Azimuth Thruster Bolt failures
Vessel was docked for first Intermediate docking. Although not required by Manufacturer an inspection of the gear case was requested by the Technical Superintendent. During this a number of broken bolts where found with the majority in the Stbd Unit.
This document pertains to an investigation by the author through inspection and discussion with attending OEM Technical. Its purpose is to provide initial feedback to allow confidence in continued operations. It will be subject to revision on receipt of investigation reports from OEM.
Thrusters have an indicated 19000 Running hours, of these it is estimated that 25% was in transit.
Without any precursor warning of failure,( Oil Analysis, Vibration etc), a number of broken bolts were found in the Bevel gear mounting arrangement on both Azimuth Thrusters.
It is likely failure was not catastrophic but occurred in a limited number of stages.
The trigger for these staged events was not continuous overload or other mal-operation but rather environmental forces for which the Thruster would have been expected capable of resisting.
The probable cause of the failure was poor design, poor assembly practices allied to use of Bolts of inconsistent quality.
It is believed the used of the spacer ring and elongated bolts of better quality will resolve this and allow for safe operations.
It is recommended that trending systems be installed to allow understanding of the relationship between environmental factors and Thruster Load transients.
It is recommended that the Thruster is opened for inspection at next docking period. During this an assessment shall be made of the bolts and where practical one or more removed for NDT inspection.
The Drive assembly consists of a Spiral Bevel gear running on rolling element bearings.
The Gearwheel is bolted to the main Shaft through M36 Socket Head Set Bolts. The Wheel is centralised radially with a spigot on the flange. Other than requirements to reasonably align clearance bolt holes in flange with threaded holes in Gearwheel there are no circumferential alignment mechanisms such as radial dowels.
The Pinion Gear is integral with Input shaft.
Drive is via the clamping force provided by the bolts at the mating face of the Gearwheel and Flange. The Bolts should not see any shear forces.
Before disassembly Mesh patterns where taken which appeared to be normal with Line of action apparently moving through a mid section pitch circle (Cone) with profiling limiting this to around 50% of tooth height and 75% of tooth length at no load.
The teeth have an involute type profile. There is no well defined tip relief although Mesh patterns probably demonstrates profiling.
The Pinion appears to have more addendum albeit sufficient deddendum to prevent root narrowing.
The Sets are handed. It is not determined if the Rotational Shear evident in the broken bolt fracture faces was also handed
There was no evidential damage to the gear teeth visually or detected as a result of Magnaflux NDT
There is visual evidence that at least one washer enter the mesh on at least one occasion.
A number of teeth, in a limited part of the circumference, on the Crown wheel where damaged on the outer edge most likely due to impact with one or more of the broken bolts. There is no evidence that a bolt entered the mesh
There were minor repeated patterns on the teeth running faces suggests ingestion of minor parts which subsequently where discharged.
There are 20 off M36 12.9 Hexagon Socket Head Allen Set Bolts in each installation. Of a total threaded length of 110mm approximately 30 mm is within the threaded portion of the Gear wheel with the remaining in the Clearance hole portion of the Shaft Flange.
At least 6 Full thread turns are engaged when the bolt is secured. This would suggest around a 2:1 ratio of material in the Thread against material in the core under tensile load suggesting that this is adequate.
The Threads are locked in position using Loctite 275, no evidence is noted of antiseize compounds so it is assumed the bolts where Torqued using ‘Dry’ values.
The bolt markings appear to indicate more than one supplier may have been used .
It is noted that all the broken bolts have the same markings suggesting they are from the same source, FKE which appears to be a Taiwanese supplier.
SCOB, the maker of a number of the intact bolts appear to be an Italian Supplier. VSP, the manufacturer of the replacement bolts is also an Italian supplier.
All recovered failed bolts appear to have failed at the mating face between the Crown Wheel and Shaft Flange.
When hammer testing remaining bolts one off failed just below the Head. This was again of the same manufacturer as with the other broken bolts.
Noticeable on this bolt was that the outer radius chamfer was noticeably greater than the average at around r=3 compared to rmean=1 reducing substantially the area of the loaded face of the head.
The shear face differed from the other broken bolts as there was less material tearing
Bolt failure Mode
Note: Contrast and sharpness have been enhanced
There is little evidence of crack propagation in any of the failed bolts. It is not likely this was the result on under-torqueing leading to fatigue through underusing the yield strength.
In each case the material appears to have a highly crystalline surface with evidence of shearing of material in a clockwise rotation looking in to the cap end of the broken surface.
There is no apparent fretting damage or other material loss due to rub. There is no apparent galling in threads on either bolt or Gearwheel threaded holes. The evident misalignment of the Gearwheel to Shaft flange may have led to excessive stressing of the bolt should contact have occurred although there is no direct evidence for this other than the visible misalignment scars of the washers.
This misalignment was also apparent when looking into the clearance holes of the flange to the threaded holes in the Gearwheel. The flange hole encroached onto the root diameter of the threads in the Gearwheel but in the cases viewed, did not appear to be more than this
There is no apparent necking of the material iwo break, it is not likely this was the result of over torqueing in the assembly phase although the twisting shear which appears evident is typically caused by this.
There was a contrasting appearance between groups of bolts, some with a highly polished surface and other with what appeared to be fresh breaks. It is possible that this is the result of where the bolt lay in the Thruster Pod after failure but more likely due to the bolts failing separately some time apart
It is likely that bolt failure mode was Impact shear although the non flat nature of the breaks in certain cases means this cannot be said with any degree of confidence. It is possible lack of homogeneity led to the raised shears and material tearing. The true failure mode will be revealed following the metallurgical analysis carried out by OEM should they disclose this.
It is not likely that the failure is due to continuous overload, the fact the unit continued to operate after losing 6 or 20 bolts demonstrates that the degree of clamping force was sufficient. Additionally there was no observed evidence of fretting between the mating faces.
Bolts where installed with a Plain washer supplied by Peiner, there is no additional information.
The use of Set bolts and what appears to be a generous clearance hole on the Washer means that it can be displaced axially to allow 2mm and above of difference in the landing area for the Bolt
Where the gearwheel has not been properly aligned with the clearance holes on the Shaft flange, this is possible and indeed likely as there are no alignment components, any deleterious effect of the displaced washer would be exacerbated. These effects may include a bending moment on the Bolt.
Bearings removed where in good condition on the running surfaces with no evidence that foreign materials and passed through
Fretting was evident on the outer landing areas on a number of bearings. The degree of this is significant and special attention should be given to this in subsequent disassembly’s.
Damaged Bevel Gear Teeth
The damaged edges of the Gear teeth will be ground back to a smooth profile. As there was no other detected damage the original sets will be re-used.
Relief is given to the full height of the tooth
16 of the 38 teeth have been ground to similar extent to that photographed with a number of others more lightly dressed
Bolts will be replaced with one of elongated design.
The use of elongated bolts reduces the bolt stresses by increasing the bend radius for the same perpendicular movement (simple beam bending)
The Bolt has a quality appearance with the landing area of the socket head flat without the variable radius of the broken ones.
A new supplier was used for the new bolts
A spacer ring of weight estimated at 50Kg is to be manufactured by Wartsila. Although a complete ring, rather than more traditional cylindrical spacers has the negative effect of increasing the rotational inertia and thereby increasing the effects of any transient loading, the removal of windage effects is perceived as being more beneficial