Sourcing Investment Castings from India: A Buyer's Guide

Most investment casting orders that go wrong don't fail because the foundry couldn't cast the part. They fail because the RFQ never said which alloy variant, which tolerance class, or which inspection method the part actually needed. The foundry just quoted its default instead of your real requirement.

Investment casting, the lost-wax process, makes it possible to cast complex, near-net-shape parts in tight sections. Forging or CNC alone can't reach those shapes economically. That same capability is why the process hides risk so well. A wax pattern can produce a beautiful first article. It can also produce a batch with internal porosity that nobody sees, until the part is machined or put into service. Sourcing this category from India well means three things. Specify tightly. Qualify the foundry's real inspection capability. Never let a single sample stand in for a production run.

When investment casting isn't the right process

Not every part that could be cast should be. Investment casting earns its cost premium on complex geometry, thin sections, and near-net shapes that would be slow or wasteful to machine from bar stock or fabricate as a weldment. It doesn't earn that premium on a simple shape with a loose tolerance, where a cheaper process does the job just as well.

A few signals point away from investment casting. The geometry is simple enough for machining or fabrication to compete on cost. The tolerance requirement is loose enough that sand casting, at a wider tolerance grade and a lower cost, would hold it. The order volume is too low to absorb the tooling lead time investment casting requires, with no design feature that actually needs the process.

If two or more of these describe your part, it's worth pricing a simpler process before committing to an investment casting RFQ.

What to specify on the RFQ

An investment casting RFQ that just says "stainless steel" or "as per drawing" invites the foundry to quote its easiest option. Not your part's real requirement. Four things belong on every RFQ, named explicitly.

Alloy and grade, not just the family. "Stainless steel" covers dozens of grades. Each has different corrosion resistance and machinability. ASTM A351 specifies requirements, including chemistry, for cast austenitic stainless grades. ASTM A957 sets common requirements for steel and alloy investment castings more broadly. ASTM A732 covers carbon and low-alloy steel investment castings specifically. If your drawing calls for CF8M, write CF8M. Not "316 stainless." Not "stainless steel."

Dimensional tolerance class. Investment casting typically achieves ISO 8062-3 tolerance grades of CT4 to CT8, depending on geometry and wall thickness, tighter than sand casting's typical CT10 to CT14 range and looser than machining. State the tolerance standard and class you expect, referenced against ISO 8062-3 or against a foundry's own process capability data. Mark which dimensions are critical to function, and which are general casting tolerance. A foundry left to guess will hold the loosest tolerance the process allows.

Surface finish requirement. As-cast surface finish varies by shell system and alloy. If a surface will mate with another part or carry a coating, specify the finish requirement instead of leaving it implied.

NDT (non-destructive testing) requirement, by defect type, not just "inspect it." This is the specification buyers skip most often, and the one with the most consequence. See the comparison below.

Which inspection catches which defect

Investment castings develop a specific, recurring set of defects. Gas porosity is rounded voids from trapped gas. Shrinkage porosity is irregular, dendritic voids where thick sections solidify last. Inclusions are fragments of ceramic shell or refractory material caught in the metal. There are also misruns and cold shuts. Each defect type is best caught by a different inspection method. Ask for "quality inspection" without naming the method, and the foundry picks the cheapest one that technically counts as inspection.

A part destined for a pressure boundary or a rotating assembly needs radiography for internal soundness. A bracket that only needs to look clean and hold a bolt pattern does not. Specify by defect risk, not by habit.

How to qualify a foundry before the first order

A foundry's certificate tells you a quality system existed on the day of the audit. It doesn't tell you whether that foundry actually runs the inspection your part needs. Not on your alloy, not at the volume you need. Three questions separate a foundry that can talk about quality from one that actually runs it.

1. "Show me a radiography or dye-penetrant report from a part in this alloy family. Not a marketing sample." A foundry that inspects routinely can produce a real report in minutes. One that can't is likely subcontracting NDT out. That's not disqualifying on its own. But it changes what you need to check: turnaround time, chain of custody, and whether the subcontracted lab is accredited.
2. "What's your typical scrap or rework rate on this alloy? Where does it usually show up?" A foundry that knows the honest answer understands its own process. One that claims zero defects doesn't know its own numbers.
3. "Walk me through what happens between wax injection and shell removal for a part this size." The lost-wax process has more manual steps than most buyers assume. Pattern assembly. Shell building in multiple ceramic coats. Dewax. Pour. Shell removal. Cutoff. Finishing. A foundry that can describe its own sequence in detail, not in general terms, is one that actually controls it.

Red flags specific to investment casting

Some red flags are true of any manufacturing category. A few are specific to this one and worth naming directly.

• No in-house NDT capability and no named accredited subcontractor. "We can arrange inspection" isn't a yes, unless there's a specific answer for who does it and to what standard.
• No heat or melt traceability. A foundry should be able to tie a batch of castings back to the specific melt (heat number) that produced them. Without that link, a chemistry failure on one part tells you nothing about the rest of the batch.
• Pattern or tooling ownership left undiscussed. Investment casting tooling, the wax injection die, is a real cost. Who owns it? What happens to it if you move the order elsewhere? Settle this before the first production run. Not after a dispute.
• A single "golden sample" presented as representative of the whole batch. This is common enough, and consequential enough, that it gets its own section below.

The sample-to-batch risk, with one worked example

A first-article investment casting is often the best casting that foundry will ever make for you. It gets extra attention. A more careful wax pattern. Closer supervision during shell building. A slower, more deliberate pour. None of that is dishonest. It's just not representative of casting number 400 in a production run, made under normal shop conditions by whoever's on shift.

Here's a case worth remembering. A buyer qualifies a stainless steel valve body casting on a golden sample. It passes a dimensional check and a visual inspection. No radiography was requested at sampling, because the part looked clean. At production volume, a batch of castings ships. Several fail at final machining. A machining cut opens up internal shrinkage porosity that was never visible from the outside, and was never checked by X-ray. The casting process didn't change between the sample and the batch. The inspection scope did. The sample passed a test that was never strict enough to catch the real failure mode. Nobody found out until real money had been spent machining defective parts.

The fix is specific. Any casting where internal soundness matters gets radiography at the sampling stage. Test it on the exact tolerance and alloy you plan to run in production. Not just a visual pass. A sample that looks right is not the same as a sample that's been checked for the specific defect that matters.

Before you finalize a foundry: a short readiness check

If more than two of these are true, your qualification process is not yet ready to commit to a production order.

• You have not seen a real NDT report on this alloy family, only a certificate.
• Nobody has confirmed which specific tolerance class the foundry's process actually holds, as opposed to what the drawing requests.
• Heat or melt traceability has not been discussed.
• Tooling ownership has not been agreed upon in writing.
• The only casting you have evaluated is the golden sample, and it has not been checked with the inspection method your risk actually requires.

Sourcing this category from India

India's investment casting market reached USD 579 million in 2025 and is projected to reach roughly USD 746 million by 2034, according to IMARC Group, a large and still-growing base. Trade bodies here include the Institute of Indian Foundrymen, established in 1950, and the Indian Investment Casting Manufacturers Association, which focuses specifically on lost-wax process technology.

Capability varies widely across that base. Some foundries run export-grade NDT and traceability systems. Others can cast the shape but can't yet document the process behind it. The specification and qualification steps above are what actually separate the two. That matters more than foundry size or output volume.

See Also

• Payment Protection For First Order
• Supplier Base Rationalization
• Should-Cost Model - Beyond the Unit Price
• Red Flags When Vetting a New Manufacturing Supplier
• Factory Vs Broker