Twiflex clutch repair and refurbishment

Background

Bartech received a failed clutch returned from an offshore platform and it was immediately taken to the Quartzelec workshop in Aberdeen to investigate the reasons for failure.

The intention was that work would be supported by the Quartzelec quality control and support engineer. The clutch was dismantled into two sections.

This revealed the extent of the damage to the clutch shoes. The Ferodo lining had been totally worn away and had started rubbing into the aluminium shoes themselves.

At this stage it became clear that rectification work would not be possible locally, so the assembly was returned to Bartech workshops in Colchester for a more comprehensive inspection and repair.

Once the unit was inspected and using locally sourced vendors, the assembly was sent to an aluminium welder to allow the shoes to be dressed and welded so the shoes could be rebuilt up to the desired thickness of lining.

Outcome

We have also sent a note to the operator to make sure the final bolts were refitted into the back of the unit with thread locking compound in readiness for installation.

Recommendations:
The run-up procedure should be followed to make sure the clutch shoes bed in correctly. If this procedure is not followed closely the clutch linings may become ineffective and not fit for use

Additional information:

The clutch shoes were aluminium sprayed to fill the damaged areas and were then over-sprayed. They were then ground back to the original profile, (the material used was SX17-aluminium coating).

It was not possible to carry out a rotational test in the timeframe of the Project so the spring tension was set as per the OEM guidelines to a figure of 5.2 mm, this gives an equivalent lift-off speed of the shoe relating to 348 RPM with an engagement/contact speed of 550 RPM.

The springs in the clutch coupling were not changed, unfortunately, there was insufficient time within the Project to have new springs made.

We completed a pressure test and all the springs retained the same strength at a test pressure of 10bar and even at this pressure all the springs would not go coil bound, and we were happy that the retained strength was more than enough to engage and disengage.