Extending Rotor life

WHEN DO ROTORS NEED TO BE REMOVED FOR MAINTENANCE OR REPLACED DUE TO DESIGN LIFE LIMITATIONS? AND WHAT ARE THE PRIMARY SOURCES OF LIFE LIMITATIONS?

As the gas turbine industry continues to move towards increasingly advanced and efficient turbines, the costs of component maintenance and replacement continue to be a crucial factor for economic viability. Most strategic planning tends to be around the refurbishment and replacement of flow path components. However, one component that can be critical to long-term financial success is the rotor itself.

Over the last couple of decades there has been a growing push to apply life limitations to entire rotors or critical rotor components (discs and wheels). These life limitations are based on either original design life targets or specific design life analysis performed by the OEM.

However, in many cases, operators have been known to continue operation of units with rotor lives well beyond the predicted lives without issues. This begs the questions: What are the primary sources of life limitations on rotors? And when should they be removed for maintenance or replaced?

Generally, there are two types of factors that can limit the life of a rotor. They can be most easily classified as those which can be seen and those which cannot be seen.

Included within the first category are all of the detectable damage modes that can typically be identified in a major rotor inspection. These include issues like corrosion and erosion, dirt build up and cooling air blockages, severe rubbing, cracks, and signs of overheating (including changes in material hardness).

Items which cannot be seen are design life limiting factors related to low cycle fatigue (LCF) and creep which result in slow but steady localized metallurgical degradation and may not be detectable until the component is near failure.

Detectable damage

Most detectable damage that a rotor may incur can be seen during a hot gas path or major inspection. Typically, the highest stress locations on the rotor tend to be within the disc serrations and disc seal arms (where applicable). Visual inspections of the rotor disc serrations can reveal several different modes of degradation.

Dirt or deposits can often accumulate in open spaces between the blade (bucket) roots and the disc roots on the unloaded sides of the serrations. These clearance gaps are meant to allow cooling air flows across the roots. Deposits clogging these locations reduce cooling flow and can have a detrimental effect on disc life.

Corrosion, erosion and wear most likely result from long periods of operation on turning gear. The clearance within the serrations can allow the airfoils to move and rattle during very low speed (turning gear) operation. This type of damage may lead to severe pitting and reduced rotor life.

Non-destructive evaluation (NDE) inspections of the disc roots can discover cracks in either the high stress zones of the root serrations or around cooling air supply holes (Figure 1).

Figure 1: NDE Evaluation of a turbine disc[/caption]

Cracks found within the serrations are probably due to either high stresses over time (creep), or high cyclic stresses. These will generally require disc or rotor replacement. Cracks around the cooling air supply holes are likely due to LCF and will be related to engine starts. Blending repairs of these cracks may be possible. However, a full engineering evaluation of the localized loads and stress patterns should be conducted to ensure that the resulting modification would be safe to run for a practical service interval.

Disassembly of the rotor may be required in some cases in order to deal with vibration or assembly issues. The cost associated with this will depend on the style of turbine in question.

However, if the rotor is disassembled, additional inspections of the interfaces between the discs as well as the interior end of cooling air supply holes can also be checked for wear or defect indications. Disassembly of the rotor will provide for the opportunity to correct assembly issues associated with vibration. However, due to the often prohibitive costs associated with this work, the decision may be made to avoid this inspection when possible.

Rotor lifing

A big question for many users arises when they appear to have only modest or limited damage as they approach the OEM defined end of life on their rotors. They wonder what would be the risk of continued operation.

The cost of replacing a rotor, or even a disc, can be prohibitive on an older engine. Life limitations provided by OEMs are usually conservative since they must be able to encompass their full fleet of turbines.

Some of these units can be assumed to be pushing the limits for numbers and types of starts, operation with wheel space and disc cavity temperatures at or above alarms, or a number of other worst-case scenarios. A rotor lifing analysis specific to their units and their mode of operation should provide enough data to make a sensible decision.

The key to life analysis is access to accurate operational history for the units. This information should encompass:

• General operational history. This should include items such as the number of hours of operation as well as the number of full starts and stops. However, improved results could be achieved if this information includes breakdowns on how the machine was operated (full power vs. part loads, number of trips from load, and so forth.)
• All available data on external or measured cooling air flow. This should includecooling air return temperatures where air coolers are in use. Cooling air flows andtemperatures are a critical aspect of component life expectancy.
• Information on startup and shutdown rates as well as spin-cool cycles. This should include ramp rates which can have a large impact on thermal gradients within components and will affect their LCF life.

A rotor life analysis should encompass all aspects of the rotor. This, in turn, demands an understanding of the design and operation of the components and loads around them. For a turbine wheel and disc, this will include thermal and mechanical loads from the airfoils, all airflows and temperature loads, contact loads with the mating components of the rotor as well as the nature and magnitude of the loads transferred through it.

These various inputs form the boundary conditions under which the analysis will be based. The formation of these conditions calls for several analyses of their own. These would include development of an aero-thermal model of the turbine gas path.

Modeling the gas path

This study effectively models the full gas path of the turbine. It can calculate the gas path temperatures, pressures, and velocities through each stage of the turbine. This, in turn, provides information on the thermal and mechanical loads seen on the airfoils as well as power outputs and torque loads applied by upstream and downstream components.

Additional elements of analysis are airfoil thermal and mechanical modeling. The gas path data from the aero-thermal model is used to develop the external thermal and mechanical loads on the airfoil. For cooled components, a full heat transfer evaluation will need to be performed on the airfoil. The primary thermal load on the disc is from heat conducted from airfoils to the blade roots.

Accurate modeling of the airfoil thermal loads is critical to the development of accurate rotor lifing results. Uncooled airfoils are more easily modeled since the primary heat balance is limited to external heating of the gas path surfaces of the airfoil and platform along with the cooling of the blade below the platform surface.

But perhaps the most critical aspect of the analysis is the development of an accurate cooling airflow network. The cooling airflow network provides the input information necessary for the thermal model of the rotor. This allows for accurate calculations on temperatures and temperature gradients. Both are critical in determining creep and LCF lives.

Figure 2: FEA results showing creep

strain distribution on a turbine disc root[/caption]

FEA (finite element analysis) of the rotor and disc will normally include the blades as part of the model (Figures 2, 3 and 4). This allows for mechanical loads and thermal loads to transfer accurately between the components. Load ramps can be based on the actual load data supplied by the user.

Figure 3: FEA temperature distribution analysis of disc and blades[/caption]

Similarly, any information provided on part-load operation and shutdown procedures will also be used in the development of thermal mechanical loads through transients. Thermal loads and pressure distribution loads calculated from the gas path model are applied to the FEA model of the blade.

Figure 4: FEA results showing

temperature distribution on the root serrations of a disc at full power[/caption]

Similarly, calculated thermal loads and pressures developed from the airflow network study are applied to all of the surfaces of the disc as well as the blade root and blade shank. Calculated contact loads are applied to any surfaces in contact with other components.

Analyses of the components are performed to calculate transient stresses and temperatures. Maximum and minimum stress levels are determined and used to determine LCF life of the components at all critical locations.

Startup, shutdown conditions

This type of analysis may need to be repeated for various types of startup and shutdown conditions including potential trips from load which may result in shortened LCF lives. In addition to these transient conditions, stresses and temperatures will also be calculated for operation in steady state conditions at load.

This information will form the basis of the creep life calculations. In both the LCF and creep calculations, extensive non-linear analysis will likely be a critical component in the development of accurate component lifing results.

This process may seem complex. However, this is based on methodology that is routinely applied to determine disc and rotor life for modern industrial and aircraft turbine engines. Thorough analytical component life modeling can provide a means of accurately determining the true life of rotor components.

Written by: Dan Stankiewicz P.Eng. is Turbine Design Engineering Manager for Liburdi Turbine Services. For more information visit: www.liburdi.com or call 905-689-0734 ext.543