Metal casting using lost wax models: main stages 

Metal casting using lost wax models: main stages of the technological process

Lost wax casting technology remains the only industrial method that makes it possible to produce metal parts of the most complex geometry with a surface roughness of Ra 1.6–3.2 μm without subsequent mechanical processing. Unlike sand casting or injection molding, where restrictions are imposed by the shape of the flask or mold, LBM gives designers the freedom to create internal channels, thin ribs and cavities that cannot be achieved otherwise. However, this versatility comes at the price of high labor intensity and strict adherence to temperature conditions at every step. Our practice shows that 80% of defects in this technology occur not at the metal pouring stage, but at the stages of assembling models or drying ceramic shells, where the human factor plays a decisive role.

This material was prepared by process engineers with more than 15 years of experience in precision casting shops. We will analyze each stage of production - from creating a master model to final cleaning of castings, based on real production cases, GOST and ISO standards, as well as defect statistics collected over the past three years. If you are planning to purchase components for the aerospace, medical technology or turbine industries, understanding these nuances will help you avoid costly mistakes when choosing a supplier.

Preparation of initial data and production of master models

The foundation of the entire process is the quality of the master model (mold), which determines the geometry of the future wax model. The error here is multiplied at all subsequent stages. In modern manufacturing, we have eliminated hand milling of aluminum in favor of High Speed ​​Machining (HSM) and Electrical Discharge Cutting (EDM) to achieve IT7-IT8 tolerances. For series of over 5,000 pieces, it is advisable to use steel molds hardened to 45–50 HRC, while for pilot batches aluminum alloys AK4 or AK6 are sufficient, which reduces the cost of tooling by 30–40%.

A critical parameter in mold design is taking into account material shrinkage. Wax contracts as it cools, and metal contracts as it hardens. The total shrinkage can reach 2.5–3.0% depending on the alloy (steel, titanium, aluminum) and the type of wax composition. Our engineers use specialized software to automatically scale 3D models, but in complex assemblies with variable wall thicknesses, we always carry out a test casting of the first batch and make adjustments to the mold geometry manually. Ignoring this rule led one of our customers to reject an entire batch of pump casings: the difference in shrinkage between the thick and thin sections caused warping that could not be corrected mechanically.

The surface of the working cavity of the mold must be polished to a mirror finish (Ra<0.2 µm). Any scratches or marks will be replicated on each wax model and will require additional grinding of the finished castings, which will increase the cost of the process. We recommend using chrome plating on the working surfaces of steel molds to increase their service life to 50,000 injection cycles. Before being put into production, the mold must be tested for the tightness of the cooling channels and the uniform heating of the plates.

Recommendation:Request a report from the supplier of the initial measurements of the master model using a coordinate measuring machine (CMM). This is the only way to ensure that the geometry matches your drawing before mass production of wax models begins.

Making wax models and assembling clusters

The process of producing wax copies begins with the injection of a molten model composition into a mold under a pressure of 0.4–0.6 MPa. The temperature of the wax is critical: overheating above 70°C leads to changes in rheological properties and increased shrinkage, and underheating causes incomplete filling of thin sections. In our production, we use automatic injection machines with a closed-loop temperature control, which ensures model weight stability within ±0.5 g. For critical parts, such as turbine blades, we use vacuum injection, which eliminates the formation of gas bubbles inside the body of the model.

After removal from the mold, each model undergoes visual inspection and measurement of critical dimensions. Defects such as underfilling, flash or warping are corrected manually using scalpels and heating tools. This operation requires highly skilled operator: inaccurate removal of flash can change the profile of the part, and excessive pressure from the tool leaves dents that will appear on the metal casting. Statistics from our workshop show that up to 15% of modellers’ working time is spent on revision, so investments in high-quality equipment pay off by reducing labor costs at this stage.

The next step is to assemble a cluster (model tree). The individual wax parts are attached to the center sprue (riser) using a heated tool or ultrasonic welding. The geometry of the cluster is calculated to ensure directional crystallization of the metal from distant points to the feeders. Incorrect pattern orientation can result in slag entrapment or shrinkage cavities in the body of the part. We use mold-fill modeling software (e.g., ProCAST or AnyCasting) to optimize the clustering scheme before producing the first sample.

The weight of the cluster usually varies from 1 to 5 kg depending on the power of the melting equipment and the type of alloy. Clusters that are too heavy are difficult to handle when depositing ceramics, and clusters that are too light reduce kiln productivity. It is important to maintain a balance between the number of parts in the wood and the quality of the resulting castings. In our practice, there was a case when an attempt to increase the yield of a suitable cluster by compacting it led to the fact that the ceramics in the center of the tree did not dry out evenly, causing cracks during firing.

Action:Check with the manufacturer for the method of connecting models to the gating system. Ultrasonic welding is preferable for highly loaded components, as it eliminates the risk of the model coming off when applying the suspension.

Molding of ceramic shells: layer-by-layer application

Creating a ceramic mold is the longest and most technologically intensive stage of the LVM process. The shell must withstand thermal shock when pouring metal at temperatures up to 1650°C (for heat-resistant alloys) and maintain dimensional stability. The process is based on the “layer cake” principle: alternating liquid suspensions and dry sands of different fractions. The first layer (front) is applied by dipping or pouring and determines the cleanliness of the surface of the future casting. Here, finely dispersed zircon powder (ZrSiO4) with a fraction of 0.05–0.1 mm, bound with ethyl silicate or colloidal silica, is used.

The quality of the front layer directly affects the need for subsequent mechanical processing. If the suspension contains large inclusions or does not wet the wax well, burns and unevenness will form on the metal. We monitor the viscosity of the suspension with a viscometer every 2 hours of line operation, since the evaporation of alcohol or water changes the rheology of the mixture. After applying the first layer, the cluster is sprinkled with electrocorundum of a fraction of 0.1–0.2 mm and dried in a chamber with controlled humidity (40–60%) and a temperature of 20–25°C for 2–4 hours. Violation of the drying regime leads to peeling of the ceramics or the formation of cracks.

Subsequent layers (5–9 layers depending on the mass of the casting) form the load-bearing part of the shell. Here, larger sand fractions (0.3–0.8 mm) and less expensive materials, such as mullite or aluminosilicates, are used. The wall thickness of the finished ceramic mold is usually 6–10 mm. Each layer requires complete drying before applying the next. An attempt to speed up the process by increasing the temperature often results in a “crust” effect, when the surface dries out, but moisture remains inside, causing the mold to collapse with further heating.

Particular attention is paid to the corners and internal cavities of the cluster. In these areas, “bridges” often occur—areas where the suspension has not completely penetrated, creating voids in the ceramic. Our operators use low pressure compressed air immediately after dipping to remove excess slurry from hard to reach areas. Neglect of this operation caused the failure of a batch of medical implants: a thin wall of ceramics in the corner collapsed under metal pressure, and the melt filled unintended cavities.

Advice:Require x-ray inspection or ultrasonic testing of ceramic molds before melting wax, especially for critical parts. This will reveal hidden delaminations that are invisible to the eye.

Wax removal and high temperature firing

After the shell is formed and completely dried, the stage of removing the model composition begins. The traditional method is autoclaving with steam under a pressure of 0.6–0.8 MPa at a temperature of 160–180°C. The wax model melts and flows through the gating system, leaving an exact negative copy inside the ceramic mold. It is critical that the wax comes out completely and quickly. If the process is slow, the melted wax can expand and rupture the ceramic that has not yet been heated. We use “fast heating” technology, in which the temperature in the autoclave rises to operating temperature in 2–3 minutes, creating a maximum pressure gradient.

The wax remaining in the pores of the ceramic (up to 1–2%) is burned out in a muffle furnace at a temperature of 800–900°C for 2–3 hours. This stage also serves as the beginning of sintering of the ceramic material, giving it the necessary strength. Gases released during wax combustion must be freely evacuated from the oven; otherwise, carbon deposits may form on the inner walls of the mold, which will then transfer to the metal surface.

The final firing is carried out at temperatures close to the pouring temperature of the metal (1000–1200°C for steels, up to 1500°C for titanium). The purpose of this stage is to complete phase transformations in the ceramic, remove residual carbon and warm up the mold to minimize thermal shock upon contact with the melt. The holding time depends on the mass of the cluster and the thickness of the mold walls. Underheating of the mold will lead to premature solidification of the metal in thin sections (underfilling), and overheating can cause deformation of the ceramic itself under its own weight.

In our practice, a temperature monitoring system has been introduced inside the cluster itself using thermocouples built into experimental forms. This revealed that standard recommendations from furnace suppliers often underestimate the holding time for large components. An increase in firing time by 15% reduced the percentage of defects due to underfilling from 4% to 0.5%.

Important:Ensure that the supplier uses kilns with a controlled atmosphere (vacuum or inert gas) to fire titanium and nickel alloy molds to prevent oxidation of the ceramic and oxygen contamination of the metal.

Pouring metal and crystallization of castings

The red-hot ceramic mold is immediately poured. For most alloys, this process is carried out in vacuum induction furnaces, which prevents oxidation of the metal and allows dissolved gases to be removed. Vacuuming the chamber to level 10^-2–10^-3Torr before melting and pouring is a prerequisite for obtaining dense metal without gas pores. The overheating temperature of the metal above the liquidus point is usually 100–150°C to ensure good fluidity.

There are two main casting methods: free casting and pressure casting (centrifugal or gas-static casting). Free pour is suitable for simple configurations and alloys with good flow (aluminium, bronze). For complex thin-walled parts made of stainless steels or heat-resistant alloys, we use centrifugal casting. Rotation of the cluster with an acceleration of 20–50 g forces the metal to penetrate into the most remote corners of the mold, compensating for the loss of fluidity during cooling.

The most difficult stage is crystallization control. The metal must solidify directionally, from distant points to the feeders, so that the shrinkage cavities go into the gating system, and not into the body of the part. The cooling rate is controlled by the mold material and environmental conditions. Sometimes the molds are placed in sand or special insulating sleeves to slow down the cooling of massive units. Rapid cooling (for example, in air) can lead to hot cracks in alloys that are prone to segregation.

After hardening, the cluster is left to cool completely (usually 1–2 hours). Sudden removal of a hot mold into cold air can cause thermal shock and cracking of castings, especially in cross-section transition zones. In one case, failure to comply with this rule led to the appearance of microcracks in the gearbox housing, which were detected only at the stage of magnetic particle testing after machining, which resulted in huge losses.

Control:Request a melting report indicating the chemical composition of each batch (spectral analysis). Conformity of the steel or alloy grade must be documented in accordance with GOST or ASTM.

Shell destruction and casting cleaning

After the cluster has cooled, mechanical separation of the castings from the gating system begins. The ceramic shell, which has become brittle after firing, is easily destroyed by vibration or hydraulic pressure (hydro-sandblasting). We use high-pressure units (up to 100 MPa), which effectively wash out ceramic residues from internal cavities and complex channels that are inaccessible to mechanical tools.

The sprues are cut using abrasive wheels, a laser or a waterjet machine. The choice of method depends on the hardness of the alloy and the requirements for the cutting area. For titanium alloys, a laser or waterjet is preferable, since mechanical cutting can cause hardening and local overheating, changing the structure of the metal. After cutting, the joints are cleaned to the level of the base metal.

Final surface cleaning includes shot blasting to remove remaining oxide film and give it a marketable appearance. The roughness after LBM is usually Ra 3.2–6.3 µm, but shot blasting can improve it to Ra 1.6 µm. Particularly critical surfaces are subjected to chemical etching or electropolishing. At this stage, the first continuous visual inspection and measurement of key geometric parameters are carried out.

However, cleaning does not end with the removal of ceramics. It is necessary to make sure that there is no abrasive or sand left in the blind holes and channels, which could damage the mechanism during operation. We practice pressurized nitrogen purging and endoscopic inspection of internal cavities for critical products. One of our customers experienced a jammed hydraulic distributor due to a grain of sand remaining in the channel after poor flushing, which emphasizes the importance of this stage.

Check:Insist on providing photographs or a video report of the cleaning process of internal cavities for parts with complex architecture.

Heat treatment and final quality control

Most castings after LBM require heat treatment to relieve casting stress, homogenize the structure, or achieve the required mechanical properties. Maintenance modes (quenching, tempering, aging, annealing) are selected individually for the alloy grade and operating conditions of the part. For example, castings made of stainless steel 12Х18Н10Т are subject to quenching at 1050°C in water to dissolve carbides, and aluminum alloys undergo artificial aging to increase hardness.

Quality control at the final stage includes a set of non-destructive methods (NDT). Visual measurement inspection (VII) reveals surface defects. X-ray inspection (RK) is necessary to detect internal pores, cavities and inclusions, especially in the areas where thick and thin walls meet. Ultrasonic testing (UT) is used to detect cracks and delaminations in massive sections. For ferromagnetic steels, magnetic particle testing (MPD) is mandatory.

All results are recorded in a quality certificate. We adhere to ISO 9001 standards and industry specific norms such as AS9100 for aviation or ISO 13485 for medicine. Each batch is accompanied by a material certificate and a test report. Lack of such documentation renders castings unsuitable for use in regulated industries.

Geometric control is carried out on coordinate measuring machines (CMMs) with an accuracy of up to 0.005 mm. The digital model of the part is compared with the real object. Deviations beyond the tolerances specified in the drawing are grounds for rejection. It is important to understand that LVM allows you to keep tolerances within IT7–IT8, but achieving IT6 requires additional machining of the base surfaces.

Result:Do not accept a shipment without a full set of NDT and heat treatment certificates. This is your insurance against hidden defects that can only appear under load.

Frequently Asked Questions

What is the maximum weight of one casting when using LVM technology?

It is technologically possible to produce castings weighing up to 50–100 kg, but the economically feasible limit for a single product is 10–15 kg. As the mass increases, the requirements for the strength of the ceramic mold and the power of the melting equipment sharply increase, which leads to an exponential increase in cost. For heavy components, sand casting is often used, followed by machining, if geometry allows.

Can recycled wax be used for new models?

Yes, wax can be regenerated. After melting from the molds, it is collected, filtered from ceramics and impurities, then mixed with fresh composition in a proportion of up to 30–40%. Complete use of recycled wax is not recommended, as repeated heating degrades its rheological properties and increases shrinkage, which negatively affects the dimensional accuracy of models.

What is the minimum print run that justifies the cost of making a mold?

The profitability threshold depends on the complexity of the part and the cost of the equipment. On average, for small parts, a print run of 50–100 pieces already justifies the cost of an aluminum mold. For large and complex units, the break-even point shifts to 200–500 units. For prototypes (1–10 pcs.), it is more economical to use rapid prototyping of master models (3D printing with resins) instead of making a metal mold.

What metals and alloys are suitable for investment casting?

The technology is universal and applicable to almost all industrial alloys: carbon and alloy steels, stainless steels, heat-resistant nickel alloys, titanium, aluminum, magnesium and copper alloys, precious metals. Restrictions apply only to metals with extremely high melting points (tungsten, molybdenum), which require special refractory ceramics and special class vacuum furnaces.

How long does the full production cycle take from order to shipment?

The standard lead time for the first batch (including tooling development) is 4–6 weeks. Manufacturing the mold itself takes 2–3 weeks, the rest of the time is spent debugging the process and producing a trial batch. Subsequent serial batches are produced faster - from 2 to 4 weeks, depending on the order volume and the complexity of clustering. Urgent orders are possible with an additional payment for the accelerated work schedule of the workshop.

Conclusion and selection of a reliable partner

Lost wax casting is a symbiosis of art and high technology, where every stage, from polishing the mold to the cooling mode of the metal, affects the final result. Understanding these processes allows the customer to speak the same language with the manufacturer, avoid unreasonable expectations and control quality at all stages. We have seen how ignorance of the nuances of shrinkage or drying conditions has led to lost months of work and significant financial losses.

When choosing a LVM service provider, pay attention not only to the price per kilogram of casting, but also to the presence of its own NDT laboratory, a fleet of modern equipment and the qualifications of technologists. A manufacturer's ability to offer die-cast part design optimization (DFM) often saves more money than bidding for a 5% discount.

For plants operating under extreme load conditions, not only the casting process itself is critical, but also a deep understanding of the properties of the alloys used in real-world industrial applications. A striking example of this approach is the companyWuxi Kaisheng Electric Power and Petrochemical Equipment Co., Ltd.. Specializing in the design and production of heat transfer equipment for the oil, gas and energy industries, they successfully integrate components obtained by the HLM method into their high-tech products. They offer titanium shell-and-tube heat exchangers, ASME high-pressure units, corrugated tube bundles in 316 stainless steel, C46400 marine brass and copper-nickel alloys, as well as complex N06625 nickel alloy assemblies. The use of precision casting allows us to create efficient tube sheets and other components from titanium, alloy steels and non-ferrous metals, ensuring high corrosion resistance and performance at high temperatures and pressure. The company's PED and ASME certified products are widely used in the marine, desalination and chemical industries around the world, demonstrating that the combination of advanced foundry technology and industry expertise produces the best results.

If you are looking for a reliable partner for the production of complex metal components with a guarantee of quality and meeting deadlines, our company is ready to offer a full range of services: from reverse engineering and tooling development to finishing machining and certification. We work with alloys of any complexity and provide transparency at every stage, just like industry leaders do for their mission-critical projects.

Contact us todayto discuss your project and receive a technical and commercial proposal. Our engineers will conduct a free audit of your design documentation and offer the optimal solution for mass production.