LP Turbine Report_Redacted.pdf

| 2016 dockside | March 22, 2016

TS Kennedy 2016 Dockside

LP TURBINE SUPPORT STRUTS

On the starboard side, in the lower casing, both the upper and lower struts have indications, with the upper strut indication through and through. Upper:

Lower:

At the bottom of the lower casing, the centerline strut is very difficult to see but it clearly appears to have an indication 1/3 of the way down from top.

The Upper Casing struts appear not to be in as bad shape, with the starboard side having visible indications.

Closest to flange

Upper case, top, center, from stbd

: From port with indication?

Upper case port flange and port upper:

Upper case starboard upper with indication at base:

History

This DeLaval Turbine was manufactured approximately in 1966. It was made for the Velma Lykes, of

Lykes Brothers Steamship Company. The ship was a C4-S-66a Cargo Vessel, built by Avondale

Shipyard. The turbine serial number is 653479-90. The casing and these struts are cast, not welded.

It is a very rigid turbine, in that, many struts were installed by DeLaval at that time in their turbines and the casings themselves were strong and heavy.

MarAd and looked through the available boxes of vessel related correspondence. The only history we were able to find was limited. In 1991, during Desert Shield, the LP Turbine had a failure.

The ship was brought to a shipyard in Lisbon, Portugal where the rotor was removed, new diaphragms were purchased, and new bearing installed. There was mention of rubbing on the turbine blades. There was no mention of the struts in that particular documentation. In 2002, at Bender

Shipyard during the conversion to a school training ship, the strut cracking problem was mentioned and plans to repair them mentioned, as well as plans to remove some of them. We only know that some work was completed, but no details are available yet.

Approach

Welding turbine struts is an approach that must be made with great consideration. Improper repair can be useless, or cause more problems. There are two welding methods that can be applied; stress relieved and non-stress relieved. Stress relieved welding offers the best potential for a long repair life, but is very complicated and time-consuming, and requires post heat of @ 1000 degrees F. (Exact temperature depends on the alloy, so even higher temperatures are possible). Multiple cycles may be necessary. Non-stress relieved welds have lower life expectancy, with the repair short lived. A preheat of over 500 degrees F is required, based on material.

In either approach, NDE inspection is highly recommended at a minimum. Radiograph inspections may be warranted if repair is the choice, as that process may reveal the full extent of the cracking. To perform any of the repairs, the upper turbine casing will have to be moved out of the way to allow the rotor to be lifted clear of the casing, which in turn, will allow the diaphragms closest to the repairs to be removed. As you can imagine, this is a large, time-consuming task.

Cracks typically occur at the steam inlet areas, where transient thermal stresses can exceed the yield point of the cast material. In more modern turbines, the OEM’s employ computer modeling and advanced alloys to reduce the likelihood of cracking. These LP Turbine struts do not have a lot of mass relative to the casing itself, so the thermal stresses will be evident more rapidly than with the casing. In addition, any turbine that has multiple start-up and secure cycles is more prone to thermal stresses and stress related problems than a turbine that is constantly running.

There are mechanical repairs that can be applied to cracks, such as stitches, but they must be properly designed to accommodate all the stresses and loading that comes into play. Usually, these types of repairs have to be designed through finite element analysis and can be time-consuming and expensive.

A strong argument can be made that in reality, the LP Turbine has created a condition itself that relieved unwanted stresses. The struts, due to their design, aren’t able to move very much, even if indications appear to be through the entire length. The readings taken by MITech show the turbine, which just came back from an extended operating period, has no signs of rubbing. All clearances are within tolerances.