In my development work on high-pressure hydraulic cylinder bodies for injection molding machine clamping systems, I treated ductile iron castings as the central material and process platform because they combine the load-bearing character of steel with the shaping freedom and damping capacity of cast iron. The component I focused on had a large outer envelope of approximately 990 mm in diameter and 625 mm in height, a mass of about 1.3 t, and a maximum local wall thickness of 135 mm. The internal hydraulic chamber and the flange transition created a severe thermal center, and the finished hydraulic chamber had to be machined to a radial surface roughness of Ra 0.8 μm while withstanding a hydraulic test pressure of 20 MPa without leakage or weeping. These requirements placed the ductile iron castings in a demanding category in which ordinary shrinkage porosity, slag inclusion, cold shut, crack, or graphite degeneration cannot be tolerated.
When I reviewed the conventional route, I found that the usual practice was to place the hydraulic chamber upward and the flange downward. In that configuration, the thermal center at the junction between the hydraulic chamber and the flange required extensive chilling and large risers or exothermic risers. Although such an approach can sometimes produce acceptable ductile iron castings, I observed several persistent penalties. The process yield was low, large risers had to be removed, many dedicated shaped chills had to be placed and controlled, cleaning and grinding were heavy, and the production cost remained high. More importantly, the upward chamber orientation often produced poor external quality on the lower side of the hydraulic chamber, increased the risk of inclusion entrapment, and created a tendency for oil seepage or leakage after machining. For ductile iron castings intended for 20 MPa hydraulic service, those defects are not acceptable because even a small connected pore network can become a leak path under pressure.
I therefore rethought the process from the standpoint of thermal gradient, mold rigidity, gas evacuation, and liquid metal cleanliness. My objective was to produce dense ductile iron castings with a uniform spheroidal graphite structure, a high nodularity, and a low level of inclusions while simplifying the mold and reducing the dependence on chromium ore sand and shaped chills. The work combined a high-stiffness core concept, a gating system with controlled area ratios and transition runners, precise molten iron chemistry, silicon carbide addition, spheroidization, and multi-stage inoculation. I then verified the result through microstructure examination, nondestructive testing, and mechanical property testing of cast-on samples.
The first technical problem I addressed was the core and mold configuration. In the conventional route, the hydraulic chamber core is often made from chromium ore sand. Chromium ore sand has a high density and a strong chilling tendency, but I found that it can create practical difficulties: the sand core may lack sufficient strength, the cooling may be non-uniform, surface sand may detach, and gas and inclusion defects may form in the hydraulic chamber. Those defects then reduce the density of the ductile iron castings and make it difficult to achieve Ra 0.8 μm after machining. A rough or porous surface can also become an initiation site for leakage under 20 MPa hydraulic pressure. I consequently designed a high-stiffness core structure in which steel tubes act as the skeleton and sand provides the supporting and filling medium. The hydraulic chamber is not formed by a fragile chromium ore sand core alone, and the dedicated chills at the flange are removed. Instead, the steel tubes reinforce the sand mold, support the core, and simultaneously act as internal cooling passages. Air can flow through the steel tube cavities after pouring, so heat is extracted from the casting in a controlled and distributed way. The result is a stronger mold, less sand consumption, better gas evacuation, and a more uniform chilling effect around the heavy hydraulic chamber.
I found that this hybrid steel-tube and sand structure is especially valuable for ductile iron castings with large thermal centers because it changes both the mechanical stability and the heat-transfer path. The steel skeleton prevents core collapse and sand wash during filling. The sand coating controls the local heat-transfer coefficient so that the chilling is not excessively abrupt. The air flowing inside the tubes removes heat from the tube wall, and the tube wall in turn extracts heat from the surrounding sand and metal. If I express the local heat removal as a simple convective relation, the heat flux from the tube inner surface can be written as
$$q = h A \left(T_{\mathrm{tube}} – T_{\mathrm{air}}\right)$$
where q is the heat flow, h is the convective heat-transfer coefficient, A is the internal tube area, T_tube is the tube wall temperature, and T_air is the air temperature. In my design, the tube layout increases A and promotes air movement, so q is distributed over a larger region rather than concentrated at a few shaped chills. This is important for ductile iron castings because non-uniform chilling can produce local carbides, hardness variation, or residual stress. The steel-tube skeleton also reduces the amount of sand that must be packed and baked, which lowers gas evolution and improves venting. I observed that the combination of lower gas pressure and higher mold stiffness reduces the probability of sand drop, slag entrapment, and gas porosity in the hydraulic chamber.
I also evaluated the thermal center using the geometric modulus concept. For a casting section, the modulus is
$$M = \frac{V}{A}$$
where V is the volume of the hot section and A is its cooling surface area. The local solidification time under conductive cooling follows the Chvorinov relation
$$t_s = B M^2$$
in which t_s is the solidification time and B is a mold constant. In the heavy flange-to-chamber transition, M is large, so t_s is long and shrinkage porosity tends to form unless I provide a favorable thermal gradient or a local heat sink. The steel-tube skeleton acts as a controlled heat sink and as a rigidity element. Because the tubes are embedded in the sand rather than directly touching all of the hot metal, the chilling is moderated, and the thermal gradient is directed toward the tube-reinforced regions. I found that this helps the ductile iron castings solidify with a more progressive front and reduces the isolated liquid pockets that cause shrinkage defects.
| Process aspect | Conventional core and chill route | High-stiffness core route I designed | Effect on ductile iron castings |
|---|---|---|---|
| Hydraulic chamber core | Chromium ore sand core with dedicated shaped chills | Steel-tube skeleton with sand filling and coating | Higher core strength, better dimensional stability, less sand-related defects |
| Flange chilling | Dedicated shaped chills at the flange transition | No dedicated shaped chills; controlled steel-tube cooling | Simpler mold control and more uniform thermal field |
| Gas evacuation | Limited by dense chromium ore sand and complex chill placement | Improved by reduced sand volume and open tube channels | Lower gas porosity and inclusion risk |
| Cooling uniformity | Often localized and difficult to balance | Distributed along the steel-tube skeleton | More uniform graphite structure and fewer shrinkage pores |
| Machining quality | Risk of rough spots and subsurface porosity | Denser chamber surface and better roughness potential | Better chance to meet Ra 0.8 μm and 20 MPa leak tightness |
The second major element I controlled was the gating system. For high-pressure ductile iron castings, the gating system must fill the mold quietly, avoid turbulence, promote slag flotation, and establish a favorable temperature field. I designed the system so that the liquid metal enters the hydraulic chamber region through gates connected to the platform on the bottom reverse side of the chamber. That location has a relatively thick section, so the metal can enter the cavity in a smooth and stable manner. If the metal enters through a thin section or drops freely, it can entrain air and oxide films, and those defects may remain in the final ductile iron castings as inclusion stringers or gas pores. I also used transition runners with controlled height differences and a filter between the first and second transition runners. The second transition runner was positioned 20 to 30 mm higher than the second runner, which encouraged slag to float and remain in the transition runner while clean metal moved into the cavity.
I specified the gating area ratios to balance filling velocity, pressure loss, and slag control. The main area ratio I used can be expressed as
$$\Sigma S_{\mathrm{sprue}} : \Sigma S_{\mathrm{runner1}} : \Sigma S_{\mathrm{runner2}} : \Sigma S_{\mathrm{ingate1}} : \Sigma S_{\mathrm{ingate}} = 1.0\!-\!1.1 : 1.3\!-\!1.5 : 1 : 1 : 1$$
and in a second controlled version I also considered an expanded ratio in which the first and second ingate groups were balanced more broadly:
$$\Sigma S_{\mathrm{sprue}} : \Sigma S_{\mathrm{runner1}} : \Sigma S_{\mathrm{runner2}} : \Sigma S_{\mathrm{ingate1}} : \Sigma S_{\mathrm{ingate}} = 1.0\!-\!1.1 : 1.3\!-\!1.5 : 1.3\!-\!1.5 : 1 : 1$$
These ratios are not arbitrary. The sprue must deliver enough metal without excessive velocity, while the runner system must reduce velocity and allow slag to separate. The ingates must fill the cavity without creating a hard jet. If the ingate area is too small, the local velocity becomes high and the mold wall may be washed; if it is too large, the filling may become too slow and cold shut or misrun may occur. I estimated the ingate velocity from
$$v = \frac{Q}{A}$$
where Q is the volumetric flow rate and A is the total ingate area. By keeping v within a controlled range, I reduced the tendency for air entrainment and oxide film breakup. I also checked the Reynolds number as an indicator of flow character:
$$Re = \frac{\rho v D}{\mu}$$
where ρ is density, v is velocity, D is a characteristic hydraulic diameter, and μ is dynamic viscosity. Although the flow in a mold is transient and not a simple pipe flow, the Reynolds number helped me compare designs. I preferred a filling regime that avoided unnecessary turbulence in the early stage and then filled the thick chamber smoothly. The filter between transition runners captured small inclusions, and the elevated second transition runner acted as a slag trap. When I poured the ductile iron castings, I observed that this gating arrangement produced a quieter cavity fill and fewer surface inclusion defects than the conventional top-chamber arrangement.
| Gating element | Design choice I used | Reason | Expected defect reduction |
|---|---|---|---|
| Sprue | Controlled area, smooth taper, low turbulence | Deliver metal without excessive velocity | Reduced air entrainment |
| First runner | Area ratio 1.3–1.5 relative to sprue | Reduce velocity and distribute flow | Less mold erosion |
| Second runner | Area ratio about 1.0 relative to sprue | Maintain controlled pressure | Stable filling |
| Transition runner | Second transition runner 20–30 mm higher than second runner | Promote slag flotation | Reduced slag inclusion |
| Filter | Placed between first and second transition runners | Capture oxide films and inclusions | Cleaner ductile iron castings |
| Ingate | Connected to bottom reverse platform of hydraulic chamber | Enter thick section smoothly | Reduced cold shut and gas porosity |
The third critical area was material chemistry and melt treatment. The component material was QT450-10, which requires a combination of tensile strength, yield strength, elongation, and hardness. Because the casting has a large thermal mass and solidifies slowly, I knew that spheroidization and inoculation could fade before the last liquid solidified. If the magnesium content falls too low, graphite may become flake-like or vermicular; if inoculation fades, carbide may form and elongation may drop. I therefore controlled the charge composition, the carbon equivalent, the sulfur level, the magnesium residual, the rare earth residual, and the antimony addition. The charge consisted of 50% pig iron, 30% steel scrap, and 20% returned scrap. I added silicon carbide at 0.7% of the total charge mass and a carburizer at 0.8% of the total charge mass. The silicon carbide was added with the metallic charge, and the carburizer was added once during charging. After melting, I added FeSi75 at 0.6% of the total charge mass. I then superheated the melt to 1 452 °C to obtain a base iron with approximately C 3.75%, Si 1.55%, Mn 0.25%, P ≤ 0.025%, S ≤ 0.023%, and the balance Fe.
I paid particular attention to the carbon equivalent because it governs fluidity, shrinkage behavior, and graphite formation. I used the standard expression
$$\mathrm{CE} = \mathrm{C} + \frac{1}{3}\mathrm{Si} + \frac{1}{3}\mathrm{P}$$
For the final iron, I achieved C 3.55%, Si 2.45%, Mn 0.25%, P ≤ 0.025%, S ≤ 0.0098%, Mg 0.037%, RE 0.008%, Sb 0.0055%, and CE approximately 4.38%. That carbon equivalent placed the ductile iron castings in a favorable range for spheroidal graphite formation and feeding. I also kept sulfur low because sulfur consumes magnesium and rare earth elements and can cause spheroidization decay. The low phosphorus level helped avoid brittle phosphide eutectic, and the controlled manganese level avoided excessive pearlite or carbide stabilization. The final magnesium and rare earth residuals were sufficient for nodularization but not so high that they caused excessive dross or shrinkage. The antimony addition, although small, helped suppress abnormal graphite and refine the structure.
| Charge material or addition | Amount | Function in the melt | Reason for the level I selected |
|---|---|---|---|
| Pig iron | 50% | Primary iron unit, low residual control | Stable carbon and low tramp elements |
| Steel scrap | 30% | Dilute residuals and adjust carbon | Reduce inherited graphite and gas risk |
| Returned scrap | 20% | Reuse sound foundry returns | Control cost and maintain known chemistry |
| Silicon carbide | 0.7% of total charge | Nucleation, silicon and carbon supply | Promote fine graphite and reduce carbide |
| Carburizer | 0.8% of total charge | Carbon adjustment | Reach target carbon equivalent |
| FeSi75 | 0.6% of total charge | Silicon adjustment and preliminary inoculation | Improve graphitization potential |
| Base iron element | Target or achieved value | Role in ductile iron castings |
|---|---|---|
| C | 3.75% | Graphite formation and fluidity |
| Si | 1.55% | Graphitization and matrix control |
| Mn | 0.25% | Matrix strength with limited carbide risk |
| P | ≤ 0.025% | Reduced phosphide brittleness |
| S | ≤ 0.023% | Limited magnesium consumption before spheroidization |
| Fe | Balance | Base matrix |
For spheroidization, I used a tundish cover method with a rare-earth-magnesium alloy. The alloy contained Mg 5.8%, RE 1.49%, Si 42.8%, Ca 2.45%, Al 0.85%, and the balance Fe. I added the spheroidizing alloy at 1.2% of the original iron mass and added pure Sb at 0.006% of the original iron mass. I then added a silicon-barium inoculant with a particle size of 3 to 8 mm and compacted it. The inoculant contained Si 72%, Ca 1.0%, Ba 2.0%, Al 0.70%, S 0.015%, and the balance Fe. The inoculation addition was 0.62% of the original iron mass. The magnesium reaction lasted about 125 s. After spheroidization, I skimmed the melt and allowed it to settle. When the temperature reached 1 295 °C, I poured the iron into the mold. During pouring, I added a final stream inoculation powder at 0.12%. This multi-stage treatment gave me both strong nodularization and late inoculation, which is essential for large ductile iron castings that cool slowly.
| Treatment step | Material or parameter | Amount or value | Purpose |
|---|---|---|---|
| Spheroidizing alloy | Rare-earth-magnesium alloy | 1.2% of original iron mass | Form spheroidal graphite |
| Antimony addition | Pure Sb | 0.006% of original iron mass | Refine graphite and suppress abnormal shapes |
| Inoculant | Silicon-barium inoculant, 3–8 mm | 0.62% of original iron mass | Increase nucleation and prevent chill |
| Magnesium reaction | Reaction time | 125 s | Ensure consistent spheroidization |
| Pouring temperature | Melt temperature at pouring | 1 295 °C | Balance fluidity and shrinkage |
| Stream inoculation | Inoculation powder | 0.12% | Restore nucleation during filling |
| Final iron element | Achieved value | Process significance |
|---|---|---|
| C | 3.55% | Supports graphite formation and fluidity |
| Si | 2.45% | Promotes ferrite and graphitization |
| Mn | 0.25% | Contributes to strength without excessive carbides |
| P | ≤ 0.025% | Limits brittle phases |
| S | ≤ 0.0098% | Protects magnesium residual |
| Mg | 0.037% | Controls spheroidization |
| RE | 0.008% | Assists spheroidization and impurity control |
| Sb | 0.0055% | Refines graphite and matrix |
| CE | 4.38% | Favorable carbon equivalent for sound ductile iron castings |
I then examined the microstructure of the cast-on test block to confirm that the process had produced the intended graphite morphology and matrix. I cut specimens from the cast-on block, ground and polished them, etched them with a 4% nitric acid alcohol solution, and observed them with a scanning electron microscope equipped with energy-dispersive spectroscopy. The microstructure consisted mainly of a ferritic matrix with uniformly distributed spheroidal graphite. The nodularity was about 94.2%, and the graphite size was approximately size 6. I also observed small lamellar pearlite colonies locally. That pearlite strengthened the matrix without making the ductile iron castings excessively brittle. The nodularity can be expressed as
$$N_{\mathrm{od}} = \frac{A_{\mathrm{spheroidal}}}{A_{\mathrm{total}}} \times 100\%$$
where A_spheroidal is the area of spheroidal graphite and A_total is the total graphite area. I also considered the graphite nodule count per unit area:
$$N_A = \frac{n}{A}$$
where n is the number of nodules and A is the examined area. A high nodule count and a uniform distribution are desirable because they reduce the mean free path for crack initiation and improve ductility. In my samples, the combination of high nodularity, size 6 graphite, and a predominantly ferritic matrix with fine pearlite explains the good elongation and strength I measured later.
The silicon carbide addition played a central role in this microstructure. Silicon carbide has a melting point much higher than the iron melt, so it does not completely dissolve immediately. The reaction can be represented as
$$\mathrm{SiC} + \mathrm{Fe} \rightarrow \mathrm{FeSi} + \mathrm{C}$$
During solidification, silicon atoms released by silicon carbide dissolution dissolve into the austenite matrix, while some carbon atoms cluster under composition and energy fluctuations. When those clusters exceed the critical nucleus radius, they can act as nuclei for spheroidal graphite. In addition, dispersed undissolved silicon carbide particles can provide heterogeneous nucleation sites, which helps increase nodularity and refine the graphite. A thin silicon dioxide film often covers silicon carbide particles and delays the dissolution reaction. That delay prevents the melt from becoming uniform too early and allows carbon clusters to persist for a longer time. I also noted that undissolved silicon carbide particles can act as heterogeneous substrates for Fe3C, increasing the nucleation density and causing Fe3C and ferrite to grow in a more compact alternating pattern. That mechanism refines pearlite spacing and lamellae, which contributes to strength and toughness in the ductile iron castings.
| EDS point | Si (%) | C (%) | O (%) | Fe (%) | Interpretation |
|---|---|---|---|---|---|
| Point 1 | 3.05 | 22.38 | 26.94 | 47.63 | Oxide-related or carbide-related region with high carbon and oxygen |
| Point 2 | 2.70 | 1.69 | 0.98 | 94.63 | Predominantly metallic matrix |
| Point 3 | 6.24 | 4.10 | 8.95 | 80.71 | Silicon- and oxygen-enriched local region |
| Point 4 | 1.45 | 32.33 | 0.78 | 65.44 | Carbon-rich graphite or carbide-related region |
After pouring and cooling, I inspected the trial ductile iron castings. Ultrasonic testing met the level 1 requirement of EN12680-3, and magnetic particle testing met the level 2 requirement of EN13697. No other casting defects were found in the hydraulic cylinder. The cast-on test block, with dimensions of 70 mm × 105 mm × 210 mm, was used for mechanical property testing. The measured tensile strength was 405 MPa, the yield strength was 278 MPa, the elongation after fracture was 18.5%, and the Brinell hardness was 148 HBW. The standard required tensile strength of at least 390 MPa, yield strength of at least 260 MPa, elongation of at least 8.0%, and hardness between 140 and 190 HBW. Therefore, all measured values met the specification. The high elongation is particularly important because hydraulic cylinder bodies must tolerate machining stresses and pressure cycling without cracking. The microstructure and chemistry together explain why the ductile iron castings reached this property balance.
| Property | Specification | Measured value | Assessment |
|---|---|---|---|
| Tensile strength | ≥ 390 MPa | 405 MPa | Passed |
| Yield strength | ≥ 260 MPa | 278 MPa | Passed |
| Elongation after fracture | ≥ 8.0% | 18.5% | Passed |
| Hardness | 140–190 HBW | 148 HBW | Passed |
| Inspection method | Standard or requirement | Result | Meaning for ductile iron castings |
|---|---|---|---|
| Ultrasonic testing | EN12680-3 level 1 | Satisfied | Indicates internal soundness and absence of harmful reflectors |
| Magnetic particle testing | EN13697 level 2 | Satisfied | Indicates surface and near-surface crack control |
| Hydraulic service requirement | 20 MPa without leakage or weeping | Satisfied in trial validation | Confirms pressure tightness of the ductile iron castings |
| Machined surface requirement | Ra 0.8 μm on hydraulic chamber radial surface | Met after process control | Reduces leak paths and improves sealing |
I believe the success of this work came from the interaction of several factors rather than from one isolated change. The high-stiffness core improved mold stability and gas evacuation while providing distributed chilling. The gating system delivered clean metal into the thick chamber bottom without turbulence and gave slag a chance to separate. The chemistry control ensured that the carbon equivalent, sulfur, magnesium, rare earth, and antimony levels were all in a favorable window. The silicon carbide addition and multi-stage inoculation provided nucleation sites and delayed fade. The result was a dense, high-nodularity structure with fine graphite and a matrix that balanced strength and elongation. For high-pressure ductile iron castings, that combination is essential because a leak is usually the result of several small defects acting together: a shrinkage pore, an oxide film, a graphite degeneration zone, or a rough machining surface. By controlling each of those risks, I reduced the probability that they would connect into a leak path.
I also considered the thermal and pressure conditions during service. The hydraulic chamber must withstand 20 MPa, which can be written as
$$P_{\mathrm{hyd}} = 20\,\mathrm{MPa}$$
Under that pressure, the wall experiences hoop stress and radial stress. A simplified thin-wall estimate for hoop stress is
$$\sigma_{\theta} = \frac{P D}{2t}$$
where D is the internal diameter and t is the wall thickness. A thick-wall cylinder would require a more complete Lamé solution, but the thin-wall estimate is useful for comparing designs. The important point is that the stress is carried by the matrix and graphite structure. If graphite is flake-like or clustered, the effective load-bearing section is reduced. If porosity is present, the local stress concentration rises. Therefore, the high nodularity and uniform graphite distribution I achieved directly support pressure integrity. The ferritic matrix gives ductility, while the fine pearlite contributes strength. The low phosphorus and controlled manganese reduce brittle phases. The low sulfur protects magnesium, and the antimony addition helps avoid abnormal graphite. Each of these details matters in ductile iron castings for hydraulic service.
I further analyzed the solidification sequence with a simple heat balance. The heat released during solidification can be approximated as
$$Q_{\mathrm{solid}} = \rho V \left(L + c \Delta T\right)$$
where ρ is density, V is volume, L is latent heat, c is specific heat, and ΔT is the superheat or undercooling range. The mold and core must remove this heat without creating isolated liquid regions. The steel-tube skeleton increases the effective heat-transfer area and provides a path for air cooling. The sand coating moderates the heat-transfer coefficient so that the metal does not freeze too quickly at the surface while the interior remains liquid. This balance is particularly important in the flange-to-chamber junction, where the modulus is high and the feeding distance is long. If I had used only chromium ore sand, the chilling would have been strong but uneven, and the sand core would have been more fragile. If I had used only sand without the steel-tube skeleton, the thermal center would have persisted and shrinkage porosity would have been more likely. The hybrid design gave me both stiffness and a controlled cooling path.
In terms of process yield, I found that eliminating large risers and dedicated shaped chills simplified the molding line and reduced cleaning labor. The steel-tube skeleton also lowered sand consumption, which reduced core-making time and gas evolution. The gating system with a filter and elevated transition runner reduced the amount of slag that reached the hydraulic chamber, so the repair and grinding workload decreased. These improvements are not only economic; they also improve quality consistency. When the process has fewer loose variables, it is easier to keep the ductile iron castings within the required window. I consider this especially important for a component that must be leak-tight at 20 MPa and machined to Ra 0.8 μm. A process that depends on many manually placed chills and large risers is difficult to control repeatedly, whereas a rigid core with a fixed steel-tube skeleton and a well-defined gating ratio is more reproducible.
I also reviewed the role of inoculation timing. In large ductile iron castings, the melt may remain liquid for a long time, so the beneficial nuclei introduced by inoculation can fade before the last eutectic cells form. My approach used a silicon-barium inoculant in the spheroidizing ladle and a final stream inoculation of 0.12% during pouring. The barium and calcium in the inoculant help preserve nucleation potential and resist fade. The rare earth elements in the spheroidizing alloy assist in controlling sulfur and oxygen, but they must not be too high or they can promote dross and Chunky graphite. The final RE of 0.008% was low but effective in combination with Mg 0.037%. The antimony addition of 0.0055% helped refine graphite without causing excessive pearlite. This balanced chemistry is one reason the nodularity reached 94.2% and the graphite size reached approximately size 6.
I can summarize the main process windows I used for the ductile iron castings in a compact parametric form. The carbon equivalent was
$$\mathrm{CE} = \mathrm{C} + \frac{1}{3}\mathrm{Si} + \frac{1}{3}\mathrm{P} \approx 4.38\%$$
The final magnesium-to-sulfur ratio was high enough to leave residual magnesium after desulfurization:
$$\frac{\mathrm{Mg}}{\mathrm{S}} = \frac{0.037}{0.0098} \approx 3.78$$
That ratio supported stable spheroidization. The nodularity target was
$$N_{\mathrm{od}} \ge 90\%$$
and I achieved about 94.2%. The graphite size target was around size 6, which is favorable for strength and ductility. The tensile strength target was
$$\sigma_b \ge 390\,\mathrm{MPa}$$
and the measured value was 405 MPa. The yield strength target was
$$\sigma_y \ge 260\,\mathrm{MPa}$$
and the measured value was 278 MPa. The elongation target was
$$\delta \ge 8.0\%$$
and the measured value was 18.5%. The hardness target was
$$140 \le \mathrm{HBW} \le 190$$
and the measured value was 148 HBW. These numbers show that the process produced ductile iron castings with a safe margin in both strength and ductility.
| Parameter | Target or design window | Achieved result | Comment |
|---|---|---|---|
| Carbon equivalent | Near 4.38% | 4.38% | Favorable for graphite and fluidity |
| Residual Mg | Sufficient for spheroidization | 0.037% | Controlled without excessive dross |
| Residual S | As low as practical | ≤ 0.0098% | Protects Mg and RE |
| Nodularity | ≥ 90% | 94.2% | High spheroidal graphite fraction |
| Graphite size | Around size 6 | Size 6 | Fine and uniform distribution |
| Tensile strength | ≥ 390 MPa | 405 MPa | Passed with margin |
| Yield strength | ≥ 260 MPa | 278 MPa | Passed with margin |
| Elongation | ≥ 8.0% | 18.5% | High ductility |
| Hardness | 140–190 HBW | 148 HBW | Good machinability and strength balance |
From a practical standpoint, I found that the steel-tube skeleton also improved the dimensional accuracy of the hydraulic chamber. A rigid core resists distortion during pouring and solidification, so the wall thickness remains more uniform. Uniform wall thickness means a more balanced thermal field, which in turn reduces shrinkage porosity. It also means that the machining allowance can be controlled more tightly. For a hydraulic chamber that must be machined to Ra 0.8 μm, a uniform and dense surface layer is much easier to achieve than a surface with scattered pores or sand inclusions. I observed that the trial castings had a cleaner chamber surface and fewer subsurface defects, which reduced the risk of leakage during the 20 MPa test.
I also evaluated the effect of the gating system on inclusion control. Oxide films and slag droplets form in ductile iron because magnesium and rare earth elements react with oxygen and sulfur. If these reaction products enter the cavity as large films, they can remain as stringers or planar defects. The filter between the transition runners captured many of these particles. The elevated second transition runner gave the slag additional time to float. The smooth ingate entry into the bottom reverse platform prevented the metal from falling and folding. I also kept the pouring temperature at 1 295 °C, which gave enough fluidity for the large casting while avoiding excessively high superheat that would increase gas absorption and shrinkage. The stream inoculation at 0.12% provided late nucleation without adding too much material that could create dross. These details collectively improved the cleanliness of the ductile iron castings.
When I compare this route with the conventional route, the difference is not just one component. The conventional route uses chamber-up orientation, chromium ore sand, dedicated shaped chills, and large risers. The route I developed uses a rigid steel-tube-reinforced core, distributed air cooling, a controlled gating system with a filter and transition runner, silicon carbide pretreatment, low-sulfur chemistry, rare-earth-magnesium spheroidization, antimony modification, and multi-stage inoculation. Each element supports the others. The rigid core supports the thermal design; the thermal design supports the gating design; the gating design supports cleanliness; the chemistry supports nucleation; and the nucleation supports the final mechanical properties. This system-level approach is what allowed the ductile iron castings to meet the hydraulic pressure and machining requirements.
I also considered the role of antimony more carefully. Antimony is a surface-active element that can refine graphite and increase nodule count when added in very small amounts. However, too much antimony can promote pearlite and reduce elongation. At 0.0055% in the final iron, the antimony addition reduced the tendency for abnormal graphite and helped refine the structure without making the matrix overly pearlitic. The final matrix was still predominantly ferritic, which is consistent with the high elongation of 18.5%. This balance is important for hydraulic components because they need both pressure tightness and damage tolerance. A fully pearlitic matrix might increase strength but reduce ductility and increase cracking risk during machining or pressure cycling. A fully ferritic matrix might give high ductility but lower strength. The mixed matrix I obtained, with fine pearlite islands in a ferritic base, gave a useful combination.
The silicon carbide addition also influenced the melt in a way that goes beyond simple chemistry. Because silicon carbide dissolves gradually, it provides a local source of silicon and carbon that can promote graphite nucleation. The reaction product FeSi and C can participate in the eutectic reaction. The undissolved particles can act as heterogeneous nuclei. The SiO2 film on the particle surface can slow dissolution and prevent the melt from becoming uniform too quickly. This means that carbon clusters can form and survive long enough to act as graphite nuclei. In large ductile iron castings, this effect is valuable because the cooling rate is slow and the number of active nuclei can decrease over time. By adding silicon carbide in the charge and combining it with late inoculation, I was able to maintain a high nodule count throughout the solidification of the heavy section.
I also controlled the pouring temperature precisely. At 1 295 °C, the melt had enough fluidity to fill the large hydraulic chamber and the flange transition, but it was not so hot that it would dissolve excessive sand binder, vaporize moisture, or increase shrinkage. A lower pouring temperature would risk cold shut and misrun, especially in the thin outer walls. A higher pouring temperature would increase gas absorption, sand burn-on, and shrinkage porosity. The 1 295 °C pouring temperature, combined with the favorable gating system, gave a stable fill. I also used the stream inoculation at the moment of pouring to compensate for the thermal and time-related loss of nucleation sites. This is a standard approach for ductile iron castings, but the amount and timing have to be adapted to the casting mass and section thickness. In my case, 0.12% stream inoculation was sufficient.
After the trial production, I confirmed that the hydraulic cylinder body met the design requirements. The nondestructive tests showed no unacceptable indications. The microstructure showed high nodularity and uniform graphite. The mechanical properties exceeded the minimum requirements. The machining trial showed that the hydraulic chamber could reach Ra 0.8 μm and be tested at 20 MPa without leakage. I therefore concluded that the process design was feasible for stable production. The key was not simply to add more chills or a larger riser, but to control the entire solidification and filling process. The high-stiffness core, the gating system, the silicon carbide pretreatment, the spheroidization and inoculation treatment, and the precise chemistry all worked together.
I can express the final process route as a sequence of controlled states. The base iron was melted to a target carbon and silicon level. Silicon carbide and carburizer were added to adjust nucleation and carbon equivalent. FeSi75 was added for silicon and preliminary inoculation. The melt was superheated to 1 452 °C. Spheroidization was performed with a rare-earth-magnesium alloy and antimony. Inoculation was performed with a silicon-barium alloy. The final chemistry was adjusted to C 3.55%, Si 2.45%, Mn 0.25%, P ≤ 0.025%, S ≤ 0.0098%, Mg 0.037%, RE 0.008%, Sb 0.0055%, and CE 4.38%. The melt was skimmed and held until 1 295 °C. It was then poured with stream inoculation at 0.12%. The mold used a steel-tube-reinforced sand core and a gating system with controlled area ratios, a filter, and an elevated transition runner. After cooling, the ductile iron castings were inspected and tested.
| Stage | Control variable | Value or action | Effect |
|---|---|---|---|
| Charging | Pig iron, steel scrap, returned scrap | 50%, 30%, 20% | Stable base chemistry |
| Pretreatment | Silicon carbide and carburizer | 0.7% and 0.8% | Nucleation and carbon adjustment |
| Silicon addition | FeSi75 | 0.6% | Silicon control and graphitization |
| Superheat | Melt temperature | 1 452 °C | Homogeneous melt and reaction control |
| Spheroidization | Rare-earth-magnesium alloy and Sb | 1.2% and 0.006% | Spheroidal graphite and refinement |
| Inoculation | Silicon-barium alloy | 0.62% | Nucleation and chill prevention |
| Pouring | Temperature and stream inoculation | 1 295 °C and 0.12% | Fluidity and late nucleation |
| Mold | Steel-tube-reinforced core and gating system | Controlled area ratios and filter | Soundness, cleanliness, and uniform cooling |
I also want to emphasize that the hydraulic chamber geometry is not a simple cylinder. The transition from the chamber to the flange creates a change in section modulus, and that change is where shrinkage porosity tends to appear. The steel-tube skeleton provides local cooling near that transition without the hard contact of a chill. The sand coating distributes the heat extraction. The gating system brings hot metal into the thick bottom platform, which helps establish a favorable temperature gradient toward the cooler tube-reinforced regions. If the hot metal were introduced at the top, the thermal gradient would be reversed and the lower flange would cool too early, leaving a liquid pocket in the chamber. My design avoids that mistake. I believe this thermal-gradient control is one of the most important reasons the ductile iron castings were sound.
In addition, I found that the steel tubes improve the exhaust path. During pouring, the mold generates steam and binder decomposition gases. If these gases cannot escape, they can be trapped in the metal and form blowholes. The open tube channels and the reduced sand volume give gases a path to move away from the hydraulic chamber. The sand permeability around the tubes is also better than a dense chromium ore sand core. Better venting means lower gas pressure at the metal front, which reduces gas porosity and surface defects. For ductile iron castings that must be leak-tight, even small gas pores are harmful because they can connect under pressure. Therefore, the venting improvement is not secondary; it is a direct quality requirement.
I also considered the influence of the steel tube on the local cooling rate. The tube wall has a higher thermal conductivity than sand, so it extracts heat from the adjacent sand and metal. However, because the tube is covered by sand, the heat extraction is not as severe as a bare metal chill. This moderated chilling is beneficial for ductile iron because it avoids the formation of hard, brittle zones while still promoting a steep enough temperature gradient to feed the thermal center. I found that the combination of moderate chilling and high mold stiffness produced a denser structure without excessive carbides. The graphite remained spheroidal, and the matrix remained predominantly ferritic with fine pearlite. This is exactly the structure needed for a high-pressure hydraulic cylinder body.
When I evaluated the final results, I used several criteria. The first was internal soundness by ultrasonic testing. The second was surface and near-surface integrity by magnetic particle testing. The third was microstructure, including nodularity, graphite size, and matrix. The fourth was mechanical properties, including tensile strength, yield strength, elongation, and hardness. The fifth was machining and pressure performance, including Ra 0.8 μm and 20 MPa leak tightness. All criteria were met. I therefore concluded that the process is reliable for producing high-quality ductile iron castings of this type. The approach can be adapted to other large hydraulic components with similar wall thickness transitions and pressure requirements.
In my view, the most important lesson is that high-pressure ductile iron castings cannot be produced by simply copying a general casting layout. The process must be designed around the specific thermal center, the pressure requirement, the machining requirement, and the risk of graphite degeneration. The steel-tube-reinforced core changes the mechanical and thermal behavior of the mold. The controlled gating system changes the filling and inclusion behavior. The silicon carbide and chemistry control change the nucleation and solidification behavior. The multi-stage inoculation changes the graphite morphology and matrix. Together, these changes produce a sound casting. I would use the same logic for other ductile iron castings: first identify the dominant defect risk, then design the mold, gating, and melt treatment to remove that risk before it can form.
For the high-pressure hydraulic cylinder body, the dominant risks were shrinkage porosity at the chamber-to-flange transition, slag and gas inclusion in the hydraulic chamber, graphite degeneration due to slow cooling, and leakage under 20 MPa. I addressed each risk directly. Shrinkage porosity was addressed by the steel-tube skeleton and controlled thermal gradient. Slag and gas inclusion were addressed by the gating system, filter, elevated transition runner, and improved venting. Graphite degeneration was addressed by silicon carbide, low sulfur, controlled magnesium and rare earth, antimony, and multi-stage inoculation. Leakage was addressed by the combination of dense microstructure, high nodularity, and machining quality. The final casting met all requirements, which confirmed that the design logic was sound.
I also recorded the main process parameters in a form that can be used for repeat production. The charge composition, silicon carbide addition, carburizer addition, FeSi75 addition, superheat temperature, spheroidizing alloy addition, antimony addition, inoculant addition, pouring temperature, and stream inoculation are all fixed. The gating area ratios are fixed. The steel-tube skeleton layout is fixed. The molding sand and coating practice are fixed. When these parameters are held within their windows, the ductile iron castings should remain consistent. I would monitor the residual magnesium, residual sulfur, carbon equivalent, nodularity, and mechanical properties as the main control indicators. If any of these drift, the process window can be adjusted before defects appear.
| Control indicator | Target window | Reason for monitoring | Action if out of range |
|---|---|---|---|
| Carbon equivalent | About 4.38% | Controls graphite and shrinkage | Adjust carburizer or silicon addition |
| Residual Mg | Around 0.037% | Controls spheroidization | Adjust spheroidizing alloy or reaction conditions |
| Residual S | ≤ 0.0098% | Protects Mg and RE | Improve desulfurization or charge cleanliness |
| Residual RE | About 0.008% | Avoids dross and abnormal graphite | Adjust rare-earth content |
| Sb | About 0.0055% | Refines graphite | Adjust small addition |
| Nodularity | ≥ 90% | Ensures ductility and pressure tightness | Review inoculation and chemistry |
| Graphite size | Around size 6 | Balances strength and ductility | Review cooling and nucleation |
| Tensile strength | ≥ 390 MPa | Meets design requirement | Review matrix and defects |
| Elongation | ≥ 8.0% | Ensures damage tolerance | Review pearlite content and inclusions |
| Hardness | 140–190 HBW | Ensures machinability and strength | Review cooling and matrix |
I also note that the use of silicon carbide and antimony is not an isolated additive strategy. It is part of a nucleation and refinement strategy. Silicon carbide provides carbon and silicon and promotes heterogeneous nucleation. Antimony refines graphite and helps suppress abnormal shapes. The rare-earth-magnesium alloy provides the primary spheroidizing effect. The silicon-barium inoculant provides late nucleation and chill prevention. The final stream inoculation provides a last increment of nucleation during filling. This sequence is important because each stage has a different time constant. If I relied on only one inoculation step, the effect might fade before the heavy section solidified. By distributing the nucleation potential across several stages, I kept the graphite spheroidal and fine throughout the casting. That is why the ductile iron castings achieved 94.2% nodularity and size 6 graphite even with a maximum wall thickness of 135 mm.
In terms of mechanical behavior, the measured elongation of 18.5% is a strong indicator that the matrix and graphite structure are healthy. Low elongation would suggest carbide formation, flake graphite, or inclusion content. The measured tensile strength of 405 MPa and yield strength of 278 MPa show that the matrix is strong enough for the hydraulic load. The hardness of 148 HBW is within the specified range and indicates good machinability. I found that these properties are consistent with a ferritic matrix containing fine pearlite and uniformly distributed spheroidal graphite. The antimony and silicon carbide additions helped refine the pearlite and graphite, while the low sulfur and controlled magnesium helped maintain spheroidization. The result is a material that can carry 20 MPa hydraulic pressure and still be machined to Ra 0.8 μm without excessive tool wear or surface damage.
I would summarize the overall process as a chain of controlled events. First, the mold is made rigid and thermally balanced by the steel-tube-reinforced sand core. Second, the gating system delivers clean metal into the thick chamber bottom with low turbulence. Third, the melt is pretreated with silicon carbide and carburizer to establish a favorable carbon equivalent and nucleation state. Fourth, spheroidization and inoculation are performed with rare-earth-magnesium, antimony, and silicon-barium additions. Fifth, the final chemistry is adjusted to low sulfur, controlled magnesium, and balanced carbon equivalent. Sixth, the melt is poured at 1 295 °C with stream inoculation. Seventh, the casting cools under distributed chilling and solidifies with a progressive front. Eighth, the ductile iron castings are inspected and machined. Every link in this chain supports the final requirement of leak-tightness at 20 MPa and surface roughness of Ra 0.8 μm.
I also found that the process is more robust than the conventional route because it removes the most fragile elements. Chromium ore sand cores are replaced by a steel-tube skeleton with sand, so core breakage and sand wash are less likely. Dedicated shaped chills are eliminated, so chill placement errors and chill-related gas defects are reduced. Large risers are reduced or eliminated, so cleaning and riser removal are simplified. The gating system includes a filter and transition runner, so inclusion control is built into the design. The chemistry includes silicon carbide and antimony, so nucleation and refinement are built into the melt. This robustness is important for production because it reduces the dependence on operator skill and manual intervention. For ductile iron castings that must meet strict hydraulic requirements, robustness is as important as the initial trial result.
In conclusion, I designed and validated a casting process for high-pressure hydraulic cylinder bodies made from QT450-10 ductile iron. The process used a high-stiffness steel-tube-reinforced sand core, a controlled gating system with defined area ratios and a filter, silicon carbide pretreatment, precise melt chemistry, rare-earth-magnesium spheroidization, antimony modification, and multi-stage inoculation. The resulting ductile iron castings achieved 94.2% nodularity, size 6 graphite, a predominantly ferritic matrix with fine pearlite, tensile strength of 405 MPa, yield strength of 278 MPa, elongation of 18.5%, and hardness of 148 HBW. Ultrasonic testing met EN12680-3 level 1, magnetic particle testing met EN13697 level 2, and the hydraulic chamber met the 20 MPa leak-tightness and Ra 0.8 μm machining requirements. I therefore conclude that the process is suitable for stable production of these high-pressure ductile iron castings and can provide a reliable basis for quality control in similar hydraulic components.