Hydrogen Embrittlement Cracking in a Batch of Steel Forgings

A.K. Das, Aircraft Design Bureau, Hindustan Aeronautics Ltd.

From: Handbook of Case Histories in Failure Analysis, Vol 2, K.A. Esakul, Ed., ASM International, 1992

Abstract: The repeated occurrence of random cracks in the fillet radius portion of low-alloy steel (38KhA) end frame forgings following heat treatment was investigated. Microstructural analyses were carried out on both the failed part and disks of the rolled bar from which the part was made. Subsurface cracks were found to be zigzag and discontinuous as well as intergranular in nature. A mixed mode of fracture involving ductile and brittle flat facets was observed. Micropores and rod-shaped manganese sulfide inclusions were also noted. The material had a hydrogen content of 22 ppm, and cracking was attributed to hydrogen embrittlement. Measurement of hydrogen content in the raw material prior to fabrication was recommended. Careful control of acid pickling procedures for descaling of the hot-rolled bars was also deemed necessary.

Keywords: Inclusions; Subsurface cracks

Material: 38KhA (Chromium alloy steel)

Failure types: Hydrogen damage and embrittlement; Surface treatment related failures


The repeated occurrence of random cracks in the fillet radius portion of low-alloy steel and frame forgings following heat treatment was investigated.


A batch of 120 end frame parts forged and machined from a 200 mm (8 in.) diam medium-carbon chromium steel rolled bar was heat treated in accordance with specification requirements. These parts are used for critical applications in helicopters, and thus utmost care is necessary to ensure quality from the melting stage onward.

Circumstances leading to failure

Magnetic-particles inspection of the initial batch of 50 heat-treated parts after cadmium plating revealed several tiny radial cracks (Fig. 1).
Figure 1

Fig. 1  Heat-treated end frame component after cadmium plating. Several tiny cracklike indications, primarily in the base fillet radial zones, were detected by magnetic-particle testing.

Pertinent specifications

Russian alloy 38KhA was sued for the fabrication of the parts.

Sample selection

One cracked sample was selected for low-power optical and high-power scanning electron microscopic (SEM) examination. Two 20 mm (0.8 in.) thick disks sliced from the rolled bar stock used for forging were also examined.

Visual Examination of General Physical Features

Examination of the parts that had been inspected by magnetic-particle testing showed several tiny radial cracks at random locations, but primarily confined to the base edge and fillet radius zones (Fig. 1).

Testing Procedure and Results

Surface examination

Scanning Electron Microscopy/Fractography. One of the tiny cracks was carefully opened. Because the depth of the crack was negligible, it was difficult to retain the fracture zone intact. The fracture zone was examined by SEM, which showed the predominantly dimple structure of a ductile failure. No clue as to the nature and type of cracking could be established.
It was suspected that the hairline cracks could be due to hydrogen embrittlement. Two disks were sliced from the rolled bar, one in the annealed (as-supplied) condition and the other in the heat-treated condition. A 50 mm (2 in.) diam hole similar to the one in the finished end frame part was bored in the center of each disk. Magnetic-particle-inspection of both disks revealed tiny discontinuous subsurface cracklike indications in the bore surface (Fig. 2). One of the subsurface cracks was carefully opened so that the very small zone of cracking could be retained undamaged for SEM examination. The fracture clearly exhibited dimpled structures associated with brittle (cleavage) flat facets, ductile hairline cracks, micropores (Fig. 3), and rod-shaped manganese sulfide inclusions (Fig. 4). All of these fracture characteristics are typical of hydrogen embrittlement.
Figure 2

Fig. 2  Curved path of the fatigue crack.

Figure 3

Fig. 3  High-magnification SEM micrograph of the fracture zone of a subsurface crack in the bore, showing brittle flat facets, ductile dimples, micropores, and hairline cracks-all indicative of hydrogen embrittlement

Figure 4

Fig. 4  High-magnification SEM micrograph of the same zone shown in Fig. 3. Rod-shaped manganese sulfide inclusions are visible, normally a preferred she for hydrogen accumulation.


Microstructural Analysis. Microexamination of sections taken across the defect indications in both disks revealed subsurface discontinuous hairline (zigzag) cracks (Fig. 5), which were intergranular in nature with no evidence of branching. The microstructure of the defective part at higher magnification also showed discontinuous intergranular cracks, with typical manganese sulfide inclusions along the path of cracking (Fig. 6).
Figure 5

Fig. 5  Optical micrograph of a polished, unetched section across a crack indication shown in Fig. 2, showing the clear presence of subsurface cracks

Figure 6

Fig. 6  High-magnification optical micrograph showing intergranular cracks with characteristic manganese sulfide and oxide inclusions in the crack path of a section across the radial cracks in the heat-treated part

Chemical analysis

The chemical composition of the end frame material was found to conform to specifications (Table 1). The cracked end frame part was also subjected to instrumental gas analysis; the hydrogen gas content was found to be on the order of 22 ppm.

Table 1   Chemical specifications for 38KhA steel

Composition, %

Mechanical properties

Impact Toughness. Charpy impact properties determined in locations both near and remote from the defective zone in the longitudinal and transverse directions were generally low—44 and 34J, or 32 and 25 ft .lbf, respectively—compared with the specification requirement of 49J (36 ft . lbf).


SEM fractography clearly established that the random occurrence of tiny cracks was caused by hydrogen embrittlement. This finding was also supported by the presence of a high hydrogen content in the material prior to heat treatment.
Initially, from the nature and location of the cracks (mostly in the fillet radius zones), faulty heat treatment was thought responsible. However, when subsequent heat treatment of the second batch resulted in identical cracks in the same location (despite satisfactory chemistry, microstructure, and hardness), the random cracking was strongly suspected to be associated with a hydrogen-induced phenomenon. Because the tiny cracks in the part could not be successfully opened, a sub-surface crack in the heat-treated disk was forced open. The fracture was examined by SEM, which revealed characteristic hydrogen embrittlement features (Fig. 3 and 4).
A 50 mm (2 in.) diam hole had been bored into the disk to develop residual stresses sufficient to cause subsurface cracking. Under high stresses, hydrogen gas pressure increases rapidly, causing flakes and fissures. A gas content on the order of 22 ppm present in the raw material was high enough to cause hydrogen embrittlement in high-strength steel without any external stresses. The low impact properties measured on the failed part also indicated the brittle nature of the material caused by hydrogen embrittlement.

Conclusion and Recommendations

Most probable cause

The basic cause of the development of tiny cracks was hydrogen embrittlement.

Remedial action

In the absence of records related to the manufacturing history of the supplied raw material, hydrogen absorption was suspected to have occurred during improper pickling of the hot-rolled bars. Careful control of the acid pickling operation within specified process parameters was recommended. Measurement of hydrogen content in the raw material prior to fabrication was also suggested.

Related Information

Fatigue Failures, Failure Analysis and Prevention, Vol 11, ASM Handbook, ASM International, 2002, p 700–727