cyclic test are understood, and the safe life can be adjusted to account for the differences. If the test is terminated by burst or complete failure, then crack initiation for this particular test may be defined using the appropriate crack growth calculations and/or fracture surface observations. An alternate approach would be to use the number of cycles at the last clean inspection to define the crack initiation point. This appr oach requires an inspection technique with a level of detection capability consistent with that used by the engine industry for rotating parts.
(7) Damage Tolerance Assessments. The applicant should perform appropriate damage tolerance assessments to minimize the potential for failure from material, manufacturing, and service-induced anomalies within the approved life of the part. Service experience with gas turbine engines has demonstrated that material, manufacturing and service-induced anomalies do occur, and they can potentially degrade the structural integrity of the life-limited parts.
Historically, the life determination process has been founded on the assumption that life-limited parts do not contain anomalies. Consequently, this methodology has not explicitly addressed the occurrence of such anomalies, although some level of tolerance for anomalies is implicitly built into the methodology using design margins, factory and field inspections, etc. Damage tolerance assessments, however, explicitly address the anomalous condition(s) and complement the safe life approach. The intent of damage tolerance as sessments is to supplement the existing safe life methodology; they are not intended to allow life-limited parts to remain in service beyond the limits established by the safe life approach or to allow rotating parts to return to service with cracks.
(a) Type of anomalies.
1 Material anomalies. Material anomalies consist of abnormal discontinuities or non-homogeneities introduced during the production of the raw material or the melting process. Examples of material anomalies are: hard alpha in titanium; oxide/carbide (slag);
stringers in nickel alloys; and contaminant ceramic particulates in powder metallurgy materials unintentionally generated during powder manufacturing.
2 Manufacturing anomalies. Manufacturing anomalies include anomalies produced in the conversion of the ingot to billet and billet to forging steps, as well as anomalies generated by the metal removal and finishing processes used during manufacture and/or factory repair. Examples of conversion related anomalie s are forging laps and strain induced porosity.
Examples of metal removal related anomalies are: tears due to broaching; arc burns from various sources; and disturbed microstructure due to localized overheating of the machined surface or to inadvertent contact with a tool or fixture.
3 Service-induced anomalies. Applicants should consider service-induced anomalies such as non-repaired nicks, dings and scratches, and corrosion. Applicants should use 12
similarity of hardware design, installation, and exposure and maintenance practice to determine relevance of the experience.
(b) Probabilistic Damage Tolerance Risk Assessments (PDTRA). The probabilistic approach to damage tolerance assessment is one of two elements necessary to appropriately assess damage tolerance. The second element is service damage monitoring (see paragraph 8.d.(7)(e) of this AC). AC 33.14-1 in cludes an example of the probabilistic process that applies to hard alpha material anomalies in titanium alloy rotor components. The probabilistic damage tolerance risk assessment is fracture mechanics-based and typically includes the following primary elements:
1 Anomaly size and frequency distributions. A key input in the damage tolerance assessment is the size and rate of occurrence of anomalies. This information may be statistical in nature and may be presented in a form that plots the number of inclusions that exceed a particular size in a specified amount of material. Anomalies should be treated as sharp propagating cracks from the first stress cycle, unless there is sufficient data to indicate otherwise.
2 Anomaly growth analysis. This determines the number of cycles for a given anomaly to grow to a critical size. This prediction should be based on knowledge of part stress, temperature, geometry, stress gradient, anomaly orientation, and material properties. The analysis approach should be validated against relevant test data. In this context, anomaly growth may be based solely on crack propagation, or a combination of crack initiation (i.e. incubation) and crack propagation, depending on the nature of the anomaly/damage.
3 Inspection techniques and intervals. Manufacturing and in-service inspections are options available to reduce the risk of fracture from inherent and induced anomalies. The applicant should identify the intervals for each specified in-service inspection.
Historical engine removal rates and module and part availability data could serve as the basis for establishing an inspection interval. The manufacturing inspections assumed in the damage tolerance assessments should be incorporated into the manufacturing plan. Likewise, the assumed in-service inspection procedures and intervals should be integrated into the service management plan and included, as appropriate, in the ALS of the ICA.
4 Inspection Probability of Detection (POD). POD of the individual inspection processes, such as eddy-current, penetrant fluid or ultrasonic, used to detect potential anomalies should be based on the statistical review of sufficient quantities of relevant testing or experience. The relevance of the data should be based on the similarity of parameters such as:
Size, shape, orientation, location, and chemical or metallurgical character of the anomaly;
Surface condition and cleanliness of the parts;
Material being inspected, such as its composition, grain size, conductivity, and surface texture;
Variation of inspection materials or equipment such as the specific penetrant fluid and developer and equipment capability or condition;
Specific inspection process parameters, such as scan index; and 13
Inspector’s capabilities and limitations, such as visual acuity, attention span, and training.
5 Risk Prediction and Allowable Risk. The inputs are integrated using risk assessment software which predicts the relative probability of failure (POF) for each part. The predicted POF is compared to specific allowable design target risk values. Note the allowable DTRs can be found in advisory circulars which address specific materials and/or anomaly types.
Designs that satisfy the allowable DTR values will be considered in compliance with the damage tolerance requirements. Part manufacturers have a variety of options to reduce the POF and achieve the level of relative risk allowed by the DTR. These options include, but are not limited to:
Component redesign Material change Material process improvements Manufacturing process improvements Manufacturing inspection improvements Enhanced in-service inspections, and Life limit reduction.
(c) Appropriate Damage Tolerance Assessments 1 Interim Compliance Options. When an industry or company specific PDTRA approach has been established and accepted by the FAA, it may be used to meet the intent of the “appropriate damage tolerance assessments” of § 33.70. The FAA acknowledges that key elements of this approach, such as the anomaly size and frequency distributions, standardized analysis techniques, and the relative allowable risk targets are not available in every case. In these instances, a range of comp liance options may be considered such as:
A simplified conservative probabilistic approach.
A deterministic fracture mechanics approach.
Comparison to successful historical experience based on crack growth rate calculations.
Design margins based on crack growth rate calculations.
Fatigue testing of simulated damage.
Application of damage tolerance concepts.
2 Appropriate Use of Interim Compliance Options and Interim Surface Damage Tolerance Requirements. When new information is developed and approved that allows the application of the probabilistic approach to a particular disk/rotor feature or anomaly type, the interim guidance provided by paragraphs 8.d.(7)(c)1 and 8.d.(7)(d) of this AC do not apply.
3 Appropriate use of Probabilistic Risk Assessment Techniques. The use of probabilistic risk assessment techniques should not be considered an alternate approach to meet the intent of § 33.70. As required by the rule, lif e limits must be based on the minimum number 14
of flights required to initiate a crack as described in the previous paragraphs, and is normally the only acceptable means to determine rotating part life limits. The exception is when the risk of failure from an anomalous condition is higher than acceptable, and the life limit is lowered to control the risk of failure. In th is case, probabilistic risk assessment tools may be used to address continued airworthiness issues associated with part life shortfalls and certain part conditions that cannot be eliminated by the best available manufacturing processes and inspections techniques.
Probabilistic risk assessment tools may not be used to relax existing well-established industry best practices. In addition, use of probabili stic risk assessment tools to reject improved manufacturing and inspection techniques, or other changes of similar nature, will require careful consideration and may need to meet higher standards. Alternate approaches must be discussed with and approved by the FAA on a case-by-case basis.
(d) Interim Surface Damage Tolerance Requirements. Currently, the required input data (anomaly size and frequency distributions, etc.) have not been developed to fully implement the probabilistic approach in all cases. This information is not available for all the various anomaly types and all required rotor features/locations. When an appropriate probabilistic approach is not available, applicants should adhere to the following interim guidance, which reflects industry benchmarked design criteria that ensure a minimum level of damage tolerance is built in to new rotor designs. Rotating component surfaces must possess a residual crack growth life equal to at least 3,000 damage tolerance cycles or a number of damage tolerance cycles equal to 50% of the part certified life, whichever is less, assuming an initial surface crack of 0.030 inches long and 0.015 inches deep and/or a corner crack equal to
0.015 x 0.015 inches. The damage tolerance cycle us ed in this assessment is the major stresscycle (min-max-min) from the missions used in the LCF certification analysis for standard day conditions. These calculations shou ld be based on the use of linear elastic fracture with cracks placed in the most unfavorable orientation and location and may use compressive residual stresses and inelastic stresses. Applicants must identify any additional assumptions associated with the residual crack growth life calculations. These assumptions must be discussed with and approved by the FAA on a case-by-case basis.
(e) Service Damage Monitoring. The overall objective of Service Damage Monitoring is to review data obtained from field operation of the type design engine to determine if there are anomalous conditions that require corrective action. Appropriate actions may include assessment of the impact of damage observed on one part/location on other parts/locations.
Applicants should determine if surface damage that has been detected is consistent with the serviceable and repairable limits and to determine if additional actions are required to prevent failure and rectify any potential unsafe conditions which may be identified. Service damage monitoring consists of the following steps:
1 Determine the serviceable and repairable surface damage limits using a process approved by the FAA and summarized within the service management plan. Damage size limits should be a function of part, part location, and damage type. Damage should include, but is not limited to, nicks, dents, and scratches. The serviceable and repairable limits should be published in the ICA.
2 Establish a monitoring process to record damage that meets all of the following criteria; (a) is inconsistent with or exceeds the repairable limits, and (b) is made available to the type certificate holder or supplemental type certificate holder through existing reporting channels. Document the monitoring pro cess in the service management plan. This activity should record at a minimum the damage size, type, and location observed during service inspections for each life-limited part.
3 Assess damage meeting the criteria defined in paragraph 2 above. This assessment should consider the following, as a minimum:
The impact of the observed damage on the life of the damaged part.
The likelihood for recurrence of similar damage.
Whether the damaged part was found during a field inspection.
Whether the damage is likely to go undetected and escape to the field.
Recommended corrective actions to identify, prevent, or eliminate the source of the damage.
4 During the service life of the part, the damage information obtained by the damage monitoring process, as well as the corrective actions implemented, should be reported to the authorities during continued airworthiness discussions.
e. Establishment of Life Limits - Static Parts.
(1) This section applies to major static structural parts whose primary failure is likely to result in one of the hazardous engine effects listed in § 33.75. It does not apply to static parts whose failure is not likely to result in a hazardous engine effect.
(2) The general principles used to establish life limits for static parts are similar to those used for rotating parts. However, for pressure-loaded static parts, the point at which a part must be removed from service may differ from rotor parts. For rotor parts, life limits are based on the initiation of a crack. For pressure-loaded static parts, the approved life may be based on the crack initiation life, plus a portion of the residual crack growth life, providing the following provisions are met:
The crack growth analysis technique is experimentally verified.
The useable portion of residual crack growth life maintains a safe margin to failure.
Mandatory inspections are incorporated into the service management plan and the ICA if there is any dependence on crack detection.
The reliability of the crack detection technique has been verified.
All regulations must be met assuming the presence of the maximum predicted size crack that can occur within the approved life of the part. In some cases, it may be necessary to limit the crack size allowed in service in order to meet regulations other than § 33.70, such as the blade containment requirement in § 33.19.
(3) Other major structural static parts besides rotating and pressure-loaded static parts may be identified as life-limited parts. Under these circumstances, applicants should discuss the methodology for determining the life of the part with the FAA certification office which must approve its use. Applicants should use the genera l principles for rotating and static pressure loaded parts as a guide.
(4) The following table provides an overview of the various lifing elements contained within the rotating part and the static part methodologies:
Table 1. Overview of Lifing Elements.
Rotating Part Element Applicable to Static Parts Comments Flight Profiles Yes Engine Performance Yes Internal Flow Analysis Yes Heat Transfer Analysis Yes Structural Analysis Yes Flight maneuver loads required.
Crack Initiation Life Yes Large number of options available. Range from a cyclic test of a single component to a fully developed crack initiation model.
Damage Tolerance Assessment No See additional comments in paragraph 8.e.(7).
Service Life Certification Yes
(5) Life-limited static parts may be repaired and returned to service providing the life limit is re-established using methods equivalent to the methods used to establish the life of a new part.
(6) The cyclic loads that must be supported by static parts also differ from those of rotating parts. For rotating pa rts, the major loads usually involve centrifugal forces and temperature gradients, while the minor loads include applied pressure loads. For pressure-loaded static parts, the major loads usually involve internal pressure and flight maneuver loads as well as thermal gradients.
(7) Certain static part construction techniques, such as welding or casting, contain inherent anomalies. For these cases, the applicant should consider a fracture mechanics-based approach as part of the methodology to establish the life of the part. This approach should be based on an understanding of the type and size of the anomalies that will be present in the part.
In these instances, the applicant may base the life limit on the crack initiation life plus a portion of the residual crack growth life, providing a safe margin to failure/burst can be maintained. The applicant should experimentally validate the cycles required to initiate a crack from an anomaly 17
and the crack growth analysis techniques. If the anomaly is assumed initially to be a crack, then only the crack growth analysis techniques require experimental verification.
(8) In determining the life of the part, the applicant should consider the temperature of the part, the temperature gradients, and any significant vibratory loads or other loads (for example, flight maneuver loads), as well as the pressure loads.
(9) Manufacturing and service inspections are options available to reduce fracture potential. The applicant shoul d incorporate the manufacturing inspections that are deemed necessary into the manufacturing plan. If the approved life limit includes reliance on the detection of cracks, the reliability of the crack detection techniques must be considered. Any dependence upon crack detection must result in mandatory inspection techniques and intervals being included in the service management plan and in the ALS of the ICA. Engine removal rates and module and piece part availability data may serve as the basis for establishing the inspection interval.
f. Certification of the Approved Life.
(1) This element requires the applicant to submit the part life, based on approved techniques, to the FAA with sufficient supporting data. The applicant is responsible for supplying sufficient data to support the part life. The FAA may approve, if appropriate, an approved life that is less than the part life. A list of the life-limited parts and their approved life must be included in the engine shop manual in the ALS of the ICA as required by § 33.4. The approved life is the mandatory retirement life.
(2) The engine shop manual must contain an approved life for each life-limited part, even if the safe life and/or the approved life are extremely long and exceed the design lifetime of the aircraft. This will ensure part usage is recorded and tracked properly.
g. Maintaining the Approved Life.
(1) At certification, the approved life of the part is based on predicted engine operation, material behavior, and environment. After certification, the applicant may need to check the accuracy of these predictions, recognizing that many aspects of the lifing system may change during its life. For example, the e ngine’s usage and its operating environment may change after a change of ownership. The appli cant should use service feedback to confirm any assumptions made in the engineering plan remain valid or to determine if modifications are required. The engineering plan should describe no t only the basis of the part’s approved life but also those actions subsequent to certification needed to ensure the approved life is appropriate throughout the operational life of the engine.
(2) The applicant must regularly review the assumptions made when establishing the approved life. The engineering pl an should detail when these reviews should occur, and what information is necessary to complete the reviews. Aspects that should be considered include, but are not limited to:
The frequency of approved life reviews.
Detailed inspection of service exposed parts including retired parts.
Review of flight profile.
Findings during maintenance.
Engine development experience.
Lessons learned from other engine projects.
Any service events.
h. Influencing Parts. Engine life-limited parts ar e part of a complex system in which other engine parts can affect the life-limited parts, including their life capability. Therefore, the engineering plan must consider these other parts and particularly any changes to them.
Examples of influencing parts include a turbine blade, a mating flange or seal, and a static part that impacts the environment (temperatures, pressures, etc.) around a life-limited part. Examples of changes to these parts include a heavier blade; a new mating part with a different coefficient of thermal expansion; orifices and clearances affecting cooling air flow; and geometric or material changes to a static part that modify the thermal response and/or the mechanical response of the part.
9. GUIDANCE FOR DEFINING A MANUFACTURING PLAN.
a. Introduction. The manufacturing plan is a portion of the overall integrity process intended to ensure the life capability of the part. The engineering plan includes assumptions about how engine life-limited parts are designed, manufactured, operated, and maintained. Each can affect the part’s life. Therefore, it is essential for the applicant to ensure the attributes required by the engineering plan remain valid and are achieved during manufacture.
b. Elements of a Manufacturing Plan. The part specific manufacturing plan should consider the attributes of the part delivered by the manufacturing process, and should highlight the processing parameters that affect the life of the part. The plan should also identify the process parameters that should not be changed without proper verification and engineering approval. Many of the parameters may be included by reference to other documents (see paragraph 7.e. of this AC). The para meters may include, but are not limited to:
Material controls including zoned areas that require special properties.
Manufacturing method specifications.
Manufacturing process steps and sequence.
Cutting parameters and scatter allowed.
Inspection method and sensitivity.
Special part rough machining methods or finishing method(s).
Methods intended to improve fatigue capability or minimize induced anomalies.
Process validation to qualify the impact of the manufacturing method variation on the part’s life capability.
Compliance with microstructural requirements.
Surface finish.
Residual stress profile
Manufacturing controls to ensure that parts are produced by a consistent and repeatable process.
Traceability records for each part.
Review of non-conforming parts to ensure the deviation does not adversely impact the life of the part.
c. Development and Verification of the Manufacturing Plan.
(1) The manufacturing plan should be reviewed and verified by the key engineering and manufacturing personnel with the following technical skills:
Engineering (Design and Lifing).
Material Engineering.
Non-Destructive Inspection.
Quality Assurance.
Manufacturing Engineering (Development and Production).
(2) These personnel should evaluate and approve process validation, the rules for change control, non-conformance disposition, and corrective actions to ensure the manufacturing product is consistent with the design assumptions of the engineering plan. The intent is to ensure:
Manufacturing processes are developed and applied with the appropriate level of oversight to ensure the part life capability required in the engineering plan is consistently achieved.
Substantiation programs are agreed to up-front and executed as part of the process validation.
Changes to the manufacturing processes and practices are visible and are made with cross-functional review and approval.
Non-conformances are reviewed by the appropriate personnel with the required skill mix prior to disposition.
Corrective action is implemented for non-conformances that have been detected.
(3) The level of detail in the plan may vary depending on the specific process step being considered, the sensitivity of the particular process step, and the level of control required to achieve the required life capability.
(4) Example. A process specification exists to control the drilling of holes. If the use of this specification produces a hole that meets the life capability requirements for a flange bolt hole, then the plan may simply note that this flange bolt hole will be produced according to the standard specification. However, if a rim air hole requires cold expansion following drilling to meet the life requirements, then it may be necessary to reference the cold expansion process specification in addition to the hole drilling specification.
10. GUIDANCE FOR DEFINING A SERVICE MANAGEMENT PLAN.
a. Introduction. The service management plan is part of the process to maintain the integrity of engine life-limited parts throughout their service life. The engineering plan includes assumptions about the way in which life-limited parts are manufactured, operated and maintained. Each can affect the life of the part. It is, therefore, essential to ensure these assumptions remain valid. The service management plan conveys the constraints for in-service repair, maintenance, and overhaul to remain consistent with the assumptions made in the engineering plan.
b. Elements of a Service Management Plan. The part-specific service management plan should consider the part attributes that engineering has identified as significant for part life, and should not be changed by the processes applied to the part during service. The service management plan should include the means to monitor the service of life-limited parts to ensure the operational assumptions remain valid. The plan may include, but is not limited to, the following information:
Maintenance and overhaul limitations.
Repair processes limitations.
Operator’s responsibility to maintain service records traceable to a particular engine and part, as required by FAA regulations.
Inspection interval (if required).
Inspection procedure (if required).
Monitoring operational flight profiles.
Damage and repairable limits.
Periodic technical reviews of service and related experience.
c. Determining the Acceptability of Repair, Maintenance and Overhaul Processes.
(1) Repair, maintenance, and overhaul processes should be reviewed by personnel with the following technical skills:
Engineering (Design and Lifing).
Material Engineering.
Non-Destructive inspection.
Quality Assurance.
Product Support Engineering.
Repair Development Engineering.
(2) The skills needed for this cross-functional review are consistent with those needed to evaluate the manufacturing plan. The review should include process validation, change control, non-conformance, and corrective actions to ensure that all repair, maintenance, or overhaul processes are consistent with engineering requirements. The following benefits are derived from the review:
Repair, maintenance, and overhaul processes and practices are developed with the appropriate level of oversight, and their possible impact on the life of the part is considered. Substantia tion programs are agreed to up-front and executed as part of the validation process.
Changes to the processes and practices are visible to all parties and are made with cross-functional review and approval.
Non-conformances are reviewed by the appropriate skill mix prior to disposition.
Corrective action is implemented for non-conformances which have been detected.
(3) To ensure the service management processes are properly implemented and controlled, the applicant should clearly define the limitations to repair, maintenance and overhaul procedures in the ICA. These procedures should also include clearly defined limitations associated with these processes and practices to ensure that engine critical life-limited parts maintain the required attributes consistent with those assumed in the engineering plan. This is necessary since inappropriate repair or maintenance could affect the integrity of the part and result in a hazardous effect.
d. Static Parts Service Management. Due to potential differences in the approach to determining the life of static parts, especially of pressure cases, the applicant should include additional information in the shop manual. The applicant should place this information, which is in addition to the approved life, in the ALS. The additional information may include, but is not limited to:
A periodic inspection interval.
The inspection method(s) to be used.
A detailed description of the area(s) to be inspected.
Acceptance and rejection criteria associated with inspection results.
Acceptable repair method(s), if applicable.
Other instructions deemed necessary to ensure that inspections are executed properly.
Required maintenance and the limitations associated with maintenance.
11. AIRWORTHINESS LIMITATION SECTION.
a. The Airworthiness Limitation Section (ALS) of the engine manuals required by § 33.4 sets out the mandatory replacement times, inspection intervals, and related procedures necessary for type certification. The engine design engine ering and service management plans required by § 33.70 also highlight the importance of the limits to in-service maintenance and repair of lifelimited parts identified in the ALS. Since any maintenance or repair could theoretically affect the integrity of a life-limited part in a hazardous manner, an applicant may want to use the ALS to advise operators of the consequences of repairs to life-limited parts. If an applicant elects to include such an advisory statement, we recommend that it read similar to the following.
“The following airworthiness limitations have been developed based on engineering analysis that assumes this product will be operated and maintained using the procedures and inspections provided in the Instructions for Continued Airworthiness supplied with this product by the type certificate holder. Repairs or alterations to engine life-limited parts using other data warrant additional analyses to assess the potential effect on airworthiness characteristics and these limitations.”
Caution: Applicants should not attempt to incorporate any statement into the ALS that limits or eliminates an operator’s option to use FAA-approved repairs, alterations, or PMA parts in an engine. Such a statement would be contrary to FAA’s policies and regulations.
b. For engines with One-Engine-Inoperative (OEI) ratings, the ALS should include a method to track the number of cycles of operation at the OEI ratings because these ratings operate at higher speeds and temperatures than non-OEI operating conditions. Applicants may accomplish this by adding a finite number of cycles to the expended life of the affected engine life-limited parts or by using appropriate life reduction factors for each of the OEI power excursions.
//signed by Francis A. Favara// Francis A. Favara Manager, Engine and Propeller Directorate Aircraft Certification Service