Flight Simulation Training Device Qualification Standards for Extended Envelope and Adverse Weather Event Training Tasks

Federal RegisterMar 30, 2016

Ask Donna

What actually matters in this document.

Text

DEPARTMENT OF TRANSPORTATION

Federal Aviation Administration

14 CFR Part 60

[Docket No.: FAA-2014-0391; Amdt. No. 60-4]

RIN 2120-AK08

Flight Simulation Training Device Qualification Standards for Extended Envelope and Adverse Weather Event Training Tasks

AGENCY:

Federal Aviation Administration (FAA), DOT.

ACTION:

Final rule.

SUMMARY:

The FAA has determined this rule is necessary to amend the Qualification Performance Standards for flight simulation training devices (FSTDs) for the primary purpose of improving existing technical standards and introducing new technical standards for full stall and stick pusher maneuvers, upset recognition and recovery maneuvers, maneuvers conducted in airborne icing conditions, takeoff and landing maneuvers in gusting crosswinds, and bounced landing recovery maneuvers. These new and improved technical standards are intended to fully define FSTD fidelity requirements for conducting new flight training tasks introduced through recent changes to the air carrier training requirements, as well as to address various National Transportation Safety Board (NTSB) and Aviation Rulemaking Committee recommendations. This final rule also updates the FSTD technical standards to better align with the current international FSTD evaluation guidance and introduces a new FSTD level that expands the number of qualified flight training tasks in a fixed-base flight training device. These changes will ensure that the training and testing environment is accurate and realistic, will codify existing practice, and will provide greater harmonization with international guidance for simulation. The amendments will not apply to previously qualified FSTDs with the exception of the FSTD Directive, which codifies the new FSTD technical standards for specific training tasks.

DATES:

Effective May 31, 2016. The compliance date of FSTD Directive No. 2 is March 12, 2019. After this date, any FSTD being used to conduct specific training tasks as defined in FSTD Directive No. 2 must be evaluated and qualified in accordance with the Directive.

ADDRESSES:

For information on where to obtain copies of rulemaking documents and other information related to this final rule, see “How To Obtain Additional Information” in the

SUPPLEMENTARY INFORMATION

section of this document.

FOR FURTHER INFORMATION CONTACT:

For technical questions concerning this action, contact Larry McDonald, Air Transportation Division/National Simulator Program Branch, AFS-205, Federal Aviation Administration, P.O. Box 20636, Atlanta, GA 30320; telephone (404) 474-5620; email

larry.e.mcdonald@faa.gov

SUPPLEMENTARY INFORMATION:

Authority for This Rulemaking

The Federal Aviation Administration's (FAA's) authority to issue rules on aviation safety is found in Title 49 of the United States Code. Subtitle I, Section 106(f) describes the authority of the FAA Administrator. Subtitle VII, Aviation Programs, describes in more detail the scope of the agency's authority.

This rulemaking is promulgated under the authority described in 49 U.S.C. 44701(a)(5), which requires the Administrator to promulgate regulations and minimum standards for other practices, methods, and procedures necessary for safety in air commerce and national security. This amendment to the regulation is within the scope of that authority because it prescribes an accepted method for testing and evaluating flight simulation training devices used to train and evaluate flightcrew members.

In addition, the Airline Safety and Federal Aviation Administration Extension Act of 2010 (Pub. L. 111-216) specifically required the FAA to conduct rulemaking to ensure that all flightcrew members receive flight training in recognizing and avoiding stalls, recovering from stalls, and recognizing and avoiding upset of an aircraft, as well as the proper techniques to recover from upset. This rulemaking is within the scope of the authority in Public Law 111-216 and is necessary to fully implement the training requirements recently adopted in the Qualification, Service, and Use of Crewmembers and Aircraft Dispatchers final rule (Crewmember and Aircraft Dispatcher Training final rule), RIN 2120-AJ00. See 78 FR 67800 (Nov. 12, 2013).

List of Abbreviations and Acronyms Frequently Used in This Document

AC Advisory Circular

AOA Angle of Attack

ARC Aviation Rulemaking Committee

AURTA Airplane Upset Recovery Training Aid

FFS Full Flight Simulator

FTD Flight Training Device

FSTD Flight Simulation Training Device

ICATEE International Committee on Aviation Training in Extended Envelopes

LOCART Loss of Control Avoidance and Recovery Training Working Group

NPRM Notice of Proposed Rulemaking

QPS Qualification Performance Standards

SOC Statement of Compliance

SNPRM Supplemental Notice of Proposed Rulemaking

SPAW ARC Stick Pusher and Adverse Weather Event Training Aviation Rulemaking Committee

UPRT Upset Prevention and Recovery Training

Table of Contents

I. Overview of Final Rule

II. Background

A. Statement of the Problem

B. NTSB Recommendations

C. Airline Safety and Federal Aviation Administration Extension Act of 2010 (Publ. L. 111-216) and the Crewmember and Aircraft Dispatcher Training Final Rule

D. Summary of the NPRM

E. Differences Between the NPRM and the Final Rule

F. Related Actions

III. Discussion of Public Comments and Final Rule

A. Evaluation Requirements for Full Stall Training

1. Aerodynamic Modeling Range

a. Aerodynamic Modeling Beyond the Stall AOA

b. Definition of the Stall AOA

2. Envelope Protected Aircraft

a. Model Validity Ranges and Associated Objective Testing

b. Validation of Stall Characteristics Using Flight Test Data

c. Required AOA Range for Normal Mode Objective Testing

3. Data Sources for Model Development and Validation

a. Define Best Available Data

b. Post-Stall “Type Representative” Modeling

c. Use of Flight Test Data and Availability

4. Qualification on FSTD Levels Other Than Level C or Level D

5. Motion Cueing System Limitations

6. Subject Matter Expert (SME) Pilot Evaluation and Qualifications

a. SME Qualifications and Experience

b. Model Validation Conducted by the Data Provider

c. NSPM Process for Evaluating and Accepting an SME Pilot

7. Alignment With the ICAO 9625, Edition 4, on Stall and Stick Pusher Requirements

8. Requirements for Previously Qualified FSTDs

a. Stall Buffet Objective Testing

b. FSTD Directive No. 2 and Grandfather Rights

9. Applicability of Stall and Upset Prevention and Recovery Training

(UPRT) Requirements on Newly Qualified FSTDs

10. General Comments on Stall Requirements

a. Testing and Checking of Stall Maneuvers

b. Interim FSTD Qualification for Stall Training

c. Aerodynamic Modeling Considerations

B. Evaluation Requirements for UPRT

1. UPRT Qualification on Lower Level FSTDs.

2. Record and Playback Requirements for UPRT

3. Instructor Operating Station (IOS) Requirements

4. Aerodynamic Source Data and Range of the FSTD Validation Envelope

a. FSTD Validation Envelope and Training Maneuvers

b. Expansion of the FSTD Validation Envelope Using Existing Flight Test Data

5. General Comments on UPRT

a. FSTD Qualification and FAA Oversight

b. Maintenance Concerns

C. Evaluation Requirements for Engine and Airframe Icing Training

1. Objective Demonstration Testing

a. Objective Demonstration Testing for Previously Qualified FSTDs

b. Icing Effects and Recognition Cues

2. Requirements for Lower Level FTDs

3. Existing Engine and Airframe Icing Requirements in Part 60

4. Applicability in Training Programs

5. Data Sources and Tuning of Ice Accretion Models

D. Evaluation Requirements for Takeoff and Landing in Gusting Crosswinds

1. Applicability on Lower Level FSTDs

2. Gusting Crosswind Profile Data Sources

3. Maximum Demonstrated Crosswind

4. Requirements for Previously Qualified FSTDs

E. Evaluation Requirements for Bounced Landing Recovery Training

1. Applicability to Lower Level FSTDs

2. Bounced Landing Modeling and Evaluation

a. Nosewheel Exceedances

b. Use of Existing Ground Reaction Models

3. Alignment With Training Requirements

4. Requirements for Previously Qualified FSTDs

F. Alignment With the ICAO 9625 FSTD Evaluation Document

1. Partial Alignment With the ICAO 9625 Document

2. New Requirements Introduced by the Proposed ICAO Alignment

a. Visual System Field of View

b. Visual System Lightpoint Brightness Testing

c. Transport Delay Testing

d. Motion Cueing Fidelity Test

e. Sound Directionality Requirement

3. Alignment With the Recently Published ICAO 9625, Edition 4 Document

4. Integration of ICAO Requirements With the Part 60 Table Structure

5. Deviation From the Part 60 QPS Using the ICAO 9625 Document

6. Level 7 FTD Requirements and Usage in Training

G. General Comments

1. Compliance Period for Previously Qualified FSTDs

2. Alternative Source Data for Level 5 FTDs

3. Objective Testing for Continuing Qualification

4. Windshear Qualification Requirements

5. Miscellaneous Comments

a. Approved Location for Objective and Subjective Testing

b. Increase the Training Credit for Time in a Simulator

H. Economic Evaluation

1. Cost of Aerodynamic Modeling and Implementation

2. Cost of Instructor Operation Station (IOS) Replacement

3. Affected FSTDs and Sponsors

4. Cost and Benefits of ICAO Alignment

IV. Regulatory Notices and Analyses

A. Regulatory Evaluation

B. Regulatory Flexibility Determination

C. International Trade Impact Assessment

D. Unfunded Mandates Assessment

E. Paperwork Reduction Act

F. International Compatibility and Cooperation

G. Environmental Analysis

H. Regulations Affecting Intrastate Aviation in Alaska

V. Executive Order Determinations

A. Executive Order 13123, Federalism

B. Executive Order 13211, Regulations that Significantly Affect Energy Supply, Distribution, or Use

VI. How To Obtain Additional Information

A. Rulemaking Documents

B. Comments Submitted to the Docket

C. Small Business Regulatory Enforcement Fairness Act

I. Overview of Final Rule

This rulemaking defines simulator fidelity requirements for new training tasks to be conducted in Level A through D full flight simulators (FFS) that were mandated for air carrier training programs by Public Law 111-216 and incorporated into 14 CFR part 121. It also addresses the potential lack of simulator fidelity as identified in several NTSB safety recommendations. This final rule establishes new and updated FSTD technical evaluation standards for full stall and stick pusher maneuvers, upset prevention and recovery maneuvers, flight in airborne icing conditions, takeoff and landing maneuvers in gusting crosswinds, and bounced landing recovery maneuvers. This final rule also partially aligns the technical standards for Level C and D (fixed wing) FSTDs that are defined in 14 CFR part 60 with the current international FSTD evaluation guidelines published in the International Civil Aviation Organization (ICAO) document 9625, Edition 4,

Manual of Criteria for the Qualification of Flight Simulation Training Devices.

This final rule will affect sponsors of previously qualified FSTDs if the devices will be used to conduct the specific training tasks defined in FSTD Directive No. 2. The FSTD sponsor has the discretion to determine if a device needs to be qualified based on whether it will be used for training the defined tasks in FSTD Directive No. 2. Additionally, because many of the technical FSTD evaluation standards in the final rule will become minimum requirements for some newly qualified FSTDs, this final rule will also affect sponsors of Level 7, Level C, and Level D FSTDs that are initially qualified after the effective date of the final rule. In addition to FSTD sponsors, this final rule will also affect data providers, FSTD manufacturers, and other entities that provide products and support to FSTD sponsors in the qualification of FSTDs for training. This final rule does not affect aviation training devices that are evaluated and approved for use outside of 14 CFR part 60.

A general summary of the applicability, compliance dates, and processes used to qualify FSTDs as defined in this rule are included in the following table:

Issue

Rule requirements

How does a sponsor determine if a previously qualified FSTD must be evaluated and qualified for stall, UPRT, engine and airframe icing, bounced landing recovery, and gusting crosswind training tasks as defined in FSTD Directive No. 2?

A previously qualified FSTD that will be used to obtain training, testing, or checking credit in an FAA approved flight training program, regardless of operational rule part, must be evaluated and qualified for the following maneuvers:

Full Stall: Training maneuvers in the recognition cues and recovery procedures from a fully stalled flight condition (including recovery from a stick pusher activation) at angles of attack beyond the activation of the stall warning system.

UPRT: Upset recovery maneuvers and unusual attitude maneuvers that are intended to exceed the parameters of an aircraft upset as defined in the Airplane Upset Recovery Training Aid (pitch attitudes greater than 25 degrees nose up; pitch attitudes greater than 10 degrees nose down, and bank angles greater than 45 degrees).

Engine and Airframe Icing: Flight training maneuvers that demonstrate the recognition cues and effects of engine and airframe ice accretion.

Takeoff and Landing in Gusting Crosswinds.

Bounced Landing Recovery Training.

How does a sponsor obtain qualification for stall, UPRT, icing, bounced landing recovery, or takeoff and landing in gusting crosswinds on a previously qualified FSTD?

FSTD Directive No. 2 contains all of the evaluation requirements for the qualification of these individual tasks on previously qualified FSTDs. FSTD sponsors will conduct the evaluations and modifications as described in the Directive and submit any required Statements of Compliance and objective testing results to the National Simulator Program (NSP) using the standard FSTD modification/notification process. The NSP will issue additional FSTD qualification for these tasks once compliance with the applicable sections of the Directive are verified and any necessary FSTD evaluations have been conducted.

How do you determine what portions of the updated qualification performance standards (QPS) appendices are applicable to previously qualified FSTDs?

As described in § 60.17(a), unless specified by an FSTD Directive, previously qualified (grandfathered) FSTDs will retain their original qualification basis under which they were originally evaluated, regardless of sponsor. All retroactive evaluation requirements for previously qualified FSTDs in this final rule are fully described in FSTD Directive No. 2.

What are the compliance dates associated with this final rule for previously qualified FSTDs?

After March 12, 2019, any FSTD being used to conduct the specific training maneuvers (as described in FSTD Directive No. 2) in an FAA approved training program must be issued additional FSTD qualification in accordance with the Directive.

How do you determine what changes in this final rule are applicable to new FSTDs that will be initially qualified after the final rule becomes effective?

With the exception of the full stall evaluation requirements, all FSTDs that are initially qualified or upgraded in qualification level after the effective date of the final rule must meet all new standards in this final rule as applicable for the particular FSTD qualification level requested.

The qualification of full stall training tasks will be optional as requested by the sponsor to support FAA approved training being conducted in the FSTD. The qualification of full stall training tasks will be included as part of the list of qualified tasks on the FSTD's Statement of Qualification (SOQ).

What is the compliance date associated with this final rule for new FSTDs that will be initially qualified after the rule becomes effective?

In general, all changes to the part 60 QPS will be effective for all FSTDs that are initially qualified after the effective date of the final rule except as permitted by § 60.15(c).

What is the process to qualify an FSTD using another standard in lieu of the part 60 QPS as permitted by the deviation authority in § 60.15?

Requests for deviation from the part 60 QPS are made to the National Simulator Program Manager (NSPM) and must include justification that demonstrates an equivalent level of safety as compared to the FSTD evaluation requirements of the part 60 QPS. Approved deviations and the supporting evaluation standards will become a part of the permanent qualification basis of the FSTD.

The FAA estimates that it will cost $72.7 million to make the necessary modifications to previously qualified FSTDs which will enable training required by the new Crewmember and Aircraft Dispatcher Training final rule. The training cost for the Crewmember and Aircraft Dispatcher Training final rule provides rental revenue to simulator sponsors which will fully compensate them for their FSTD modification expenses. These simulator revenues were accounted for as costs of the additional training and were fully justified by the benefits in that final rule. The FAA estimates it will cost $1.3 million for the evaluation and modification of engine and airframe icing models which will enhance existing training requirements. If these modifications prevent only one severe injury the benefits will exceed the costs. The estimated cost of $6.9 million to align standards with ICAO will result in improved safety and cost savings.

The costs and benefits of this rule are presented in the table below.

Present value at a 7% rate

Present value at a 3% rate

FSTD Modifications for New Training Requirements:

Cost

$72,716,590

$63,610,049

$68,562,049

Benefits

Rational simulator owner will choose to comply.

Icing provisions:

Cost

$1,256,250

$1,098,926

$1,184,476

Benefits

Only one prevented severe injury valued at $2.5 million makes the icing benefits exceed the costs.

Aligning Standards with ICAO:

Cost

$6,875,000

$5,356,979

$6,132,690

Benefits

Improved safety and cost savings.

Total Cost

$80,847,840

$70,065,954

$75,879,215

II. Background

A. Statement of the Problem

In order to mitigate aircraft loss of control accidents and to comply with the requirements of Public Law 111-216, the FAA has issued new and revised flight training requirements in the Crewmember and Aircraft Dispatcher Training final rule for flight maneuvers such as full stall and upset recovery training. In support of this effort, the FAA participated in a number of collaborative industry and government working groups that examined loss of control training requirements and the flight simulation training device (FSTD) fidelity needed to support such training. These working groups included the International Committee on Aviation Training in Extended Envelopes (ICATEE), the Industry Stall and Stick Pusher Working Group, the Stick Pusher and Adverse Weather Event Training Aviation Rulemaking Committee (SPAW ARC), and the Loss of Control Avoidance and Recovery Training (LOCART) Working Group.

Through participation in these working groups and in consideration of the formal recommendations received from the SPAW ARC, the FAA determined that many existing FSTDs that could be used by air carriers to conduct such training may not adequately represent the simulated aircraft for the required training tasks. Additionally, the FAA evaluated several recent air carrier accidents and associated NTSB accident reports and determined that low FSTD fidelity or the lack of ability for an FSTD to adequately conduct certain training tasks may have been a contributing factor in these accidents.

A potential lack of simulator fidelity could contribute to inaccurate or incomplete training on new training tasks that are required by the Crewmember and Aircraft Dispatcher Training final rule, which could lead to a safety risk.

Some of these accidents include the 1996 Airborne Express DC-8-63 loss of control accident, the 2001 American Airlines flight 587 A300 loss of control accident, the 2009 Colgan Air flight 3407 DHC-8-400 loss of control accident, and the 2008 Continental flight 1404 Boeing 737-500 runway excursion accident.

Furthermore, since the initial publication of the part 60 final rule in 2008, the international FSTD qualification guidance published in the ICAO 9625 document has been updated to incorporate general improvements to new aircraft and simulation technology and the introduction of new FSTD levels that better align FSTD fidelity with required training tasks. The ICAO 9625 document is an internationally recognized set of FSTD evaluation guidelines that was developed by government and industry experts on flight simulation training and technology and has been used as a basis for national regulation and guidance material for FSTD evaluation in many countries. Internationally aligned FSTD standards facilitate cost savings for FSTD operators because they can reduce the number of different FSTD designs, as well as reduce the amount of redundant supporting documentation that are required to meet multiple national regulations and standards for FSTD qualification.

This final rule was developed using recommendations from the SPAW ARC

and the international FSTD qualification guidelines that are published in ICAO 9625, Edition 3 and the newly published ICAO 9625, Edition 4.

The requirements in this final rule are primarily directed at improving the fidelity of FSTDs that will be used in air carrier pilot training to conduct extended envelope training tasks, but will also have an added benefit of improving the fidelity of all FSTDs initially qualified after the final rule becomes effective.

A copy of the SPAW ARC final report has been placed in the docket for this rulemaking.

International Civil Aviation Organization (ICAO) publications can be located on their public internet site at:

http://www.icao.int/.

B. National Transportation Safety Board (NTSB) Recommendations

This proposal will incorporate changes into part 60 that address, at least in part, the following NTSB Safety Recommendations through improved FSTD evaluation standards to support required training tasks:

1. Stall training and/or stick pusher training (Recommendations A-10-22, A-10-23, A-97-47, A-07-3, and A-10-24);

2. Upset Recognition and recovery training (Recommendations A-04-62 and A-96-120);

3. Engine and airframe icing training (Recommendations A-11-46 and A-11-47)

4. Takeoff and landing training in gusting crosswind conditions (Recommendations A-10-110 and A-10-111); and

5. Bounced landing training (Recommendations A-00-93 and A-11-69).

C. Airline Safety and Federal Aviation Administration Extension Act of 2010 (Pub. L. 111-216) and the Crewmember and Aircraft Dispatcher Training Final Rule

On August 1, 2010, President Obama signed into law Public Law 111-216. In addition to extending the FAA's authorization, Public Law 111-216 included provisions to improve airline safety and pilot training. Specifically, section 208 of Public Law 111-216, Implementation of NTSB Flight Crewmember Training Recommendations, pertains directly to this rulemaking in that stall training and upset recovery training were mandated for part 121 air carrier flightcrew members.

On November 12, 2013, the FAA published the Crewmember and Aircraft Dispatcher Training final rule, adding the training tasks required by Public Law 111-216 that specifically target extended envelope training, recovery from bounced landings, enhanced runway safety training, and enhanced training on crosswind takeoffs and landings with gusts, which further requires that these maneuvers be completed in an FSTD. As a result, revisions to all part 121 training programs will be necessary prior to March 12, 2019 and the revisions to part

60 in this final rule are required to ensure FSTDs are properly evaluated in order to fully implement the flight training required in the Crewmember and Aircraft Dispatcher Training final rule.

D. Summary of the Notice of Proposed Rulemaking (NPRM)

On July 10, 2014, the FAA published an NPRM (79 FR 39461), proposing changes to the flight simulation training device (FSTD) technical evaluation standards. The primary purpose of the NPRM was to establish and update FSTD technical evaluation standards to address new training tasks required by the Crewmember and Dispatcher Training final rule, including full stall training, upset prevention and recovery training, and other new training tasks. Additionally, the NPRM proposed the incorporation of FSTD evaluation criteria as defined in the International Civil Aviation Organization (ICAO) 9625,

Manual of Criteria for the Qualification of Flight Simulation Training Devices

(Edition 3) document. Significant changes to the part 60 qualification performance standards (QPS) were proposed in the following areas:

Full Stall Evaluation:

Minimum requirements were introduced to include aerodynamic modeling of a full stall and stick pusher activation (where equipped) up to ten degrees of angle of attack (AOA) beyond the stall AOA, subject matter expert (SME) pilot evaluation of the FSTD's stall characteristics, and improved objective testing to validate the FSTD's performance and handling qualities in the stall maneuver.

Upset Recognition and Recovery:

New requirements were proposed for the qualification of upset recognition and recovery training tasks including the evaluation of a minimum set of upset recovery maneuvers against the defined FSTD validation envelope, providing a means to record and playback upset recovery maneuvers conducted in the FSTD, and providing the instructor with a minimum set of feedback tools on the instructor operating station (IOS) that gives information on the FSTD's expected fidelity, aircraft operational limitations, and student flight control inputs.

Engine and Airframe Icing:

Modifications were proposed to the existing part 60 Level C and Level D FSTD qualification requirements for engine and airframe icing. The proposed amendments included requirements for ice accretion models based upon aircraft original equipment manufacturer (OEM) data or other analytical methods that incorporate the aerodynamic effects of icing as well as objective tests on the FSTD that demonstrate the effects of icing.

Takeoff and Landing in Gusting Crosswinds:

New amendments were proposed that would require that realistic gusting crosswind profiles must be available to the instructor and the profiles must be tuned in intensity and variation to require pilot intervention to avoid runway departure during takeoff or landing roll. A Statement of Compliance (SOC) would be required to describe the source data used to develop the crosswind profiles.

Bounced Landing Recovery:

New requirements were proposed to complement existing part 60 ground reaction requirements to support bounced landing recovery training. The updated requirements added that the effects of a bounced landing must be modeled and evaluated and include the effects of nosewheel exceedances and tail strike where appropriate.

ICAO 9625 Alignment:

In the NPRM, the FAA proposed alignment with the updated ICAO 9625, Edition 3, FSTD evaluation document for similar FSTD levels that are defined in the part 60 QPS (Appendices A and B). This included incorporating updated technical standards for Level C and Level D FSTDs to align with that of the ICAO Type VII FSTD and creating a new high fidelity fixed-base flight training device (the Level 7 FTD) that is based upon the similar Type V device as defined in the ICAO document. This alignment also included adopting the ICAO language and numbering format for some of the technical requirements tables as well as integrating the existing legacy part 60 FSTD levels into these tables to maintain continuity with the current part 60 defined hierarchy of FSTD levels.

In general, the proposed amendments to the part 60 QPS would only be applicable to FSTDs that are initially qualified or upgraded in qualification level after the final rule becomes effective. Because many previously qualified FSTDs will likely be used to accomplish the training tasks required by the Crewmember and Dispatcher Training final rule, the FAA also proposed an FSTD Directive in order to retroactively apply evaluation requirements for those previously qualified FSTDs that will be used to conduct certain training tasks, including full stall, upset prevention and recovery training, engine and airframe icing, takeoff and landing in gusting crosswinds, and bounced landing recovery training.

On September 16, 2014, the FAA extended the comment period of the NPRM for an additional 90 days (79 FR 55407). The comment period closed on January 6, 2015. The FAA received approximately 675 individual comments in response to the NPRM. Commenters included air carriers, simulator training providers, FSTD data providers, FSTD manufacturers, the NTSB, labor organizations, trade associations, aircraft manufacturers, and individuals.

E. Differences Between the NPRM and the Final Rule

As a result of the comments received on the NPRM, the FAA made several changes to the final rule. A summary of significant changes as a result of comments are highlighted in the following table:

Issue

Significant changes

Full Stall Evaluation

(a) Improved the definition of the stall AOA for the purposes of defining the required aerodynamic modeling range. Clarifies specific issues concerning stick pusher equipped aircraft and envelope protected aircraft.

(b) Made clarifications concerning acceptable source data for stall aerodynamic models. Clarified that data sources other than the aircraft manufacturer may be acceptable if they meet the modeling and SME pilot evaluation requirements.

(c) Improved the qualification requirements for subject matter expert (SME) pilots that subjectively evaluate the stall model. Adds deviation authority if an acceptable SME pilot cannot be located. Allows for SME evaluation to be conducted on an engineering or development simulator where objective proof-of-match test cases are provided that verifies the model implementation on the FSTD.

(d) Removed the proposed requirement for all newly qualified FSTDs to be evaluated and qualified for full stall training tasks. Full stall qualification will only be required for FSTDs that will be used to conduct this training as requested by the FSTD sponsor.

(e) (Previously qualified FSTDs) Removed the proposed objective testing requirements for stall maneuvers where validation data may not exist for some older FSTD data packages (cruise and turning flight stall). These conditions will still require aerodynamic modeling and subjective evaluation by a SME pilot.

Upset Prevention and Recovery Training (UPRT) Evaluation

(a) Removed the proposed minimum FSTD evaluation requirements for Level A and Level B FSTDs.

(b) Removed the proposed specific requirements for features and malfunctions necessary to drive upset scenarios.

(d) Improved the definition of required instructor operating station (IOS) parameters and feedback mechanisms. Allows for methods other than graphical displays to be used where the required parameters are provided to support the training program.

(e) Expands the definition of UPRT to include unusual attitude training in which scenarios are introduced that are intended to exceed the defined parameters of an aircraft upset. This change better differentiates UPRT from the existing part 60 unusual attitude evaluation requirement in Table A1B.

Engine and Airframe Icing Evaluation

(a) Clarified that specific icing effects are only required to be introduced where such effects are representative of the particular aircraft being simulated.

(b) Revised the existing part 60 engine and airframe icing special effects test (Table A3F) to remove references to gross weight increments and to better align with the updated requirements.

(c) Clarified that flight test data is not necessarily required for the development of icing models. Engineering and analytical methods may be used to develop representative icing models.

(d) Added provisions to allow for supplemental tuning of icing models using an SME pilot assessment.

Gusting Crosswind Evaluation

(a) Removed references to the windshear training aid for gusting crosswind model development. Recommend use of gusting crosswind profiles provided by the FAA in guidance material.

(b) Removed the proposed minimum qualification requirement for Level A and Level B FSTDs.

Bounced Landing Recovery Evaluation

(a) Removed the proposed ground reaction requirement to compute nosewheel exceedances.

(b) Clarified the requirements to emphasize the effects and indications of ground contact due to landing in an abnormal aircraft attitude and that aircraft dynamics in a bounced landing recovery maneuver are already adequately covered in the existing part 60 rule.

Alignment with the ICAO 9625 Document

(a) Restored the general requirements table (Tables A1A and B1A) format, numbering system, and content to the existing part 60 versions. Appended the proposed ICAO 9625 (Edition 3) requirements from the NPRM into their applicable sections.

(b) Restored the existing part 60 visual system field of view (180°x40°) and system geometry requirements for Level C and Level D FSTDs.

(c) Adopted the less restrictive visual system lightpoint brightness tolerance (5.8 ft.-lamberts) from the updated ICAO 9625, Edition 4, document.

(d) Adopted the less restrictive transport delay tolerances (100 ms for instrument and motion system response; 120 ms for visual system response) from the updated ICAO 9625, Edition 4, document.

(e) Modified the objective motion cueing test (OMCT) description to not require testing for continuing qualification evaluations, removed minimum tolerances, and further moved much of the technical test details into guidance material.

(f) Aligned language where practical for similar stall, UPRT, and icing requirements from the ICAO 9625, Edition 4, document.

(g) Added deviation authority for the FAA to accept alternate FSTD evaluation standards where no adverse impact to the fidelity of the FSTD can be demonstrated.

(h) Reorganized the flight training device (FTD) requirements in Appendix B to restore the existing part 60 table structure and better separate requirements for the new Level 7 FTD and the legacy part 60 FTD levels.

(i) Clarified the Level 7 FTD's minimum qualified training tasks in Table B1B to better align with the ICAO 9625 guidelines.

(j) Removed minimum requirements for extended envelope training tasks for the Level 7 FTD that are not included in the ICAO 9625, Edition 4 document for the Type V device.

F. Related Actions

As a result of information gathered from various working groups, the FAA has taken action on loss of control training and simulator fidelity deficiencies by issuing the following voluntary guidance material:

1. FAA Safety Alert for Operators (SAFO 10012)—Possible Misinterpretation of the Practical Test Standards (PTS) Language “Minimal Loss of Altitude.” The purpose of this alert bulletin is to clarify the meaning of the approach to stall evaluation criteria as it relates to “minimal loss of altitude” in the Airline Transport Pilot PTS;

2. FAA Information for Operators Bulletin (InFO 10010)—Enhanced Upset Recovery Training. This information bulletin recommends the incorporation of the material in the AURTA into flightcrew training. The AURTA contains guidance for upset recovery training programs for air carrier flightcrews, as well as the evaluation guidance for FSTDs used in such training;

3. FAA Information for Operators Bulletin (InFO 15004)—Use of Windshear Models in FAA Qualified Flight Simulation Training Devices (FSTDs);

4. FAA National Simulator Program (NSP) Guidance Bulletin No. 11-04—FSTD Modeling and Evaluation Recommendations for Engine and Airframe Icing;

5. FAA National Simulator Program (NSP) Guidance Bulletin No. 11-05—FSTD Evaluation Recommendations for Upset Recovery Training Maneuvers;

6. FAA National Simulator Program (NSP) Guidance Bulletin No. 14-01—FSTD Evaluation Guidelines for Full Stall Training Maneuvers;

7. AC 120-109A—Stall and Stick Pusher Training;

8. AC 120-111—Upset Prevention and Recovery Training; and

9. Airline Transport Pilot Practical Test Standards (Change 4).

Portions of the above guidance material provide FSTD operators with recommended evaluation methods to improve FSTD fidelity for selected training tasks. To ensure that all FSTDs used to conduct such training are evaluated and modified to a consistent standard, the applicable part 60

technical requirements must be modified as described in this final rule.

III. Discussion of Public Comments and Final Rule

A. Evaluation Requirements for Full Stall Training Tasks

The existing FSTD evaluation requirements for stall maneuvers are generally limited to the evaluation of stall speeds with little emphasis on the actual aircraft performance and handling characteristics as the aircraft exceeds the stall warning AOA. As a result, FSTDs used for such training may not provide the necessary cues and associated performance degradation needed to train flight crews in the recognition of an impending stall as well as training the techniques needed to recover from a stalled flight condition. In the NPRM, the FAA proposed updated general requirements, objective testing requirements, and functions and subjective testing requirements for the evaluation of full stall training maneuvers to support air carrier training as required in the Crewmember and Aircraft Dispatcher Training final rule.

Aerodynamic Modeling Range

a. Aerodynamic Modeling Beyond the Stall AOA

In order to support the required training objectives, the proposal included requirements for the modeling and evaluation of the FSTD's stall characteristics up to 10 degrees beyond the stall AOA.

CAE, Inc. (CAE) commented that the 10 degrees beyond the stall AOA requirement should be further reviewed, since application of the recovery should immediately lead to a reduction in AOA and therefore is inappropriate to relate the requirement to the 10 degrees beyond the stall AOA. CAE recommended that the 10 degree requirement be removed where rationale is provided for the upper limit of AOA modeling in the required SOC.

The NTSB is generally supportive of the modeling requirements, citing that a peak AOA growth of about 10 degrees beyond the stall is typical for most incidents and accidents it has investigated. However, it did note that stick pusher response dynamics could cause a higher AOA overshoot and this dynamic behavior is a “critical cue to a stall, which pilots must be trained to recognize.” The NTSB also noted in its comments that the Colgan flight 3407 accident resulted in an AOA that extended to 13 degrees beyond the stall AOA.

In addition, the NTSB stated that the required aerodynamic modeling for aircraft equipped with a stick pusher should not be limited to that of the stick pusher activation and that the aerodynamic modeling range include the flight dynamics that may occur where a pilot resists the stick pusher in training.

See NTSB accident report, Loss of Control on Approach, Continental Connection Flight 3407, February 12, 2009, NTSB Accident Report, NTSB/AAR-10/01; page 87, “After the stall, the AOA oscillated between 10 deg and 27 deg . . . .”.

The FAA disagrees with CAE that the 10 degree requirement be removed in select cases. The 10 degree AOA range was initially recommended by the SPAW ARC as necessary to accomplish full stall training. Furthermore, this 10 degree AOA range is currently a recommended practice for simulator aerodynamic modeling in the International Air Transport Association (IATA) Flight Simulation Training Device Design and Performance Data Requirements document

and has been a recommended practice since the second edition of the IATA document that was published in 1986. Finally, the FAA notes that an unpublished simulator investigation conducted by ICATEE in conjunction with NASA on their Enhanced Upset Recovery model showed that the 10 degree AOA range should be sufficient to capture most overshoots in AOA during various stall recovery maneuvers.

International Air Transport Association (IATA) Flight Simulation Training Device Design and Performance Data Requirements Document, 7th Edition (2009), sections 3.1.1.2 and 3.1.1.3 addresses stall entry and recovery as well as required angle of attack ranges for supporting data.

The FAA agrees with the NTSB that pilots can benefit from experiencing the aircraft dynamics involved in a stick pusher activation and recovery maneuver in training. The FAA has reviewed the NTSB accident reports and supporting data on two loss of control accidents in which pilots resisted the activation of a stick pusher and encountered an aerodynamic stall. In the Pinnacle Airlines Flight 3701 accident, the initial stick pusher activation occurred at approximately 10.5 degrees AOA at the start of the aircraft upset and the AOA subsequently oscillated from approximately −6 degrees to +14 degrees over three successive stick pusher activations with some instability evident in the roll axis.

Only until just before the fourth activation of the stick pusher system (approximately eleven seconds after the initial stick pusher activation) did the AOA exceed the proposed aerodynamic modeling range (of 10 degrees beyond the stall AOA) for FSTD evaluation purposes.

See NTSB accident report, Crash of Pinnacle Airlines Flight 3701, October 14, 2004, NTSB Accident Report, NTSB/AAR-07/01 and supporting flight data recorder factual report on the NTSB public docket (NTSB accident identification number DCA05MA003).

For this aircraft, since the aerodynamic stall occurs after the stick pusher is designed to activate, the stall identification is provided by the stick pusher system activation and aerodynamic modeling would be required up to at least 20.5 degrees AOA for this configuration.

In the Colgan 3407 accident, aerodynamic stall occurred before the stick pusher activation

at approximately 14 degrees AOA which included an initial roll off to about 50 degrees of bank angle. After the initial stick pusher activation at about 17.5 degrees AOA, the subsequent AOA overshoot remained within 24 degrees as the aircraft rolled through 100 degrees of bank angle in the opposite direction of the initial roll off. The peak AOA value of approximately 27 degrees (10 degrees of AOA beyond the stick pusher activation where stall identification should have occurred) was not recorded until after multiple incorrect column responses by the pilot against the stick pusher over a time period of 30 seconds after the pilot's initial incorrect response to the stall warning.

According to the NTSB accident report, the stick pusher on this aircraft is designed to activate after the aerodynamic stall.

The FAA considered the comments and based on a review of industry recommendations and best practices, has determined that aerodynamic modeling to at least 10 degrees beyond the stall AOA is necessary so that the modeling does not abruptly end should the pilot overshoot the stall recognition and recovery in training. The FAA recognizes that the 10 degree AOA range may not be sufficient to capture all of the flight dynamics involved with multiple severe divergent pitch oscillations where the pilot repeatedly resists a stick pusher system; however, training should not normally be allowed to continue significantly beyond the point where a trainee initially resists the stick pusher before recognizing the stall identification cues and executing the recovery procedures. As demonstrated by the AOA oscillations experienced in the Colgan and Pinnacle accidents, the FAA has determined that aerodynamic modeling to 10 degrees beyond the stall AOA should be sufficient to capture aircraft dynamics in instances where a pilot initially resists the stick pusher activation in training. The data from these accidents suggests that the 10 degree AOA aerodynamic modeling requirement would adequately cover an

AOA range that includes several seconds of inappropriate pilot responses to a stick pusher activation. The FAA has determined this range is sufficient to meet the training objective of teaching a pilot to not resist a stick pusher system activation.

b. Definition of the Stall AOA

In the NPRM, the FAA defined the required aerodynamic model validity range for full stall qualification as 10 degrees of AOA beyond the stall/critical AOA and not as a function of when the stall identification cues are present.

Airbus commented that the definition of stall or full stall should emphasize “heavy buffet” as an important cue. Airbus further cited the ICAO 9625, Edition 4, document

states that a stalled flight condition may be recognized by continuous stall warning activation accompanied by at least one of the following: (1) Buffeting, which could be heavy at times; (2) lack of pitch authority and/or roll control; or (3) inability to arrest the descent rate.

See section III.F.3 concerning changes made to address the recently published ICAO 9625, Edition 4 document.

The FAA concurs with Airbus' comment that heavy buffet can be an important cue of a stall. The FAA has further considered the definition of stall as described in the ICAO 9625 document to determine an appropriate definition for stall with respect to the modeling requirements necessary to support the training objectives. The FAA does not fully agree, however, with the ICAO 9625 definition of stall; specifically the criteria of “lack of pitch authority and/or roll control” to define the stall since the part 25 airplane certification requirements state that the pilot must be able to control the aircraft in pitch and roll up to the stall. While control effectiveness can be reduced, it would be incorrect to say that it is lacking for certified airplanes.

Two fundamental objectives of the stall training requirements are to train pilots to recognize the cues of an impending stall as well as to reinforce to pilots that the stall recovery procedures learned during stall prevention training are the same recovery procedures needed to recover from an unintentional full stall. To determine the extent of FSTD aerodynamic modeling necessary to conduct this training, the stall identification AOA must be defined as the point in which the pilot should recognize that the aircraft has stalled and that the stall recovery procedures must be initiated. The FAA has considered both the aircraft certification (part 25) definition of a “clear and distinctive” indication of a stall, as well as the ICAO 9625, Edition 4, stall definition. In order to provide a more consistent definition of the stall AOA to ensure that the required aerodynamic modeling range covers potential overshoots in AOA during stall training, the FAA has amended the final rule to better define stall identification:

i. No further increase in pitch occurs when the pitch control is held on the aft stop for 2 seconds, leading to an inability to arrest descent rate;

ii. An uncommanded nose down pitch that cannot be readily arrested, which may be accompanied by an uncommanded rolling motion;

iii. Buffeting of a magnitude and severity that is a strong and effective deterrent to further increase in AOA; and

iv. The activation of a stick pusher.

Since AOA awareness is a fundamental element of stall training, the instructor must be provided with feedback at the IOS concerning the aircraft's current AOA as well as the stall identification AOA. This feedback will not only provide the instructor with additional awareness concerning the aircraft's current AOA and proximity to the stall, but will also assist the instructor in determining when the aircraft has stalled and that the stall recognition cues have been provided as necessary to support the training objectives. In the final rule, the FAA has amended the IOS feedback requirements for upset prevention and recovery training to include AOA and stall identification AOA parameters.

The FAA further notes that the stall identification cues exhibited by an aircraft can, and often do, vary depending upon the aircraft's configuration (

e.g.

weight, center of gravity, and flap setting) and how the stall is entered (turning flight or wings level stall entry). Where differing stall identification cues are present on the aircraft, the FSTD's aerodynamic model should be capable of providing these cues and variation of stall characteristics for training purposes. The FAA also points out that, while this requirement was implied in the stall model evaluation requirements in the NPRM, ICAO 9625, Edition 4, further clarifies this issue with additional language which states that “. . . the model should be capable of capturing the variations seen in the stall characteristics of the aeroplane (

e.g.,

the presence or absence of a pitch break).” The FAA has determined that the ability to show these variations would be valuable in training and has included similar clarifying language in Table A1A, section 2.m. of the final rule.

Envelope Protected Aircraft

a. Model Validity Ranges and Associated Objective Testing

In the NPRM, the FAA included provisions that did not specifically require objective validation testing at an AOA beyond the activation of a stall identification (stick pusher) system through recovery. The primary purpose of including this provision was to not require the collection of flight test validation data at an AOA that could result in an unrecoverable and dangerous stalled flight condition.

Empresa Brasileira de Aeronautica S.A. (Embraer), Airbus, and an individual commenter questioned why computer controlled aircraft with stall envelope protection systems are treated differently from aircraft equipped with stick pusher systems with respect to model validity ranges and associated objective testing. Delta Airlines, Inc. (Delta) further questioned whether such modeling and testing will be required for an Airbus A350 aircraft that has part 25 special conditions on stall testing for airplane certification.

The FAA notes that Public Law 111-216 and the Crewmember and Aircraft Dispatcher Training final rule require training to be conducted to a stall. The primary purpose for the training is to provide flight crews with experience in recognizing the cues of an impending stall, as well as reinforcing the recovery techniques learned in stall prevention training. To expose flight crews to these stall identification cues, envelope protections systems must typically be disabled in training. Unlike most envelope protection systems, stick pushers are typically installed to either compensate for an inability of the aircraft to meet the part 25 stalling definitions in § 25.201 or the stall characteristics requirements in § 25.203. Where a stick pusher is installed to meet the stall identification requirements of § 25.201, the activation of the stick pusher provides the pilot with a clear and distinctive indication to cease any further increase in AOA. This “clear and distinctive” indication of a stall is necessary to accomplish the training objectives and simply reaching the AOA limits of the envelope protection or “alpha floor” on an envelope protected aircraft will not provide the stall recognition cues that a pilot needs to learn to prevent and recover from a full stall in the event that the envelope protection systems fail. The accident and incident record contains multiple instances of stall envelope protection

system failures in the past, some of which progressed into a full stall situation where recognition cues of the stall were not identified by the flight crews.

One such example is the June 2009 crash of Air France flight 447, an Airbus A330-203 that experienced failure of the high angle of attack (stall) protection system due to the loss of airspeed data as a result of pitot probe blockage. See “Final report on the accident on 1 June 2009 to the Airbus A330-203 registered F-GZCP operated by Air France flight AF 447 Rio de Janeiro—Paris”; Bureau d'Enquêtes et d'Analyses (BEA); Paris, France. Another example is the December 2014 crash of Indonesia Air Asia flight 8501, an Airbus A320-216, where flightcrew actions to correct a malfunctioning flight augmentation system resulted in the loss of stall protection. See “Aircraft Accident Investigation Report; PT. Indonesia Air Asia; Airbus A320-216; PK-AXC”; Komite Nasional Keselamatan Transportasi (KNKT), Republic of Indonesia 2015.

The FAA further notes that the FSTD qualification requirement for objective and subjective testing of the stall is not new with this rulemaking. The part 60 standard published in 2008 contains both objective and subjective testing of the stall to include the “g-break” and is required for computer controlled aircraft in a non-normal operational mode.

Furthermore, the FAA's FSTD qualification standards dating back to AC 121-14C (1980) have also had both objective and subjective testing requirements for stall.

As a result, virtually all of the currently qualified Level C and Level D FSTDs for transport category aircraft have objective testing already in place for stall maneuvers in their FAA approved Master Qualification Test Guide (MQTG) and most of these objective tests are validated against flight test data collected up to and including the stall. The FAA finds that reducing these requirements would not support the full stall training requirements in the Crewmember and Aircraft Dispatcher Training final rule and therefore maintains that the requirements set forth in this final rule are necessary.

See 14 CFR part 60 (2008), Appendix A, Table A2A, test 2.c.8 (Stall Characteristics) and Table A3A, test 6.a. (High angle of attack, approach to stalls, stall warning, buffet, and g-break . . . .”.

Advisory Circular (AC) 121-14C (1980), “Aircraft Simulator and Visual System Evaluation and Approval”.

b. Validation of Stall Characteristics Using Flight Test Data

In the NPRM, the FAA proposed objective testing of stall characteristics for computer controlled aircraft in both normal mode and non-normal mode flight conditions up to the full stall through recovery to normal flight.

Embraer commented that during the developmental flight test campaign, full aerodynamic stalls that are considered hazardous or impractical can only be done if the aircraft is equipped with additional safety features, such as a tail parachute or other equivalent device, and those features obviously change the aircraft behavior during stall recovery if they are employed. Additionally, Embraer emphasized that for safety reasons in the certification flight test campaign, depending upon the aircraft's aerodynamic characteristics during stalls; full aerodynamic stall flight tests are not done in control states in which the stall protection system is not available. Embraer recommended that flight testing for validation should not be required for objective testing in non-normal control states where the stall protection system is not available.

As previously stated, the non-normal control mode objective testing to a full stall has been required in the existing part 60 stall characteristics objective tests as well as in previous FSTD evaluation standards dating back several years and the FAA has not significantly changed this requirement in this rulemaking. The FAA agrees with Embraer that aerodynamic stall flight testing may be hazardous or impractical to conduct in some circumstances (on both envelope protected and non-envelope protected aircraft) and this rulemaking has not specifically required additional flight test validation data to be collected at an AOA beyond where it is reasonably safe to do so.

As described in the NPRM, the FAA has included allowances for aerodynamic stall models to be developed and validated using engineering and analytical methods. While the FAA agrees with the commenter that some airplane certification flight test data collected in a stall maneuver may not be suitable for simulator modeling and validation purposes (such as where a tail parachute has been deployed as mentioned by the commenter), other flight testing conducted to investigate the stall characteristics of the airplane during the aircraft certification program may be used to develop engineering simulator models. Where significant safety issues would prevent flight testing at an AOA beyond the activation of a stall protection system, engineering simulator validation data will be acceptable for FSTD objective testing purposes. The FAA has made amendments in the final rule to make this clarification.

c. Required AOA Range for Normal Mode Objective Testing

In the NPRM, the FAA did not specify a particular AOA range to support the normal mode testing requirements for stall characteristics on computer controlled aircraft.

Delta and Airlines for America (A4A) requested clarification on what will be the required AOA range for objective testing on aircraft with highly automated systems where the aircraft does not reach aerodynamic stall in “normal control state.”

The FAA has not specified a particular AOA range to support the normal mode testing requirements in this final rule, as this will be a subset of the AOA range required for non-normal mode testing. Public Law 111-216 and part 121, subparts N and O, require training for recoveries from stalls and stick pusher activations, if equipped. In order to conduct stall recovery training, the protections of an envelope-protected aircraft must be disabled. As such, aerodynamics outside of the envelope protections up to ten degrees beyond the stall AOA must be considered to allow for stall recovery training in the event the envelope protections fail.

3. Data Sources for Model Development and Validation

a. Define Best Available Data

In the NPRM, the FAA proposed that where limited data is available to model and validate the stall characteristics of the aircraft, the data provider is expected to develop a stall model through analytical methods and the utilization of the “best available data”.

Bihrle Applied Research (Bihrle), A4A, and an anonymous commenter stated that the term, “best available data” (with regards to the aerodynamic data used to model and validate the stall model) is ambiguous and open to interpretation. American Airlines (American), FlightSafety International (FlightSafety), A4A, JetBlue Airways (JetBlue), and Delta further requested clarification from the FAA on whether a “non-OEM” provided source of data would be acceptable to the FAA to meet the representative stall model requirements.

The FAA notes that there is not a specific requirement currently in part 60, nor has a new requirement been introduced in this final rule that mandates FSTD sponsors use the original equipment [aircraft] manufacturer's (OEM) data to develop and validate the aerodynamic and flight control models in qualified FSTDs. As described in § 60.13(b), “The validation data package may contain flight test data from a source in addition to or independent of the aircraft manufacturer's data in support of an FSTD qualification . . .” There are

numerous FSTDs that have been qualified up through Level D where the FSTD manufacturer or other third party data provider has instrumented and flight tested an aircraft in order to collect flight test data to develop and validate their own aerodynamic and flight control models to support FSTD evaluation and qualification.

The FAA has considered the issues involved with requiring aircraft OEM data to develop and validate stall models for the purpose of conducting full stall training. While flight test data collected by the aircraft manufacturer will generally be the preferred source of data to model and validate FSTDs for training, the FAA has determined that “non-OEM” sources of aerodynamic data must be considered for the following reasons:

i. Restricting the development of stall models to that of the airplane manufacturers could impose a high cost on the FSTD sponsors and may not be possible in some instances where the airplane manufacturer does not support a simulator data package or is no longer in existence;

ii. Recommendations by the SPAW ARC, ICATEE, and other working groups have supported the use of analytically developed “type representative” stall models for training purposes; and

iii. An FAA simulator study

has supported the SPAW ARC's findings and found that analytically derived “type representative” stall models that are developed by third party data sources and thoroughly evaluated by a SME pilot can be effectively used to support stall training tasks in a simulator.

Schroeder, J.A., Burki-Cohen, J., Shikany, D.A., Gingras, D.R., & Desrochers, P. (2014). An Evaluation of Several Stall Models for Commercial Transport Training.

AIAA Modeling and Simulation Technologies Conference.

For these reasons, the FAA finds that it would not be practical to require FSTD sponsors to use an aircraft manufacturer's high AOA/stall model to meet the requirements of this final rule and other source data may be acceptable. Furthermore, Boeing, A4A, and an anonymous commenter stated that “flight test data should be noted as the preferred source of data, if available, with other data sources to be used if acceptable to the FAA.” The FAA concurs with this statement. To manage unknown risks, an aircraft manufacturer provided stall model developed with flight test data will generally be the preferred source of data; however, the FAA has concluded that there is not sufficient evidence to warrant mandating a particular source of data for model development. The FAA acknowledges that the term, “best available data” is ambiguous and has removed that language in the final rule.

b. Post Stall “Type Representative” Modeling

In the NPRM, FAA indicated that flight crews should be provided with practical experience in recognizing a full stall should the stall warning system become ineffective. To support this objective, the FSTD must provide critical aircraft type-specific stall recognition cues to enable the crew to recognize the onset of a stalled flight condition. Where data limitations and aircraft behavior may prevent conducting precise objective validation of post-stall behavior in the FSTD, the FAA included provisions in the proposal for “type representative” modeling and validation. To distinguish between the objectively validated “type specific” pre-stall modeling and post-stall modeling that may be developed through engineering analysis and SME pilot evaluation, the FAA used the term “type representative” in the NPRM.

Delta, FlightSafety, and A4A requested that the FAA better define the term, “type representative” with regards to post stall model fidelity.

In defining the FSTD fidelity requirements for full stall behavior, the FAA considered the primary training objectives for such training. The first objective of stall training is to provide flight crews with practical experience in recognizing a full stall should the stall warning system become ineffective (either through malfunction or human error). To support this objective, the FSTD must provide critical aircraft “type specific” recognition cues of an impending stall. Examples include cues such as reduced lateral/directional stability, deterrent stall buffet, and reduced pitch control if the particular aircraft has these cues.

The second objective of stall training is to reinforce to flight crews that the recovery procedures learned during stall prevention training are the same procedures needed to recover from a full stall. From an aerodynamic modeling standpoint, this presents a more significant challenge for two reasons. First, aircraft behavior in an aerodynamic stall may not be stable and is often sensitive to initial conditions, which creates the impression of non-repeatable chaotic behavior. Second, because this occurs in a flight regime with reduced stability, there can be practical limitations on the amount of flight test data that can be safely collected for simulator modeling and validation purposes. It is for these reasons that objectively validated “type specific” behavior at an AOA beyond the aerodynamic stall may not be a reasonable goal for defining fidelity in a training simulator.

The FAA has determined that the primary training objective for stall training is to have a pilot learn the proper stall recovery procedure in response to the variety of stall cues that a particular aircraft presents. Owing to the reduced stability, unsteady aerodynamics, and surface and rigging variations that occur with use, an aircraft will respond differently from stall to stall. However, the physics of what can happen in a stall are known, accepting that they can differ from aircraft to aircraft. The FAA has concluded that if a pilot can demonstrate applying the stall recovery technique for the general characteristics of what might occur for an aircraft type, the precise characteristics are not required. That is, if an airplane typically rolls 10 degrees left or 20 degrees right in a stall does not matter as long as the pilot does not incorrectly apply the stall recovery technique by responding to that roll before reducing AOA. What is important is to present roll if an aircraft has rolling tendencies to ensure that a pilot responds properly.

In order to avoid confusion with other uses of the word “representative” with respect to simulator fidelity, and to remain consistent with the ICAO 9625 definitions, the FAA has changed the description of the post-stall fidelity requirements to “sufficiently exemplar of the airplane being simulated to allow successful completion of the stall entry and recovery training tasks.” For the purposes of stall maneuver evaluation, the term “exemplar” is defined as a level of fidelity that is type-specific of the simulated airplane to the extent that the training objectives can be satisfactorily accomplished.

c. Use of Flight Test Data and Availability

In consideration of the recommendations of the SPAW ARC as well as the results of the FAA stall study, the FAA proposed that the necessary levels of simulator fidelity (including type specific pre-stall behavior and type representative post-stall behavior) can be achieved through a combination of engineering analysis, SME pilot assessment, and improved pre-stall objective testing through the use of existing stall flight test data that is already required by part 60 and

previous simulator standards.

Furthermore, the FAA proposed additional objective testing requirements for stall characteristics to include turning flight stall and high altitude cruise stall. In the proposal, these tests were also included in the FSTD Directive as applicable to previously qualified FSTDs.

14 CFR part 60 (2008) currently requires stall characteristics objective testing that extends to the full stall and “g-break”. Similar requirements exist for grandfathered simulator standards dating back to AC 121-14C (1980).

Dassault Aviation (Dassault) commented on the availability of full stall flight tests and that flight test points may not be available for some conditions where aircraft certification does not require them. Dassault further commented that corresponding flight test points might be implemented in the devices where partial data is available; however, no extension or extrapolation should be considered as type representative because this might lead to a very different behavior. An anonymous commenter made similar comments in that “unless there is a source of flight test data in every possible combination of conditions that might exist in a full stall, a demonstration of recovery techniques in a given set of conditions is the only plausible solution.”

FlightSafety further questioned whether there would be a release from liability should a stall model developed through engineering judgment and analytical methods prove to be inadequate.

As stated in previous sections, the FSTD qualification standards have had objective testing requirements for flight maneuvers up to and including full stall since 1980, so nearly all currently qualified full flight simulators (FFS) already have full stall flight test points that are used for simulator validation purposes. For previously qualified FSTDs, this data could be used to further improve existing stall models to meet the requirements of this final rule. The FAA does recognize, as Dassault points out, that additional flight test validation data may not readily exist to validate the new stall maneuvers introduced in the objective testing requirements (

e.g.,

cruise stall and turning flight stall). To address this concern, the FAA has amended the FSTD Directive for previously qualified FSTDs to remove the objective testing requirements for both the cruise condition and the turning flight stall condition and replaced them with subjective evaluation by an SME pilot. The remaining required objective testing stall characteristics tests (second segment climb and approach or landing conditions) are already required under the existing part 60 rule and should have existing validation data that can be used to meet the new objective testing requirements. Where limitations exist in the stall aerodynamic model due to the lack of data or reliable analytical methods, the data provider may declare these limitations as part of the required aerodynamic modeling SOC for the purposes of restricting the FSTD to certain stall maneuvers.

In response to FlightSafety's comment, the FAA notes that engineering judgment and analytical methods are used extensively in other areas of a simulation model besides stall and these models are used for training in conditions and situations that vary from the flight conditions used to validate the model. This practice has proven satisfactory, as known physical principles are used by FSTD manufacturers and data providers to represent the training conditions that vary from the flight-validated conditions. The FAA issues standards for FSTD evaluation, but generally does not prescribe specific methods for developing simulation models. The FAA does not have the authority to declare a release from liability.

4. Qualification on FSTD Levels Other Than Level C and Level D

In the NPRM, the FAA proposed modifications to the Level A and Level B stall qualification requirements to include stick pusher system force objective testing and updated objective and subjective testing requirements for the approach to stall flight conditions for newly qualified FSTDs.

Boeing, Delta, and A4A commented that while the FAA proposed modifications to the Level A and Level B stall qualification requirements, the Crewmember and Aircraft Dispatcher Training final rule does not permit such training in these devices and therefore these requirements should be removed. Delta and Boeing had additional comments concerning new requirements proposed for the “approach to stall” objective tests on Level A and Level B simulators (including additional configurations, tolerances, and subjective testing of the autoflight/stall protection systems) with one commenter stating that there is no apparent explanation why the approach to stall characteristics objective test has changed for Level A and Level B simulators and it should remain unchanged to be consistent with the ICAO 9625 document.

The FAA concurs with the commenters in that § 121.423 requires extended envelope training be conducted in a Level C or Level D simulator and has removed the associated minimum requirements for full stall on Level A and Level B simulators. However, the FAA notes that such devices are qualified to conduct stall prevention training at AOAs below that of the activation of the stall warning system and improving the validation of these FSTDs in the approach to stall flight condition would be beneficial to this training. Where new testing requirements were proposed for Level C and Level D simulators for AOAs below the activation of the stall warning system, these testing requirements were carried over to Level A and Level B simulators to provide better validation of the simulator to conduct stall prevention training tasks. The FAA further notes that these requirements for Level A and Level B simulators are not retroactive requirements defined in the FSTD Directive and will only be required for Level A and Level B simulators that are initially qualified after this final rule becomes effective. The FAA does not believe these changes for Level A and Level B FSTDs will have an impact on the alignment with the ICAO document since the Level A and Level B FSTD levels in part 60 have no equivalent ICAO device level.

5. Motion Cueing System Limitations

In the NPRM, the FAA included provisions to allow the FSTD manufacturer to limit the maximum buffet based on “motion platform capabilities and limitations” (see Table A2A, Entry No. 2.c.8). A similar provision was also included in the ICAO 9625, Edition 4.

The FAA received several comments that the FSTD sponsors, in addition to the device manufacturers, should be allowed to limit maximum buffet based upon motion platform capabilities and limitations. Furthermore, Delta, Boeing, FlightSafety, A4A, JetBlue, and United Parcel Service (UPS) commented that FSTD sponsors should have the ability to tune down or otherwise reduce motion vibrations due to maintenance and reliability aspects, personnel safety, and limitations of other simulator components, such as visual display systems and other hardware onboard the simulator. Boeing additionally commented that other simulator systems, such as the visual system, may also limit the buffet levels.

With regards to reducing or otherwise limiting motion vibrations that are within the motion platform's capabilities and limitations, the FAA has determined not to include specific

provisions to allow for arbitrary reductions in stall buffet from the levels that are evaluated through SME pilot assessment or objective testing. On many aircraft, the stall buffet is an important cue of an impending stall and, in some cases, may be the only distinctive cue a pilot will receive before or during an actual stall. In an FAA stall study on its B737-800 simulator

in which the magnitude of the stall buffet cues had been modified and increased significantly, all ten of the participating test pilots who had stalled the B737 noted the importance of accurately presenting the strong buffet cues as a stall progresses. Furthermore, the importance of stall buffet in training has been emphasized numerous times by the various working groups that provided recommendations to the FAA on stall training and associated simulator fidelity. As such, the FAA has determined that to accomplish the intended training objectives to provide flight crews with accurate recognition cues of an impending stall, the stall buffet characteristics should be provided in the FSTD at a level that is representative of the aircraft as evaluated by an SME pilot.

Schroeder, J.A., Burki-Cohen, J., Shikany, D.A., Gingras, D.R., & Desrochers, P. (2014). An Evaluation of Several Stall Models for Commercial Transport Training. AIAA Modeling and Simulation Technologies Conference.

Furthermore, as cited in A4A's and American's comments, Schroeder did acknowledge in his paper that buffet levels are sometimes reduced in a simulator to extend component life; however, no such reduction in stall buffet was implemented for this experiment. In fact, overall buffet gains were increased by a factor of 2.5 in the simulator with no adverse effects noted after the completion of the five week experiment.

The FAA's CAE simulator was operated for an average of 8 hours per day for five weeks to conduct approximately 700 stall maneuvers which had significant buffet levels. The FAA estimated that this simulator was exposed to approximately 67 total minutes of stall buffet over this five week period of time, which is comparable to what a typical part 121 carrier's simulator may be exposed to over an entire year under the new training rule. There were no reports of equipment damage after the completion of the experiment.

The FAA acknowledges that the potential exists for increased maintenance and reliability issues due to the repeated exposure of the FSTD to stall buffet. The FAA concurs with Boeing's comment in that other simulator systems (

e.g.,

visual systems) may limit the maximum buffet levels that are possible in a simulator and the FAA has made changes in the final rule to reflect this. Particularly with visual display systems, notch filters are frequently employed to reduce the vibration output of the motion platform at or around a resonant frequency that would cause damage to visual system components such as a Mylar mirror. These methods have been employed in the past and will continue to be permissible to protect the simulator and its occupants from known system limitations where damage is likely to occur or occupant safety may be compromised.

Furthermore, given that these standards may be applied to previously qualified FSTDs where the original FSTD manufacturer may not be accomplishing and evaluating the modifications of the FSTD, the FAA agrees with the commenters that the ability to limit the maximum buffet due to motion platform and other simulator system capabilities and limitations should be extended to the FSTD sponsor. The FAA has amended the final rule to allow for the FSTD manufacturer or the FSTD sponsor to limit the maximum motion buffet levels as described in this section.

6. Subject Matter Expert Pilot Evaluation and Qualification

a. SME Qualifications and Experience

In the NPRM the FAA proposed that the SME pilot who conducts the subject evaluation of the FSTD's stall characteristics must have “. . . acceptable supporting documentation and/or direct experience of the stall characteristics of the aircraft being simulated” and have “knowledge of the training requirements to conduct the stall training tasks.” The additional requirements proposed in Attachment 7 of the NPRM further stated that that the SME pilot must have experience in conducting stalls in the type of aircraft being simulated and, where not available, experience in an aircraft with similar stall characteristics.

The FAA received several comments concerning the experience and qualification requirements for SME pilots. American, A4A, Delta, and FlightSafety requested clarification on whether the required SME must be a pilot who has flown a full stall in the airplane or a pilot who only has knowledge of training requirements to conduct the stall tasks. Delta and A4A also questioned whether there are any other SME experience requirements beyond conducting stalls in the aircraft being simulated, or in an aircraft with similar stall characteristics. A4A, Delta, and FlightSafety, further requested clarification on whether an SME pilot can gain the necessary stall experience in an audited engineering simulator or on another Level D FFS that has already been qualified for stall maneuvers.

The FAA maintains that the subjective evaluation of the aerodynamic stall model is a critical component in ensuring that the FSTD's stall characteristics are representative of the aircraft and support the training objectives. The FAA further maintains that for such a subjective assessment to have credibility, the pilot must have direct experience in conducting stall maneuvers in the aircraft being simulated or in a similar aircraft that is expected to share the same general stall characteristics.

The FAA acknowledges that the SME requirements in the NPRM were not clearly defined and has revised Attachment 7 of Appendix A of the final rule to better define these requirements. In particular, rather than just stating the stall experience must be in the “type of aircraft being simulated”, the FAA clarified this by stating that the experience must be “. . . direct experience in conducting stall maneuvers in an airplane that shares a common type rating with the simulated aircraft.” In instances where the stall experience is in a different make, model, and series of aircraft within a common type rating, the FAA clarified that differences in aircraft specific stall recognition cues and handling characteristics must be addressed using available documentation such as aircraft operating manuals, aircraft manufacturer flight test reports, or other documentation that describes the stall characteristics of the aircraft.

Particularly for aircraft that are no longer in production, the FAA recognizes that there may be practical limits in finding SME pilots with the required experience to conduct the stall model evaluations. In instances where an acceptable SME cannot be reasonably located, the FAA has included deviation authority in the final rule for a sponsor to propose alternate methods in conducting the SME pilot evaluation of an FSTD's stall model.

In response to the comments concerning whether the SME pilot is required to have experience in the stall characteristics of the aircraft or knowledge of the training requirements to conduct the stall training tasks, the FAA has determined that the SME pilot must have both aircraft experience and knowledge of the training requirements, with the exceptions on experience as noted previously. While an important element of the subjective assessment is the comparison of the FSTD's performance against that of the aircraft, knowledge of the training tasks to be conducted in the FSTD should be

considered when conducting these evaluations. The recognition cues and handling qualities of an airplane can change significantly as a function of the aircraft configuration and how the stall is entered. To ensure the model can support the training objectives as well as to communicate any known or potential deficiencies in the model, the SME pilot conducting this subjective evaluation should focus the evaluation on those general aircraft configurations and stall entry methods that will likely be used in training. The FAA has clarified this language in the SME pilot evaluation requirements in Attachment 7.

The FAA has considered whether an SME pilot can gain experience in an audited engineering simulator or another Level D FFS that has been qualified for full stall maneuvers and has concerns that the effectiveness of an SME pilot evaluation may be diminished when making such comparisons from simulator to simulator without an objective measure to ensure that the aerodynamic model from the engineering simulator has been properly implemented on the training simulator. For these reasons, the FAA maintains that the SME pilot conducting the subjective evaluation of the FSTD or associated stall model must have direct experience of the stall in the aircraft. A pilot cannot gain the necessary aircraft experience required to be a SME in an engineering simulator or another FFS that has been qualified for full stalls.

b. Model Validation Conducted by the Data Provider

Boeing and Airbus commented that in lieu of an SME pilot evaluation being conducted on the individual FSTDs for initial and recurrent evaluations, the model validation with the SME pilot can be conducted by the data provider where objective stall data is provided to validate the individual FSTDs. Delta and A4A made similar comments. The FAA agrees with the commenters and notes that provisions to conduct the SME pilot evaluation on an engineering simulator were included in the proposal in Attachment 7 to Appendix A. The FAA maintains that where objective proof of match tests are provided to verify the models have been properly implemented on the training FSTD (including stall characteristics and stall buffet objective testing), the FAA will accept an SOC from the data provider that confirms the integrated stall model has been evaluated by an SME pilot on an engineering simulator or other simulator acceptable to the FAA. Furthermore, there is no intent to require that this SME evaluation be conducted annually, and the SOC that confirms this SME assessment has taken place will remain valid as long as the stall model remains unmodified.

c. NSPM Process for Evaluating and Accepting an SME Pilot

In the NPRM, the FAA proposed that an SOC be provided to the FAA that confirms that the FSTD has been evaluated by an SME pilot. This requirement was proposed to apply to both newly qualified FSTDs as well as previously qualified FSTDs that are evaluated under the requirements of FSTD Directive No. 2.

Delta and A4A requested clarification on this process that the NSPM follows to evaluate and accept an SME pilot.

As described in FSTD Directive No. 2 and Attachment 7 to Appendix A, the process for the qualification of stall maneuvers requires that the sponsor submit an SOC to the NSPM confirming that the FSTD has been evaluated by a SME pilot with the required experience. The NSPM will review this SOC to verify that the evaluating SME pilot has the required experience as specified in the rule before issuing additional qualification for full stall training tasks. Additionally, requests for deviation from the SME experience requirements as described in Attachment 7 should be submitted to the NSPM when requesting additional qualification for full stall training tasks. Where specific questions arise, the NSPM will contact the sponsor or data provider directly for clarification.

7. Alignment With ICAO 9625, Edition 4, on Stall and Stick Pusher Requirements

The FAA's proposal for the stall and stick pusher requirements were primarily based upon the recommendations from the SPAW ARC, as well as other working groups such as ICATEE and the LOCART working group. After the FAA first initiated this rulemaking, the ICATEE recommendations that were considered by the FAA in developing the proposal were also considered by ICAO for updating the ICAO 9625 document to include FSTD evaluation standards for stall and upset prevention and recovery training.

The FAA received numerous comments that some of the general requirements and objective testing requirements in the proposal did not align with the ICAO 9625, Edition 4 requirements, which became available following the publication of the NPRM. A4A, Boeing, and an anonymous commenter indicated that the stick pusher requirements (Table A1A, Entry No. 2.1.7.S) in the NPRM should be relocated to the flight controls section where they are more applicable. Boeing and A4A also commented that the stall buffet onset measurements in the stall characteristics objective tests (Table A2A, Entry No. 2.c.8) are based upon speed rather than AOA like ICAO 9625, Edition 4. Delta, A4A, and an anonymous commenter indicated that the control force tolerances in the stall characteristics test should be applicable only to aircraft with reversible flight control systems. Finally, A4A and Boeing commented that the required test conditions for the stall buffet motion characteristics test (test 3.f.8 in Table A2A of the NPRM) do not include the same conditions as ICAO 9625, Edition 4.

The FAA was unable to fully participate in the ICAO deliberations due to ex parte concerns as the agency was engaged in this rulemaking proceeding. The FAA has had an opportunity to review the final release of the ICAO 9625, Edition 4, document and has found that only minor differences exist with regards to the stall qualification requirements as compared to the final rule. As such, in order to maintain alignment with the ICAO document as identified by the commenters, the FAA has incorporated the ICAO language into the final rule to the maximum extent possible. The FAA has amended the final rule by adopting much of the ICAO language for high AOA/stall modeling minimum requirements (Table A1A, Entry No. 2.m. in the final rule) as well as the stall characteristics objective test tolerances and flight conditions (Table A2A, Entry No. 2.c.8.a in the final rule).

The FAA did not, however, amend the required conditions for the stall buffet tests to align with the ICAO 9625 standard. As recommended by the SPAW ARC report, stall buffet evaluation should include a broader range of flight conditions than what is currently evaluated. The FAA has determined that the inclusion of the second segment climb condition is important to evaluate the differences in stall buffet vibrations at high power settings, particularly for turboprop airplanes. As a result, the FAA has maintained this is as a required condition for the stall buffet characteristic vibrations test (Table A2A, Entry No. 3.f.5).

While the FAA has aligned a majority of the general requirements and the objective testing requirements with the ICAO document, specific differences must be maintained in the final rule to address comments received on the proposal as well as retroactive FSTD

evaluation requirements that are required to support the mandated training for United States (U.S.) air carriers.

8. Requirements for Previously Qualified FSTDs

a. Stall Buffet Objective Testing

In the proposal, the retroactive requirements for previously qualified FSTDs, as described in FSTD Directive No. 2., did not include objective testing for stall buffets.

Boeing, Delta, A4A, and an anonymous commenter stated that the general requirement and objective testing requirements (Table A1A and Table A2A, respectively) for stall buffet vibration measurement state that these tests are required for all FSTDs qualified for stall training tasks. This is in conflict with the proposed FSTD Directive No. 2, which specifically states that stall buffet objective vibration testing is not required for previously qualified FSTDs.

In recognizing the potentially high cost of gathering additional flight test validation data for stall buffets, the FAA did not include this requirement in the proposed FSTD Directive No. 2 retroactive requirements for previously qualified FSTDs. Since changes to the QPS tables are not typically applicable to previously qualified FSTDs, changes to Table A1A or Table A2A are not necessary since all of the retroactive requirements are defined in FSTD Directive No. 2. The FAA has added language in FSTD Directive No. 2 in the final rule to clarify the retroactive testing requirements.

b. FSTD Directive No. 2 and Grandfather Rights

In FSTD Directive No. 2, previously qualified FSTDs that will be used to conduct full stall, UPRT, and other specific training tasks will be required to meet certain sections of the general requirements, objective testing requirements, and subjective testing requirements of the updated QPS tables in order to obtain qualification for these training tasks.

A4A requested clarification on whether FSTDs that are “upgraded” to provide extended envelope training would also have to comply with the proposed ICAO alignment requirements as well (such as the new visual display system requirements). American and A4A further noted that some sections within the QPS tables appear to have been mistakenly applied to all simulators instead of those qualified after the effective date of the final rule.

The FAA notes that the only new QPS requirements applicable for previously qualified FSTDs are those that are described in FSTD Directive No. 2. As described in § 60.17 and paragraph 13 of Appendix A, previously qualified FSTDs will continue to hold grandfather rights and the changes to the QPS tables will not generally be applicable to previously qualified devices unless specifically stated in an FSTD Directive. The FAA has reviewed FSTD Directive No. 2 and made amendments in the final rule to clarify which sections of the QPS appendices will be applicable to previously qualified devices.

The FAA further notes that an “upgrade,” as defined by part 60, is an “improvement or enhancement of an FSTD for the purpose of achieving a higher qualification level.” FSTDs that are upgraded in qualification level will generally have to comply with the standard that is in effect at the time of the upgrade. It is important to note, however, that compliance with FSTD Directive No. 2 does not require a change in qualification level and is not considered an “upgrade” under part 60. As a result, the other changes made to the QPS appendices, including the general changes made to align with the ICAO document, will not be applicable to previously qualified FSTDs unless upgrading in FSTD qualification level.

9. Applicability of Stall and UPRT Requirements on Newly Qualified FSTDs

In the NPRM, the FAA proposed that the minimum requirements for the evaluation of full stall maneuvers and UPRT maneuvers would be applicable for all fixed wing Level C and Level D FSTDs that are initially qualified after the final rule becomes effective.

Dassault commented that while UPRT and full stall training will become mandatory for part 121 operators, it is not clear if this applies to part 135 and part 91 operators as well. Dassault further questioned whether the objective testing requirements for full stall maneuvers would be required for an FSTD that will not be used for full stall training. Finally, Dassault commented that they would prefer the requirements to be applied to new or modified aircraft types instead of new FSTDs since this would allow collecting necessary data at the time of the type certification flight tests.

CAE made similar comments that point out that the FSTD Directive (for previously qualified devices) is only applicable for those FSTDs that will be used to conduct such (UPRT and stall) training, however, the requirements in the QPS appendices are mandatory for newly qualified FSTDs regardless of whether they are used in an air carrier or a non-air carrier training program. CAE recommended that operators of newly qualified FSTDs (that are initially qualified after the final rule becomes effective) who are not subject to the Crewmember and Aircraft Dispatcher Training final rule should also be given the same option on whether or not to invest in the additional features that support extended envelope and other tasks as required under the final rule. CAE further stated that this would provide an option to those operators who may have multiple devices to limit such updates to certain equipment that will be utilized to conduct such training.

FAA agrees with the commenters that the requirement for FSTD modifications and data collection should not be imposed on sponsors who will not use those FSTDs to conduct full stall training and have no mandate to conduct such training. Similar to the FSTD Directive for previously qualified FSTDs, the FAA has amended the final rule to make the qualification of full stall maneuvers optional for newly qualified FSTDs. This will allow flexibility for operators to decide how many FSTDs need to be evaluated for full stall maneuvers to support training requirements.

FAA has, however, maintained the minimum requirements for UPRT evaluation on newly qualified Level C and Level D FFSs. The FAA has estimated that the addition of such IOS feedback tools to support UPRT would add little to no incremental cost to that of a newly qualified FSTD and will enhance instructor awareness in support of the existing part 60 unusual attitude qualification requirement.

14 CFR part 60, Appendix A, Table A1B, Entry No. 3.f., “Recovery From Unusual Attitudes”. This minimum qualification requirement covers maneuvers that are “within the normal flight envelope supported by applicable simulation validation data.”

In order to ensure that only FFSs that are evaluated and qualified for stall training tasks are used for such training, compliance with the stall and UPRT evaluation requirements will be tracked by the FAA through modifications to the FSTD's Statement of Qualification (SOQ).

10. General Comments on Stall Requirements

a. Testing and Checking of Stall Maneuvers

Boeing commented that stall training beyond the stick shaker activation does not require testing or checking in part 121 and references made to testing and checking in FSTD Directive No. 2 should be removed.

FAA agrees with Boeing's comment and has modified the language in FSTD Directive No. 2 accordingly.

b. Interim FSTD Qualification for Stall Training

A4A commented that the FSTD Directive (for previously qualified FSTDs) requires evaluation by the NSPM for additional qualification and should allow a draft SOQ to be issued until the next scheduled evaluation.

FAA notes that FSTD Directive No. 2 does not require an update to the FSTD's permanent SOQ before stall training can be conducted in an FAA approved training program. A positive response from the NSPM to the FSTD modification notification confirming that the requirements of the Directive have been met will, in most cases, serve as an interim update to the FSTD's SOQ until the next scheduled FSTD evaluation. In some instances, however, additional FSTD evaluations conducted by the FAA may be required before the modified FSTD is placed into service. FAA has added clarifying language to the FSTD Directive that this response will serve as interim FSTD qualification for stall training tasks until the next scheduled FSTD evaluation where additional FSTD evaluations conducted by the FAA have been determined to not be required.

c. Aerodynamic Modeling Considerations

Frasca International (Frasca) commented that AOA rate is a significant contributor to stall behavior and should be considered as part of the requirement for aerodynamic stall modeling. FAA agrees with Frasca's comment and has added AOA rate to the list of aerodynamic modeling considerations in Attachment 7.

B. Evaluation Requirements for Upset Prevention and Recovery Training Tasks

In order to support UPRT that was introduced in the Crewmember and Aircraft Dispatcher Training final rule, the FAA proposed new FSTD evaluation requirements for these training tasks. The proposed requirements were based upon recommendations from the LOCART and ICATEE working groups as well as from the guidance in the Airplane Upset Recovery Training Aid (AURTA), and included new standards to better define the FSTD's aerodynamic validation envelope. The proposal also included requirements to improve the feedback at the instructor operating station (IOS) concerning the FSTD validation envelope limits, aircraft operational limits, and flight control inputs by the trainee.

1. UPRT Qualification on Lower Level FSTDs

In the NPRM, the FAA proposed minimum qualification requirements for full stall and UPRT in the newly defined Level 7 flight training device (FTD) (Table B1A of Appendix B).

TRU Simulation and A4A commented that the proposal requires extended envelope modeling for the Level 7 FTD, but the part 121 training requirements have a minimum requirement that this training must be conducted in a Level C or higher simulator. In addition, A4A commented that this is inconsistent with ICAO 9625, Edition 4, where UPRT training is only qualified on a Type VII device. Finally, Air Line Pilots Association, International (ALPA) commented that training could be negatively impacted if allowed to be conducted on a Level A or Level B FFS as the proposal states and this is inconsistent with the recommendations of the SPAW ARC.

FAA agrees with A4A and TRU Simulation regarding UPRT qualification on a Level 7 FTD. This was an error in the proposal and the FAA has amended the final rule to remove minimum qualification requirements for both full stall and UPRT on the Level 7 FTD.

The FAA has reconsidered the qualification of Level A and Level B FFSs for UPRT tasks that involve no bank angle excursions, such as nose-high or nose-low upsets, as defined in the NPRM, and amended the final rule by removing references to full stall and UPRT evaluation requirements for Level A and Level B FFSs in the FSTD Directive.

The FAA notes that the primary differences between the Level A and Level B minimum qualification requirements compared to the Level C and Level D qualification requirements are generally limited to ground reaction modeling, visual system field of view requirements, and minimum motion cueing requirements. The ground reaction modeling requirements have no impact on UPRT or stall training given that training is typically conducted well outside of ground effect. There are significant differences in the motion cueing abilities between Level A and Level B FFSs versus Level C and Level D FFSs that impact the ability for effective full stall and upset training to be conducted in the lower level devices. Level A and Level B FFSs have a 3 degree-of-freedom (DOF) motion cueing system compared to the 6-DOF motion cueing requirement for Level C and Level D FFSs. Typically, a 3-DOF motion cueing system includes motion cues in the pitch, roll, and heave axes.

For wings-level maneuvers, such as the nose-high or nose-low upsets, the dominant motion cues during the stimulation of such an upset will typically be limited to the pitch and heave axis with little activity in the other axes. Because there may be considerable variation in how each pilot responds to an upset in training, other cues may be introduced during the recovery maneuver that are outside of the capability of a Level A or Level B FFS. Furthermore, a wings-level stall entry may result in considerable lateral-directional accelerations on airplanes that are unstable at the stall. These cues will generally be outside the capability for a Level A or Level B FFS with a 3-DOF motion cueing platform to reproduce; therefore, evaluation of full stall and upset in these devices would not be appropriate in most cases.

See 14 CFR part 60, Table A1A, entry 5.b.

FAA adds that while the qualification of extended envelope training tasks will generally be applicable only to Level C and Level D simulators, operators of other FFSs have the option to apply for FAA consideration of a deviation from the use of a Level C or Level D simulator for extended envelope training tasks as described in § 121.423(e). Since the approval of such a deviation will be linked to the training program and the alternate means that are proposed to achieve the required learning objectives, approvals to deviate from the Level C or higher requirements in § 121.423 will have to be reviewed on a case-by-case basis under the deviation authority.

2. Record and Playback Requirements for UPRT

In its proposal, the FAA included minimum requirements for a means to record and playback audio and video as well as a means to record and playback certain parameters for the qualification of UPRT maneuvers.

American, Boeing, Delta, A4A, FedEx, JetBlue, and an anonymous commenter stated that the requirement for record and playback functionality is outside the scope of the part 60 rule and does not provide additional benefits to the training scenario. While the commenters generally agreed with having parameters available to the instructor during the scenario, such as the aerodynamic validation envelope and the aircraft operational limits, the recording and playback of parameters, particularly the recording and playback of audio and video, should be left to the discretion of the operator. Both ALPA and A4A further commented that there are union and collective bargaining agreements to

consider with videotaping flight crews in training. Additionally, several commenters noted that there is a high cost burden with requiring the audio and video playback functionality and the requirement should be removed.

The FAA has reconsidered the instructor feedback requirements and agrees with the commenters that effective UPRT can be conducted without audio and video playback capabilities or with the use of an instructor off-board debriefing system located outside of the simulator for the purposes of replaying the training scenario after its conclusion. While the use of off-board debriefing tools and audio/video playback may enhance such training, the FAA recognizes that operators can still conduct effective training without them and has amended the final rule to remove the audio and video record and playback requirements.

3. Instructor Operating Station (IOS) Requirements

In the NPRM, the FAA proposed minimum requirements for a feedback mechanism, located on the IOS and available to the instructor, that provides a minimum set of parameters to display to determine expected FSTD fidelity, aircraft structural/performance limitations, and student flight control inputs. The FAA provided example IOS feedback displays in the information section of Attachment 7 to Appendix A. The proposal also included requirements for features or malfunctions to support the training of crew awareness, recognition, and recovery from an aircraft upset.

American and A4A commented that the UPRT requirements for upset “awareness” and “recognition” features and/or malfunctions are outside of the scope of the rule and emphasis should be placed on recovery from an upset. JetBlue made similar comments on this topic. Boeing further commented that how the training requirements are met should be at the discretion of the training program and is not pertinent to FSTD qualification. Since these features are not prescribed, they should appear in the information/notes column and not in the requirements column of Table A1A. Frasca additionally questioned what would be some examples of relevant data sources with respect to externally driven upset scenarios.

Regarding the IOS requirement to display “Cl-max”, A4A, Boeing, and an anonymous commenter stated that “Cl-max” is not an explicit output of most aerodynamic models and is not available for plotting on the IOS display. Similar comments concerning the use of “Cl-max” as an example of a limit were made by the NTSB. Boeing and FlightSafety also recommended changing the IOS feedback requirement from showing “aircraft structural/performance limitations” to showing “aircraft operating limits”. FlightSafety further commented that aircraft structural and performance limitations are not likely to be known or provided to simulator manufacturers or operators. Delta commented that as an alternative to the record and playback functionality, enhancing existing IOS functionality to include “FSTD crash” and freeze when g-load or control input parameters are exceeded would provide immediate information to the instructor. UPS made similar comments in that a flag could be added to the IOS for envelope excursion and a maximum load indication and that other feedback mechanisms are cost prohibitive and not needed.

The FAA agrees with the commenters in that mandating specific features and malfunctions to drive upset scenarios is generally outside the scope of part 60 and has removed these requirements in the final rule. The FAA further notes that specific guidance material on developing UPRT scenarios has been published as part of Advisory Circular (AC) 120-111, Upset Prevention and Recovery Training.

The FAA maintains that minimum feedback requirements have been found necessary to provide meaningful information to the instructor in training and evaluating pilots in UPRT maneuvers. The FAA recognizes that FSTD sponsors and operators may have other means to display this information and the example IOS displays provided in Attachment 7 are included in an information section as guidance material and are intended to be examples that could be used if desired. Digital or discrete IOS feedback mechanisms may prove to be acceptable for some or all parameters as Delta and UPS have suggested and, consequently, the FAA has not mandated a particular solution. The FAA has amended the final rule to allow FSTD sponsors the discretion to determine a feedback mechanism design that provides the required parameters needed for UPRT and supports their particular training programs and FSTD capabilities.

The FAA has further amended the final rule to remove the “structural/performance limitations” terminology and replaced it with “aircraft operational limitations” as suggested by the commenters. Additionally, the FAA has removed the feedback parameter, “Cl-max” as suggested by the commenters and replaced it with “stall speed” and “stall identification angle of attack” since these are more useful parameters for instructors to directly provide feedback to crew members when conducting UPRT and stall maneuvers.

4. Aerodynamic Source Data and Range of the FSTD Validation Envelope

a. FSTD Validation Envelope and Training Maneuvers

In the NPRM, the FAA proposed requirements to define the limits of the FSTD's validation envelope and test the FSTD against a minimum set of standard upset recovery maneuvers as defined in the AURTA.

Boeing, A4A, and an anonymous commenter stated that the term “extended envelope” in the general requirements is redundant because “modeling to the extent necessary. . . .” defines the requirement adequately. Boeing further commented that this phrase is a misnomer and implies that the flight model may need to be extended. For some upset recovery training, the existing model may be sufficient to support the training needs. A4A made similar comments stating that its experience has shown that the current data appears to be sufficient for conducting upset recovery training.

Airbus further commented that the evaluation of the FSTD should take into consideration the training practices recommended by the aircraft OEM. An anonymous commenter additionally stated that it is imperative that the validation limits are defined by the aerodynamic data provider since they are the only credible source for these limits.

FAA agrees that the term, “extended envelope” may be redundant in this particular context and has amended the final rule accordingly. The FAA recognizes that many aerodynamic models on existing FSTDs may currently be capable of conducting UPRT maneuvers within their AOA versus sideslip validation envelope with no need to be extended further as the commenters suggest. However, the range of validation envelopes can vary significantly between FSTDs as a function of the extent of flight test data, wind tunnel data, and other data used to develop the model. Since those validation envelopes have not been transmitted by the data providers to the FSTD operators in most cases, the FAA has determined that the comments are unsupported and have concluded that operators need to obtain the validation envelopes and ensure that their training maneuvers remain within them.

The FAA agrees with Airbus in that the evaluation of the FSTD should consider the training that will be conducted in the device. However, this rulemaking only addresses FSTD qualification standards and the FSTD evaluation requirements were primarily developed to support training as required by the Crewmember and Aircraft Dispatcher Training final rule and public law. In developing the FSTD evaluation standards for UPRT, the SPAW ARC recommendations, as well as the AURTA recommendations, were reviewed to define a standard set of upset recovery maneuvers that were needed to minimally qualify an FSTD for such training. This set of maneuvers is considered to be the minimum required for FSTD qualification that will provide a baseline evaluation of the FSTD's capabilities to conduct UPRT, but in no way limits an FSTD sponsor's decisions concerning which upset recovery maneuvers they incorporate into their training programs.

The FAA further notes that the qualification requirements for UPRT in this final rule exceeds the current part 60 FSTD qualification requirement for “recoveries from unusual attitudes” which limits maneuvers to “within the normal flight envelope supported by applicable simulation validation data.”

If a training provider, regardless of operational rule part, performs unusual attitude training

maneuvers that exceed the parameters that define an aircraft upset, that FSTD must be evaluated and qualified for UPRT. The FAA does not believe this will impose an additional cost burden on sponsors of previously qualified FSTDs since UPRT qualification is only required if the training provider chooses to conduct unusual attitude training that exceeds the defined upset conditions.

14 CFR part 60, Appendix A, Table A1B, Entry No. 3.f., “Recovery From Unusual Attitudes”.

Unusual attitude training is required training for an instrument rating, an airline transport pilot certificate, and an aircraft type rating.

The FAA generally agrees that the validation limits are best defined by the aerodynamic data provider and has provided clarification in Attachment 7 in Appendix A of the final rule; however, there may be instances where the original aerodynamic data provider cannot directly provide this information (the original data provider is either no longer in business or no longer supports the model) and the FSTD sponsor must determine the validation envelope using data supplied with the original aerodynamic data package. The FSTD sponsor will be required to define such aerodynamic data sources in the required SOC.

b. Expansion of the FSTD Validation Envelope Using Existing Flight Test Data

In the existing part 60 rule, the objective testing requirements found in Attachment 2 of Appendix A requires that testing be conducted in weights and centers of gravity (CG) conditions that are typical of normal operations. Furthermore, where such testing is conducted at one extreme weight or CG condition, a second test must be provided at “mid-conditions” or as close as possible to the other extreme condition.

Airbus and Boeing commented that the existing part 60 requirement for objective testing to be predominately conducted in mid-weight/mid-CG flight conditions is outdated and a wider coverage of the alpha/beta (

e.g.,

AOA versus sideslip) envelope may be accomplished using critical flight conditions testing during aircraft certification at extreme weight and CG combinations. Boeing additionally stated that while the current regulation supports this, it requires testing at the opposite extreme conditions which increases the burden on the sponsor. Airbus additionally commented that there is no need to have a global requirement for this because the weight/CG requirements can be specified for each test where relevant. CAE made similar comments on this issue.

FAA agrees with the commenters and supports allowing flexibility in providing the best range of data to support not only extended envelope training, but all training conducted in an FSTD. Where weight and CG configuration is critical for validating a particular flight maneuver (such as in some of the takeoff objective tests), those conditions are described as a test requirement for that particular test. In general, the FAA recognizes that weight and CG effects on the aerodynamic model are well known and requiring redundant test conditions at varying weight and CG ranges has questionable benefit for FSTD validation in some required objective tests. The FAA has amended the final rule as recommended by the commenters to allow for greater flexibility in determining appropriate weight and CG conditions for some of the required objective tests that do not have specific requirements contained within Table A2A.

5. General Comments on UPRT

a. FSTD Qualification and FAA Oversight

ALPA commented that while they support the requirements associated with the simulator providing feedback to the instructors and evaluators, they believe that only simulators that can perform all aspects of the new training required in the Crewmember and Aircraft Dispatcher Training final rule should be qualified. In addition, ALPA further stated that since the proposed rule only requires FSTD evaluation for those FSTDs used to conduct the additional training tasks, a robust oversight system will be needed to ensure that only the simulators qualified for this training are used in the required training.

In developing the proposed requirements in the NPRM, the FAA considered the economic costs and benefits of mandating FSTD modifications and evaluations to support training requirements. With the considerable cost in the implementation of new aerodynamic stall models on previously qualified FSTDs, the FAA could not justify imposing this cost on FSTD sponsors who currently do not have a mandate to conduct such training. Furthermore, the FAA determined that some FSTD sponsors that do have a training mandate for stall and UPRT may realize some cost savings by not having to qualify all of their FSTDs where the training can be accomplished on a lesser number of devices. Finally, with the large number of FSTDs that will require evaluation to meet the part 121 compliance date of March 2019, this may provide some practical relief in having to qualify all FSTDs within a relatively short amount of time.

The FAA appreciates ALPA's concern for proper FAA oversight to ensure that the FSTDs are evaluated and qualified before extended envelope training is conducted. The FAA notes that an oversight system to track FSTD qualifications is already in place with the list of qualified tasks that is currently required on the part 60 required SOQ for all FAA qualified FSTDs.

In the final rule, the FAA maintained the requirement in FSTD Directive No. 2 that the individual training tasks are to be reflected on the FSTD's SOQ once qualified. The FSTD's SOQ will then serve as a tracking mechanism to ensure the FSTD has been properly evaluated and qualified by the FAA NSP to conduct the individual training tasks. Furthermore, the FAA

will coordinate internally with Principal Operations Inspectors (POIs) to ensure that only FSTDs that are qualified in accordance with FSTD Directive No. 2 are approved for use in training those specific tasks as part of an FAA approved training program.

See § 60.17(b)

b. Maintenance Concerns

A4A commented that further testing is needed to ensure that the reliability and availability of FSTDs due to maintenance issues is unchanged with the addition of UPRT training.

The potential for stall vibrations to cause FSTD maintenance issues has been acknowledged and discussed in a previous section on stall buffet. The FAA acknowledges that conducting UPRT maneuvers in an FSTD can produce significant motion system excursions, however, the FAA is not aware of any evidence that the addition of general UPRT maneuvers will introduce significant maintenance issues that would affect the overall reliability and availability of an FSTD beyond what is normally seen in existing training. As with motion system tuning in general, the FAA expects that FSTD sponsors will employ limits and protections within their motion system hardware and software that will protect the FSTD from dangerous excursions that could damage the FSTD's equipment or injure its occupants. The exposure to stall buffet likely has the greatest potential for affecting an FSTD's reliability and the FAA has addressed this issue in the stall requirements sections.

C. Evaluation Requirements for Engine and Airframe Icing Training Tasks

In the NPRM, the FAA proposed changes to the general requirements for engine and airframe icing qualification as well as adding a new objective demonstration test for ice accretion effects for newly qualified FSTDs. The changes were based upon new icing requirements in the ICAO 9625 document, as well as recommendations made by the SPAW ARC, and were intended to improve upon the existing engine and airframe icing requirements in part 60. The proposed changes focused on requirements for improved ice accretion models that represent the aerodynamic effects of icing rather than estimating icing effects through gross weight increments.

1. Objective Demonstration Testing

a. Objective Demonstration Testing for Previously Qualified FSTDs

In the proposal, the FAA introduced new objective testing requirements for the demonstration of icing effects on Level C and Level D FFSs. The objective tests are intended to demonstrate that the aerodynamic effects of ice accretion are present in the simulation with the icing model active as compared to the simulation where no ice is present. Due to the potential cost impact for previously qualified FSTDs, these tests were not retroactively required in FSTD Directive No. 2.

Boeing commented that the objective demonstration test for engine and airframe icing is not required in FSTD Directive No. 2 (for previously qualified FSTDs) and recommended that text should be added to Table A2A (Entry No. 2.i.) to clarify that this test is not required for previously qualified FSTDs.

FAA agrees with Boeing in that this demonstration test for engine and airframe icing is not required for previously qualified FSTDs and has added clarifying language in FSTD Directive No. 2. As with comments in previous sections concerning stall buffet testing, previously qualified FSTDs will maintain grandfather rights and the modifications to Table A2A will generally not be applicable to previously qualified FSTDs unless specified in an FSTD Directive. As a result, FAA has not added additional text in Table A2A concerning previously qualified FSTDs because it will be adequately addressed in the FSTD Directive.

b. Icing Effects and Recognition Cues

In the proposed icing effects objective demonstration test, the FAA included specific icing effects that may be present and evaluated as applicable to the particular airplane type. This list included both aerodynamic effects of ice accretion as well as engine effects that may also be present with the icing model activated in the simulation.

Boeing commented that the objective demonstration test for icing includes engine effects, but the general requirement for icing does not specifically identify engine effects and this should be removed from the objective testing requirement. An anonymous commenter stated that it may be necessary to show engine effects and airframe effects of icing separately because the test will not differentiate between thrust losses and drag increases. Another anonymous commenter pointed out that changes in control effectiveness and control forces are limited mainly to reversible systems on certain airframe configurations and the FSTD should only introduce these changes when they are representative of the specific make and model of aircraft. Additionally, an anonymous commenter stated that there is “very little guidance on what engine icing effects should be represented and most manufacturers state there are little effects on engine indications for current turbofans. Based upon the data we do have for engine inlet icing, the effects are often very subtle, yet the requirements seem to ask for something more dramatic. If we modify our icing models to favor dramatic effects, do we risk training pilots to miss looking for the subtle indications?”

Concerning Boeing's comment, the general requirement for engine and airframe icing (Table A1A, Entry No. 2.j.) does include modeling the effects of icing on the engine, where appropriate, as does the current requirement in part 60. While the information section in the demonstration test does state “aerodynamic parameters,” the intent of the test is to demonstrate the effects of the icing model integrated into the simulation. If the sponsor designated icing model used for the demonstration test has an effect on relevant engine parameters (such as thrust reduction or other effects), these effects should also be shown as part of the test. FAA has amended the test details in the table to clarify this. Other icing models that may be optionally developed by the FSTD sponsor to train recognition of engine effects due to icing will not require separate objective demonstration testing.

The FAA agrees that icing effects should only be introduced where representative of the specific make and model of aircraft and has clarified this in Table A2A (test 2.i.) and Attachment 7 of the final rule. The FAA does not intend for a simulator operator to artificially insert dramatic icing effects that are not representative of the aircraft. While the FAA is aware that the cues of ice accretion can vary significantly depending upon the nature of the icing event and the aircraft's characteristics, the icing models developed for simulation and training purposes should support the general recognition of icing cues that are typical for the aircraft being simulated.

2. Requirements for Lower Level FTDs

In the NPRM, the FAA proposed general requirements and objective demonstration testing for engine and airframe icing as part of the new Level 7 FTD requirements in Appendix B.

TRU Simulation commented that in the proposal for ICAO 9625, Edition 4, only a Type VII is allowed for use in UPRT and this item (icing) is identified as only being required on devices where UPRT will be trained. TRU Simulation requested that the FAA confirm applicability on a Level 7 FTD and

remove the requirement if not. TRU Simulation and A4A further commented that the objective demonstration test for icing is not required for an ICAO 9625 Type V device and should be removed from the Level 7 FTD requirements. TRU Simulation and A4A additionally commented that a new requirement for Level 6 FTD was introduced to have the anti-icing system operate with appropriate effects upon ice formation on airframe, engines, and instrument sensors.

FAA reviewed ICAO 9625 Edition 4 and found that the general requirement for the modeling of icing (Appendix A, Entry No. 2.1.S.e.) is a minimum requirement for an ICAO 9625 Type V device and has therefore maintained this requirement for the FAA Level 7 FTD. FAA confirms that the objective demonstration testing for icing is not required for an ICAO 9625 Type V device and therefore has removed this requirement for the FAA Level 7 FTD in Table B2A to maintain consistency with the ICAO document.

Regarding the addition of anti-icing effects to a Level 6 FTD, FAA has removed the ICAO numbering system in the general requirements table that was published with the NPRM and restored the existing part 60 requirements for Level 6 FTDs. The FAA notes, however, that the existing part 60 functions and subjective testing requirements for Level 6 FTDs includes “operations during icing conditions” and “effects of airframe/engine icing” in Table B3A of Appendix B. The FAA has not changed these requirements in the final rule.

3. Existing Engine and Airframe Icing Requirements in Part 60

In the existing part 60, the subjective evaluation requirements in Appendix A includes a table of special effects (Table A3F) that contains additional requirements for the qualification of engine and airframe icing. In the NPRM, the FAA maintained this table with no changes to it.

Boeing, A4A, and NTSB commented that the requirements for icing evaluation in Table A3F (special effects) include the evaluation of increased gross weight due to ice accumulation. The commenters noted that the pilot has no means to recognize if the simulated aircraft's weight has increased and an increased gross weight due to ice accumulation is typically an insignificant effect of icing. Boeing further commented that this test requires a “nominal altitude and cruise airspeed and is likely to result in a flight condition where icing does not occur for large commercial transport category airplanes. This flight condition will also likely result in trimming at a low AOA where the effects of ice, even with the anti-ice system deactivated, are small (a few tenths change in pitch attitude or a few percent change in thrust to maintain level flight). In the lower AOA range, the aerodynamic effects of ice are relatively small. For large commercial transports one might expect to see a few tenths of a degree change in pitch attitude or a few percent change in thrust to maintain level flight with the addition of ice. This proposed new test will likely result in generating unnecessary questions when the expected (larger) results are not seen.”

FAA agrees with the commenters and has removed references to increased gross weight in the final rule as that table entry for icing special effects (Table A3F, Entry No. 2) was inadvertently retained in the proposal. Furthermore, the FAA has amended this table to remove the “nominal altitude and cruise airspeed” requirement and made additional changes to better align this section with the general requirements for engine and airframe icing in Table A1A, Entry No. 2.j.

4. Applicability in Training Programs

In the NPRM, the proposed updated requirements for engine and airframe icing were applied to all Level C and Level D FFSs, regardless of the type of aircraft or operator. This is consistent with the engine and airframe icing requirements in the existing part 60 and previous FSTD evaluation standards. The FAA notes that “engine and airframe icing” simulation is not a new FSTD qualification requirement that was introduced by this rulemaking. In fact, the “effects of airframe icing” has been a minimum FSTD qualification requirement for Level D (Phase III) FFSs since the publication of AC 121-14C, Aircraft Simulator and Visual System Evaluation and Approval, published in 1980. Similarly, the “effects of airframe and engine icing” is currently an FSTD qualification requirement in the existing part 60 rule (published in 2008) for Level C and Level D FFSs.

Delta commented that the de-icing and anti-icing systems are very effective on turbojet airplanes. The accidents referenced in NTSB reports are turboprops with significantly less performance available. Delta added there are no useful training objectives to be taught to pilots of commercial turbojet airplanes in icing conditions. A4A commented that stall ice effects are not required by Public Law 111-216 or the Crewmember and Aircraft Dispatcher Training final rule and should be deleted from this final rule. Delta, A4A, and FlightSafety further questioned whether the FAA has a specific list of airframes that are impacted by icing or are vulnerable to a specific type of ice accretion.

The FAA points out that Section 208(b)(1) of Public Law 111-216 addressed increasing the familiarity of flight crewmembers with, and improving the response of flight crewmembers to icing conditions. However, irrespective of statutory direction, the FAA believes the understanding of the effects of icing on aircraft performance is essential for professional crewmembers particularly as it relates to stall AOA.

The FAA agrees with Delta that de-icing and anti-icing systems are generally very effective on turbojet airplanes. However, every airplane is susceptible to icing to some extent and therefore, there are useful training objectives to be taught to pilots of turbojet aircraft. While the FAA recognizes that turboprop airplanes are generally more susceptible to ice accretion, accidents and incidents on turbojet aircraft have occurred in the past. In the case of the Circuit City Cessna 560 (a turbojet aircraft) accident in Pueblo, Colorado on February 16, 2005,

the flight crew did not comply with de-icing procedures during approach which led to an aerodynamic stall from which they did not recover. While it is unknown if the crew recognized the effects of icing before the aerodynamic stall occurred, enhanced simulator training on de-icing and/or anti-icing procedures with representative effects of ice accretion may have increased their awareness that ice accretion was occurring.

Crash During Approach to Landing; Circuit City Stores, Inc.; Cessna Citation 560, Pueblo, Colorado, February 16, 2005. Accident Report NTSB/AAR-07/02. National Transportation Safety Board.

With respect to engines, while turboprop and propeller aircraft engines are generally more susceptible to the effects of ice accretion than turbojet engines, power loss events due to core icing have been known to occur on multiple models of aircraft and engines (including large turbojet aircraft). In research conducted in 2009, it was found that engine power loss events due to ice accretion were occurring at a rate of about one event every 4 months.

While these events often occurred in conditions that pilots considered benign with no airframe ice accreted, there were recognition cues present and it was noted that each engine appeared to

have a different manifestation of the icing event. While this final rule does not require specific engine icing models such as these, providing flight crews with representative cues of engine icing, where present during a typical in-flight ice accretion event, could aid in its recognition during line operations.

Mason, J., “Current Perspectives on Jet Engine Power Loss in Ice Crystal Conditions: Engine Icing,” Presentation at 2008 AIAA Atmospheric and Space Environments, June 23rd, 2009.

The FAA has not prescribed specific types of ice accretion models to be implemented in the final rule. The intent is to provide flight crews with representative recognition cues of ice accretion for the aircraft being simulated. Where the accident and incident record indicates that a particular airframe may be susceptible to a particular type of ice accretion, the simulation of the cues associated with that type of icing should be considered when developing a representative icing model. While the accident record has some general examples of this (such as supercooled large droplet icing or tailplane icing on some aircraft), the aircraft manufacturer will likely be the best source of information as to a particular type of icing scenario that may enhance training in recognizing and exiting icing conditions for that aircraft.

5. Data Sources and Tuning of Ice Accretion Models

In the proposal, the FAA introduced updated engine and airframe icing requirements that included a requirement to use “aircraft OEM data or other acceptable analytical methods” to develop ice accretion models.

An anonymous commenter stated that the cost of purchasing icing data, if it exists, could be prohibitive. Due to the availability of SME's who have flown the subject aircraft in icing conditions, the requirement should allow SME pilot validation of icing models. Both A4A and CAE made similar comments that some SME pilot tuning and validation of icing models should be allowed in the requirements.

Dassault further commented that flight test data obtained through the aircraft certification process is limited with larger amounts of ice accretion. Engineering tests might be conducted in those conditions; however, Dassault claimed it would be unable to provide an SOC because there is no flight test data to support it.

The FAA maintains that icing models may be developed using analytical or other engineering methods, incorporating flight test data where available. This process may include supplemental SME pilot assessment to tune and subjectively validate the models. Furthermore, the objective demonstration test does not require the use of flight test data or other data to validate the model. The demonstration test is for the purpose of demonstrating that the expected icing recognition cues are present as compared to the simulation with no ice present. The FAA has added clarifying language in Table A1A and Attachment 7.

The FAA agrees with Dassault that flight test data gathered during the aircraft certification process will generally be limited to ice shape testing conducted to demonstrate performance limits. Like the current part 60 requirements for the simulation of airframe and engine icing, engineering and analytical methods may be used to develop representative icing models that support the intended training objectives. While the use of flight test data would certainly assist in developing such models, engineering analysis supported with subjective assessment and tuning of the icing models for the expected recognition cues will be acceptable in lieu of flight test developed models and should not be as costly.

D. Evaluation Requirements for Takeoff and Landing in Gusting Crosswinds

In order to support the new gusting crosswind training requirements in the Crewmember and Aircraft Dispatcher Training final rule, the FAA proposed new minimum requirements for Levels A, B, C, and D FFSs to include the programming of realistic gusting crosswind profiles. The FAA notes that in the existing part 60 and previous FSTD evaluation standards, there is no requirement for any FSTD to simulate gusting crosswinds. These proposed requirements also included updated ground handling characteristics to be evaluated with crosswinds and gusting crosswinds up to the aircraft's maximum demonstrated crosswind component. The FAA further included guidance material in the information section of the proposal that recommended the use of the Windshear Training Aid or other acceptable source data in the development of the gusting crosswind profiles.

1. Applicability on Lower Level FSTDs

In the proposal, FSTD evaluation requirements for gusting crosswind profiles were made applicable for all FFS levels in Appendix A as well as the Level 7 FTD defined in Appendix B.

TRU Simulation and A4A commented that a new gusting crosswind requirement was added for the Level 7 FTD and questioned whether this was appropriate for a Level 7 FTD. Boeing additionally commented that the requirement for gusting crosswinds are proposed for Levels A, B, C, and D FFSs, but crosswind takeoff and landing tasks are not minimum requirements for Level A simulators in Table A1B. Finally, A4A and Delta commented that gusting crosswind requirements have been added for both Level A and B simulators, but should be removed due to lack of alignment with the ICAO 9625 FSTD device type categories.

With regards to the Level 7 FTD, FAA has examined the ICAO 9625 requirements for the Type V device and found that instructor control of “surface wind speed, direction, and gusts” is a minimum requirement for this device level (see ICAO 9625, Appendix A, section 11.4.R,G). In order to maintain consistency and alignment with the similar ICAO device, FAA has maintained this requirement in the general requirements and functions and subjective testing tables for the Level 7 FTD, but removed the more detailed requirement for realistic gusting crosswind profiles and the associated SOC that was proposed in the NPRM.

FAA agrees with Boeing's comment concerning the qualification of the Level A simulator for takeoff and landing tasks and has removed this requirement in the final rule. Additionally, due to the lack of required side force motion cueing in a Level B simulator that would enhance the simulation of a realistic and dynamic gusting crosswind scenario, the FAA has also removed this minimum requirement for Level B simulators in the final rule.

2. Gusting Crosswind Profile Data Sources

In the NPRM, the FAA proposed requirements for FSTD sponsors to develop a realistic gusting crosswind profile for use in training. The FAA was not prescriptive in this requirement and only required that the profile be “realistic” and “tuned in intensity and variation to require pilot intervention to avoid runway departure during takeoff or landing roll.” The FAA additionally provided guidance in the information column of the proposal recommending the use of the Windshear Training Aid or other acceptable data sources to develop the gusting crosswind profiles.

The FAA received several comments concerning the data sources needed to develop realistic gusting crosswind profiles to meet the rule requirements. American, JetBlue, and A4A commented that FAA should provide an appropriate gusting crosswind model as recommended by the NTSB in its safety recommendation. Boeing commented that the Windshear Training Aid does not provide the necessary data to effectively model gusting crosswinds. Delta and A4A further commented that

the FAA should define “other acceptable source data” to help sponsors be consistent in programming the gusting crosswind scenarios. Additionally, A4A commented that the FAA should permit carriers to use crosswinds with gust data from multiple sources because doing so will provide flexibility, more compliance options, and reduce compliance burdens. Finally, an anonymous commenter stated that all references in the NPRM to “gusting crosswinds” lack definition of what is considered a “gust”. “Without a definition such as “10 percent increase over steady state wind speed for x seconds, repeated randomly”, this is an entirely subjective condition and as such is subject to every inspector's idea of what a wind gust should or could be. If the FAA cannot provide subjective guidance similar to the Windshear Training Aid, which does not provide adequate information for this scenario, the gusting crosswind scenarios should be treated as `demonstration only' and not for training credit.”

While the FAA would generally agree that a defined wind gust model could provide standardization for FSTD qualification purposes, such a generic model may not be realistic unless tuned for the particular aircraft and training scenario. Similar to the Windshear Training Aid's windshear profiles, subjective tuning would be required to adjust the model as a function of the aircraft type/configuration and ambient conditions to provide the cues and aircraft performance needed to accomplish the training objectives. In the proposal, the FAA required that such wind gust models be “realistic” and have been “tuned in intensity and variation to require pilot intervention to avoid runway departure.” Like many other areas in the simulator qualification standards, this allows for the FSTD sponsor to develop solutions that meet the needs of their particular training program without the FAA prescribing a specific solution. While realistic baseline wind gust models may be derived from aircraft operational data, meteorological data, or other data, a certain amount of subjective tuning will be required in many cases to ensure the gusts are adequate enough to require pilot intervention to avoid runway departure or otherwise do not exceed the crosswind capabilities of the simulated aircraft and supporting aerodynamic and ground model data. Due to the wide range of aircraft and associated crosswind capabilities, the FAA has found that specifying a certain gust characteristic for FSTD qualification would not be practical and has maintained the requirements as proposed.

In response to the NTSB safety recommendation

and commenters' requests for an FAA developed gusting crosswind model, the FAA conducted an analysis of the extracted wind data from the Continental (CO) 1404 accident

and developed two wind gust models that may be used by FSTD sponsors to meet the requirements for a realistic gusting crosswind model. The first model was developed using the CO 1404 accident data to closely replicate the wind gust that was experienced by the flight crew in that accident. While this model was tested by FAA on a Boeing B737-800 simulator and was found to provide a subjectively acceptable training scenario, it is expected that the model will need to be tuned by the sponsor for different aircraft and operator specific training scenarios.

NTSB safety recommendation no. A-10-110.

Runway Side Excursion During Attempted Takeoff in Strong and Gusty Crosswind Conditions, Continental Flight 1404, December 20, 2008, NTSB Final Report, NTSB/AAR-10/04.

A second model was developed using a simplified linear estimation of the CO 1404 accident data using maximum wind rates of change as referenced in the Windshear Training Aid and the Joint Airport Weather Studies (JAWS)

. Similar to the continuous wind gust model, this model may also require tuning by the sponsor for different aircraft and operator specific training scenarios.

The maximum wind rates published in the Windshear Training Aid are based upon the Joint Airport Weather Studies (JAWS) and were calculated from accident flight data recorder and Doppler radar measurements of microburst events.

FAA recognizes that sponsors may desire to implement their own wind models that may be more suitable for their particular training programs and has not mandated the above described wind gust models as a condition of FSTD qualification. These models will be provided with the final rule as guidance material in a National Simulator Program (NSP) Guidance Bulletin and may be used as one method to develop realistic gusting crosswind profiles to satisfy the requirements of the rule. As suggested by A4A, this will provide operators with flexibility to develop other wind gust models from multiple sources to meet the FSTD qualification requirements.

3. Maximum Demonstrated Crosswind

In the proposal, the FAA included general requirements for Level C and Level D FFSs that included ground handling characteristics for crosswinds and gusting crosswinds up to the aircraft's maximum demonstrated crosswind component.

Delta and A4A requested clarification if the maximum demonstrated crosswind value includes the gusting component, or is the intent to require the gusting component in addition to the maximum demonstrated crosswind value.

The FAA has not prescribed a specific wind magnitude and direction to be implemented in the gusting crosswind model requirements. The wind gust models that will be provided by the FAA in guidance material were designed to allow for tuning of the gust characteristics as needed for the particular training scenarios (such as steady state wind conditions and runway direction) and aircraft type being simulated. The tuning of gust models should be conducted in consideration of the maximum crosswind capabilities of the aircraft in order to provide operationally realistic scenarios that are survivable in training. The specific aircraft crosswind capabilities, to include the addition of gust factors, are determined by the aircraft OEM. If this information is not clear in the aircraft flight manual, the FSTD sponsor should consult with the aircraft OEM. Additionally, the FSTD sponsor should coordinate with the data provider to ensure that gust models do not exceed the capabilities of the simulator's aerodynamic and ground models. The FAA has added information material in Table A1A (entry no. 2.d.3) to the final rule for clarification.

4. Requirements for Previously Qualified FSTDs

In the proposal, the updated ground handling and ground reaction requirements in Table A1A included information that stated “tests required” for these particular sections. The FAA notes that this text was derived from the similar sections in the ICAO 9625 document as part of the alignment process.

Delta and A4A pointed out that the general requirement for gusting crosswind (Table A1A, Entry No. 3.1.S in the NPRM) states “tests required” and requested clarification if additional objective testing is required under the FSTD Directive for previously qualified FSTDs.

In the final rule, since the FAA restored the existing part 60 format for the general requirements table as compared to the ICAO format in the proposal (including sections for ground reaction and ground handling

characteristics), the text for “tests required” was removed from the ground handling requirements in Table A1A, Entry No. 2.d.3. in the final rule. No additional objective testing for ground reaction and ground handling characteristics was intended for previously qualified FSTDs in FSTD Directive No. 2. The FAA further notes that all required objective testing is fully described in Table A2A, making any such “tests required” notations in the information column redundant.

E. Evaluation Requirements for Bounced Landing Recovery Training Tasks

In the proposal, the FAA included updated FSTD evaluation requirements for ground reaction characteristics to support the bounced landing recovery training task that is required in the Crewmember and Aircraft Dispatcher Training final rule. The new requirements included ground reaction modeling to simulate the effects of a bounced or skipped landing as well as the indications of a tail strike or nosewheel exceedances as appropriate for the simulated aircraft and conditions.

1. Applicability to Lower Level FSTDs

In the proposal, the new requirements for bounced landing recovery evaluation were included for Level C and Level D FSTDs in Appendix A as well as for the new Level 7 FTD in Appendix B.

TRU Simulation and A4A commented that the bounced landing requirements were added for the Level 7 FTD and questioned whether it was appropriate for this device.

Given the Crewmember and Aircraft Dispatcher Training final rule requirement that a Level C or higher FSTD be used to conduct bounced landing recovery training tasks, the FAA has removed the additional FSTD evaluation requirements in the final rule for bounced landing recovery from the Level 7 FTD minimum requirements in Appendix B.

2. Bounced Landing Modeling and Evaluation

a. Nosewheel Exceedences

As part of the bounced landing recovery requirements in the proposal, the FAA included requirements to include indications of a tail strike and nosewheel exceedances.

Boeing commented that the requirement for “nosewheel exceedances” needs to be more clearly defined (

e.g.,

limit, yield, or ultimate loads) and suggested changing the rule text to read “effects and indications of ground contact. . .”. An anonymous commenter further stated that calculation of structural loads on the nose gear is not a common feature in current FSTDs. Any nose first landing is considered abnormal and could be flagged on the IOS.

The FAA agrees with the commenters and has removed the nosewheel exceedances requirement from the final rule as it is not necessary to accomplish the training objectives for bounced landing recovery training tasks. This language was replaced with “the effects and indications of ground contact due to landing in an abnormal aircraft attitude . . .” since information on aircraft attitude during the landing and go-around sequence will be more useful to the instructor in evaluating bounced landing recovery training tasks.

b. Use of Existing Ground Reaction Modeling

In the NPRM, the FAA proposed that ground reaction modeling must simulate “. . . the effects of a bounced or skipped landing (to include indications of a tail strike or nosewheel exceedances) as appropriate for the simulated aircraft and conditions”.

Delta and A4A commented that the existing part 60 requires verification of ground reaction and ground effects by minimum unstick speed, ground effects, and takeoff and landing performance objective tests. An SOC from the data provider and an affirmation that the model has been implemented correctly should be adequate. There is no need for additional subjective verification by a qualified pilot. A4A further commented that at least one data provider has implied that their current data and model meets the proposed requirements. CAE commented that the strut system simulation (damper/spring) and its geometry are already properly modeled and should provide the appropriate forces and moments during a bounce.

As described in the proposal, the FAA agrees with the commenters that much of the aerodynamic and ground reaction modeling is currently required and validated in several required objective tests for FSTD qualification. As such, the FAA has not required any additional objective testing for the qualification of bounced landing recovery training tasks in this final rule. In order to support bounced landing recovery training, the FSTD must have the ability to provide the instructor with the effects and indications of ground contact as a result of the FSTD being landed or conducting a go-around at an improper aircraft attitude. In addition to pitch attitude information, other parameters such as indications of nosewheel contact and indications of a tailstrike would provide useful information to the instructor in evaluating a bounced landing recovery maneuver. FAA agrees with the commenters that the use of a qualified SME pilot to evaluate these indications may be of limited value because they may not have any direct experience in the indications of a tailstrike in the airplane to base such an evaluation on. The FAA does recognize, however, that a tailstrike and other indications of ground contact can be computed in software using the geometric dimensions of the airplane and these indications will provide the instructor with additional feedback to assist in determining whether the aircraft landed in or a go-around was attempted in an unusual aircraft attitude. These indications and the ability of the modified FSTD to perform the intended training tasks are what should be evaluated by the sponsor's designated pilot as described in the FSTD Directive and § 60.16(a)(1).

The FAA has reviewed the current part 60 ground reaction and ground handling requirements along with associated objective testing that are already required for Level B through Level D FFSs and has determined that adequate requirements already exist in part 60 to evaluate and validate the aircraft dynamics necessary to support bounced landing recovery training tasks.

In order to improve the instructor's evaluation of an abnormal aircraft attitude during the bounced landing recovery maneuver, the FAA has amended the current ground reaction requirement for Level B through Level D FFSs to include appropriate effects during bounced or skipped landings, including the effects and indications of ground contact due to landing in an abnormal aircraft attitude.

In addition to objective testing requirements for maneuvers such as takeoff, landing, minimum unstick speed, and ground effect, the current part 60 ground reaction general requirements (Table A1A, Entry No. 2.d.2.) already requires ground reaction modeling that generally supports bounced landing recovery training.

3. Alignment With Training Requirements

As noted in the NPRM, the FSTD evaluation requirements for bounced landing recovery maneuvers were introduced both to support new requirements in the Crewmember and Aircraft Dispatcher Training final rule as well as to address comments concerning potential deficiencies in FSTD fidelity in this flight regime.

An anonymous commenter stated that “there is no bounced landing training task listed in Table A1B (Table of Tasks v. Simulator Level). It is agreed that a

Level D simulation should produce a bounced landing if appropriate, however that does not translate into a training requirement. There is currently no approved pilot training program that includes bounced landing. At most, it could be a required demonstration element, but it should not be a required training maneuver.”

A4A commented that Boeing has already addressed the bounced landing recognition and recovery procedure in their operating manuals and in recurrent simulator training and that the FAA should review simulator data it currently receives to determine if recurrent training programs implemented due to the NTSB recommendations were effective. A4A and JetBlue further commented that “the training final rule limits new training requirements to recovery from bounced landing because carrier training programs currently include bounced landing training as recommended in FAA's InFO 08029 . . . simulator modeling for this final rule should be limited to enhancement to train recovery methods; it should avoid introducing elements that might induce negative training associated with `teaching to bounce'.” In addition, CAE made similar comments concerning the potential of a transfer of negative training in introducing a bounced condition during landing.

The FAA notes that bounced landing recovery is a training requirement for air carriers under § 121.423. While the minimum qualified task list in Table A1B does not specifically list bounced landing tasks, the final rule will require an amendment to the FSTD's SOQ that the FSTD has been evaluated for bounced landing recovery training tasks. As addressed in the Crewmember and Aircraft Dispatcher Training final rule, the FAA is aware of the incorporation of bounced landing recovery training by operators in response to the FAA's InFO and SAFO bulletins. To support the new training requirements in § 121.423 for bounced landing recovery training, the FSTD qualification standards were revised in this rule to ensure the FSTDs used to conduct such training have been properly evaluated for the training tasks.

The FAA agrees with commenters in that the purpose of bounced landing recovery training is to train bounced landing recovery methods and not to teach a pilot how to bounce the aircraft. While the simulation should support the ability to reproduce a bounce where the flight conditions dictate, the primary objective of training is to train recovery techniques should the landing result in an inadvertent bounce. The FAA agrees with the commenters in that these recovery techniques can be taught without stimulating an actual bounce during the landing sequence and rather “calling a bounce” to initiate the recovery maneuver. The FAA has amended the final rule to emphasize that the FSTD evaluation requirements are on the aircraft dynamics resulting from the bounced landing recovery and not in stimulating a bounce during the landing sequence.

The FAA further emphasizes that the FSTD evaluation requirements in the final rule that support bounced landing recovery training tasks are essentially a consolidation of existing requirements within part 60

and will further support the instructor evaluation of other landing training tasks where the simulator may be inadvertently landed in an abnormal aircraft attitude.

See 14 CFR part 60 (2008), Appendix A: Table A1A, Entry No. 2.d.2 (ground reaction modeling); and Table A3D (motion system effects), Entry no. 7 (main and nose gear touchdown cues), and Entry No. 13 (tail strikes and engine pod strikes).

4. Requirements for Previously Qualified FSTDs

Delta, FlightSafety, and A4A pointed out that the general requirement for ground reaction modeling (Table A1A, Entry No. 3.1.S in the NPRM) states “tests required” and requested clarification if additional objective testing is required under the FSTD Directive for previously qualified FSTDs.

This text is long and has been trimmed here. Open the source document for the complete record.

This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.