Ammunition Stockpile and Ammunition Stockpile and ServiceService--life Reliability:life Reliability:

Ammunition Stockpile and Ammunition Stockpile and ServiceService--life Reliability:life Reliability:ServiceService life Reliability: life Reliability: Improvement Efforts at Improvement Efforts at

US Army ARDECUS Army ARDEC

ServiceService life Reliability: life Reliability: Improvement Efforts at Improvement Efforts at

US Army ARDECUS Army ARDECPresented for Precision Strike Association Firepower ForumPresented for Precision Strike Association Firepower ForumPresented for Precision Strike Association Firepower ForumPresented for Precision Strike Association Firepower Forum

Jason L. Cook, Ph.DChief, QE&SA Sciences DivisionQE&SA – ARDEC – RDECOM

Jason L. Cook, Ph.DChief, QE&SA Sciences DivisionQE&SA – ARDEC – RDECOM

The ProblemThe Problem

• Testing for reliability through the life of a smart-munition is not financially feasible– Firing 100+ rounds from each strata– Every 3-5 years– For the life of the item

• Waiting until the item is bad does not provide enough time to buy moreprovide enough time to buy more– 2 to 6 year cycle time from need to field

The Solution: The Solution: Predictive Predictive Stockpile ManagementStockpile Management

Failure Mode and

Mode-MechanismModelSensors

Predictive Stockpile +

=

+CONOPS &

Data Repository+Mechanism Management

p y

An Example…An Example…

• Identify Failure Mode– What fails?

• Identify Failure Mechanism– What causes the failure?

• Determine rate of degradationH l d it t k t f il?– How long does it take to fail?

• Correlate and synthesize When will it fail?– When will it fail?

– When should I produce more?– Which items are at risk?– Which items are not?Which items are not?

Ammunition Stockpile ReliabilityAmmunition Stockpile ReliabilityProgramProgram

Elements of the ASRP:

Design for Storage LifePredictiveTesting Design for Storage Life

Predictive Engineering

Ammunition Surveillance Program

Testing

g

Function (Reliability) Testing

Laboratory testing programFunctionTesting

LabTesting Laboratory testing program Testing Testing

Ammunition Surveillance

Storage Life PredictionsStorage Life Predictions

• Reactive (Fielded Items)• Proactive (Development ( )– Perform function testing per

ASRP Plan– Analysis of variance

Proactive (Development Items)– Analogy based analysis to

determine at risk, life limiting • Age• Lot• Manufacturer

St l ti /t

, gitems

– Accelerated life testing to predict storage life

• Storage location/type• Design revisions

– Detect reliability degradation trends

• Controlled • Uncontrolled

– Determine design changes or mitigations to e tend life trends

– Predict breach of lower reliability threshold

or mitigations to extend life

Predictive Technology (ALT) can be used for fielded items also

InitiativesInitiatives

• Policy – Army Regulations and local installation application policies

• Process – Lean Six Sigma Green Belt Project to refine methodsmethods

• Data - Predictive Summary Report and BenchmarkingData Predictive Summary Report and Benchmarking

• Application – Synergistic programs addressing multiple pp y g p g g pitems or classes of items

Goal – Enable Predictive Stockpile Management

ASRP PolicyASRP Policy

• Memo documenting policy requirementsAmmunition Stockpile Reliability Program– Ammunition Stockpile Reliability Program

• AR 702-6– Ammunition Surveillance

• AR 740–1, AR 702–12, and AR 700-142 – Required at time of MR

• Key responsibilities of PM and ARDECKey responsibilities of PM and ARDEC– Baseline performance and reliability– Identify life-limiting components– Identify acceptable limits of degradation– Design and build unique inspection/test equipment

SSGB ProjectSSGB Project

• Objectives:Develop process map for creation of

• Approach:– Define current process– Develop process map for creation of

ASRP Plan– Improve timeliness and value of the

ASRP Plan and its execution• Completed at time of MR

– Improve quality of plans to include:

Define current process– Measure and Analyze results of current

process and adherence to AR– Improve and Lean process to provide

more value and synergy across ammo classesImprove quality of plans to include:

• Greater use of predictive engineering and accelerated life testing

• More item and failure mode unique testing and inspections

• Add Ammunition Peculiar Testing Equipment

– Institute Controls to ensure continual improvement

Equipment• Add detailed test procedures

– Institute Configuration Management• Approval routing• Revision Management• Document MaintenanceDocument Maintenance• Define how ECP and MIF information is

added to ASRP Plan

Unified ASRP ApproachUnified ASRP Approach

PredictiveTestingR(t) Testing(t)

CI

FunctionTesting

LabTesting

time

CI

Testing Testing

Ammunition Surveillance

Practical ALTPractical ALT

Define:Failure Mode

HALT

Measure:Sensitivity to Stresses

DoE based ALT

Analyze:Correlation to CI’s

Improve:

ANOVA

FRACAS Improve:Design out if possible

Control:Monitor stresses & CI’s

FRACAS

Sensor & Lab test Monitor stresses & CI s

CBM for Ammo

Predictive Summary ReportPredictive Summary Report

• Compilation and update of tests and analyses capturing environmentally susceptible items and componentsSo rces• Sources:

– ASRP function testing– ASRP surveillance inspections– DIF/MIF reports

FAT/LAT lt– FAT/LAT results– Predictive Engineering/Aging Studies

• Motivation– Identify common causes and risk for LCMC managed items– Provide repository of data to expedite MR process and avoid duplication of

effort– Determine candidates for further investment and investigation

• Aging program• In-situ sensing• In-situ sensing • Telemetry• Additional functional, lab, or surveillance sampling

Sensor Sensor and Database Projectand Database Project

• Investigate COTS sensors – Literature review and continued work with UMD Consortium– Identify customer requirements (cost, size, IO, resolution)– Classes of sensors

• Cheap and simple for cheap and simplep p p p• Ensure CBA/ROI is favorable

• Qualify one or more from each class– Durability - Sensor can’t fail before round– Accuracy – Sensor data can’t drift with time– Interoperability(E3) – Eliminate interference/safety concerns

• Data Analysis and Warehouse– Open Architecture– Tailorable

Self definable models– Self-definable models

• Application guidance– Common I/O and data collection methods– Coordination with JMC QASAS

Value of Temperature DataValue of Temperature Data

Predicted Life

Life Estimate between

2 and 22 years

LOW88B081-006

LOW88B081Y006

PXC-89-459

PXC-89-60

PXC-90-62 Pressure Correlation

450.0

500.0

Data PointsFailure ThresholdPrediction Threshold

LOW82B021Y006

LOW84M026-013

LOW85B026-021

LOW86C028-016

LOW86J029-005

Lot #

350.0

400.0

Act

ual

0 5 10 15 20 25

KC-22-14

LOW82B021-006

YearsHand Held Signal Device 250.0

300.0

Failure Predictable 200.0

250 300 350 400 450

Predicted

withOnly 1.7% escapes

Electronics Predictive Electronics Predictive ProjectProject

• Reliability Ch t i ti f SMT

PBGA

Solder balls

Characterization of SMT components in temperature cycling environment 18 M 502 5

Histogram of C1Normal

Printed Wiring Board

environment

• Predictive algorithm development to identify

Freq

uenc

y

18

16

14

12

10

8

6

4

Mean 502.5StDev 72.45N 100

development to identify incipient failures

D t ti ( )

C1650600550500450400350

4

2

0

• Demonstration sensor(s) from Low Cost sensor program (if funded)

Predictive Stockpile Mgmt. Predictive Stockpile Mgmt. RoadmapRoadmap

SLow

Cos

PS

ES st S

ensorssS

M

ODSODS

Sum

mar

y M

T Predict

ODSODSODSODSODS

PE

Sive P

rogra

Healthy: H0Degraded: H2 Degraded: H1

16

am

Degraded: H3σ1=Vσ2

Healthy: H0

Degraded: H4σ2= σ2/V

Degraded: H3σ1=Vσ2

Healthy: H0

Degraded: H4σ2= σ2/V

Degraded: H3σ1=Vσ2

Healthy: H0

Degraded: H4σ2= σ2/V

-M M0

0

POCs:Jason.Cook1@us.army.milKelley.Caflin@us.army.mil

Questions?Questions?