NASHUA, N.H. — Keeping an aircraft in service can become complicated when one critical metal part needs replacing and the original manufacturing setup disappeared years ago.

The problem extends beyond finding a drawing or recreating the shape of the component. For a cast part, engineers may also need to reconstruct the process that produced it. The original tooling may no longer exist, suppliers may have left the market, and some of the manufacturing knowledge may have disappeared with them.

Digital casting combines process simulation with additively manufactured sand molds and cores to give aerospace manufacturers another way to approach that problem.

A casting requires more than the shape of the finished part

Metalcasting sounds straightforward at its most basic: molten metal enters a mold and cools into the desired shape. Producing a complex, repeatable casting requires a lot more control.

Engineers have to account for how metal travels through the mold and how different sections cool. For example, thick areas may remain hot after surrounding material starts to solidify. Metal also contracts as it cools, which can contribute to shrinkage, porosity and other defects if the mold does not feed material where it is needed.

The mold package therefore includes more than a cavity matching the finished component. Gates, runners and risers help control the flow and supply of molten metal throughout the process.

For a new casting, engineers can develop that process alongside the component. A legacy part presents a different challenge. Having a surviving component or its geometry does not necessarily reveal exactly how the original manufacturer produced it.

Simulation moves some trial and error ahead of the pour

Traditionally, foundries can learn about a casting process by producing a part, inspecting it, adjusting the process and trying again. That physical iteration becomes expensive when each attempt consumes material, labor and foundry capacity. Simulation gives engineers a way to investigate some of those problems before pouring metal.

Casting software can model fluid flow and solidification within a proposed mold package. Engineers can use those results to identify areas likely to remain hot, determine how metal feeds different regions and find locations that may be vulnerable to shrinkage or other issues. Then, they can modify the design and run the simulation again. 

This does not guarantee a successful casting, and teams still have to validate the physical process. It can, however, move part of the learning cycle into a digital environment where changing a design does not require scrapping tooling or another unsuccessful pour.

Printed sand molds expand casting options

Simulation and additive manufacturing solve different parts of the casting problem. A simulation may identify a mold configuration that improves metal flow or solidification, but conventional moldmaking still has physical constraints. Patterns, parting lines and separately manufactured cores can limit how teams translate a digital design into actual tooling.

3D-printed sand molds and cores remove some of those restrictions. Manufacturers can produce the mold package directly from CAD data. This lets engineers create more complicated gating arrangements and internal geometries without first making a corresponding pattern or core box.

Printed cores can also consolidate geometries that conventional manufacturing might divide among several pieces. Fewer assembled sections can minimize opportunities for alignment problems while allowing engineers to design internal features around specific requirements.

For a difficult component, simulation helps determine how the mold should work, while additive manufacturing gives engineers greater freedom to produce that design.

Legacy aircraft parts change the manufacturing equation

Those capabilities become particularly interesting when an aircraft needs a component that no longer has an active supply chain. Recreating conventional tooling can make sense when a team expects a long production run. The economics look different when an operator or supplier needs a limited quantity of replacement castings for an aircraft that may have entered service decades earlier.

A digital process can allow engineers to begin with available part data, develop a new mold package and refine it for production before printing the sand mold for a physical casting.

Arc Impact is applying that workflow through its Digital Casting Services, which combine mold-package design, gating and rigging development, fluid-flow and solidification simulation, and 3D-printed sand molds and cores. The company is using the process for complex castings as well as obsolete and difficult-to-source parts.

For aerospace sustainment, the potential benefit is a more practical way to produce low-volume replacement castings without recreating a conventional production line.

A successful pour is only part of the process

Digital tools do not eliminate the requirements that come with producing critical aerospace hardware. Teams still have to prove that the finished casting meets the requirements of its application. Depending on the component, that can include heat treatment, dimensional inspection, nondestructive evaluation, material testing and other qualification steps.

Repeatability matters as well. Producing one acceptable component does not automatically establish a manufacturing process capable of consistently producing qualified parts. That distinction becomes especially important in aerospace, where the consequences of a defective component can go far beyond the cost of the casting itself.

Digital casting instead changes where some of the development work takes place. Engineers can investigate flow, cooling and mold geometry before committing foundry resources to a physical iteration. Then, they can use printed tooling to put the resulting design into practice.

For aging aircraft, that could make the difference between knowing what an obsolete component looks like and having a practical way to manufacture it again.