A Rudder Stock Failure Investigation

Introduction

One Eighty was appointed by a marine engineering company to investigate the failure of a rudder stock which has been reported while the vessel was at sea. The rudder stock fractured between the stock lower part and rudder upper casting section as seen in Figure 1. The rudder stock was reported to have been eight years old at the time of the incident. A section of the stock lower part shaft was provided by the marine engineering company for further analysis to determine the mechanism of failure.

Methodology

One Eighty has a very thorough and holistic approach to root cause failure analysis. In this case, our investigation involved a number of elements:

• Conducting a thorough analysis on the rudder stock lower part section and determine the most appropriate methodology to identify the root cause of failure.
• The rudder stock provided to One Eighty was used to conduct metallurgical analyses by way of
  • spectrographic analysis
  • microstructural examination
  • macro-observations
  • visual observations of the fracture surface
  • hardness testing.
• Documents provided by the marine engineering firm, including drawings and images of the failure, were to be used as the basis of all calculation parameters. Metallurgical analysis and calculated results, along with the age, condition and operational procedures will provide the evidence required to identify the mechanism of failure for the rudder stock shaft.

Investigation and results

Visual observations

Visual observations of the fracture surface shows that the failure mechanism was due to fatigue. The fracture features indicate that the fatigue was uniaxial with low nominal bend stresses as was observed within the fatigue pattern. The fatigue initiated on the edge of the shaft on the O-ring groove located near the sleeve edge. The addition of the groove would consequently create a stress raiser on the shaft. Features of the groove, such as sharp corners, would exacerbate the stress concentrator. Additionally, the shaft was observed to be significantly corroded as is typical of forged steel in a marine environment.

Technical drawings analysis

One Eighty conducted a thorough analysis of the drawings of the rudder stock construction (shown below), as well as the general part lists and notes.

Drawings of the rudder stock construction (Figure 5) show the layout of the ship’s rudder stock profile where the failure occurred. The failure occurred on the shaft stock lower part, above the rudder upper casting. The shaft has a slight taper between the interface of the upper casting and shaft section where the fracture occurred as seen in the drawings below.

The design of the rudder stock would enable the shaft to achieve sufficient alignment. Various design features such as shaft tapering, dual rudder horn mount fixtures (stock and pintle), multi-layered bush assemblies and steering gear connection would contribute to correct alignment during the vessel’s construction.

Spectrographic analysis

Spectrographic analysis was conducted on both the shaft and sleeve material. The results from the analysis confirmed that the material’s chemical composition was acceptable – as designated in the construction drawings. The shaft and sleeve were identified to be a LR grade forged steel and 316L stainless-steel, respectively.

The electrochemical phenomenon known as galvanic (bimetallic) corrosion occurs when two dissimilar metals are within physical or electrically bridged contact with one another. The difference in SCE voltage between the two different materials determines which metal acts as the cathode or cathode, with the difference value determining the magnitude of the corrosion rate. The difference between the forged steel and stainless steel allows for a galvanic reaction to take place, with the forged steel section acting as the sacrificial anode in this case. The rate of corrosion would be slow due to the voltage difference being low yet will remain active over a prolonged period. The use of sacrificial zinc anodes is typically used to accommodate the effect of galvanic corrosion by protecting other susceptible metals.

Optical observations

Macro-examination and microstructural analysis conducted on the sleeve and shaft of the rudder stock showed that the material was homogenous and annealed. Light microscopy images on the cross-section near the fracture surface showed corrosion to have affected the shaft significantly. Corrosion would allow for the shaft surface to deteriorate and create additional stress raisers. The sleeve was observed to have been negligibly affected by corrosion.

Hardness testing

Vickers hardness results indicate that the material is within specification, with no variation of hardness detected through the cross-section of both the sleeve and shaft material.

Theoretical calculations

The endurance limit of a shaft is typical calculated to determine the minimum strength required to facilitate fatigue propagation. Endurance limit strength of the rudder stock shaft design near the location of failure was calculated to be 58 MPa. This value is significantly low for a shaft of this size due to the addition of the groove with small radii corners.

Conclusion

In conclusion, the failure of the rudder stock shaft was the result of fatigue. Fatigue was initiated by the addition of an O-ring groove which functioned as a stress raiser within the shaft, consequently lowering the endurance limit strength of the design. Additionally, the use of a stainless-steel sleeve near the groove resulted in slow galvanic corrosion that wore away at the shaft which would further reduce the endurance limit strength of the material. The wear of the material near the groove would have likely initiated the crack.

The period until failure occurred would have been slow and insidious due to the low nominal stresses seen on the fracture surface fatigue pattern. The addition of low galvanic reaction between the stainless-steel and forged steel would have exacerbated the problem by hiding any flaws beneath surface corrosion. Visual inspection would not have been able to detect the flaw before failure occurred.