Updated: Dec 18, 2020
Written by: Dr. Conor McCaughey, Metallurgist, Institute of Spring Technology
In the aerospace sector, like a lot of other industries, springs are everywhere. They’re in the seats and luggage bins of commercial planes, the control panels of jet fighters and in the antenna of satellites. Although there are hundreds of different spring shapes and designs flying above our heads every day, they all have something in common; they can fail. Although a lot of springs might not be considered “safety critical” in the aerospace sector, failure, especially systemic failures, are not acceptable. Would you feel safe on a plane if there were broken components in the cabin?
This article will look at an extension spring, manufactured from stainless steel wire, which failed during fatigue testing. The spring was part of a safety critical system used in the control panel of a plane. The end user sent several springs for fatigue testing, half of which were failing the fatigue requirements. The failures were all located in the same position, within the hook of the spring (Figure 1). Multiple failed samples were supplied to us to gather evidence from, along with an unfractured sample for comparison.
Our standard failure analysis consists of 4 main stages; visual examination, optical metallography, hardness test and scanning electron microscope and EDX analysis. Figure 2 shows images taken during the visual examination. Image A shows the fracture face from one of the fractured springs, showing a typical fatigue failure. Initiating from the inside hook position (region of highest stress in the hook) the fatigue crack propagated roughly halfway through the wire concluding in a final ductile overload. Examination of the fracture faces noted multiple suspected initiations coalescing into one propagation. Image B shows the surface of the wire at the inside hook position. This shows that the surface of the wire was damaged during operation, most likely due to wear. Closer examination of the wire surface also showed a suspected longitudinal defect, away from the flattened region on one of the fractured springs. However, damage was also noted at the inside hook position on the spring which did not fail, indicating that there could be another reason for the springs not reaching the designed fatigue life.
Figure 2: Showing optical images taken of the fracture face and surrounding wire of one of the fractured springs
Microstructural and hardness evaluation of the springs were as expected for this grade of material (ASTM A313 grade 631) and showed no significant microstructural or hardness variations between the broken and unbroken samples. It was therefore concluded that there was no issues with the structure of the material. However, during the microstructural examination surface defects were seen throughout the circumference of the wire.
The majority of stainless steel grades do not have any limits on the depth of surface defects, beyond being visually free from pits, seams etc.. Therefore although defects were seen throughout the circumference none of them were deemed to be beyond what is expected for this grade of material.
Finally, the failed springs were imaged in a SEM, some of the images taken can be seen in Figure 3. Image A shows the initiation of one of the failures. This was in the flattened region on the inside hook position, as expected based on the information gathered so far. However, although longitudinal striations, typical of wear, can be seen in the flat, the initiation points are located where the damage on the surface seems more comparable to the defects seen throughout the surface of the wire. This was also the case for one of the other failures examined, but not for the last failure. This fracture face showed that an initiation took place away from the flattened inside hook position at a longitudinal defect, seen in the visual examination. These observations indicate that the failure was not ultimately due to the wear and flattening of the inside hook position, but due to surface defects inherent to the wire. EDX analysis showed no evidence that corrosion had been a factor in the failure.
Figure 3: SEM images of two of the fracture faces examined.
With the evidence compiled in the report it would be easy to conclude that the failure was caused by surface defects, accelerated by wear, and leave it at that. This, however, isn’t very helpful for our customer. As we have already stated, there are no strict requirements for the level of surface defects allowed in this grade of stainless steel. It leaves the customer knowing the problem but not knowing how to fix it. This is where it pays to have an expert in a component perform the failure analysis.
We were able to make several recommendations to the customer as to how this could be remedied. Although there are no limits for surface defects within the standard for this grade, these can be requested by the customer. Our knowledge of dynamic grades allowed us to give a reasonable recommendation for the depth of surface defects. This should increase the fatigue life, but also be achievable by a wire manufacturer.
Beyond this, we were also able to inform the customer that although the predictions for the fatigue life of the spring were in excess of what was required, this is only for the body of the spring, not the hooks. The hooks of extension springs are under different stresses to the body of the spring. Bending and torsional stresses in the hooks compared to purely torsional stresses in the body. Depending on the design of the spring and hooks the stresses in the hooks can exceed those on the spring, reducing the fatigue life of the component. We are able to calculate the stresses in the hooks and give a predicted fatigue life based on experimental data produced in our lab. This would allow us to look at the design of the hooks and possibly suggest redesigns if the surrounding fixtures allow.
Finally, we also recommended shot peening the springs. Shot peening will impart beneficial residual stresses into the surface of the spring as well as reducing the defects seen on the surface. Although shot peening is not usually done on extension springs, as the benefit to the body of the spring is usually minimal. In this case where the failure is occurring in the hook, shot peening should increase the fatigue life.
Providing a thorough and accurate service not only determines why this failure occurred, but gives reassurance to the customer that future failures can be eliminated. Giving our customer the confidence to put these springs into aircrafts where hundreds of lives could be reliant on a spring not failing prematurely.