Waspaloy-Class GH4698: Why a ‘U’-Shaped Die Beat an ‘H’ by 7.4× on Stress-Rupture Life

Forging & Heat Treat · Field Note from FUSHUN METAL

Waspaloy-Class GH4698: Why a ‘U’-Shaped Die Beat an ‘H’ by 7.4× on Stress-Rupture Life

A Chinese Waspaloy-class nickel-superalloy disk forging, in a recently published trial, took 376 hours to rupture at 650 °C / 705 MPa after a ‘U’-shaped die was used. The same bar, the same heat, the same chemistry, formed in an ‘H’-shaped die, ruptured in 51 hours. Same alloy. Same heat-treat. 7.4× difference in stress-rupture life — decided by the shape of the forging. That is the number this note is built around, and it has direct consequences for anyone buying or specifying GH4698 / ЭИ698 (EI698) disks, rings, or shaft components for 750–800 °C service.

1. Grade cross-reference: GH4698 and its international equivalents

The trial under discussion uses the Chinese GB/T 14992 designation GH4698 (old name GH698, master grade code H46980). On paper the grade is registered as the Chinese equivalent of the Russian ÉÌ698 (EI 698 / EP 698), developed in the 1950s. In the international wrought-disk family, it sits between Inconel 718 and Waspaloy on temperature capability — closer to Waspaloy in working envelope, but without the cobalt bill.

The chemical composition reported on the trial bar (Table 1 in the source) is the practical fingerprint buyers and metallurgists should match. The lot used in the forging trial came in well inside spec on every controlled element:

The combination of high Cr (oxidation resistance), high Mo (solid-solution strengthening), and Al + Ti + Nb (joint γ′ / γ″ precipitation) is the “fingerprint” that separates Waspaloy-class alloys from 718-class alloys. Nb at 2.08 % also stabilises a fine γ′ population at 750 °C, which is the single biggest reason GH4698 sits 100 °C above 718 on the long-term ceiling.

2. The result that should not be possible

The published data come from Cao Yuru and Cao Yigao at Forging & Stamping (2023, No. 19). The trial material is a φ300 mm VIM + VAR bar of GH4698. The bar passed incoming inspection on chemistry, macrostructure, grain size (ASTM 3), ultrasonic testing (φ3.2 mm flat-bottom hole, no excess reflections), and the room-temperature, 750 °C tensile, and 650 °C / 720 MPa stress-rupture properties required by GJB 3165.

Two forging die shapes were designed for the same finished-machined disk: an ‘H’ shape and a ‘U’ shape. The H-die used 360 kg of billet, the U-die 340 kg. Both forgings went through the same three-stage heat treatment: 1120 °C × 8 h AC, 1000 °C × 4 h AC, 700 °C × 16 h AC. Both were ultrasonic-tested before and after heat treatment, both passed the same φ1.2 mm FBH acceptance.

The mechanical-property delta between the two forgings, on the same heat, is the point of the paper:

Two things stand out. First, the U-die forging beats the bar on every room-temperature metric — strength is up ~2 %, but ductility (A) is up 25 % and reduction of area (Z) is up 46 %. Second, the stress-rupture result is on a different scale: the bar passes the ≥ 24 h spec at 720 MPa, the U-die forging runs 376.42 h at a slightly lower stress (705 MPa). Even after normalising for the small stress difference, the U-die forging is in a different life regime.

3. What changed: die shape, not chemistry

The chemistry is identical, the heat is identical, the forging press is the same 400 MN hydraulic press, the heat-treat furnace is the same. The only two variables are the forging-die cross-section (‘H’ vs ‘U’) and the punch profile. A DEFORM-based process simulation was run before any steel was touched, and it predicted the result with uncomfortable accuracy.

The line that decides the outcome is the effective-strain row. In an H-die forging, large volumes of the finished-machined zone see strains as low as 0.375 — below the 0.6 threshold for full dynamic recrystallisation in GH4698. In a U-die forging, almost the entire machined zone sits between 0.625 and 1.0, well inside the recrystallisation window. The U-die punch is also a large-draft-angle (B) profile rather than a spherical (A) profile, which avoids the part wrapping the punch on cooling.

4. From bar to disk: the 7.4× number, decomposed

Three things have to be true at once for a Waspaloy-class disk to give this kind of stress-rupture life:

1. Every part of the machined zone has to be fully recrystallised. Below ~0.6 effective strain, the original cast / VIM-VAR billet structure is not broken up, the grain size is uneven, and the ductility / rupture life of the finished forging is dominated by the weakest unrecrystallised pocket. The H-die forging has pockets sitting at 0.375 effective strain — the U-die forging does not.
2. The grain-flow lines have to follow the principal stress direction in service. A Waspaloy-class disk in service sees centrifugal + thermal stresses that are tangent to the disk axis. The H-die produces swirled flow with vortex regions; the U-die produces flow that follows the flash direction, which is closer to the tangent stress direction. The U-die forging therefore has fewer flow-line discontinuities where a crack can open under cyclic loading.
3. The forging temperature window has to be hit, every cycle. The published process window is heating to 1120 °C, transfer ≤ 45 s, forging speed 5 mm/s, finish-forge ≥ 1000 °C. A U-die distributes the strain more uniformly, so the part does not locally overheat or chill; the post-forge temperature field is uniform, and the standard three-stage heat treat (1120 °C × 8 h AC / 1000 °C × 4 h AC / 700 °C × 16 h AC) actually delivers the solution-and-aging microstructure it was designed to deliver.

The mechanical numbers track that. The 25 % jump in room-temperature elongation and the 46 % jump in reduction of area between the bar and the U-die forging are not normal forging deltas — they are the signature of a fully recrystallised, fine-grain microstructure. The 7.4× jump in stress-rupture life is the high-temperature signature of the same microstructural change, plus the absence of vortex flow lines.

5. The three knobs that made it work

For a buyer or a process engineer, the published recipe can be reduced to three decisions, each of which is independent and each of which can be verified:

The process parameters that did not change but had to be controlled are: 400 MN hydraulic press, 1120 °C furnace set-point, ≤ 45 s transfer, 5 mm/s ram speed, ≥ 1000 °C finish-forge temperature, three-stage heat treatment (1120 / 1000 / 700 °C, all air-cool). These are the published operating window for GH4698 / ЭИ698. They were not the variables; they were the constants.

6. What this means for disk forgings

For a Waspaloy-class disk buyer — aero-engine turbine or compressor disk, marine gas-turbine disk, large industrial disk for 700–800 °C service — the published data argue for three changes to the usual procurement conversation:

1. Ask for the die-shape study, not just the chemistry. A chemistry certificate and a heat-treat certificate are necessary but not sufficient. Ask the forge shop for the DEFORM simulation that supports their die geometry, the macro-etched flow-line map, and the corner-strain map on the production forging. If the simulation does not put the machined zone above 0.6 effective strain, the forging will not behave like a Waspaloy disk regardless of what the chemistry says.
2. Treat stress-rupture as the design variable. The published bar passed ≥ 24 h at 720 MPa, which is the spec — and the bar would have been a perfectly acceptable delivery on that basis. The U-die forging runs 376.42 h at 705 MPa. That 7.4× difference is the difference between a disk rated for one maintenance interval and one rated for several. If the design is rupture-limited, ask for the stress-rupture data on a forged sample, not on the bar.
3. Standardise on U-shaped (or equivalent) geometry where the part allows it. The U-die forging is also 20 kg lighter on a 360 kg forging, which is a 5.5 % material saving on a high-nickel, high-molybdenum disk. For a Waspaloy-class disk at 2026 nickel prices, the material saving on a single disk is in the same order of magnitude as the forging cost itself.

A Waspaloy-class disk is not a bar that happens to be round. The die shape, the metal flow, and the strain distribution in the forged shape are the variables that decide what the chemistry and heat-treat are actually worth. Specifiers who buy on chemistry + heat-treat alone are buying a 51 h disk. Specifiers who buy on chemistry + heat-treat + die-shape + flow-line evidence are buying a 376 h disk. The data are now in the public domain; the procurement spec just has to catch up.

7. The point of view

In a Waspaloy-class disk, the alloy is the easy part. The hard part is the forging. A die that flows the metal in one direction and distributes 0.6–1.0 effective strain through the machined zone will give seven times the stress-rupture life of a die that swirls the metal and leaves pockets below 0.6. That is what the published GH4698 trial shows, and it is what the procurement spec should be measuring.

Source: Cao Yuru, Cao Yigao. Forming-process study of “U”-shape die forging of GH4698 superalloy. Forging & Stamping, 2023, (19). (Original Chinese; forging trial at unspecified Chinese aerospace forge, 400 MN hydraulic press.)