| |
|
 |
|
 |
|
|
|
|
A second exposure
of the same metal samples reported in Testing Series
B was performed over a longer period of 42 days. The purpose again
was to measure the relative corrosion rate of the most commonly used
CorrView ® models in 1-1/2 in. and 3/4 in. ASTM 1018
steel, and commercially available ASTM 1010 and 1018 corrosion coupons
typically used in coupon racks to estimate pipe wall losses.
As previously
defined, a salt spray test booth was constructed. Design of the test booth
substantially followed ASTM Designation B 117 - 95: Standard Practice for
Operating Salt Spray (Fog)
Apparatus.
Coupons of ASTM 1010 and
1018 mild carbon steel, typically used in the corrosion rate evaluation of HVAC
piping systems, were purchased from Metal Samples. Samples of actual
ASTM A53, A106, and A795 pipe were also acquired on the open market, and
fabricated into the approximate length and width of the corrosion coupons.
Weights of the individual samples were not recorded, since an ultrasonic
measurement of true wall loss would be used as the basis for corrosion rate,
rather than weight loss.
 |
|
Side View Of Test
Assembly |
The wall
thickness of one end of each corrosion coupon and pipe sample coupon was
measured ultrasonically at 15 individual locations according to a standard grid
pattern of specific dimension. Both CorrView ® models
were similarly measured according to a specified grid at 15 individual
locations at its center. All wall thickness data was recorded into a
spreadsheet to establish a baseline.
A
glass enclosure was constructed to house the metal test specimens. All
specimens were positioned on a wooden base for isolation and placed into the
fog chamber. Specimens were physically arranged at an angle to the fog spray
according to the requirements of ASTM B 117 - 95.
A 5% salt spray
solution was introduced into the test chamber via ultrasonic atomization and an
overhead header configured to provide uniform and indirect introduction of the
salt fog. Our corrosion simulation procedure took exception to maintaining a
test chamber temperature of 95 ° F or to maintain pH, as the purpose of the
test was to make a side by side comparison of different metal types rather than
produce a standardized corrosion rate environment. All test subjects were
exposed to the same conditions during its entire test
period.
Testing was continued and the
results observed. On the first cycle of exposure at 42 days, the samples were
removed and allowed to dry, then photographed. Deposits from the measured end
of each test sample were then brushed free of rust deposits and rinsed.
Following the removal of this area of deposits, ultrasonic testing was
performed at the same area of each sample, and along the same grid pattern to
provide a second set of wall thickness measurements. Wall loss and corrosion
rates were calculated for each test specimen. Additional photographs were taken
of the cleaned test samples.
Results showed a
general agreement of corrosion rate statistics within what we would consider
reasonable limits. The salt spray environment produced an extremely aggressive
attack at all metal surfaces as evidenced by visual inspection of the exposed
metals, and of their underlying surfaces once deposits were removed (shown
below). Moderate to high pitting was found in all examples, requiring the use
of "echo to echo" ultrasonic measurement technique in order to
accurately measure the base dimension of such
areas.
Corrosion rates of between 31.2 MPY
and 61.8 MPY were measured; higher than those identified in
Testing Series
B and Testing Series
C. The highly corrosive environment produced by the salt spray
chamber would be expected to exaggerate minor differences in metal chemistry of
the steel samples, and therefore produce a wider variation in result than under
more typical conditions of a 1-3 MPY corrosion rate found at an open cooling
tower system. The differences noted between the steel pipe samples alone help
support this position.
We would have
expected lower corrosion rates for this longer sample exposure given that
virgin steel corrodes at an aggressive rate for its first initial few days, and
then declines to a more constant rate after some period of time. However, we
can also raise the possibility that variances in temperature of the test setup
between both series of tests, and possibly other different physical conditions,
may have contributed to this difference. More important than differences
between series of tests, however, is that the set of samples comprising each
test set were exposed to identical conditions and therefore offer a comparative
view of their respective corrosion
susceptibility.
Nevertheless, this series
of tests showed a maximum variance of only 36.5% from the average corrosion
rate measured, which we found reasonable. This highest variance in result was
again found at sample # 5 of ASTM A53 pipe. The average variance of all samples
tested from the mean corrosion rate was 22.5% - minor in comparison to the
large discrepancy often found between corrosion coupons and true corrosion
losses.
Test results for the standard
corrosion coupons of ASTM 1010 and ASTM 1018 mild steels showed similar rates
of corrosion to the actual pipe samples measured. Installed in externally
located corrosion coupon racks, however, their reported corrosion rates are
typically far below what actually exists at the interior pipe wall - the result
of their being isolated from most of the corrosive effects existing within the
actual piping system.
A corrosion rate
determination for a condenser water system of 0.5 MPY using coupons, when the
actual measured loss of pipe wall is 0.105 in. over 12 years or 8.75 MPY, is
a substantial 1,750 % difference or under reporting of actual wall loss.
Such extreme error in corrosion coupon rates, often found to exist between 100%
and 1,000% in actual UT testing comparisons, shows the current 36.5 % variance
in corrosion rate results of all samples tested to be nearly
insignificant.
The below graph well
illustrates the variation in wall loss for the nine test subjects evaluated
under this exposure period.
The below set of
tables offers a visual and statistical comparison of corrosion rates for the
two corrosion coupons, four pipe sample coupons, and two CorrView
® products tested. Shown at the left is the original metal
sample prior to testing, and in its original form. The center photograph is the
exposed and rusted sample as removed from the salt fog booth. The far right
photograph shows the same exposed sample after wire brushing.
Our testing produced an average corrosion
rate of 43.3 MPY at the five samples of actual steel pipe - ASTM A106, A795,
A135, and A53, and an average corrosion rate of all nine metals tested of 43.3
MPY. For purposes of this evaluation, we then compared the corrosion rates of
the commercially available steel corrosion coupons and CorrView
® products, finding an average 20.4 % and 9.8 % variance in
their corrosion rates from the true pipe samples respectively. Both the
corrosion coupons and CorrView products slightly over reported the
corrosion rates found at the steel samples during this test series - opposite
of what was found in Testing Series
B.
The percentage of variation
in corrosion rate from the average measured value of 43.3 MPY measured at the
five steel pipe samples is also provided below. Further testing is in progress,
and those additional results will be presented when available.
Metal Sample # 1 -
Corrosion Coupon 1010 Steel
Metal Sample # 2 -
Corrosion Coupon 1018 Steel
Metal Sample # 3 -
Steel Pipe ASTM A106
Metal Sample # 4 -
Steel Pipe ASTM A795
Metal Sample # 5 -
Steel Pipe ASTM A53
Metal Sample # 6 -
Steel Pipe ASTM A53
Metal Sample # 7 -
1-1/2 in. NPT CorrView Monitor
Metal Sample # 8 -
3/4 in. NPT CorrView Monitor
Additional
corrosion rate comparison testing is presented on this Internet site under the
above Testing
heading, and is currently underway. A duplicate of this testing procedure has
been commissioned to an independent laboratory. Results will be presented here
as soon as it is available. A third set of results from this group of test
samples will be available on approximately 10/1/04.
©
Copyright |
|
|
|
|
|
|
|
|
|
 |
|
|
