ISO/DTR 19852

ISO/DTR 19852:2025(en)
ISO /TC 2/SC 14
Secretariat: DIN
Date: 2025-09-25xx
The Neutral salt spray test — Results of an international
interlaboratory test and conclusions for practical application
Essai au brouillard salin neutre — Résultats d'un essai interlaboratoire international et conclusions pour une
application pratique
© ISO 2025
All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication
may be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying,
or posting on the internet or an intranet, without prior written permission. Permission can be requested from either
ISO at the address below or ISO’s member body in the country of the requester.
ISO copyright office
CP 401 • Ch. de Blandonnet 8
CH--1214 Vernier, Geneva
Phone: + 41 22 749 01 11
EmailE-mail: copyright@iso.org
Website: www.iso.org
Published in Switzerland
ii
Contents
Foreword . iv
Introduction . vi
1 Scope . 1
2 Normative references . 1
3 Terms and definitions . 1
4 Symbols and abbreviated terms . 1
4.1 Symbols . 1
4.2 Abbreviations . 2
5 Test samples . 2
6 Test procedure . 5
6.1 Normative operating parameters. 6
6.2 Corrosion test panels . 6
6.3 Assessment of the bolts subjected to salt spray testing . 7
6.4 Evaluation of normative operating parameters . 8
6.5 Evaluation of the findings for the bolts . 9
6.6 Calculation of proficiency testing of the laboratories . 10
7 Test results . 11
7.1 Normative operating parameters. 11
7.2 Determination of corrosivity . 18
7.3 Evaluation of the findings for the tested bolts . 24
7.4 Determining the precision data . 31
7.5 Laboratory proficiency assessment . 33
8 Summary of the conformity assessment . 33
9 Summary . 36
10 Conclusion . 37
Annex A (informative) Companies participating in the interlaboratory test . 39
Bibliography . 41

iii
Foreword
ISO
iv
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards
bodies (ISO member bodies). The work of preparing International Standards is normally carried out through
ISO technical committees. Each member body interested in a subject for which a technical committee has been
established has the right to be represented on that committee. International organizations, governmental and
non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the
International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
The procedures used to develop this document and those intended for its further maintenance are described
in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the different types of
ISO documents document should be noted. This document was drafted in accordance with the editorial rules
of the ISO/IEC Directives, Part 2 (see www.iso.org/directives).
Attention is drawnISO draws attention to the possibility that some of the elementsimplementation of this
document may beinvolve the subjectuse of (a) patent(s). ISO takes no position concerning the evidence,
validity or applicability of any claimed patent rights in respect thereof. As of the date of publication of this
document, ISO had not received notice of (a) patent(s) which may be required to implement this document.
However, implementers are cautioned that this may not represent the latest information, which may be
obtained from the patent database available at www.iso.org/patents. ISO . ISO shall not be held responsible
for identifying any or all such patent rights. Details of any patent rights identified during the development of
the document will be in the Introduction and/or on the ISO list of patent declarations received (see ).
Any trade name used in this document is information given for the convenience of users and does not
constitute an endorsement.
For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and expressions
related to conformity assessment, as well as information about ISO's adherence to the World Trade
Organization (WTO) principles in the Technical Barriers to Trade (TBT), see www.iso.org/iso/foreword.html.
This document was prepared by Technical Committee ISO/TC 2, Fasteners, Subcommittee SC 14, Surface
coatings.
Any feedback or questions on this document should be directed to the user’s national standards body. A
complete listing of these bodies can be found at www.iso.org/members.html.
v
Introduction
0.1 General
The neutral salt spray test (NSS test) has been used for at least one century for the corrosion evaluation of
various protective layers, including coatings for fasteners. This test is well-known and used worldwide by all
involved parties (chemical suppliers, job coaters, manufacturers of components including fasteners,
purchasers .), due to its relatively short testing time compatible with industrial production.
The NSS test is generally considered a suitable method for determining the effectiveness of corrosion
protection. The ability to expose samples and components quickly makes it possible to identify weaknesses,
pores and damages in organic and inorganic coatings. It is to be noted here that by the use of a salt spray with
a defined composition (5 % sodium chloride, NaCl) and by the material-related formation of protection layers
there is no direct correlation to other media capable of inducing corrosion. Instead, the results obtained using
this method serve to compare different surface conditions under specified and constant conditions. As such,
the salt spray test is an effective tool for the quality control of products exposed to corrosive operating
conditions.
In terms of the reproducibility of the test results, it is beneficial if the coatings and layers are sufficiently
similar. In addition, there are numerous manufacturers that offer different types of salt spray cabinets. The
various construction concepts available on the market, as well as factors such as the loading of the cabinet or
the use of samples with different geometries, influence the salt spray formation inside the cabinet. This makes
it difficult to conduct a comparative assessment of the corrosion test results obtained using different test
equipment. For this reason, a project group was established within the Working Group Surface Protection
Coating Systems of Deutscher Schraubenverband e. V. with the objective of compiling salt spray test results
from as many laboratories as possible and then preparing a comparative assessment.
The samples consisted of bolts that were electrolytically coated using two different zinc-based variants. A total
of 39 participants, consisting of bolt manufacturers, job coaters, users and institutes with 75 cabinets from
11 countries were involved in the tests, see 1Annex A. The large number of participants from all sectors
throughout the value chain of a coating process enables a reliable statistical analysis of the test results.
The organizers would like to thank all test participants for supporting this project with their extensive data
sets. We also thank the company Gevag for providing the bolts, the company Leist for coating the bolts, LISI
Automotive from France for providing hot-dip galvanized test panels, Artema France for measuring the
thickness of the test panels, and Frédéric Raulin (Coventya France) for supplying us with degreasing agents
for cleaning the panels. In particular, we would like to thank Jürgen Böttner, EJOT SE & Co. KG, for supporting
with the statistical evaluation of the test results.
0.10.2 Executive summary and conclusion
One objective of the interlaboratory test was to conduct a neutral salt spray test in accordance with ISO 9227
on two coating variants of bolts M6x50M6×50. Time of occurrence of gray veil, white rust and red rust was
documented. The corrosivity of the salt spray was determined by means of assessment of the mass loss of an
uncoated steel panel as specified in ISO 9227 as well as the determination of the time until appearance of red
rust on hot-dip galvanized steel panels in accordance with ISO 4042 and ISO 10683.
Another objective was to compare the two methods used to determine the corrosivity of the salt spray in order
to establish the suitability of these methods by comparing the corrosion assessment results obtained for the
bolts.
In addition, the normative operating parameters (temperature in the test cabinet, collection rate, pH and
density or NaCl concentration of the solution collected) were documented to ascertain whether there is any
correlation with the results of the corrosion assessment performed on the bolts.
vi
The main objective of the interlaboratory test was the determination of the reproducibility of the salt spray
test.
A statement has been included in the introduction toof the current version of ISO 9227:
“When interpreting test results (e.g. minimum time to damage or corrosion) for product quality control or
acceptance specifications, it is important to note that salt spray testing may have low reproducibility,
especially for manufacturing parts that are tested in different laboratories.”
These findings are supported by the interlaboratory study. The results are summarized as follows:
— — The mass losses of the standardized test panels as per ISO 9227 and the corrosion behaviorbehaviour
of standardized, hot-dip galvanized test panels as per ISO 4042 do not correlate with the corrosion
behavior observed on the zinc-electroplated and zinc-nickel electroplated bolts M6x50 with transparent
or black passivation that were tested in parallel.
— — There seems to be no clear connection between the criterion of “compliance with the normative test
parameters as per ISO 9227”, which is used to classify the corrosion cabinets as “compliant” or “non-
compliant”, and the assessed corrosion behaviorbehaviour of the test panels or the corrosion
behaviorbehaviour of the bolts tested. Accordingly, compliant operation in line with normative test
parameters does not lead to a reduced scatter of the times to failure recorded for the selected coated bolts
examined as part of this interlaboratory test.
Additional information about the context to understand these results can be found in 8Clause 8.
The authors of this report and the experts within ISO/TC 2/SC 14, Surface Coatings, concluded from the results
of this interlaboratory test that current methods to evaluate corrosivity of test chambers, e.g. ISO 9227, ISO
4042, are insufficient to ensure reproducibility of results and that the salt spray test in accordance with
ISO 9227 is therefore not suitable as an acceptance criterion.
Alternatives to salt spray testing, such as cyclic testing procedures, are established in the market but are
mostly customer specific: they need different testing equipment and/or dedicated settings for each type of
test (environmental cabinets also need specific skills and experts). Such cyclic tests are useful, however,
according to the experience of the committee members, do neither solve the question of correlation between
the cabinet parameter settings and the results on tested samples nor of observed scattering of the results.
It is the opinion of the experts within ISO/TC 2/SC 14 that salt spray testing in accordance with ISO 9227 and
ISO 4042 should still be applied for:
— — production process monitoring and verification for the coating process (but not for process control,
especially if based on a statistic approach)
— — comparison with different parts using the same coating and the same coating process.
Salt spray testing is however not exhaustive and advantageously accompanied by other tests specified in
relevant standards.
vii
The Neutral salt spray test — Results of an international
interlaboratory test and conclusions for practical application
1 Scope
The objective of this project wasdocument is to conduct a neutral salt spray test in accordance with ISO 9227.
The test is a proven method for assessing the corrosion protection of coatings of components such as bolts.
For this reason, two coating variants were chosen for conducting the tests on hexagon bolts with a size of
M6 × 50. The bolts were examined at specified points in time and the time of occurrence of grey veil, white
rust and red rust was documented. The corrosivity of the salt spray was determined by means of two methods
and also documented in an evaluation form. These two methods are the assessment of the mass loss of an
uncoated steel panel as specified in ISO 9227 as well as the determination of the time until appearance of red
rust on hot-dip galvanized steel panels in accordance with ISO 4042 and ISO 10683.
The processing steps wereare specified in detail in a manual that was provided to the test participants. If
followed precisely, these instructions allow for a comparative analysis of the results from the individual labs
and make it possible to determine the reproducibility of the salt spray test. Another objective wasis to compare
the two methods used to determine the corrosivity of the salt spray in order to establish the suitability of these
methods by comparing the corrosion assessment results obtained for the bolts. In addition, the normative
operating parameters (temperature in the test cabinet, collection rate, pH and density or NaCl concentration
of the solution collected) were documented for every inspection date in order to ascertain whether there is
any correlation with the results of the corrosion assessment performed on the bolts.
2 Normative references
There are no normative references in this document.
23 Terms and definitions
No terms and definitions are listed in this document.
ISO and IEC maintain terminology databases for use in standardization at the following addresses:
— — ISO Online browsing platform: available at https://www.iso.org/obp
— — IEC Electropedia: available at https://www.electropedia.org/
34 Symbols and abbreviated terms
3.14.1 Symbols
F(t) Weibull distribution function
b Shape parameter or Weibull slope
e Exponential function
k1 Factor for zU score
k2 Factor for zU score
m Hampel estimator
pH value Potentia hydrogenii
Ra value Arithmetic mean roughness
sR Reproducibility standard deviation
s Repeatability standard deviation
r
T Characteristic life time
t Time to failure
t0 Displacement parameter (failure-free time)
t Time to 10 % probability of failure
z score Score for the assessment of a laboratory mean value
U
3.24.2 Abbreviations
ASTM American Society for Testing and Materials
DIN German Institute for Standardization (Deutsches Institut für Normung)
DSV Deutscher Schraubenverband e. V. (German Fastener Association)
D0–D4 Scale for the degree of coverage
EN European Standard
G0–G4 Gray veil scale
ISO International Organization for Standardization
M Metric thread
n Lab/cabinet No.
NaCl Sodium chloride
XRF X-‑ray fluorescence spectroscopy
W0–W4 White rust scale
45 Test samples
The following test and reference samples were used, 0Figure 1 to 0Figure 4::
Corrosion test panels in accordance with ISO 4042 or ISO 10683:2018, Figure 1:
Material: CR24 as per ISO 6932
Dimensions: 190 mm × 90 mm
Surface condition: Hot-dip galvanized, (11 ± 1) μm
Corrosion test panels in accordance with ISO 9227:2022, Figure 2:
Material: CR4 as per ISO 3574
Dimensions: 150 mm × 70 mm
Surface condition: Uncoated, dull, degreased, roughness Ra = (0,8 ± 0,3) μm
Test bolts silver coloured, 0Figure 3::
Material: Quenched and tempered steel
Dimensions: M6 × 50
Surface condition: Zinc electroplated, thick-film passivated, sealed
Coating thickness: 9,3 µm (min. 9,0 µm, max. 9,6 µm)
Test bolts black coloured, 0Figure 4::
Material: Quenched and tempered steel
Dimensions: M6 × 50
Surface condition: Zinc-nickel electroplated, black passivated, sealed
Coating thickness: 9,0 µm (min. 8,2 µm, max. 9,6 µm)
The coating thickness of the test bolts was determined for 10 samples each at the point of force application on
the bolt head using X-‑ray fluorescence spectroscopy (XRF) in accordance with ISO 3497. The thickness of the
sealers was not determined. All test samples were shipped in shrink-wrap packaging in order to prevent any
prior corrosive damage and reduce the risk of impact damage to the films. The cleaning solution was also
shipped in powder form. The participants cleaned and masked the hot-dip galvanized panels themselves.
19852_ed1fig1.EPS
[SOURCE: Atotech Deutschland GmbH & Co. KG]
Figure 1 — Hot-dip galvanized corrosion test panel in accordance with ISO 4042
19852_ed1fig2.EPS
[SOURCE: TU Darmstadt]
Figure 2 — Corrosion test panel in accordance with ISO 9227
19852_ed1fig3.EPS
[SOURCE: TU Darmstadt]
Figure 3 — Silver coloured test bolts: zinc electroplated, thick-film passivated and sealed
19852_ed1fig4.EPS
[SOURCE: TU Darmstadt]
Figure 4 — Black coloured test bolts: zinc-nickel electroplated, black passivated and sealed
56 Test procedure
Prior to conducting the tests, the degree of coverage of the head and the non-threaded shank of the black bolts
was determined and documented in accordance with DIN 34804 on assessing the appearance of black surface
coating systems. From the 30 bolts supplied, 20 bolts were selected based on subjective evaluation, with the
coating surface as intact as possible. The normative operating parameters of the cabinet, the test panels and
the test bolts were inspected and assessed during the test and the results documented. If possible and unless
specified otherwise (see 7.26.2),), the operating parameters and the bolts were inspected every 24 h. This
inspection could be skipped on weekends and public holidays. The following chapters provide a detailed
overview of the inspection intervals. The results were entered in the report table provided by the test
organizers. The participants made sure that the cabinet was not left open for more than 1 hourh during the
inspections. The corrosion test panels and the test bolts were inserted into the cabinet at an angle of 20° to
the vertical axis. The use of suitable holding devices was not mandatory, but recommended. A possible
positioning of the collectors and the test panels and test bolts can be seen in 0Figure 5.
19852_ed1fig5.EPS
[SOURCE: TU Darmstadt]
Figure 5 — Example of a 1 000 Litersl cabinet equipped with plastic panels with M6 threaded holes
and bolts at the beginning of the test
The test panels were positioned in the same area of the salt spray cabinet as the test bolts. In addition, the
other specifications of ISO 9227 applied in order to ensure a freely circulating salt spray in the cabinet. This
also included observing the specified minimum distance to the wall. The samples were not positioned directly
in line of the spray jet and did not shield one another so that the salt solution film on the surface could not drip
from one sample onto another. Furthermore, only the samples to be tested were positioned freely in the salt
spray cabinet, if possible. To accommodate the participants’ daily operating processes, however, it was also
acceptable to insert other samples that were not part of the interlaboratory test into the test cabinet.
5.16.1 Normative operating parameters
The documentation of the normative operating parameters to included the temperature in the test cabinet,
the amount of salt solution accumulated in the two collectors as well as the pH value, density or NaCl
concentration of the solution. In addition, the parameters of the test solution (conductivity of the water used,
pH value, density or NaCl concentration) were recorded and entered into the report table prior to using the
solution in the cabinet. This also applied to all test solutions replenished into the salt spray cabinet over the
entire course of the test.
5.26.2 Corrosion test panels
5.2.16.2.1 Evaluation in accordance with ISO 4042
At the beginning of the interlaboratory test, the hot-dip galvanized test panels were subjected to salt spray
testing for 120 h in accordance with ISO 4042. Prior to testing, the panels were cleaned using the provided
cleaning agent in line with the test instructions and then masked, 0Figure 1. Finally, the panels were
positioned in the cabinet at an angle of 20°, see 0Figure 5,, and the test started within 24 h.
The first assessment of the panels took place after 72 h. The cabinet remained closed during the first 72 h, i.e.,
the operating parameters (see 7.16.1)) were not recorded after 24 h and 48 h and the bolts were not checked.
At the inspection intervals (after 72 h, 96 h and 120 h), the panels were assessed in a wet, unrinsed condition.
To this end, the delivered control mask was placed on the panel, 0Figure 6,, and all boxes with visible red rust
were counted and documented in relation to the total number of boxes. The time in the assessment at which
at least 7 boxes (5 %) show visible signs of red rust was also documented in the report table. If fewer than
7 boxes were counted after 120 h, the value “> 120 h” was entered in the table.
19852_ed1fig6.EPS
[SOURCE: Atotech Deutschland GmbH & Co.KG]
Figure 6 — Control mask on hot-dip galvanized panel for red rust evaluation
After a test duration of 120 h, the test panels were removed from the salt spray cabinet and testing of the bolts
continued.
5.2.26.2.2 Evaluation in accordance with ISO 9227
The reference panels in accordance with ISO 9227 were delivered in a clean condition. They were marked on
the back and then weighed (accuracy of ±1 mg). After masking the back of the panels, they were inserted into
the salt spray cabinet and put into the same positions as the hot-dip galvanized panels tested before (see
7.2.16.2.1).). The cabinet was then to be operated for 48 h without opening.
After 48 h, the test panels were removed, cleaned in accordance with the test instructions and pickled with a
diammonium (hydrogen) citrate solution. The panels were left to dry in air, and then weighed again to an
accuracy of ±1 mg. The weight was recorded in the report table.
Furthermore, the test participants were entered the results of their most recent internal corrosion test on
plain steel reference panels in accordance with ISO 9227 that was performed prior to the interlaboratory test.
5.36.3 Assessment of the bolts subjected to salt spray testing
Prior to conducting the test, the 20 bolts with the best surface finish as determined per visual inspection were
selected from the 30 bolts supplied for each variant. For the black bolts, the homogeneity of the black aspect
was determined. The homogeneity of the bolts was assessed (K0 to K4 in accordance with DIN 34804) and the
respective number of bolts was entered into the report table.
The bolts were positioned in compliance with the clearances (to walls and other test samples) stipulated in
ISO 9227. The bolts were then be screwed into the corresponding plastic panels, which were positioned such
that the bolt heads were at an angle of approx. ~20° to the vertical, 0Figure 5. The cabinet was operated
continuously (also during weekends and public holidays). The joint visual assessment and documentation of
the normative operating parameters was conducted at the relevant inspection interval. The maximum opening
time of the cabinet was 1 hour per day. The visual inspection for black spots, grey veil, white and red rust or
other anomalies/discolorations was performed on the bolt head as well as on the non-threaded shank. For
this purpose, the bolts first were rinsed carefully with water and dried in a stream of clean air.
In addition, the black bolts were classified in terms of grey veil (G0 to G4) and white rust (W0 to W4) as per
DIN 34804. The number of bolts with relevant anomalies were entered into the report table at the respective
inspection interval.
Bolts showing signs of red rust were be removed from the salt spray cabinet. The test concluded after 1 008 h.
The test duration was extended if no signs of red rust were observed on at least 2 silver bolts, or if no signs of
white rust were identified on at least 2 black bolts.
5.46.4 Evaluation of normative operating parameters
The normative operating parameters were visualized for all labs in the form of box plots, 0Figure 7. This form
of graphic representation makes it possible to illustrate the scatter of values over the test duration for each
lab and to identify any data that does not comply with the specification. The plot consists of a box, with a
horizontal line inside the box indicating the median of the data set. The boundaries of the box depict the upper
and lower quartiles. The lower and upper whiskers, both plotted in the form an error bar, indicate the
minimum and maximum respectively. This means that the lowest 25 % of the measured values for a parameter
are in the interval from the whisker end to the lower quartile, the next 25 % are in the interval between the
lower quartile and the median, and so on. The box depicts the middle 50 % of measured values, while the
interval between the lower to the upper whisker end is also referred to as the range. The arithmetic mean
value is indicated with a diamond symbol. For symmetric distributions, the arithmetic mean value is close to
the median, while asymmetric distributions can result in greater deviations.
19852_ed1fig7.EPS
Key
1 upper whisker
2 upper quartile
3 arithmetic mean
4 median
5 lower quartile
6 lower whisker
1)
Figure 7 — Box-and-whisker plot
5.56.5 Evaluation of the findings for the bolts
For the silver bolts, the number of bolts with discolorations or visible signs of grey veil, black spots and white
or red rust were documented in the report table for each inspection interval. For the black bolts, the
percentage of grey veil or white rust on the total surface was also evaluated and documented. In line with
DIN 34804 for assessing the appearance of black surface coating systems, five grades (G0 to G4, and W0 to
W4) were selected for this purpose. For the red rust criterion, only the number of bolts with visible signs of
red rust at the relevant inspection interval were documented. Due to deviations when entering data into the
report table, it was decided not to conduct a statistical analysis of the probability of failure for all individual
grades of the black bolts. Only the occurrence of grey veil and white rust was taken into account, regardless of
its extent.
The statistical analysis of the findings was limited to calculating the time for a 10 % probability of occurrence
of a specific anomaly on the bolts examined. The time of occurrence of an anomaly was defined as an input
value for each of the 20 bolts and taken into account in determining the probability of failure. Due to the
intervals between the individual inspection interval, half the duration was used as the input value if an
anomaly was observed. This is due to the fact that the exact time at which the anomaly occurred could not be
verified due to (at least) 24 hoursh interval between two inspections. In the next step, the individual
probabilities of failure were plotted as a 2-‑parameter Weibull distribution function F(t) (0(Equation (1),,
taking into account the characteristic life time T as well as the documented failure times t. The values were
entered in a Weibull plot, 0Figure 8. The initial displacement parameter t0 (failure-free time) of the
3-‑parameter Weibull distribution function was set to 0.
(1)
19852_ed1fig8.EPS
𝑡−𝑡
0 𝑏
−[ ]
𝑇−𝑡
𝐹(𝑡)=1−𝑒 (1)
Key
1)
See Hedderich, Jürgen; Sachs, Lothar: Angewandte Statistik: Methodensammlung mit R. Berlin Heidelberg New York:
Springer-Verlag, 2018.
X time till occurrence of white rust (silver-coloured bolts) t [h]
WR
Y1 linear transformed frequency sum ln[-[‑ln(1-F)]
Y2 frequency sum F [%]
[SOURCE: Jürgen Böttner, EJOT SE & Co. KG]
Figure 8 — Example of a Weibull plot grid with regression lines
In the final step, a logarithmic regression calculation of the data points was performed in order to determine
the time t relating to the 10 % probability of failure. This value was used in the comparative analysis of the
results. The slope b (shape parameter) of the Weibull distribution function was also determined through the
regression function and used in the evaluation. This parameter serves to approximate the function to other
distribution patterns.
5.66.6 Calculation of proficiency testing of the laboratories
The evaluation procedure described in ISO 13528 is based on robust statistical methods that include
numerical calculation processes. This approach has the advantage that even some outliers only have a minor
impact on the standard deviation and the median value.
The approach is briefly described in the following. As a general rule, the number of laboratories and the
number of values for a given parameter can vary.
The reproducibility standard variation, sR, was calculated in the first step. This is based on the difference
between the values recorded in one lab and all other values recorded in all other labs. Each pair of values was
only counted once. Differences between the values recorded in one lab were not taken into account.
In the second step, the repeatability standard deviation, s , was calculated. This value is based on the
r
differences between the values recorded in a single lab. Each pair of values was only counted once. Differences
between the labs were not taken into account.
Sophisticated algorithms (here: the Q method) assign outliers less statistical weight for s and s . If there are
R r
outliers, the normal standard deviation is larger than sR and sr.
In a third step, the arithmetic mean values were calculated for every lab.
The fourth step consisted of determining a robust mean value, also referred to as Hampel estimator m for all
labs. All lab mean values that deviate substantially from the robust mean value were given less weight. The
reproducibility standard deviation, sR, was used to assess the deviation. Values between ±1,5sR were weighted
fully. The weighting was reduced continuously to zero for ±4,5s and higher.
R
In the fifth step, the z scores were calculated for each laboratory. To this end, a “true” value and a “true”
standard deviation were determined. The “true” value (usually m) and the “true” standard deviation (usually
sR) were used to calculate the lower and upper tolerance limit. The lower tolerance limit was defined as
(m – k · s ) and the upper tolerance limit as (m + k · s ).
1 R 2 R
The z score is the difference between the laboratory mean value and the “true” value, divided by the “true”
standard deviation. Unless specified otherwise, the Hampel estimator m and the reproducibility standard
deviation s were used as the best approximation to the ”true” value and the ”true” standard deviation,
R
respectively. However, if the majority of the labs produced incorrect values, the few labs with correct values
would have failed the proficiency test. For this reason, if a “true” value was known in advance (e.g.,. from
certified reference material), it could be used for the assessment. If s was unusually high (e.g.,. too many
R
outliers, several outlier labs, large differences between the labs, insufficient number of participating labs,
strong scatter of individual values), a common standard deviation for the method could be selected for the
assessment in advance.
In a sixth step, the z scores for all values that can only be positive due to their nature were converted to z
U
scores by using the factors k < 1 (for negative scores) and k > 1 (for positive scores). These depend on the
1 2
relative standard deviation.
The z and z scores were interpreted as follows: If their absolute is greater than 2, the statement “the lab did
U
not work correctly” can be made with a confidence level of 95 %.
For this reason, laboratory proficiency tests were usually given the following assessments:
Labs with |z | ≤ 2 have passed the proficiency test.
U
Labs with |z | > 2 have failed the proficiency test.
U
NOTE In general, the clear statement “failed” is usually given for |z | > 3, and the zone 2 < |z | ≤ 3 is considered as a
U U
“grey” zone in which a clear statement is difficult to give. Here, the “grey” zone is already considered as “failed” as the
test is not clearly “passed”.
67 Test results
6.17.1 Normative operating parameters
The normative operating parameters include the amount of collected salt solution as well as its pH value and
density (see 6.1). These parameters were assessed and compared by means of box plots (see 7.46.4).). A test
duration of 1 008 h was used in the evaluation.
6.1.17.1.1 Collected test solution
The box plot diagram in 0Figure 9 shows the amount of test solution collected in all labs as well as the limit
values of (1,5 ± 0,5) ml/h (relating to a collecting area of 78,5 cm , which is derived from the collector
diameter of 100 mm) as specified in ISO 9227.
Only one value each was submitted for cabinets No 67 and No 72, which is an indication that the test solutions
of the two collectors were mixed. This approach means that deviations of the normative operating parameters,
which are caused by the different collecting points within the salt spray cabinet, are not visible. No fluctuations
in the values stated were observed for cabinets No 19 and No 51. In these cases, the values are all identical
over a period of 6 weeks, which is considered unlikely. Four values were submitted for cabinet 67. There is no
data available for cabinet 31. For the remaining cabinets, two values for the collection rate were submitted in
line with the specifications.
Taking into account the statistical distribution, 45 test cabinets are in compliance with ISO 9227 and,
ASTM B117 and 0-18, Figure 10. A total of 30 cabinets do not fully comply with these standards, see
Commented [eXtyles1]: eXtyles Inline Standards Citation
Match has detected that the standard reference
0Figure 11.
"ASTM B117-18, Figure 10" refers a specific part of an
undated standard. Because part numbers may change between
editions, please check the part number for accuracy or change
to a dated reference.
19852_ed1fig9.EPS
Key
Y collection rate (ml/h)
19852_ed1fig9_key1.EPS lower tolerance limit
19852_ed1fig9_key2.EPS
specified value
19852_ed1fig9_key3.EPS
upper tolerance limit
Figure 9 — Documentation of the test solution collected in all cabinets, including the tolerance range
defined in ISO 9227 and ASTM B117-18
19852_ed1fig10.EPS
Key
Y collection rate (ml/h)
19852_ed1fig9_key1.EPS upper tolerance limit
19852_ed1fig9_key2.EPS
specified value
19852_ed1fig9_key3.EPS
upper tolerance limit
Figure 10 — Documentation of the test solution collected in all cabinets for which compliant values
were submitted
19852_ed1fig11.EPS
Key
Y collection rate (ml/h)
19852_ed1fig9_key1.EPS lower tolerance limit
19852_ed1fig9_key2.EPS
specified value
19852_ed1fig9_key3.EPS
upper tolerance limit
Figure 11 — Documentation of the test solution collected in all cabinets for which non-compliant
values were submitted
6.1.27.1.2 pH value
ISO 9227 and ASTM B117-18 specify a pH value of 6,5 to 7,2 for the collected test solution. 0Figure 12 shows
the evaluation for all 75 test cabinets. 61 cabinets are within these specifications, while 14 do not fully comply,
as shown in 0Figure 13 and 0Figure 14. No data is available for cabinets No 1 and No 31. For cabinets No 46
and No 68, all values submitted are identical and do not exhibit any scattering over 6 weeks, which is
considered unlikely.
19852_ed1fig12.EPS
Key
Y pH value
19852_ed1fig12_key1.EPS lower tolerance limit
19852_ed1fig12_key2.EPS
upper tolerance limit
Figure 12 — Documentation of the pH value of all cabinets, including the tolerance range defined in
ISO 9227 and ASTM B117-18
19852_ed1fig13.EPS
Key
Y pH value
19852_ed1fig12_key1.EPS lower tolerance limit
19852_ed1fig12_key2.EPS
upper tolerance limit
Figure 13 — Documentation of the pH value in all cabinets for which compliant values were
submitted
19852_ed1fig14.EPS
Key
Y pH value
19852_ed1fig12_key1.EPS lower tolerance limit
19852_ed1fig12_key2.EPS
upper tolerance limit
Figure 14 — Documentation of the pH value in all cabinets for which non-compliant values were
submitted
6.1.37.1.3 Density and concentration of the test solution collected
In order to determine the NaCl content in the collected test solution, it was possible to enter the density or
concentration of the solution. 42 participants entered density values (specification as per ISO 9227:
3 3 3 3
1,029 g/cm to 1,036 g/cm ; specification as per ASTM B117-18: 1,025 5 g/cm to 1,040 g/cm ).
19 participants submitted concentration values in g/l (specification as per ISO 9227: (50 ± 5) g/l), while
8 participants stated the concentration values in percentage by weight (specification as per ASTM B117-18:
(5 ± 1) percentage per weight). No density data was submitted for 6 cabinets.
A comparison chart with statistical cabinet data including the density values provided is shown in 0Figure 15.
The data for 29 cabinets is in accordance with ISO 9227. A total of 9 cabinets do not fully comply with these
specifications. However, these cabinets are compliant to ASTM B117-18. The values submitted for 4 cabinets
do not comply with either ISO 9227 or ASTM B117-18. The values provided for cabinets 21 and 46 are
identical and without scattering, which is considered unlikely.
The results for the concentration values in g/l are presented in 0Figure 16. The values of 16 cabinets meet the
specifications of ISO 9227. The values of 3 cabinets are partially outside the tolerance range. However, after
converting the value from g/l to g/kg, as stated in ASTM B117-18, the cabinets would meet ASTM
specifications. The values submitted for cabinet 51 are identical and without scattering, which does not seem
very realistic.
6 of the 8 cabinets for which the NaCl content was provided in percentage per weight meet the specifications
of ASTM B177 – 18. It is assumed that there was a transmission error in the data submitted for cabinets 17
and 30. The values were thus multiplied by a factor of 1 000. These two cabinets are highlighted in the box
plot in 0Figure 17. Two further cabinets did not conform to either ISO 9227 or ASTM B117-18.
19852_ed1fig15.EPS
Key
Y density (g/cm )
19852_ed1fig15_key1.EPS lower tolerance limit ASTM B117
19852_ed1fig15_key2.EPS
lower tolerance limit ISO 9227
19852_ed1fig15_key3.EPS
upper tolerance limit ISO 9227
19852_ed1fig15_key4.EPS
upper tolerance limit ASTM B117

Figure 15 — Documentation of the density of the test solution including the tolerance range defined
in ISO 9227 and ASTM B117-18
19852_ed1fig16.EPS
Key
Y NaCl conc
...