Hangbo Alloy Group
Shanghai · Est. 2012
Non-metallic inclusions seed fatigue cracks and pitting in nickel alloy while chemistry tests still pass. This buyer's guide shows how to rate cleanliness with ASTM E45, macroetch and UT, set a limit and write the purchase-order clause.
Non-metallic inclusions are tiny particles of oxide, sulfide or nitride trapped inside nickel alloy metal during melting and solidification. On their own they are measured in micrometres, but they are among the most under-rated causes of premature fatigue and corrosion failure in high-value nickel alloy parts. A heat that passes chemistry and tensile tests can still carry a cleanliness problem that no standard tensile specimen happened to sample. This guide is written for buyers who specify rotating, cyclic-loaded or corrosion-critical nickel alloy components and who need to know how to specify, measure and verify metallurgical cleanliness before a dirty heat reaches the shop floor.
We explain what inclusions are, why they seed fatigue cracks and pitting, how they are rated under ASTM E45 and related methods, the physical checks that reveal a dirty heat, the standards that let you specify a cleanliness limit, and a ready-to-use set of purchase-order clauses. The prevention steps reflect how Hangbo Alloy Group controls cleanliness inside its own Shanghai plant, where every heat is built from prime virgin raw material, double-melted under vacuum, descaled and consolidated before working, then sampled at multiple stages so a contaminated heat never reaches a customer. For material context, our Inconel 625 guide, our Hastelloy C-276 composition guide and our nickel versus stainless selection guide describe the alloys whose service life inclusions can quietly shorten.
Short answer: inclusions are microscopic non-metallic particles — oxides, sulfides, nitrides and silicates — trapped in the metal matrix during melting, deoxidation and solidification.
They form when elements in the melt react to make compounds that do not dissolve in nickel. Aluminium, calcium, magnesium and silicon oxides are common after deoxidation; manganese sulfides and titanium nitrides appear from residual sulfur and nitrogen; complex silicates can remain from slag entrainment. In a single-melt or poorly controlled heat they are simply unavoidable debris, while in a vacuum double-melted heat their population is driven down by orders of magnitude because the reactive impurities are removed before the metal freezes.
The practical point for a buyer is that inclusions are a cleanliness metric, not a chemistry metric. Two heats can share an identical UNS composition and tensile report while one carries ten times the inclusion population of the other. Cleanliness is a separate property that must be specified and measured on its own.
For the buyer this means two bars with the same UNS number and the same mill certificate can behave very differently in service depending on how carefully the supplier controlled the melt. A plant that melts from scrap or skips the vacuum remelt step will tend to carry more oxide and nitride debris than one that starts from prime material and refines it twice under vacuum. The nominal grade on the certificate cannot tell you which you received, which is the core reason cleanliness must be a measured, specified property rather than an assumed one.
Short answer: an inclusion is a hard, brittle, non-load-bearing speck inside a ductile matrix, so under cyclic stress it becomes a stress raiser and a crack-initiation site, and under corrosion it becomes a pit nucleation point.
Fatigue cracks almost always start at a surface or a near-surface defect, and an inclusion is a near-perfect defect. The particle does not stretch with the matrix, so a microscopic void opens at its edge on the first load cycles, and that void grows into a crack. In turbine discs, shafting, springs and fasteners — components whose lives are defined by fatigue — a single large inclusion at a critical location can cut the service life dramatically. Aerospace alloys such as those described in our Inconel 718 guide and our Waspaloy guide are rated precisely on how clean the starting stock is.
Corrosion tells the same story in a different way. An inclusion sitting at the surface breaks the passive film and gives a pit somewhere to begin, and a pit is the starting point for both crevice-related attack and corrosion-fatigue, the coupled mechanism our corrosion-fatigue guide covers. Our PREN and crevice corrosion guide shows how a clean surface defends the alloy's corrosion rating, which an inclusion undermines from within.
This is why aerospace and turbine grades are bought on cleanliness specifications rather than on chemistry alone, and why a buyer of any cyclic-loaded part should treat inclusions as a design input, not a metallurgical footnote. The cost of one early fatigue crack in a disc or shaft routinely exceeds the entire premium of specifying and verifying a clean heat at the outset.
Short answer: inclusions are quantified by microscopic comparison against standard charts — most commonly ASTM E45 — which grades oxide, sulfide and silicate populations by type and severity.
| Method | What it measures | Output | Buyer use |
|---|---|---|---|
| ASTM E45 | Oxide, sulfide, silicate and globular inclusions by chart comparison | Severity ratings A, B, C, D per field | Set a maximum rating in the PO |
| ASTM E112 grain size | Supporting grain structure context | Grain size number | Pair with cleanliness for fatigue parts |
| Macroetch (ASTM E340 / E381) | Centerline, porosity, segregation at low magnification | Macro cleanliness rating | Reveals cluster zones a micro method misses |
| Ultrasonic inclusion scan | Large isolated inclusions as reflectors | Indication count and size | Screens long, critical forgings |
| SEM / EDS | Chemistry and type of individual particles | Particle identity | Root-causes a failure or a high rating |
The rating methods are complementary rather than interchangeable. A chart count catches the distributed population; a macroetch catches clustered bands; an ultrasonic scan catches the rare very large particle that a small sampled area would miss. For a fatigue-critical part you generally want more than one, because each sees a different scale of defect.
Short answer: require a documented inclusion rating on the actual heat, sample more than one location, and for critical parts add a macroetch and an ultrasonic scan rather than trusting a single tensile coupon.
| Check | What to require | Method | What a clean heat shows |
|---|---|---|---|
| Inclusion rating | ASTM E45 rating on the delivered heat | Microscopic chart count | Ratings at or below the PO limit |
| Multi-location sampling | Rating from more than one position | Several prepared fields | No hotspot above the limit |
| Macroetch | Centerline and band check on a transverse slice | Acid etch, low-power view | Even structure, no clustered band |
| Ultrasonic scan | Large-inclusion screen on long parts | UT with recorded gates | No isolated large reflector |
| Melt route disclosure | Named vacuum double-melt route | Certificate and route statement | VIM+ESR or VIM+VAR declared |
The trap is sampling. A single transverse coupon can be drawn from the cleanest part of a billet and miss a segregated core entirely, which is why a one-coupon cleanliness claim is worth little for a rotating or cyclic-loaded component. The same principle that makes a narrow or poorly gated inspection hide internal defects also applies to inclusion screening, so specify multiple sampling locations and a macroetch rather than rely on one slice. For a long or large component a recorded ultrasonic screen is the only practical way to catch a rare very large particle that a small prepared field would never contain.
Short answer: ASTM E45 gives you the rating language, and the melt-route and product standards (AMS, ASTM, ISO) let you require a vacuum double-melt route that makes a low inclusion count achievable in the first place.
| Document | Covers | Buyer action |
|---|---|---|
| ASTM E45 | Inclusion rating by type and severity | Name a maximum rating, e.g. thin series limits |
| ASTM E340 / E381 | Macroetch macro-cleanliness | Require a macroetch on critical forgings |
| AMS melt-quality specs | Vacuum double-melt for aerospace grades | Cite the AMS class for disc/shaft stock |
| ASTM / ISO product specs | Material form and condition | Choose a grade whose route supports cleanliness |
| EN 10204 3.1 | Heat-specific inspection certificate | Require heat-level, not generic, data |
Cleanliness is mostly won at the melt, not at inspection. A vacuum induction melt followed by electroslag or vacuum arc remelting removes the dissolved gases and reactive impurities that become inclusions, so specifying the route is the most reliable clause you can write. The rating methods then confirm what the route already delivered.
A practical note: the cheapest place to buy cleanliness is in the purchase order, not in the inspection report. Once a dirty heat has been forged, machined and delivered, the only remedy is rejection and replacement, with all the schedule and cost that implies. Writing the melt route and the rating limit up front steers the supply toward heats that are clean by construction, and the inspection then becomes confirmation rather than discovery.
Short answer: state a maximum inclusion rating by method, require multi-location sampling, mandate a macroetch and ultrasonic screen for critical parts, and require a named vacuum double-melt route on the certificate.
| Clause | What to require | Why it matters |
|---|---|---|
| Cleanliness limit | Maximum ASTM E45 rating by type | Puts a number on acceptable cleanliness |
| Sampling plan | More than one location, stated in the PO | Stops a single clean coupon hiding a dirty core |
| Macroetch | Transverse macroetch on forgings and bars | Reveals clustered and centerline bands |
| Ultrasonic screen | Recorded UT gate for large inclusions | Catches rare very large particles |
| Melt route | VIM+ESR or VIM+VAR named on cert | Makes low inclusion count achievable |
| Heat-level data | EN 10204 3.1 with heat-specific results | Prevents a generic cleanliness claim |
For parts in sour or cryogenic service the cleanliness requirement is even tighter, because a pit or a crack starter is exactly what those environments exploit. Tie the clause to our sour-gas selection guide for downstream oil and gas, and to our cryogenic and LNG guide where a low-temperature toughness loss from a crack starter is unacceptable. Weldments are not exempt either; our welding guide notes that a dirty base can seed defects in the heat-affected zone.
Put these into every order for rotating, cyclic-loaded, corrosion-critical or safety-class nickel alloy parts:
Usually not. A tensile coupon samples a tiny volume and is dominated by the matrix, so a heat with a high inclusion population can still pass composition and tensile limits. Cleanliness has to be measured separately with rated microscopy, a macroetch and, for critical parts, an ultrasonic screen, none of which a tensile test substitutes for.
No. Porosity and shrinkage are gas or solidification voids in the metal itself, while inclusions are separate non-metallic particles embedded in the matrix. They have different causes and different fixes, but all three are caught more reliably by a macroetch and an ultrasonic scan than by a single coupon.
It makes a low inclusion count achievable and is the single most reliable control, but it is not a substitute for measurement. Slag entrainment, contaminated consumables or a broken process can still introduce particles, so the route should be paired with a rated cleanliness check on the delivered heat.
More than one location, drawn to represent the cross-section the component is cut from, with a macroetch to expose clustered bands and an ultrasonic scan to catch isolated large particles. A single transverse coupon is not enough, because inclusions are not uniformly distributed and the critical one can sit where you did not look.
Yes, though less dramatically than for fatigue. An inclusion at the surface is a pit nucleation site and a break in the passive film, which is why our PREN and crevice corrosion guide treats surface cleanliness as part of the corrosion defence. For components where a single pit can start crevice or corrosion-fatigue, the rating still earns its place in the order.
You can try, but you should not. Without a named method such as ASTM E45 the supplier can choose the most lenient scheme, so always state the method, the inclusion type and the maximum severity, and require the result on a heat-specific certificate so the number is tied to your actual lot.
Reject or quarantine it against the purchase order, retain the sample and certificate, and request a replacement heat with heat-specific cleanliness data. Because the defect is distributed through the metal, a high rating is not something that can be machined or repaired away, and the safe move is to source a verified-clean heat.
If you specify rotating, cyclic-loaded or corrosion-critical nickel alloy components and want stock whose cleanliness is measured rather than assumed, send us the grade, standard, form, dimensions and service environment. We will confirm the right vacuum double-melt route, rate the delivered heat by ASTM E45 with multi-location and macroetch checks where the part demands it, and supply prime-virgin, fully traceable stock from our Shanghai plant. Response time is typically ten minutes, and we ship worldwide.
Email: serena@hangboalloy.com
Phone: +86 166 2109 2936
WhatsApp: +86 166 2109 2936
Send us the grade, standard, form, dimensions and service environment. We will confirm the right vacuum double-melt route, rate the delivered heat by ASTM E45 with multi-location and macroetch checks where needed, and quote prime-virgin, fully traceable stock from our own plant.