(7) The applicant should use analyses, or testing, or both, to show that changing to, and operating in, alternate modes has no unacceptable effect on engine durability or endurance. This should not be interpreted as the need to run multiple §33.87 endurance tests. Short segments of testing in combination with analysis may be sufficient for showing of compliance.
(8) The applicant should demonstrate the durability and reliability of the control system in all modes, primarily through the engine system and component testing of §§ 33.53 and
33.91. Performing some portion of the engine certification testing in the alternate mode(s) and during transition between modes can be used as part of the system validation required under § 33.28(b)(1).
(9) Engine test considerations.
(a) The engine certification tests defined in part 33, Subparts D and F, are often performed using only the ECS’s primary mode. If the applicant requests approval for dispatch in an alternate mode under TLD, then the applicant must demonstrate that the engine can meet the defined test-success criteria in any alternate mode that is proposed as a dispatchable configuration. If a configuration is to be dispatchable, then that mode should be demonstrated to operate in all foreseeable environmental conditions. This demonstration may use analysis or test and should include capabilities such as operability in rain or hail or bird ingestion.
(b) Some capabilities, such as operability in rain or hail or bird ingestion, may be lost in some control modes that are not dispatchable. Modes that are not dispatchable do not require engine test demonstration as long as the engine installation and operating instructions reflect this loss of capability.
(10) Availability. Applicants should demonstrate through testing or monitoring, that any backup mode will be available when needed. The applicant should document in the ICA the minimum frequency of inspection or testing needed to ensure its availability.
b. Crew training modes. As part of the System Safety Assessment of 33.28(e), applicants should assess crew training modes. These modes are usually installation, and possibly operator specific and need to be negotiated on a case-by-case basis. For example, one common application of crew training modes is simulation of the ‘fail-fixed’ mode on a twin-engine rotorcraft. The applicant must describe training modes in the engine installation and operating instructions as appropriate. Also, the applicant must take precautions in the design of the ECS and its crew interfaces to prevent inadvertent entry into any training modes. The applicant must assess crew-training modes, including lockout systems.
c. Non-dispatchable configurations and modes.
(1) For control configurations that are not dispatchable, but for which the applicant seeks to take credit in the system LOTC/LOPC analysis (this analysis and the LOTC/LOPC concept are covered in full in Chapter 6 below), specific operating limitations may be acceptable.
This means that the system will not be charged with an LOTC/LOPC event when the fault is covered by a backup means that allows the system to continue to function safely. The system, 15
however, would not be dispatchable in this configuration. Compliance with § 33.28(b)(1) in non dispatchable configurations does not imply strict compliance with the operability requirements of § 33.65 and § 33.89, if applicants demonstrate that in the intended installation no likely pilot control system inputs will result in engine surge, stall, flameout, or unmanageable delay in power recovery or engine controllability issues. For example, in a twin-engine rotorcraft, a rudimentary backup system may be adequate since frequent and rapid changes in power setting when operating in the backup system may not be necessary. In this example, it is very unlikely that dispatch would be allowed with the engine controlled by this rudimentary backup system.
(2) In addition to these operability considerations, the applicant should consider the following factors in assessing the acceptability of a reduced-capability backup mode:
(a) The installed operating characteristics of the backup mode and the differences from the primary mode.
(b) The likely impact of the backup mode operations on pilot workload, if the aircraft installation is known.
(c) The frequency of transfer from the primary mode to the backup mode (i.e. the reliability of the primary mode). Frequencies of transfer of less than 1 per 20,000 engine flight hours are acceptable.
d. Control transitions.
(1) Section 33.28(c) requires that any control transitions that occur as a result of fault accommodation occur in an acceptable manner.
(2) In general, the ECS should transition to alternate modes automatically. However, systems in which pilot action is required to engage the backup mode are also acceptable. For instance, a fault in the primary system may result in a “fail-fixed” fuel flow that requires some action by the pilot to engage the backup system to modulate engine power. When pilot action is required, applicants should ensure that any reliance on a pilot executing a manual transition does not pose an unacceptable operating characteristic, unacceptable crew workload, or require exceptional skill.
(3) The applicant must review the transient change in power or thrust associated with transfer to alternate modes for compliance with § 33.28(c). If available, consider input from the installer.
(4) Applicants should consider at the minimum, the following items when reviewing control mode transitions:
(a) The frequency of occurrence of transfers to any alternate mode and the capability of the alternate mode. Applicants should support computed frequency-of-transfer rates with data from endurance or reliability testing, in-service experience on similar equipment, or other appropriate data.
(b) The magnitude of the power, thrust, rotor or propeller speed transients may affect the dynamics of the aircraft. Therefore these transients should be carefully assessed to assure acceptability for installation. The applicant should coordinate these characteristics with the intended installer.
(c) Successful demonstration, by simulation or other means, that the ECS controls the engine during the transition. In some cases, particularly those involving rotorcraft, applicants may be unable, analytically or through simulation, to show that the ECS can control the engine during mode transition. Therefore, applicants should propose a flight test program to demonstrate that the ECS controls the engine acceptably during in-flight control mode transitions.
(d) An analysis to identify those faults that cause either automatic or pilot initiated control mode transitions.
(e) For turboprop or turboshaft engines, the control mode transition should not result in excessive over-speed or under-speed of the rotor or propeller. These speed transitions could cause emergency shutdown, loss of electrical generator power, or setting off warning devices.
(f) Including in the engine installation instructions, a declaration of the power or thrust change associated with a transition. If the change is negligible, declare it as such.
(5) Time delays. Applicants should include in the engine installation and operating instructions a description of any observable time delays associated with control mode, channel or system transitions, or in re-establishing the pilot’s ability to modulate engine thrust or power.
The acceptability of these delays may need to be assessed during aircraft certification.
(6) Annunciation to the flight crew.
(a) If annunciation is necessary to comply with § 33.28(c)(1)(iii), the type of annunciation to the flight crew must be commensurate with the nature of the transition. For example, the form of annunciation for a reversion to an alternate mode of control when the transition is automatic and the only observable changes in engine operation are different thrust control schedules would be very different than if timely action by the pilot is required to maintain control of the aircraft.
(b) The intent and purpose of the cockpit annunciation must be clearly stated in the engine installation and operating instructions, as appropriate.
CHAPTER 6. SECTION 33.28(d) – ENGINE CONTROL SYSTEM FAILURES.
6-1. Rule Text. Section 33.28(d) reads:
“d. Engine control system failures. The applicant must design and construct the engine control system so that:
(1) The rate for Loss of Thrust (or Power) Control (LOTC/LOPC) events, consistent with the safety objective associated with the intended application can be achieved;
(2) In the full-up configuration, the system is single fault tolerant, as determined by the Administrator, for electrical or electronic failures with respect to LOTC/LOPC events;
(3) Single failures of engine control system components shall not result in a hazardous engine effect; and
(4) Foreseeable failures or malfunctions leading to local events in the intended aircraft installation such as fire, overheat, or failures leading to damage to engine control system components do not result in a hazardous engine effect due to engine control system failures or malfunctions.”
6-2 Guidance: Engine Control System Failures .
a. Objective. Section 33.28(d) establishes ECS integrity requirements that are consistent with the operational requirements of the various installations. The introduction of EECSs should provide at least an equivalent level of safety and reliability for the engine as achieved by engines equipped with hydro-mechanical control and protection systems and magneto systems (refer to SAE ARP5107B and FAA Policy Memorandums on TLD in paragraph 3.b.). In chapters 9 and 15 of this AC, the concept of development errors being introduced within the software, or AEH designs ,or both, and the development assurance processes are addressed. This chapter and chapter 7 discuss the development error concept at the system level for an ECS.
b. Criteria for an LOTC/LOPC event. Industry practice over the past 30 years has resulted in the following generally accepted criteria for defining an LOTC/LOPC event. The applicant may propose another set of criteria if it defines an LOTC/LOPC event.
(1) For turbine engines not intended for rotorcraft, we consider an LOTC/LOPC event occurs when the ECS:
(a) Has lost the capability to modulate thrust or power between idle and 90% of maximum rated power or thrust,
(b) Suffers a fault that results in a thrust or power oscillation greater than the levels given in Paragraph 6-2c of this AC, or 18
(c) Has lost the capability to govern the engine in a manner that allows compliance with the operability regulations given in §§ 33.65 and 33.89.
(2) For turbine engines intended for helicopters, an LOTC/LOPC event occurs when the ECS:
(a) Has lost the capability to modulate power between idle and 90% of maximum rated power at flight condition and, if appropriate, at rotorcraft operating condition, or
(b) Suffers a fault that results in a power oscillation greater than the levels given in Paragraph 6-2 c of this AC, or
(c) Has lost the capability to govern the engine in a manner that allows compliance with the operability regulations given in §§ 33.65 and 33.89. The inability, however, to meet the operability regulations in the alternate modes may not need to be included as LOPC events.
1 Single-engine rotorcraft will normally be required to meet the operability regulations in the alternate modes. Engine operability in alternate modes is a necessity if the control transitions to the alternate mode occur more frequently than the acceptable LOPC rate, or normal flight crew activity requires rapid changes in power to safely fly the aircraft.
2 The LOPC definition typically includes the ability to meet the operability regulations in the alternate mode(s). For multiengine rotorcraft, it may be acceptable if one engine control transitions to an alternate mode that may not have robust operability. That engine can be left at reasonably fixed power condition so that the engines with normally operating controls can change power as necessary to complete aircraft maneuvers and safely land the aircraft. Where this potential exists, i.e., differing operability levels, applicants may be required to demonstrate the ECS continues to meet regulatory standards under all conditions at aircraft certification.
(3) For reciprocating engines intended for part 23 Class I., II., and III. Aircraft, an LOPC event occurs when the ECS:
(a) Has lost the capability to modulate power between idle and 85% of maximum rated power at all operating conditions, or
(b) Suffers a fault that results in a power oscillation greater than the levels given in Paragraph 6-2 c. of this AC, or
(c) Has lost the capability to govern the engine in a manner that allows compliance with the operability regulations in § 33.51.
(4) For engines incorporating functions for propeller control integrated in the EECS, the applicant should consider the following faults or failures, either individually or in combination with another, as LOPC events:
(a) Inability to command a change in pitch,
(b) Uncommanded change in pitch,
(c) Uncontrollable propeller torque or speed fluctuation.
c. Uncommanded thrust or power oscillations. Any uncommanded thrust or power oscillations should not be of such a magnitude as to affect aircraft controllability. For example, thrust or power oscillations less than 10% peak-to-peak of take-off power, or thrust, or both, are acceptable where oscillation affects one engine only, unless the installer defines more or less restrictive requirements. Regardless of the levels discussed here, if the flight crew has to shut down an engine because of unacceptable thrust or power oscillations caused by the control system, we would deem such an event as an in-service LOTC/LOPC event.
d. Acceptable LOTC/LOPC rate. The applicant may propose an LOTC/LOPC rate other than those below. The applicant should substantiate such a proposal in relation to the criticality of the engine and control system relative to the intended installation. The intent is to show equivalence of the LOTC/LOPC rate to existing systems in comparable installations. Refer to SAE ARP5107B and FAA Policy Memorandum, PS-ANE100-2001-1993-33.28TLD-R1 for additional information.
(1) For turbine engines, the EECS should not cause more than one LOTC/LOPC event per 100,000 engine flight hours.
(2) For reciprocating engines intended for part 23 Class I., II., and III. aircraft, an LOPC rate of 45 per million engine flight hours (or 1 per 22,222 engine flight hours) represents the upper limit of an acceptable level for the most complex EECS. As a result of the architectures used in many of the EECS for these engines, the functions are implemented as independent system elements. These system elements or subsystems can be fuel control, ignition control, or others. For systems with only one element, such as a fuel control, we recommend applicants use a maximum system LOPC rate of 15 LOPC events per million engine flight hours (PMEFH). For systems that contain more than one element, we recommend applicants use a LOPC rate of 15 events PMEFH for each, but only up to 45 events PMEFH. So, if two elements are present, for example a fuel control and an ignition control, the appropriate rate is 30 events PMEFH. If three or more, then the LOPC rate would be 45 events PMEFH.
e. LOTC/LOPC analysis.
(1) Since the LOTC/LOPC analysis is part of the required System Safety Assessment of §33.28(e), applicants must perform and submit an LOTC/LOPC analysis for the ECS. An ECS LOTC/LOPC analysis may take the form of a system reliability analysis (refer to SAE ARP5107B). We recommend a numerical analysis such as a Markov model analysis, fault tree analysis, or equivalent analytical approach.
(2) The LOTC/LOPC analysis should address all components in the system that can contribute to LOTC/LOPC events. This includes all electrical (including wiring), mechanical, hydromechanical, and pneumatic elements of the ECS.
(3) The engine fuel pump is usually considered part of the fuel delivery system and, therefore, not included in an LOTC/LOPC analysis.
(4) An LOTC/LOPC analysis should include those sensors or elements that may not be part of the engine type design, but that may contribute to LOTC/LOPC events. An example is the throttle or power lever transducer, which is usually supplied by the installer. The LOTC/LOPC analysis should include the effects of loss, corruption, or failure of aircraft-supplied data. The engine installation instructions should include the assumed reliability and interface requirements for these nonengine type design elements. Within the aircraft system safety analyses, we recommend that installers ensure no double counting of the rate of failure of nonengine parts occurs.
(5) The LOTC/LOPC analysis should also consider all faults, both detected and undetected. The applicant should include any periodic maintenance actions needed to find and repair both covered and uncovered faults to meet the LOTC/LOPC rate in the engine instructions for continued airworthiness.
(6) See chapter 7 of this AC for additional guidance on how to perform the System Safety Assessment required under § 33.28(e).
f. Reliability assessment plan (RAP). We recommend the applicant prepare, and be ready to show us, a RAP (refer to SAE ARP5890A for a framework). A RAP documents the applicant’s controlled, repeatable processes for assessing the reliability of systems and equipment. It also helps assess the reliability of systems and equipment during design and operational life. The results of the RAP are important inputs to many safety assessment and analysis tasks. The results of a RAP can be used as a part of:
(1) Reliability program planning and monitoring,
(2) Safety assessments and analyses,
(3) Certification analyses,
(4) Equipment design decisions,
(5) System architecture selection, and
(6) Continued airworthiness assessments.
g. Commercial or industrial grade electronic parts.
(1) The grade and handling of electronic parts is an important contributor to the reliability of the EEC. Two examples of industry documents that provide guidance on the application of commercial or industrial grade components are:
(a) IEC/TS 62239-1, Process Management for Avionics – Preparation of an Electronic Components Management Plan, and
(b) IEC/TR 62240-1, Process Management for Avionics – Use of Semiconductor Devices Outside Manufacturers’ Specified Temperature Ranges.
(2) The applicant should prepare, and be ready to show us, their Electronic Component Management Plan (ECMP).
(3) When applicants specify as part of the engine type design commercial or industrialgrade electronic components, which are not manufactured to military standards, we recommend considering data similar to the following, as applicable:
(a) Reliability data for each commercial and industrial grade electrical component specified in the design.
(b) The applicant’s procurement, quality assurance, and process control plans for the vendor-supplied commercial and industrial grade parts. These plans should ensure that the parts will be able to maintain the reliability level specified in the approved engine type design.
(c) Unique databases, for similar components, obtained from different vendors because commercial and industrial grade parts may not all be manufactured to the same accepted industry standard.
(4) Temperature ranges for commercial or industrial grade parts are typically a narrower range than those for military grade parts.
(a) If the applicant’s declared temperature environment for the ECS will result in the detail parts exceeding the stated capability of the commercial or industrial grade electronic components, the applicant should ensure through test and analysis, that the proposed range of the specified components is suitable for the intended ECS environment. The applicant should also show that the failure rates used for those components in the System Safety Asssessment (SSA) and LOTC/LOPC analyses are appropriately adjusted for the extended temperature environment.
(b) Sometimes commercial or industrial parts are used in an environment beyond their specified rating and cooling provisions are required in the design of the EECS. The applicant should specify these provisions in the engine installation instructions, if action by the installer is required to provide cooling. This provision ensures that the cooling is not compromised. The cooling provisions included in the EECS design may have failure modes. If 22
the failure modes could result in exceeding temperature limits, then applicants should account for the probability of these failures in their SSA and LOTC/LOPC analyses.
(5) When any electrical or electronic components are changed, the applicant should review their SSA and LOTC/LOPC analyses with regard to the impact of any changes in component reliability. Component, subassembly or assembly level testing may be needed to evaluate a change that introduces commercial or industrial part(s). However, such a change would not be classified as “significant” with respect to § 21.101(b)(1).
h. Single fault accommodation.
(1) The following guidance clarifies the meaning of “single fault tolerant.”
(a) The applicant may show compliance with the single fault regulations of § 33.28(d)(2) and (3) by test and analysis. According to § 33.28(d), single failures or malfunctions in the ECS’s components, in its fully operational condition and all declared dispatchable configurations, must not result in a hazardous engine effect (refer to § 33.75(g)(2) for a definition of “hazardous engine effects”). In addition, § 33.28(d) requires that in its full-up configuration, the control system must be essentially single fault tolerant of electrical/electronic component failures with respect to LOTC/LOPC events.
(b) We recognize that achieving true single fault tolerance could require a triplicated design approach or one with 100% fault detection. Currently, systems have been designed with dual, redundant channels, or with backup systems that provide what has been called an “essentially single fault tolerant” system. Although these systems may have some single faults (that are not covered faults) that lead to LOTC/LOPC events, they have demonstrated excellent in-service safety and reliability and have proven to be acceptable.
Therefore, configurations such as these may be found to be compliant.
(2) Dual-channel or backup system configurations cover the vast majority of potential electrical and electronic faults. However, omitting some coverage because detection or accommodation of some electrical or electronic faults may not be practical, may be acceptable.
Single, simple electrical or electronic components or circuits can be employed in a reliable manner. In those cases, requiring redundancy may be unnecessary. In those cases, failures in certain single-electrical or electronic components, elements, or circuits may result in an LOTC/LOPC event. These systems, which are referred to as “essentially single fault tolerant,” are acceptable.
(3) Single failures that result in a high thrust failure condition, with no throttle response, may be catastrophic for some aircraft operating conditions. As a cautionary note, engine certification applicants should be aware that in this case, either a modification of the engine control or an independent aircraft system will be needed for aircraft certification.
i. Local events.
(1) Under § 33.28(d)(4), foreseeable failures or malfunctions leading to local events, such as engine or installation-related failures that could lead to damage to control system electrical harnesses or connectors or to the control units, must not result in a hazardous engine event. We recommend that applicants analyze local events to ensure a hazardous engine event will not occur. These events include:
(a) Overheat conditions, for example, those resulting from hot air duct bursts,
(b) Fires,
(c) Maintenance and foreseeable maintenance errors such as using a wire bundle as a hand hold,
(d) Fluid leaks, and
(e) Mechanical disruptions that could lead to damage to control system electrical harnesses, connectors, or control unit(s).
(2) These local events are normally limited to one engine. A local event is not usually considered a common mode event, and common mode threats, such as HIRF, lightning and rain, are not considered local events. Examples of a single, common mode fault in systems are single source batteries in multiengine applications and the use of identical software in multiengine, dual-channel systems. In these and similar cases, the applicant should take extra design, testing or maintenance precautions to ensure safety.
(3) Although they are limited to one engine, local events affecting the ECS should not generate hazards to the aircraft. Local events, in particular fire and overheat, are dependent on installations and may also be addressed as a part of aircraft certification. Coordination with the installer is highly advised.
(4) Invalid assumptions of independence between failures, as well as failure to recognize common cause failure modes, are leading reasons for invalid conclusions within safety analyses. In the assessment of local events, emphasis should be placed on identifying and ensuring critical functional or physical isolation is maintained. The applicant should ensure no common cause event is present that would violate any assumptions of independence between failures. For example, fuel control components whose failure could result in overspeed are not affected by the same events that could cause loss of overspeed protection.
(5) Whatever the local event, the behavior of the EECS must not cause a hazardous engine effect in any declared dispatchable mode.
(6) When demonstrating that no hazardous engine effect exists based on the assumption that another function exists to provide the necessary protection, the applicant should show that the other function is not rendered inoperative by the same event (including destruction of wires, ducts, or power supplies).
(7) An overheat condition exists when the temperature of the system components is greater than the maximum safe design operating temperature declared by the engine applicant in the engine installation instructions. The ECS must not cause a hazardous engine effect when the components or units of the system are exposed to an overheat or over-temperature condition, or when it cools down. Applicants may use specific design features or analysis methods to show prevention of hazardous engine effects. We may require testing when this is not possible. For example, due to the variability or the complexity of the failure sequence. Refer to SAE ARP5757 for an example of this type of testing.
(8) The ECS, including the electrical, electronic, and mechanical parts of the system, must comply with the fire regulations of § 33.17. This rule applies to the elements of the ECS that are installed in designated fire zones. Refer to SAE ARP5757 and AC 33.17-1 for additional guidance.
(9) If an ECS component is located so that it could present an ignition source for flammable fluids or vapors, the applicant should conduct an explosion proof demonstration to verify that the component cannot be the source of ignition for an explosion. Refer to SAE ARP5757 for this type of demonstration.
(10) Applicants should consider all foreseeable local events when complying with § 33.28(d)(4). We recognize, however, that it is difficult to address all possible local events in the intended aircraft installation at the time of engine certification. Therefore, the applicant should use sound engineering judgment to identify reasonably foreseeable local events.
Compliance with this regulation may be shown by considering the end result of the local event on the ECS. Well documented local events and their analysis will aid in engine installation certification.
(11) The following guidance applies to ECS wiring.
(a) Test or analyze each wire or combination of wires interfacing with the EECS that could be affected by a local event. The assessment should include opens, shorts to ground, and shorts to power (when appropriate). The results should show that faults result in specific responses and do not result in hazardous engine effects. Any EEC system component connector that becomes disconnected while the engine is operating must not result in a hazardous engine effect. Nor should it endanger the continued safe flight and landing of the aircraft.
(b) The applicant should test or analyze engine control unit aircraft interface wiring for shorts to aircraft power. These “hot” shorts should result in a specific and non-hazardous engine effect. Where aircraft interface wiring is involved, the engine installation instructions should inform the installer of the potential effects of shorts in the interface wiring. The installer should ensure that no wiring faults exist that are not detectable and not accommodated. Also, these wiring faults must not result in a hazardous engine effect.
(c) Where practical, wiring faults should not affect more than one channel. The engine applicant should include any assumptions regarding channel separation in the LOTC/LOPC analysis.
(d) Where physical separation of conductors is not practical, the engine applicant and the installer should coordinate to ensure that the potential for common mode faults is minimized between channels on one engine and eliminated between ECSs.
(e) The applicant should test and analyze the effects of fluid leaks impinging on EECS components. Impingement must not result in a hazardous engine effect, and the fluids should not impinge on circuitry or printed circuit boards, nor result in a potential latent failure condition.
(f) If the installation of the engine could be subject to part 25 rules, refer to the applicable sections of 14 CFR 25, Subpart H, Electrical Wiring Interconnection Systems (EWIS).