By Craig J. Schroeder

The failure of parts or assemblies can affect the delivery of goods, result in costly repairs and down time, and jeopardize the safety of people. Whenever critical parts fail unexpectedly, it is a clear sign that something has gone wrong. They may have been used beyond their intended useful life. They may have been abused. There may be a flaw in their design or in the materials of which the parts are made.

It is the job of metallurgical failure analysis—the minute examination of materials—to determine the root cause of failure. Knowing why something failed is the key to preventing future failures.

The following is an example of a field failure that resulted in costly down time for the customer of the manufactured part. The case study will be used to demonstrate the metallurgical failure analysis process.
The manufacturer of a hitch was encountering a high rate of cracks in hitch arms in use. The company, which had used our services in the past, called on Stork Technimet to analyze the part.

In this case study, a hitch arm that was removed from a hitch was received for metallurgical failure analysis. Numerous other hitches, which were manufactured in a similar manner but were not submitted, also reportedly had cracks. The specified material of the hitch arm was Grade 1045 steel.

Visual Inspection

The first step in most failure investigations is to perform a visual inspection, to examine the part for clues that can help determine why it failed. The types of clues that the investigator will look for include cracks, general damage, and residues from corrosion products or other sources of contamination. If the primary fracture is readily visible, the fracture origin region is determined and documented, when possible. The broken hitch arm sent as a sample is shown in Figure 1.

The fracture origin region of the arm is shown in Figure 2. The fracture origin region is determined by studying the texture of the fracture surface. Flow lines on the fracture surface often indicate the direction of crack progression. In this case, the flow lines in the texture of the fracture surface indicated that the fracture origin region of the arm was at the bottom portion of the part in Figures 1 and 2. A curved region with a matte gray coloration was present at the bottom edge of the part. Locations 1A and 1B were at the ends of the curved region and were therefore selected for closer inspection. Location 1C, in the rust-colored region with a rough texture and Location 1D, in a bright gray region with a smoother texture, were selected for closer inspection and for comparison to Locations 1A and 1B. Linear features are visible under the blue paint. The linear features are indicative of a laser or plasma cut surface. The red line in Figure 2 indicates Section M that was selected through the origin region for metallographic analysis.

Microscopy

After visual inspection of the fracture surface, it is common to examine the fracture origin region with a low-power binocular microscope or a scanning electron microscope, or both. The magnified views of features help determine the mode of fracture. In most cases, confirmation of the fracture mode can only be achieved via scanning electron microscopy.

In this case, the fracture surface of the arm was cleaned in a mild alkaline detergent under ultrasonic agitation in order to remove loose foreign deposits. The fracture was then examined with a scanning electron microscope. The fracture surface of the arm at Location 1A, as shown in Figure 2, was examined.

A scanning electron micrograph showing the fracture at Location 1A is presented in Figure 3. The texture has a “rock candy” appearance that is indicative of brittle intergranular cracking. This means that the crack progressed along the grain boundaries of the metal through this region.

A scanning electron micrograph showing the fracture at Location 1B is presented in Figure 4. The fracture morphology changed to transgranular cleavage-type cracking in this region.

The fracture surface of the arm was studied at Locations 1C and 1D. The fracture morphology at Locations 1C and 1D consisted of cleavage-type cracking, similar in comparison to Location 1B.

Metallography

Metallographic analysis of the failed part is an important step in the metallurgical failure analysis process. Metallographic analysis is used to determine if defects or deleterious microstructural features are present in the metal. Special attention is often given to the origin region of the failure of the part.

In this case, a metallographic cross section, identified as Section M for the purposes of this study, was prepared through an area of the fracture through Location 1A, as indicated by the red line in Figure 2.

The microstructure at the origin region is presented in Figure 5, where the fracture surface is oriented toward the bottom, and the fracture origin region at Location 1A of the arm is at the left side of the image. A heat affected zone was present at Location 1A. It was judged that the heat affected zone was the result of a weld in this region. A dark etching layer of tempered martensite was also present along the laser or plasma cut surface, as shown on the top left of Figure 5. Locations 1 through 4 were selected for microhardness testing.

A bright etching region of martensite needles and retained austenite was present in the fracture origin region, as shown in Figure 6. The presence of retained austenite is indicative of carbon enrichment that, in this case, was judged to occur during the cutting process. Grain boundary bainite was also present in the origin region. Bainite is a transformation product that results from the transformation of austenite at temperatures below the pearlite range but above the martensite start temperature.

The microstructure near Location 1C consisted of pearlite and ferrite, as shown in Figure 7. The presence of pearlite and ferrite is indicative of a normalized condition. The transgranular cleavage fracture through the ferrite and pearlite is evident. The microstructure at Location 1C indicates that the part was originally normalized, but the microstructure at 1A was transformed due to heating from the weld and the cutting process.

The microstructure along the cut edge above the origin region was further inspected. A light-etching phase was present along the cut surface, as shown on the left side of Figure 5.

A higher magnification view of the left center of Figure 5 at Location 2 is presented in Figure 8, where continuous grain boundary carbides and retained austenite are evident. The presence of continuous grain boundary carbides and retained austenite indicates that carbon diffused into the surface during the cutting process. Grain boundary carbides provide brittle paths for cracks to follow and explain, in part, the intergranular morphology of the crack at the fracture origin region.

Chemical Analysis

Another common step in the metallurgical analysis of a failed component is the determination of the base metal chemical composition to determine whether the specified material was used in the manufacture of the part. The chemical composition of the hitch arm was determined via optical emission spectroscopy, and the results are presented in Table 1. Chemical analysis of the section revealed that it did not meet the minimum carbon requirement of SAE J403 Grade No. 1045 steel. Although the carbon level was lower than specified, it was judged in this case not to have contributed significantly to the failure of the part.

Table 1 — Chemical Analysis Results
(Percent by Weight)

Element	Hitch Arm	SAE J403 Grade No. 1045 Requirements
Carbon	0.40	0.43 - 0.50
Manganese	0.69	0.60 - 0.90
Phosphorus	0.013	0.030 max.
Sulfur	0.003	0.050 max.
Silicon	0.18	N.S.
Chromium	0.07	N.S.
Nickel	0.09	N.S.
Molybdenum	0.02	N.S.
Aluminum	0.019	N.S.
Copper	0.26	N.S.
Vanadium	0.002	N.S.
Titanium	0.002	N.S.

Analysis completed using optical emission spectroscopy (CS-5, 04-07).
N.S. = Not Specified.

Hardness Test Results

Hardness testing is typically performed in a failure investigation to determine whether the specified or expected hardness of the material has been met. Deviations in hardness are often indicative of improper material processing. The hardness of the arm was tested in accordance with ASTM E 10-07 and was 187 HBW, which is typical for the normalized condition of Grade 1045 steel.

The microhardness of the arm was tested at the heat affected zone and Locations 1 through 4, as shown in Figure 5. The results are presented in Table 2. The microhardness of the weld heat affected zone, along with Locations 1 and 2, were the hardest, averaging 56 to 58 HRC equivalent. This indicates these regions were hard and brittle due to thermal alteration of the microstructure from the welding and cutting processes. The brown and black regions at Locations 3 and 4 were 49 HRC equivalent, which is typical of high carbon tempered martensite. The base metal was 92 HRB, which is very soft in comparison to the other locations and is typical of normalized Grade 1045 steel.

Table 2 — Microhardness Test Results

Location	Average Microhardness, HK500	Approximate Equivalent Rockwell Hardness, HRC (base metal HRB)*
Weld HAZ	653	56
1	696	58
2	646	56
3	526	49
4	478	49
Base metal	210	92

Tested in accordance with ASTM E 384-07.
* Per Tables 1 and 2 of ASTM E 140-07.

Tensile and Charpy Impact Test Results

Determination of the mechanical properties of the metal can play an important role in the failure analysis of a part. Mechanical testing can help determine the inherent properties of the metal for comparison to the expected or specified properties of the part. The fracture of the tensile specimens and impact specimens can also be compared to the field fracture surface to help verify the type of fracture the part experienced in the field.

Tensile testing was performed on a specimen excised from the hitch arm in the orientation shown in Figure 1. Results of the test are shown in Table 3. The mechanical properties of the base metal were generally typical for normalized Grade 1045 steel except for the yield strength which was slightly low. The slightly low yield strength was judged not to be significant to the failure of the part.

Table 3 — Tensile Test Results

Property	Arm Tensile Test Results	Typical Grade 1045 Steel — Normalized Condition
Yield strength, 0.2% offset, psi	51,600	60,000
Tensile strength, psi	95,500	95,000
Elongation, %	22	23
Reduction in area, %	44	45

Tested in accordance with ASTM A 370-08a.

Three Charpy impact tests were performed on specimens excised from the arm in the orientation shown in Figure 1. Results of the tests are shown in Table 4. The toughness of the base metal was low for Grade 1045 steel with a hardness of 187 HBW. The expected toughness of Grade 1045 steel at 187 HBW should be in excess of 20 ft-lbf. The results indicate that the metal is inherently brittle. The reasons for the low toughness of the metal are not evident from this study.

Table 4 — Impact Test Results

Specimen No.	Energy, ft-lbf
1	9
2	14
3	15
Average	13
Test Temperature, °F	68
Specimen Orientation	Longitudinal

Tested in accordance with ASTM A 370-08a and E 23-07a.

CONCLUSIONS

• By following the typical steps in the metallurgical failure analysis process, it was determined that the hitch arm failed due to deleterious microstructures that formed along the cut surface and at the heat affected zone of a weld. This led to a brittle condition on the surface and made the metal prone to cracking.

• Metallographic inspection of the hitch arm at the laser or plasma cut surface revealed a mixture of tempered martensite with isolated areas of continuous grain boundary carbides and retained austenite. The presence of grain boundary carbides and retained austenite indicated that carbon diffused into the surface of the metal during the cutting process. The grain boundary carbides are hard and brittle in comparison to the tempered martensite and provided a brittle path for crack propagation.

• The fracture morphology at the crack origin was characteristic of brittle intergranular cracking while fracture through the base metal consisted of cleavage morphology which is indicative of fast, brittle fracture. The transition in fracture morphology coincided with the change in microstructure.

• Impact testing of the base metal revealed lower than typical toughness values for the specified material at the measured hardness. The hard, brittle surface layer along with the poor toughness of the base metal made the part susceptible to brittle overload failure.

Craig J. Schroeder, P.E., is senior metallurgical engineer at Stork Technimet in New Berlin, Wis. (www.storksmt.com).