I have spent many years developing and controlling foundry processes for steel components that must resist pressure, wear, and cyclic loading. Among the methods available to me, skin-dried sand mould casting has proved to be one of the most practical and reliable routes for high-pressure castings when the production batch is small or medium and when the part geometry demands strong mould surfaces, stable cores, and clean machining surfaces. I do not treat skin-dried sand mould casting as a simple variation of green sand moulding. I treat it as a controlled process in which the mould cavity surface is dried to a defined depth while the bulk of the sand mould remains green, resilient, and economical.
In our steel foundry, skin-dried sand mould casting has been used for various industrial and mining replacement parts. Later, I applied the same method to a family of high-pressure castings for oilfield wellhead equipment. These castings belong to petroleum machinery and must satisfy strict leak-tightness requirements. The main family contains three parts, and the largest of them is a four-way block with large and small flanges. I will describe how I selected process parameters, how I designed the feeding and gating systems, how I prepared and dried the sand moulds, and how I controlled pouring and testing so that the castings met the drawing requirements and passed hydrostatic testing.
The phrase sand mould casting covers a wide range of moulding technologies. In my work, the specific variant is skin-dried sand mould casting. The mould is made from a moulding sand mixture similar to that used in green sand moulding, but after moulding and coating, the cavity surface is dried by a controlled bake. The result is a mould with a dry, strong surface layer and a green, less expensive body. This combination is particularly useful for high-pressure steel castings because the surface layer resists metal erosion, reduces gas-related defects, and helps prevent burn-on and scabbing.
Why I Selected Skin-Dried Sand Mould Casting
When I compare skin-dried sand mould casting with green sand mould casting, the first difference I observe is surface strength. A green sand mould has adequate strength for many general castings, but its surface contains a relatively high amount of moisture. When high-temperature steel enters the cavity, that moisture can produce steam, and the resulting gas can create blowholes, pinholes, and surface defects. A skin-dried mould has a dry surface layer, so the amount of moisture immediately available at the metal-mould interface is much lower. This reduces the tendency for gas holes, burn-on, scab, and erosion.
When I compare skin-dried sand mould casting with fully dry sand moulding, the economic advantage is clear. A fully dry mould requires the entire mould to be baked, which consumes fuel and electric power and lengthens the production cycle. A skin-dried mould requires only the cavity surface to be dried to a controlled depth. Therefore, I save energy, shorten the production cycle, reduce the workload in the drying area, and improve the working environment and labor conditions. For small and medium batch production, this balance between quality and cost is very attractive.
| Comparison Item | Green Sand Mould Casting | Skin-Dried Sand Mould Casting | Fully Dry Sand Mould Casting |
|---|---|---|---|
| Surface strength | Moderate | High in the dried surface layer | High throughout the mould |
| Surface moisture | High | Low at the cavity surface | Very low |
| Gas-hole tendency | Higher | Lower | Low |
| Burn-on and scab tendency | Higher | Lower | Low |
| Erosion resistance | Moderate | High | High |
| Fuel and electric power consumption | Low | Moderate and localized | High |
| Production cycle | Short | Moderate | Long |
| Working environment | Good but moisture-related issues can occur | Improved compared with dry moulding | More heat, dust, and handling |
| Best application | General castings | High-pressure, dense castings in small and medium batches | Special high-quality castings where full baking is justified |
I also found that skin-dried sand mould casting is not merely a compromise. It can produce pressure-tight castings when the process is designed around the solidification behavior of the steel. The dry surface layer is only one part of the story. The feeder must be correctly sized, the gating system must be non-turbulent, the cores must be supported and vented, the chills must promote directional solidification, and the pouring temperature and time must be controlled. When these factors are combined, skin-dried sand mould casting becomes a reliable production method for high-pressure components.
Part Family and Service Requirements
The high-pressure castings I refer to are used in oil extraction wellhead equipment. The main components are three parts, and the largest is a four-way block. This four-way block has large flanges, small flanges, and intersecting passages. It must be dense because it contains sealing surfaces and is subjected to hydrostatic testing. A small internal defect near a sealing ring groove can cause leakage, sweating, or pressure drop, so the casting must be sound not only in the bulk section but also at the surfaces that will be machined and sealed.
The material is a cast steel grade suitable for pressure service. The technical requirements are strict. First, the casting must receive stress-relief annealing to remove casting stresses. Second, after rough machining, the part must be quench-and-tempered to achieve the specified hardness. Third, the sealing ring groove must not contain any small sand holes, gas holes, or other defects that can affect sealing performance. Fourth, the part must pass a hydrostatic test for a specified duration without leakage, sweating, or pressure drop.
| Requirement | Technical Meaning | Process Response in Skin-Dried Sand Mould Casting |
|---|---|---|
| Stress-relief annealing | Reduce residual stresses from solidification and cooling | Control cooling after shakeout and perform a defined annealing cycle |
| Quench and temper after rough machining | Obtain specified hardness and mechanical properties | Provide sufficient machining allowance and avoid surface defects that deepen during machining |
| Sealing ring groove soundness | No small sand holes, gas holes, or porosity affecting sealing | Use skin-dried surface, clean coating, controlled pouring, and proper feeding near the flange |
| Hydrostatic test | No leakage, sweating, or pressure drop within the test duration | Ensure directional solidification and eliminate shrinkage porosity in pressure-containing sections |
For this type of casting, I cannot rely on inspection alone. I must design the process so that defects are prevented. The skin-dried sand mould casting method gives me a strong cavity surface and a low-moisture interface, but it does not automatically eliminate shrinkage. The feeding system must remain liquid long enough to compensate for liquid and solidification shrinkage. The gating system must fill the cavity smoothly and avoid direct impingement on cores. The chills must create the correct temperature gradient. The pouring practice must avoid re-moistening the mould surface and must avoid excessive temperature that causes burn-on and gas absorption.
Parting Line, Core Support, and Venting
I selected a symmetrical parting line for the four-way block. The large and small flanges are placed on the side of the mould, so all four through-openings lie in the parting plane. This arrangement has several advantages. It makes the moulding process simpler because the flange faces and passages are accessible. It supports the cores more effectively because the core prints can be located at the parting plane. It also improves venting because the core gases can escape along the parting line and through the core prints. In sand mould casting, core support and venting are critical for high-pressure castings because a shifted core or trapped gas can produce an internal defect that is difficult to repair.
I use horizontal moulding with the parting line passing through the four openings. The large flanges and small flanges are positioned so that the metal can flow into the cavity without excessive turbulence. The cores are designed with adequate core prints and with internal vent passages. I also ensure that the core assemblies are stable during pouring. If a core is not well supported, the metallostatic pressure of the steel can cause movement, and the resulting wall thickness variation can create either a thin section or a heavy section. Both are undesirable. A thin section may cause premature solidification and incomplete filling, while a heavy section may create a hot spot and shrinkage porosity.
The parting line also affects the placement of feeders. Because the large flanges are on the side, their upper sections become natural locations for open feeders. I place the feeders on the top of the two large flanges. This location is close to the heavy flange sections and allows the feeder to feed the flange and the intersecting body. The small flanges are filled through the gating system and are not used as primary feeder locations. This choice keeps the feeder count low and concentrates feeding where the thermal mass is greatest.
| Design Decision | Selection | Reason |
|---|---|---|
| Moulding position | Horizontal moulding with symmetrical parting | Simplifies moulding, core setting, and venting |
| Flange position | Large and small flanges on the side | Places all four through-openings at the parting plane |
| Core support | Core prints at the parting plane | Improves stability and reduces core shift |
| Venting | Vents along parting line and through core prints | Allows core gas to escape and reduces blowholes |
| Feeder location | Top of the two large flanges | Feeds the heaviest sections and promotes directional solidification |
| Gating location | Bottom flask flange area | Avoids direct metal impingement on cores and reduces turbulence |
Pattern Allowances and Shrinkage
For sand mould casting, pattern dimensions must account for liquid and solidification shrinkage, machining allowance, draft, and core print clearance. I use separate external and internal shrinkage allowances. The external dimensions of the casting are controlled by the pattern, while internal passages are controlled by the core. Because the steel shrinks during cooling, the pattern must be larger than the finished casting for external dimensions. For internal dimensions, the core must be made so that the hole or passage reaches the correct size after shrinkage. In practice, I apply an external shrink allowance and an internal shrink allowance that are selected according to the steel grade, section size, and restraint from the mould and cores.
$$L_{\text{pattern}} = L_{\text{casting}} \left(1 + s_e\right)$$
$$L_{\text{core}} = L_{\text{internal}} \left(1 – s_i\right)$$
Here, \(s_e\) is the external shrinkage allowance and \(s_i\) is the internal shrinkage allowance. The exact values depend on the steel and the casting geometry, but for the high-pressure steel castings in this family, I keep the external allowance somewhat larger than the internal allowance. I also add machining allowance on the flange faces, sealing surfaces, and other surfaces that will be machined. The draft angle is selected so that the pattern can be withdrawn without damaging the skin-dried surface. The core prints are made strong enough to support the cores and to provide a clear vent path.
| Allowance Type | Symbol | Applied To | Purpose |
|---|---|---|---|
| External shrinkage | \(s_e\) | Outside dimensions, flange diameters, length, width | Compensate for steel contraction after solidification |
| Internal shrinkage | \(s_i\) | Bores, passages, through-holes | Ensure internal dimensions after cooling |
| Machining allowance | \(m\) | Sealing surfaces, flange faces, machined bores | Remove surface defects and achieve final dimensions |
| Draft | \(\alpha_d\) | Pattern sides | Allow clean pattern withdrawal from the sand mould |
| Core print clearance | \(c_p\) | Core prints and core seats | Prevent core damage and allow assembly |
I do not rely only on a single shrinkage factor for every dimension. The flange sections are heavy, and the intersecting body has a different cooling rate. If I use one arbitrary value, the flange diameter may be correct while the passage spacing is wrong, or the sealing surface may clean up on one side but not on the other. Therefore, I record the actual shrinkage behavior from the first articles and adjust the pattern for the production batch. In sand mould casting, this feedback is important because the mould and core materials restrain contraction differently from a fully dry mould or a permanent mould.
Feeder Design and Calculation
The feeder, or riser, is the most important factor for eliminating shrinkage porosity in high-pressure castings. I use a waist-conical open feeder placed on the top of each large flange. The open feeder allows the feeder cavity to be filled from the top and permits the addition of insulating materials. The waist-conical shape provides a favorable solidification profile. I calculate the feeder using a proportional method and verify it with the modulus concept. The basic requirement is that the feeder modulus must be larger than the casting modulus at the hot spot.
$$M = \frac{V}{A}$$
$$M_f \ge 1.2 M_c$$
In these equations, \(M\) is the modulus, \(V\) is the volume, \(A\) is the heat-transfer surface area, \(M_f\) is the feeder modulus, and \(M_c\) is the casting modulus at the section to be fed. For steel castings, I usually require the feeder modulus to exceed the casting modulus by a factor that ensures the feeder remains liquid longer than the hot spot. The exact factor depends on the feeder shape, insulation, and pouring temperature. In this high-pressure casting, I used a factor of at least 1.2 and adjusted it with the measured hot-spot size.
I first estimate the average wall thickness and the height of the heavy section. Then I determine the hot-spot circle at the junction between the flange and the conical body. The feeder width is selected from the hot-spot diameter and the height-to-width ratio. For a waist-conical open feeder, I use the following relationships:
$$W_f = k D_h$$
$$C = 2\sqrt{2 R W_f – W_f^2}$$
$$W_t = W_b + 2 H_f \tan \alpha_f$$
Here, \(W_f\) is the feeder width, \(D_h\) is the hot-spot circle diameter, \(k\) is an empirical coefficient, \(C\) is the feeder chord length, \(R\) is the radius of the flange circle, \(W_t\) is the top width, \(W_b\) is the bottom width, \(H_f\) is the feeder height, and \(\alpha_f\) is the feeder taper angle. I round the calculated values to practical dimensions and then check the feeder volume against the shrinkage requirement.
| Feeder Parameter | Symbol | Design Basis | Process Target |
|---|---|---|---|
| Feeder type | — | Waist-conical open feeder | Good feeding and easy insulation |
| Feeder location | — | Top of the two large flanges | Feed the heaviest sections |
| Number of feeders | \(n_f\) | Two large flange hot spots | Two feeders |
| Feeder width | \(W_f\) | Hot-spot diameter and height ratio | Selected from proportional calculation |
| Feeder chord | \(C\) | Flange circle geometry | Adjusted to fit the flange top |
| Feeder height | \(H_f\) | Required feeding distance and pressure head | Enough to feed the hot spot |
| Feeder taper | \(\alpha_f\) | Pattern and sand mould release | Controlled taper for easy withdrawal |
| Feeder modulus | \(M_f\) | \(M_f \ge 1.2 M_c\) | Feeder remains liquid after casting section |
After the feeder dimensions are established, I calculate the yield. The yield is the ratio of the casting weight to the total metal weight, including feeders and gating system. For high-pressure castings, a lower yield is acceptable if the feeder efficiency is high and the casting is sound. I would rather use a slightly larger feeder with good insulation than a smaller feeder that produces shrinkage porosity. However, I do not increase the feeder blindly. I combine the calculation with chills, proper gating, and top feeding to improve the feeding effect.
$$Y = \frac{W_c}{W_c + W_f + W_g} \times 100\%$$
In this equation, \(Y\) is the process yield, \(W_c\) is the casting weight, \(W_f\) is the total feeder weight, and \(W_g\) is the gating system weight. I monitor this value from batch to batch. If the yield drops without an improvement in quality, I review the feeder size and insulation practice. If the yield is high but hydrostatic testing reveals leakage, I increase the feeding efficiency rather than simply increasing the feeder volume.
Feeding Aids and Directional Solidification
To ensure that the feeder remains liquid and feeds the casting, I use several feeding aids. After the steel is poured to a certain height in the feeder, I stop pouring, plug the sprue with sand, and pour additional hot steel directly into the feeder from the top. This top-pouring practice raises the temperature of the feeder metal and compensates for the shrinkage that occurs during the initial pour. When the feeder is nearly full, I add charcoal powder and straw ash or a similar insulating cover. The cover prevents the feeder surface from solidifying too early and keeps the feeder metal in contact with the atmosphere. After a short interval, I re-pour the feeder once. After re-pouring, I cover it again with ash. This practice keeps the feeder hot and improves its feeding efficiency.
| Feeding Aid | Purpose | Practice |
|---|---|---|
| Top pouring into the feeder | Raise feeder temperature and compensate initial shrinkage | Stop pouring at the selected feeder height, plug the sprue, and pour hot metal into the feeder top |
| Charcoal powder | Provide an insulating and reducing cover | Add when the feeder is nearly full |
| Straw ash or ash cover | Prevent early skin formation | Cover the feeder after topping and re-pouring |
| Delayed re-pouring | Compensate feeder shrinkage | Re-pour once after a short interval, then cover again |
| Open feeder | Allow atmospheric pressure to act on the feeder metal | Keep the feeder open and avoid premature crusting |
| External chills | Promote directional solidification | Place on the outer lower sides of the large flanges |
I also place external chills on the outer lower sides of the two large flanges. These chills increase the cooling rate at the flange surface and help establish a temperature gradient from the feeder toward the remote sections. The chills must be clean, flat, and free of rust. If the chill surface is contaminated, it can produce gas defects or incomplete fusion. The chill thickness is selected according to the section thickness. A common relationship is:
$$t_c = 0.4 t_s$$
Here, \(t_c\) is the chill thickness and \(t_s\) is the local section thickness. I also size the chill area so that it extracts enough heat without chilling the section excessively. If the chill is too small, it has little effect. If it is too large, it may cause a cold shut or a hard zone. Therefore, I use the chill together with the feeder calculation, not as a substitute for feeding.
Gating System Design
The gating system must fill the mould cavity smoothly, avoid direct impingement on cores, and prevent slag and sand from entering the casting. I placed the gating system in the bottom flask flange area. I used an open gating system with a ratio that reduces the velocity in the runner and ingates. The ingates are directed downward and tangentially. This tangential downward direction reduces direct impact on the cores and helps the metal fill the cavity without excessive turbulence. In sand mould casting, turbulence is dangerous because it can erode the mould surface, entrain air, and produce gas defects.
$$A_{\text{sprue}} : A_{\text{runner}} : A_{\text{ingate}} = 1 : 2 : 3$$
The exact area ratio depends on the number of ingates and the casting geometry, but the principle is to keep the sprue area smallest, the runner area larger, and the total ingate area largest. This provides a controlled pressure drop and reduces jetting. I use a sprue brick for the down sprue to protect the mould and maintain a clean channel. The sprue diameter is selected so that the metal fills the cavity in the required time without excessive velocity.
The pouring time is controlled by the casting weight and section thickness. For steel castings, a common form is:
$$t_p = K \sqrt{G}$$
In this equation, \(t_p\) is the pouring time, \(G\) is the total metal weight, and \(K\) is an empirical coefficient that depends on the steel grade, wall thickness, and casting complexity. I keep the pouring time within a narrow range. If pouring is too slow, the metal may freeze before the cavity is filled. If pouring is too fast, the metal may erode the mould and cores. For these high-pressure castings, I use a low-temperature, fast-pouring practice. The exact temperature is recorded on the heat card, and I keep it below the range that would cause excessive moisture absorption and burn-on.
| Gating Parameter | Symbol | Design Choice | Reason |
|---|---|---|---|
| Gating location | — | Bottom flask flange area | Avoid direct impingement on cores |
| Ingate direction | — | Downward and tangential | Reduce turbulence and core erosion |
| Gating type | — | Open gating system | Lower velocity and smoother filling |
| Sprue | — | Sprue brick | Protect mould and maintain clean channel |
| Area ratio | \(A_s:A_r:A_i\) | \(1:2:3\) | Control pressure and velocity distribution |
| Pouring time | \(t_p\) | \(t_p = K \sqrt{G}\) | Fill the cavity before freezing without erosion |
| Pouring temperature | \(T_p\) | Low-temperature fast pouring | Reduce gas absorption and burn-on |
Moulding Sand and Coating
For skin-dried sand mould casting, the moulding sand must be coarser than the sand used for some green sand applications. A coarse sand provides better permeability, which helps vent steam and core gas. At the same time, the sand must have enough green strength to hold the mould shape during handling and enough dry strength after the surface layer is baked. I use a mixture based on silica sand, clay, and an organic binder. The exact formulation depends on the sand source and the casting size, but the general requirements are consistent.
| Component | Function | Typical Addition Range | Control Point |
|---|---|---|---|
| Silica sand | Refractory base | Balance | Coarse grain size for permeability |
| Clay | Green bond | Controlled percentage | Green compressive strength |
| Organic binder | Dry strength and surface toughness | Small addition | Dry surface strength |
| Water | Activate clay and form mould | Moderate | Moisture content |
| Additives | Improve surface finish and reduce defects | Small amount | Coating adhesion and finish |
The coating is as important as the moulding sand. I apply a coating made from refractory powder, fine sand, clay, and a binder. The coating protects the mould surface from the steel, reduces burn-on, and improves the surface finish of the casting. The coating must be applied in two layers. After the first coating, I wait about one hour before applying the second coating. Then I allow the mould to dry naturally for several hours. This natural drying removes part of the water before the controlled bake. If the coating is too thick, it may crack and flake. If it is too thin, it may not protect the surface. Therefore, I control the coating consistency and the number of layers.
$$M_w = \frac{m_{\text{water}}}{m_{\text{sand}}} \times 100\%$$
$$S_g = \frac{F}{A}$$
In these equations, \(M_w\) is the moisture content, \(m_{\text{water}}\) is the mass of water, \(m_{\text{sand}}\) is the mass of dry sand, \(S_g\) is the green strength, \(F\) is the applied force, and \(A\) is the cross-sectional area. I test the sand regularly so that the moulding properties remain within the specified range. A sand mould casting process is only as good as its sand control. If the moisture is too high, the skin-drying step may not remove enough water. If the green strength is too low, the mould may collapse during pouring. If the permeability is too low, gas may be trapped and produce blowholes.
| Property | Test Method | Target | Effect on Sand Mould Casting |
|---|---|---|---|
| Moisture content | Loss on drying | Moderate and stable | Too high increases gas defects; too low reduces green strength |
| Green compressive strength | Compression test | Sufficient for handling | Prevents mould deformation and erosion |
| Permeability | Permeability tester | High enough for venting | Reduces trapped gas and blowholes |
| Dry surface strength | Surface hardness or scratch test | Strong dry layer | Prevents scab, erosion, and burn-on |
| Coating thickness | Visual and mass measurement | Uniform two-layer coating | Protects surface and improves finish |
| Coating adhesion | Visual inspection | No flaking or cracks | Prevents coating defects and inclusions |
Skin Drying and Mould Baking
After the mould is repaired and coated, I dry the skin under controlled conditions. The first stage is natural drying, which removes part of the water from the coating and the surface sand. The second stage is baking with a diesel burner. I bake the mould surface in two steps. The first bake is relatively short. Then I allow an interval and set the cores. After core setting, I perform a second bake. The second bake dries the mould surface again and removes moisture introduced during core setting. After another interval, I check the dried layer thickness. The dried layer must be thick enough to resist metal erosion and moisture release, but not so thick that the mould becomes brittle or wastes energy.
| Step | Action | Interval | Control Target |
|---|---|---|---|
| 1 | Coat the cavity surface | — | Uniform first coating |
| 2 | Apply second coating | About 1 hour | Uniform second coating |
| 3 | Natural drying | Several hours | Partial moisture removal |
| 4 | First diesel burner bake | Defined time | Dry the cavity surface |
| 5 | Core setting | Short interval after first bake | Stable cores and clean seats |
| 6 | Second diesel burner bake | Defined time | Remove moisture from core setting and coating |
| 7 | Check dried layer thickness | Short interval after second bake | Required dry layer depth |
| 8 | Clean cavity and close mould | Immediately after inspection | No loose sand, ash, or debris |
The dried layer thickness is a key variable. If the layer is too thin, the steel may penetrate the coating and react with the moist sand beneath. If the layer is too thick, the surface may crack and spall. I check the thickness by probing or by visual inspection of the baked zone. In general, the target is a dry layer that covers the cavity surface and extends a few millimeters into the sand. The exact depth depends on the section thickness and pouring temperature. I use a simple relation to describe the growth of the dried layer:
$$d_d = k_d \sqrt{t_b}$$
Here, \(d_d\) is the dried layer depth, \(k_d\) is a coefficient that depends on the sand and burner, and \(t_b\) is the baking time. This relation helps me adjust the baking time when the sand moisture or burner output changes. I do not treat baking as a fixed time only. I treat it as a heat-transfer process that must produce a measurable dry layer.
Mould Closing and Pouring Practice
After the second bake and inspection, I clean the cavity with compressed air and remove any ash, loose sand, or coating debris. Then I close the mould and prepare for pouring. The mould must be poured within a limited time after closing. If it remains open or closed for too long, the skin-dried surface can absorb moisture from the air and lose its low-moisture advantage. If a long delay is unavoidable, I open the mould and bake it again before closing. This rule is essential for skin-dried sand mould casting because the dry surface layer is relatively thin and can re-humidify quickly.
$$M(t) = M_\infty – \left(M_\infty – M_0\right)e^{-kt}$$
In this equation, \(M(t)\) is the moisture content at time \(t\), \(M_0\) is the initial moisture content, \(M_\infty\) is the equilibrium moisture content, and \(k\) is a rate constant. Although the equation is simple, it captures the practical behavior: moisture regain is fast at first and then slows. I use this concept to set a maximum holding time between closing and pouring. If the holding time exceeds the limit, I re-bake the mould.
The pouring temperature must not be too high. High-temperature steel can absorb moisture from the mould surface, produce gas holes, and cause burn-on. Burn-on makes cleaning difficult and may require excessive grinding, which can expose subsurface defects. I therefore use a low-temperature, fast-pouring practice. The pouring temperature is controlled within a narrow range, and the pouring time is kept short enough to fill the cavity before the steel freezes. For the steel used in these high-pressure castings, I keep the pouring temperature in the low range appropriate for the section thickness and keep the pouring time within the calculated window.
| Pouring Variable | Symbol | Practice | Reason |
|---|---|---|---|
| Pouring temperature | \(T_p\) | Low but safe for filling | Reduce gas absorption, burn-on, and sand erosion |
| Pouring time | \(t_p\) | Fast within a controlled range | Fill the cavity before premature freezing |
| Mould holding time | \(t_h\) | As short as possible after closing | Prevent moisture regain in the skin-dried layer |
| Sprue plugging | — | Plug after initial fill | Allow top feeding into the feeder |
| Feeder covering | — | Charcoal powder and ash | Keep feeder hot and open to atmosphere |
| Re-pouring | — | One controlled re-pour | Compensate feeder shrinkage |
Defect Prevention in Skin-Dried Sand Mould Casting
I have found that most defects in skin-dried sand mould casting can be prevented if the process variables are controlled together. A dry surface alone will not prevent shrinkage porosity. A large feeder alone will not prevent gas holes. A good coating alone will not prevent core shift. The process must be viewed as a system. The table below summarizes the main defects I watch for and the countermeasures I apply.
| Defect | Main Cause | Countermeasure in Skin-Dried Sand Mould Casting |
|---|---|---|
| Gas holes | Moisture, coating gas, core gas, turbulent filling | Skin-dry the surface, vent cores, use open gating, control pouring temperature |
| Burn-on | High pouring temperature, reactive sand surface, insufficient coating | Apply two-layer coating, bake the surface, lower pouring temperature |
| Scab | Surface layer expansion and weak dry strength | Increase surface ramming, control moisture, bake to proper depth |
| Sand erosion | High metal velocity, direct ingate impingement | Use tangential downward ingates, open gating, smooth runner |
| Shrinkage porosity | Insufficient feeding, poor directional solidification | Calculate feeder modulus, use chills, top-pour and insulate feeders |
| Core shift | Weak core prints, unstable core assembly | Use symmetrical parting, strong core prints, proper chaplets if needed |
| Leakage at sealing groove | Subsurface porosity or inclusion near the machined surface | Ensure dense feeding at flange, clean mould, proper gating, adequate machining allowance |
| Pressure drop during hydrostatic test | Through-wall or interconnected porosity | Verify feeder design, chill placement, pouring practice, and heat treatment |
The sealing ring groove is especially demanding. A defect that is harmless in a general casting can cause leakage in a high-pressure casting. I therefore give extra attention to the flange hot spots. The feeder must be close enough to feed the flange, and the chill must promote solidification from the outside toward the feeder. The gating system must not introduce slag or sand into the flange area. The coating must be clean and well bonded. The mould must be free of loose sand before closing. These details are not optional. They are part of the sand mould casting process for pressure-tight components.
Hydrostatic Testing and Acceptance
After heat treatment and machining, the high-pressure castings are subjected to hydrostatic testing. The test pressure is selected according to the working pressure and the applicable specification. A common relationship is:
$$P_t = \eta P_w$$
Here, \(P_t\) is the test pressure, \(P_w\) is the working pressure, and \(\eta\) is a safety factor specified by the product standard. During the test, the casting must show no leakage, no sweating, and no pressure drop. A pressure drop indicates that a leak path exists or that the test setup has a leak. If the setup is sealed correctly and pressure still drops, the casting is rejected or repaired according to the approved procedure.
$$\Delta P = P_0 – P_f \le 0$$
In this expression, \(P_0\) is the initial test pressure and \(P_f\) is the final pressure after the specified duration. If \(\Delta P\) is greater than zero, the casting fails the pressure-tightness requirement. I also inspect the sealing surfaces after testing. A sweat mark may indicate a very fine porosity network that does not produce a visible leak but still violates the acceptance criteria. For high-pressure service, such a defect is not acceptable.
| Test Item | Symbol | Requirement | Result |
|---|---|---|---|
| Test pressure | \(P_t\) | \(P_t = \eta P_w\) | Set according to specification |
| Hold time | \(t_t\) | Specified duration | No leakage, sweating, or pressure drop |
| Pressure drop | \(\Delta P\) | \(\Delta P \le 0\) | Pass or fail criterion |
| Sealing ring groove | — | No sand holes, gas holes, or porosity | Visual and dimensional inspection |
| Surface condition | — | No burn-on, scab, or erosion defects | Visual inspection after cleaning |
| Hardness | \(H\) | Specified after quench and temper | Verified on test coupon or casting |
I have produced several thousand four-way blocks using this skin-dried sand mould casting process. They have passed hydrostatic testing and met the drawing requirements. This result confirms that the process is successful for this family of high-pressure castings. It also shows that skin-dried sand mould casting can produce pressure-tight steel castings when the process is designed with proper feeding, gating, chilling, drying, and pouring control.
Process Control Summary
For me, the most important lesson is that skin-dried sand mould casting for high-pressure castings is a calculated process. I do not simply increase the feeder size and hope for the best. I calculate the feeder dimensions, design the gating system, place chills, control the mould surface, and monitor the pouring temperature and time. The following table summarizes the key control variables and their purpose.
| Control Variable | Symbol | Purpose | Typical Action |
|---|---|---|---|
| External shrinkage | \(s_e\) | Compensate external contraction | Adjust pattern dimensions |
| Internal shrinkage | \(s_i\) | Compensate internal contraction | Adjust core dimensions |
| Feeder modulus | \(M_f\) | Ensure feeder feeds the hot spot | Keep \(M_f \ge 1.2 M_c\) |
| Feeder width | \(W_f\) | Match hot-spot size | Calculate from hot-spot diameter |
| Feeder taper | \(\alpha_f\) | Allow pattern withdrawal and proper feeding | Use waist-conical open feeder |
| Chill thickness | \(t_c\) | Promote directional solidification | Use \(t_c = 0.4 t_s\) as a guide |
| Gating ratio | \(A_s:A_r:A_i\) | Control filling velocity | Use open gating with \(1:2:3\) |
| Pouring time | \(t_p\) | Fill before freezing | Use \(t_p = K \sqrt{G}\) |
| Dry layer depth | \(d_d\) | Resist erosion and moisture release | Bake to a measured depth |
| Mould holding time | \(t_h\) | Prevent moisture regain | Pour within the allowed time or re-bake |
| Pouring temperature | \(T_p\) | Reduce gas and burn-on | Use low-temperature fast pouring |
I also keep batch records. For each heat, I record the steel grade, pouring temperature, pouring time, sand moisture, green strength, coating batch, baking time, dried layer depth, feeder dimensions, chill placement, and hydrostatic test result. These records allow me to trace any defect back to the process variable that may have caused it. In sand mould casting, traceability is not a paperwork exercise. It is a tool for continuous improvement. When a casting passes the hydrostatic test, I want to know why it passed. When a casting fails, I want to know which variable moved out of range.
Comparison with Other Moulding Routes for High-Pressure Castings
I have evaluated skin-dried sand mould casting against other routes for high-pressure steel castings. The choice depends on batch size, geometry, pressure rating, and available equipment. The table below presents my general comparison.
| Moulding Route | Surface Quality | Dimensional Stability | Energy Use | Production Flexibility | Suitability for High-Pressure Castings |
|---|---|---|---|---|---|
| Green sand mould casting | Moderate | Moderate | Low | High | Limited for severe pressure and thick sections |
| Skin-dried sand mould casting | High at the cavity surface | Good | Moderate | High | Very suitable for small and medium batches |
| Fully dry sand mould casting | High | Good | High | Moderate | Suitable but more expensive |
| Resin-bonded sand mould casting | High | High | Moderate to high | Moderate | Suitable with proper feeding and coating |
| Investment casting | Very high | Very high | High | Lower for large parts | Suitable for smaller and more complex parts |
For the four-way block and similar wellhead components, I prefer skin-dried sand mould casting because it provides a strong surface without the cost of fully drying the entire mould. It also allows me to use strong core assemblies and to place chills and feeders exactly where they are needed. The process is flexible enough for small and medium batch production, and it can be repeated consistently when the sand, coating, baking, and pouring variables are controlled.
Practical Rules I Follow
Over many production runs, I have developed a set of practical rules for skin-dried sand mould casting of high-pressure castings. These rules are simple, but they are based on the physics of solidification and the behavior of the mould.
First, I never treat the skin-dried layer as a cosmetic feature. It is a functional barrier that controls heat transfer, moisture release, and surface erosion. I specify the dry layer depth and verify it. Second, I calculate the feeder before I cut the pattern. I do not wait until the first casting fails. Third, I keep the gating system open and non-turbulent. A smooth fill reduces gas entrapment and sand erosion. Fourth, I place chills where they create a favorable gradient, not where they are convenient. Fifth, I control the pouring temperature. A high temperature may improve fluidity, but it also increases gas defects and burn-on. Sixth, I keep the time between closing and pouring short. A skin-dried mould can regain moisture if it is left too long. Seventh, I record every heat and every test result. The records are the basis for improvement.
I also pay attention to the human factors. The moulding crew must understand that the skin-dried surface is fragile before baking and must be handled carefully. The coating must be applied uniformly. The burner must be moved at a controlled speed. The core setter must not damage the dried layer. The pourer must follow the temperature and time instructions. In sand mould casting, small variations in handling can produce large variations in quality. Therefore, training and communication are part of the process.
Conclusion
Skin-dried sand mould casting has allowed me to produce high-pressure steel castings that meet strict density and pressure-tightness requirements. Compared with green sand mould casting, it provides a stronger surface, lower surface moisture, and a lower tendency for gas holes, burn-on, scab, and erosion. Compared with fully dry sand mould casting, it saves fuel and electric power, shortens the production cycle, and improves the working environment. For high-pressure castings in small and medium batches, it is a practical and reliable route.
The key to success is not the name of the process. The key is the integration of design and control. I calculate the shrinkage allowances, design a symmetrical parting line, support and vent the cores, size the waist-conical open feeders, place chills on the large flanges, use an open gating system with tangential downward ingates, prepare a coarse permeable sand, apply a two-layer coating, bake the mould surface in two stages, control the dried layer depth, pour at a low temperature with a fast pouring time, and test the final casting hydrostatically. When all of these elements are controlled together, skin-dried sand mould casting can produce sound, pressure-tight castings.
I have applied this approach to several thousand four-way blocks and related wellhead components. The castings have passed hydrostatic testing and met the drawing requirements. This experience confirms that skin-dried sand mould casting is a valid and dependable method for high-pressure castings. It also provides a foundation for future work on similar pressure-containing steel components. For single-piece and small-batch production of pressure-resistant parts, it remains one of the most effective routes I can recommend.