ELECTRONIC COMPONENT
COOLING ALTERNATIVES:
COMPRESSED AIR AND LIQUID NITROGEN
by .
Stephen C. Schmitt and Robert F. Olfenbuttel
Battelle
Columbus, Ohio 43201
Contract No. 68-CO-0003
Work Assignment No. 2-36
Project Officer
Johnny Springer, Jr.
Waste Minimization, Destruction, and Disposal Research Division
Risk Reduction Engineering Laboratory
Cincinnati, Ohio 45268
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
)Printedon Recycled Paper
-------
NOTICE
This material has been funded wholly or in part by the U.S. Environmental Protection
Agency (EPA) under Contract No. 68-CO-0003 to Battelle. It has been subjected to the Agency's
pear and administrative review and approved for publication as an EPA document. Approval does
not signify that the contents necessarily reflect the views and policies of the U.S. Environmental
Protection Agency or Battelle; nor does mention of trade names or commercial products constitute
endorsement or recommendation for use. This document is intended as advisory guidance only to
solvent-using industries in developing approaches to waste reduction. Compliance with environ-
mental and occupational safety and health laws is the responsibility of each individual business and
is not the focus of this document.
-------
FOREWORD
Today's rapidly developing and changing technologies and industrial products and
practices frequently carry with them the increased generation of materials that, if improperly dealt
with, can threaten both public health and the environment. The U.S. Environmental Protection
Agency (EPA) is charged by Congress with protecting the Nation's land, air, and water resources.
Under a mandate of national environmental laws, the agency strives to formulate and implement
actions leading to a compatible balance between human activities and the ability of natural systems
to support and nurture life. These laws direct the EPA to perform research to define our environ-
mental problems, measure the impacts, and search for solutions.
The Risk Reduction Engineering Laboratory is responsible for planning, implementing,
and managing research, development, and demonstration programs to provide an authoritative,
defensible engineering basis in support of the policies, programs, and regulations of the EPA with
respect to drinking water, wastewater, pesticides, toxic substances, solid and hazardous wastes,
Superfund-related activities, and pollution prevention. This publication is one of the products of
that research and provides a vital communication link between the researcher and the user
community. ,
Passage of the Pollution Prevention Act of 1990 marked a significant change in U.S.
policies concerning the generation of hazardous and nonhazardous wastes. This bill implements the
national objective of pollution prevention by establishing a source reduction program at the EPA and
by assisting States in providing information and technical assistance regarding source reduction. In
support of the emphasis on pollution prevention, the "Waste Reduction Innovative Technology
Evaluation (WRITE) Program" has been designed to identify, evaluate, and/or demonstrate new
ideas and technologies that lead to waste reduction. The WRITE Program emphasizes source
reduction and on-site recycling. These methods reduce or eliminate transportation, handling,
treatment, and disposal of hazardous materials in the environment. The technology evaluation
project discussed in this report emphasizes the study and development of methods to reduce waste
and prevent pollution.
E. Timothy Oppelt, Director
Risk Reduction Engineering Laboratory
in
-------
ABSTRACT
The goal of this study was to evaluate topis used to troubleshoot circuit boards with
known or suspected thermally intermittent components. Failure modes for thermally intermittent
components are typically mechanical defects, such as cracks in solder paths or joints, or broken
bonds, such as interconnections inside integrated circuit packages or capacitors. Spray cans of
refrigerants (R-12 [CFC-12] and R-22 [HCFC-22]}, which are commonly used in electronics
manufacturing and repair businesses for this purpose, served as the benchmark for the evaluation.
A promising alternative technology that was evaluated in this study is a compressed-
air tool that provides a continuous stream of cold air that can be directed toward specific
components. Another alternative technology that was considered is a Dewar flask that dispenses
cold nitrogen gas as the cooling agent. Critical parameters were measured for each cooling method
to provide a basis for comparison of compressed air and liquid nitrogen with spray cans of
refrigerant. These parameters are accuracy, electrostatic discharge risk, cooling capability,
technician safety, pollution prevention potential, and economic viability.
This study was performed in accordance with the Quality Assurance Project Plan for
Cold Compressed Air for Electronic Component Cooling Study, dated August 1991. Although the
plan was written specifically for the evaluation of compressed air, the test plan was written to
include an evaluation of liquid nitrogen because test site staff were interested in evaluating this
technology. The liquid nitrogen evaluation showed that it could be a viable alternative. Therefore,
with the concurrence of the Project Officer, this final report includes the results of both com-
pressed air and liquid nitrogen.
Newark Air Force Base, in Ohio, was the site for evaluating compressed-air technolo-
gy. Electronic circuit boards from a variety of Air Force Systems are tested and repaired on a daily
basis. A percentage of these circuit boards demonstrate thermally intermittent failure modes and
were used for comparison testing.
This report was submitted in partial fulfillment of Contract Number 68-CO-0003, Work
Assignment 2-36, under the sponsorship of the U.S. Environmental Protection Agency. This report
covers the period from June 1991 to February 1993 and work was completed as of September
1993. :
iv
-------
CONTENTS
Page
NOTICE ...............'.".... ii
FOREWORD ; . Hi
ABSTRACT '....• „ jv
FIGURES . vjjj
TABLES . . . ix
ACKNOWLEDGMENTS . . . . -.....- x
SECTION 1
PROJECT DESCRIPTION . |
INTRODUCTION 1
PROJECT OBJECTIVES . . . 1
DESCRIPTION OF THE TECHNOLOGY . 2
DESCRIPTION OF THE SITE .5
SUMMARY OF APPROACH 6
Accuracy 6
Electrostatic Discharge Risk 9
Cooling Rate and Absolute Temperature Drop 10
Technician Safety , 12
Pollution Prevention Potential . . . 12
Estimation of Economics 12
SECTION 2
ACCURACY EVALUATION .14
RESULTS . . ... . . . 14
Test Article 1: 1.2 KHz Inverter 15
Test Article 2: 1.2 KHz Inverter 15
Test Article 3: FMC Tank Processing Unit 18
Test Article 4: IFMP Primary Microprocessor 18
Test Article 6: Pitch Gimbal Buffer 21
Test Article 7: FMC Primary Microprocessor 21
Test Article 9: Carousel Instruction Processing Unit , 21
Test Article 10: FMC Tank Processing Unit , . . : . 25
Test Article 13: FSAC Central Processing Unit 25
Test Article 14: FMC Tank Processing Unit 25
Test Article 15: FMC Tank Processing Unit 28
Test Article 16: FMC Tank Processing Unit 28
Test Article 17: FMC Tank Processing Unit 28
* INTERPRETATION 33
-------
CONTENTS (contihued)
Page
SECTION 3
ELECTROSTATIC DISCHARGE RISK EVALUATION 34
RESULTS „ 34
Circuit Board Tests . . .; „ 34
Nozzle Tests „ 35
INTERPRETATION .......'.'.'.'.'.'.'.'.'.'.'.'.'.'.'. 35
SECTION 4
COOLING RATE AND ABSOLUTE TEMPERATURE DROP EVALUATION 36
RESULTS 36
Absolute Temperature Drop 35
Cooling Rate 41
INTERPRETATION '.'.'.'.'.'.'.'. 41
General Component Cooling Characteristics . 41
Sensitivity to Application Parameters 55
SECTION 5
TECHNICIAN SAFETY EVALUATION 57
RESULTS . . • .'.""" 57
INTERPRETATION 57
SECTION 6
POLLUTION PREVENTION POTENTIAL EVALUATION 58
RESULTS 58
INTERPRETATION ......'.'.'.'.'.I'.'.'.'.'.'.'. 58
SECTION 7
ESTIMATION OF ECONOMICS 60
RESULTS ...........! 60
Cooling Material Costs 60
Investment Costs . 62
INTERPRETATION . -. '..'.'.'.'.'.'.'.'.'.'.'.'.'.'. 62
SECTION 8
QUALITY ASSURANCE i 63
LIQUID NITROGEN EVALUATION 63
ACCURACY EVALUATION 63
R-12 Substitution ! . . 63
Completeness 66
ELECTROSTATIC DISCHARGE RISK EVALUATION 66
R-12 Substitution 66
Test Location and Nozzle Test Meter Change 66
Nozzle Electrostatic Charge Buildup: Completeness . . . ' 67
Nozzle Electrostatic Charge Buildup: Precision 67
Nozzle Electrostatic Charge Buildup: Accuracy 68
Circuit Board Electrostatic Charge Buildup: Steel Aerosol Nozzle Evaluaton 68
Circuit Board Electrostatic Charge Buildup: Completeness 69
Circuit Board Electrostatic Charge Buildup: Precision 69
Circuit Board Electrostatic Charge Buildup: Accuracy 69
VI
-------
CONTENTS (continued)
Page
COOLING RATE AND ABSOLUTE TEMPERATURE DROP EVALUATION . 71
Unit of Measure Change 71
R-12 Substitution . . . 71
Data Acquisition Methodology Description 71
Cooling Rate: Completeness 72
Cooling Rate: Precision . 72
Cooling Rate: Accuracy 74
Absolute Temperature Drop: Completeness 77
Absolute Temperature Drop: Precision 77
Absolute Temperature Drop: Accuracy 78
Compressed-Air Pressure: Completeness 79
Compressed-Air Pressure: Accuracy 80
Compressed-Air Temperature: Measurement Method Change 80
Compressed-Air Temperature: Completeness and Accuracy . 80
Ambient Air Temperature: Completeness 80
Ambient Air Temperature: Accuracy 81
TECHNICIAN SAFETY EVALUATION . . 81
Sound-Level Measurement Procedure Change 81
Sound Level: Accuracy 81
Sound Level: Precision and Completeness . . 82
POLLUTION PREVENTION POTENTIAL . . . 82
R-12 Substitution 82
CFC Released: Completeness . 82
CFC Released: Accuracy 82
ESTIMATION OF ECONOMICS : 83
R-12 Substitution 83
Compressed-Air Release Time: Completeness 83
Compressed-Air Release Time: Accuracy 83
Compressed-Air Pressure: Completeness 83
Compressed-Air Pressure: Accuracy 83
SECTION 9
DISCUSSION ^ . . 85
SECTION 10
DATA REDUCTION 87
•' ACCURACY EVALUATION 87
ELECTROSTATIC DISCHARGE RISK . 87
COOLING RATE AND ABSOLUTE TEMPERATURE DROP . . 87
SAFETY . 89
POLLUTION PREVENTION POTENTIAL 89
ESTIMATION OF ECONOMICS . 89
APPENDIX A
COMPONE-NT TEMPERATURE CONTROL: LIQUID NITROGEN .90
APPENDIX B
MEASUREMENT PRECISION OBJECTIVES 92
VII
-------
CONTENTS (continued)
FIGURES
Figure 1 Compressed-Air Tool Operating Principle 4
Figure 2 Typical Compressed-Air Tool Dimensions „ „ „ 4
Figure 3 Typical %-L Liquid Nitrogen Dispenser 5
Figure 4 Test Site Compressed-Air Filter System 7
Figure 5 Electrostatic Charge Measurement Method 11
Figure 6 Test Board Design 11
Figure 7 Test Article #1 ...!..; 16
Figure 8 Test Article #2 '.'.'.'. 17
Figure 9 Test Article #3 19
Figure 10 Test Article #4 20
Figure 11 Test Article #6 , 22
Figure 12 Test Article #7 23
Figure 13 Test Article #9 24
Figure 14 Test Article #10 26
Figure 15 Test Article #13 27
Figure 16 Test Article #14 ............................ 29
Figure 17 Test Article #15 30
Figure 18 Test Article #16 31
Figure 19 Test Article #17 . 32
Figure 20 Cooling Material Application Parameters . . 37
Figure 21 Cooling Rate Comparison for Integrated Circuits: Distance %", Direction A to C . . 42
Figure 22 Cooling Rate Comparison for Integrated Circuits: Distance 1", Direction A to C ... 42
Figure 23 Cooling Rate Comparison for Integrated Circuits: Distance %", Direction D to B . . 43
Figure 24 Cooling Rate Comparison for Capacitors: Distance %", Direction A to C 43
Figure 25 Cooling Rate Comparison for Capacitors: Distance 1", Direction A to C ........ 44
Figure 26 Cooling Rate Comparison for Capacitors: Distance %", Direction D to B 44
Figure 27 Integrated Circuit (H-3-1} Time/Temperature Plot 46
Figure 28 Integrated Circuit (H-6-1) Time/Temperature Plot 46
Figure 29 Integrated Circuit (H-9-1) Time/Temperature Plot ........................ 47
Figure 30 Integrated Circuit (N-3-1) Time/Temperature Plot ....................... 47
Figure 31 Integrated Circuit (N-6-1) Time/Temperature Plot ........................ 48
Figure 32 Integrated Circuit (N-9-1) Time/Temperature Plot 48
Figure 33 Integrated Circuit (A-3-1) Time/Temperature Plot 49
Figure 34 Integrated Circuit (A-6-1) Time/Temperature Plot 49
Figure 35 Integrated Circuit (A-9-1) Time/Temperature Plot 50
Figure 36 Capacitors (H-3-1) Time/Temperature Plot 50
Figure 37 Capacitors (H-6-1) Time/Temperature Plot . 51
Figure 38 Capacitors (H-9-1) Time/Temperature Plot . . . . 51
Figure 39 Capacitors (N-3-1) Time/Temperature Plot . 52
Figure 40 Capacitors (N-6-1) Time/Temperature Plot . ; 52
Figure 41 Capacitors (N-9-1) Time/Temperature Plot 53
Figure 42 Capacitors (A-3-1) Time/Temperature Plot . : ........................... 53
Figure 43 Capacitors (A-6-1) Time/Temperature Plot . ; 54
Figure 44 Capacitors (A-9-1) Time/Temperature Plot 54
viii
-------
CONTENTS (continued)
. ,
TABLES
Table .1 Accuracy Evaluation Summary for Cooling Methods ....................... 14
Table 2 Electrostatic Charge Measurements: Circuit Board Tests .34
Table 3 Electrostatic Charge Measurements: Nozzle Tests 35
Table 4 Minimum Temperature Achieved: %" Distance, Direction A to C ..... .--. ....... 38
Table 5 Minimum Temperature Achieved: 1" Distance, Direction A to C . 39
Table 6 Minimum Temperature Achieved: %" Distance, Direction D to B .."...... 40
Table 7 Component Cooling Rate Comparison 45
Table 8 Target/Adjacent Component Temperature Difference 55
Table 9 R-12 Refrigerant Usage 58
Table 10 Cooling Material Usage and Cost 61
Table 11 Investment Cost and Payback . . 62
Table 12 Revised Quantitative QA Objectives 64
Table 13 Performance Against Revised Quantitative QA Objectives . . 65
Table 14 Electrostatic Charge Buildup — Measurement Precision for Nozzle Tests 68
Table 15 Electrostatic Charge Buildup — Measurement Precision for Circuit Board Tests ..... 70
Table 16 Rate of Cooling — Measurement Precision .' . 73
Table 17 Rate of Cooling — Measurement Accuracy 75
Table 18 Absolute Temperature Drop — Measurement Precision 78
Table 19 Absolute Temperature Drop — Measurement Accuracy .79
IX
-------
ACKNOWLEDGMENTS
The assistance of the following is acknowledged: Don Hunt, Captain Vern Milholen,
Tim Winkler, Robert Hanlin, Jeffrey Heim, and Mike Thomas at Newark Air Force Base, Ohio; Don
Gray of Vortec Corporation, Cincinnati, Ohio; and Michael Bryne of Brymill Corporation, Vernon,
Connecticut. The New Jersey Department of Environmental Protection also provided support for
this study.
-------
SECTION 1
PROJECT DESCRIPTION
INTRODUCTION
The objective of the U.S. Environmental Protection Agency (EPA) Waste Reduction
Innovative Technology Evaluation (WRITE) Program is to evaluate, in a typical workplace environ-
ment, examples of prototype technologies that have potential for reducing wastes at the source or
for preventing pollution. In general, for each technology to be evaluated, three issues should be
addressed.
First, it must be determined whether the technology is effective. Because pollution
prevention or waste reduction technologies usually involve recycling or reusing materials or using
substitute materials or techniques, it is important to verify that the quality of the materials and the
quality of the work product are satisfactory for the intended purpose.
Second, it must be demonstrated that using the technology has a measurable positive
effect on reducing waste or preventing pollution.
Third, the economics of the new technology must be quantified and compared with the
economics of the existing technology. It should be clear, however, that improved economics is not
an absolute criterion for the use of the prototype technology. There may be justifications other than
saving money that would encourage adoption of new operating approaches. Nonetheless,
information about the economic implications of any such potential change is useful for understand-
ing the overall jmpact of implementation.
PROJECT OBJECTIVES
The goal of this study was to evaluate cold compressed-air tools and liquid nitrogen as
methods for cooling electronic components while searching for the causes of thermally intermittent
electronic circuit failure. Aerosol cans of refrigerant (i.e., R-12 and R-22), which have been used
commonly in electronics manufacturing and repair businesses for this purpose, served as the bench-
mark for the evaluation. The questions to be answered by this study were:
-------
1. Would the technicians' ability to find causes of failure.be degraded by use of the
alternatives?
2. How did the cooling characteristics of the alternatives compare to aerosols with
those of aerosols used by technicians?
3. Would the risk of electrostatic damage to electronic components be increased by
use of the alternatives? ;
4. Would the noise generated during compressed-air tool operation be an occupa-
tional safety hazard?
5. How much refrigerant release would be avoided by using the alternatives?
6. What were the economics of implementing either alternative?
The first two issues are related but required different approaches. The cooling
characteristics of the alternatives were known to differ from each other and from refrigerant
aerosols but were not well understood. Also not understood was the effect the characteristics
would have on the troubleshooting process. For example, while it was known that the compressed-
air tool could not cool thermocouples as low as R-12, it was not known whether the temperature
difference would affect the technicians' ability to find causes of thermally intermittent circuit
failures. The cooling characteristics could be compared using fabricated test boards, but a variety
of active circuit boards with various real thermally intermittent failure modes were the best method
to address the first issue. Approaches used to address all six issues are discussed later in this
section.
DESCRIPTION OF THE TECHNOLOGY
Trouble-shooting circuit boards with known or suspected thermally intermittent compo-
nents is a common operation in the electronics manufacturing and repair industries.. If, for example,
an electronic device works when first turned on but fails as it warms up in operation, a technician
may spray refrigerant towards board areas or on specific components to reduce temperatures until
the device begins to work again. Failure modes for thermally intermittent components are typically
mechanical defects, such as cracks in solder paths or joints, or broken bonds, such as interconnec-
tions inside integrated circuit packages or capacitors. Thermally intermittent failures can occur
when temperature changes and material expansion or contraction aggravate the mechanical failure
to create an electrical discontinuity condition. The component that, when cooled, causes the failure
mode to appear or disappear is replaced.
Finding the causes of thermally intermittent circuit failures is often a difficult task. It is
not uncommon to test, replace a component, and retest a circuit several times before eliminating
the failure mode. In some cases, the cause of failure cannot be determined and the circuit board is
-------
condemned. Even with trouble-shooting tools such as freeze compound, it is a trial-and-error
operation.
• m • ~i •- -• • . .:
As trouble-shooting tools, aerosol cans of refrigerant {R-12 and R-22) a.re very
common. They can be used easily to cool an entire circuit board or a single solder connection, are
portable, and are relatively inexpensive. However, as recognized in the Montreal Protocol of 1987,
chlorine released by decomposing chlorofiuorocarbons (CFCs), such as R-12, decreases stratospher-
ic ozone. The protocol calls for the elimination of CFC manufacture in the future. As a result, many
businesses are seeking technologies that will replace current uses of CFCs. Hydrochlorofluorocar-
bons (HCFCs) such as R-22 also will be phased out, although they have lower stratospheric ozone
depleting potential.
The first alternative technology evaluated was a compressed-air tool that provides a
continuous stream of cold air that can be directed towards components. A schematic of how the
tool operates is shown in Figure 1; a drawing of a typical compressed-air tool is shown in Figure 2.
Compressed air enters a tangentially drilled stationary generator, which forces the air to spin down
the long tube's inner walls toward the hot-air control valve. A percentage of the air, now at
atmospheric pressure, exits through the needle valve at the hot-air exhaust. The remaining air is
forced back through the center of the sonic-velocity airstream where, still spinning, it moves at a
slower speed, causing a simple heat exchange to take place. The inner, slower moving air gives up
heat to the outer, faster-moving air column. When the slower inner air column exits through the
center of the stationary generator and out the cold exhaust, it has reached an extremely low
temperature. To obtain temperatures in the range of -35°C to -40°C, the tool requires clean,
dry, room temperature air flowing at 15 scfm at 100-psi pressure.
The second alternative technology evaluated uses liquid nitrogen. A half-liter Dewar
flask (illustrated in Figure 3) can be used with a release valve that allows a stream of nitrogen gas
and liquid-droplets to be directed through a small-diameter stainless-steel nozzle. As the valve and
nozzle are cooled by the nitrogen flow, the portion of the stream that is droplets increases and the
output stream drops in temperature. A variety of valves, nozzles, and heat exchangers are available
to tailor the delivery and cooling characteristics of .the stream of nitrogen. The Dewar flask can be
refilled from a bulk container of liquid nitrogen.
-------
Compressed
Air In (70°F)
Control
Valve
Cold Air
Out (-46°F)
(-46°C)
Vortex-Generation Chamber
Hot Air
Out (212°F)
(100°C)
Source: Vortec Catalog
Rgure 1. Compressed-air tool operating principle.
9-9/16'-
{243mm)
l/8i-27NPT(Femol€)lnlef
1-9/16'
(40mm)
Diomter
1 /8' (3mm) Cold Air Dischorg«
Source: Vortec catalog
Rgure 2. Typical compressed-air tool dimensions.
-------
Figure 3. Typical %-L liquid nitrogen dispenser.
DESCRIPTKDN OF THE SITE
Newark Air Force Base (NAFB), in Ohio, was the site at which compressed-air and
liquid-nitrogen alternative technologies were evaluated. During the study, it was announced that
Newark AFB would be closed; the exact fate of the work performed there was unclear. Electronic
circuit boards from a variety of Air Force systems are tested and repaired at NAFB daily. Examples
are inertial guidance systems used in KC-135,.C-5, and C-141 aircraft and a fuel saver advisory
system used in the KO135. A percentage of the circuit boards tested demonstrate thermally
intermittent failure modes; during the test period, these boards became test articles for comparison
testing. R-12 was used for this study as the benchmark.
Each repair shop at NAFB is responsible for specific systems, such as the KC-135 fuel-
control system. Because compressed air typically is not available at the test stations where cooling
materials are needed, it was necessary to select one shop for the study. After evaluating several
shops, the Carousel Shop was selected as the test site for the following reasons:
• Test stations included fixtures capable of reducing circuit board temperature (using
carbon dioxide) while the board is tested. This feature provided confirmation that
thermally intermittent failure mode existed but did not provide a trouble-shooting
capability because the entire board was cooled at one time.
-------
• The systems repaired in the Carousel shop contained circuit boards in a variety of
sizes, component densities, and component varieties.
• Installation costs to deliver compressed air could be minimized because the three
test stations utilized for the study are in close proximity.
The compressed-air system utilized for the study consisted of a large industrial
compressor with a refrigeration system to chill the compressed air as it passed into a storage tank.
The air passed through approximately 50 ft of half-inch line with nonrestrictive couplings to three
outlets. A filtration and drying system, as described in Figure 4, was installed approximately 20 ft
from the test stations.
SUMMARY OF APPROACH
Accuracy
An objective of the study was to compare the effect of using alternative cooling
materials on technician ability to find causes of thermally intermittent circuit failures using
refrigerant aerosol as the benchmark. This parameter of the cooling materials was termed
"accuracy" because it is a measure of the accuracy with which technicians could find causes of
circuit failures. Two key elements of accuracy that were unknown were how differences in cooling
material dispensing characteristics would affect technician ability to isolate circuit failure causes and
if the temperatures to which active circuit components could be cooled would be low enough to
cause circuit failures to appear or disappear, depending on the failure mode. Standard measures or
measurement methods of this parameter did not exist, sp they were devised so that, in addition to
fulfilling the study objective, they fit within constraints irnposed by the site selected and the study
schedule.
As described in the project objectives, active circuit boards with thermally intermittent
failure modes rather than a fabricated test circuit board were needed to compare the accuracy of
the cooling methods. Building a test circuit board to simulate a circuit board with a thermally
intermittent failure mode was not considered feasible, primarily because the temperature to which
active circuit components must be cooled to eliminate the failure mode was unknown and was
expected to vary among circuit boards. A comparison of the three cooling methods through testing
of active circuit boards with real thermally intermittent failure modes was expected to provide the
most useful data to readers.
-------
jO
u.
r
1
i « -
»_
a> ro o
i "5 e
t!- • £
o ro *;
• a
(0
IT
£
•a
w
CO
in
O
u
to
E
^«.
QJ
4_»
ro
t. > o
£^~
**•%
to —• c
•— o — .F
n
ons
O
to
o>
o
Q>
C
o
Q
E
o
E
o>
*->
to
£
.*•«
(_
ro
a.
o
_o
Si
i E
- ^™ A.
Q> "•"•
ii in
C o
^y> *~
j— c.
— o>
?§i
. JJ O
fc <= c
ft* «u C
** £ o>
=: c t-
«- O £_
?° ^
i ~ ro
o o >
«o _ —
A> E *•"
•- w o
CO 4> .
° i
>^s^
ll^
o"0 e
C E —
tu
ra
iT
ro
Q.
• O
I-
S§
«2- o
o —
o o
E
-sg
ss°
O — t_
o^ * k
. . CO ~ O
"Ik
^ „ o
0> m <°
2^ 2 "O
z E<
i i- ro
C _ J-» O>
O i_ «- 0>
O O < (X
.•^ in \6
-------
The decision to us'e real circuit boards required a test site that encountered such circuit
boards in reasonable quantity and variety and which could support testing with all three cooling
materials. The Carousel Shop at Newark AFB met these criteria, although it imposed constraints on
the project. The constraints were as follow:.
1. Although circuit boards with thermaljy intermittent failure modes were identified
routinely in the shop, the number that would be identified during the test period
was unknown due to fluctuations in workload. Past experience in the shop
indicated that it was unlikely that the number would exceed that sufficient to
meet the study needs.
2. There were only three technicians in the Carousel Shop with one working each of
three shifts.
3. It would not be feasible to track the test/repair/retest process of every test article
through to conclusion during the study test period. Delivery cycles for replace-
ment components are routinely long enough that for many test articles, the repair
and retest steps would occur after the end of the test period.
The first two constraints imposed by the site selection affected the experiment design.
The limited number of test articles meant that each had to be tested with all three cooling methods.
Each of the three tests had to be performed by a different technician to avoid prior knowledge of
the suspected cause of the circuit failure. Because the technician factor could not be held constant
by having one technician test each article with all three cooling methods, the assignment of cooling
method to technicians for each test article was randomized. Variability of test results caused by
technicians was also minimized because all three had at least eight years experience and all three
had opportunities to become familiar with the alternative cooling methods prior to the test period.
With this experiment design, it was expected that comparisons could be made between cooling
methods even though variability, although minimized as much as possible, in technicians was a
factor.
The selection of a measure of accuracy was affected by the first and third constraints.
The measure of accuracy could not be based on the results of the circuit board repair process
because it was likely that many test articles would not have been repaired and retested by the end
of the test periods. If an abundance of test articles were expected, it would have been feasible to
plan to drop out any which had not been retested at the end of the test period. Because this was
not the case, the measure of accuracy had to be based solely on the results of the initial test step.
The measure of accuracy which was devised to use the results of the initial test step
was a subjective evaluation by the technician of the probability that the cause of the circuit failure
had been identified. During testing, the technician searches a circuit board by cooling progressively
8
-------
smaller areas to find the likely cause of the circuit failure. At some point, the technician stops
searching and decides what repair action to take. This point usually occurs when the technician is
••
-------
• Compressed-air tool with a single-section plastic nozzle
• Liquid nitrogen Dewar flask with a straight stainless-steel nozzle approximately 4-in
long.
The measurements for these experiments were taken following general laboratory practices used for
evaluation of equipment, supplies, and worker apparel related to electronic equipment manufacture
and repair.
The first experiment measured the electrostatic charge generated on the nozzle during
release of cooling material. During a 10- to 12-second material release, the nozzle was held parallel
to and approximately one inch from the platen of a Monroe Electronics, Inc., Model 175 Charged
Platen Monitor*, which measured charge buildup. Two measurements were taken for each cooling
method/nozzle combination.
The second experiment measured electrostatic charge buildup when cooling material
was dispensed towards circuit boards placed on the platen of a Monroe Electronics, Inc. Model 175
Charged Platen Monitor. The dispenser was held so that the nozzle was approximately 0.5 inch
from the edge of the circuit board, both horizontally and vertically, and at approximately 45 degrees
relative to the horizontal surface of the circuit board (see Figure 5). Six circuit boards were
evaluated, with two measurements taken for each cooling method/nozzle combination. The six
circuit boards were selected to provide component and density variety.
Cooling Rate and Absolute Temperature Drop
The characteristics measured for each method were cooling rate and absolute tempera-
ture drop. An experiment was designed to estimate the rate of change of component temperature
by using thermocouples buried inside components at which cooling materials were dispensed. Two
test boards were fabricated, one having integrated circuits and the other having wound-film
capacitors. Each test board contained three components with thermocouples {TC-1, TC-2, TC-3)
and one exposed (TC-REF) thermocouple (see Figure 6).
During tests, all four thermocouples on a test board were connected to a Yokogawa
LR4110 four-channel data logger, which simultaneously recorded temperatures of all four
thermocouples as cooling material was directed at the target component. For each test board,
cooling material was applied from two directions and two distances. Two measurements were
taken for each combination of test board, cooling method, direction, and distance. Before each
measurement for R-12 and compressed air was taken, the cooling material was dispensed directly at
* Mention of trade names and products does not constitute endorsement for use.
10
-------
Circuit
Board
(Resting on
the platen
Platen
.Dlspenser Nozzle
Test Meter
Approximately 45 degrees
Figure 5. Electrostatic charge measurement method.
CHANNEL •:•
CHANNEL *Z*
CHANNEL "4"
CHANNEL ~3m
D
— /
_>
/
/
V
Ul
/
/
/
U3
/
B.2S8"
A *
B.8BB--OJ
' Tr . /$•
* ^ ' S ,<-«•»»•
• ••••••« — *
' ^
TC-REF 8.8BB-
- •' " ' '*' 1
B.33S-
— ~« TC-3 <^
c
\
<3-
B. .„
^
a
-o
Z?
B8- -e
INTEGRATED CIRCUIT >
J FIXTURE -SZS"^- /
>4
y
V
/
/
J<3_ B.SBB-
C
_
V TC-l
/.,
^ TC-S
/ «
[o-
' TC-fiEF
/^ TC-3
B.3SB"
1
-^r
— i
.800-
T
i-^ -V—
S.D9S"
L
* 0.B85"
r.
_
C3 .
^
1 CAPACITOR
FIXTURE
CHANNEL •«
CHANNEL -
Figure 6. Test board design for absolute temperature
drop and cooling rate experiment.
11
-------
the exposed thermocouple to determine the absolute lowest temperature that could be achieved
given the test distance, direction, and cooling method. This was not necessary for liquid nitrogen
because it was known that the thermocouple would reach the lowest measurement limit of
-175°C.
Understanding the characteristics of and differences between cooling methods will
enable technicians to use alternate cooling materials effectively. If, for example, the distance
between the applicator nozzle and the component does not significantly affect the cooling rate of
aerosol cans of R-12 but is a significant factor in the cooling rate provided by compressed air, a
technician should be aware of the difference. ;
Technician Safety
Exposure to sound created by operation of the compressed-air tool was the only safety
concern that required measurement. To assess the potential safety hazard, sound-level measure-
ments were taken by personnel from the Newark AFB Bioenvironmental Engineering group during
operation of the air tool. Other safety concerns associated with the alternative cooling methods
include handling pressurized air and liquid nitrogen, both of which are readily addressed by providing
safety training and using appropriate equipment. Other than potential air tool noise problem, neither
alternative was considered to pose safety risks greater than using refrigerant in aerosol cans.
Pollution Prevention Potential
The purpose of replacing aerosol cans of refrigerant is to reduce the amount of
pollutants released into the atmosphere. As indicated in the discussion of accuracy (see page 6),
the weight of R-12 released during evaluation of each circuit board with thermally intermittent
failure modes was determined. These data provide a measure of the average pollution per circuit
board that could be avoided if either of the alternative cooling methods were adopted.
Compressed air and nitrogen are released to the atmosphere by the alternatives, but
neither are considered pollutants. Pollution generated during the production of either liquid nitrogen
or electricity to power air compressors was beyond the scope of the study.
Estimation of Economics
The approach used to estimate operating 'costs was to measure the volume of each
cooling material used during test article accuracy evaluations and calculate a per-board material
cost. Although material costs are only one aspect of operating costs, it was the only aspect that
12
-------
could be measured during the tests. Material costs for each cooling method were estimated using
the following methodologies: ,
-*; " -fe f
• R-12 cost was estimated by dividing the total weight of R-12 used by the weight of
one can and then by the number of articles tested to obtain the average cans
required per test article. An average cost for a can of R-12 was used to estimate an
R-12 cost per test article. The weight of the empty can was subtracted during
calculations.
• Compressed-air cost was estimated by multiplying the air tool operation time
{release time) by the tool's consumption rate (15 scfm at 100 PSI) to obtain the
volume of air used. Dividing the volume of air by the number of test articles and
multiplying by an average cost to generate compressed air yielded an average cost
of compressed air per test article.
• Liquid nitrogen cost was estimated by dividing the total weight of liquid nitrogen
used by the number of test articles and converting to liters to obtain the average
volume used per test article. Multiplying the volume by an average price of liquid
nitrogen per liter gave the average cost of liquid nitrogen per test article.
The approach to estimating investment cost focused on the cost of dispensers, which
is the only significant investment for the liquid nitrogen alternative. For compressed air, investment
cost is expected to range widely because the condition and capacity of existing compressed-air
supplies at test stations will vary widely. Some sites may not have any existing air supply.
Potential users will need to determine what, if any, investment is required to obtain compressed air
in the quantities and quality required.
13
-------
SECTION 2
ACCURACY EVALUATION
RESULTS
Test article characteristics and Component Identification Confidence (CIC) scores are
summarized in Table 1. Seventeen circuit boards with thermally intermittent failure modes were
identified during the 5-month test period. It was determined later by the Newark AFB test engineer
that 4 circuit boards (Test Articles 5, 8, 11, and 12) should not be included in the evaluation
because they were not thermally intermittent (e.g., loose connector) or because the defective
components were known from previous experience with a specific model circuit board. The latter
type of circuit board would have given the technicians prior knowledge of the cause of failure and
TABLE 1. ACCURACY EVALUATION SUMMARY FOR COOLING METHODS
Test
Article
1
2
3
4
6
7
9
10
13
14
15
16
17
Circuit Board Characteristics
Component
Density
High
High
High
High
Low
High
High
High
High
High
High
High %
High
Component
Variety
High
High
High
High
High
High
Low
High
Low
High
High
High
High
Width
4.50
5.50
6.50
9.00
4.63
6.25
4.50
7.25
5.75
6.25
6.50
6.25
6.50
Length
6.25
6.00
11.50
8.00
5.63
10.50
6.25
10.75
6.59
10.50
10.50
10.50
10.50
Component Identification
R-12
100%
33%
0%
33%
100%
100%
67%
33%
33%
0%
33%
33%
33%
Liquid
Nitrogen
0%
O%
0%
100%
100%
33%
33%
100%
1 00%
1 00%
67%
33%
0%
Compressed
Air
33%
67%
67%
67%
100%
100%
67%
33%
67%
67%
100%
100%
0%
14
-------
would not have been a valid test of the cooling methods. Evaluation results specific to. each test
article are described below; component identification confidence ratings for each evaluation are
provided in parentheses (see the discussion of accuracy on page 6). If repairs were made and retest
data are available, these data are provided. However, in several cases replacement components
were still on order at the end of the test period and it was not possible to make final determination
of accuracy.
Test Article 1: 1.2 KHz Inverter .
The module exhibited sine source output with fluctuating amplitude. The output was
displayed on an oscilloscope during evaluation.
The compressed-air evaluation identified capacitor C51 of the amplitude feedback
control as the suspected component (33%). The R-12 evaluation identified an output power
transistor, Q5, on the opposite side of the circuit board as being the defective component (100%).
The liquid nitrogen evaluation was unable to identify any components (0%). The failure mode was
not corrected by replacement of transistor Q5; therefore, the module was submitted for additional
testing. , .
A photograph of Test Article 1 with suspected defective components indicated is
included as Figure 7.
• /
Test Article 2: 1.2 KHz Inverter
The module exhibited a failure mode similar to that of Test Article 1.
The R-12 evaluation identified a group of 2 resistors and 2 diodes suspected of
containing the defective component(s) (33%). The compressed-air evaluation selected one
component, A4, as the defective component (67%). A4 is an operational amplifier used to control
the phase and frequency of the output and is located adjacent to the group of components identified
during the first evaluation. During the liquid nitrogen evaluation, the technician was unable to cause
the circuit to fail when the entire module was cooled. Therefore, the liquid nitrogen evaluation of
test article #2 could not be performed. Component A4 was selected as the most likely cause of the
failure mode, but replacement was delayed until a new component could be requisitioned.
A photograph of Test Article 2 with suspected defective components indicated is
included as Figure 8.
-------
JS
"o
£
O)
16
-------
17
-------
Test Article 3: FMC Tank Processing Unit
This module is used to gauge the amount of fuel in a tank. The failure mode involved
an erroneous quantity being given when the module was cold.
The compressed-air evaluation identified a capacitor, C116, as defective (67%).
During the liquid nitrogen evaluation, the circuit failed when a large area of the module was cooled;
by probing in this area, the technician identified an operational amplifier, U105, as defective (0)%.
Because cooling alone did not enable the technician to identify a component, the component
identification confidence of 0% was applied. When the R-12 evaluation was performed, a defective
component could not be identified (0%). Because C116 is a decoupling capacitor for U105, it is
likely that the defective component is in this area. The module was placed on engineering hold for
additional evaluation because the R-12 evaluation was unable to identify a failure.
A photograph of Test Article 3 with suspected defective components indicated is
included as Figure 9.
Test Article 4; 1FMP Primary Microprocessor .
The module was causing loss of primary functions of the Integrated Fuel Management
Panel (IFMP) when cold. The module was tested by running the IFMP while the components were
being cooled.
The R-12 evaluation identified a group of components suspected of containing the
defective component (33%). The liquid nitrogen evaluation identified a random access memory
(RAM) device, U18, as the defective component (100%). During the compressed-air evaluation,
the technician was able to make an experience-based selection of diode CR19 (67%). The
subsequent repair process of Article 4 replaced diode CR19 and eliminated the failure mode.
Possible reasons for selection of U18 during the liquid nitrogen evaluation are: CR19 was cooled
enough to cause the circuit failure when cooling material was directed at U18 or the low tempera-
ture of U18 resulting from liquid nitrogen spray may have temporarily changed the access time,
which would cause loss of primary function.
A photograph of Test Article 4 with suspected defective components indicated is
included as Figure 10.
18
-------
-------
-O
20
-------
Test Article 6: Pitch Gimbal Buffer
——^-——————^————^—^— • »• _..,. i , •, ai.. . -
The module was causing the pitch gimbal to drive to the physical stop when the
system was turned on. The "cage" output was saturated.
The liquid nitrogen evaluation identified transformer T2 as the defective component
(100%). The other two cooling methods identified transformer T1 as the defective component
(100%). Replacement of T1 corrected the thermally intermittent failure mode. "
A photograph of Test Article 6 with suspected defective components indicated is
included as Figure 11.
Test Article 7: FMC Primary Microprocessor
The module was causing the FSK communication link between the fuel management
computer and the IFMP to drop off.
The compressed-air evaluation identified U45 as the defective component (100%).
The liquid nitrogen evaluation selected a data buffer, U37, from a group of components (33%).
The R-12 evaluation identified a resistor array, U40, as the defective component (100%). Because
the output of U40 is directly linked to the FSK signal, U40 was selected for replacement. After
replacement, the failure mode remained, and the module was returned for additional testing. All
three components are in an area approximately 2 inches by 2 inches, with U40 and U45 separated
by about 0.5 inch.
A photograph of Test Article 7 with suspected defective components indicated is
included as Figure 12.
Test Article 9: Carousel Instruction Processing Unit
The module is two-sided and uses flat-pack integrated circuits with surface-mount
solder joints. It was removed from a Carousel INU because of a history of computer-related
failures.
AH three cooling methods identified components on the B side of the circuit board as
defective. It was determined that the cause of failure was corrosion between component leads and
circuit traces throughout the board. The circuit board was condemned.
A photograph of Test Article 9 is shown in Figure 13. The failure mode was related to
the circuit board itself rather than to any specific components.
21
-------
22
-------
£•>» .
P-
(0
(0
(3)
iJI
23
-------
.l>l!uill »te* i > i h i »onTiTTn ^»
24
-------
Test Article TO: FMC Tank Processing Unit
~- . ~~* ~ : —~. -V . * ,-f*. • -
The module was causing a particular tank to indicate dashes, which means that the
value being calculated is "unreasonable'1!*"^** _ *y • . '
The compressed-air evaluation identified a group of two operational amplifiers, U5 and U6,
as containing the defective component (33%). The R-12 evaluation selected U6 as the suspected
defective component (33%). The'liquid nitrogen evaluation selected R42, which is a gain resistor for
U6, as the defective component (100%). All components are located in an area approximately 1-inch
by 1.5 inch. Because replacement of R42 did nbt eliminate *he failure mode, U5 and U6 were replaced.
Because the failure mode still remained, the module was delivered to the engineering group for further
evaluation. .
A photograph of Test Article 10 with suspected defective components indicated is
included as Figure 14.
Test Article 13; FSAC Central Processing Unit
The module is the microprocessor for the Fuel Savings Advisory Computer (FSAC). It
failed on the automated module tester during cold soak.
The liquid nitrogen evaluation selected a 4-bit latch, U12, as the defective component
(100%). The R-12 evaluation selected an large-scale integrated circuit (LSI) 4-bit latch, U6, from a
group of components (33%). The compressed-air evaluation selected another 4-bit latch, U18,
from a group of components (67%). U12 and U18 are adjacent to each other, but U6 is located at
the other side of the circuit board. Any of the three components could cause the failure mode.
A photograph of Test Article 13 with suspected defective components indicated is
included as Figure 15.
Test Article 14: FMC Tank Processing Unit
The module caused a particular tank to read 0 when cold.
The compressed-air evaluation selected an operational amplifier, U208, from a group
of components (67%). The R-12 evaluation was unable to identify a defective component (0%).
The liquid nitrogen evaluation identified another operational amplifier, U204, as the defective
component (100%). The two identified components are approximately 0,5 inches apart and either
could cause the failure mode. .
25
-------
-------
CO
OS
3
in
en
27
-------
A photograph of Test Article 14 with suspected defective components indicated is
included as Figure 16.
Test Article 15: FMC Tank Processing Unit
The module caused a tank to read too high when cold.
The R-12 evaluation selected a 4-bit multiplexer, U109, from a group of components
(33%). "The liquid nitrogen evaluation selected an operational amplifier, U103, from a group of
components using technician experience (67%). The compressed-air evaluation identified two
integrated circuits, U104 and U108, as defective (100%). All four components are located in an
area approximately 3 inches by 1.5 inches in size.
A photograph of Test Article 15 with suspected defective components indicated is
included as Figure 17.
Test Article 16; FMC Tank Processing Unit
The module caused a particular tank to drift and then to read dashes.
The R-12 evaluation identified four feedback capacitors, C113, C114, C115, and
C121, as suspected of containing the defective component(s) (33%). The compressed-air
evaluation identified an FET switch, U107, as defective (100%). The liquid nitrogen evaluation
identified a group of two operational amplifiers, U105 and U106, as defective (33%). All seven
components are located in an area approximately 2 inches by 1 inch in size.
A photograph of Test Article 16 with suspected defective components indicated is
included as Figure 18.
Test Article 17; FMC Tank Processing Unit
The module caused a tank reading value to be too high.
The liquid nitrogen and compressed-air evaluations were both unable to identify a
defective component (0%). The R-12 evaluation was able to selected an FET switch, U7, from a
group of components (33%).
A photograph of Test Article 17 with suspected defective components indicated is
included as Figure 19.
28
-------
-------
30
-------
-------
4t
Q>
!s
<
*••
w
£
d
«•••
en
i£
32
-------
INTERPRETATION
To be an effective trouble-shooting tool, a ^ipoling material must be capable of cooling
components to the temperature where a failure occurs (or disappears). However, it must not damage
components with low temperatures, and it must be able to isolate the defective component(s). Although
the number and variety of test articles were less than hoped for, the results of the accuracy evaluation still
provide important information to potential users of the alternatives and make possible the following
interpretations:
• The data obtained during the Absolute Temperature Drop/Cooling Rate experiment (see
Section 4) indicate that compressed air fails to cool components to the levels obtained with
R-12. In 12 of 13 circuit boards tested during the Accuracy Evaluation, the CIC obtained
with compressed air exceeded 0%. Therefore, the cooling capability of compressed air was
sufficient, in all but one case, to reproduce circuit failures.
• A potential problem related to liquid nitrogen temperatures may have been identified
during testing of Article 4. The 100%-confident identification of a RAM chip as defective
when a diode proved to be the defective component may be a case where the low tem-
perature temporarily made the device appear to be the cause of the thermally intermittent
circuit failure. Potential users of liquid nitrogen may want to consider temperature control
strategies to avoid low temperatures that could temporarily change component functions or
even damage components. Several potential strategies are discussed in Appendix A.
• As shown by the CIC data in Table 1, in 8 of the 13 test articles, liquid nitrogen enabled the
technicians to identify the components having thermally intermittent failure mode with an
equal or greater confidence level than that for R-12. Given that the temperatures attainable
with liquid nitrogen are much lower than those with R-12, variability of application
technique is the most likely explanation for the four test articles with a liquid nitrogen CIC
level of 0%. The results seem to point to a need for technician understanding of the
characteristics of alternative cooling methods. These characteristics are discussed in Section
4.
33
-------
SECTION 3
ELECTROSTATIC DISCHARGE RISK EVALUATION
Section 3 contains the results and interpretations of the results of measuring electro-
static charge buildup. A description of the procedures used is provided in Section 1, page 9
{Electrostatic Discharge Risk).
RESULTS
Circuit Board Tests
Table 2 summarizes the electrostatic charge measurements obtained as cooling
materials were dispensed towards the six circuit boards selected. Using averages of each pair of
measurements:
• For all six test articles, the compressed-air alternative generated lower charge
buildup than did R-12 dispensed through a plastic nozzle.
• For four of the six test articles, the liquid nitrogen alternative generated lower
buildup than did R-12. i
TABLE 2. ELECTROSTATIC CHARGE MEASUREMENTS: CIRCUIT BOARD TESTS
Electrostatic
Charge Buildup
Test Board #1
Test Board #2
Test Board #3
Test Board #4
Test Board #5
Test Board #6
Aerosol R-1 2
w/Plastic Nozzle
(volts)
-251
158
-411
-1366
-143
-138
Aerosol R-1 2
w/Steel Nozzle
(volts)
623
443
-666
-900
-139
• -40
Compressed
Air Tool
(volts)
-58
-1
-6
-80
-80
-45
. Liquid
Nitrogen
Dewar (volts)
152
28
133
92
300
174
34
-------
• For three of the six test articles, R-12 generated lower buildup dispensed through
steel nozzles than R-12 dispensed through plastic nozzles.
Nozzle Tests
Table 3 summarizes the electrostatic charge measurements obtained as cooling
materials were dispensed to the atmosphere. Both the compressed air and liquid nitrogen alterna-
tives generated lower electrostatic charge buildup than did R-12 through either plastic or steel
nozzles. R-12 dispensed through steel nozzles generated lower electrostatic charge than when
dispensed through plastic nozzles.
INTERPRETATION
The electrostatic charge buildup data do not support a conclusion that electrostatic
discharge risk is increased by using either of the alternative component cooling technologies.
However, the quality of compressed air should be considered (see the filtering and water separation
system described in Section 1, Description of the Site, page 5) because it is the contaminants in
flowing air (e.g., oil, water, and particulates) that cause electrostatic charge to buildup. If aerosol
cans of R-12 have been utilized successfully, either compressed air or liquid nitrogen should be
acceptable alternatives.
TABLE 3. ELECTROSTATIC CHARGE MEASUREMENTS: NOZZLE TESTS
I Electrostatic
- Charge Buildup
Aerosol R-1 2
w/Plastic Nozzle
(volts)
376
Aerosol R-1 2
w/Steel Nozzle
(volts)
10
Compressed
Air Tool
(volts)
2
Liquid
Nitrogen
Dewar (volts)
3
35
-------
SECTION 4
COOLING RATE AND ABSOLUTE TEMPERATURE DROP EVALUATION
The procedure used to obtain temperature vs time data is described in Section 1,
Cooling Rate and Absolute Temperature Drop (page 10). In Section 4, the data are presented and
interpreted. Figure 20 illustrates the application parameters that were varied during the tests.
RESULTS
In all tests, the cooling material dispensers were positioned and aimed manually. Using
visual feedback from the data logger chart to determine when a stable minimum temperature was
reached, the technician adjusted the angle of elevation slightly to ensure that minimum tempera-
tures were obtained for each application direction and distance. Different angles of elevation result
in underspray or overspray of cooling material, thus changing the cooling rate and the difference in
temperature between the target component and other components on the test fixtures. As a result,
the absolute temperature drop data presented are used for direct comparison of cooling materials;
but cooling rate and temperature difference data, while they indicate performance that may be
obtained in actual use, are not used for direct comparisons.
Absolute Temperature Drop
Tables 4, 5, and 6 summarize the minimum temperatures achieved using different
cooling materials, components, and application directions and distances. Minimum temperatures for
both the target component and the exposed thermocouple are provided to show the effect of
component mass. Using Type K thermocouples with the data logger, temperatures below -175°C
could not be measured. After cooling the exposed thermocouple with liquid nitrogen during the
Initial tests, it was obvious that the minimum measurable temperature would be reached each time.
Therefore, the step was eliminated for subsequent tests.
36
-------
Q
O
u *j
— o
< Q
(ft
I
a)
IB
<5
Q.
.
a.
a
Te
a
u>
"5
o
o
0
CM
1
en
tl
to
en
Q
0*
Of
o
•1
en E
i
*
<
c
i
c
r
j
U
3>
O
H-
d
O
^^ {Jj L*""*lfc
§<
SI
<
< Q.
37
-------
O
o
Ul
tr
ui
o
1
S
o
ui
X
u
<
UI
EC
EC
Ul
n.
UJ
S
z
i
•
Ul
m
c
CD
f $
2 z
T3 CO
3 £
^
^
•DT
CD CO
in '
CO ^
CL S
o
U
CM
T™ ^"
1 1
£E CO
CO 4->
O CO
CD "
X) J£
CD C» |2
O..E o
-S H g
UI §;
^™*
O
D
4-1
CO —
IS
E
CD
H-
*o *^
CD CD "5
CO c c
»F g
UJ «
CD
3
CO — -
0.°
E
1=
"o ^S
CD CD ^
CO c C
D..E O
••" "^ Qj
UJ w
£
4W
CO <^»
ig
E
CD
•
q
Jo
q
in
^
1
0
O)
CM
in
CM
1
O
00
^~
q
in
1
Target
Component
1
q
in
i—
1
1 ;
**J
in :
CO
1
I
in
in
1
CD
Q.
Exposed
Thermocou
CO
S H-
CO 4_J
™1
£ »-
^ o
q
in
q
CO
T—
1
q
CM
in
^3
1
in
f*^
in
CO
in
1
Target
Component
1
q
in
»—
1
1
q
in
CO
1
|
in
O)
in
1
CD
Q.
Exposed
Thermocou
E
t P
~° '5
3 S S
g & S
5 U H
c !
CD
C
o
Q. _
ECD
_ a.
o >.
U H
CO
CD
O
(0
o.
CD
U
•a
o
o
T3
CD
CO
CD
o
o
in
JD
S
•a
CD
E
CO
CO
CO
n.
E
CD
4*1
_JD
Q.
O
U
O
E
CD
+^
E
1
CD
^«
O
38
-------
O
H-
<
&
O
K
D
ul
O
1
CO
Q
Q
UJ
O
ui
tc.
QC
UJ
EL
1
3
E
UJ
03
c
CO
O5'-
2«b
z •*•
<#-*
T3 en
'^ CD
.l*~
-3
IT
0) CD
ID .
OH
CO
0
o
ID
n.
CD
u
•a
o
c
o
•o
09
CO
CO
J3
O
o
in
I
CD
•a
03
CO
CO
CD
CD
a.
a.
i
o
03
'c
39
-------
03
o
2
O
g
in
EC
O
UJ
o
I
CO
Q
O
UI
O
Ul
EC
ec
UJ
o.
UJ
i
CD
III
03
DJ «~
Sen
•*-* i
•D w
f£
(b.
•o *T
ffi O
CO •
CO ^
£ 4-
c" ffi
E \—
0
CNJ
T"» fl™
t I
£E en
o i
CO 4-1
O CO
m *
<
' ill
» ffi ^
" c 5 u>
n. .£ o II
— II
« II
3
2 r?
Jo U
c^ •»•
E
0
H
-o ^
ffi » ^
CO t~ c
a. E o
CD P 0
ii
q
in
^^
^
1
q
oo"
CO
^ — II
£
3
to - m
^1!
E
u
H-
1 gl
eo
l
o
&l 8 | S
.2 1- g r*
~
1-h
ig =.
i
i
Target
Component
1
*
1
0
CO
CO
1
'
1
q
CO*
in '
I
Exposed
Thermocouple
ICQ
•o §>
2 H~
CD i_f
033
ffi O
4^ ^
o
6
en
O
CD
in
1
q
0
CN
1
q
r*v
in
CN
in
1
Target
Component
I
i
*
i
o
in
CO
1
I
1
in
10
'
Exposed
Thermocouple
j=
Hr cs
•o ~
§ If
5 U H
4W
C
ffi
0
E ffi
Q.
O >
0 H
CO
03
U
CD
Q.
CD
O
13
O
O
CO
CO
JO
O
o
in
03
X!
•a
03
3
CO
CO
a
03
a.
eo
a.
o
o
o
40
-------
Cooling Rate
~: .'ffk . i
t°:- - ' . 33 • ;
Figures 21, 22, and 23 compare the cooling rates of the three methods for three
combinations of direction and distance from the target integrated circuits. Figures 24, 25, and 26
present similar information for wound-film capacitors. Table 7 summarizes calculated cooling rates
over approximately the first 50% of the temperature range.
Figures 27 through 44 compare the cooling rates of one exposed thermocouple and
three thermocouples embedded in components on test boards (see Figure 1). A legend provided
with each figure describes the test number and other parameters. The final data point for each
thermocouple represents the stable temperature level reached as cooling material was directed at
the target component. Thermocouples in nontarget components typically reached a stable minimum
temperature before the target component thermocouple reached a stable minimum temperature.
Table 8 summarizes temperature differences between the target and the adjacent
components for each cooling rate test. The temperature difference between these components is
an indicator of the ability of a cooling method to isolate a component with a thermally intermittent
failure mode. The temperature differences were determined from the data logger charts at the point
when the target component reached -10°C. These differences were adjusted to allow for the
difference in starting temperatures; for example, the difference was reduced if the adjacent
component started at a higher temperature.
INTERPRFfATlON
The purpose of the cooling rate and absolute temperature drop tests was to obtain
cooling characteristic information for each material. This information will help potential users who
are experienced with aerosol cans of R-12 use the alternative methods effectively.
General Component Cooling Characteristics
The three cooling materials differed in how they cooled components as described
below:
• As R-12 was sprayed towards components, it built up a "slush" on and around the
component. When the spray of R-12 was stopped, the slush continued to evaporate
and lower the component temperature even further. The fastest initial cooling rates
were obtained with R-12, although the cooling rate decreased as component
temperature dropped. •
41
-------
100
Compressed Air
•ZOO
10
1
40
20 30
Elapsed Time (Seconds)
Figure 21. Cooling rate comparison for integrated circuits: distance V*", direction A to C.
100-
-100-
-200
Liquid Nitrogen
80
20 40 60
Elapsed Time (Seconds)
Figure 22. Cooling rate comparison for integrated circuits: distance 1", direction A to C.
42
-------
100
'•"——« Com pressed Air
-•fl-12
60
80
-200
0 20 40
Elapsed Time (Seconds)
Figure 23. Cooling rate comparison for integrated circuits: distance %", Direction D to B.
100
a Compressed Air
-200
50 100
Elapsed Time (Seconds)
150
Figure 24. Cooling rate comparison for capacitors: distance %", direction A to C-
43
-------
R-12
-150
•20
100
120
40 60 80
Elapsed Time (Seconds)
Figure 25. Cooling rate comparison for capacitors: distance 1", direction A to C.
100-
—r. Compressed Air
R-12
•200
20
40 60
Elapsed Time (Seconds)
Liquid Nltrogon
100
Figure 26. Cooling rate comparison for capacitors: distance %", direction D to B.
44
-------
o
(A
EC
o
a
ULJ
o
o
u
O
a,
S
O
o
GO
o> f
.£ » 03
o co 42
n fT" f>
w it l_>
(J 0
03 03 "5
|E |
E jg
£
lft
.03
Ins
052
=5 S u
uu £
CD
w
o) 5
C to —
** i: U
w o
1 E ~
03
sl
2 g
CO >-
0 0
c
03
C 03
O O-
Q. >-
E l-
o
CJ
05 75
c •—
.— Jr
0 4=
0 to
1.
00 f*^ ^
co r^ co
" CM *-
in in in
CO r- CM
in in q
000
o o o
CM CM CM
i i i
in in o
CM CM CM
"A'
o m
S. o ' S
< < °
-J ^. S.
T- r-
CO CO O)
I I- I
•c
05
co t .
03 '5
03 o
.E 'o
CM
a:
en CD en
CO CO CM
in in o
«— T— »-
m o o
rt ^ d
O 0 0
o o o
CM CM CM
1 1 1
in o o
t «- o
CM CM CM
u o m
S o S
? < f
5 ^
co CD en
XII
E
V o
3 CO
0 0.
> CO
> 0
in in CD
CO ^J- CO
in in in
m o in
CM CM CM
q q q
d co ^
en 03 o>
O O o
o o o
o o o
O CO T-
CM CM CM
0 CO
2 o . S
^ < f
•«4- ^}-
»~ *~
CO CD O)
2 Z 2
S
£ *-
05 '5
03 CJ
IT'S
o
-tf
c ,
•a
§• e
Zi as
oo r^ in
CO i- CO
in o in
*fr co in
CM in CM
q o o
co' d o"
en en o
O O O
O O 0
r* r* r*.
000
CO O O
CM CM CM
O CO
O O
*- o **
^ 4-« -^
^ "^
^ s ^>
r- ,-
i i t
CO CD O)
222
E
"r 5
'u jt^
3 CO
> «°
> O
CO O i-
CD i- T-
000
CO «- »-
CM CM
q in o
en d co
*- CM CM
O O O
0 O O
O in o
en o eo
t- CM CM
2 » 1
J» *J -^ .
^t fl
S =
^(- S ^>
*- *-
CO CD O>
09
CO _
0) 3
03 C3
*- tl
.£ 'o
•o
03
to
CO
03
a,
E
O i-
O '5
en to o
0 0 T-
O in in
CM CO O
CM, CO CM
moo
d ^ ^
CM *- CM
O O O
o o o
mo o
o *- T-
CM CM CM
°, « =
* < f
»- «-
CO CD O3
E
H= ^
T3 ±i
C U
3 CO . .
00- - •
> CO
> 0
45
-------
Distant
Component
-60
10
20 30 i 40
Eiap*ed Time (Seconds)
50
60
Figure 27. 1C (H-3-1) time/temperature plot.
Distant Component
Target Component
-60
Exposed Thermocouple
40 60
Elapsed Time (Seconds)
Figure 28. 1C (H-6-1) time/temperature plot.
46
-------
-60
100
-200
Adjacent Component
•o Target Component
~—« Distant Component
Exposed Thermocouple
20
40 60
Elapsed Time (Seconds)
80
Figure 29. 1C (H-9-1) time/temperature plot.
"Adjacent Component
Target Component
Exposed Thermocouple
20 40 60
Elapsed Time (Seconds)
Figure 30. 1C (N-3-1) time/temperature plot.
—i
80
47
-------
100
-200
100
O
I
a
&
I
, -100-
•200
Distant Componant
Adjacent Component
20
80
40 60
Elapsed Time (Seconds)
Rgure 31. 1C (N-6-1) time/temperature plot.
100
-^Distant Componant
Exposed Thermocouple
Adjacent Componmt
10
40
20 30
Elapsed Time (Seconds)
Rgure 32. 1C (N-9-1) time/temperature plot.
so
48
-------
-30
«
o>
s
a>
I
.10
—» Distant Component
Adjacent Component
10
40
20 30
Elapsed Time (Seconds)
Figure 33. 1C (A-3-1) time/temperature plot.
so
—s Distant Component
Adjacent Component
Elapsed Time (Seconds)
Figure 34. 1C (A-6-1) time/temperature plot.
49
-------
-20
40
20-
S
0-
'"•"-* Distant Component
Adjacent Component
20
80
40 60
Elapsed Time (Seconds)
Figure 35. 1C (A-9-1) time/temperature plot.
100
-20-
-40-
-60
Adjacent Component
Target Component
•Exposed Thermocouple
20
40 60
Elapsed Time (Seconds)
so
100
Figure 36. Capacitors (H-3-1) time/temperature plot.
50
-------
•60
20
40 60 80
Elapsed Time (Seconds)
100
120
Figure 37. Capacitors (H-6-1) time/temperature plot.
v-~« Distant Component
-60
Target Component
—^Adjacent Component '
Exposed Thermocouple
40 60
Elapsed Time (Seconds)
Figure 38. Capacitors (H-9-1) time/temperature plot.
51
-------
100
«
o
-100-
-200-
Adjacent Component
ExpoMd Thermocouple
100i
I
-100-
-200
v« Distant Component
Target Component
20 40 60
Elapsed Time (Seconds)
Figure 39. Capacitors (N-3-1) time/temperature plot.
80
-—& Distant Component
Adjacent Component
Target Component
20
40 60 80
Elapsed Time (Seconds)
100 120
Figure 40. Capacitors (N-6-1) time/temperature plot.
52
-------
100
o
-------
-30
Targat Component
*"""^~« Distant Component
Adjacent Component
Exposed Thermocouple
20 40 60
Elapsed Time (Seconds)
Figure 43. Capacitors (A-6-1) time/temperature plot.
80
a
I
-10-
20
40 60
Elapsed Time (Seconds)
—a Distant Component
Adjacent Component
80
100
Figure 44. Capacitors (A-9-1) time/temperature plot.
54
-------
TABLE 8. TARGET/ADJACENT COMPONENT TEMPERATURE DIFFERENCE
Component
Cooling
Material
R-12
Compressed
Air
•
Liquid
Nitrogen
Component
Type
Integrated
Circuits
Wound-Film
Capacitors
Integrated
Circuits
Wound-Film
Capacitors
Integrated
Circuits
Wound-Film
Capacitors
Test
H-3-1
H-6-1
H-9-1
H-3-1
H-6-1
H-9-1
A-3-1
A-6-1
A-9-1
A-3-1
A-6-1
A-9-1
N-3-1
N-6-1
N-9-1
N-3-1
N-6-1
N-9-1
Application
Direction
AtoC
AtoC
D to B
A to C
AtoC
Dto B
A to C
A to C
D to B
A to C
AtoC
D to B
AtoC
AtoC
D to B
A to C
AtoC
D to B
Application
Distance
%"
1"
K"
%"
IV
%"
V*"
1"
%"
%"
1"
% "
1/«"
1"
%"
%"
1"
%'
Component
Temperature
Difference
/OQj(a)
-11.0
31.5
31 .0
13.5
8.5
8.5
25.5
(b)
24.0
12.0
(b)
• 11.5
30.0
30.0
28.0
25.0
15.0
22.0
(al Negative difference indicates that the adjacent component was colder than the target
component when the target component reached -10°C; positive difference indicates
warmer adjacent component.
(b)
Target Component did not reach -10°C during test.
55
-------
• Liquid nitrogen provided the coldest temperatures of the three cooling materials. In
contrast to R-12, an accelerating cooling rate was obtained when liquid nitrogen was
used (see Figure 28). The cooling material consists of nitrogen gas and droplets of
liquid nitrogen; as the dispensing valve and nozzle cools, the proportion of droplets
increases. The increase in droplets could be heard as increased "sputtering" of
cooling material during material release. Frost buildup on the components during
cooling was minimal.
• Compressed air provided the least cold temperatures and the slowest cooling rate.
As with R-12, the cooling rate decreased as component temperature dropped.
Compressed-air cooling resulted in a slight frost buildup on the components.
Sensitivity to Application Parameters
The three cooling methods varied in their sensitivity to parameters such as component
type, application distance, and application direction. Evaluation of minimum target component
temperature data in Tables 4, 5, and 6 indicate that:
• For all three combinations of application distance and direction, both liquid nitrogen
and compressed air provided lower temperatures with integrated circuits. R-12 was
less sensitive to the type of component cooled; minimum temperatures for capaci-
tors and integrated circuits were not significantly different under each application
distance/direction combination.
• The component cooling capabilities of both compressed air and liquid nitrogen are
sensitive to distance from the target component. A comparison of temperature data
in Table 5 to data in Tables 4 and 6 reveals that, as the distance from the compo-
nent to the nozzle increased from 0.25 inch to 1 inch, the minimum component
temperature decreased for both alternative methods. This relationship does not exist
for R-12, indicating that it is not as sensitive to distance.
• Comparing component minimum temperature data in Table 4 (A to C direction) to
Table 6 (D to B direction) indicates that R-12 is not sensitive to application direction.
Lower component temperatures for integrated circuits were obtained with com-
pressed, air, but liquid nitrogen yielded lower component temperatures for wound-
film capacitors. The most likely explanation of this difference is the variability
resulting from manual positioning of the dispensers.
56
-------
SECTION 5
TECHNICIAN SAFETY EVALUATION
RESULTS
Personnel from the Newark AFB Bioenvironmental Engineering Office took sound-level
measurements during operation of the compressed-air tool. A sound level of 81 dBA was recorded
at the operator work position. Because the sound levels did not exceed 84 dBA, additional
measurement was not required by the Air Force and, in accordance with Air Force Regulation 161-
35, hearing conservation precautions were deemed unnecessary.
INTERPRETATION - ' •
Sound level during operation of the compressed-air tools is not expected to represent a
hazard to operators.
57
-------
SECTION 6
POLLUTION PREVENTION POTENTIAL EVALUATION
RESULTS
Table 9 summarizes the amounts of R-12 dispensed to evaluate each of 13 test
articles. The data collection process is described on pages 9 and 10.
TABLE 9. R-12 REFRIGERANT USAGE
Test
Article #
1
2
3
4
6
7
9
10
13
14
15
16 '
17
Total
Average
Dichlorodifluoromethane — R-12
R-1 2 Released
(grams)
249.31
153.72
360.91
386.83
167.58
34.86
. 65.53 "
239.27
399.48
331.19
163.76
204.94
267.07
' 3024.45
232.65
Equivalent
1 5-ounce cans
0.76
0.47
1.10
1.18
0.51
0.11
0.20
0.73
1.22
1.01
0.50
0.63
0.82
9.23
0.71
INTERPRETATION
A total of 3024.45 grams (6.67 pounds) of R-12 were released to test the 13 articles.
The average release per article was 232.65 grams (0.51 pounds). The variability of R-12 released
per circuit board is related to -the difficulty of finding the suspected cause of circuit failure.
With the adoption of either alternative technology, release of R-12 would be eliminated
along with the wastestream of empty aerosol cans. The pollution prevention potential of wide-
58
-------
spread adoption of one or both of these technologies cannot be estimated with any confidence
**¥ , . - .»;
because neither usage nor production information for the United States was available when this
, ' ;'. <-|i- ;*
report was written. The quantities consumed vary by user, ranging from a few cans per month in
repair shops to over a thousand cans per year in production operations.
59
-------
SECTION 7
ESTIMATION OF ECONOMICS
RESULTS
It was not possible in this evaluation of alternative component-cooling materials to measure
all potential impacts on operating costs, particularly those for direct labor and materials. If an alternative
cooling material is less able to isolate the specific component causing a thermally intermittent circuit,
components may be replaced unnecessarily. Each component replacement adds cost in the form of
direct labor for replacement and retesting, component costs, and risk of circuit board damage. If a
cooling method is unable to identify a component, a circuit board may be condemned unnecessarily.
The comparison of the ability of cooling materials to isolate thermally intermittent components was
addressed in the accuracy evaluation (Section 2) and the absolute temperature drop/cooling rate
evaluation (Section 4). Cooling material costs, estimated based on actual use during the accuracy
evaluation, are the basis for the economic evaluation performed in this section.
Cooling Material Costs ,
Cooling material costs for each cooling method are summarized in Table 10. Labor costs
were not considered. Cooling material costs are based on the usage data collected during the accuracy
evaluation of thirteen test articles. Usage data were converted to cost data as follows:
• R-12 cost was based on a cost of $7.50 per 16-ounce aerosol can. Purchase price of
R-12 or R-22 freeze compound ranges from $6 to $15; $7.50 was selected as a
conservative estimate.
* Compressed-air cost was calculated using an air tool consumption rate of 15 cfm at
100 psi and an estimated compressed-air generation cost of $0.26 per thousand cubic
feet. The generation cost will vary based on power costs and other factors and should
be verified by potential users. . •
® Purchase cost of liquid nitrogen varies widely; $0.25 per liter was used as a typical cost.
Potential users should obtain price quotations from local suppliers.
60 .
-------
W5
•o
.o
0
•z.
<.
ui
00
o
o
p
d
i—
1LI
-J
(fi
<
•g
W
to
1
1
c
to
a>
0
Z
•D
'5
cr
_l
CM
BC
O
CO
o
o
o
3
:=
!
•g
5
•0 0
0 tO
*; w 5*
o to -*•
g &.<«
"1 8
0 s
c
*O 'c
ffi 5
.t | "5
< g w
1!
n
111!
I|P1
0 <
°1 Is
o = ^«-
Z©
J«t
|l|
Co00
3 "ft
C "O -~
•n (D O
c3°£
<®
c <» £
0 g S
e g g»
> 1 • »
2 "> 0 CM
g - co
u S @
•o _
_. s> «
CM CO c
»- ra t
• a)
1- <
CM
O
O
CO
(0
1^
00
o
10
CO
o
o
«>•
—
o
l"»
10
O)
CO
r»
in
<»
to
r>.
o
Jo
O)
t
CM
*»
O
O
0
4>
o
o>
r>.
CM
o
o
«>
oo
CO
o
CM
03
f.
0
,co
CM
in
CO
•v>
rx
•*
o
CM
t—
CO
in
*—
CM
CO
O
O
4f>
in
00
O
CO
r-
in
O
CM
(O
o
O
CM
CM
O
O
CM
6
CO
(O
CM
00
«>
o
V—
O)
o
(O
CO
CO
CM
O
O
in
r>
o
CD
in
O
(O
CO
o
0
«>
Tt
CO
o
o
*
1^
CM
(O
CD
CO
<»
00
CO
oo
(O
00
CO
^f
O)
o
o
•co-
in
CO
CO
CO
CM
CM
CM
O
0
O
«>
O
o
o
o
in
Tj-
00
CO
•w-
T™
in
o
CO
in
!•»
to
to
CO
o
«>
g
o
00
*f
fM
(O
CM
CO
00
o
•a>
O
rv
o
CO
CM
iv
CO
in
O
oo
0
«>
1—
o
(O
CO
•*
CO
r^
CM
O
O
•o
in
CM
CM
to
in
•*
TJ-
O
o
•«>
in
»—
O
o
en
Tt-
CM
O
in
«»
O
CM
O
CO
in
in
ID
05
^
,- O
4A-
r»
CM
t
O
^t
in
en
to
CM
CO
o
0
•co-
rn
o
O
r»
in
O
V—
CO
t
in
«>
CO
r»
O
r»
CM
en
CO
CM
o
CM
«—
o
in-
to
r*
CO
in
•«t
in
*
8-
o
o
fl>
<*•
o
o
o
en
Jo
in
en
«>
CM
CM
CO
•
o
o
(O
in
O
O
en
en
(O
CO
CO
in
r>-
•
O
en
CO
CO
t
cS
O
«>•
0
CM
in
CO
I*
o
»—
in
o
O
fi>
CM
O
6
co
CO
CO
in
r-»
co
-to-
CS
in
O
CO
r-
co
CD
in
CD
O
o
^f
CO
CO
CO
en
CM
CM
CM
'CO
0
O
i—
•?-
O
en
o
en
en
CD
**•
«>
CO
CO
o
S
g
CM
CD
CM
O
to-
Tt-
CM
CM
en
r«
CM
00
CM
in
r«.
o
o
4>
CD
CM
0
O
in
CM
CM
,_
CD
•CO-
CM
CO
o
r~
0
r-
CD
CM
r»
"3-
cn
O
•CO-
O
CO
CO
en
in
CO
r»
p>
en
CO
CM
00
CO
o
•CO-
CM
r~
CM
O
o
in
en
CM
•
CO
CO
CO
rv
CM_
•*
• t
CO
in
O
6
«/>
CM
O
in
00
CO
CD
CO
co
in
•CO-
^
0
in
CO
CM
CO
CM
09
o
n
<
61
-------
Investment Costs
There is no investment cost for R-12. Costs for the alternative cooling material dispensing
equipment are as follows:
Implementation of compressed air requires, at a minimum, investment in the air tools at
approximately $200 per unit. The investment required to generate and deliver 15 scfm at 100 psi to
the tools at a work position will vary with each potential user. Assuming no compressed air is available
In a shop, the minimum equipment required to supply one air tool is a 5-horsepower compressor, oil-filter
and desiccant filters, and nonrestrictive air lines, connectors, and valves. Purchase and installation costs
also will vary for each potential user. \
Implementation of liquid nitrogen would require approximately $500 for each half-liter Dewar
flask. Heat exchangers or other accessories would be additional costs. Cylinders for bulk liquid nitrogen
generally are provided by the suppliers at no charge. If use rate is low, suppliers may require a leasing
arrangement for the bulk containers.
INTERPRETATION
Based on cost data in Table 10, a material cost savings of $5.28 per circuit board can be projected
if testing is done with liquid nitrogen instead of R-12. This would result in payback of a $500
investment after 95 circuit boards have been tested.
For a shop with an existing adequate air supply, the average operating cost savings of $5.26
per board would pay back a $200 air-tool investment after testing 38 circuit boards. The payback period
would be extended if additional investment were required to compress and deliver air to the work
positions.
Table 11 summarizes investment and payback fgures for each alternative technology.
TABLE 11. INVESTMENT COST AND PAYBACK
Cooling Method
Compressed Air
Liquid Nitrogen
Investment
$200
$500
Payback
(circuit boards tested)
38
95
62
-------
.«*•
SECTION 8
QUALITY ASSURANCE
LIQUID NITROGEN EVALUATION
This study was performed in accordance with the Quality Assurance Project Plan for Cold
Compressed Air for Electronic Component Cooling Study, dated August 1991. Although the QAPP was
written specifically for evaluation of compressed air, the test plan was written, with concurrence from
the Technical Project Manager, to include an evaluation of liquid nitrogen. Data collection procedures
were the same for both alternative cooling methods. The results of the liquid nitrogen study indicated
that it is a viable alternative, offering both advantages and disadvantages when compared to compressed
air and to refrigerant. The Technical Project Manager approved a request from the Battelle Study Leader
to include the liquid nitrogen evaluation results in the final report.
Adding the liquid nitrogen evaluation necessitated changes and additions to the Quantitative QA
Objectives {QAPP Table 2-1). Table 12 shows the original objectives and changes, as well as objectives
related to one additional measurement: Liquid Nitrogen Released. Performance against revised
Quantitative QA Objectives is summarized in Table 13 and discussed in this section.
ACCURACY EVALUATION
R-12 Substitution •
A change to the QAPP was authorized by the Battelle Study Leader. This change was the
substitution of R-12 freeze compound (CFC, dichlorodifluoromethane) for R-22 (HCFC, chlorodifluoro-
methane) freeze compound for the accuracy evaluations specified in the QAPP. R-12 and R-22 freeze
compounds are available under the same Federal Stock Number; R-12 was the compound in stock when
the Newark AFB Experiment Coordinator obtained freeze compound from the NAFB supply area. Because
R-12 and R-22 freeze compounds generally are used interchangeably, the freeze compound substitution
was not expected to affect accuracy evaluation results.
63
-------
I
u
Ul
s
o
§
UI
g
o
ui
CO
Ul
m
JS
•o*
«* 0 W
•Sf 8
0 C C
cc o
0
' S
§
.S
ll
ll
c
o
I
D.
•o &
II
CC ^
>.
§
S
'_.
IS
0
Unit of
Measure
1
£
1
00
CO
ll
I
g
Is
*
h
ls
o
Z
g
&5
B a
Z
i
(D -Q
2
•I
-H
IQ S
5
is
s g
c*"0
Component
Idontlllcatlon
Conlldenca
i?
CO
to
CM
10
CM
a?
S«
*"
1
ll
O
Z
CO
to
ll
3
W
1
Nozzle
Electrostatic
Charge Buildup
CO
CO
Tf
W
1C
OJ
m
01
S
^
*"
e
u
I
«
0 A
o
I
Circuit Board
Electrostatic
Charge Buildup
5 1
11*
.£ ^2 J2
UJ Q CC
to
CO
S
o
o
SP
CM
se
CM
B
U
eta
a -o
o
Z
5
II
i
U
m
U.
o
HJ
CC
f
0
5
to
CO
CM
O
o
CM
c*
CM
cb
tj
ll
o
j
ll
0
U.
o
Absolute
Temperature
Drop
CM
OJ
CB
U
II
5 :
i5 '
Hi
1 :
CO
#
CO
u
ll
I
IB
o. •=
CO A
O
"35
n.
Compressed Air
Pressure
OJ
01
ID
U
ll
^
CO
ll
I
IN
ES
CM
cb
o
ll
o
Z
J
to 2
*
u.
o
Compressed Air
Temperature
«
S
s
If
eo -O
2
s
ll
o
Z
CM
g5
CM
B
Q. CD
*
I
a_o
i
LL.
0
Ambient
Temperature
I
° s
0 B c S 0 S
(J nr ^ J )_ Q
o
CO
8
o
o
•o
CM
^
•o
CM
•H
B
_O
B 3
o
CO
^0
ll
o
m
•o
Sound Level
to
-------
•S
I
LU
m
O
2
>
§
§
Q
Ul
CO
cn
UJ
U
DC
O
LL.
ec
iu
a,
m
o
-2 c
ra Q} M m
JIM
CL
tn
S «
"g c >
.1 1 s
1 1 f. '
u
CD
ill
£ t:
2
|||
"> "o 0
= £§
CD
> 0
• S S
2 5 E
o> to •&
.250
0 ° Jj1
K < O
_i'|
5 .—
o
"5 3
** to
3 5
^ "
Measuremen
0
"0
ra
ra
Q_
CO
00
ipplicable
to
I
0
XI
ra
o
0.
Q.
re
» c>
Z
o
1
u
"a-
a.
ra
o"
Z
n
.2
Ti
a.
(0
I
0
.0
ra
u
a.
ra
0
Z
o
u
c
o
S
g-
u
Component
Identification
Confidence
Accuracy
00
CO
O JJ
£ —
CC w
in
CM
CM
^
»- .
5
ft)
.g
1
to
0
Z
S
o
.2
to
Nozzle Electrost
Charge Buildup
_o
IEIectrosta
Discharge
CO
CO
CO
CO
o u>
5 £
•m -0
If
as
CM
CM,
3^
«-
CD
A
CD
•!•
CO
0
Z
ta
"o
Circuit Board
Electrostatic
Charge Buildup
at
in
E
CO
CO
8
0 JO
5 5
"5S ^
= -|2
o
CM
o j;
0 '•
• 1
= H
0
S
ra
o
a
S
.Z
8
to
tL
O
Cooling Rate
o
'CD <
1 Cooling R
and Abso
CO
CO
CO
CO
0 CO
£ -
cc £
o
CM
o fn
£ °>
o "S
= |2
0
S
ra
0
CL
ra
O
Z
u.
o
S-
Absolute
Temperature Dr<
0
3
m
0
= K O
."is
£
ipplicable
re
o
Z.
5
X)
a
CL
ra
S
-Z
a?
CM
gE
CO
J2
1
a.
CD
•g
Z
__
o.
Compressed Air
Pressure
'-j:
^
CM
CM
pplicable
re
S
_0
m
u
1
o
z
(0
p^.
0
cn
ra
a
a>
0
CO
^
CM
5
ra
Q.
ra
*J
'z
u.
o
Compressed Air
Temperature
's
s
pplicable
re
0
o
X)
ra
.a
a.
a.
re
S
Z
CM
CO
^
D
0
X)
re
CL
a
o
Z
u.
o
Ambient
Temperature
-
o
C*5
CM
CO
0
O)
ra
Q.
to
-------
Completeness
The objective of the accuracy evaluation was to compare the effectiveness of the three cooling
methods. The site for the evaluation was selected because,the systems tested there contained a variety
of circuit boards that, it was hoped, would provide a variety of test articles during the test period.
Variety in the test articles would have allowed investigation of the possible effects of circuit board
characteristics on cooling method effectiveness. At the end of the test period, however, the hoped-for
variety had not occurred.
A total of 17 circuit boards with thermally intermittent failure modes were identified by test sets
duringthe five-month test period. Of these circuit boards, only 13 were determined to have failure modes
that could be tested using the three component cooling methods. Of the 13 test articles, 11 had high
component density and component variety; one board (Sample #6) was of low density and one board
(Sample #9} was of low variety. Six of the 13 test articles were the same model circuit board, an FMC
Tank Processing Unit. . .
The impact of the actual quantity and variety of test articles was that conclusions would be more
applicable for potential users who test similar high-component-density/high-component-variety circuit
boards. Without samples of low-component-density or low-component-variety, it was not possible to
determine how accuracy of alternative cooling materials might change with varying circuit board
attributes.
ELECTROSTATIC DISCHARGE RISK EVALUATION
R-12 Substitution
R-22 freeze compound was not available through the base supply during the study period (see page
63). The Battelle Study Leader authorized the substitution of R-12 freeze compound for both nozzle and
circuit board tests because the two compounds generally are used interchangeably.
Test Location and Nozzle Test Meter Change
The test location for measurement of electrostatic charge buildup on both nozzles and circuit
boards was changed from the Newark AFB Electrostatic Discharge (ESD) Laboratory to the Carousel
Shop. The change was authorized by the Battelle Study Leader because the compressed air supply
in the ESD Lab could not provide the 90 psi specified for air-tool operation during tests.
As a result of the location change, the nozzle electrostatic charge buildup measurements could
not be made with the Ion Systems monitor specified in the QAPP because the monitor was not portable.
66
-------
The consensus of Newark AFB and Battelle staff was that a Monroe Electronics meter could be
substituted if the component cooling tools were" held so that dispensing nozzles were parallel to the
meter platen. Charge buildup on the platen was measured as the cooling material (R-12, nitrogen, or
cold air) was released. The Battelle Study Leader authorized the test meter substitution.
Nozzle Electrostatic Charge Buildup: Completeness . ' '
The completion of two measurements each for three cooling methods {R-12 aerosol with a steel
nozzle, R-'12 aerosol with a plastic nozzle, and compressed air) resulted in a completeness parameter of
100%. The additional testing of liquid nitrogen increased the number of measurements by two.
Nozzle Electrostatic Charge Buildup: Precision
The QAPP required a quantitative objective for measurement precision for electrostatic charge
buildup. To satisfy this requirement, each measurement was repeated one time and a precision measure,
RPD, was calculated using the following formula:
Precision = RPD = (A~B) x 100%
(A+BJ/2
where A, 13 = Results from repeated tests.
Precision calculations for nozzle tests are included in Table 14.
No potential user-driver precision limit was identified during the study. Due to budget constraints,
no preliminary testing was performed to gain experience with the precision capability of the measurement
method. The QAPP objective for measurement precision, 25%, was based solely on the knowledge
that electrostatic charge measurements were sensitive to many factors. Measurements were actually
more sensitive than expected as evidenced by the calculations in Table 14. The measurement precision
experienced does not indicate problems with measurement method but rather indicates that the
electrostatic charge buildup is highly variable. This variability was caused by manual positioning of the
cooling material dispenser. Because manual positioning of dispensers would be used in production,
the variability experienced during the evaluations would occur in production also. Nozzle electrostatic
charge buildup results would be directly applicable to all potential users.
67
-------
TABLE 14. ELECTROSTATIC CHARGE BUILDUP
PRECISION FOR NOZZLE TESTS
MEASUREMENT
Electrostatic
Charge Buildup
(volts)
Test 1
Test 2
Measurement
Precision (%)
Aerosol
R-12
w/Plastic
Nozzle
376
1445
117.4
Aerosol
R-12
w/Steel
Nozzle
10
13
26.1
Compressed
Air
Tool
2
3
40.0
Liquid
Nitrogen
3
3
0.0
Nozzle Electrostatic Charge Buildup: Accuracy
Because no potential user-driver accuracy objective was identified during the study, the QAPP
objective was based on using measuring equipment typically found in testing laboratories such as that
at Newark AFB. Preliminary information indicated that both the Ion Systems Model 200* and the
Monroe Electronics Model 175* measurement devices provide measurement accuracy of ±1%.
According to the manufacturer's specifications, the Monroe Electronics meter used for the nozzle
electrostatic charge buildup evaluations (see discussion page 67) actually provides accuracy of 2% of
full-scale measurement ± 1 digit. The measurement accuracy of the Monroe Electronics meter should
be acceptable to potential users of alternative cooling methods.
Circuit Board Electrostatic Charge Buildup: Steel Aerosol Nozzle Evaluation
Additional data were collected during the circuit board tests to confirm earlier Newark AFB tests,
which had indicated that steel nozzles used with R-12 aerosol cans reduced electrostatic charge buildup
when compared to plastic nozzles provided with the aerosol cans. The Battelle Study Leader authorized
the additional data collection.
* Mention of trade names and products does not constitute endorsement for use.
. 68
-------
Circuit Board Electrostatic Charge Buildup: Completeness
f . JLI !-.
4 - ': • -. ,
The completion of two measurements each for six circuit boards with R-12 and compressed-air
cooling methods resulted in a completeness parameter of 100%. The additional testing of liquid nitrogen
and R-12 with steel aerosol nozzles required an additional 24 measurements. The additional testing was
suggested by the Battelle Technician and authorized by the Battelle Study Leader.
Circuit Board Electrostatic Charge Buildup; Precision
The QAPP required a quantitative objective for measurement precision for electrostatic charge
buildup. To satisfy this requirement, each measurement was repeated one time and a precision measure,
RPD, was calculated using the following formula:
Precision = RPD =
-------
TABLE 15. ELECTROSTATIC CHARGE BUILDUP — MEASUREMENT
PRECISION FOR CIRCUIT BOARD TESTS
Electrostatic Charge
Buildup (volts)
Test
Board
#1
Test
Board
#2
Test
Board
#3
Test
Board
#4
Test
Board
#5
Test
Board
#6
Test 1
Test 2
Measurement
Precision (%)
Test 1
Test 2
Measurement
Precision {%)
Test 1
Test 2
Measurement
Precision {%)
Test 1
Test 2
Measurement
Precision (%)
Test 1
Test 2
Measurement
Precision {%)
Test 1
Test 2
Measurement
Precision (%)
Aerosol
R-12
w/Plastic
Nozzle
-251
-341
30.4
158
1250
155.1
-411
162
86.9
-1366
-1100
21.6
-143
-907
145.5
-138
-63
74.6
Aerosol
R-12
w/Steel
Nozzle
623
110
140.0
443
102
125.1
-666
-1380
69.8
-900
-470
62.8
-139
-165
:17.1
-40
-19
71.2
Compressed
Air
Tool
-58
-63
8.3
-1
0
200.0
-6
-16
90.9
-80
-69
14.8
-80
-174
74.0
-45
-50
10.5
Liquid
Nitrogen
152
205 '
29.7
28 -
26
7.4
133
45
98.9
92
25
114.5
300
254
16.6
174
247
34.7
70
-------
the Monroe Electronics meter used for the electrostatic charge buildup evaluations actually provides
accuracy of 2% of full-scale measurement? ± 1-digit. The measurement accuracy of the Monroe
Electronics meter should be acceptable to potential users of alternative cooling methods.
COOLING RATE AND ABSOLUTE TEMPERATURE DROP EVALUATION
Unit of Measure Change
The QAPP (Table 2-1) specified all temperature measurements in Fahrenheit: This objective was
not incorporated into the test plan, and staff performing the temperature measurements set the data
logger to measure in degrees Celsius. Conversion of data could have been performed to comply with
the QAPP, but another step would have been added to the trail from time/temperature plots to the final
report data. The Battelle Study Leader approved the use of degrees Celsius as the unit of measure
because it is interchangeable with Fahrenheit and because it avoids a conversion step.
R-12 Substitution
As described on page 63, only R-12 freeze compound was available at Newark AFB at the time
of the cooling rate and absolute temperature drop experiments. Although R-22 is expected to cool
components to lower temperatures than R-12, both materials are used commonly and are generally
interchangeable. The substitution of R-12 for R-22 in the cooling rate and absolute temperature drop
evaluations was authorized by the Battelle Study Leader.
Data Acquisition Methodology Description ' .
Absolute temperature drop and cooling rates are determined from thermocouple time/temperature
plots recorded by a four-channel data logger connected to thermocouples. Elapsed time was obtained
by dividing the distance measured on the data logger strip chart by the feedrate of the log paper. A
template with 1 second demarcations was used to obtain measurements. Elapsed times were
determined for each thermocouple at 10°C intervals beginning at 20°C and descending until stabilization
occurred at a minimum temperature. Elapsed times were measured from the initial dropoff of the target
component temperature and were rounded to the nearest half second. During recording, the data logger
was set so that the physical pen offsets were not reflected in the line plots. Temperature levels were
read at the demarcation lines for 10°C increments; starting and minimum temperature levels were
rounded to the nearest half degree.
71
-------
Cooling Rate: Completeness
Twelve evaluations each were performed forthe R-12 and forthe compressed-air cooling methods,
resulting in a completeness measure of 100%. An additional 12 evaluations of liquid nitrogen were
performed, for a total of 36 evaluations.
Cooling Rate; Precision
The QAPP required a quantitative objective for cooling rate measurement precision. To satisfy
this objective, each measurement was repeated once, and a precision measure, RPD, was calculated
using the following formula:
Precision = RPD = (A-B) x lOQQ/o ;
(A+B1/2
where A, B ** Results from repeated tests.
Precision calculations for cooling rates were included in Table 16. Cooling rates were calculated
for the first half of the temperature drop range., A temperature of -20°C was used as the minimum
for R-12, -70°C for liquid nitrogen, and 0°C for compressed air (except in wound-film capacitor tests
A-6-1 and A-6-2, where 10°C was used). The ranges were selected by the Battelle Study Leader
because they are expected to be the area of most concern to potential users of the alternative cooling
methods.
No potential user-driven objectives for precision were identified during the study, nor was
preliminary testing performed to gain experience with the precision capability of the measurement
method. The QAPP objective, 10%, was established solely on the knowledge that the compressed-ail-
cooling method would be sensitive to application distance and direction. The data in Table 16 indicate
that precision exceeded the objective in 11 of 18 evaluations.
The precision of the cooling rate measurements does not indicate problems with the measurement
method, but rather it indicates that cooling rate was more sensitive to application distance and direction
than expected. This was particularly true for the compressed air evaluations where the precision
objectives were greatly exceeded in five of six evaluations. Because the variability of cooling rates was
caused by manual positioning of the cooling material dispensers during material release and because
the same positioning method is used in production, potential users of alternative cooling methods could
expect comparable variability.
72
-------
TABLE 16. RATE OF COOLING - MEASUREMENT PRECISION
Test
ICH31
ICH32
ICH61
ICH62
ICH91
ICH92
ICN31 '
ICN32
ICN61
ICN62
ICN91
ICN92
ICA31
ICA32
ICA61
ICA62
ICA91
ICA92
CAPH31
CAPH32
CAPH61
CAPH62
CAPH91
CAPH92
CAPN31
CAPN32
CAPN61
CAPN62
CAPN91
CAPN92
CAPA31
CAPA32
CAPA61
CAPA62
CAPA91
CAPA92
Stan
Temp ("O
24.5
22.5
21.5
22.0
21.0
21.5
20.0
21.0
23.0
21.0
21.0
25.0
19.0 '
21.0
20.5
20.5
23.0
19.0
24.5
20.5
21.0
20.5
20.0
22.0
23.0
22.0
20.0
19.0
20.0
21.0
20.5
21.0
21.0
21.0
21.0
24.5
End
Temp (°C)
-20.0
-20.0
-20.0
-20.0
-20.0
-20.0
-7o.o :
-70.0
-70.0
-70.0
-70.0
-70.0
0.0
0.0
0.0
0.0
0.0
0.0
-20.0
-20.0
-20.0
-20.0
-20.0
-20.0
-70.0
-70.0
-70.0
-70.0
-70.0
-70.0
0.0
0.0
10.0
10.0
0.0
0.0
Delta
Temp (°C)
44.5
42.5
41.5
42.0
41.0
41.5
90.0
91.0
93.0
91.0
91.0
95.0
19.0
21.0
20.5
20.5
23.0
19.0
44.5
40.5
41.0
40.5
40.0
42.0
93.0
92.0
90.0
89.0
90.0
91.0
20.5
21.0
11.0
11.0
21.0
24.5
Elapsed Time
(sec)
6.5
6.0
1.5
1.5
2.5
4.0
25.5
29.0
20.5
20.5
25.5
23.5
5.5
3.0
21.0
12.0
21.0
12.0
11.5
11.5
11.5
11.5
14.0
14.0
24.5
27.0
53.0
45.0
25.5
27.5
22.0
18.5
36.5
16.0
20.5
25.0
Cooling
Rate
(°C/sec)
6.8
7.1
27.7
28.0
16.4
10.4
3.5
3.1
4.5
4.4
3.6
4.0
3.5
7.0
1.0
1.7
1.1
1.6
3.9
3.5
3.6
3.5
2.9
3.0
3.8
3.4
1.7
2.0
3.5
3.3
0.9
1.1
0.3
0.7
1.0
1.0
Precision
(%)
4.3
1.1
44.8
12 1
2.2
10.5
66 7
51.9
37:0
10.8
2.8
3.4
11.1
16.2
5.9
20.0
80.0
0.0
73
-------
Cooling Rate: Accuracy
The data logger and type K thermocouples provided a calculated worst-case error of ±3.21 °C
at 20°C and ±4.59°C at -175°C (based on manufacturer data). Temperature measurement accuracy
varies with temperature. Additional error can be introduced also by the method used to read temperature
levels from the data chart or by chart paper alignment in the data logger. Because temperature levels
for 10°C increments were read from chart demarcation lines, error should be negligible. Starting
temperatures and minimum temperature levels that fell between chart demarcation lines were rounded
to the nearest half degree. After chart paper loading, paper alignment was checked using the data logger
routines; because pens stopped at the extreme chart ends, error from paper alignment can be ignored.
The accuracy of elapsed time data is determined by the accuracy of the chart feed and the
accuracy of the measurement tool and method used to measure elapsed times data. Elapsed times were
determined by measuring the distance from the beginning of cooling, using a template with 1-second
demarcation lines; elapsed times were rounded to the nearest half second. Chart feed accuracy is
specified at ±0.1 % for recordings over 1 meter. Accuracy for recordings of less than 1 meter in length,
which includes all recordings made during this experiment, is not specified but is presumably worse
due to feed motor start characteristics. It is reasonable to expect that worst-case error of ± 0.5 second
for elapsed time data covers the combined error of chart feed and measurement error.
Cooling rate accuracy calculations are summarized in Table 17. The lower limit {slowest cooling
rate) was calculated using the greatest temperature drop (start temperature at upper limit and end
temperature at the lower limit) and the shortest elapsed time. The upper limit (fastest cooling rate) was
calculated using the smallest temperature change and the longest elapsed time. Accuracy was expressed
as a percentage of the calculated cooling rate by dividing the absolute difference between the calculated
limit and the calculated cooling rate. The calculated accuracies represent the absolute worst case
conditions.
No potential user-driven accuracy objectives were identified during the study. The QAPP accuracy
objective was based solely on the 2% temperature measurement accuracy expected from the data logger
and thermocouple. Calculated accuracies exceed the maximum accuracy objective because the
temperature measurement accuracy actually was worse than anticipated and because accuracy of
elapsed time measurement was included.
Although the cooling rate accuracy objectives were not met under worst case cooling conditions,
conclusions that would be meaningful to potential users could still be made. Because all measurements
were made using the same data logger on the same day and only two target component thermocouples
were used (one for capacities and one for integrated circuits), the measurement accuracy should be
much greater than worst case calculations indicate. Given these measurement conditions, comparisons
based on cooling rates of alternative cooling methods should be acceptable to potential users.
74
-------
EC
U
u
UI
UJ
cc
tn
<
UJ
o
o
u
u.
o
ui
ec
F-»'
V
UI
CO
jj&g
lill
o 2
lg&£
-1 <
•2 1 If
5 1=1
c ~ c
a£| I
«•-•= E
11 i 1
' *u
•5 "5 E
-1 1
ii |
uj 5 g
2 S §
in 2 E
Is 1
E 2 E
f-E
§£o
O £_
•o
§ ? u
tc co
1- J2
UJ
2 S-rj
s 5 o
•0 P 0
Sft.
tt g-rj
s;,2 i.
1
m
PJ
. CM
in
CO
in
o
CM
V
tn
f-s
CO
CM
ri
CM
00
to
r*.
r^.
CM
CO
CD
CO
in
CO
in
^
O
«
in
^
CM
CO
X
u
in
in
CN
en
00
to
CM
CO
tn
in
CO
in
in
CM
co*
CM
cb
CO
r*
in
CM
CO
CD
^
r*
O
CO
in
CM
0
0
CN
in
CN
CM
PJ
X
CJ
pi
r*.
en
r*.
v
CD
CO
n
in
!*•!
V-
CN
-
CM
P>"
CN
00
to
r>
^
CM
co
CO
fx
CM
in
in
^*
o
o
CM
tn
CM
CO
- X
o
en
pi
r*.
^j-
CO
^~
^*
CO
n
CO
r*^
CM
-
CM
ri
CM
00
CO
CM
in
CM
00
03
o
CO
CM
tn:
o
CM
0
o
CM
o
CM
CM
CO
X
o
tn
W
•W
p*
PS
CM
r*.
en
CM
in
CO
CM
CM
p>
CM
CO
CD
CM
cn
00
CN
ri
CO
CM
tn
CM
CM
ri
r*
CO
CO
CN
n
CM
00
CO
in
PJ
in
tn
CM
O
O
o
o
f*
o
o
CM
y
en
CO
V
P
CO
00
CD
CM
in
oi
CM
in
CO
CM
CM
ri
i*.
00
CO
CM
<*
CM
CO
r*
^
p>
O
en
CM
O
^»
O
s
o
CM
i
o
in
0>
0
in
^
CJ
y
CS
S
CN
CO
r»
00
fn-
CM
CD
CM
00
cn
in
^~
in
8
o
m
o
s
0
P)
CM
CO
'1
'.. r*
cri
en
^i
CM
cn
q
*i
CM
0
CM
CM
?
00
CM
s
00
r*.
*t
^r
in
s
o
*-
o
s
o
CM
CD
(J
en
cn
CO
00
CO
p>
ri
CO
CM
in
CM
CM
ri
r*
CO
CO
CM
*fr
CM
CO
r*
CO
PJ
tn
in
CM
O
^
o
s
o
CN
en
1
0
cn
*t
•^i
r*.
CO
T*.
ri
^t
CM
P)
in
PJ
CM
O
tn
0
s
o
tn
CM
en
i
P*
*t
tn
«
cn
CO
CM
CD
tn
CM
ri
CM
PJ
fM
CM
CM
CO
tn
tn
p>
tn
tn
0
en
0
o
o
en
1
CD
CD*
tn
0
tn
o
*»
CM
V
in
ri
tn
CM
CM
CO
CN
P)
CM
^>
CM
CD
r«»
O
r*
O
CO
0
*-
o
o
o
CM
CO
5
IT
^•*
CO
CO
cn
CM
CO
f-
6
in
CM
in
o"
CM
CM
ri
CM
CO
r*
CO
CM
PJ
r«
O
0
CM
in
O
o
o
in
O
CN
CO
s
o
r-
CO
CO
CM
O
*
CO
-
tn
CM
in
CM
ri
CM
CO
r*
CO
CM
CO
P-.
p».
0
CM
tn
o
o
o
in
3
CO
5
q
n
*
CO
0>
01
00
o
tn
CN
in
o'
CM
CM
CO
CM
CO
CM
CD
CM
00
cn
0
CM
o
CO
o
0
0
CO
CM
cn
<
u
in
3
CM
CM
^
S
q
in
CM'
in
CM
CO
CM
CO
CM
CM
CM
00
in
CO
o
CM
O
cn
o
0
o
cn
o>
S
75
-------
Z
o
u
m
|j&i
1III
I fr#
3*S-
iiii
|iii
iiii
"iiE
£3
?l i
bj 2 S
T? 3 1
0,5
>it
III
f_ t?
li
o •_
1
III
« S-rr
If"
ex_.
i5 |l-
&
«|s.
1
(O
01
CO
^
o
e>
CM
n
CM
..
CM
CO
CO
o
I*.
r*
CM
co
CM
O1
«
in
*-
in
v
o
s
In
•»
CM
TO
E
<
o
„
CM
CO
«
TO
at
CO
CM
CN
CM
TO
eo
CM
to
g
CM
n
X
g
01
g
CO
*t
*-
Cn
at
CM
N
CM
n
«
CD
N
^
CM
CD
CD
n
tn
«—
o
~-
o
g
0
CN
to
X
g
..
CM
n
*r
TO
01
O
CM
CM
CM
CO
m
CO
r*.
n
CM
CO
to
CO
to
*~
10
o
o
s
tn
o
CM
s
g
ro
s
»
CO
en
m
CO
CM
to
»f
in
CO
CM
CO
CD
CD
CM
CO
CM
m
en
CM
O
*"
o
o
o
8
0
o
CM
i
g
to
en
CD
CO
CM
CO
IO
CM
to
V
in
CO
CM
CO
CO
CD
CM
to
CM
CD
O
CO
o
•"•
o
CM
O
g
0
CN
CN
CN
cn
X
g
_
O)
^
«*
r-»
CD
in
CO
tn
«g-
n
w
00
CD
CM
S
09
m
ro
in
CN
O
CO
o
0
0
CO
CN
CO
z
<
u
o>
CO
r^
CO
CP
CD
^f
CO
in
r^
tn
CO
CM
CO
CO
CO
CN
tn
CM
CO
^
CO
o
CM
o
CM
o
o
0
CN
CN
CM
i
g
_
CD
CO
O
CD
CO
to
CO
to
CM
CM
CO
CD
CD
CM
CO
CM
CD
r*.
o
to
O
o"
o
o
0
o
CM
09
z
g
^
CD
CN
CM
CO
CO
in
in
in
*t
CM
CO
CD
CD
CN
CM
CM
CO
o
CM
O
"»
o
en
o
o
0
o
CN
CO
z
g
CM
cn
cn
CO
en
CD
CM
CO
CD
to
CM
CO
CO
CD"
CM
CO
CM
CD
10
CO
in
CM
O
°
o
o
0
o
CM
en
•z
g
o
en
CO
CO
r^
a
!
b
CO
eb
r^
CM
CO
CD
CO
CM
^
CM
CO
CO
TO
in
CM
o
"
o
o
,r>
0
CM
CM
en
Z
g
CO
*±
CO
CO
CO
n
(O
o
in
CN
tn
CM
CO
CN
CO
r-
co
CM
CO
0
o
0
CN
in
O
o
o
tn
O
CN
CO
<
<
u
CN
^>
CO
tn
^
CN
CO
O
cn
CD
CM
CO
CM
CO
CN
<<*
CN
CD
t_
in
CO
0
•-
o
o
0
CN
CM
CO
<
<
U
in
o
CO
in
O
n
CO
^
O
^
s
CD
CO
CM
CO
CM
»
CM
m
CO
O
IO
CD
CO
o
—
o
o
0
CM
1
<
U
*f
CO
CO
CO
e>
CO
0
in
co"
in
in"
to
CD
CM
TO
CM
*f
CM
CO
r^
o
0
CO
0
—
o
o
0
CM
CM
1
g
CD
CO
CO
«s-
CM
CM
r*.
O
o
CM
CO
CM
TO
CM
•»
CM
CO
o
IO
o
CM
O
«-
O
o
0
CM
1
<
U
r*
GO
CM
CO
CO
r*.
f*
o
in
tri
in
^
CM
CO
CM
CO
r*
r»
CM
n
CM
0
O
4O
CM
in
^f
0
o
in
^f
CM.
CN
1
g
76
-------
Absolute Temperature Drop: Completeness
.
Twelve evaluations were performed for both the R-12 and the compressed-air cooling methods;
therefore, absolute temperature drop parameter completeness was 1 00%. An additional 1 2 evaluations
of liquid nitrogen were performed, for a total of 36. Of the twelve, temperature of the thermocouple fell
below the minimum measurable temperature of. - 1 75 °C.
Absolute Temperature Drop: Precision
The QAPP required a quantitative objective for cooling rate measurement precision. To satisfy
this objective, each measurement was repeated once and a precision measure, RPD, was calculated using
the following formula:
Precision = RPD = (A-B)x100%
(A + B)/2
where A, B = Results from repeated tests.
Precision calculations for absolute temperature drop are included in Table 1 8. No potential user-
driven objectives for precision were identified during the study. Due to budget constraints, preliminary
testing to gain experience with the precision capability of the measurement method was not performed.
The QAPP objective, 10%, was established solely on the knowledge that the compressed-air cooling
method would be sensitive to application distance and direction. The data in Table 18 indicate that
precision exceeded the objective in 7 of 16 evaluations.
The precision of the cooling rate measurements does not indicate problems with the measurement
method but does indicate that cooling rate was more sensitive to application distance and direction than
expected. This was particularly true for the five of six compressed air evaluations in which the precision
objectives were greatly exceeded. The variability of cooling rates was caused by manual positioning of
the cooling material dispensers during material release. Because the same positioning method is used
in production, potential users of alternative cooling methods could expect similar variability in cooling
rates.
77
-------
TABLE 18. ABSOLUTE TEMPERATURE DROP - MEASUREMENT PRECISION
Test
ICH31
ICH32
ICH61
ICH62
ICH91
ICH92
ICN31
ICN32
ICN61
1CN62
ICN91
ICN92
ICA31
ICA32
ICA61
ICA62
ICA91
ICA92
CAPH31
CAPH32
CAPH61
CAPH62
CAPH91
CAPH92
CAPN31
CAPN32
CAPN61
CAPN62
CAPN91
CAPN92
CAPA31
CAPA32
CAPA61
CAPA62
CAPA91
CAPA92
Target Component
Minimum
Temperature (°C)
-45.0
-50.0
-53.0
-54.0
-51.0
-55.0
-175.0
-175.0
-134.0
-151.0
-175.0
-175.0
-27.5
-28.5
-6.0
-7.0
-18.5
-16.5
-53.5
-53.0
-57.5
-55.0
-52.5
-51.5
-134.0
-139.0
-101.0
-105.0
-150.0
-152.0
-11.5
-14.0
6.0
1.0
-12.0
-14.0
Precision
(%)
10.5
1.9
7.5
• •
11.9
• •
3.6
15.4
11.4
0.9
4.4
1.9
3.7
3.9
1.3
19.6
142.9
15.4
Exposed Thermocouple
Minimum
Temperature (°C)
-54.5
-58.0
-55.0
-55.0
-58.0
-55.0
-175.0
9
«
«
O
•
-35.5
-34.5
-12.0
-18.5
-36.0
-35.5
-59.5
-59.0
-55.5
-55.0
-57.5
-57.0
-175.0
-175.0
•
*
*
*
-35.0
-35.0
-22.0
: -14.0
-35.0
-35.0
Precision
(%)
6.2
0.0
' 5.3
»*
• *
*•
2.9
' 42.6
1.4
0.8
0.9
0.9
»•
• « .
• •
0.0
44.4
0.0
' Measurement not taken. Assumed to be-175°C.
• Precision not calculated — no measurement data or measurement method at minimum limit of -175°C.
Absolute Temperature Drop; Accuracy
Absolute temperature drop accuracies for each evaluation are summarized in Table 19.
Measurement accuracy calculations are based on the additive (worst case) accuracy of the data logger
and the type K thermocouple. Accuracy of both components of the measurement system are stated in
terms of the measured temperature.
No potential user-driven accuracy objective for absolute temperature drop measurement was identified
during the study. The QAPP accuracy objective of 2% was established based on planned use of
78
-------
TABLE 19. ABSOLUTE TEMPERATURE DROP - MEASUREMENT ACCURACY
Test
ICH31
ICH32
ICH61
ICH62
ICH91
ICH92
ICN31
ICN32
ICN61
ICN62
ICN91
ICN92
ICA31
ICA32
ICA61
ICA62
ICA91
ICA92
CAPH31
CAPH32
CAPH61
CAPH62
CAPH91
CAPH92
CAPN31
CAPN32
CAPN61
CAPN62
CAPN91
CAPN92
CAPA31
CAPA32
CAPA61
CAPA62
CAPA91
CAPA92
Target
Component Min.
Temp. (°C)
-45.0
-50.0
-53.0
-54.0
-51.0
-55.0
-175.0
-175.0
-134.0
-151.0
-175.0
-175.0
-27.5
-28.5
-6.0
-7.0
-18.5
-16.5
-53.5
-53.0
-57.5
-55.0
-52.5
-51.5
-134.0
-139.0
-101.0
-105.0
-1 50.0
-152.0
-11.5
-14.0
6.0
1.0
-12.0
-14.0
Measurement
Accuracy
{%)
7.2
6.5
6.1
6.0
6.3
5.9
2.6
2.6
2.8
2.7
2.6
2.6
11.7
11.3
53.4
45.8
17.3
19.4
6.0
6.1
5.6
5.9
6.1
6.3
2.8
2.8
3.0
3.0
2.7
2.7
27.9
22.9
53.4
320.1
26.7
22.9
Exposed
Thermocouple Min.
Temp. (°C)
-54.5
-58.0
-55.0
-55.0
-58.0
-55.0
•1 75.0
*
»
*
«
9
-35.5
-34.5
-12.0
-18.5
-36.0
-35.5
-59.5
-59.0
-55.5
-55.0
-57.5
-57.0
-175.0
-175.0
*
0
*
*
-35.0
-35.0
-22.0
-14.0
-35.0
-35.0
Measurement
Accuracy
(%)
5.9
5.6
5.9
5.9
5.6
5.9
2.6
« *
* * * . - '
•o a
9.1 '
9.3
26.7 :
17.3
8.9
9.1
5.4
5.5
5.8
5.9
5.6
5.7
2.6
2.6
*'*
«• »
9.2
9.2
14.6
22.9
9.2
9.2
* Measurement not taken. 'Assumed to be — 175°C
"* Accuracy not calculated — no measurement data or measurement method at minimum limit of.-175°C.
the data logger only. The accuracy of the type K thermocouple erroneously was not included. The
accuracy provided by the temperature measurement system should be acceptable to potential users of
alternative cooling methods because the same data logger was used for all measurements and the same
thermocouples were used for capacitor and wound-film measurements (target component and exposed).
Compressed-Air Pressure: Completeness
All twelve planned measurements were taken; completeness of the parameter was 100%.
79
-------
Compressed-Air Pressure; Accuracy
Air pressure was regulated so that the pressure at the work position gauge read 100 psi. The
pressure gauge used provides an accuracy of ±2%, which is within the QAPP objective of ±3%. The
gauge accuracy of ±2% is valid in the range from 40 to 120 psi.
Compressed-Air Temperature: Measurement Method Change
The QAPP specified that the exposed thermocouple would be used to measure the temperature
of compressed air as it is released with a blowoff tool. Temperatures obtained using this measurement
method are not representative of the temperature of the air before release. The compressed air used for
the experiment includes an air chilling system to cool air as it is delivered from the compressor to the
storage tank. The chiller reduces the temperature of the air to approximately 80°F. When the compressed
air arrived at the air tool, it was assumed, to be at or slightly above shop ambient temperature. This
assumption is supported by the fact that temperatures achieved when cooling exposed thermocouples
with compressed air are very close to advertised cooling capability; if the compressed air was significantly
above ambient temperature, minimum thermocouple temperature would have been warmer as well.
Compressed-Air Temperature: Completeness and Accuracy
A method for measuring compressed-air temperature at the air tool was not available during the
study; completeness of the parameter was 0%. Accuracy for the parameter is no longer applicable;
compressed-air temperature was known to be close to the ambient shop temperature of approximately
20°C.
Ambient Air Temperature; Completeness
All 24 planned measurements were taken. Completeness of the parameter was 100%.
80
-------
Ambient Air Temperature: Accuracy
•5*~ - . ' • . rSft
Ambient air temperature was obtained using the data logger and the exposed thermocouple.
As discussed in the accuracy discussion for absolute temperature drop, the QAPP accuracy objective
of 2% was based on using the data logger and the effect of the type K thermocouple accuracy was
omitted erroneously. At 20°C, the accuracy of the measurement system is calculated at 3.21%,
assuming worst-case condition with the accuracy of both data logger and thermocouple added together.
TECHNICIAN SAFETY EVALUATION
Sound-Levol Measurement Procedure Change
The QAPP specified an extensive test plan for evaluating sound levels during compressed-air too!
operation. Before Newark AFB technicians were permitted to operate the compressed-air tool. Base
bioenvironrnental engineering personnel performed an evaluation that included sound-level measuremerrt.
TSgt Earl Matthews performed the measurement, following Air Force procedures, and determined that the
sound levels were well below the threshold of 84 dBA, where more extensive measurements would be
necessary to characterize operator exposure hazards. The Battelle Study Leader cancelled the extensive
testing specified in the QAPP because it would add unneeded cost to the study while providing unneeded
information. The measurement equipment and measurement techniques, and therefore the sound-level
measurement data, should be acceptable to potential users of the compressed-air tools.
Sound Level: Accuracy
TSgt Matthews performed the measurements using a General Radio 1565B Sound-Level Meter *
The meter was calibrated immediately before use with a General Radio 1562 Sound-Level Calibrator,
which was calibrated on February 6, 1992. The calibrator provide measurement accuracy ±0.3 dB
at 500 Hz and ±0.5 dB at other frequencies. The instructions for calibrating the sound-level meter
are to ensure meter measurement is within 0.5 dB of the calibrator at 500 Hz, within 1.0 dB at 125,
250, and 1000 Hz, and within 2.0 dB at 2000 Hz. It is not necessary to convert available instrument
accuracy information to a dBA error for comparison to the QAPP objective of ±2% dBA because the
equipment used is standard for sound-level measurements and the accuracy is sufficient for potential
users of this study.
Mention of trade names and products does not constitute endorsement for use.
81
-------
Sound Level: Precision and Completeness
Following Air Force sound-level measurement procedures, TSgt Matthews measured sound levels
at the work position during air tool operation. During a period of approximately 10 seconds, the peak
sound level observed was 81 dBA. Because, as discussed above, the recorded sound level was below
a threshold of 84 dBA, the extensive testing specified in the QAPP was not performed and the precision
and completeness objectives no longer were applicable.
POLLUTION PREVENTION POTENTIAL
R-12 Substitution
As described on page 63, R-12 was used instead of the R-22 specified in the QAPP. With respect
to the amount of cooling material used in circuit board evaluations, any difference between R-12 and
R-22 is expected to be insignificant.
CFC Released: Completeness
As described on page 66, the 13 test articles evaluated represent a completeness of 72.2%.
The impact of the actual quantity and variety of test articles on the Pollution Prevention Potential
evaluation was similar to the impact on the accuracy evaluation.
CFC Released: Accuracy
The scale used to weigh aerosol cans of R-12 during accuracy circuit board evaluations was a
Mettler PC440 Electronic Top Loading Balance*. For measurements in the range of 450 grams (the
weight of a full can of R-12), the tolerance of the scale is 0.01 grams. This level of accuracy is within
the QAPP objectives.
Mention of trade names and products does not constitute endorsement for use.
82
-------
ESTIMATION OF ECONOMICS
R-12 Substitution
The cooling material use rate differences between R-12 and R-22 are expected to be insignificant,
as was discussed on page 66. The government procures both under one stock number at one price.
Therefore, cooling material costs for R-12 and R-22 can be considered equivalent.
Compressed-Air Release Time: Completeness
As described on page 66, only 13 compressed-air release time measurements were obtained,
representing a completeness measure of 72.2%. The impact of the actual quantity and variety of test
articles on the Estimation of Economics evaluation was similar to the impact on the accuracy .evaluation.
Compressed-Air Release Time: Accuracy
The stopwatch used to measure release time against a known standard was checked and found
to gain 4 seconds per 24-hour period. The sum of stopwatch error and inaccuracy related to nonsimul-
taneous activation of the stopwatch and the air-tool switch by the operator are assumed to be within
the 5% QAPP objective. • •
Compressed-Air Pressure: Completeness
As described on page 66, only 13 compressed-air release time measurements were obtained,
representing a completeness measure of 72.2%. The impact of the actual quantity and variety of test
articles on the Estimation of Economics evaluation was similar to the impact on the accuracy evaluation.
Compressed-Air Pressure: Accuracy
Air pressure was set at the pressure regulator so that pressure at the work position gauge read
100 psi. The pressure gauge utilized provides an accuracy of ±2%, which is within the QAPP objective
of ±3%. The gauge accuracy of ±2% is valid in the range from 40 to 120 psi. The regulator was
a new unit purchased by Newark AFB specifically for this study. It was not calibrated prior to the study,
but that should not have affected the study results because the air pressure gauge at the work position
was used to measure line pressure at the tool. Air-pressure gauge inaccuracy could affect the cooling
characteristics of the air tool and the volume of air consumed during accuracy evaluations. The air-
83
-------
pressure gauge was not calibrated, but the fact that the absolute temperature drop temperatures for
the exposed thermocouples (Tables 4, 5, and 6) were consistent with tool specifications indicates that
excessive air pressure gauge inaccuracy did not exist during the evaluations.
84
-------
SECTION 9
DISCUSSION
The objective of this study was to characterize aerosol cans of refrigerant, compressed air and
liquid nitrogen as methods for cooling electronic component cooling during testing. Data obtained from
testing were used to compare alternative cooling methods in terms of accuracy, electrostatic discharge
risk, cooling performance, technician safety hazards, pollution prevention potential, and economics.
Conclusions drawn from this study are as follow:
» The compressed-air tool evaluated during the study was unable to cool components to
the temperature level that was obtained with either R-12 or liquid nitrogen. However, the
results of the accuracy test indicate that during all but one test, temperatures achievable
with the compressed-air tool were low enough to reproduce failures.
•. Liquid nitrogen has the capability to readily cool components to below -175°C if
dispensed closely enough. At such temperatures, components may fail from temporary
changes in output signals or fail permanently from physical damage. Two methods to
control the temperature of components are to maintain dispensing nozzle distance and to
slow the cooling rate of the dispenser by adding heat exchangers or smaller orifices. Both
methods rely on a technician to a greater extent than either compressed air or R-12.
Further discussion of component temperature control with liquid nitrogen is provided in
Appendix A.
o
Neither alternative is expected to increase safety risks to technicians when compared to
aerosol refrigerants. Noise levels are higher during compressed-air tool operation than with
R-12 or liquid nitrogen, but they are not high enough to pose a health hazard to users.
Handling of liquid nitrogen presents a safety risk in the form of exposure to low
temperatures, but technician training and proper safety procedures and equipment are
expected to minimize risk. As with any aerosol, release of refrigerants under pressure
presents a safety risk that is controlled through training.
Replacement of aerosol refrigerant prevents emissions of substances that deplete the
stratospheric ozone layer as well as accumulation of empty aerosol cans requiring landfill
disposal. With liquid nitrogen, only nitrogen is emitted and refillable bulk containers and
dispensers are used. Compressed air generates a small amount of pollution in the forms
of waste compressor oil and filter elements; however, the incremental increase in these
wastestreams following adoption of the compressed-air method is not expected to be
significant.
Material costs of either alternative are expected to be lower than R-12 or R-22 at current
prices. Prices of R-12 and R-22 will undoubtedly escalate. Eventually, these materials
will be unavailable due to regulatory prohibition.
85
-------
• Investment cost to implement liquid nitrogen is expected to consist of the price of dispensing
Dewar flasks at approximately S500 each in the half-liter size. Compressed-air tools cost
approximately $200 each. The cost of equipment to deliver compressed air that is clean, dry, and
near room temperature in the volume and pressure required to achieve maximum cooling
capability will depend on existing equipment and the number of tools to be utilized.
• The results of this study led Newark AFB personnel to conclude that either of the cooling methods
tested were viable alternatives to aerosol cans of refrigerants, recognizing that control of electronic
component temperatures when using liquid nitrogen required resolution.
86
-------
SECTION 10
DATA REDUCTION
ACCURACY EVALUATION
Component identification confidence measurements were read directly from Accuracy Experiment
Data Collection Packages, which were filled out by Newark AFB technicians during test article
evaluations.
ELECTROSTATIC DISCHARGE RISK
Nozzle Electrostatic Charge Buildup measurements were transferred directly from data collection
sheets to tables in this report. Comparisons of recorded voltage levels for four nozzle/cooling material
combinations were made. Precision of repeated measurement was calculated as described on page
67.
Circuit board electrostatic charge buildup measurements were transferred directly from data
collection sheers to tables in this report. Comparisons of recorded voltage levels for four nozzle/circuit
board combinations and six circuit boards were then made. Precision of repeated measurements was
calculated as described on page 69.
COOLING RATE AND ABSOLUTE TEMPERATURE DROP
Cooling rates were calculated from temperature change overtime measured as cooling material
was dispensed toward thermocouples embedded in target components. Elapsed time was obtained
from data logger plots on plotter paper. A clear template demarcated in 0.5 second lines was laid over
the plot with the start line (time = 0) located at the time where material release began and the template
demarcation lines visually parallel with the time demarcation lines of the plotter paper. The elapsed
time then was read from the template by finding the line closest to the point where the data logger
plotter brown line crossed the temperature demarcation line on the plotter paper. Cooling rates were
calculated by dividing the temperature change by the elapsed time.
Precision of cooling rates for repeated tests were calculated as described on page 72. Accuracy
was estimated using worst case conditions. The upper limit of cooling rate was calculated using the
87
-------
longest temperature change (based on temperature measurement accuracy) and the shortest elapsed
time (based on temperature measurement accuracy). The lower limit was calculated using the shortest
temperature change and the longest elapsed time.
Absolute temperature drop was determined from data logger paper plots. The lowest point that
the brown plotter pen traveled represented the lowest temperature reached by the target thermocouple
during cooling material release. The temperature reached was determined using the demarcation lines
of the plotter paper.
Precision of absolute temperature drop measurement was calculated as described on page 77.
Accuracy of absolute temperature measurements was calculated by assuming worst-case conditions
and adding the accuracy of the data logger and type K thermocouple. Data logger accuracy was
calculated using the following formula:
±(.0005 x observed temperature °C) + 1 °C '
Thermocouple accuracy was calculated using the following formula:
maximum of: ±2.2°C or .02 x observed temperature °C
Figures 21 to 44 in this report were created using the following methodology:
(1) Starting temperatures for each thermocouple were read from the time
temperature plots using the plotter paper demarcation lines. Elapsed times
for each thermocouple to reach 10°C increments (as cooling materials were
dispensed) were determined using the template described earlier in this
section. The minimum temperature for each thermocouple was determined
using the plotter paper demarcation lines, and elapsed time was determined
using the template. All temperature and elapsed time data were recorded
on worksheets along with appropriate test information.
(2) Temperature and elapsed time data were entered into a spreadsheet program
(Cricket™) and plotted using the spreadsheet capabilities. The plot was then
imported into a computer-aided design program (MacDraw™). Test descriptive
infprmation was then added and the plots were printed.
No data reduction was required for either the compressed-air pressure or the compressed- air
temperature measurements. Ambient temperature measurements were read from data logger plots using
the red pen plots, which represented the exposed thermocouples on test boards. Accuracy of the
temperature measurements was calculated as described on page 74.
88
-------
SAFETY
Sound-level measurements were read from the sound-level meter by TSgt Matthews of Newark
AFB. The peak value observed was recorded in a letter, a copy of which was provided to the Battelle
Study Leader. No data reduction was required.
POLLUTION PREVENTION POTENTIAL
CFC released during the accuracy evaluation of each test article was determined from beginning
and ending aerosol can weights recorded by technicians in the Data Collection Packages. Weight in
grams was converted to cans using a conversion factor of 328 grams per can. An average use per
test article was obtained by dividing the total CFC released during the accuracy evaluation by the number
of test articles.
ESTIMATION OF ECONOMICS
Compressed-air release time measurements were recorded by technicians in the Data Collection
Packages during the accuracy evaluation. Release time was converted to compressed-air consumption
using a factor of 15 scfm. The published specifications of the air tool used are 15 scfm at 100 psif
which was the air pressure used in the accuracy evaluation. Consumed compressed air was converted
to cost using an estimate of $0.26 per 1000 scf. This estimate was provided by the air tool
manufacturer; it is acknowledged that the cost will vary with-geographic location and compressed-air
system. An average cost per test article was determined by dividing total compressed-air cost by the
number of test articles.
Liquid nitrogen released was determined from start and finish Dewar weights recorded by
technicians in the Data Collection Packages for the accuracy evaluation. The weight of the liquid
nitrogen was converted to volume using a chemical handbook factor of 814 grams/liter. An estimate
of $0.25 per liter was used to convert liters of liquid nitrogen to cost. Liquid nitrogen cost will vary
depending on numerous factors. An average cost per test article was determined by dividing total liquid
nitrogen cost by the number of test articles.
Compressed-air pressure was recorded in the Data Collection packages during accuracy
evaluations of test articles. No data reduction was performed.
89
-------
APPENDIX A
COMPONENT TEMPERATURE CONTROL: LIQUID NITROGEN
All of the tested cooling methods will achieve a steady component temperature level if held in
the same orientation and dispensed for a long enough time. With R-12 and compressed air, the coldest
level a component will achieve is near —60°C and -40°C, respectively. However, liquid nitrogen
dispensed towards a component eventually will reduce the temperature of the component to near that
of liquid nitrogen, possibly degrading component performance temporarily or permanently. The minimum
component temperature can be controlled by four techniques described in the following paragraphs.
Holding the nozzle away from the component will warm the stream of material somewhat before
it reaches the component. The drawback is that the time required to cool the component to a given
temperature level also increases. In the ideal process, the nozzle would be moved away from the
component as the temperature of the material stream dropped, thus obtaining fast cooling without
exceeding a desired minimum component temperature. It is conceivable that a technician could learn
to operate the dispenser in such a manner, but it would remain an imprecise control method.
Limiting the length of time material is released will prevent the valve and nozzle, and therefore
the material stream, from exceeding some minimum temperature level. The drawback to this approach
is that the technician must control the release times and allow sufficient time .between releases for the
valve and nozzle to return to ambient temperature.
Orifices can be used to control the rate at which material is released so that the valve and nozzle
never exceed a desired steady-state minimum temperature. The drawback to this approach is that the
cooling rate is slowed as the volume of material released is reduced. Three orifices supplied by Brymill
Corp. were evaluated using the data logger to record temperature levels over time. The integrated circuit
test board described on page 11 was used with only the target component connected to the data logger.
All tests were performed at an application distance of .25 inch. With the smallest orifice, size C, the
minimum target component temperature stabilized at approximately -90°C after a 5-minute release
time. Both of the larger orifices, sizes A and B, allowed the target component to achieve the lowest
recordable temperature (approximately — 140°C) after 1.5 and 2.5 minutes, respectively. When
compared to data in Table 4, which show that without a restrictive orifice liquid nitrogen cooled the
target component to — 175°C in 51 seconds, these results demonstrate that all three orifices reduce
the coolmg rate.
90
-------
A heat exchanger can be attached to the nozzle to slow the cooling rate of the liquid nitrogen.
Eventually the stream will approach liquid nitrogen temperatures unless an orifice is also used to restrict
* -;- ' ' -I - ^
flow of material. The drawbacks are the same as yvith the reduced release rate alternative. Two heat
• . .. ''&. •'( '
exchangers were evaluated using the test method described above. One was the standard unit provided
by Brymill, and the other was a standard unit that had been modified by removing approximately half
its length. When used with the A, B, or C orifices, the heat exchangers slowed the cooling rate.
However, in the case of the A and B sizes, the target component temperature still reached the minimum
recordable temperature, The standard heat exchanger slowed the cooling rate more than the modified
unit. When the standard heat exchanger was used without an orifice, approximately 90 seconds were
required to reduce the target component from the ambient temperature of approximately 20°C to 0°C.
Without the heat exchanger, as shown in Figure 30, the elapsed time for a similar temperature reduction
was about 5 seconds.
In conclusion, restrictive orifices or a mechanical standoff are the only control methods for liquid
nitrogen that can ensure that some desired minimum temperature is not exceeded. With experimenta-
tion, an orifice or a standoff could be sized to the minimum temperature required. As noted above,
the drawback to these approaches is a slowed cooling rate. All other control techniques will rely on
the technician to control the component temperature level using only knowledge of the cooling
characteristics of liquid nitrogen with a specific dispenser apparatus and visual cues, such as frost
buildup on the dispenser. This may be a viable approach if a heat exchanger is used that slows the
cooling rate so that long dispense times are required before unacceptable temperatures are achieved.
As with the first two control methods, slowing the cooling rate is a drawback that may reduce the
effectiveness of liquid nitrogen as a trouble-shooting tool to identify electronic components with
thermally intermittent failure modes.
91
-------
APPENDIX B .
MEASUREMENT PRECISION OBJECTIVES
Measurement precision objectives were established in the QAPP for five parameters:
• Nozzle electrostatic charge buildup
• Circuit board electrostatic charge buildup
• Cooling rate
• Absolute temperature drop
• Sound level
Precision of sound-level data was not calculated because a single measurement was performed
(see the discussion on page 82). Measurement precision calculations for the other four parameters are
summarized in Tables 14, 15, 16, and 18.
Precision was calculated using the following formula:
Precision= RPD = (
where A, B - measurements from repeated tests.
92
------- |