How to Specify a Wire EDM Vendor in India: GD&T, Recast Layer Limits, and the QC Checklist That Prevents a Failed Batch

Srihari Maddula • Founder & Technical Lead, Eurth Techtronics Pvt Ltd 

Category: Manufacturing Realities 

Estimated Reading Time: 8 min

A hardware team sources a precision wire EDM part from a locally-qualified vendor, receives a batch that looks visually correct, integrates it into an assembly — and discovers weeks later, once the assembly is under real operating stress, that a meaningful fraction of the batch fails prematurely at exactly the machined feature the EDM process was supposed to produce reliably. The root cause, traced back, is almost never that wire EDM as a process was the wrong choice — it's that the specification handed to the vendor left critical parameters (recast layer limit, specific inspection method, acceptance criteria) implicit rather than explicit, and a competent vendor executed exactly what was actually specified, which turned out to be looser than what the application actually needed. This post covers how to specify a wire EDM job correctly the first time.

Overview: Why Wire EDM Specification Fails Silently

Wire EDM is a mature, well-understood precision manufacturing process, and most vendors offering it in India's major industrial hubs (Hyderabad's Balanagar Industrial Estate is one well-known concentration, along with similar clusters in other major manufacturing cities) are genuinely competent at executing a clearly-specified job. The failure mode this post addresses isn't vendor incompetence — it's specification incompleteness, where a drawing that looks complete to an engineer unfamiliar with EDM-specific parameters is missing exactly the details (recast layer limit, specific surface finish measurement method, dimensional tolerance stack-up at the feature that actually matters) that determine whether the delivered part meets the application's real functional requirement, not just its nominal dimensions.

Technical Details & Specifications: What a Complete Wire EDM Spec Actually Requires

A wire EDM specification that actually protects the buyer needs to go meaningfully beyond a standard mechanical drawing's usual dimensional tolerances, because EDM has process-specific failure modes a generic machining drawing doesn't address.

Specification element	Why generic drawings miss it	What to specify instead
Recast layer depth	Not a parameter that shows up in a standard dimensional drawing at all	Explicit maximum recast depth (e.g. ≤ 2µm for a fatigue-critical feature), with a defined inspection method
Surface finish (Ra) at the specific critical feature	A drawing may specify a general surface finish callout that doesn’t distinguish the critical feature from non-critical surfaces	Feature-specific Ra callout on the actual functional surface, not a blanket value across the whole part
Number of EDM passes	Rarely specified explicitly, left to vendor discretion	Minimum pass count (rough + finish passes) where recast control matters, since a single-pass cut trades speed for exactly the recast depth a fatigue-critical feature can’t tolerate
GD&T at the critical feature	General tolerancing may not call out true position, profile, or perpendicularity at the specific feature that actually matters functionally	Explicit GD&T callouts (true position, profile tolerance) at the functionally critical feature, not just a linear dimension with a ± tolerance
Material certification	Assumed but not always verified	Mill certificate requirement confirming actual alloy composition and, where relevant, grain orientation (plate vs. extrusion)

EXAMPLE SPEC CALLOUT BLOCK (what should appear on the drawing or

in an accompanying process spec document, not left implicit):

FLEXURE HINGE FEATURE [ref. detail view B]:

  - Cross-section: 0.10mm +/- 0.005mm

  - Surface finish: Ra <= 0.4 micrometers, measured per [specified

    standard, e.g. ISO 4287], on the hinge face specifically

  - Recast layer: <= 2 micrometers, verified via cross-sectional

    metallurgical inspection on first-article and per defined

    sampling frequency thereafter

  - EDM process: minimum 3-pass (1 rough + 2 finish), finish pass

    parameters to prioritize recast minimization over cut speed

  - Material: 6061-T6 PLATE STOCK (not extrusion) -- mill

    certificate required confirming form and alloy

  - GD&T: true position 0.01mm at datum [A], profile tolerance

    0.005mm on hinge face relative to datum [A|B]

THE RULE:  If recast layer depth, surface finish measurement method, and GD&T at the critical feature aren't explicitly on the drawing or spec document, they're not part of the contract — a vendor executing exactly what's specified, and nothing more, isn't a quality failure on their part. It's a specification failure on the buyer's.

Advantages of Rigorous Specification: What It Actually Buys

A complete specification, done once at the start of a vendor relationship, pays back across every subsequent order — it removes ambiguity about what's acceptable, which reduces the back-and-forth rejection-and-rework cycle that an incomplete spec produces when a delivered batch technically meets the drawing's stated dimensions but doesn't meet the application's actual functional requirement. It also gives the buyer a real, objective basis for rejecting a non-conforming batch — an implicit expectation ("the surface should obviously be smooth enough for a precision application") is not enforceable in a vendor dispute the way an explicit Ra callout with a defined measurement method is.

A rigorous specification also functions as a genuine vendor-qualification filter: a vendor capable of understanding, quoting accurately against, and reliably meeting a spec with explicit recast-layer and GD&T requirements is demonstrating real process capability at the precision tier the application needs. A vendor who balks at, misunderstands, or quotes vaguely against these requirements is signaling — usefully, before an order is placed rather than after a failed batch — that they may not be the right fit for this specific tolerance tier, even if they're a perfectly competent general machining shop for less demanding work.

Challenges: Where Rigorous Specification Gets Genuinely Hard

Writing a complete EDM-specific specification requires knowledge most mechanical engineers, trained broadly rather than in EDM process specifics, don't have by default — recast layer behavior, appropriate pass-count strategy, and realistic achievable tolerance at a given feature size are genuinely specialized knowledge, and a first-time buyer without in-house EDM expertise faces a real chicken-and-egg problem: writing a correct spec requires knowledge that's usually gained through prior EDM experience, which a first-time buyer doesn't yet have. The practical mitigation is treating the first vendor engagement as a genuinely collaborative specification process — sharing the functional requirement (fatigue life, measurement resolution, whatever the actual downstream need is) rather than just a dimensional drawing, and asking a qualified vendor to propose the process parameters that would meet that functional requirement, then reviewing and formalizing their proposal into an explicit spec for future orders, rather than attempting to write a fully specified drawing from a position of limited EDM-specific expertise on the first attempt.

Inspection capability is the other genuine challenge: verifying recast layer depth requires either cross-sectional metallurgical inspection (destructive, meaning it consumes a sample part rather than inspecting the actual delivered unit) or a specified non-destructive method, and not every vendor has this capability in-house — it may require a separate metallurgical lab relationship, adding cost and lead time to the qualification and ongoing quality-control process that a buyer needs to budget for explicitly rather than assuming inspection capability is bundled into standard EDM service.

THE RULE:  Recast layer verification is destructive by its most reliable method — budget for sacrificial sample parts as a real, recurring quality-control cost, not a one-time first-article expense, if recast depth genuinely matters for the application's reliability.

Case Study: A Specification Gap Caught Before It Became a Failed Batch

During specification development for a precision flexure component, an early draft drawing specified the hinge cross-section dimension and a general surface finish callout, but left EDM pass count and recast layer depth unspecified — a gap that, in a less careful process, would likely have gone unnoticed until a delivered batch either failed fatigue testing or, worse, passed initial testing but failed prematurely in the field after the specific recast-layer-driven fatigue degradation this post has covered throughout. Catching this gap required specifically cross-referencing the drawing against the functional requirement (millions of fatigue cycles over instrument service life) and recognizing that a general surface finish callout doesn't constrain recast layer depth — these are related but genuinely distinct parameters, and a drawing that specifies one without the other has a real, exploitable gap a vendor could execute against without technically violating the spec.

The fix — adding explicit recast layer limits, minimum pass count, and a defined inspection method to the specification before it went to any vendor for quoting — meant the resulting vendor quotes reflected the true process requirement from the start, rather than requiring a re-quote cycle after an initial vendor's lower quote turned out to reflect a single-pass process that wouldn't have met the actual recast-layer requirement. This is the value of catching a specification gap at the drafting stage rather than at the receiving-inspection stage: the cost of adding a missing requirement to a specification document is negligible; the cost of the same gap surfacing as a failed batch, after tooling, quoting, and a full production run have already happened against an incomplete spec, is not.

Implementation Plan: Specifying and Qualifying a Wire EDM Vendor

• Start from the functional requirement (fatigue life, measurement resolution, whatever the actual downstream need is), not just a dimensional drawing — this is what surfaces whether recast layer, specific surface finish, or GD&T at a particular feature actually matters for this specific part.

• For a first-time EDM buyer without in-house process expertise, treat the first vendor engagement as collaborative specification development — share the functional requirement and ask a qualified vendor to propose parameters, then formalize their proposal into an explicit spec.

• Include recast layer limit, minimum pass count, feature-specific surface finish (not a blanket value), explicit GD&T at the critical feature, and material certification requirements on every EDM specification going forward, not just for the first, highest-scrutiny order.

• Confirm the vendor's inspection capability for every specified parameter before placing an order — specifically ask how recast layer depth will be verified, and whether that requires a separate metallurgical lab relationship the vendor doesn't have in-house.

• Budget for destructive sample inspection (sacrificial parts pulled from each batch for cross-sectional recast verification) as an ongoing quality-control cost for any application where recast depth genuinely drives reliability, not a one-time first-article formality.

• Qualify a new vendor against a real trial order with full specification and inspection before committing to a production-volume order — a vendor's general reputation or capability claim is not a substitute for verified performance against this specific spec.

• Document any specification gap found during review, however it's caught, and add it to the standard specification template used for future orders — this is exactly the kind of institutional knowledge that shouldn't need to be rediscovered on every new project.

Conclusion: The Specification Is the Actual Deliverable

A competent wire EDM vendor will reliably deliver exactly what's specified — which means the single highest-leverage action a buyer can take to prevent a failed batch isn't more aggressive vendor vetting or a stricter incoming-inspection process, though both have their place; it's writing a specification complete enough that "exactly what's specified" and "what the application actually needs" are the same thing. The gap between those two is where nearly every precision-manufacturing quality failure in this category actually originates, and closing it is unglamorous, detail-heavy work — recast layer limits, feature-specific surface finish, explicit GD&T, verified inspection methods — that pays for itself many times over against the cost of a batch that technically meets an incomplete spec and fails anyway.

EurthTech delivers AI-powered embedded systems, IoT product engineering, and smart infrastructure solutions — Hyderabad, India. www.eurthtech.com