AIAA 91-1978

A Parametric Study of Factors Affecting the

Corrosion of Stainless Steel by MON Oxidiser

B. Melior and C.L. Smith

Royal Ordnance plc
Westcott, England

I. Kalsch
European Space Research and Technology Centre,
Noordwijk, The Netherlands

I AIAA/SAEI A.S ME
27th Joint Propulsion Conference
June 24-26, 1991 ISacramento, CA

for permission to copy or republish, contact

American Institute of Aeronautics and Astronautics
370 L'Enfant Promenade, S.w., Washington, D.C. 20024-2518
 A PARAMETRIC STUDY-OF FACTORS AFFECTING THE CORROSION OF
STAINLESS STEEL BY MON OXIDISER

B. Mellor1 and C. L. Smith-

Royal Ordnance, Westcott, England

European Space Research and Technology Centre, Noordwijk, The Netherlands

ABSTRACT

The interaction of nitrogen tetroxide oxidiser (as Any assessment of the risk of Flow Decay requires
NTO or MON) with stainless steel (CRES) has been data on two aspects of the chemical interaction
the subject of several previous studies. However, between CRES and NTO:
most of the investigations performed to date have
tended to be project specific, with very limited • Those factors which influence the RATE of
objectives. Consequently, even after thirty years use, corrosion (i.e. the rate at which iron dissolves
the CRES-NTO corrosion database has proved to be from the CRES into the NTO),
inadequate, as shown in the assessments made prior
to the launch of ESA's OLYMPUS spacecraft and • Those factors which affect the SOLUBILITY
NASA's GALILEO probe. This paper presents the limit of the iron corrosion product in the NTO,
results of an extensive parametric study of CRES- (i.e. the iron concentrations which can be
NTO corrosion, intended to provide a reliable and tolerated before precipitation may occur).
comprehensive database to support spacecraft design
and operation. Data has been obtained showing the During the mid-1980's two major space projects
effects of propellant composition (~O, NO, Cl), prompted earnest attempts to use the data then available
steel specification, temperature, SN ratio, and surface to assess the risk of Flow Decay. In Europe, the
passivation processes on corrosion rates. This data imminent launch of the large OLYMPUS
provides a basis for the optimised use of NTO in communications spacecraft was the motivator. In the
propulsion systems utilising CRES in their USA, the mission of NASA's interplanetary spacecraft
construction. GALILEO was causing concern. The existing
database for CRES-NTO interaction proved to be
inadequate; best-case and worst-case assessments
1. Introduction could be made to show that Flow Decay was either
very unlikely, or very likely. An analysis of the risk
Nitrogen tetroxide (NTO) gained acceptance as a to GALILEO performed at the US Jet Propulsion
high performance storable liquid oxidiser in the early Laboratory concluded, "there is not currently enough
1960's. By the 1980's, the requirements for a storable specific, quantitative data available with which to
oxidiser for long-life orbital and interplanetary develop a sound Flow Decay avoidance strategy"
spacecraft were being met entirely by NTO. Despite (Ref. 1).
its extensive and generally successful utilisation,
there has always remained an element of risk involved As a consequence of the experience gained on the
in its use due to corrosion-related problems. The OLYMPUS project, ESA's Space Technology and
reaction between NTO and stainless steel (CRES) Research Centre (ESTEC) initiated a programme to
results in dissolution of trace quantities of iron (Fe) establish a reliable database for CRES- NTO corrosion
from the alloy and the formation of low solubility chemistry. This study is being performed by Royal,
corrosion products. These can precipitate from the Ordnance and some of the results gained in this work
liquid and deposit inside oxidiser flow passages, are presented in this paper.
restricting the flow of the propellant. This effect is
usually referred to as "Flow Decay".

1 Head, Liquid Propellant R & D, Senior Member AIAA 2 Senior Propellant Chemist
3 Project Officer, Chemical Propulsion Section

Copyright © 1991 Royal Ordnance plc

Published by the American Institute of Aeronautics and Astronautics with permission

1
 2. Structure of the NTO Study been of very short duration, lasting no more than 2-
14 days. Such data is of very limited use in predicting
The overall programme was divided into three key propellant behaviour over spacecraft lifetimes of
activities related to CRES- NTO corrosion chemistry: 3,000-5,000 days.

• A Literature Review The review concluded that no parametric study of

NTO corrosion chemistry had ever been undertaken
• Formulation of an Experimental Test Plan and and that this was considered to be the only effective
Test Procedures approach to generating a useful database.

• The Experimental Test Programme On completion of the literature review two critical
requirements for the new study were established:
This sequence of work was considered to be
particularly important; NTO chemistry had been (i) an experimental test matrix was required
investigated for thirty years and the database was still which would enable the effects of several key
inadequate. Consequently, an independent Literature parameters to be quantified,
Review was considered essential in order:
and
(i) to identify those areas where data was lacking,
or unreliable (ii) the type of test performed would have to be
capable of generating experimental data which
and would be directly applicable to spacecraft
operation.
(ii) to appraise critically the experimental
methods which could be used in the study of The technical approach to these tasks is described
CRES-NTO interaction. below.

3. Summary of the Literature Review 4. Experimental Test Plan

Approximately seventy papers concerning every 4.1 General

aspect of NTO propellant chemistry were reviewed.
The existing database was found to be very much a The rationale on which the Test Plan was formulated
'patchwork' body which lacked coherency. Much of may be summarised as follows:
the available data had been obtained for specific
projects and could not easily be applied more • As a very generalised view, it seems probable
generally. Several parameters were known (or that Flow Decay is most likely to occur when
believed) to influence the rate of attack of NTO on the iron content of the oxidiser reaches the
steels. However, few studies had been performed to solubility limit, i.e. the propellant becomes
quantify the effects of any one parameter in controlled saturated with iron (Ref. 2).
tests. For example it might be possible to find
corrosion rate data at different temperatures, but such • In order to be able to predict the onset of Flow
data might come from different studies, and usually Decay, data is required on both the RATE at
some of the secondary parameters (e.g. propellant which iron dissolves from the steel into the
composition, steel type) varied from study to study. propellant and the SOLUBILITY limit of the
Consequently, it was rarely possible to make iron.
deductions about the effects of any one parameter,
without also making some assumptions or allowances The experimental study therefore included tests to
for the effects of some other parameterts). In some investigate the various parameters which were
instances, direct comparative tests had been believed to influence corrosion rates and corrosion
performed, but it was often the case that these had product solubility (only the corrosion rate work is
described in this paper).

2
4.2 Factors Affecting Corrosion Rates very carefully assessed and a numerical procedure
was developed to make a correction and normalise
As a result of the Literature Review, the following the measured iron content of the oxidiser to that
factors were identified as those most likely to influence produced by a constant SN ratio. All of the SN
corrosion rates: ratios quoted in the test matrix are based on geometric
surface areas rather than true surface areas (the true
PROPELLANT COMPOSITION surface area of the steel meshes is much greater than
I the geometric area).
I I I I
Water Nitric Oxide Chloride Iron
The oxidiser used in the tests was prepared in the
content content content content
laboratory by distillation through molecular sieves
(MONl/MON3)
and was subsequently stored in titanium alloy vessels
prior to use. Consequently, the pre-test iron content
was always $0.1 ppm, (Le. the detectable limit). The
STEEL SPECIFICATION
I
baseline test conditions used throughout the
I I I programme involved the use of MON1 with a water
Type Fonn Surface Condition content of 0.04%.
(301.304L.etc) (sheet/mesh) (passivation)
In order to provide oxidiser containing the necessary
levels of minor constituents (HP, NO and Cl),
SERVICE CONDITIONS required by the test matrix, batches were produced by
I
I I I controlled additions as appropriate (note: chloride
Temperature Vapour/liquid Surface area! was added as chlorine).
phase corrosion Volume ratio
(SN) In the majority of the corrosion tests, the CRES
specimen was a piece of fine mesh material in type
Ten experimental Test Series were performed to 304L steel. Additional tests were also performed on
investigate these factors, each Test Series investigating sheet materials in both cryoformed and annealed
one specific variable over a range of values of practical conditions (see Annex B).
significance. The test matrix is summarised in Table
1. A number of control experiments were also The general analytical procedures for the analysis of
performed to assess the influence of the particul ar test the propellant used in this study are listed in Annex
methods on the results obtained. A and are the same as reported in previous work by
Royal Ordnance (Refs. 3 and 4).
4.3 Test Procedures

The corrosion rate tests were performed by immersing 5. Results

test specimens of the CRES alloys in oxidiser stored
in thick-walled glass tubes. The tubes were fitted 5.1 Effect of water content
with a closure and needle valve, both of which were
fabricated in a fluoropolymer. Periodically, small Five comparative tests were performed using oxidiser
samples (approx 3 ern") of the propellant were taken with water contents of 0.01, 0.02, 0.04, 0.08 and
from the tests via the valve and analysed for iron 0.15%. The results are shown in Figure 1 and
content, enabling a plot of iron content-v-time to be illustrate the direct relationship between the water
drawn. All of the tests were performed in duplicate. content and the rate of iron dissolution from the steel.
Consequently, the iron-v-time curves presented in
1.8
Section 5 consist of data points showing the mean + 0.15 wt% H,O
+
o 0.08 wt% H,O
values of the duplicate tests, and a 'best-fit' line has 1.5 V 0.04 wt% H,O
been drawn through the data, based on the geometric fj. 0.02 wt% H,O
o o.oi
expression y ax". = E
0.
0.
12
wt% H,O

-;;
i:; +
The tests were originally planned to continue for a 0.9
period of 250 days, with sampling after 10, 30,60, ~
::E
180 and 250 days. [Note: all of the tests have now 0.6
+
completed this duration and a number are being
v
maintained to 600 days]. 0.3

It was recognised that the removal of the oxidiser 0.0

100 200 300
samples would cause a reduction in the liquid volume TIME. days

and affect the SN ratio of the test. This effect was Figure 1 Effect of water content on corrosion rates of 304L in MON 1

3
 5.2 Effect of Nitric Oxide Content 5.4 Effect of Iron Content

Comparative tests were performed using oxidiser Two corrosion tests (A and B) were performed with
containing 1 and 3% nitric oxide (i.e, MON1 and MON1 having an initial iron concentration of <0.1
MON3). The results are shown in Figure 2 and show and 0.43 ppm respectively. The increase in iron
that the rate of iron dissolution was greatest in concentration over the test period of 50 days was 0.47
MONl. ppm in test A and 0.40 in test B (see Figure 4). The
differences are believed to be within experimental
0.4
o 1.04 wt% Nitric Oxide error, indicating that the initial iron content of the
A 2.99wt%NitrlcOJcide
oxidiser has little, if any, effect in subsequent corrosion
0.3
rate.
E
a.
a.
t. Experiment B
~~ 0.2
1.0 0 Experiment A

::.
A
t.
0.8

0.1 E
a.
~ 0.6
!;;.

~ o
0.0 ::. 0.4
n
100 200 300
TIME. days

Figure 2 Effect of Nitric Oxide on corrosion rates of J04L in MON 1 0.2

0.0
I,
o 10 20 30 40 50 60
TIME. days
5.3 Effect of Chloride Content
Figure 4 Effect of iron content on corrosion rates in MON1
Tests were performed using MON1 with chloride
levels of <0.01,0.02 and 0.04%. The results shown
in Figure 3 suggest that any variation in corrosion 5.5 Effect of Steel Specification (material
rates were within experimental error. specifications are given in Annex B)
1.0
o 0.040 wt% Cl-
Corrosion rate tests were performed on the two types
11 0.024 wt% ct-
n 0.005 wt% C1- of steel mesh (fine and coarse weaves) commonly
0.8
employed in the construction of surface tension
E
propellant management structures. As shown in
a. 0.6
11
a.
Figure 5, the fine mesh gave rise to only slightly
A
~ 0 0
higher levels than the coarse mesh.
~
::. 0.4 0

0.5
o 304L fine mesh
0.2 A 304L coarse mesh

0.4
0
0.0
0 100 200 300
E
TIME. days Et 0.3

Figure 3 Effect of Chloride on corrosion rates in MON 1 ~ 0

~
::. 0.2

0.1 A

0.0
100 200 300
TIME. days

Figure 5 J04L mesh corrosion rates

4
Corrosion tests were also performed on a number of The results are shown in Figure 7 and illustrate the
specimens of sheet steels, including grades 301, clear variations in iron dissolution rates resulting
304L and 321. Steel samples were provided for the from the different surface treatments.
tests by three major manufacturers of propellant
tanks. The results of this comparison are summarised 0.8
o As received
in Figure 6, which indicates the iron dissolution from 6 PassivateA
0.7 + Passivate 0
the best and worst materials, together with a 'mean' o Passivate C

result for the seven forms of sheet tested. 0.6

0
0.5 0
Ee,
<> Worst c,
1.0 t,. Mean " 0.4
o Best ::;
~ 0.3
::;;
0.8
e 0.2

E
g; 0.6 0.1

~
0.0
~
::;; 0.4
TIME. days

o o Figure 7 Effect of passivation procedure on corrosion rates in MON 1

0.2
o
o 0

0.0
100
TIME. days
200 300
The elecropolishedsurface showed optimum corrosion
resistance, giving rise to iron levels in the oxidiser of
Figure 6 Sheet steel best, worst and mean corrosion rates approximately one-fifth the values generated by the
non-passivated steel.
A detailed assessment of the complete data suggested
that the most important factor which influenced the 5.7 Effect of Temperature
spread of results was the surface passivation treatment
applied to the steel (see 5.6 below). Corrosion tests were performed at temperatures of 5,
25 and 35°C (41, 77 and 95°F). Figure 8 shows the
5.6 Effect of Surface Passivation measured iron levels, indicating the direct influence
of temperature on corrosion rate.
In order to obtain a direct comparison of the effects
of different passivation procedures on steel corrosion, 0.5
<> 35·C
comparative tests were performed using test A 25·C
05 ·C
specimens of one particular ~ of steel (i.e. type 0.4
321) but passivated using one of three respective
procedures, along with control specimens which E
g; 0.3
were not subjected to any form of passivation. The
three passivation treatments used were: ~
::;;0.2

(i) electropolishing
0.1

(ii) passivation in nitric acid

0.0
100 200 300
(iii) pickling in nitric/hydrofluoric acid, followed
by passivation in nitric acid.
Figure 8 Effect of temperature on corrosion rates of 304L in MON 1

5
 A plot of the propellant iron content (after 250 days)
as a function of temperature is shown in Figure 9, and
the relationship appears to be near-linear.

0.4

0.3 o

~ 0.2
o
3 4 6
0.1 SN Ratio. cm:'

Figure J J Iron levels at 250 days as a function of SIV ratio

0.0
10 20 30 40
Temperature °C
5.9 Comparison of Liquid and Vapour Phase
Figure 9 Iron levels at 250 days as a function of temperature Corrosion

Comparative tests were performed in which the steel

5.8 Effect of SN Ratio specimens were either, (a) immersed in liquidoxidiser
or, (b) suspended in the vapour above the liquid. The
Corrosion rate tests were performed at SN ratios of build-up of iron in the liquid in each case is shown in
004,2.0 and 7 cm:'. The limits of this range correspond Figure 12.
to the values that would be found in a small bore (6
mm o.d.) pipe, and that found in a partially full 0.6
o Liquid phase corrosion
spacecraft tank. The results are shown in Figures 10 I:!. Vapour phase corrosion
0.5
and 11 and show that the iron level present in the
oxidiser was directly related to the SN ratio of the o
0.4
test, i.e. higher SN ratios give rise to higher iron
contents. ~
4: 0.3

j 0.2
c SN - 7.0 cm-'
l'. SN - 2.0 cm-'
o SN - O.4cm-' 0.1

0.0
100 200 300
TIME. days

~ 2
Figure 12 Effect of liquid/vapour phase on corrosion rates in MON1

Iron levels in the experiment in which the specimen

was immersed were approximately 30% higher than
o
100 200 300 the levels in the vapour phase test.
TIME. days

5.10 Control Experiments

Figure J 0 Effect of SIV ratio on iron build-up from 304L in MON 1
Four test vessels were filled with MON at the start of
the tests, but without the addition of any steel corrosion
specimens. Throughout the duration of the test
programme, samples of MON were periodically
removed from these control tests and analysed for
iron content. As expected, the iron levels remained
at the same value as the start of the tests, i.e. ::;;0.1
ppm.

"
11
 6. Discussion days, especially if the oxidiser has a significant water
content. The use of steel for propulsion system
On completion of the 250 day study, it is believed that components giving rise to high SN ratio must be
the tests have provided a very clear indication of how approached with caution.
the various parameters affect the rate of attack of
NTO on CRES alloys. The comparison of corrosion rates in the liquid and
vapour phases gave results which showed that
The rate attack of NTO was markedly dependent on corrosion in the vapour phase occurred at only slightly
the water content of the oxidiser. The rate of attack lower rates than in the liquid oxidiser. It may be
was significantly higher in MON1 than MON3. important to note here that because of the way the
Neither the initial iron content of the MON nor the experiments were performed, the test vessels were
chloride level appeared to affect the rate of attack, at inverted on the six occasions when propellant samples
least within the conditions investigated in this study. were taken for analysis. At such times, the test
specimens in the "vapour phase" tests would have
The steel meshes gave rise to iron levels which were become immersed in propellant, and consequently
higher than steel sheet (for a given geometric area). the tests were not exclusively vapour phase tests.
This is a natural consequence of the fact that the true However, it must be remembered that in a spacecraft
surface area of the mesh (consisting of very many tank, micro-gravity conditions would result in
strands of fine wire) is greater than the true surface movement of the propellant around the tank (except
area of the sheet. within the propellant management device).
Consequently the tank's walls would be expected to
Of the different samples of steel sheet tested, the be alternatively in contact with both the vapour and
lowest and highest iron levels generated in the MON the liquid. As such, it is considered that the corrosion
after 250 days varied by a factor of - 3. This variation test performed was not unrealistic.
could not be correlated to the ~ of steel used (i.e.
301, 304L and 321), all of which appeared to be very Previous studies of corrosion rates of steels by NTO
close in performance when grouped together on the have often produced very scattered and erratic results.
basis of manufacturer. However, it is believed that In the present study, it was considered important to
the reason for the significant difference between the obtain a quantitative assessment of the accuracy and
best and worst materials was related to the method of reproducibility of the data. A total of sixty-eight
surface passivation. The best material had been corrosion tests were performed and within the matrix,
electropolished and the worst appeared to have been fourteen individual tests (parts of separate
aggressively pickled. This appraisal was supported experiments) were performed under identical
by the separate study in which one type of steel sheet conditions.
(unpassivated) was subjected to three respective
passivation procedures in our laboratories. The results The iron values measured in these tests after 250 days
(Figure 7) showed clear variations in subsequent are tabulated below:
compatibility of the material with the oxidiser, the
electropolished specimens being the most corrosion 0.42 0.30 0.35 0.37 0.37 0.45 0.37
resistant.
0.32 0.30 0.31 0.29 0.29 0.75 0.26
As expected, the tests at higher temperatures showed
higher rates of iron dissolution. This confirms that, as The mean value of the 14 tests was 0.37 ppm with a
far as is practicable, storage of the propellant at lower standard deviation of 0.12 ppm. It may be noted,
temperatures will help keep iron levels low. however, that the fourteen values include one
exceptionally high result which was believed to be
The tests investigating the effect of SN ratio clearly caused by contamination of one of the tests. If that
showed the expected result that high iron levels result value is excluded, the remaining thirteen tests provided
in conditions of high SN ratio. The solubility of iron a mean value of 0.34 ppm, with a standard deviation
in MON oxidiser has been previously reported (Ref. of 0.05 ppm. This level of accuracy has provided
5) to be in the range 1-5 ppm, depending on exact confidence in using the data to extrapolate the iron
conditions (see below). At high SN ratios (i.e. inside levels in propellant over typical spacecraft lifetimes,
small-bore pipes) the iron level in the oxidiser can i.e. 3,000-5,000 days. This work is still in progress
exceed the solubility limit within a period of <100 and is not reported here.

7
l~
In looking for an effective means of comparing the Work is also now in progress to utilise the corrosion
data obtained in this study with previous work, use rate data (extrapolated to spacecraft lifetimes) and
has been made of the assessment performed for the the solubility data, to assess the risk of Flow Decay
GALILEO project (Ref. 1). Figure 13 shows a under a range of conditions. If a general assumption
comparison of the upper and lower iron levels is made that Flow Decay is unlikely to occur until the
predicted in NTO by the JPL assessment and the propellant is saturated with iron, it is possible to use
experimental data obtained in the present study. the above data to calculate the "time-to-saturation",
i.e. the operational lifetime during which Flow Decay
8.0 would not be expected. The preliminary assessment
7.0 FSA Study: 0.15% 1-1,0 _ ..-
suggests that under optimum conditions, i.e. the use
..- of MON 3 with low water content, the use of effective
6.0
passivation procedures and the avoidance of CRES
5.0 materials for components giving rise to a high SN
Iron
level 4.0 ratio, MON oxidiser can be used safely over periods
(ppm)
--- far in excess of current spacecraft requirements (15
3.0
years). However, in the worst circumstances, i.e. the
2.0
converse of the above, precipitation of iron nitrate
1.0 from the oxidiser, and the risk of Flow Decay, might
begin within a few months.
30 60 90 120 150 180 210 240
llMEda)'l'

Figure 13 Comparison of JPL data audit and ESA experimental data 7. Conclusions
for NTO-CRES iron levels
1) The study has succeeded in establishing a
It can be seen that the upper and lower limits for iron wide-ranging database concerning CRES-NTO
levels are actually more divergent in this study than interaction.
theJPL limits. However, the current data enables for
the first time the rationalisation of the various factors 2) A number of factors have been shown to have
which cause the divergence: a particularly important a critical effect on the rate of attack of NTO on steel.
factor being the water content of the oxidiser. Consequently, it is clear that the risk of Flow Decay
may be reduced by:
As indicated in Section 4.1, on completion of the
corrosion rate studies we have now started to (a) the use of NTO with very low water contents
investigate the solubility behaviour of iron in both
MON1 and MON3. A limitation exists with the data (b) the use of MON3 in preference to MON1
previously available on solubility, that it does not
cover propellant compositions containing <0.08% (c) the use of acid pickling/passivation or
water (Ref. 5). Currently, most spacecraft grade electropolishing on steel surfaces.
NTO contains :5;0.04%water, and the new work will
investigate solubility in MON with water contents 3) The solubility limit of iron in NTO oxidiser is
ranging from 0.01-0.10%. of the order of 0.4-4.0 ppm. The corrosion rate data
shows that under best-case conditions (as implied by
The tests completed so far have given measured (2) above) iron levels will stay well below the
solubility valued which are much lower than those solubility limit for periods far in excess of spacecraft
predicted by extrapolation of the results of the existing lifetimes. In worst-case conditions, however, iron
data (in some cases, a factor of 10 lower). Values levels will exceed solubility limits within spacecraft
ranging from 0.4-4.0 ppm have been determined, lifetimes and the risk of Flow Decay would then
depending on the propellant composition ,temperature appear high.
and the pore size of the filter used to filter the
propellant. The variation of iron solubility is now
being determined over a test matrix based on these
parameters.

8
 8. Recommendations ANNEXES

1) For long-life, orbital (and interplanetary) Annex A - Summary of Analytical Methods

spacecraft, NTO in the form of MON3, and with
water content of <0.04%, should be used. A standard
specification for this 'High Purity' grade, should be VARlABLE METHOD
formulated and this specification adopt the use of Hp NIR Spectrophotometry
proven analytical methods for MON and its
constituents. NO UV -VIS Spectrophotometry

2) Optimum conditions for the passivation of Cl UV -VIS Spectrophotometry

steels should be established and accepted as standard Fe Atomic Absorption Spectrophotometry
procedures.

3) Although the corrosion database for steel-

NTO is now well established, comparable data is
now required for other alloys. Test work should be Annex B - Summary of Materials Tested
initiated on titanium and aluminium alloys.
Fine Mesh: 325x2300 TDDW weave in 304L steel

9. Acknowledgements Coarse Mesh: 165x800 TDDW weave in 304L steel

All of the work presented in this paper has been Cryofonned steels: types 301 and 304L, provided by
performed by Royal Ordnance as part of the ESA 2 manufactures, surface finishes
Study on Nitrogen Tetroxide Chemistry (ESTEC including acid pickled and
contract no. 7945/88). We are grateful to the Directors electropolished.
of ESTEC and Royal Ordnance for their permission
to publish this paper. Annealed steels: type 321

10. References

(1) T. W.Cole,R.H.FrisbeeandA.H. Yavrouian,

"Analysis of Flow Decay Potential on Galileo".
AIAA paper 87-2016, 23rd Joint Propulsion
Conference, San Diego (June/July 1987).

(2) 1. Kalsch and H. Keil, "Recent Investigations

on NTO Oxidiser Flow Decay AIAA Paper 85-
1297,AIAA/ASME/SAE/ASEE21stJointPropulsion
Conference, Monterey, California, July 8-10, 1985.

(3) B. Mellor, M. A. Moore and C. L. Smith,

"Factors Affecting the Corrosion Behaviour of 304L
Steel in MON Oxidiser". AIAA Paper 89-2951
AIAA/ASME/SAE/ASEE 25th Joint Propulsion
Conference, Monterey, Califomia, July 10-12,1989.

(4) B. Mellor and C. L. Smith, "Production and

Assessment of High-Purity MON Oxidizer, AIAA
Paper 90-2757, AIAA/ASME/SAE/ ASEE, 26th Joint
Propulsion Conference, Orlando, Florida, July 1990.

(5) 1. D. Smith, "Test Report, Iron Nitrate

Contamination in NP4". Document No. TR-248-
001 NASA-JSC/WSTF, Las Cruces, New Mexico,
30 July 1982.

9
 ~
CRES
Propellant ...,'"':::; 0
..;j
a:I..,.
Mean
Experimental Composition a:I
Measured
Corrosion '"'0 ~ S
Objective ~o ~ <.,) [Fe] ppm
Specimen Fe
NO% H2O% Cl% S U1 at 250 days
[nom] ~
Determine effect of 304L mesh 0.96 0.15 0.008 ~0.1 25 0.4 1.70
water content on 0.96 0.08 0.008 ~0.1 25 0.4 0.45
corrosion rates 0.96 0.04 0.008 ~0.1 25 0.4 0.37
0.96 0.02 0.008 ~0.1 25 0.4 0.30
0.96 0.01 0.008 ~0.1 25 0.4 0.22

Determine effect of 304L mesh 0.98 0.04 0.003 ~0.1 35 0.4 0.35
temperature on 0.98 0.04 0.003 ~0.1 25 0.4 0.30
corrosion rates 0.98 0.04 0.003 ~0.1 5 0.4 0.20

Comparison of 304L, fine mesh 0.92 0.04 0.005 ~0.1 25 0.4 0.33
corrosion rates of 304L, coarse mesh 0.92 0.04 0.005 :50.1 25 0.4 0.31
different types of 301, #1 sheet 0.92 0.04 0.005 ~0.1 25 1.2 0.88
steel in MON1 304L, #2 sheet 0.92 0.04 0.005 ~0.1 25 1.2 0.73
301, #3 sheet 0.92 0.04 0.005 ~0.1 25 1.2 0.36
304L, #4 sheet 0.92 0.04 0.005 ~0.1 25 1.2 0.48
301, #5 sheet 0.92 0.04 0.005 ~0.1 25 1.2 0.26
304L, #6 sheet 0.92 0.04 0.005 ~0.1 25 1.2 0.28
321, #7 sheet 0.92 0.04 0.005 ~0.1 25 1.2 0.34

Determine effect of 304L mesh 1.04 0.04 0.003 ~0.1 25 0.4 0.33
SN ratio on 1.04 0.04 0.003 :50.1 25 2.0 1.68
corrosion rate 1.04 0.04 0.003 ~0.1 25 7.0 3.35

Determine effect of 304L mesh 1.04 0.04 0.003 ~0.1 25 0.4 0.33
(NO) on corrosion 2.99 0.04 0.003 ~0.1 25 0.4 0.20
rate

Determine effect of 304L mesh 0.97 0.04 0.005 ~0.1 25 0.4 0.60
chloride on corrosion 0.97 0.04 0.024 ~0.1 25 0.4 0.53
rate 0.97 0.04 0.040 ~0.1 25 0.4 0.42

Comparison of effect 321 sheet

of passivation Control 1.00 0.04 0.005 :50.1 25 0.4 0.51
procedures on Passivated 1.00 0.04 0.005 ~0.1 25 0.4 0.40
corrosion rate Electropolished 1.00 0.04 0.005 ~0.1 25 0.4 0.11
Pickled & Passivated 1.00 0.04 0.005 ~0.1 25 0.4 0.29
Comparison of liquid 304L mesh liquid 1.02 0.04 0.005 ~O.l 25 0.4 0.31
& vapour phase 304L mesh vapour 1.02 0.04 0.005 ~0.1 25 0.4 0.24
I corrosion rates

Determine effect of 304L mesh 0.97 0.08 0.004 ~0.1 25 0.4

0.43 0.4
t
I
iron level on
corrosion rate
0.97 0.08 0.004 25 N/A

Control experiments 0.96 0.04 0.008 ~0.1 25

for corrosion rate 0.96 0.04 0.008 ~0.1 25
tests N/A 0.96 0.04 0.008 ~0.1 25 N/A N/A
0.96 0.04 0.008 :50.1 25

Table 1 Experimental parameters summary

10