Can Fault Tree Analysis Technique Resolve Fault Isolation Ambiguity? - - PowerPoint PPT Presentation

Reliability Maintenance and Managing Risk Conference 2019

Alpana Gangopadhyaya Supportability Systems Engineer Northrop Grumman Alpana.Gangopadhyaya@ngc.com Phone Number 630 853-9974

Can Fault Tree Analysis Technique Resolve Fault Isolation Ambiguity?

RMMR 2019, Oct. 15 - 16, San Antonio, TX 1

Reliability Maintenance and Managing Risk Conference 2019

Outline of Abstract

We propose an advanced methodology for resolving fault isolation (FI) ambiguity by applying Fault Tree Analysis (FTA) technique in conjunction with Built-In Test (BIT). FI ambiguity occurs whenever any fault is associated with multiple replaceable units, which can cause false replace and remove maintenance

• activities. As a result, we experience substantial and unnecessary support cost

along with reduced system availability. BIT is considered to be a potential technique for fault detection and fault isolation (FDFI) that is inherent in system design. However, for many systems, BIT by itself cannot isolate the fault to a single unit without additional external diagnostic efforts. FTA is a deductive technique where a top-level undesired system failure is analyzed logically to find the basic cause of that failure. FTA generates a graphical representation of the logical relationship between an undesired system failure and the basic fault events associated with it. This presentation will describe how to resolve FI ambiguity by combining FTA technique with BIT information as a diagnostic instrument.

RMMR 2019, Oct. 15 - 16, San Antonio, TX 2

Reliability Maintenance and Managing Risk Conference 2019

Agenda

RMMR 2019, Oct. 15 - 16, San Antonio, TX 3

• Problem Description / Challenge
• Built-In-Test (BIT) Design Background
• An Affordable Methodology For FI Diagnosis:

Apply Fault Tree Analysis (FTA) in Conjunction With BIT

• How To Develop A Fault Diagnosis Methodology ? : FTA + BIT
• A Case Study for A Sample Hypothetical System
• Case Study Conclusion: Is Unambiguous Fault Isolation

Possible?

• Benefits And Drawbacks of FTA + BIT Methodology
• Summary

Reliability Maintenance and Managing Risk Conference 2019

Problem Description / Challenge

RMMR 2019, Oct. 15 - 16, San Antonio, TX 4

Fault Isolation (FI) Ambiguity: FI Ambiguity occurs when:

• A fault is associated with multiple

replaceable units

• When fault occurs on interface of

units

FI Ambiguity Consequences:

• Maintenance events ““No Fault

Found” / “Cannot Duplicate” occurs

• Increases in Support cost
• Decreases in System availability.

The Challenge:

Significant Demand for Unambiguous Fault Isolation in Maintenance Technology

Reliability Maintenance and Managing Risk Conference 2019

Built-In Test Design Background

• Built-In Test (BIT)

– A potential technique for fault detection and fault isolation (FDFI) integrated within system design – Designed to detects potential failures using Failure Mode Effect And Criticality Analysis (FMECA)

• BIT Coverage Limitation

– Does not always guarantee a capability to identify the cause of a fault – Does not alone accurately fault isolate to a single unit

• BIT limitation

– FMECA process may have different potential failures with very different risk priority numbers ▪ Discrepancy makes it difficult to identify and prioritize failures ▪ Lacks reliability and validity when used in isolation

Require an Additional Diagnostic Implementation to Resolve FI Ambiguity

RMMR 2019, Oct. 15 - 16, San Antonio, TX

Reliability Maintenance and Managing Risk Conference 2019

Apply Fault Tree Analysis (FTA) in Conjunction With BIT

• Fault Tree Analysis (FTA)

– FTA is a top-down approach, which provides a logical framework for understanding the ways a system can fail for an undesired system failure – FTA is a deductive technique where a top-level undesired system failure is analyzed logically to find the basic cause / event

• New FI Diagnosis Methodology

– BIT design uses Failure Mode Effect and Criticality Analysis (FMECA) to capture potential fault detection – FMECA is bottom-up inductive analytical approach, whereas FTA is a top- down – Apply FTA in conjunction with BIT during preliminary design phase – Provides an opportunity to overcome any BIT deficiencies – Is extremely effective and affordable

Apply “FTA + BIT” Method During Design Phase to Reduce FI Ambiguity

RMMR 2019, Oct. 15 - 16, San Antonio, TX

Reliability Maintenance and Managing Risk Conference 2019

Top Event Construct Fault Tree Determine a set of Basic events that Causes the top event Incorporate BIT with each cause Associate BIT with respective component Is Basic cause associated with a single component? Unambiguous FI achieved YES NO FTA FTA + BIT

“FTA + BIT” Method Flow Chart

RMMR 2019, Oct. 15 - 16, San Antonio, TX

Reliability Maintenance and Managing Risk Conference 2019

“FTA + BIT” Method Steps:

Top- Event

• Select Top-level event and set physical boundary

FTA

• Perform FTA and construct fault tree for top-level event

FTA + BIT

• Transform the generated fault tree knowledge into BIT

Preliminary Design FI

• Unambiguous FI occurred?
• If not, then find all hidden faults to mitigate design

RMMR 2019, Oct. 15 - 16, San Antonio, TX

Reliability Maintenance and Managing Risk Conference 2019

Infrared Sensor Imaging Process Comm_2

System Processor Sensor Image Processor Assembly Thermal Imaging Assembly LRU Controller Sensor Camera

LRU1 LRU3 LRU2 Comm_1 Video Image

A Case Study for A Hypothetical System

RMMR 2019, Oct. 15 - 16, San Antonio, TX

Reliability Maintenance and Managing Risk Conference 2019

A Case Study (Cont.)

Preliminary BIT Design and Definition for IR Sensor Using FMECA

BIT Definition For IR Sensor Process Developed with use of FMECA

BIT Error Code BIT Test Name BIT Test Description PFL MFL T1 Sensor_Comm_Err Fault in receiving Sensor image from LRU2 to LRU1 IR_Sensor_Fail LRU 1, LRU2 T2 Video_Image_Fail Fault in receiving Video image from LRU 3 to LRU2 IR_Sensor_Fail LRU2, LRU 3 T3 Comm1_Err LRU1 to LRU2 Interface Communication fault IR_Sensor_Fail LRU1, LRU2 T4 Comm2_Err LRU3 to LRU2 Interface Communication fault IR_Sensor_Fail LRU2, LRU 3 T5 Sensor_Ctl_Comm_E rr Fault in Sensor Camera and Sensor Controller interface IR_Sensor_Fail LRU 3 T6 Carousel_Err Fault in Carousel IR_Sensor_Fail LRU2, LRU 3 T7 Thermal_Image_Com m_Err Fault in Thermal Image processing IR_Sensor_Fail LRU2, LRU 3 T8 Sensor_Motor_Err Fault in Sensor Motor IR_Sensor_Fail LRU2, LRU 3 T9 Sensor_Temp_Err Temperature fault in sensor IR_Sensor_Fail LRU2, LRU 3 T10 Loop_back_Comm1 Fault in communication interface loopback check IR_Sensor_Fail LRU1 T11 Comm1_Intrf_Err Fault in communication interface in LRU2 IR_Sensor_Fail LRU2 T12 Sensor_Init_Fail Fault in sensor initialization IR_Sensor_Fail LRU 3 T13 Image_Proc_Init_Err Fault in Image processor initialization IR_Sensor_Fail LRU2

RMMR 2019, Oct. 15 - 16, San Antonio, TX

Reliability Maintenance and Managing Risk Conference 2019

A Case Study (Cont.): Fault Tree for Top Event “IR Sensor Fail”

Basic Events

RMMR 2019, Oct. 15 - 16, San Antonio, TX

Reliability Maintenance and Managing Risk Conference 2019

A Case Study (Cont.): Fault Tree for Top Event “IR Sensor Fail”

Ambiguity in Basic Event

BIT Preliminary Design has Ambiguity in Fault Isolation

RMMR 2019, Oct. 15 - 16, San Antonio, TX

Reliability Maintenance and Managing Risk Conference 2019

A Case Study (Cont.):

Unambiguous Fault Isolation Captured Hidden Faults

T14 Carousel_BIT_Err Fault check for carousel controller T15 Carousel_FDBK_Err Fault check for carousel feedback message T16 Sensor_Motor_BIT_ Fail Fault check for sensor controller T17 Sensor_Motor_FDB K_Err Fault check for Senosr motor feedback message T18 Cooler_Ctl_BIT Fautl check for cooler controller T19 Cooler_FDBK_Err Fault check for cooler feedback message

RMMR 2019, Oct. 15 - 16, San Antonio, TX

Reliability Maintenance and Managing Risk Conference 2019

A Case Study (Cont.):

Improved BIT Design using “FTA + BIT” Method

“FTA + BIT” Method Resolved FI Ambiguity

BIT E rror Code BIT Test Name BIT Test Description PFL MFL T1 Sensor_Comm_Err Fault in receiving Sensor image from LRU2 to LRU1 IR_Sensor_Fail LRU 1, LRU2 T2 Video_Image_Fail Fault in receiving Video image from LRU 3 to LRU2 IR_Sensor_Fail LRU2, LRU 3 T3 Comm1_Err LRU1 to LRU2 Interface Communication fault IR_Sensor_Fail LRU1, LRU2 T4 Comm2_Err LRU3 to LRU2 Interface Communication fault IR_Sensor_Fail LRU2, LRU 3 T5 Sensor_Ctl_Comm_E rr Fault in Sensor Camera and Sensor Controller interface IR_Sensor_Fail LRU 3 T6 Carousel_Err Fault in Carousel IR_Sensor_Fail LRU2, LRU 3 T7 Thermal_Image_Com m_Err Fault in Thermal Image processing IR_Sensor_Fail LRU2, LRU 3 T8 Sensor_Motor_Err Fault in Sensor Motor IR_Sensor_Fail LRU2, LRU 3 T9 Sensor_Temp_Err Temperature fault in sensor IR_Sensor_Fail LRU2, LRU 3 T10 Loop_back_Comm1 Fault in communication interface loopback check IR_Sensor_Fail LRU1 T11 Comm1_Intrf_Err Fault in communication interface in LRU2 IR_Sensor_Fail LRU2 T12 Sensor_Init_Fail Fault in sensor initialization IR_Sensor_Fail LRU 3 T13 Image_Proc_Init_Err Fault in Image processor initialization IR_Sensor_Fail LRU2 T14 Carousel_BIT_Err Fault check for carousel controller IR_Sensor_Fail LRU2 T15 Carousel_FDBK_Err Fault check for carousel feedback message IR_Sensor_Fail LRU 3 T16 Sensor_Motor_BIT_ Fail Fault check for sensor controller IR_Sensor_Fail LRU2 T17 Sensor_Motor_FDB K_Err Fault check for Senosr motor feedback message IR_Sensor_Fail LRU 3 T18 Cooler_Ctl_BIT Fautl check for cooler controller IR_Sensor_Fail LRU2 T19 Cooler_FDBK_Err Fault check for cooler feedback message IR_Sensor_Fail LRU 3

Additional BIT These were missing in Preliminary BIT Design which was developed by using FMECA

RMMR 2019, Oct. 15 - 16, San Antonio, TX

Reliability Maintenance and Managing Risk Conference 2019

Case Study Conclusion: Is Unambiguous Fault Isolation Possible?

IR Sensor BIT

IR Sensor Preliminary BIT Design has FI Ambiguity BIT alone cannot actually fault isolate to a single replaceable unit

FTA + BIT Technique

Allowed unambiguous FI at no additional cost

Reduced FI ambiguity

Moving Forward: Start Using FTA + BIT for Fault Diagnosis Methodology During Design Phase

RMMR 2019, Oct. 15 - 16, San Antonio, TX

Reliability Maintenance and Managing Risk Conference 2019

Benefits And Drawbacks of “FTA + BIT” Methodology

• The fault tree explicitly represents all the various

relationships that are necessary to evaluate the top-level event

• Identifies all hidden failures which are overlooked by BIT,

therefore provides an opportunity to overcome any BIT drawbacks

• Allows easy mapping of BIT information into fault tree
• Applying this method during early design phase will be cost

effective

• May lead to a very large tree if the analysis is extended in

depth

• Relies on the skills of an analyst
• Can be costly in time & effort if there is a lack of thorough

system knowledge and may not resolve FI ambiguity

RMMR 2019, Oct. 15 - 16, San Antonio, TX

Reliability Maintenance and Managing Risk Conference 2019

Summary

• BIT has limitation for unambiguous fault isolation
• FTA is a deductive technique where a top level undesired system

failure is analyzed logically to find the basic cause of that failure

• Applying FTA in conjunction with BIT is an innovative, affordable

and effective technique for resolving FI ambiguity and overcoming BIT limitations

• FTA + BIT technique does have a capability to resolve fault

isolation ambiguity

• Applying FTA + BIT method during design phase will reduce

maintenance support cost and increase system availability

• Moving Forward: Apply FTA in conjunction with BIT for affordable

fault isolation diagnosis instrument

RMMR 2019, Oct. 15 - 16, San Antonio, TX

Reliability Maintenance and Managing Risk Conference 2019

RMMR 2019, Oct. 15 - 16, San Antonio, TX