3. RELIABILITY INTEGRATION
“the process of seamlessly
cohesively integrating reliability
tools together to maximize
reliability and at the lowest
possible cost”
4. Reliability vs. Cost
Intuitively, one recognizes that there is some minimum total
cost that will be achieved when an emphasis in reliability
increases development and manufacturing costs while
reducing warranty and in-service costs. Use of the proper
tools during the proper life cycle phase will help to
minimize total Life Cycle Cost (LCC).
CRE Primer by QCI, 1998
5. Reliability vs. Cost, continued
TOTAL
COST
OPTIMUM CURVE
COST
POINT RELIABILITY
PROGRAM
COSTS
COST
WARRANTY
COSTS
RELIABILITY
6. Reliability vs. Cost, continued
In order to minimize total Life Cycle Costs (LCC),
a Reliability Engineer must do two things:
choose the best tools from all of the tools available
and must apply these tools at the proper phases of a
product life cycle.
properly integrate these tools together to assure that
the proper information is fed forward and backwards
at the proper times.
7. Reliability vs. Cost, continued
As part of the integration process, we must choose a
set of tools at the heart of our program in which all
other tools feed to and are fed from. The tools we
have chosen for this are:
HALT and HASS
8. HALT and HASS Summary
Highly Accelerated Life Testing (HALT) and Highly
Accelerated Stress Screening (HASS) are two of the best
reliability tools developed to date, and every year engineers
are turning to HALT and HASS to help them achieve high
reliability.
9. HALT and HASS Summary,
continued
In HALT, a product is introduced to progressively higher
stress levels in order to quickly uncover design weaknesses,
thereby increasing the operating margins of the product,
translating to higher reliability.
In HASS, a product is “screened” at stress levels above
specification levels in order to quickly uncover process
weaknesses, thereby reducing the infant mortalities,
translating to higher quality.
11. HALT - Highly Accelerated
Life Test
Quickly discover design issues.
Evaluate & improve design margins.
Release mature product at market introduction.
Reduce development time & cost.
Eliminate design problems before release.
Evaluate cost reductions made to product.
Developmental HALT is not really a test you pass or fail,
it is a process tool for the design engineers.
There are no pre-established limits.
12. HALT, How It Works
ss
re
St
Start low and step up the
stress, testing the product
during the stressing
13. HALT, How It Works
Fa
ilu
ss re
re
St
Gradually increase
stress level until a
failure occurs
14. HALT, How It Works
Fa
ilu
ss re
re
St
s is
aly
An
Analyze
the failure
15. HALT, How It Works
Fa
ilu
ss re
re
St
s is
Im
aly
pr
ov
An
Make
temporary e
improvements
16. HALT, How It Works
Increase
stress and Fa
start
ilu
re s s
process
e)
re re
as
over
St
( inc
s is
Im
aly
pr
ov
An
e
17. HALT, How It Works
Fa
ilu
re s s
e)
re re
as
St
( inc
s is
Im
aly
pr
ov
An
e
18. HALT, How It Works
Fa
ilu
re s s
e)
re re
as
St
( inc
s is
Im
aly
pr
ov
An
e
19. HALT, How It Works
Fa
ilu
re s s
e)
re re
as
St
( inc
s is
Im
aly
pr
ov
An
e
20. HALT, How It Works
Fa
ilu
re s s
e)
re re
as
St
( inc
s is
Im
aly
pr
ov
An
e
21. HALT, How It Works
Fa
ilu
re s s
e)
re re
as
St
( inc
s is
Im
aly
pr
ov
An
e
22. HALT, How It Works
Fa
ilu
re s s
e)
re re
as
St
( inc
s is
Im
aly
pr
ov
An
e
23. HALT, How It Works
Fa
ilu
re s s
e)
re re
as
St
( inc
s is
Im
aly
pr
ov
An
e
24. HALT, How It Works
Fa
ilu
re s s
e)
re re
as
St
inc
Fundamental
(
Technological
s is
Im Limit
aly
pr
ov
An
e
25. HALT, Why It Works
Classic S-N Diagram
(stress vs. number of cycles)
S0= Normal Stress conditions
S2
N0= Projected Normal Life
S1
S0
N2 N1 N0
26. HALT, Why It Works
Classic S-N Diagram
(stress vs. number of cycles)
Point at which failures become non-relevant
S0= Normal Stress conditions
S2
N0= Projected Normal Life
S1
S0
N2 N1 N0
36. Developmental HALT Process
STEP 1: Planning a HALT
Meet with design engineers to discuss product.
– Determine stresses to apply.
– Determine number of samples available.
– Determine functional tests to run during Dev. HALT.
It is essential that the product being tested be fully exercised
and monitored throughout HALT for problem detection.
– Determine what parameters to monitor.
– Determine what constitutes a failure.
Develop Test Plan
37. Developmental HALT Process
For each stress, we use the
Step Stress Approach
Stimuli
Continue until operating & destruct limits
of UUT are found or until test equipment
limits are reached.
D
C
B
A
0:00 0:10 0:20 0:30
Time (hour:minute)
38. Developmental HALT Process
STIMULI VIBRATION HIGH TEMP LOW TEMP
START 3-5 G’s +20° C +20° C
INCREMENT 3-5 G’s 5 to 10° C 5 to 10° C
DWELL TIME 10 min* 10 min* 10 min*
END Destruct Limit or Test Equipment Limitation
* In addition to functional test time
OTHER STIMULI:
• Voltage/frequency margining
• Power cycling
• Combined environment (Temp/Vib)
• Rapid transitions up to 60oC/min on the product
39. Developmental HALT Process
STEP 2: Setting up for HALT
Setup
– Design vibration fixture to ensure energy transmission to the
product (different from electrodynamic vibration fixtures).
– Design air ducting to ensure maximum thermal transitions on the
product.
– Tune chamber for product to be tested.
– Apply thermocouples to product to be tested.
– Setup all functional test equipment and cabling.
40. Developmental HALT Process
STEP 3: Executing a HALT
- Perform Step Stress on each individual stress
- Then Combine Stresses together
41. Developmental HALT Process
STEP 4: Post Testing
– Determine root cause of all failures that occurred.
– Meet with design engineers to discuss results of Developmental
HALT and root cause analysis.
– Determine and implement corrective action.
– Perform Verification HALT to ensure problems fixed and new
problems not introduced.
– Periodically evaluate product as it is subjected to engineering
changes.
42. When to Perform HALT?
Feasibility Development Qualification Launch
P1- P2 → Late P2 → P3 →
Perform HALT Perform HALT ♦Demonstrate ♦Tracking
on 1 to 2 early on more 100% reliability reliability
prototypes. samples. target @ 80% through field
These samples These samples C.L. data
may be hand- will be closer to ♦Shipping /
made and test final product Packaging test
coverage may and functional ♦Validation
be low, but we tests will be HALT can be
can still get more refined performed here
clues as to with higher test
gross design coverage.
issues.
Lessons learned feedback to next
generation product
43. When to Perform HALT?
Some would say that you must do in P1 in order to affect the
design.
Others would say that you cannot perform HALT until P2
when the design is more mature and functional tests are more
complete.
Still others would argue that we cannot perform HALT until
P3 during device qual.
Which is correct?
44. When to Perform HALT?
Both and Neither, really!
Best solution is really to do both to get the benefits of early
testing (P1) but also to test later in the design (P2) when test
coverage is higher and samples are cleaner and then use
Verification HALT to validate the design (P3).
And if we are working with a modular design, we should be
testing the subassemblies as they become ready (P1) rather
than waiting for system test at the end – may be too late.
45. HALT Equipment Commonly
Used
Combined
Temperature/Vibration
Equipment
Pneumatic Vibration (to
provide the random
vibration) with Wide
Frequency Spectrum
Fast Thermal Rates of
Change and Wide
Thermal Range
46. HALT
Cost Benefits
Reduced product time to market
Lowered warranty cost through higher MTBF
Faster DVT with fewer product samples
Accelerated screening (HASS) allowed
47. Words of Wisdom From IVAC...
In order to see the failure modes that
must be eliminated, we can,
Test 100 units for one
year under normal
conditions
OR
Perform HALT on six
units for one week! From IVAC Tympanic Thermometer Model 2090 pamphlet
48. Case Study: HALT on a
Medical Infusion Pump
An infusion pump company was developing an n+1 design (they
had a product in the marketplace and this project was to upgrade
the present design, correct issues that had been discovered, and
add new features). Early in the program we determined what
was different from the previous design.
49. Case Study: HALT on a
Medical Infusion Pump
We determined that the main differences between the two product
iterations were the following:
1. A new motor technology. The previous design had one motor –
for infusing; the new design was to have three – one for the
infusing, one for the cassette insertion and removal, and one for
the air sensor to pinch the tubing in order to detect air bubbles
in the line
2. A new power supply
3. A touch screen
4. An 802.11 wireless feature
5. A new battery design to allow for longer total life
6. New software to handle all of these new features
50. Case Study: HALT on a
Medical Infusion Pump
We then performed an FMEA on each of these subsystems and
determined which testing method was most appropriate for
proving the new subsystem.
For items 1, 3, and 5, we felt that ALT was the best approach
and for items 2 and 4, we felt that HALT was a better test
vehicle. For item 6, we decided that the best approach was to
perform a Software FMEA followed by a rigorous review of the
requirements followed by development of test cases to exercise
the requirements.
51. Thank You for Listening !
Mike’s book on reliability titled
“50 Ways to Improve Your
Product Reliability”
is due out in the spring.
CONTACT INFO:
Mike Silverman, C.R.E.
Managing Partner, Ops A La Carte
mikes@opsalacarte.com
(408) 472-3889