Weinke posted to this list last November about the same subject, but he
reported that 70% was the apparent threshold for increased combustibility in
tests conducted at Los Alamos.
Gary

-----Original Message-----
From: Bill Wolk [mailto:BillWolk@ea*.ne*]
Sent: Tuesday, September 04, 2001 9:14 AM
To: Iain Smith; Techdiver; IB2.ian@ca*.fr*.co*.uk*;
Gregory.Porter@AR*.Bo*.co*; FLTechDiver@mikey.net;
vbtech@ci*.co*
Cc: deepdive@xt*.co*.nz*
Subject: Re: CO detection


Dave Dalton was kind enough to post a Nasa/Johnson Space Center study on
regulator contamination to his website. The original links are below.
The test was done at White Sands and concluded that even dirty, non-02
cleaned regs were safe to use with mixes up to 50% oxygen. Although this
test was performed with 50% O2 and not 100%, I thought it worth posting
to the list to put the issue of O2 cleaning in perspective.

>
>http://home.quixnet.net/~dmdalton/nasa.htm
>
>or if you want to see the version with the signatures you can see it at
>http://www.dnax.com/
>
>look under the link for Nitrox Research.

-----


Document no: TR-900-001

December 22, 1997

Component Testing and Clean Verification of SCUBA Equipment for the NASA
JSC

Neutral Buoyancy Laboratory

Lyndon B. Johnson Space Center

White Sands Test Facility

P.O. Box 20

Las Cruces, NM 88004

(505) 524-5011


Abstract

The NASA White Sands Test Facility was requested by NASA Johnson Space
Center (JSC) to perform component testing and clean verification on the
self-contained underwater breathing apparatus (SCUBA) diving equipment to
be used in the JSC Neutral Buoyancy Laboratory (NBL).

The NBL facility requires divers to breathe a nitrogen/oxygen (NITROX)
gas mixture with an oxygen concentration ranging between 2 I and 49
percent. The increased oxygen concentration warrants testing of the
apparatus in its worst-case use environment to evaluate potential
ignition mechanisms. Three SCUBA equipment assemblies were tested, and
included a first-stage regulator with flex hoses to two second-stage
regulators, a submersible pressure gauge, and a buoyancy control device.
Sixty pneumatic impact cycles were performed at 20.7 MPa (3000 psi) in a
50-percent NITROX gas mixture on each of the three test articles. Each
test article passed the impact test with no failures or evidence of
ignition, The clean verification indicated that gross amounts of
particulate and non-volatile residue were present in the oxygen-wetted
portions of the assemblies.

1.0 Introduction

WSTF was requested by NASA Johnson Space Center (JSC) to perform
component testing and clean verification on the self-contained underwater
breathing apparatus (SCUBA) diving equipment to be used in the JSC
Neutral Buoyancy Laboratory (NBL}. The component testing and clean
verification were performed in response to recommendations made by an
earlier oxygen hazards analysis of the JSC NBL SCUBA equipment (Forsyth
1996).

2.0 Objective

The objective of the component tests was to determine the susceptibility
of the NBL SCUBA equipment to ignition by pneumatic impact in a
50-percent nitrogen/oxygen (NITROX) mixture at worst-case operating
conditions. The objective of the clean verification was to determine the
approximate clean level of the oxygen-wetted portions of the SCUBA
assemblies.

3.0 Background

The JSC NBL facility will require divers to use a special NITROX gas
mixture instead of conventional breathing air in their cylinders.
Conventional breathing air has an oxygen concentration of approximately
21 percent; it ranges between 21 and 49 percent for the NITROX gas
mixture. Previously, WSTF had performed an oxygen hazards analysis on
this SCUBA equipment to consider the materials flammability and potential
ignition mechanisms with the new gas mixture (Forsyth 1996).

The SCUBA equipment analyzed in the oxygen hazards analysis included an
oxygen cylinder, an oxygen cylinder valve, a first-stage regulator, a
second-stage regulator, a submersible pressure gauge, and a buoyancy
control device. The analysis recommended that component testing be
performed on the first-stage regulator and downstream assembly, at
worst-case use conditions, to evaluate the equipment's vulnerability to
ignition by pneumatic impact with gaseous oxygen. The analysis also
recommended that clean verification be performed annually on a sample
population of the SCUBA equipment to evaluate the cleanliness level.

4.0 Approach

The approach for the component tests was to randomly select four SCUBA
assemblies that had been in service for approximately one year from the
JSC Weightless Environment Training Facility. These assemblies were
delivered to WSTF for testing and evaluation.

The test articles were placed in a test fixture to secure them for impact
and then allow for a remote venting of the test article. Each component
assembly was tested individually in its use configuration by rapidly
impacting the first-stage regulator for 60 cycles at the worst-case
operating pressure. Three component assemblies were tested by pneumatic
impact to assess the repeatability of the data.

Clean verification tests were performed on four SCUBA assemblies. The
assemblies were functionally checked, disassembled, sampled for
particulate and non-volatile residue (NVR), and re-assembled. The SCUBA
assemblies were then returned to the customer.

5.0 Experimental

5.1 Test Articles

The SCUBA assemblies tested were the same as those analyzed in the
hazards analysis, excluding the oxygen cylinder and oxygen cylinder
valve. The assembly included a first-stage regulator, two second-stage
regulators, a submersible pressure gauge, a buoyancy control device, and
their connecting flexible hoses (Figure 1). A summary description of the
test articles is shown in Table l.

Only three of the assemblies were tested by pneumatic impact (SCUBAs 2,
3, and 4). The fourth assembly, SCUBA 1, was a demonstration unit but was
included in the clean verification tests. The demonstration unit was the
only assembly that was ultrasonically cleaned before shipping. In three
of the assemblies, NITROX 0-ring kits were installed in the regulators
before delivery to WSTF (SCUBAs I, 3, and 4). These kits had Viton®
0-rings. In the other assembly (SCUBA 2), no pretest preparation was
performed; the original 0-rings were left installed, and the assembly was
not cleaned before shipping.

Table 1

Summary Description of SCUBA Test Articles

First Stage Second-Stage Second-Stage

Name Regulator Regulator 1 Regulator 2 0-Rings Clean

SCUBA Mares MR 12 III Mares MR 12 III Mares MR l2 III NITROX Ultrasonic

1 S/N 651378 S/n W29S/N W30

07168648 08036509

SCUBA Mares MR 12 Mares MR 12 Mares MR 12 III Standard None

2 DFC S/N E70335 Navy S/N 06204829

S/N 513713

SCUBA US Divers US Divers US Divers NITROX None

3 Conshelf 21/22 Conshelf 14 Conshelf 14

S/N ALB0524 S/N B76K2310 S/N JY6K0735

SCUBA US Divers US Divers US Divers NITROX None

4 Conshelf 21/22 Conshelf 14 Conshelf 14

S/N ALB1100 S/N B76K119 S/N JY6K1S57

 

 

5.2 Test Method

The test method used to evaluate these SCUBA assemblies was pneumatic
impact, or adiabatic compression. When a dead-ended test article filled
with ambient pressure oxygen is rapidly pressurized, the pressure can
increase too quickly for the heat of compression to be dissipated to the
surroundings. Most of the thermal energy generated is contained in the
compressed gas, resulting in a small slug of hot gas at an elevated
pressure located at the dead end. In this test configuration, the gas is
rapidly compressed against the first stage regulator seat, which is the
"dead-end." In theory, the gas temperature generated by adiabatic
compression should reach the autogenous ignition temperature of the
polymer seat material in 100-percent oxygen. However, in this case, the
first- stage regulator allows flow upon impact, so is not a true
dead-end. Also, the gas is a 50-percent oxygen concentration, which
decreases the severity of the test environment.

5.3 Test System

The tests were conducted using a WSTF pneumatic impact component test
system that simulates a sudden pressurization event such as what could be
caused by a fast-opening system valve. The fast opening system valve
allows pressurization rates to the test article of approximately 20 ms
with system pressures up to 69 MPa (10,000 psi). A VME computer system
was used to control the pressurization of the test articles during the
pneumatic impact cycles. The VME software was written to allow computer
control of the fast-acting impact valve, the vent valve, and the upstream
isolation valve. Otherwise, all valves were controlled by the test
conductor. The VME system was also used for data acquisition. Standard
video recordings were made of all tests.

A test fixture was designed and implemented to remotely vent the gas
through the second-stage regulator (Figure 3). A Bimba actuator provided
a piston to push the vent button on the second-stage regulator, The test
fixture also secured the SCUBA regulators and flexible hoses and provided
a mount for the test article pressure gauge so it could be monitored by
video.

5.4 Test Conditions

The test conditions were chosen to closely simulate the actual
application. A 50-percent NITROX

gas mixture was used in all tests. The pretest temperature was ambient.
The test pressure up to the

first-stage regulator was from ambient to maximum use pressure, 20.7 k
0.7 MPa (3000 k 100 psi).

The pressurization time to the first-stage regulator was consistent at
approximately 20 ms.

 

 

5.5 Procedures

5.5.1 Pneumatic Impact Tests

The test articles were treated as clean hardware while at WSTF. Latex
gloves were used in handling the test articles at all times, and each
component was double bagged to maintain clean whenever the test articles
were not in the test system.

The test articles were inspected prior to installation in the test
system, and any anomalies were noted. The test article was then mounted
and secured in the test fixture. The first stage regulator was connected
to the system's test article interface (Figure 3). An aluminum blast
shield and cinder blocks were placed around the test article to protect
the test system from damage.

After installation, the test article was leak checked with gaseous
nitrogen (GN2). The vent on one second-stage regulator was then opened,
and the test article assembly was purged with low pressure NITROX to
adequately displace the GN2, The vent was then closed, and the system was
pressurized up to the fast-acting isolation valve with NITROX at test
pressure. The data acquisition and video equipment were then started.
Upon command, the software opened the fast-acting valve, creating a rapid
compression on the first-stage regulator, followed by a lower-pressure
surge to the second-stage regulators, The pressure was held for
approximately 18s to determine if leaks were present, indicating a
possible ignition. Ignition was also determined by a flash on the video
or by an audible report. If no reactions or failures were evident, the
pressure was vented through the second-stage regulator vent and upstream
of the first-stage regulator, and the test article was impacted again. A
total of 60 pneumatic impacts were performed on SCUBAs 2, 3, and 4.

5.5.2 Clean Verifications

First, a functional check was performed on each SCUBA assembly by
pressuring the assembly up to its maximum operating pressure with GNp and
recording the pressure decay over 5 min. The SCUBA assemblies were then
carefully disassembled to the piece-part level in a Class 100 flowbench
to ensure cleanliness was maintained before flushing, Only the
oxygen-wetted pieces of the assemblies were tested for cleanliness. The
metal pieces were separated from the polymer pieces prior to flushing,
and different procedures were used for the metal and polymer pieces. All
metal pieces were flushed with CFC l13, and both particulate and NVR
samples were evaluated. All polymer pieces were flushed separately with
distilled water, and only a particulate sample was evaluated. No NVR
levels were analyzed for polymer pieces because many polymers are
composed of hydrocarbons that could be released into the sample and
affect the results. For this reason, the flexible polymer hoses were
flushed only with water and evaluated for particulate.

The metal pieces of the first-stage regulator that were individually
flushed included the internal trim pieces, the internal regulator body,
and the inlet filter. The first-stage regulator soft goods were flushed
together as one sample. The metal pieces of the second-stage regulators
that were individually flushed included internal trim pieces and the
internal regulator body. The second-stage regulator soft goods were
flushed together as one sample, and the flexible hoses were flushed
individually.

6.0 Results and Discussion

6.1 Pneumatic Impact Tests

No indication of ignition or failure was observed in any piece of the
three SCUBA assemblies tested. Even test article SCUBA 2, which did not
have a NITROX-compatible soft good kit, did not sustain any ignitions.
There were no flashes or audible reports in any tests, nor did any test
article leak pressure during the pressure hold phase of the impact cycle.
In each test, the pressurization rate to the test article was
consistently 20 ms and the peak pressure was approximately 20.7 MPa (3000
psi).

Figure 4 shows a pressure vs. time curve for test article SCUBA 1, cycle
31. The pressure shown was read at the test article interface. This curve
is typical of all pneumatic impact cycles performed on the three SCUBA
assemblies. Data recording began before opening the fast-opening valve,
when the test article pressure is ambient. Upon opening the fast-opening
valve, the test article pressure reaches its peak pressure and is he1d
for approximately l8 s before being vented back to ambient pressure.
Because the first-stage regulator allows flow-through up to its set
pressure, the upstream pressure bleeds through to the second-stage
regulators, creating a slow pressure rise after the initial impact. This
effect is more evident when only the impact phase of the cycle is plotted
(Figure 5). The initial impact on the first-stage regulator, shown as the
first pressure peak in Figure 5, consistently reached approximately 19.3
MPa (2800 psi) during each test, After the initial impact, the pressure
reached its maximum at 20.7 MPa (3000 psi) after approximately 0.8 s.

Posttest functional checks with GNp showed no anomalies in any of the
four SCUBA assemblies.

Each assembly allowed minimal leakage over the 5 min hold at 20.7 MPa
(3000 psi). Data sheets for each assembly recorded the pressure decay on
two gauges, the test console gauge standard and the first-stage regulator
gauge of the SCUBA assembly. The data sheets for SCUBAs 1 through 4 are
included Appendices A through D, respectively.

6.2 Clean Verification Tests

After the functional checks were completed, the SCUBA assemblies were
disassembled for the clean verification tests. The particulate level and
NVR quantity data for the piece parts of the first and second-stage
regulators and connecting hoses of each assembly were recorded in detail
on the data sheets in the Appendices. Table 2 summarizes the respective
particulate and NVR levels for each component of the four SCUBA
assemblies. The NVR levels recorded in Table 2 are calculated values
based on the quantity in milligrams (mg) of NVR weighed in the sample,
divided by the approximate surface area, in square feet (ft2), of the
component piece(s) flushed. The unit of measure, mg/ft2, was chosen to
allow easy comparison with data from previously reported literature. The
surface areas of the component pieces used for calculation are only rough
approximations and, in most cases, represent larger-than-actual surface
areas to give best-case, lowest contaminant levels.

The first-stage regulator of each of the test articles was found to be
highly contaminated in each of the four assembly parts, both in
particulate and NVR levels. In SCUBA 1, the first-stage regulator parts
each failed a Level 1000 particulate analysis, meaning that particles
larger than 1000 pm were discovered in each of these parts after flushing
(Table 2). The internal metal pieces in SCUBAs 3 and 4 had particulate
levels higher than what could be counted. The particles were described as
possibly "Krytox, plastics, metals, glass, and grit" (Appendices C and
D), In SCUBA 2, the internal metal pieces failed a Level 200 particle
count, meaning that at least one particle larger than 200 pm was found.
The analysis did not state if higher particle levels would have been
passed. The internal soft goods of each test article were flushed
together, and large particulate was also found. SCUBA 3 had a particulate
level higher than what could be counted. The inlet filter of SCUBA 2
passed a Level 300 particle level. No particles larger than 300 pm, and
an allowable distribution of smaller particles, were found in the filter
after back-flushing.

NVR levels in the first-stage regulators were also excessive. The
internal metal pieces of SCUBA 4 showed the lowest NUR quantity of the
four SCUBA assemblies analyzed, at 49.0 mg/ft2 (Table 2). The body cavity
NVR data for SCUBA 1 were not listed in Table 2 because of an error
discovered during the sampling process which invalidated the NVR
quantity, and the body cavity NVR data for SCUBA 3 were not analyzed.
However, the YVR quantities for this part in SCUBAs 3 and 4 were very
similar, at 102.7 mg/ft2 and l 18.7 mg/A~, respectively. The inlet filter
of SCUBA 1 was back flushed, and a large amount of NVR was discovered,
equaling 122.2 mg/ft2 Higher contamination levels in the filter were
expected to a certain extent, because the filter is exposed to the
outside environment during connecting and disconnecting the SCUBA
assembly to its gas source. For SCUBA 2, the NVR quantity was
significantly lower, at 24.4 mg/ft2.

The second-stage regulators of each test article were also found to be
highly contaminated in each of their assembly parts. Particulate levels
found on the internal metal pieces of each second-stage regulator failed
level 750 and above in each test article. The particles were described as
"mostly metal and fiber" (Appendices). Particulate analysis of the soft
goods of each second-stage regulator indicated that one sample failed
Level 300 and one failed Level 500, but the rest failed Level 1000 or
higher. Similar high levels of particulate were found in sampling the
second-stage regulator bodies. The particulate in one second-stage
regulator body was described as "mostly clear and metal chunks and some
clear fibers" (Appendix C). Some particulate samples from the flexible
hoses were less contaminated. One flexible hose sample passed Level 200,
and one passed a level 500. Most other hose samples, though, contained
large particles. The particulate was characterized as "mostly metal"
(Appendix C).

NVR levels were taken only for the internal metal pieces of the
second-stage regulators. These ranged from 150.7 mg/ft2 in Scuba 03 to
469.2 mg/ft2 in Scuba K. NVR levels were not taken for the second-stage
regulator bodies because it was difficult to separate the polymer pieces
from the metals.

 

 

7.0 Conclusions

No ignitions were sustained by the three SCUBA assemblies tested by
pneumatic impact in a 50 percent oxygen NITROX mixture at 20.7 MPa (3000
psi). Even test article SCUBA 2, which did not have a NITROX-compatible
soft good kit, did not sustain any ignitions. Functional tests performed
after the completion of the pneumatic impact tests confirmed that each
SCUBA assembly held pressure. The clean verification tests showed gross
amounts of particulate and NVR in most component parts. Despite these
high levels of contamination, the pneumatic impact tests demonstrated
that the component design is robust enough to withstand ignition under
worst-case NITROX operating conditions. It should be noted, however, that
the results of this testing do not negate the need to pre-clean systems
for oxygen service or the need to maintain these systems clean during
use. Prudent knowledge and practice of oxygen system safety is
recommended for all oxygen systems, even for those operating in
50-percent NITROX environments.

 

8.0 Recommendations

It is recommended that each SCUBA assembly be initially cleaned and then
maintained clean to an acceptable level during annual disassembly and
cleaning and during intermediate maintenance procedures. The acceptable
level will be approved by JSC QARSO and Materials Processing personnel.


Best regards --

Bill

--
Send mail for the `techdiver' mailing list to `techdiver@aquanaut.com'.
Send subscribe/unsubscribe requests to `techdiver-request@aquanaut.com'.

--
Send mail for the `techdiver' mailing list to `techdiver@aquanaut.com'.
Send subscribe/unsubscribe requests to `techdiver-request@aquanaut.com'.

Navigate by Author: [Previous] [Next] [Author Search Index]
Navigate by Subject: [Previous] [Next] [Subject Search Index]

[Send Reply] [Send Message with New Topic]

[Search Selection] [Mailing List Home] [Home]