An electroslag-remelted (ESR) CrNiMoV steel cut its average non-metallic inclusion size from 20 μm to 10 μm and the maximum inclusion from 50 μm to 28 μm relative to a vacuum-degassed reference, with ductility up 10%, contraction up 18%, and notched toughness at 20 °C up 46% [S1]. For tool and die buyers, the practical question is whether that cleanliness gain justifies the roughly $250–$300 per ton premium ESR carries over conventional electric-furnace melting [S2].
The decision turns on the failure mode you are designing against. ESR is a secondary remelting route, not a grade, so two bars labeled the same AISI or DIN number can behave very differently in fatigue, polishability, or cross-grain toughness depending on whether they were air-melted, vacuum degassed (VD), vacuum induction melted (VIM), or ESR-refined [S3].
What ESR Actually Changes in the Inclusion Population
ESR pulls a consumable electrode through a molten, electrically heated slag bath held in a water-cooled copper mold, so every droplet of metal reacts with the slag before it enters the solidifying pool [S3].
The inclusion chemistry also shifts, not just the size distribution. In a 2025 tool-steel study using two different fused-slag systems, slag choice controlled the population of Mg- and Al-based oxide inclusions, while CaS-type sulfides appeared in every ESR sample regardless of slag [S4]. That matters for corrosion-sensitive tooling: the same study linked slag-derived inclusion populations to localized corrosion behaviour in ambient exposure, meaning the slag system is a design variable, not an internal mill detail.
Sulfur removal is part of the package as well, because the reactive slag absorbs sulfide inclusions during droplet transit, and ESR steel is widely cited for longitudinal-transverse isotropy because directional solidification suppresses the fiber-flow anisotropy that plagues large conventional ingots [S2].
ESR vs Air-Melt vs VAR: Decision Criteria Side by Side
For specification work, the useful comparison is not "ESR good, air-melt bad"; it is which route solves which problem. On a property-gain basis, ESR and vacuum arc remelting (VAR) both lift steel performance by roughly 20% over an equivalent air-melt, primarily through lower inclusion counts and fewer subsurface crack starters [S5]. On route selection, the geographic convention is that VAR is the typical Type 1 aerospace route in North America while ESR is the typical Type 2 route in Europe, and the practical difference is that VAR controls purity by excluding the atmosphere in a vacuum chamber, whereas ESR controls purity by tuning the slag chemistry [S5].
On inclusion chemistry, ESR and VIM are often paired: VIM strips dissolved gases during primary melting, then ESR removes sulfur and refines inclusions in the secondary step, which is the typical aerospace, defense, and medical-implant route [S7]. On cost, air-melt is the cheapest baseline, ESR sits roughly $250–$300 per ton above electric-furnace conventional, and a specialty air-melt such as Aubert & Duval XD16N can actually beat a remelted grade on corrosion and fatigue for non-flight-critical parts, so the premium is not always justified [S2][S5]. On ingot soundness, ESR's bottom-up directional solidification in a water-cooled mold suppresses shrinkage cavities and macrosegregation, which is the central defect mode in large conventional ingots [S2][S8].
A condensed decision frame, drawn only from the cited sources:
Criterion 1, inclusion size: air-melt worst, VD intermediate (20 μm avg / 50 μm max reference), ESR best (10 μm avg / 28 μm max) [S1]. Criterion 2, atmosphere control: VAR excludes it, ESR manages it via slag, air-melt is fully exposed [S5]. Criterion 3, premium over electric-furnace: air-melt baseline, ESR about +$250–$300/ton, specialty air-melts can undercut both on cost for non-critical use [S2][S5]. Criterion 4, isotropy in forged bar: ESR ingots give consistent longitudinal and transverse properties after hot working, while conventional cast ingots show fiber-flow anisotropy [S2].
Where ESR Pays Off, and Where It Is Overkill
ESR is the right call when the failure mode is inclusion-driven. Polished plastic-mold and optical-mold surfaces are a textbook case, because non-metallic inclusions at the cavity face pull out during EDM or polishing and leave pinholes and pitting, and ESR's cleaner inclusion population is what enables a defect-free mirror finish [S2]. For the same reason, cold-work die steels, bearing steels, and high-stress fatigue components benefit from the lower inclusion count and tighter size distribution [S3][S5].
ESR also fits when the drawing, AMS, or customer standard literally requires it. Aerospace, defense, and medical-implant specifications often call out Type 1 VAR or Type 2 ESR by name, in which case the route is non-negotiable regardless of cost, and substitution requires written engineering approval [S3][S5][S7]. The same logic applies to any component where sub-surface crack initiation, rolling-contact fatigue, or transverse toughness is a design constraint, because those are the properties that move the most with inclusion control [S1][S5].
ESR is overkill, or at least non-obvious, for non-critical wear parts, general-purpose structural tooling, and any application where the part is dominated by bulk hardness rather than inclusion-initiated failure. A specialty air-melt such as XD16N, designed for corrosion and fatigue resistance in non-flight-critical use, is the standard counter-example: it can out-perform a remelted grade on those two metrics at lower cost [S5]. If the failure mode is not inclusion-driven, the $250–$300 per ton buys margin the part will not use.
Process Limits and Failure Modes Engineers Should Watch
ESR is a refining step, not a magic step. The consumable electrode's chemistry, cleanliness, dimensions, and surface condition carry through, so a bad electrode makes a bad ESR ingot; the process reduces and modifies inclusions rather than literally removing every one [S3]. Process control also matters at the droplet stage: melt rate, electrode immersion depth, slag temperature, and electrical input all influence the molten pool shape and the resulting solidification structure, which is why ESR is computer-controlled and can take hours to days per heat [S2][S5].
Slag selection is a real engineering variable, not a fixed internal choice. The 2025 tool-steel study showed that two different fused-slag systems produced measurably different populations of Mg- and Al-based oxide inclusions and different localized-corrosion behaviour in ambient exposure, with CaS sulfides present in every sample [S4]. For corrosion-sensitive tooling, that means the ESR designation alone is not enough; the slag system, or the mill's published inclusion rating, is part of the spec.
ESR also has variant atmospheres that change the cost-benefit case. Pressurized, inert-gas, and vacuum-ESR variants exist for alloys sensitive to nitrogen or hydrogen pickup, and they push the cost above standard ESR but below the equivalent VAR-plus-VIM route in some alloys [S5][S7]. For shop-floor decisions, the relevant boundary is that ESR competes with VAR on premium cleanliness, with VD on mid-tier cleanliness, and with specialty air-melts on cost-sensitive wear applications; it does not replace the primary melt.
Specifying ESR Cleanliness in a Purchase Order
Engineers buying ESR tool steel should write the route, the inclusion rating, and the test method into the PO rather than relying on the word "ESR" alone. Common acceptance references are ASTM E45 (inclusion rating by microradiograph or field count) and the equivalent DIN/SEP methods for non-metallic inclusions, and aerospace work typically references AMS 2300/2301 cleanliness plus the specific Type 1 or Type 2 callout [S3][S5].
Useful numeric anchors for the PO line, all from the cited sources: ESR average inclusion 10 μm vs 20 μm for VD reference; ESR maximum inclusion 28 μm vs 50 μm for VD reference; ESR Charpy-V at 20 °C improvement +46% and at −40 °C +30% versus the same reference; ESR/VAR general property gain roughly +20% over equivalent air-melt; cost premium roughly $250–$300 per ton over conventional electric-furnace tool steel [S1][S2][S5]. For buyers comparing tool and die steel grades against an air-melt alternative, those numbers are the defensible reference points.
Trackable signals to watch over the next reporting cycle: published slag-system disclosures on ESR tool-steel mill data sheets, inclusion-rating revisions in AMS 2300/2301, and any new aerospace specification that pairs Type 2 ESR with explicit inclusion-size limits rather than leaving them to the mill's process control. Related procurement reading for adjacent decisions lives in our DBA vs rebar for welded stud connections spec-driven piece and in the pitless truck scale foundation choice guide for site-side steel decisions.
Component reference pages worth checking: air pick, and air impact wrench.