FIDES: A Physics of Failure Solution for Reliability Predictions ASTR Workshop, St. Paul 10 th to 12 th September 2014 Franck Bayle Thales Chanthachith Souvanthong ON Semiconductor
Why reliability prediction? Customer requirements Reliability MTBF, Probability of Mission Success Supplier needs Sub assembly Reliability allocation Testability Fault coverage / isolation rate, False alarm rate. Safety Probability of Feared Events Maintenance Direct maintenance costs Spare parts number assessment Through-life support assessment & Repairs Number Prediction Stocks assessment & Obsolescence
Design For Reliability Concept Reliability MTBF, Probability of Mission Success Safety Probability of Feared Events Testability Fault coverage / isolation rate, False alarm rate. Preliminary Reliability Prediction at system level Components Reliability allocation Reliability Prediction at component level Why reliability prediction? Production & Operations Through-life support assessment & Repairs Number Prediction Stocks assessment & Obsolescence Maintenance Direct maintenance costs Final Reliability Prediction at equipment level Spare parts number assessment Final Reliability Prediction at system level OEM System Provider Component Supplier
Why FIDES? Existing predictive reliability methods are inadequate q MIL-HDBK-217 is obsolete The available Reliability Data book prediction are obsolete because they don t cover actual component technologies and state of the art of acceleration laws (i.e. MIL-HDBK-217 is not maintained since 1995) q IEC 62380 TR Ed.1 (RDF 2003) or PRISM / 217+ don t propose complete answers o Data not updated since 2003 New economic challenges need to know realistic system reliability to engage many years support contracts with customers
Assumptions of reliability prediction Reliability prediction is only one step of the reliability process. Some assumptions must be checked during the Design For Reliability activities performed during the development phase as : q Product robustness to reduce induced failures (overstresses) Derating analysis Worst case analysis Highly Accelerated Stress (HALT) q Life limited component analysis to reject wear-out failure mechanisms after the product service life q Accelerated tests for specific components q Optimized screening (HASS) q Life profile (stress type, stress level, operational phases)
MIL-HDBK-217 Weaknesses Not maintained since 1995 Only valid for on phases Physic of Failures not adapted (temperature only) 90% component failure, 10% life cycle Not adapted for recent technologies life cycle not taken into account F Difficulties for the customer to select a supplier when adjustments are different from one supplier to the other and not justified. F Need an adjustment from the field to be realistic
MIL-HDBK-217 Weakness MIL-HDBK-217 cannot be directly used but many correction factors are needed: Package adjustment Adjustment from the field at equipment level ( C1 Π + C Π ) Π Π Krec λ = 2 T E Q L Die Complexity adjustment Activation energy not representative of technology Mission profile adjustment Component family adjustment Integrated circuits failure rate
IEC TR 62380 Weaknesses Not maintained since 2003 Unable to sort by COTS manufacturers Physic of Failures not adapted (missing humidity & vibration) Not adapted for recent technologies Package FIT rate includes both solder joints and bonding failure modes life cycle not taken into account F Customer difficulties for the supplier choice when adjustments are different from a supplier to another and not justified F Need an adjustment from the field to be realistic
FIDES is filling the gaps Valid for all operational phases Adapted for recent technologies Based on Physic of Failures Take product life cycle into account Can be used by any suppliers Provide realistic prediction Regularly reviewed and maintained Available for components, boards & sub-assemblies
FIDES global modelisation Origin cause of failure Components Process impact Process Process failure distribution versus life Cycle Spécification Exploitation/ Maintenance Design Intégration Manufacturing components technologies TECHNOLOGY RELIABILITY Technologies Electronic Component family Classification Detailed technologies =>Recent technologies covered Profile Description Phases breakdown Duration of each phase Functioning condition (on/off) Ambient temperature (constant) Thermal cycling (duration, frequency, amplitude) Relative Humidity Vibrations Chemical (pollution) Life cycle Audit 170 Questions PROCESS USE System Life Profile
FIDES global modelisation! The general equation is : λ = λ = λ λ Physique Physical π π Part _ manufacturing π Pr ocess Part_ manufacturing π Process λphysical π Part _ manufacturing Is the physical contribution (technology and stresses) Is the quality and technical control of the item s manufacture π Pr ocess Is the quality and technical control of the processes of development, manufacture and operation of the product containing the item 11
Physical failure rate λ Physical = Physical _ contributions ( λ ) 0 Πacceleration Πinduced λ 0 The basic failure rate is based on part technologies Π acceleration The environmental acceleration factor indicates the sensitivity to a physical contributor : electrical, temperature, thermal cycling, humidity, mechanical and chemical (rated stresses) Πinduced Is the induced factor contribution (overstress), classically viewed in a given application 12
Physic failure rate Physic of Failures laws Mission profile λ physic = ( λo _ TH Π + λo _ Tcy Π + λo _ RH Π + λo _ grms Π ) TH Tcy RH grms
Induced failures are catalectic failures that can occur because of : An usage of the product out of its specification range; This could be due to: Bug in the design Cascading failure Unexpected use of the product Wrong manipulation Induced factor in FIDES are of mechanical (MOS), electrical (EOS) and thermal (TOS) origin. Π induced
ΠPart_Manufacturing Π part manufacturing characteristics take into account the quality and technical control of the item s manufacture, Discriminate between the manufacturers, Applicable to the models components, Electronic boards and Sub-assemblies Definition of the parameters Manufacturer quality assurance level Qa manufacturer Component quality assurance level Qa component Component reliability assurance level RA component Experience that the buyer of the item may have of his supplier ε Π Part_Manufacturing = δ1. 1 Part_Grade δ 2 e
ΠProcess There is a constant improvement of the hardware component reliability itself. The part corresponding to the failures related to the life cycle of the product is becoming the most important contributor. Components Life cycle π = e δ.(1 Pr ocess _ Grade ) Pr ocess Specification Production Integration Maintenance Design Usage
At equipment level FIDES models experimentation 100 Prediction / Field return MIL-HDBK-217F FIDES 10 1 Single Aisle civilian Long range civilian Military mirage 2000-5 Global Express regional jet civilian Hélicopter EC135 DASH8 civilian winged aircraft 0.1
At component family level FIDES models experimentation 1000.0000 100.0000 Field return Lower bound Field return Field return upper bound FIDES/Field return 10.0000 1.0000 0.1000 0.0100 0.0010
Support of FIDES q FIDES is distributed for free (except UTE version of the document) q A maintenance and development structure was created for FIDES in 2007 This structure is part of a French association dedicated to reliability and safety : the IMDR (www.imdr.eu) q This structure is divided in 3 Working Groups : FIDES evolutions FIDES international promotion and communication Guides of Best practices and application to other domains q More than 35 companies registered q Only French members for the moment, but the whole group agrees to become international