In my work with full mold casting, I have repeatedly seen that shrinkage porosity and shrinkage cavities are among the most persistent internal defects. They are not merely cosmetic problems. In full mold casting, a shrinkage defect can reduce pressure tightness, fatigue resistance, wear resistance, machinability, and final service life. When I examine a failed full mold casting, I therefore begin with the solidification path: how the liquid metal entered the cavity, where it stopped feeding, which region froze last, and whether the gating system had enough stored liquid to compensate for liquid and solidification contraction. My purpose here is to present the defect features, locations, mechanisms, causes, and practical prevention methods that I use in full mold casting production, with tables and formulas that summarize the key relationships.
My Classification of Shrinkage Defects
In full mold casting, I classify shrinkage defects by morphology, position, and feeding condition. A concentrated cavity with rough internal surfaces, irregular shape, and visible dendritic branches is what I call a concentrated shrinkage cavity. If it breaks the outer surface, I treat it as an external shrinkage cavity. If it remains inside the casting, I treat it as an internal shrinkage cavity. If it forms in a re-entrant corner or sharp concave region, I treat it as a corner shrinkage cavity. Although these defects appear in different places, I have found that their root cause is usually the same: a region of liquid metal solidifies last and cannot receive sufficient feed metal from a hotter, still-liquid reservoir.
| Defect type | Typical appearance | Usual position in full mold casting | Feeding condition |
|---|---|---|---|
| External shrinkage cavity | Open, rough, irregular depression | Top surface, riser contact area, hot spot exposed to air | Last liquid at surface, no feed after skin formation |
| Internal shrinkage cavity | Enclosed irregular void with dendrites | Thermal center, heavy section, isolated thick region | Liquid enclosed by solidified skin |
| Corner shrinkage cavity | Cavity at concave corner or junction | Re-entrant corner, rib junction, local heat pocket | Poor local heat extraction and late freezing |
| Shrinkage porosity | Dispersed micro-voids among dendrites | Wide freezing range alloys, dispersed hot spots | Insufficient local feeding and gas precipitation |
I also distinguish between macro-shrinkage and micro-shrinkage. Macro-shrinkage is visible to the eye or by low-magnification inspection after sectioning. Micro-shrinkage is distributed between dendrite arms and often requires dye penetrant, radiographic, or microscopic examination. In full mold casting, both forms may occur in the same casting when the thermal gradient is weak and the feeding path is interrupted.
Locations I Check First
When I investigate a full mold casting for shrinkage, I do not inspect the whole part randomly. I first identify the last-freezing regions. These are normally thermal centers, heavy sections, isolated thick bosses, junction zones, and areas remote from the gating system. I also check how the inner gate position, local heat extraction, gating system heat content, and module arrangement affect cooling. In full mold casting, module arrangement can be as important as the individual casting geometry because neighboring parts can preheat the shell and slow local cooling.
I use three practical methods to estimate shrinkage locations. The first is the isothermal line method, in which I sketch solidification isotherms and look for the last region to reach the solidus. The second is the inscribed circle method, in which I draw the largest possible circles in section and compare their diameters. The third is sectioning or simulation. With modern foundry software, I can observe filling and solidification and predict shrinkage cavities and porosity before making tooling. In full mold casting, this predictive step saves far more money than it costs.
| Location indicator | What I measure | Why it matters in full mold casting |
|---|---|---|
| Thermal center | Last region to solidify | Requires direct feeding from gate or riser |
| Heavy section | Large modulus relative to neighbors | Stores heat and freezes late |
| Isolated thick boss | No clear feed path | Forms internal cavity |
| Concave corner | Poor local heat loss | Creates corner shrinkage |
| Gate contact zone | Hot metal concentration | Can become a shrinkage site if gate freezes early |
| Module edge or center | Radiation and conduction balance | Changes local cooling rate |
Contraction Mechanism I Use
In full mold casting, the casting contracts in three stages from pouring to room temperature: liquid contraction, solidification contraction, and solid contraction. Liquid and solidification contraction determine the volume deficit that must be fed. Solid contraction affects final dimensions and stress but does not directly create the internal void if feeding is already complete. I summarize the liquid contraction as follows:
$$ \varepsilon_{V,l} = \alpha_{V,l}(t_p – t_l) \times 100\% $$
Here, \(\varepsilon_{V,l}\) is the liquid volumetric contraction, \(\alpha_{V,l}\) is the average volumetric contraction coefficient of the liquid from pouring temperature \(t_p\) to liquidus temperature \(t_l\), and \(t_p-t_l\) is the superheat. For carbon steel, I commonly use an average value near \(1.6\times10^{-4}\ \mathrm{^\circ C^{-1}}\). This value depends on temperature and chemical composition, so I do not treat it as a universal constant in full mold casting.
The solidification contraction is the volume change between liquidus and solidus. It depends strongly on carbon content and alloy system. For a medium-carbon steel, I often take a solidification contraction near \(3\%\), but I always verify it against the actual grade. The solid contraction from solidus to room temperature can be written as:
$$ \varepsilon_{V,s} = \alpha_{V,s}(t_s – t_r) \times 100\% $$
$$ \varepsilon_l = \alpha_l(t_s – t_r) \times 100\% $$
Here, \(\varepsilon_{V,s}\) is solid volumetric contraction, \(\varepsilon_l\) is linear contraction, \(\alpha_{V,s}\) is the volumetric contraction coefficient in the solid state, \(\alpha_l\) is the linear contraction coefficient, \(t_s\) is the solidus temperature, and \(t_r\) is room temperature. For steel, I usually relate solid linear contraction to volumetric contraction by a factor of about one third, with a common value near \(2.4\%\). The total contraction can be summarized as:
$$ \varepsilon_{V,total} = \varepsilon_{V,l} + \varepsilon_{V,c} + \varepsilon_{V,s} $$
For a carbon steel with \(100\ \mathrm{^\circ C}\) superheat, a typical estimate can be:
$$ \varepsilon_{V,total} = 1.6\% + 3.0\% + 7.2\% = 11.8\% $$
In real full mold casting, this total is reduced by gas precipitation, shell restraint, core restraint, and casting geometry. If gas precipitates during solidification and cannot escape, it may offset some volume loss, but it also increases dispersed porosity and reduces density. Therefore, I never treat gas porosity as a substitute for proper feeding. In full mold casting, the target is controlled feeding, not accidental gas compensation.
| Carbon content (%) | Solidification contraction (%) |
|---|---|
| 0.10 | 2.0 |
| 0.35 | 3.0 |
| 0.45 | 4.3 |
| 0.70 | 5.3 |
| Carbon content (%) | Pre-pearlite contraction (%) | Eutectoid expansion (%) | Post-pearlite contraction (%) | Total solid linear contraction (%) |
|---|---|---|---|---|
| 0.08 | 1.42 | 0.11 | 1.16 | 2.47 |
| 0.14 | 1.51 | 0.11 | 1.06 | 2.46 |
| 0.35 | 1.47 | 0.11 | 1.04 | 2.40 |
| 0.45 | 1.39 | 0.11 | 1.07 | 2.35 |
| 0.55 | 1.35 | 0.09 | 1.05 | 2.31 |
| 0.90 | 1.21 | 0.01 | 0.98 | 2.18 |
I also consider the effect of alloying elements on solid contraction. In full mold casting, elements such as carbon, manganese, silicon, chromium, and molybdenum change both the freezing range and the solid-state contraction. I therefore avoid using a single shrinkage allowance for all grades. The correct approach is to combine alloy-specific contraction data, casting modulus, and the actual shell preheat schedule.
How a Concentrated Shrinkage Cavity Forms
In a cylindrical section, the heat is mainly removed through the shell. A temperature gradient forms from the surface to the center. The outermost liquid reaches the liquidus first and forms a thin solid skin. As the skin grows inward, the remaining liquid cools and contracts. Because the skin is already solid, the liquid level falls. The volume loss from liquid contraction and solidification contraction is larger than the solid contraction of the skin, so the liquid separates from the top surface and a cavity grows. At the end of solidification, an inverted cone-like cavity remains in the upper part. Later solid cooling slightly reduces both the outer dimensions and the cavity volume.
I summarize the competition as:
$$ V_{\text{cavity}} \approx V_{\text{liquid loss}} + V_{\text{solidification loss}} – V_{\text{feed}} – V_{\text{gas}} – V_{\text{expansion}} $$
Where \(V_{\text{cavity}}\) is the final cavity volume, \(V_{\text{liquid loss}}\) is the liquid contraction volume, \(V_{\text{solidification loss}}\) is the solidification contraction volume, \(V_{\text{feed}}\) is the volume supplied by the gating or riser, \(V_{\text{gas}}\) is the volume occupied by precipitated gas, and \(V_{\text{expansion}}\) is any expansion, such as graphite expansion in cast iron or eutectoid expansion in steel. In full mold casting, my goal is to maximize \(V_{\text{feed}}\) and place the final cavity in the gating system rather than in the casting.
| Solidification stage | What happens | Effect on shrinkage | My control action in full mold casting |
|---|---|---|---|
| Liquid cooling | Liquid contracts before liquidus | Lowers liquid level and creates volume deficit | Control superheat and pouring speed |
| Solidification | Liquid to solid volume change | Main source of concentrated cavity | Provide directional freezing toward gate or riser |
| Solid cooling | Solid casting contracts | Changes dimensions and stress | Control shell temperature and cooling uniformity |
| Gas precipitation | Gas forms during freezing | Can reduce macro-cavity but increase porosity | Deoxidize and control melt cleanliness |
Root Causes I Examine in Full Mold Casting
I group the causes of shrinkage defects into seven practical categories: gating design, casting structure, alloy contraction, pouring conditions, shell cooling capacity, local heat extraction, and pressure head. When a full mold casting shows shrinkage, I work through these categories in order rather than guessing.
Gating System Design
An improperly designed gating system prevents directional solidification and adequate feeding. Directional solidification means that the metal freezes progressively toward the gating or riser, so that the last liquid remains in the gating system and feeds the adjacent late-freezing region of the casting. If the gating system freezes too early, or if the gate is too small, the casting cannot be fed. If the gate is too large, the gate itself may remain liquid too long and create a shrinkage cavity at the gate-casting junction.
| Gating defect | Result in full mold casting | Correction I apply |
|---|---|---|
| Gate too small | Early gate freeze, no feeding | Increase gate area or use insulated gate |
| Gate too large | Hot spot at junction, shrinkage at contact | Reduce gate modulus or relocate gate |
| Gate placed away from hot spot | Thermal center remains isolated | Move gate to thermal center |
| No riser or insufficient riser | Cavity remains in casting | Add or enlarge riser |
| Riser too small | Cavity moves only partly into riser | Increase riser modulus |
| Riser without taper | Shrinkage tail extends into casting | Add taper and proper riser neck |
In full mold casting, the gating system normally includes a pouring cup, sprue, runner, and inner gates. When needed, I add slag traps, buffers, and vents. The pouring cup receives metal and establishes pressure for filling and feeding. I usually keep the top of the pouring cup about \(70\) to \(100\ \mathrm{mm}\) above the wax pattern level. The sprue is the structural support during shell making and often acts as a riser. To ensure feeding, I design the sprue cross-section as about \(1.4\) times the total inner gate cross-section:
$$ A_s = 1.4 A_i $$
For the runner, the function determines the area. If the runner is intended to feed, I use:
$$ A_r = (1.1 \sim 1.3)A_s $$
If the runner is not intended to feed, I use:
$$ A_r = (0.7 \sim 1.0)A_s $$
The inner gate is the most critical element in the gating system of full mold casting. It controls filling, solidification, feeding, stress, shrinkage, hot tearing, distortion, cleaning, machining, and surface quality. I therefore define its position, number, shape, length, and cross-section with care.
| Gating element | Primary function | Design rule I use | Full mold casting consideration |
|---|---|---|---|
| Pouring cup | Receive metal and establish pressure | Top surface \(70\) to \(100\ \mathrm{mm}\) above pattern | Prevents turbulence and improves fill |
| Sprue | Support shell and act as riser | \(A_s = 1.4 A_i\) | Must remain liquid long enough to feed |
| Runner | Distribute metal, trap slag, feed | \(A_r = (1.1 \sim 1.3)A_s\) if feeding; \(A_r = (0.7 \sim 1.0)A_s\) if not | Controls module-to-module distribution |
| Inner gate | Connect gating to casting | Position at hot spot; size by hot spot or modulus | Most critical for shrinkage control |
| Vent | Release gas | Place at last-filled high points | Reduces back pressure and gas defects |
Inner Gate Position, Number, Shape, Length, and Size
I position the inner gate from two viewpoints. For filling, I avoid direct impingement on cores and thin shell sections, prevent turbulence and eddies, and avoid local overheating that softens or deforms the shell. For feeding, I place the gate at the thermal center so that directional solidification proceeds toward it. In full mold casting, the gate position often decides whether the shrinkage cavity ends up in the casting or in the gating system.
The number of inner gates depends on pattern making, shell making, dewaxing, pouring, cleaning, feeding, distortion, and hot tearing. I prefer one gate when possible. When the casting has several isolated hot spots, high density requirements, or complex geometry, I use two or more gates. The shape can be rectangular, circular, square, or fan-shaped, depending on the contact area and feeding direction. For easy-break gates, I use a length of about \(8\) to \(12\ \mathrm{mm}\). For gas-cut gates, I use about \(10\) to \(15\ \mathrm{mm}\).
For cross-section, I use two methods. The first is the hot-spot circle ratio method:
$$ D_i = K D_h,\quad K=0.7 \sim 0.9 $$
Where \(D_i\) is the inner gate diameter or equivalent diameter and \(D_h\) is the hot-spot circle diameter. The second is the equivalent modulus method. I convert the casting hot spot into one or more simple units, such as a cylinder, so that the unit has the same weight and solidification modulus as the hot spot. Then I determine the inner gate size by ratio. If the inner gate is too large, the gate neck remains liquid until the end and no other liquid is available to feed it, so a shrinkage cavity forms at the gate-casting junction. If the inner gate is too small, it freezes early and the casting cannot be fed.
| Inner gate variable | My practical rule | Risk if violated |
|---|---|---|
| Position | At thermal center; avoid core impingement | Isolated hot spot, turbulence, shell damage |
| Number | One preferred; more for dispersed hot spots | Insufficient feeding or excessive distortion |
| Shape | Match contact geometry and feed direction | Hot spot or early freeze |
| Length | \(8\) to \(12\ \mathrm{mm}\) easy-break; \(10\) to \(15\ \mathrm{mm}\) gas-cut | Cleaning difficulty or poor feeding |
| Cross-section | \(D_i=KD_h\), \(K=0.7\) to \(0.9\) | Shrinkage at junction or in casting |
| Modulus | Gate modulus lower than feeding reservoir | Gate freezes before casting |
Casting Structure Design
I have found that many shrinkage defects in full mold casting are designed into the part before the foundry ever sees it. A casting with abrupt section changes, isolated thick sections, inadequate or excessive fillets, or a geometry copied directly from a forging or weldment is difficult to feed. When section size changes too abruptly, the thin section freezes faster than the thick section and interrupts the feed path. When an isolated thick section has no connection to a riser, it cannot be fed. When the fillet is too small, the transition is too sharp. When the fillet is too large, it creates a new hot spot. When a part is converted from forging or welding to casting without redesign, the solidification path may be unfavorable.
| Structural feature | Problem for full mold casting | My design change |
|---|---|---|
| Abrupt section change | Thin section freezes first and blocks feed | Gradual transition, generous fillet |
| Isolated thick section | No feed path | Add rib, core, or redesign as uniform section |
| Small fillet | Sharp transition and stress concentration | Increase fillet radius |
| Excessive fillet | New thermal center | Reduce fillet or add feeding |
| Forging-type geometry | Poor castability | Redesign for directional solidification |
| Welded assembly geometry | Uneven wall thickness | Integrate ribs and bosses for feeding |
When I review a casting for full mold casting, I first identify how many hot spots exist and how they can be divided into feeding zones. If feeding is difficult, I adjust the machining allowance, thicken or thin local areas, and make wall thickness as uniform as practical. If these changes are not enough, I use padding. Padding is an added thickness on the side near the riser to establish a positive temperature gradient toward the riser. It increases the feeding channel and improves the feeding ability of the sprue or riser. For top pouring, when wall thickness is fixed, the pad width increases with height. When casting height is fixed, a thinner wall requires a wider feeding pad. However, padding increases machining allowance, and excessive padding reduces the economic advantage of full mold casting.
Alloy Contraction and Gas Effects
The larger the liquid contraction and solidification contraction, the larger the potential shrinkage cavity. For alloys that do not expand during solidification, and when no feeding occurs, the liquid contraction is proportional to the contraction coefficient and the superheat:
$$ \varepsilon_{V,l} \propto \alpha_{V,l}(t_p-t_l) $$
However, I do not simply lower pouring temperature to reduce liquid contraction. If the gating system is responsible for feeding, lowering the pouring temperature can reduce the feeding ability of the gating system, or even freeze the gating system early. The result can be a larger cavity. In some cases, a moderate increase in superheat improves feeding enough to reduce the final cavity, even though liquid contraction increases. In full mold casting, I optimize superheat together with gate modulus and shell temperature.
Solidification contraction also increases cavity volume. If the melt is not fully deoxidized, gas precipitates during freezing. If the gas cannot float or enter an existing cavity, it remains as dispersed micro-porosity between dendrite arms. This may reduce measurable macro-cavity volume, but it expands the shrinkage porosity zone. I therefore deoxidize thoroughly and control melt cleanliness. In full mold casting, gas and shrinkage porosity often interact, so I treat them together.
| Factor | Effect on macro-cavity | Effect on micro-porosity | My action |
|---|---|---|---|
| Higher liquid contraction | Increases cavity volume | May increase feeding demand | Optimize superheat and feeding |
| Higher solidification contraction | Increases cavity volume | Raises local demand | Select alloy and control cooling |
| Incomplete deoxidation | May reduce macro-cavity | Increases dispersed porosity | Full deoxidation and clean melt |
| Gas precipitation | Can offset volume loss | Expands porosity zone | Vent and control gas content |
| Eutectoid expansion | Reduces final cavity | May increase stress | Use controlled cooling |
Pouring Conditions
Pouring speed has a strong effect on shrinkage in full mold casting. If the casting does not freeze locally during pouring, a faster pour gives less time for heat removal, so the metal in the cavity remains hotter at the end of pouring. This increases liquid contraction and enlarges the final cavity. A faster pour also means that the first metal to enter has less time to receive feed from later metal. Therefore, the final cavity can be larger. I often use the principle of high-temperature tapping and low-temperature pouring. For large modules, I use a fill strategy of fast first, slow later, and then a final top-up. This helps directional solidification and allows the gating system to feed the casting.
For top pouring, I tend to use higher pouring temperature and slower pouring speed. For bottom pouring, I tend to use lower pouring temperature and faster pouring speed. The exact values depend on alloy, section thickness, shell temperature, and module weight. In full mold casting, I record pouring time, temperature, and shell temperature for every critical part because these variables determine whether the feeding path remains open.
| Pouring variable | Effect | My preferred practice |
|---|---|---|
| High pouring temperature | Increases liquid contraction but improves feeding | Use only when gating can feed |
| Low pouring temperature | Reduces liquid contraction but may freeze gate | Avoid if gate freezing risk is high |
| Fast pouring | Raises final metal temperature and shrinks feed time | Use controlled fast fill then slow top-up |
| Slow pouring | Lower final temperature but may cause cold shut | Use for thin sections or top pouring |
| Top pouring | Higher temperature gradient | Higher temperature, slower speed |
| Bottom pouring | Lower turbulence | Lower temperature, faster speed |
Shell Cooling Capacity and Local Heat Extraction
The poorer the shell cooling capacity, the slower the metal cools and solidifies. A slow-cooling shell means that earlier liquid contraction is less compensated by later liquid, so the final cavity can be larger. If I can make the solidification time equal to the pouring time, the casting can freeze without a shrinkage cavity, which is the principle behind continuous casting. In full mold casting, I cannot make the whole shell act like a continuous caster, but I can control shell preheat, shell insulation, and local cooling.
Local heat extraction is also critical. A concave corner loses heat more slowly than a flat or convex surface. The metal there freezes later. If it cannot be fed, a corner shrinkage cavity forms. I improve local cooling by adjusting module arrangement, keeping wax patterns at proper spacing, and placing hot spots near the module edge when possible. In some cases, I use chill blocks in the shell-making process. I prefer brass or stainless steel chills because carbon steel can oxidize during shell firing and may not bond well. Local chills improve the local temperature gradient and feeding ability, reducing or eliminating corner shrinkage in full mold casting.
| Cooling condition | Solidification effect | Shrinkage risk | My corrective action |
|---|---|---|---|
| Poor shell cooling | Slow freezing, long liquid period | Large cavity | Adjust shell preheat and insulation |
| Concave corner | Local late freezing | Corner cavity | Use local chill or relocate gate |
| Crowded modules | Mutual preheating | Dispersed porosity | Increase spacing and improve arrangement |
| Hot spot at center | Last freezing away from edge | Internal cavity | Move hot spot to edge or add feed |
| Uneven shell thickness | Uneven heat extraction | Local shrinkage | Control dipping and shell build |
Pressure Head and Feeding Distance
Feeding pressure head affects the flow rate of feed metal. A higher pressure head improves feeding. However, for ordinary castings, the effective feeding distance decreases as wall thickness increases. Increasing riser diameter and height alone may not achieve effective feeding. If the riser is too high, it may only feed the lower half of the casting, leaving a secondary shrinkage cavity in the upper part. I therefore combine insulated risers with covering flux or exothermic agents. The covering agent reduces and delays solidification of the top liquid surface in the riser and prevents a secondary cavity from forming under a top crust. When necessary, I switch to centrifugal or pressure-assisted pouring.
| Feeding condition | Effect on full mold casting | My solution |
|---|---|---|
| Insufficient pressure head | Slow feed flow, poor feeding | Increase sprue height or use pressure pouring |
| Excessive riser height | Feeds only lower part | Use insulated riser and exothermic cover |
| Long thin section | Feed distance limited | Add multiple gates or increase gradient |
| Thick section with small riser | Riser freezes before casting | Increase riser modulus |
| Secondary cavity in riser | Top crust blocks feeding | Use cover flux and hot topping |
My Prevention Rules for Full Mold Casting
The basic principle I follow is simple: make the casting solidify directionally and leave the shrinkage cavity in the gating system. I implement this principle through gating design, structure improvement, contraction reduction, pouring control, local cooling, and pressure head. The following table summarizes the prevention actions and their expected effects.
| Prevention area | Action | Expected effect | Full mold casting note |
|---|---|---|---|
| Gating | Design sprue, runner, and gate with proper modulus | Feeds hot spot and moves cavity to gating | Most direct control method |
| Structure | Uniform wall, proper fillets, padding | Improves feed path | Requires customer approval |
| Alloy | Select lower contraction alloy when possible | Reduces total volume deficit | Must meet service requirements |
| Pouring | Optimize temperature and speed | Balances contraction and feeding | Use high-temperature tapping, low-temperature pouring |
| Melt | Full deoxidation | Reduces gas-related porosity | Improves density |
| Cooling | Local chills and module spacing | Sharpens temperature gradient | Brass or stainless steel chills preferred |
| Pressure | Increase pressure head or use insulated riser | Improves feed flow | Combine with exothermic cover |
Balancing Directional Solidification and Residual Stress
I have learned that directional solidification solves shrinkage but can create a new problem: large temperature differences between regions. These differences can produce residual stress, distortion, and even cracking. In full mold casting, I therefore do not pursue the sharpest possible gradient everywhere. I pursue the minimum gradient that still gives feeding to the last-freezing region. After the shrinkage path is secured, I adjust shell temperature, cooling rate, and post-pour handling to reduce stress. The best solution is a balanced process: feeding is adequate, but the casting does not develop harmful residual stress.
I often use the following qualitative relation during process review:
$$ \text{Process quality} = f(\text{feeding sufficiency},\ \text{thermal gradient balance},\ \text{stress control}) $$
A process with excellent feeding but uncontrolled stress is not a robust full mold casting process. A process with low stress but insufficient feeding will also fail. My target is the intersection of these requirements.
Application Case: Rare-Earth Modification in Full Mold Casting
I once worked on a critical full mold casting used in mining equipment. The part weighed about \(16\ \mathrm{kg}\) and was expected to achieve a service life above ten thousand meters of advance. In practice, it failed by wear after only three to four thousand meters. Inspection showed severe shrinkage cavities and axial shrinkage porosity on the wear surface, reducing wear resistance and causing early failure. I addressed the problem by modifying the steel with rare-earth elements. Rare earth additions improve fluidity, increase feeding ability, reduce shrinkage cavities and porosity, and improve mechanical properties. They also purify, modify, and alloy the steel.
I selected a rare-earth-modified grade with the following chemical composition.
| Material | C (%) | Si (%) | Mn (%) | Cr (%) | Mo (%) | P (%) | S (%) | Re addition (%) |
|---|---|---|---|---|---|---|---|---|
| ZG20CrMnMoRe | 0.15-0.25 | 0.40-0.75 | 0.90-1.20 | 1.10-1.40 | 0.20-0.30 | \(\le 0.04\) | \(\le 0.04\) | 0.15 |
Rare-Earth Addition Method and Amount
I tested three addition methods in a \(500\ \mathrm{kg}\) coreless induction furnace with an acid lining. The results showed that the ladle injection method gave the highest recovery and the most stable rare-earth residual. The mechanical properties were also more stable. I summarize the methods in the table below.
| Addition method | Rare-earth recovery (%) | Rare-earth residual (%) |
|---|---|---|
| Furnace insertion | Decreases with pouring time | Decreases with pouring time |
| Furnace addition | 15-25 | 0.02-0.04 |
| Ladle injection | 50-80 | 0.08-0.12 |
For the addition amount, I used an estimate based on oxygen, sulfur, residual rare earth, and losses. One useful starting relation is:
$$ \text{Re addition} = 5.85([O]_0-[O]_r)+2.93([S]_0-[S]_r)+\text{Re residual} $$
This relation does not include all losses, so I also consider deoxidation, desulfurization, burn-off, and residual requirements. In practice, I tested rare-earth additions of \(0\), \(0.15\%\), \(0.30\%\), and \(0.40\%\) using the ladle method. The samples were normalized at \(900\) to \(920\ \mathrm{^\circ C}\) for one hour, tempered at \(650\) to \(680\ \mathrm{^\circ C}\) for two hours, carburized at \(920\) to \(940\ \mathrm{^\circ C}\) for fourteen hours, quenched at \(830\) to \(850\ \mathrm{^\circ C}\) for twenty minutes, and tempered at \(180\) to \(200\ \mathrm{^\circ C}\) for one hour. The mechanical results are summarized below.
| Rare-earth addition (%) | Tensile strength, sample type 1 (MPa) | Tensile strength, sample type 2 (MPa) | Impact toughness, sample type 1 (J/cm²) | Impact toughness, sample type 2 (J/cm²) |
|---|---|---|---|---|
| 0 | 217 | 212 | 1.08 | 3.07 |
| 0.15 | 236 | 312 | 1.58 | 5.17 |
| 0.30 | 243 | 319 | 1.63 | 5.73 |
| 0.40 | 234 | 288 | 1.50 | 3.80 |
I selected \(0.15\%\) rare-earth addition by ladle injection as the best balance of quality, cost, and benefit. At this level, rare-earth recovery was \(50\) to \(80\%\), and residual rare earth was \(0.08\) to \(0.12\%\).
Pouring Trials and Density Results
I poured the same heat into castings and test bars both with and without rare-earth addition. For the rare-earth-modified full mold casting, gas in the liquid formed rare-earth compounds, so less gas precipitated before solidification. Feeding ability also increased. As a result, gas-related rejects decreased and density improved. The shrinkage porosity was significantly reduced.
| Casting type | Number poured | Gas-related rejects | Density after nitric acid etching |
|---|---|---|---|
| Without rare earth | 10 | 2 | Severe shrinkage cavity and porosity visible |
| With rare earth | 15 | 0 | Slight shrinkage porosity visible |
| Material | C (%) | Si (%) | Mn (%) | Cr (%) | Mo (%) | S before Re (%) | S after Re (%) | Re residual (%) |
|---|---|---|---|---|---|---|---|---|
| ZG20CrMnMo | 0.24 | 0.35 | 1.00 | 1.32 | 0.23 | 0.029 | 0 | 0 |
| ZG20CrMnMoRe | 0.24 | 0.57 | 1.00 | 1.32 | 0.23 | 0.029 | 0.015 | 0.102 |
Mechanical and Wear Performance After Rare-Earth Modification
After normalizing, tempering, carburizing, quenching, and tempering, I tested bending strength and impact toughness. I distinguish between true carburized samples, representing the carburized layer, and false carburized samples, representing the core structure. The rare-earth-modified steel showed higher bending strength and much higher impact toughness. The true carburized impact toughness increased by about \(50\%\), and the false carburized impact toughness increased by nearly three times.
| Material | Bending strength, true carburized (MPa) | Bending strength, false carburized (MPa) | Impact toughness, true carburized (J/cm²) | Impact toughness, false carburized (J/cm²) |
|---|---|---|---|---|
| ZG20CrMnMo | 187 | 218 | 1.2 | 2.2 |
| ZG20CrMnMoRe | 198 | 288 | 1.8 | 6.4 |
Wear tests showed that the rare-earth-modified steel was \(10\) to \(20\%\) more wear resistant than the unmodified steel. I summarize the wear data below.
| Material | Hardness (HRC) | Revolutions | Mass loss (g) | Dimension loss (mm) |
|---|---|---|---|---|
| ZG20CrMnMo | 59-61 | 5000 | 0.2087 | 0.0825 |
| ZG20CrMnMo | 59-61 | 20000 | 1.7770 | 0.3360 |
| ZG20CrMnMo | 59-61 | 40000 | 3.7872 | 0.8810 |
| ZG20CrMnMoRe | 60-62 | 5000 | 0.1557 | 0.0500 |
| ZG20CrMnMoRe | 60-62 | 20000 | 1.2368 | 0.2250 |
| ZG20CrMnMoRe | 60-62 | 40000 | 3.2678 | 0.7170 |
Service Life Improvement in the Full Mold Casting Application
The field test was conducted under hard rock conditions with rock hardness \(F=12\) to \(16\), air pressure of \(5\) to \(6\ \mathrm{kg/cm^2}\), and water pressure of \(5\) to \(7\ \mathrm{kg/cm^2}\). Two rare-earth-modified full mold castings were installed and monitored for more than ten months. Both exceeded ten thousand meters of advance. This was more than twice the life of the unmodified casting. Later, by reducing wall thickness and machining allowance, I further improved the qualified rate and increased service life to more than three times the original value. The modified practice was then incorporated into normal production.
| Run | Jun | Jul | Aug | Sep | Oct | Nov | Dec | Jan | Feb | Mar | Total |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Run 1 | 1000 | 1148.5 | 766 | 2267.5 | 1112.5 | 2016 | 1882 | 1228.3 | 965.5 | 745.5 | 13131.7 |
| Run 2 | 1020 | 1183 | 233 | 1237 | 134.2 | 1357.5 | 1389 | 635.5 | 1514 | 774.7 | 10675.7 |
My Diagnostic Matrix for Shrinkage in Full Mold Casting
When a shrinkage defect appears, I use a diagnostic matrix to move from symptom to cause to action. This prevents random changes to the process. In full mold casting, the most common error is to increase pouring temperature without checking gate modulus, or to increase gate size without checking whether the gate itself becomes a hot spot.
| Symptom | Likely cause | My first check | Corrective action |
|---|---|---|---|
| External cavity on top surface | Riser too small or top crust formed | Riser modulus and hot topping | Increase riser, add exothermic cover |
| Internal cavity at thermal center | No feed path or gate froze early | Gate modulus and position | Relocate gate, increase sprue area |
| Corner cavity | Poor local cooling | Corner geometry and module spacing | Add chill, adjust gate, change fillet |
| Dispersed porosity | Gas and weak gradient | Deoxidation and shell cooling | Improve melt cleanliness and cooling |
| Cavity at gate junction | Gate too large | Gate-to-casting modulus | Reduce gate size or change shape |
| Cavity in upper half only | Riser height ineffective | Feeding distance and taper | Use insulated riser and taper |
| Multiple cavities in isolated bosses | Structure not feedable | Hot spot map | Redesign, pad, or add gates |
Quantitative Checks I Use During Process Design
I use several quantitative checks before releasing a full mold casting process. The first is the hot-spot diameter check. The second is the gate-to-hot-spot ratio. The third is the sprue-to-gate area ratio. The fourth is the modulus comparison between casting hot spot and feeding reservoir. The fifth is the pouring time and shell temperature check. These checks are not a substitute for simulation, but they catch many errors quickly.
| Check | Formula or rule | Acceptance target |
|---|---|---|
| Liquid contraction | \(\varepsilon_{V,l}=\alpha_{V,l}(t_p-t_l)\times100\%\) | Estimate volume demand |
| Solidification contraction | \(\varepsilon_{V,c}\) from alloy table | Estimate cavity potential |
| Total contraction | \(\varepsilon_{V,total}=\varepsilon_{V,l}+\varepsilon_{V,c}+\varepsilon_{V,s}\) | Compare with feeding capacity |
| Inner gate diameter | \(D_i=KD_h,\ K=0.7\sim0.9\) | Gate feeds hot spot without becoming hot spot |
| Sprue area | \(A_s=1.4A_i\) | Sprue remains liquid for feeding |
| Runner area with feeding | \(A_r=(1.1\sim1.3)A_s\) | Distributes and feeds |
| Runner area without feeding | \(A_r=(0.7\sim1.0)A_s\) | Distributes without excess heat |
| Pressure head | Height above pattern and sprue height | Enough feed flow |
| Shell temperature | Recorded before pour | Balances filling and freezing |
Practical Implementation Checklist
I use the following checklist in full mold casting production. It is written as a sequence because each step affects the next. If I skip a step, I often create a shrinkage problem that appears only after sectioning or field failure.
| Step | Action | Question I answer | Output |
|---|---|---|---|
| 1 | Review casting geometry | Where are the hot spots and isolated thick sections? | Hot-spot map |
| 2 | Select feeding zones | Can each hot spot be fed? | Feeding plan |
| 3 | Design gating | Will the gate freeze after the casting? | Gate and riser dimensions |
| 4 | Check structure | Can wall thickness be made more uniform? | Design change list |
| 5 | Select alloy and melt practice | Is contraction and gas content controlled? | Melt procedure |
| 6 | Set pouring conditions | What temperature and speed balance contraction and feeding? | Pouring card |
| 7 | Set shell and cooling practice | Where do I need a sharper gradient? | Shell and chill plan |
| 8 | Simulate or section | Does the last liquid lie in the gating system? | Predicted shrinkage map |
| 9 | Run trial and inspect | Are cavities or porosity present? | Trial report |
| 10 | Standardize | Can the process be repeated? | Production standard |
My Conclusions on Shrinkage Control in Full Mold Casting
I conclude that the fundamental method for reducing or eliminating shrinkage cavities is to achieve directional solidification. In full mold casting, this means that the casting must freeze progressively toward the gating system, and the last liquid must remain in the gating system. To achieve this, I consider the casting structure, gating design, machining allowance or padding, shell temperature, pouring conditions, local cooling, and pressure head. These factors are linked. A change in one factor often changes another. For example, increasing superheat may improve feeding, but it also increases liquid contraction and can worsen residual stress. Increasing gate size may improve feeding, but it can create a new hot spot at the junction.
I also conclude that directional solidification can create residual stress, distortion, or cracks if the temperature difference is too large. Therefore, I do not maximize the gradient blindly. I first secure feeding, then reduce stress by controlling shell temperature, cooling rate, andpost-pour handling. The best full mold casting process is not the one with the largest riser or the highest pouring temperature. It is the one with a stable feeding path, controlled thermal gradient, clean melt, and repeatable production variables.
Finally, I have seen that rare-earth modification can be a powerful supplement in full mold casting. It improves fluidity, increases feeding ability, reduces shrinkage cavities and porosity, and improves mechanical properties and wear resistance. In the application case, rare-earth modification more than doubled service life, and further structural and machining improvements raised it to more than three times the original value. This confirms my practical rule: solve the shrinkage mechanism first, then use alloy and process modifications to strengthen the result. When I combine directional solidification, correct gating, clean melting, controlled pouring, and local cooling, I can greatly reduce shrinkage defects, lower cost, and improve the reliability of full mold casting.