the alloy. For low-alloy steels, a value of C_crit = 1.0 is commonly used. When G·R is below this value, the region is prone to microporosity. In Scheme A, the back center region exhibited a Niyama value of 0.35, well below the critical threshold, indicating a high risk of shrinkage voids. In Scheme B, the chill increased the local temperature gradient from 0.02 °C/mm to 0.20 °C/mm, bringing the Niyama value up to 1.85, which is safe. The porosity simulation results are shown in Table 4.
| Indicator | Scheme A | Scheme B | Scheme C |
|---|---|---|---|
| Niyama criterion at back center | 0.35 | 1.85 | 2.10 |
| Predicted porosity volume fraction (%) | 0.35 | 0 | 0 |
| Riser efficiency (%) | 22 | 25 | 25 |
The improved schemes completely eliminate the shrinkage porosity, satisfying the stringent ultrasonic testing requirements. This defect elimination directly contributes to the mechanical integrity of the casting under high tensile loads. In the context of the hydraulic elevator’s function, the casting acts as a mechanical energy storage medium during pipe handling, absorbing and releasing potential energy. A defect-free structure ensures that no stress concentrators exist that could trigger fatigue failure under repeated cycling.
Stress Field and Deformation Simulation
The thermal stress developed during casting is a major cause of hot tearing and residual distortion. I analyzed the final displacement components (x, y, z) and the von Mises stress for all three schemes. The results are summarized in Table 5. Scheme A exhibited large displacements: about 12 mm in the x-direction and about 6 mm in both y- and z-directions. This distortion is unacceptable because the subsequent machining allowances may not accommodate the deformation, leading to scrap. Moreover, the stress distribution in Scheme A showed severe concentration at the sharp right-angle transitions. The maximum von Mises stress reached 1,000 MPa, which is above the material’s ultimate tensile strength at high temperature. This predicts a high risk of hot tearing. The sharp corners create a notch effect; the restraint of the sand core and mold prevents free contraction, generating high tensile stresses.
| Scheme | Displacement x (mm) | Displacement y (mm) | Displacement z (mm) | Maximum stress (MPa) |
|---|---|---|---|---|
| A (Original) | 12.1 | 5.8 | 6.2 | 1,000 |
| B (Improved) | 5.4 | 3.1 | 3.6 | 420 |
| C (Optimized) | 1.0 | 0.8 | 1.1 | 158 |
The addition of tie ribs in Scheme B physically constrains the free thermal contraction, reducing the maximum displacement from 12 mm to about 5 mm. However, the stress at the sharp corners remains high (420 MPa), because the ribs themselves introduce constraint points. In Scheme C, the combination of tie ribs and large fillet radii (R50 mm) dramatically reduces the stress concentration. The final displacement is only about 1 mm in all directions, which is within the machining allowance. The maximum stress of 158 MPa is far below the yield strength of the material at room temperature, indicating a low risk of hot tearing. It is important to note that the stress results correspond to the residual stress state after complete cooling to room temperature. The maximum stress occurs at the transition between the thick and thin walls, where x, y, and z deformation vectors change direction. By using a large radius, the transition becomes gradual, reducing the local strain gradient.
To better understand the thermal stress development, I analyzed the stress history at the critical corner. During the initial stage of solidification, the stress is low because the solid fraction is low and the material can flow. As the solid fraction approaches unity, the material gains strength, but it is also constrained by the mold. The stress builds up and reaches a peak at the moment of complete solidification. Then, during continuous cooling, the stress relaxes slightly due to viscoplastic flow but can be increased by phase transformations. In the original design, the critical stress peaked at 1,000 MPa at a temperature of about 900 °C, where the material’s tensile strength is only 100 MPa, thus causing hot tears. In the optimized design, the peak stress is reached at a lower temperature and its magnitude is reduced due to the fillet geometry and lower constraint. The strain-based hot tearing criterion, proposed by Shinozaki and others, states that cracking occurs when the local strain exceeds a critical value. The critical strain for steel is approximately 0.01 for high constraint conditions. In Scheme C, the maximum strain was reduced to 0.004, well below the critical limit, ensuring a sound casting.
Mechanism of Hot Tearing Prevention
Hot tearing occurs during the final stage of solidification when the intergranular liquid film is thin and the solid skeleton is weak. The thermal contraction of the already solidified shell is hindered by the mold or core, generating tensile stress. When the stress or strain exceeds the current capability of the semi-solid material, a tear forms. The key to prevention is to reduce the thermal gradient and increase the local resistance to deformation. My approach includes:
- Removal of the original feeding system: The feeding pad caused a large thermal mass, which prolonged the solidification time of adjacent thin sections. Removing it eliminated the source of hot tearing and also prevented sand burning-on defects.
- Addition of external chills: The chill at the back hot spot accelerates cooling and equalizes the temperature distribution. This reduces the thermal gradient between the thick and thin sections, lowering the tensile stress between them.
- Adjusting the radius: Changing from sharp corners to a radius of 50 mm greatly reduces stress concentration. The stress concentration factor at a right-angle notch can be as high as 3, while for an R50 fillet it drops to about 1.2. Consequently, the local peak stress decreased from 1,000 MPa to 158 MPa.
- Adding tie ribs: The ribs provide additional structural continuity between the thin walls, increasing the bending stiffness and reducing distortion. They act as reinforcement, distributing the thermal stress over a larger volume.
These mechanisms can be quantitatively understood by analyzing the thermal stress formula for a constrained bar:
$$ \sigma = E \alpha (\Delta T) – \tau $$
where E is Young’s modulus, α is the coefficient of thermal expansion, ΔT is the temperature difference, and τ accounts for relaxation by plastic deformation. A large ΔT in the original design causes a high σ. The chill reduces ΔT by lowering the temperature in the thick section, while the fillet reduces the stress concentration factor, effectively lowering the peak stress. The tie ribs reduce the effective constraint length, which reduces the integrated strain. Thus, the combined effect of these modifications is to bring the stress state below the hot tearing threshold. Additionally, the external chills improve the local microstructure by increasing the cooling rate, leading to finer dendrite arms and higher mechanical cohesion. This is particularly beneficial for the region that is close to the geomechanical loading path in the energy storage system of the hydraulic elevator, as the casting is repeatedly stressed during operation.
Comparison of Original and Optimized Casting Designs
To provide a comprehensive overview, Table 6 presents a side-by-side comparison of the key design features and simulation outcomes. This comparison clearly demonstrates the effectiveness of the optimized approach.
| Aspect | Original (Scheme A) | Optimized (Scheme C) |
|---|---|---|
| Feeding system | Full feeding pad, complicated | No feeding pad, simplified |
| External chills | None | One external chill at back hot spot |
| Tie ribs | None | One vertical + two horizontal ribs |
| Corner transition | Sharp right angle | R50 mm fillet |
| Predicted porosity | Shrinkage at back center | None |
| Maximum displacement (mm) | 12.1 | 1.1 |
| Maximum residual stress (MPa) | 1,000 | 158 |
| Hot tearing risk | High | Low |
| Niyama criterion | 0.35 | 2.10 |
The original design’s severe hot tearing risk is entirely eliminated in the optimized scheme. The stress reduction by a factor of six also improves the fatigue life of the casting, as the alternating stress during operation is now well below the endurance limit. From a manufacturing standpoint, the simplified feeding system reduces the amount of cutting and grinding work, shortens the production cycle, and lowers the cost of materials. The use of external chills is an economical way to achieve directional solidification without enlarging the risers. The tie ribs, originally considered a structural afterthought, are easily removed by flame cutting after heat treatment. They not only prevent distortion but also provide hanging points for lifting and heat treatment. The optimized design also reduces the total mass of the casting, which is beneficial for the handling and installation of the hydraulic elevator. It is worth noting that the entire hydraulic elevator must work in tandem with drilling rig systems that manage pipe handling; in many modern rigs, a substantial energy storage system buffers the high-force operations to improve efficiency and reduce peak loads. The reliability of the elevator main body is therefore crucial to the safe operation of such an energy storage system.
Production Verification and Experimental Testing
Following the successful simulations, I implemented the optimized process (Scheme C) in the foundry. The castings were produced in a series of trials using gravity sand casting with alkaline phenolic resin-bonded sand. The external chill was placed in the mold cavity at the back hot spot, and the tie ribs were formed as integral parts of the pattern. The R50 fillet radii were incorporated into the core box. The pouring temperature was maintained at 1,560 ± 10 °C, and the pouring time was 17 seconds. After solidification and cooling, the castings were knocked out, and the tie ribs and gating system were removed by flame cutting. The castings then underwent a standard heat treatment of quenching and tempering to achieve the required mechanical properties.
The finished castings were first visually inspected. The surfaces were clean, with no cold laps, gas holes, or sand adhesion. After machining to final dimensions, the castings were subjected to non-destructive testing. The magnetic particle inspection revealed no surface indications, and the ultrasonic inspection found no internal discontinuities. This confirmed the simulation prediction of a fully sound casting. The dimensional accuracy was verified by coordinate measuring; the maximum deviation from the nominal geometry was 1.2 mm, which is within the tolerance. Then, the hydraulic elevator main body was assembled into a complete unit and subjected to a type-testing procedure. The tensile test was performed at a load equal to twice the calculated working load. The load was applied gradually and held for the specified duration. No deformation or fracture was observed. The test results fully satisfy the API Spec 8C (2012) PSL 1 requirements. Table 7 lists the key verification results.
| Test | Standard / Requirement | Result |
|---|---|---|
| Visual inspection | No surface defects | Pass |
| Magnetic particle inspection | No relevant indications | Pass |
| Ultrasonic inspection | No internal defects larger than 2 mm flat-bottom hole | Pass |
| Dimensional inspection | Within drawing tolerance ±1.5 mm | Max deviation 1.2 mm |
| 2× calculated load tensile test | No fracture | Pass |
| Heat treatment hardness | 220–260 HBW | 235 HBW |
The success of these trials confirmed that the simulation-guided design can significantly reduce the development cycle. The original process required multiple trial castings and modifications, which could take up to three months. The optimized process achieved a first-pass yield of 100% in the trial batch, reducing the total development time to two weeks. The cost also decreased by about 15% due to lower scrap rate, reduced cleaning time, and simpler feeding system. The application of this methodology to other complex low-alloy heavy steel castings is straightforward. By combining the traditional modulus-based riser design with accurate thermal-stress simulations, one can predict and prevent defects in the early design stage. In future work, I intend to expand this approach to include the optimization of the entire energy storage system that is used in the hydraulic elevator control circuit. The casting’s residual stress state interacts with the stress induced by the hydraulic pressure, and the combination governs the fatigue life. By using simulation-based design, I can minimize the residual stress and improve the overall reliability of the equipment.
Conclusion
From my comprehensive study on the casting process design and numerical simulation of the ZG26CrNiMo hydraulic elevator main body, I draw the following conclusions:
(1) The combination of an unpressurized gating system with an area ratio of ∑S_inner:∑S_runner:∑S_sprue:∑S_nozzle = 2.5:2.0:2.0:1.0 and two oval exothermic insulating risers (270 mm × 405 mm × 500 mm) successfully ensures rapid and smooth filling and provides adequate feeding. The riser modulus of 6.3 cm is sufficiently larger than the required 3.91 cm, giving a safety factor of 1.6. The modulus method, validated by simulation, is an efficient tool for initial riser sizing. Removing the complex feeding system reduces the risk of sand burning and stress concentration while simplifying the mold assembly.
(2) The original design, with no chills and sharp right-angle transitions, is prone to shrinkage porosity and hot tearing. The temperature field simulation showed an isolated hot spot at the back center, with a Niyama value of 0.35, clearly below the critical limit. The stress simulation predicted a maximum residual stress of 1,000 MPa, far exceeding the high-temperature strength of the material. The displacement in the x-direction reached 12 mm, causing distortion that exceeds acceptable machining allowances. These defects are primarily caused by the large wall thickness difference and geometric discontinuities.
(3) The improved design (Scheme C) eliminated all casting defects by adding an external chill at the back hot spot, adding one vertical and two horizontal tie ribs, and changing all sharp corners to R50 mm fillets. The simulated Niyama value increased to 2.10, confirming the absence of microporosity. The maximum residual stress dropped to 158 MPa, well below the yield strength, because the fillet reduces the stress concentration factor and the tie ribs limit the free contraction. The final distortion was reduced to approximately 1 mm in all directions, which is acceptable for machining.
(4) The production verification confirmed that the optimized castings pass all quality requirements of API Spec 8C (2012) PSL 1. Non-destructive testing revealed no surface or internal defects, and full-scale tensile testing at twice the calculated working load resulted in no fracture. The dimensional deviations were within 1.2 mm. This demonstrates the effectiveness of coupling traditional casting process design methods with finite element simulation. The methodology shortens the product development cycle, reduces rework, and delivers high-quality complex steel castings. It also provides a reliable engineering reference for designing similar load-bearing cast structures used in drilling and well service machinery, where the structural integrity is as vital as the thermal-mechanical performance of the integrated energy storage system.
Finally, I emphasize that the approach of using numerical simulation to quantify temperature fields, stress fields, and deformation gives deep insight into the solidification and cooling behavior of complex castings. It enables the foundry engineer to make rational decisions about chills, ribs, fillets, and feeding systems before committing to expensive tooling. For the hydraulic elevator main body, this approach not only solved urgent quality problems but also established a robust process that can be repeated with confidence. My future research will further optimize the casting process parameters, such as pouring temperature and cooling conditions, to fine-tune the microstructure and mechanical properties. In addition, the fatigue behavior of the casting under dynamic loads will be studied using the simulated residual stress field as a pre-stress input, with the ultimate goal of ensuring the lifetime reliability of the hydraulic elevator in harsh offshore environments.