You are on page 1of 8

Implementing Reliability Tools in the existing

Process flow for New Product Development, to


enhance the Product Reliability
Pavan Kumar. P
Product Validation and Reliability Department
Amara Raja Electronics Limited
Diguvamagham, Chittoor, India
padi@amararaja.co.in

Abstract— In this paper we present steps taken to incorporate


the Reliability Tools in the regular Process flow development of a
New Product. The ultimate goal of this activity is to ensure that
the Robust design with high reliability is delivered to the end
customer, and as a result attain zero field complaints. Also,
attached 2 illustrations that has been used during calculating the
MTBF for the products designed at AREL.

Keywords— Reliability Tools, Robust Design, Process


Enhancement, Chi-Squared Distribution, Weibull Analysis.

I. INTRODUCTION
Generally the strategy of any Manufacturing Industry is to
focus on Quality of the Product. In this regard many initiatives
like Quality Control Circles, Green Belt, Black Belt Projects,
Continual Improvement Programs are been taken. All the
employees concerned to the process are well trained and
groomed to perform the same. But ultimately, Quality refers to
instantaneous Performance of the product (i.e., whether the
Product is functioning good at that moment). Quality is
everything until put into operation, whereas Reliability is
ability of an item to perform a required function under stated
conditions for a specified period of Time. Hence by
implementing the Reliability tools in the regular Process Line,
we will be able to enrich Customer Satisfaction, eliminate
Failures in Line and field. In short, Reliability is Quality over
time.

II. REGULAR PROCESS FLOW


The following flow Chart (Figure – 1) defines the Process
flow for New Product development activity that was being
followed.

Figure 1: Existing Process Flow Chart for New Product Development


III. RELIABILITY DEFINITION V. PROPOSED PROCESS FLOW CHART FOR NEW PRODUCT
Reliability is the probability that an item will perform a DEVELOPMENT
required function under stated conditions for a specified period
of Time.
The reliability process is applicable for all stages from
design to Product reaching the end Customer.

Figure 2: Bath Tub Curve

IV. RELIABILITY PROCESS FLOW


The reliability process spans all stages of design and
production – from requirements, through design, component
testing, production and sub-contractor evaluation, and finally
integrative stress testing to prove the overall reliability of the
system.
Reliability tests require expensive resources and precious
time. There are estimation tools to evaluate the life time of a
system. However, the available tools are notorious for
neglecting various failure mechanisms and stresses. The
ultimate goal is achieved by a multi-level reliability plan which
is described in Figure 3.

Figure 4: Proposed flow Diagram for new Product Development in-


line with Reliability Tools

VI. COMMUNICATION AND TRAINING


To initiate a Reliability Process, Communication is the first
Step. This is ongoing and need to be delivered to all levels of
Figure 3: Applicable Tools for different stages of the the Team. It is not just enough to define the Tools for each
Product development
stage but also required to allot the person dedicated to the For the purposes of our calculations an activation value of
activity. The Person responsible has to be well aware of 0.53 eV is used. This agrees with the thumb rule of doubling
utilizing the tool. So, a list is drafted in Annexure – 1 where the equivalent product usage hours for every 10°C above the
tools, reference standards in each stage is explained in detail. A normal operational temperature. For additional information on
training Calendar is framed to train the Employees, which is the reliability of electronic components reference MIL-HDBK-
attached as Annexure - 2. 217.
Currently two Solar Products are being traded. These
Products are paved way to Dispatch only after fulfilling the D. Acceleration Factor, Af
Reliability assessment.
The above mentioned tools were implemented during the
design development stage and the final product is subjected to
Reliability Testing.
After Design freezing, the samples were tested for
Reliability and the calculations are detailed as illustrations:
Af = acceleration factor
Ea = activation energy in electron-volts (eV) = 0.53 eV
VII. ILLUSTRATION – 1
k = Boltzmann’s constant = 8.617 x 10-5 eV/Tk
A. Objective: Ts = Temperature of normal operation, in degrees Kelvin
= 318°K (45°C)
To determine the MTBF of the ABC Product that can
perform a required function under 45⁰C condition for specified Tt = Temperature of operation during test,
mission time of 500 hours. The samples are put on accelerated
life test for the fixed Test time. in degrees Kelvin= 333°K (60°C)
Tk = Kelvin Temperature
B. Data:
e = 2.71828 (mathematical constant)
To Sample Size : 5 No’s Af = (2.71828)^[6150.632* {(1/318) – (1/333)}]
Test Environment : 60⁰C Af = 2.3898
Test Start Date : May 28, 2015 E. MTBF Calculation:
Test End Date : June 16, 2015
Mission Time : 500hrs
No. of Failures Observed: NIL
Reliability Methodology: Chi-Squared Distribution.
where:
C. Activation Energy, Ea:
MTBF= Mean time between failures
It is an empirical value that the minimum energy required 2 is the Numerical value
to initiate a specific type of failure mode that can occur within
a technology type. Oxide defects, bulk silicon defects, mask N is the Number of elements = 5 SHLS Samples
defects; electro-migration and contamination are some T = Total test time = 384 hours
examples of such failure modes, each with a unique associated
Ea. Af = Acceleration factor from Arrhenius equation = 2.390
α = 0.1
(We considered 90% confidence level for this purpose:
(for Consumer Electronics, Confidence level from 80% to
90% can be considered)
n = number of failures (used to calculate degrees of
freedom); we observed zero failures during Test duration. So, n
=0
χ²[α,(2(n+1))] = Chi-squared distribution function = 4.605

Figure 5: Activation Energy for different Failure Modes


Substituting in the above formula, ƞ – Slope / Scale Parameter
MTBF = 2594.79 hours Г – Gamma Function
Failure rate, λ hr = 0.00038539 λ – Failure rate

F. Result: E. Weibull Analysis:


The MTBF of the ABC Product to perform a required
function under 45⁰C condition for specified mission time of Table 2
500 hours is 2595 hours. Time to Failure Failure Order Median Rank
F(T)
in Hours Number %
VIII. ILLUSTRATION – 2
407 1 12.96 0.13
A. Objective: 2 31.48 0.31
521

To determine the Reliability of the XYZ Product that can 736 3 50.00 0.50
perform a required function under 50⁰C condition. The samples 4 68.52 0.69
762
are put on a life Test and run until failure.
936 5 87.04 0.87
B. Data:
Sample Size : 5 No’s
Test Environment : 50⁰C
Mission Time : 500Hrs
Where, MR = Median Rank
No. of Failures Observed: Given below
i = Failure Order number
Reliability Methodology: Weibull Analysis
N = Sample Size
C. Failure Details: Table 3
Y X Y
Table 1: Failure Details
(1/(1- ln(ln(1/(1- ln[-ln(1-
Time to Failure F(T) T ln(T)
Type of Failure observed F(T))) F(T)))) F(T))]
in Hours

407 LCD Fail 0.13 1.15 -1.97 407 4.56 -1.97

521 Controller Fail 0.31 1.46 -0.97 521 4.79 -0.97


736 Controller Fail
0.50 2.00 -0.37 736 4.91 -0.37
762 LCD Fail
0.69 3.18 0.14 762 5.15 0.14
936 Controller Fail
0.87 7.71 0.71 936 5.43 0.71

D. Referecne Terminologies:

Hrs – Hours
LCD – Liquid Crystal Display
MR – Median Rank
MTBF – Mean Time between Failures
Pdf – Probability Density function
R(t) – Reliability Function
t – Mission time
β – Shape Parameter

Figure 6
X. REFERENCES
Slope=Beta = β = 3.074 (from Excel formula)
Intercept = -20.3779 (from Excel formula) 1. CFR – CRP Program course Notes
Eta = ƞ = Exp (-Intercept/Slope) 2. “Tests for Validity of a Constant Failure Rate Assumption
= 756.2277 hours – Chi - Square Test ”, IS:8161 (Part 6/Sec 1) – 1983
Time, t = 500 hrs (considered) 3. M. Levin and T. Kalal, “Improving Product Reliability”,
John Wiley, 2003
Generally,
4. Reliasoft, webpage
If β = 1, Exponential Distribution
5. “Practical Reliability Engineering”, 4th Edition, Patrick
If β = 2, Rayleigh Distribution D. T. O’Connor, 2002
If β > 3, Normal Distribution 6. “Improving Product Reliability: Strategies and
Here, β > 3, so Normally Distributed Implementation”, Mark A. Levin and Ted T. Kalal, 2003
So, using 2-parameter Weibull pdf

MTBF =

= 756.23. Г ((1/3.074) + 1)
= 756.23. Г (1.33)
= 756.23* 0.89338
(from Gamma function Table)
= 675.60 hours
Reliability, R(t) =
= exp [-(500/ 756.23) ^ 3.074]
= 0.7556
Reliability for the mission time of 500 hrs, R(500) =
75.56%

F. Result:

The Reliability of the XYZ Product to perform a required


function under 50⁰C condition for specified mission time of
500 hours is 75.5%.

IX. CONCLUSION

Though it is tedious to incorporate new procedures in a


well-established line, with due adherence to Reliability Tools
and concept and strong commitment to accept the change, we
can support highly reliable Product to the highly competing
Market.
ANNEXURE – 1:
Process Flow / Department wise applicable Tools for ensuring Product Reliability:

ACTIVITY DEPT. TOOL USED / SKILL REQUIREMENT REFERENCE STANDARDS / Documents

Technical knowledge Individual competency

Stress Analysis Bellcore (Telcordia)

Reliability Prediction MIL HDBK 217F


Software based (or) Analytical calculation MIL HDBK 108
with the available data MIL HDBK 251
System Design RAD
MIL STD 1629A,
FMEA, FMECA - Design Level IEEE STD 7858)Task 204)
MIL STD 470A (Task 204)
Product Testing and Verification ARE-RAD-Verification-Report-format

IEC 62040, IEC 60950, MIL STD 882B, Ingress


Complying System Safety requirements
protections

Technical knowledge Individual competency

Product Testing and Validation ARE-PVR-OBS-report-format

Thermal Cycling ARE-PVR-OBS-report-format


Product Validation PVR
Highly accelerated Life Testing MIL HDBK 344A
Environmental Stress Screening MIL STD 2164

MIL STD 882B, IEC 62040, IEC 60950, ingress


confirming System safety requirements
Protections

Line Testing RAE-RAD-Line-test-Procedure


PCB, System MIL HDBK 344A
Production Highly Accelerated Stress Screening
Assembly, Magnetics MIL STD 2164
Burn-in Test ARE-RAD-Burn-in-Test-Procedures

MIL STD 1629A,


FMEA, FMECA - Operations level IEEE STD 7858)Task 204)
QAD and Control Plan MIL STD 470A (Task 204)
Product Quality
Process Engineering Process Parameter setting AREL-Apex-Manual
Internal AREL-QA-Procedures-manual
AREL-PE-Procedures manual

Package Engineering, Logistics Packaging MIL STD 2073

FRACA - Failure Reporting, Analysis and


Corrective action System
Visual Inspection
Field Failures
Electrical Testing
Reporting and Servicing, QAD, PVR MIL STD 2155
Radiography
Analysis
Decapsulation
Internal Visual Inspection
Scanning Electron Microscope
Availability,
Maintainability, Servicing, PVR Maintainability Prediction MIL HDBK 470, 471, 472
Reparability
ACTIVITY DEPT. TOOL USED / SKILL REQUIREMENT REFERENCE STANDARDS / Documents

RQT (Reliability Qualfication Testing),


MIL STD 781A
Reliability Testing PVR PRAT (Production Reliability Acceptance
MIL STD 781D
Testing)

Reliability MIL STD 781A


PVR RDT (Reliability Demonstration Testing),
Demonstration MIL STD 781D
ANNEXURE – 2:

Training Schedule:

You might also like