DTD911C

Download as pdf or txt
Download as pdf or txt
You are on page 1of 44

AD-775 197

EVALUATION OF THE DOW 17 TREATMENT


FOR MAGNESIUM ALLOYS
K. G. Adamson, at al
Ministry of Defence
London, England
February 1973
DISTRIBUTED BY:
National Tecbnical Information Service
U. S. DEPARTMENT OF COMMERCE
5285 Port Royal Road, Springfield Va. 22151
.1 Unlmite
D~. Hat RL%,wrt No. 192 Iaef. ZA,;9/O54
Febrary, 19703
M4INISIRY OiF DEMC~XE (rPOCURE9RENlr C1v)
I ~DMfECTORAIE OF RES.A.RCH JWLIERIAIJS 2
EVXALTATION OF TIM DCW 17 TREXPIEN'1
by
K.G. Adamson
J.F. King
W. L'nsivorth
JM3GNTS1154 iaXKTRON LIMITED
FEBRUARY 1973
Uxxlimi
ted.
- D,- Mat Retprt 192
pA
FOREWORD
This report dec-ribes work carried out by Magnesium Elektron
Limited, -Stino't, Uanchester, =.ider Ministry ef Defence (PE)
Cont-aee KFAL/0579/CB.43AZ2 4
n _T e wo.k was suervised by Rt, NF4 nq-.dr tVe di--ection of
U l-'
SUnlidcited
-r
3;
Unliuli ted
D. Mat Report 192
February 1973
EVALUATION OF TME DOW 17 TREATMENT FMR MAGNZSIM ALLOYS
The operating parameters for the Dow 17 surface treatment of =agnesium
alloys have been determined. The thickness of Dow 17 coatings can be
satisfactorily controlled from the thickness-time relationship at
specific current densities.
The cleaning effect of Dow 17 has been compared with that of fluoride
anodising. The results obtained were not consistent, but suggested that
no significant deleterious effe,:t would result from substituting a
1.0 mil thick Dow 17 treatment for fluoride anodising, chromic acid
stripping and chromating.
Various thicknesses of Dow 17 pretreatment have been compared with
standard DTD 911C pretreatment. The assessment included corrosion
tests on the various pretreatments with surface sealing, painting
to DTD 5580, and a full DTD 911C procedure. The results indicated
that Dow 17 pretreatments in excess of 1.0 mil thick, were as good
as the DTD 911C pretreatment.
The effect of various thicknesses of Dow 17 coatings on the fatigue
strength of ZW3 and RZ5 alloys has been determined under rotating
bending conditions. Signiificant reductions in the fatigue strength
of ZW3 .iere observed with increasing thickiness of Dow 17. However,
comparable reductions were not observed with 1Z5.
Unlimi ted
C ONT E NT S
Page
SU?4,ARY
c
I.1. I NTRODLUTION
1
2. INVESTIGATION OF OPERATING PARAMINERS OF DOW 17 BATh 1.
E- 2. 1 Installation and Preparat,.on of Bath 1
2.2 Experimental Details 2
2.3 Results
2
2.4 Discussion 3
ME :. COMPARISON OF DOW 17 TREATMENTS WITH DTD 911C
Awl
PRtEPTREATMENT
3.1 VINTRODUCTI ON
3.2 COMPARISON OF THE CLEANING EFFECT OF TE 17 4
TREATMENTS WITH FLUORIDE ANODISING
3.2.1 Introduction
4
3.2.2 Experimental Procedure 4
-; 3.2.3 Results
3
3.2.4 Discussion
5
3.3 COMPARISON OF THE PROTECTIVE VALUE OF DOW 17 6
TREATe&N'IS WITH DTD 911C PREETREATMh-EN?
3.3.1 Material and Sample Treatment 0
3.3.2 Evaluatin 7ests
7
3.3.3 Resulte
7
3.3.4 Discussion
3
4. COMPARISCN OF VARIOUS CURRENTLY USED PROTECTIVE SY-TEMS -3
BASED ON DOW 17 VITH DTD 911C PROCEDURE
4.1 INTRODUCTION 9
N 4.2 5"EV-ALT.WTION OF t3OW 17 TREATMENTS ON SI LE PANELS J
-Of 4.2.1 Material and Sample 1eatment:
4.2.2 Protective Schemes Evaluated
4.2.3 Evaluation Teats
4.2.4 Results
10
4.2,5 D scussion
U
4.3 ,VALUATION OF DOW 17 T11EAi-N"MllQN FLANGE
1I
ASS3L I ES
4.3.1 Material. and Sample Treatment
ii
4.3.2 Protective Schemcs Evaluatted
1I
4.3.3 Cirrosicon Tests 12
4.3.4 Resul ts
12
4.3.5
DDisussion
5. COMPARISON OF THE FATIGUE PROPERTIES OF' DOW 17 14
TR.EA -ME.7-N
;5 Introdiction
14
5.2 Mterial and Sample Ta ,nt
14
5.3 Results
14
5.4 %)sctu'sion
1
6. GENERAL CONCL1%XO7;8
15
7. REtFE?1ECES
17
8. TABLES
18 - 25
9. FIGU.ES
2 - 35
10. APPENDICES
35
3
is
77
V/
At the tirat of inceptiun of' the work described in this report (Mjay :L968)
the U.K, a~ircraft iztdustry was ia-creatiingly undertaking joint projects
with otner European -fianufacturers, and an~ increasing nurmber of American
built, or Amarican designed aircraft or components were being used. As
a r:asult, th" Dow 17 process, widely used in the U.SA, and Eurc-pe, was
assLrning increasing importance in the U.K.
Various t'ser' adbpted different conditions for the Dow J7? treatment, and
no comparattve studies of their effectiveness as pretreatments had been
made. In particuIrr, no direct comparison had been mad& 1'etween Dow 17
pretreatwents and fluoride a~nodising followed by stripp,,ng and chrom-ating,
as required in M1 911C.
The pu~rpose of this investigation wras to study the Dow 17 process ado
revaluate ttie coating in comparison with equiv-alent, currently used
rpretreatments, Th~e following main areas "er- investi&-ated.
(a) Irvestigation of operating- parameters okp the Dow 17 bath.
(h) Comparison of Dow 17 treatments with the DT'D 911C required
pretreatment.
(C) C0Me LiSo0i -f vai'ious c-urrently ua'o-1 proteictive systems kased
on Dow 17y with DTD 911C proc-diire.
(d) Comrlris-an of fatigiie p-cplftrti-s bfter Dow 17 treatments.
i-'requeent reference, is made Liroughout this report to the thickness of
Daw 17 films. Film thicrnesses, were obtained by micrometer measurement,
and are actually increases in dimeensi'on. The true film th-icknesses
would be somewhat greater due to inward growth of the filmn. 17he %term
"thickness" is used, however for simplicity,
2. INVESTIGATION OF' OPERAIVJ~r PA:F'A.-ATFP.S OF' DOW 17 BATH
2.1 INSIPALEAT!ON AND PREPARATION OF BAVh
A Dow I? pi-lot plant with an elettrolytie capacity of 56 gallons ;ras
designed and installed. *f4he pl;xnt cionsisted of a mild steel tank
w4.th o:uternal electrical heating c.- il-i j'2 Kw, and a thermostat.
As the tank could not be earthed diuring operatioa it was surrourded
by a wooden guard, with removeabJle lid operating a safety cut-out on
tl-e aod.9ig auuret va i is.din. aluminiumn bus bars to ths
sph eroles. HFrn fumesi evapraiomn fram alesourfaidedo te bathe bats
kw1
56 galsena of electrolyte were made up to the following formulation
in tap watcer.
Scodim i~i chromate (aCr 0 .2 2l0% 107"W/v
Phosphoric Acid (85%HP 9% V/v
22EXPERIMEW'AL DETAILS
P-5 2.2.1 Establishment of Pairameters
ZW3(Hr 04 aloypanels, dimensions 4ins.x 2ins, were used
comosi'.or.oZZW3andoter !'-ov usd n tisevaluatien are
notd, nAmicrometer measuremients on panels before arid after
txevnet idictedthe increase in dimensions due to anDdising.
Thae relationship between the various paramete.-s was then
determined.
2.3 RE~SULTS
2.3. Thiknes ofCoatiulls
Thve reainhpbetween thickness awl time could be interpretec!
as apoiaeylinear up to 2 roils (0.002 ins.). For thicknross-.C
in eces of2 ml6,par~ticularly that produced at high current
denitisthe thickness i.ncreased more rapidly with time, (see
Fi.1'. %Swol be expected, a greater film thickness was
obtan~dfor a gixven time wxith inarpasing current density,
The plot of thickness agairt3t term.inal voltage in Fig. 2 showed
that a rapid increase in thicknese occurred at terminal voltages
in excess of 100, particularly with higher current densities.
Higher current densities gave higher, terminal voltages for
equivalent film build-up, although behaviour at 50 A/Wt was
anomolous in this reapett for lower film build-ups. (Fig. 2)
Nr- simple relationship between thicknress, current density and
terminal voltage could be- seen.
The thickness off Dow 17 filmrs urp to approximately 1.5 mil, could
with a little experi-!nce be very aceurately judged from the
colour of th-im po~cd. The colour of both sides o.f! each
sataple, and all" samples from at batch of trea ted sanples, was in
every case similar, indica~ing that the throwing power of the
treatment Vas exceptiona' ly go,(>J.
2.3.2 Anodisirng of Test Panels
Having establishead a relationship between film build-up and
AF anodising time at various current densities, batches of panels
were anodised wi th specific thicknesses 2f Dow 17 for subsequent
tests. Batches were anodised at 30 A/ft on thi) basis of data
Pi Vshown i~n Figure 1. Actual thickness runges obtained are shown
in Figures 1 and 2. The thicknesses were slightzly greater than
predicted by the earlier tests, and in addition the terminal
voltages were higher. These variaticas were probably due to
changesin bath composition with increasing usage.
2.4 DISCUSSION
From the rbove results, the most pronising method or controlling film
~~ ~thickrxe;;s during anodisinig was by prediction from the t~hicinetime
curves at specified current densities (Figure 1). Dow literatrex
i.ndicated that for a specified alloy, the total ampere minutes
required was independent of current density. 'n Figu~re 3 thicknesses
wo2re plotted against total ampere minutes for all the trial runs
carried out. Up to 1.5 mil, the relationship 'was approximately
r_ linear. However, some variation with current density was evident,
U ~and the use of the total ampere minute requirement alone to prodicr't
~ film thickness was not thought to be sufficiently valid,
When the initially determined film thickness-time velationship at
30 A/ft
2
was used to predizt film thickness for sub~sequent batches
of panels (2.3.2.), actual thicknesses produced w-ere greater than
those predicted. This was attributed to changes in bath composition
F ~ ~~wlth increasing usage_ Subsequent a.oin wokcrid oti h
bath indicated that actual thickness had drifted to an even higher
l.evel t~han those shoown in Figure 1. In order to compensate for this,
the slope of the pre~diction Iin-z was increased empirically to the
position inidicated by a broken line in Figure 1. This was used
successfully in later anodising work. it seems likely that parameters
would have to be re-determined for a fresh bath or for a scaled uip
plant, but the present me~thod has been found adequate for small scale
work.
3. COMPARISON OF DOW .0 TM7ATMEN1TS WITH DTD 911C PREMhEATNIENT
s_ :3.1 1M. ODUCTION
The purpose of this invesziga~.iorn was to compare variu.us thicksiesses
:,f Dow 37 film with p*.'etreat'ynent in accordance with 0TD 911C. The
Dow 17 t-eatment had potential advantages in that, giv'en good cleanIng
R4_ ~ and corrosion resistant proj.ertt.es and compatab-ility with stoved epoxy
resins, the single process coLld iplace the rather lengthy, and more
expensive procedure of fluoride anodising, chrom-ic acid stripping,
activation and chromating. 7The investigation inoluded the assessment
of the cleaning effect produced by the applica~tion of various thicknesses
SF of Dow 17 film in comparison wiJ.h fluoride anodising and the comparsion
of the ccrrosion resistance of surface sealed Dow 17 films with that
of the current DTO 911C pretreatment.
3.2 COMPARISON OF THE CLEANI.NG EFFECT OF TIE DOW 17 TREAMULINTS WIh
3.2.1 Introduction
Tvo separate tests were carried out. In the first the cleaning
effect of a 1.0 mil Dow 17 film was compared with fluoride
anodising on shot blasted A8 (BSL 122) alloy. In the second
test ZREI (BSL 126) and ZW3 (BSL 507) alloys were used, and a
range of Dow 17 and fluoride anodising treatments were applied.
3.2.2. Experimental Procuedure
T9-st 1
The investigation was carried out using a single I ins. thick cast
A8 alloy plate, with a relatively smooth cast surface. 3 ins. x
3 ins. specimens were cut from the plate and heavily blasted on
both surfaces with steel shot. Quadruplicate specimens were then
given the following treatments;
1. Fluoride anodised 15 minutes aZ 120 V.
2. As (,) then stripped in boiling 106 chrotAc acid.
3. Dow 17 treated to give a nominal 1 mil per surface.
4. AS (3) then stripped in boiling 10% chromic acid.
Specimens ve-re weighed to ob" a the increase in weight due
to anodising, and re-weighed after stripping.
Corrosion testing was carried out in 3% sodium chloride. solution
saturated with Mg(OH)
2
. Corrosion rates were deter-$ ned after
76 and 414 hours.
Test 2
Four 2RE1 alloy cast plates, 8 ins. x 7 ins. x *in. thick,
and * in. thick rolled ZWl plates were used. Both materials
were heavily shot blasted with steel shot before citting into
4 in. x 2 in, test panels. Panels were s%lected ,c-r treatment
so that the whole range of pretreatments was carried out on
panels from a single plate, in order to prevent inaccuracies
resulting from differences between plates.
panels from both alloys were given the following treatments:
(1) Untreated
(2) 0.5 mil Dow 17
(3) 1.0 mil D-w 17
(4) 1.5 mrol 1ow 17
(5) Fluoride anodised - 15 mins. at 120 volts.
L(6) .Luor~de anodised - 15 mins. at 90 volts.
Details of these treatzints are given in Appendix I. Half
the samples in each condition were then stripped in boiling
10% chromium trioxide solution.
Duplicate samples were weighed and immersed, in a X sodium
chloride solution saturated with magnesium hydroxide. Corrosion
rates were then determined after l6E hours.
3.2.3 Results,
The results of corrosion tests are shown in Taole 2.
In test 1, on AS alloy specimens, the lowest corrosion rates
were shown by specimens which hPd been fluoride anodised and
stripped, followed by specimens Dow 17 treated and stripped.
Specimens from which the films had not been remov I showed
higher corrosion rates, with Dow 17 treated specimens the
highest. In test 2, on ZREI and ZWJ alloy specitens, the
results were less denisive. The conclusions could be
summarised as follows:
(1) The corrosion rate of ZRE1 was higher than that of
Z'6l under all treatment conditions.
(2) Subsequent stripping of the anodised film made a
major contribution to the ,reduction of corrosion
rates after both Dow 17 and fluoride anodising,
particularly with ZWL samples.
(3) 120 volt fluoride anodising showed less cleaning
effect han the 1.0 mil thick Dow 17 treatment
9
which was generally the most efficient cleaner
of the Dow 17 treatments. Cleaning effects of
0.5 and 1.5 mil Dow 17 treatments were roughly
comparable with that of the 120 volt fluoride
anodising.
(4) 90 volt fluoride anodising gave less cleaning effect
than the 120 volt fluoride anodising and was poorer
than any of the Dow 17 treatments.
3.2.4 Discussion
"I The distinct superiority of fluoride anodising as a cleaning
treatment observed in test 1 *was not confirmed in test 2. The
wrider range of treatments indicated a greater variation in
cleaning effect although. both treatments had a considerable
effect particularly after stripping.
STe two equivalent practical processes under consideration were
fluoride anctdising followed by stripping in chromic acid, and
I Dcw 17 without stripping. Both processes had a significant
cleanirng effect, although the relative met it varied from test I
to test 2. In test 2 the best cleaning effect was given by a
1.0 mil Dow 17 film. Thicker or thinner films gave less cleaning
effect.
F
In retrospect a more valid comparison would have been between
a Dow 17 pretreatment ard a pretreatment consistuig of iluoride
anodising, stripping in boiling chromic acid and then chrome
manganese treatment. The effect of this would be to further
improve the corrosion resistance of the fluoride anodised
samples.
For all practical purposes, the results of both tests urdertaken
suggest that the cleaning effects obtained by Dow 17 treatment
or by fluoride anodising and stripping are of the same order,
M aad application of a Dow 17 pretreatment as a single process
would not show serious disadvantages. from the cleaning aspect,
compared to the 3 stage DID 911C procedure.
3 3.3 COM-PAIISONJ OF MEE PROTECTIVE VALUE OF DOW 17 TREATMENTS WITH
I DTD 911C PRETMEAThENif
3.3.1 Material and Sample Treatment
'Die evaluation was carried out on sand cast high purity AS
r (BSL 121) alloy plates and rolled sheet specimens in ZW3
(BSL 505). All sheet samples were cleaned by inmersion in
boiling chromic acid followed by a 10 second dip in 54,, nitric
acid. Batches of sheet sample- were then given the following
tceatments:
(1) 0.5 mil build-up of Dowv 17
(2) 0.8 mil build-up of Dow 17
(3) 1.3 mil build-up of Dow 17
(4) 1.5 mil build-up of Dow 17
(5) DTD 911C pretreatment. (i.e. fluoride anodised,
chromic acid stripped
and 2 hours chrome manganese
bath treatment)
Details of treatments are given in Appendix 1. All samples
Swere surface sealed with Araldite 985F rosin. The coating
nroduced a film thickness r" 1.5 roll on the non-porous chrome
m;tnganese film and a correspondingly smaller build-up on the
more porous Dow 17 films.
"it was rather more difficult to vbtain a predetermined thict ei.i
of Dow 17 fil& on the sand cast A8 samples, as accurate micrometer
!F -measurements could not be made on the rough sand cast suJrfaces,
and no experience of anodising conditions for alloys other than
"Z.3 was available. !hatohes of shot blasted AC alloy plates wrec
tnerefore Now 17 treated to give nor='nally thin, medium and
thick ccztinAg, by comparing the film weights obtained on AS
with those obtained on ZW3 alloy.
Mf
r-." Tese samples wer-e also .iurface sealed with Arallite 985E.
.1, - 6
S3.3.2 E-valuation Tests
Panels were subjected to the follo~ing tests:
R-
Physical tests
1. Shot test
2. Bend test
Corrosion tests
1. Salt Spray and Humidity Test - 12 months
2. R.AE. Seawater Spray Test - 12 months
3. Atmospheric Exposure Test - 12 months
4. Distilled Water Column Test - 12 rtonths
Dotails ,f these tests are given in Appendix I1.
3.3.3 Results
Phnvsical Tests
The results of the bend and shot tests are given in Table 3. The
adhesion of the Araldite 985E resin film to a chrome manganese
film was better thani that to the thin (0.5 - 0.8 mnil) Dow 17 films.
Thicker Dow 17 films tended tL craze on the inside of the bend,
A and the resin was detached. Where this occurred part of the
anodised film was removed aith the resin, leaving, in the case of
* Me thinner film, a very thin light buff coloured coating. With
thicxcer films, a thin greenish-buff coloured film with minute
areas of the darker green film remairrd.
The shot test also indicated this tendency for failure to occur
within the Dow 17 film, rather than at the metal-film or film-
resin interfaces. The supe-ior adhesion of Araldite 985E to
the chrome manganese films was again evident.
Cor-osion Tests
S~Salt Spray and H~umidity Test
Samples previously used for the Shot test were used for this part
of the evaluation. Details of examination after 12 months exposure
are given in Table 4. On as cast A8 samples co-siderable asperity
attack occurred on the thinner coatin-s. Medium and thicker Dow 17
coatings were less affected by asperity attack, and the thick Dow 17
film showed good resistance to creepage corrosion from all points
of damage. 9TD 911C procedure gave be.ter protection than the
thin or medium Dow 17 with surface sealing but 16hick Dow 17 with
surface sealing showed .he best protection.
-t
: - 7-
Or 7.W3 pnl, all brealdowr was by creepage beneath the surface
st)sling, resin and &1l the Dow 17 filmns gave better protection
:1: Uti.3 DIED 9flC proceadure. The extent of creepage corrosion
decreased with -Ixrreusing- thickness of Dow 17.
R. A. 7-3. Sealwacter- Spray Ts
De.tail1s cf the examinati 'rt aftzI.r 12 months expos'ure are gi',on in
'lnb1* 4.
The cor3roion oc-curring during this test: was more severe than
that obtainer' in the humidity te-st although the amotut of
cr--epage CorrusiC'n was less. on cast pankils, thin and medium
Now 17 f ilmPs showed severe! asperity attack, DIM! 91V, Pr, treatments
showed slight a'-perity attaclz- while the Puick ')ow 17film was
jirtually unattacired. On roiled pagiets, t',ie corresion resi4stance
of" Duw 17 films inczreased witn inexlacsing thickness. UT!) 911W
treated panels ware slightly warri? than the thinnest Dow 17
treated panels.
AtMaSPheric Exrasurse Te-st
"Attle breakdown oc'unied on any of the panels although lightening
in col~ur cn aspe;-tias2 of cast panels could indicate slight
asperity corrosion, TAhis was apparent on all the cast panels,
althougiz slgigtly leszs mark--d on the DrO 911C treated panels.
~2A Rolled panels showed slighft pitting along scribed cro~-ses and
siki;%p eoge-s of coding, but t-her-e was no significant c~ifference.
Li oistilied Water Column Tst
(:.111 tue Ps cast panels showrad some elect-rical conduction shortly
after t)hc start of the test, but this did not increase sign~ificantly
iiduring the test. Ther DT!) 911C treated, pan-els show-od slightly less
conduc-tion titan the Dow 17 treated panels, but no truze breakdown
or corrA~tosJionocre on an Pnlsa e 12 montns.
3.3.4 Discuss'ion
The inabil-ity of ar unpigsiented resir. to adequately cover the
;asperities of cast surfaces was demionstrated by the humidity and R.A.E.
seaw*:ater spray- tuests.
The a6-ouM- of cree-page from- po.ints of damaag~e dec~reased with increasing
taickness of Dow 17 coating. Thle adhesioni of surface sealing S~r!
bette.--r to the nthicker, rather porous, Dow 131 Coatiinga, that, it
was to the thinner, smoother and more com -t Do-:w 17caigshr
the resin was not able to fully -impregnate the very fine pores,.
On the smoother surZace of 0?!) 911W treated rolled panels, the
k- anxaunt of croepage from points of damage was considerable, This
was undoubtedly due to the reduced mechanical k;eyiw'-, compared to
that obtained on a cast surface, and the resilient nature or the
resin,
4. C-UOARISON OF VARIOUS CU N-Tf_,, USED PROIE-CIVE SYST%2'S - BASEn ON
DON 17 V!Trh j D 9 110 PflocEURx
4,1 INi.ZDWTI ON
SThe effect of Dow 17 pretreatment, as an. alterntive to DTi) 910
procedure on the propertLes of the ctmplete protective system was
nvest.ted. Th=e coating sy3%tem used was DTD 5580 r:Ath and without investigrted. Th e-edgsI
prior surface sealing. Dow 17 treatments by Procol in France and
M.E.L. were inve-tismt&
4.2 EVALUATION OF LX)W !7 TrREATIMENT ON SIMPLEP-ANE-LS
4.2.1 antel and Sample Treatment
-. The evaluation was carried out on 3 ins. squre by 6 swg. ancd
4 ins. x 2 ins. x 16 swj.. rolled paneis in ZW3 alloy, The
panels were deburred, deireased and cleaned in !% c.roliun,
trioxIde solution followed bv a 10 second dip in cold 5% nitric
acid solution. Dow 17 treatment was carrI-ed out at ,0 amps/sq.
Sft; the times being adjusted to give the -rez!quired film .hic!nesses,
Sce sngples wre Dow 17 treated oy procol, France. These
samples had a very thin (0.15 il) greenish buff coloured film.
4.2.2 Prote-tive Schemes Evaluated
Batches of panels wre given thle folloing surface treatments:
1. 0.15 idi Procol Dow 17 + Piinting to DTD 5580.
2. 0.15 rail Proool Dow 17 + Surface Sealing + Painting to UTD 5580.
3. 0.5 mil Dow 17 + painting to D9M 55B0.
4. 0.5 mil Dow 17 + Surface Sealing + Painting to DTD 5580.
5. 0.8 mil pow 17 + Painting to DTD 5580.
6. 0.8 mil Dow 17 + Sw-face Sealing + Painting to CT0 5580.
t 7. 1.7 mil Dow 17 + Painting to DTD 5580.
8. 1.7 mil Dow 17 + SL-uface Secaling - Painting to DTD 5580.
9. Fluoride Anodised, Stripped in boiling Chromic Acid,
Chrome Manganese treated, Surface Sealed + Painting to DTD 5580.
IDtails of the treatmemts are given in Appendix I.
4.2.3 Evaluation Tests
Sample panels have been subjected to the following tests:
(a) Physical tests (bend tests, shot tests)
(b) Salt Spray - Humidity Test (Using shot test specimens)
- 12 months
(c) R.A.E. Seawater Spray Test - 12 months
Md Distilled Water Collmn Test* - 15 mno.-t."
The tests are fully described in Appendix 11.
4.2,4 Results
(a% Physical Tests!
The result,3 of the shot and bend tests are, given in Tatle 5. In
al! cases whierc fnilure oco.-ui-ed on tio inside ..7 tc bend L~he
underside of the -etacherd paint cr resin wzas greert a coloulr.
and some o the Dow 1?7 film coulc4 be seen adhering to it-. A
thin green-Ish-huff film remained on the metapl surf ace. Wtth
the 0.35 mil Dow 17 Ejil, Very l-ittle could be seen adhei-ing to the
resin or paint, but with Uie thick film- almost all the Dov 17
fIlm was dached vith the o.r&nic fitm.
Al though the UMh 5580 sysitepm wax- fufffciently flexib]; to wil"itan'I
bending to tlh.e point of metal "ailure arouxd the autside of the
bernd when applied tp a L~ow 17 f il.' i~t sho:,wed pocr adhesion over
tie Araildite 985-- resin rilm and e - her !Fpalled or cusuld be pe-eled
off from the entir-e deformed area. No breakdown of th-e surface
F V sealing occurred on tho outjeide oZ the bend.
Iii ths shot test~s, paihals with paint applicd Oirectly or. the
Do~ 17 film spallea within the Dow 17 f "im except tor very tbin
Ptoc-ol Dow 17 treated panels, -in which the failure appeared to
he at the nairnt/Dow 17 interface. The poor adhesion of the
DTD 5580 paint to surface sealiing was again zhovwn. There was
no spalling of the surface sealing.
(b) Salt F-rsay-flumidityTs
'fhe resultts tir this test ave given in Table 3. Thbe thtn. Dow 17
tr-eated pane -s whicij were not surf ace seaLed! chowed .ecr crep
corroaion and some blistering. l-rotectiv%' scl e~s biaaed or., the
thicker Dow 17 coat-ings had improved corrosion resistance.
Although the ad;oesion of' the DTDV 5-5SO paint scheme to surface sealing-
was poor, the benefit of surface spaling was evidIent, ev-en on the
thin Dow 17 treated panels.
Mledium and thick Dlow 17 films with aad without surface sealing
gave better protection than the 0Th 911C system but thin Dow 17
was inferior to DTD 911C. in both conditions. niest protection
was given bY the surface sealed medium (0.8 mil) andu thick (1.7 mil)
Dow 17 films.
(c) R.A.E. Siawater Spray Test
The results of this test are- also shown in Table 6.
romparatively little breakdowan oc-curred. Thlat which did occur was
A ~pitting at damage points with little creepage. Agnain redium and
k ~trsick Dow 17 f ilms gavQ goad protection, bujt the thin f Uss
N pi-rticularly without surface sealing were inferior to the DTD 911C
proctedure.
(d) Distilled Water Column Test
7No apparent deterioration of any cne of the schemes occurred
I + durtwig the 15 month expozoure.
4,2.5 Discussion
AU thi tests shorWd an improvement in the protective value of
f.l ow 1V films by curface sealing. In addition, the schemes based
on. thick and medium Dow 17 films with surface sealing were
Z:
initcan-alty supwTior to the DID 911C scheme. However, the
-DiD 9C110n ohe'e was superior to the surface sealed thin Dow 17
"s..xseme,
ihe DTD 5580 paint sc=eme showed poor -adhesion to surface sealing
""uid ehib!lIted bliseturing t-ader con0itions of high htbaidity,
pa-tiiaurly in the absence of surface sealing. This indicated
sme degree of water permieability in the polyurethane scheme.
4.3 EVALUATIU- OF 'OW 17 TREA'.L-rNT ON FLANGE ASSEMBLIES
4.3. Material atd- S eapla Treatment
The evlation was carr-ied out on standard M.E.L. flange assemblies.
Each sand cast flange, wvas approximately 4*- diameter and had a
centrul boss, drilled apzi tapped to accommodate a nominally 1 ins.
dinametcr pp. paised spot faces equally spaced around the flange
ware drilled to accommodate 5/16 ins. diameter bolts. Flanges
were machined then given the appropriate protectt-ve treannent. Twv
such flapes, one in AS alloy! and the other in ZRSI alloy, were
us ed in each aasE-mbly. They were bolted together using 5/16 ins.
diameter BSF :admium plated nutax, bolts annd washers. A 3 ins. length
of mild stee pine was screwed into the boss of the A8 alloy flange
and a similar length of chromated aluminium alloy (Tr30W'P) was
SI:sewed in the ZREI alloy flange.
z 4.3=2 Protective Scheme- Evaluated
All the flanges were thoroughly deburred an-d degfeased in
t-iehloroet~hylene vapour. Batc-hes of AS and ZRqE! flanges were
given the following treatment:
1. 0.15 mil Dow 17 (Procol)
2, 0.15 mil Dow 11 (procol) + Surface xeaiing,
2. 0.8 mil Dow 17
4. 0,8 m-'i bow X7 + Surface Sealing
5. 1.5 mil Dow 17
6. 1-5 m!l Dow 17 + Surface Sealing
ELt
- 1
7, DTD 911C Pretreatment (Includlng Chrome Manganese)
8. DTD 911C Pretreatme:nt (Including Chrome Manganese)
Surface Sealing
9. Dm 911C Pretreatment (Including Chrome Mainganese ) +
J. Hails 588/0066
10. H.A.E.
1. H.A.E. + Surface Sealing
Details of these treatments are given in Appendix IT.
The wild steel and almininun alloy tubes were screwed into
the flanges using Palycast Type 2 sealing compound. Additional
Pairs of sisilarly treated flanges were baited topether; the
nuts being tightened to a torque of 12 ft. lbs. Polycast was
again used ior assembly and caulking the joints and bolts.
All flange assemblies with the exce-ition of those. having the
J. Halls 588/0066 coating were ther. painted to UTD 5580. Those
having the J. Halls treatment were pair.ted to DTD 5555. Details
of the paint treatments are given in Appendix 1.
4.3.3 Corrosi-ot Tests
SAssem-blies were subjected to the following tests.
(a) Salt Spray ar4 Humidity - 1 year
(b) R.A.S. Seawater Spray - I year
(c) Marine Atmospheric Expoetre (ieaunaris) - 24 years
Details or these tests are given in Appendix I!I.
4.3.4 Results
(a) Salt Spray and Hrumidity Test
~ Details of the qen-ination of the assemblies after 12 months
exposure are su=marised in Table 7. The assemblies have been
given a corrosion rating,; A to E, based on visual assessment
of zhe degree of blister and breakdown. A1l the assGmblies which
F: had been oe-r.osted with D1') 5580 paint scheme showd bl0stering
to sa'e extent. Corrosion breakdown had occurred beneath some
blisters, particularly adjacent to spot faces and on flange edges=
rBlistering and breakdowns werm more pronounoed -,n non-surfooe
I .. xeied a-an-- as c be seen by comparing Figures 4 and 5 whIch
= show the maximu= blistering observed, and the impr-ovement produced
by surface sealing, Fewer blister -. ere evident on the H.A.E.
treated assem---bly-
The flange assembly acated rith J. 'Halls system and overcoated
with rTD 5555 showed no blistering, hbt some breakdown had occurred.
I 1K
1-2I
All the non-surface sealed assemblies showed poor resistance to
high humidity conditions. Systems incorporating surface sealing
Sshowed a higher order of corrosion protection. The best protection
was given. by H.A.8., surfZace sealed and overoainted with 07Y) 5580,
which showed no breakdown an' only minute localised blistering.
The standard DTD 911C syn-tew fell into the same general categnary
as the other systems enie.odying surface sealing.
(b) R.A.E. Seawater Spray Test
Table 7 sumnlarises the results of 12 months exposure. A corrosion
rating 'has again been given. Severe c-orrosion break-downs occurred
on all assemblies Which had not been surface sealed. The largest
a,-d most nure-rous breakdowns occurred on flanges with the very
zhin Procol Dow 17 pretreatment (Fig. 6) and became less severe
wit%. increas-ing thickness of Dow 17 fzlm'i No significant breakdown
occurred on any flange assemblies which had been surface sealed,
aitnaugh single isolated points of breakdown were observed on the
assemblies with 0.15 (Fig. 7) and 0.8 mil Dow, 17 coatingi. )..5 mil.
Dow 17, H.A.E. and DTP 911C pretreatmnerts with DTID 5580 top coats
all showed no breakdown.
The extint of break-down was generally more severe on AS than on
Z.RE~ flanges. The j. Hall's 588/0066 plus DTD 5555 system
s~howed breakdown on the A8 flange, although the corresponding
ME1 flIange show.~d none. The severe blistering which occurre-d
with DTD 5580 coated assemblies in the humiditv test did not occur
in tne R.A.E. test.
(a) Marirne Atmospheric _Epo2sure
Flange assemblies were examined after 21 years ex<posure. The A8
alloy flange of the 0.15 mill Dow 17 'Procol) plus DTD 5580 treated
asaerabi~y had three corrosion creepages. None of the other masemblies
showed any blistering- or break-downs.
s!igh deradation of the paint i tself waIs evident, as shown by
cracking of the L01D 5580 and DTI) 5555 paints on the corners of
the siteel washer6 etc. The epnoxy finish on the J. Malls 588/10066
plus painting to 0TD 5555 scheme had chalked considerabl1y.
I-435 Discussion
The results of the RHumidity and R.A.E. Seawater S3pray Tests indicated
in a most eqphatic manner the necessity to surface seal comlmnents
operating in corrosive environments. The poor corrosion --sistance
of all the pretreatmetnts ex~mined without surface sealing may be
due to the water permeability and poor adhesion of the DTO 5580
system, as indicated by the severe blistering occurring tin the
humidity test. As would be expic-ted, increase in thickness of anodic
coating improved corrosion resistance, although it did not itself
prevent corrosion, even In the case of the II.A.E. film. Surface
sealing prevented significant corrosion on all th-- pretreatmients
studiod and no real differentiation could be made between the ~.arious
systems zlthough ear-lier work (3,4.2) on simple panels indicated
[.that thinner Dow 17 films (up to 0.5 mil) were inferior to DID 911C
pret:,earhnc-t even w~hen surface sealed.
-13-
The marine atmospheric exposure te; t at Beaumaris proved to be teo
If milJ a corrosive environment to be useful in the evaluation of
ceaplete protective systems,
5. COMP.A-RISON OF MIE FATIMS! PRO' MRTIES OF DOW 17 TRIMI-IENTS
The ebject of the investigation was to determine the effect .f 'arying
thi.nesses of Dow 17 on the fatigue strengths of RZ5 and Z'W3 alloys,
using an Avery 0302 Wohler-type fatigue tvsting machine. The S/N
curves obtained %uld be comparable with those previohisly obtained for
%he same !lovawith a wide range of surface treatments.
1
5.2 LATE-q.IJL AND SAMPLE TREAATIMEN
50 sand cast RZ5 (L 128) alloy DTD bars were produced and machined to
staudard A'ery type, Wohier fatigue b-rs with a gauge diameter of
0.2629". 40 ft. of 1 ins. diameter extruded bar in Z,13 (L 505) alloy
was also produced. The chemical compositions and mechanical properties
of the alloys are given in Table 8. Stringent nrecautions w'.-e taken
to obtain lengths of extrusion free of defect, a;,d t.ese were tIen
machined to fatigue bars. The fatigue bars in both z.1loys were randomised
prior to surface treatment to avoid any progre*iive deviations arising
from factors associated with casting or extrusion of the material.
eatzhes of 12 fatigue bars from each alloy were Dow 17 treated to
p)roduce a film build-up of 0.5, 1.0 and 1.5 mils. The test area only
was anodised; the remainder was blanked off durng anodising.
Tnickness of film produced by anodising was obtained by micrometer
measurement at the point of maximum stress before and during treatment
until the desired thickness was achieved. Bars we e also retained in
the un-anodised condition for comparison.
5.3 RESIULTS
S/N curves for plain pumiced, thin (0.5 mil), medium (1.0 mil) and
thick (0.5 mil) Dow 17 treated specimens in Heat Treated RZ5 (5SL 128)
mad extruded ZW3 (BSL 505) have been obtained. The results are plotted
in Figs. 8 and 9. The curves show that Dow 17 treatments had no
significant effect on the fatigue properties of RZ5 alloy. On the
other hand, a marked reduction in the fatigue properties of ZW3 was
produced. The effect varied from a reduction in fatigue strength of
approximately 95 for the 0.5 m1 Dow 17 coating to 21% for the 1.5 mu
thick Dow 17 coating at 5 x 10 cycles.
Spalling of the Anodic coating occurred on fracture with all three
Dow 17 coatings on Z93 but no such spalling occurred on RZS.
414
Examination of Nta va tographic sections taken from the fractured fatigue
bars showed rnat t.. D5ow 17 treatments produced a roughening of the
metal surface; the mnri'lhening increased with increasing film thickness.
There was no diseernun ' 'n ifference in the surfac' roughness produced
by similar Dow 17 treat .esvl on RZ5 and ZW3.
5.4 DISC7jSSION
it is known that treatrxuts ic.ulting in pittinr or roughening of a
Smetal surface result in the redu?,;.sn Li the fvtignie strengths of the
material, and that tn.e great+" the :,nui.ening effect, the greater the
nrduction in fatigue strength. Sin'.%% any Dow 17 treatment of magnesium
produces some degree of 3urface "-oughkui!ng, the treatment is likely
to reduce the fa tigue properties of maegeaium alloys. Although this
was shown to be the case with ZW3 allo} whcn reductions in fatigue
strengths of up to 21% were obtained, the eltoct oZ Dow 17 treatments
on RZ5 was negligible.
The notch sensitivity of magnesium casting alloy-s Is generally lower
than that of wrought alloys. Consequently the eZec-c of Dow 17
treatments on the fatigue strength of RZ5 should be less than that
on ZWY3. However, some reduvictioan in the fatigue strength of " -
w-auld be expected. It was not possible to exp~aiu this &tppart.t anomoly
from the results and observations mazde.
A comparison of these results with those of previous work ahowed that
the effect of Dow 17 and H.A.E. n.nodiixng on the fatigue" properties of
ZT3 was, thickness for tticknesc, very similar. The reduction in
fatigue properties of ZW3 alloy resulting from a 0.5 mil thick Do 17
treatment was similar to that produced by a chrome manganese tra-atm!-nat.
However, the chrome msinganese treatment had no significant effec',t on tie
fatfgiue properties of R75 alloy.
6. GENMSAL CONCLUSIONS
The following coraclusions can be drawn from the results of the work.
INVESTIGATION OF OPERATING PP'AMETEIS OF DOW 17 BATH
1. It is possible to control the thickness of Dow 17 film with a
reasonable degree of accur-acy by controlling the anodising time
at specific current densi ties in accordance with the graphs
shown in Fig. 1.
COMPARISON OF DOW 17 TREATMENT q lfTh DTD 911C PRETREATMENT
1. No significant deleterious effects wouid result from substituting
a 1.0 nil thick Dow 17 treatment for fluoride anodizing.
2. 1.0 nil, and thicker, Dow 17 coatings are comparable to the DTD
911C pretreatment as bases for further organic protection e.g.
surface sealing. ThMinner coatinq-s, particularly the very thin
coatings. comonly used in the U.S.A. and i-ivrope, are inferior to
DMi 911C procedure, and their use is not rec.ommended.
a
3. Dow 17 coatings have been showu o be porous and consequently must
be surface sealed for optimum proteazt on.
Z- ... .
COLAP-ARISOH %)F VATUOUS CLJRRENTM' USED) PHUTELM. I S-STEMS MAOMD ON
DOW 17 WItIJ DTD 911C PROCED!IPE
1. Systems based or surface sealed Dow 17 coatings were ccmpiprablc
W1.0 Di Th 91.1C procedure. DOW 17 films, without surface spealing
2.The use of ,,t assemb1l, tetu-Miq ies were esqential to avoid the
-~ galvanic corrosion of miagn-.siuu. in complex assemblies.
3.. 'Ibe D-Th 5580 paint sc~heme e-hibited Wistering =nder high humaidity
test conditi~ons indi-cating that the paint was to some extt-~
water oermeabie.
C-CAPARISC\ OF' IIIE. FATIGUE PROPERTIES OF DOW 17 T'REA'PENTS
1. "Ibe catigue, rtrenigth of 2X3 was considerably reaucced byDowi17
treatm~ents. No similar deletvrious effect was obtained with
p25 aad- further workc is requitred to explain 'this apparent anomo3.y.
.- 1k
7. REFirERENC-S
R1.
EmleY, E.F. Princeiples of MagnesiUm Tl~ch-nology, 657 -660
Pergamon Press (1966)
1-k
Bl
B&S
-17
0
z 0
0 4
p!
U o4 H
ot
N M
TABLE 2: RESULTS OF TESTS TO COMPARE CLEANING EFFECT OF DOW 17
TREAiMENTS WITH FLUORIDE ANODISING
T 1Section 3.2
FR Test- 1
Average Corrosion Rate AVerage
2
Corrosion Rate
Alloy Treatnent (mgs/cm
2
/day) After 76 lirs. (mgs/cm /day) After 414 hrs.
Snstrippe41 Stripped in CrO
3
Unstrippedj Stripped in CrO
3
120 v
Fluoride 3.74 1.51 5.76 3.10
Anodise
1 rmil Dow 17 5.
3 4
2.14 9.33 4.04
__ .__ __I._____ __ _
Average Corrosion Rate (mffs/cm-/day) After 168 flours
Aloy Traten - Unstripped Stripped in Cro
3
None 39.3 43.0
0.5 rii 32.2 1.6.1
SDow 17
w.0 1l 10.5 15.1
i1Dow 17
1.b oil
Z R E116.9 18.2
Dow 17
120 v
Fluoride 30.3 18.5
Anodising
90 v
Fluoride 34.8 21.4
Anfodising
None -46 14,3
0.5 rail
0 i9.2 0.4
Dow 17
S1.0 mil
Dow 17 3.2
0.7
z w 1 1.5 Wil 8.!.0
Dow 17
_--120 v
Fluoride 7.9 5.7
S; Anodising
90* v
F9-Aoride 28.1 10.7
Anodising
"Specimens washed and scrubbed only.
INot cleaned in chronic ecid
Wide Spread in corrosion rates.
I
0: Q: 00 a
0 F41
all ~0 ~ 0 0
ci.
4*~0
bato
0 :30 2
00
t- l -4
r .6
0 x ~ "I r* V
a. - c Z. 0
0. a00
44V
0
1
to940
r i 1 4 -41 H -Q0
00
-d0 s. 4-42
:1 0 Lt 40 r.,
Jo
2: <-
0.
rr
H
be
100
0 40 44t
S
4
-4
.- A-- - -~m
TAME 4: REWLZTS (.If 2 WOThj UtMtIOIlY TESLAT I.- WIM)'ES_.A. I-AAEI PAYEST '(N VARL- -3 mlThTNSSESs
OF DOW 17P YILM WINh -SL(RFACE 8FLI W
Section 3.3
12 Mdonths shot. Salt Spray Mt-sdity Teat 12 Montris R.A.E. Seawater Spray
9Atcta SUrf&C '.e r4at~ntft
A3. Crze-,ae from CXtVV I CrOepage from shtGeneral Surtauce 0. Oeeiage from 1Gen~eral Surfaces
I I 1~0~Cross
D". 17 (0.46 oz/ytt ) I1)General asperity breakdown or. front face with cree-pago. 3) Gzoner" heav'y Mlmeroua breakdoon
Srf cot- staling 2) Creepage 2 rox shot.* and cracks on rrar up to 10 cmi,. 4) corrosion to W both saurftoea
eo 1 0.4Cr,"page froza chat, croia, car~tar to 10 =%a. on frontfact, Is ons. !pto 15 sma.
Dc 7(.4Wl 9 Largeeaest15=sLoer o16s.Apiy b.d. aro~und 11 Pit ting up to %~u .d. on
S acsealing
&,o o1 m ma =. max. >.oth faces to
5 10 IF*.to1 3 tma. max. Axperity b.d. at shot 12 pittint to 5 =v Several b.d. on
L Ito 10 ma. max. rAX. >noth sides to
IiCreepage trom c rucks 7 namaa max.
ii _______________ jon rear. _______
Do. 1.7 (1.3 zyd 2 7 1 ema.I. SIlght anpority bhA. 19 4 b.d. (I am. 4b.d. on rear
+ ufc 31n 3lrshot. Creepage !fan. In 4 ma.
~ ~ hi-a creck3 or. rear
I LI Ilfractl to 13 ams.
I.. 1Ia bd
Ifrom cracks on rear lam.
pitoridg Aroviiaed - 3 6Z La m. 31 va. )l1gt talperityrb.d. 21 1b.8. I G. Ibothfuzasperity
seo . urtosh 14I-on rot face c.1r. 27j ~ i b.d. or. both fat~e
&h~~# Sufc I
I C on.
Sclng.0 I4to I exv~. 3 1~tifasperity b.d. 28 6 Sb.d.( <I c. Several asperity
I I fronZt fact only, Ib.d. on both face
I - - -to 5 Ms.. max.
Dw17 (0.5 all) 4 1 65 8t 13 =an. max. ~1Is fma. -AX. remYfrom edge on 167 24 b.d. t. 4 sole.m'an 01
I~ crrepagee an rear
Surf ace aC~~~1l~~g ~ar(c i to 13 7Ain n. oig o aa
6AH 12 to 1!7m=a. max. 17 mas. man.on z-mar
1
. 130 b~d. to 3 aM1..b ron
23~ ens..
Saurfso ntftlinw -IAt rear to 4 mcZ, and at holes a.nd
124 6 to 11 mos. C-AX. 10 ;Ma. ann. Crtecpgce fr.m darage 36 10 b..d. 1 m:%. Sligtt b.d. at
_____ ja
t
rear to 7 mona. I ito 2 ca. oea odn
Derv 17 (1.3 a",) . 1 ZSeverftl <0.:1 7 =an. =aX. Creepahv from damage 43 7b mm 1a. S2ight b.d. at
Surzacel aIling at rear to *7 ana.. holes. codtng
U 1 I to 5 zasn. I to 1W It 0&3. ftax. I -- Ilage trog daitago [44 7 b.d./u l V. AA far 43
Jat rear to a me&. ______
nw17 '.1.3 all) 49 3 to 1, =.ma.4 w-%&.~ max Lrjpg fo t C 151. 20 b.d. /4 1 a=. Ilightib.d. at
qur.
atra o j..hla oding
at roar to 3 cotn. m
or nuorl odt Io Mcatoroug to 15 1s 23 mnn. W.X. Qrtnpsgo from damage Z? 1Ioaorout to 3 amb.d. at b*)-es.
Mj hM max. jat rear to 13 ams. " ~ Max. Qdges CodinLg to
.fooede o 0 an on front face
385 vj-Muma. to 19 a=n. 213 UMa. max. Creeciogze from danzage 60 ,Naimrous to 5 Imn.R.d. at holes,
V AX. at rear to 1.0 btn. j~x edgen. 2cod-
.*h e13 e~a. Cili b.d. on
[.............. ___________________ __rear face 4 ma.
TA=LE St RESULTS OF Mi'D A14D SlO?" TESTS OX VAUMIS3 1TI1CKEN.SSES OF 0 17
FILM WITM AND W*.1t"I',, S7IFACE SAIA'r, PLUS PAINTING TO VM 55.0
Sact ion 4.2
-RESMLTS OF HLI',i TESTS j SULTS OF S1rO T-msfS
r nInsitie btnd Outsd- Dend
Duscription 1- Decrpti__ MaII- F'_-_nt ___V Pea Fac
10.15 ail Dow 17 4. Splitnc of paint up to 1Itela of pant up to
o.15 nil Dow 17 N o.D. N.D. Spallino of paint t1 SpallIng of paint td - S.S. + VTD 53S0
sur-tace sealing up to sur.face seolling; u~p
0.5 altl Dow 17 i- Datach=ont of* ;mint Dow 17 so brakdow Srllilne of paitn to 12 an SllinC of prlat to 1-'-
DID 5580 to m--os. Max. /paint
0.5 oil Dow 17 Ootac)mnt of total Do. 17 Dotamltent of panz S.S,/ Splling of paint only SPIitlin of palit to 4
S.S. + DaD 5580 film up to 6 =A. B.S. only up to P cu. pint to 50 MS. M.a Cracking of paint
aax. to 2". mms.
0.8 ail Dow 17 , Dot..chment or puint In No brP-J,own Spoiling of paint to Spalling of pain' to 4
DTD 5580 + scom Dow 17 to Dow 13 cms. =r.. Ftrthar cracklno
a to# 1I of paint to 9 =m.
0.8 Nil DOw 17 Dotach:ennI of paint In Dtachment o. paint S.S..; Spallang of paint only spalting or Paint only to
.S. 5 D 331.0 $.3. r-d 3oa'e Dow 17 Dow only up to 0 Pas. piant to 40 cma. 23 ms. . spallinT of
to a ms 17 !,S. to 3 s$. froc crackl
1.7 nll Dow 17 * Detaclhnt of naInt In N breakdown - Spoli ng of paint to Spmlling of paint to 4
DIM 5350 some Dow 17 tO DoL 23 runs. -A's. Further cracktng of
a mo. 17 /pint to 12 VAS.
1.7 all Dow 1? * Detachment ofpalnt, !n Detactoeont of paint S.S., Spael Ag of pAint only to Spelling af point onsy to
BS.. - amD 53W0 S.S, and sm" Dow 17 t - only to 0 man. paint 38 mas. some craing . f 25 imm. rurthei crAckinz
to 5 =2. Further 17 expcoed S.S. to 12 a--. Crazing old
c;Acklag to 6 cas. vpal;lng ct 8.S. tu 4 m-.
Chiaoe Xinganesm Dotachment of paint ChArrA Dotchv-nnt of paint S.S., S'Plling of paint only Spalling of pNtnt onl$ to
IS.o * D+ D 5380 * S.S. to 2 s. -a"/ o0.17 to 9 a1. ptin', to 15 ma. zomerally but 4 aa. Furthr crackin
4x. ,S. 43 v-1onC cracka of paint to 23 mo.
.. S. a .. .. acea. =- with Araeite 90U
TABLE.F 6: 2:M.LtES girS OF2 Mais2l MIfZ1O1Y TP-c" Awn 12 ka"I I R.A.E. SE"-o1ATER sflwAy r
ON VAIRIOUS 7M1CM=SSM (F fOM. 17 YJLUS WIIhANi v1hni3? 2S-F~x EA~LLII
PLUS MY,) TO fiTO SWO
______________
'~:ion4.2
12 )lomths Shot. sal1t spriky . iru~idity Toet f 12 Io t.h R.A.E. 3eaeat..r -:)rav
crs Frotec
crs ro' Shot
-
0.15 =il Dow 17 4,T pt 5mc. 5=.S=
M:
%VLw* W4str
12 tip~g ro _ b to 32.5: W =aa-
IR DD 5580 1 1p 10 7 au. .5 V
0.5clDw2 pto 1.5 =a 1 a. 1bm,% rsfw 2u oIx roakd4n-
1.7 1311 Dow 17 3 1,~ Crepg brac 2 o oV. hooicd,.
DID, 5590
(ron do' edps to
80u to 1 3 cas. - - No & frdo m 0 to 6 A* fl. "o br".ak'oMz
~ ~ - 1AX. +. 8 e xo brott dom hole. to*v~
point ..r .17 o 2~dw
0. ail 1o 1* 0 =2 IS .v breakdow.n 8b.d. ato 2. na o relco
$ 1 to 23ca.=M o ntldvn 1 to.d 1 .5 mx8. No broakdowin
I. nlord DnOWte -7 10 =5. CorIrosinkof~ 4 to 21 10* na. o bu~redowu
film. strapa
S to 2~u. CN. - Sligto 17aatn
j0 toon 13~ %M.mx
to 2 ems. c~ N
sevralt I =m_ _ __okdw 0t I --a_ _brvd-
B.S. -w0 DctiD with partit of8ros t
~1.
TABLE 7: RESULTS OP 12 MONMh HUMIDITY TEST AND 12 VON-hS R,A.E.
EAWATER SPRAY TES7S ON F7LAGE ASBEJ3mLIES
""ction
4.3
12 Month Salt Spray- 12 Month R.A.E. Seawrater
Surface IHmidity Test Spray Test
Tfreatment No........
4
No. C, R. No. C.R. _
N 0.15 rmil Dow 17
(Procol) + 2609 D 2809 E
DI9 5580 2/8 1/7
S0. 15 mil Dow 17
i (Procol) + 2809 B 2809 B
Surface Seal-ng + 5/11 4/10
DTD 5580
0.8 mil Dew 17 + 2848 D 2848 E
ITD 5580 38/62 37/61
0.8 mil Dow 17 +
irf ace Sealing + 2848 C 2848 C
01D 5580 41/65 40/64
S1.5 nil Dow 17 + 2848 D 2848 0
DT&D 5580 44/56 43/55
L1.5 mil Dow 17 +
Surface Sealing + 2848 B 2848 A
DTD 5580 47/59 4C/"
Fluoride -Anodised,
S2ri pped, 848 C 2848 A
Chrome Manganese 50/68 49/67
SArface sea'i:n4 +
_ DT;D 5580
S~~~Fluoride AnodisedP48D84E
Chrome-Ma nganese
SDTD
558053/715217
Fluoride Anodised,
Stripped, j 2866 C 2866 E
Chrom-ce-Mangainese +[ 2/4
1/13
J. ilalls 588/0066 I-
DTD 5555
H.A.E. Anodised + 2866 D 2266 C
UMD 5580 5/17 4/16
H.A.E. Anodisode+ i 2866 A 2"866
Surflce Sealing +
DTD 5580
1 8/21
7/20
IC.R. = Corrosion Rating as folicws:
A No significant deterioration
B' B No visible corrosion. Slight Blistering only
C SmaUl isolated corrosion breakdowns only with slight blistering
0 More general corrosion breakdown and/or general blistering
RE
r.4 0
I0
ImI
N tc
zi~
Z&
IFE
41.-.=-. 20 amp;/ft
2
F 4 0 3n . .
S.. 30 .. (Amendad)
..
a 40 ...
V &-- AO
-i I Ranges for samples
$3-0
-
iI.
Z
EI
0
!0
20
30
40
S~Time
(mins)
- Fig: 1. Increase in Sheet Thickness/Surface
v Tiume for Dow 17
i
~ treatment at various curre:nt densi ties.
S~Section
2.
-2e
S.~
N
20 amps/ift2
4.!4 0 30,, ,
.W -- 4-050 v .
%
Ranges
fcr samples
I
t.
3
2.
0
J
M
3-0 -
tEt
IC,
C
0 1"
I
Ii
70 80 90o 10 1oo10
Terminal Voltage
A Fig'. 2: increase in Sheet Thickness/Surf ace v Terminal Voltage for
Dcw 17 Treatment at Var-ious Current Densities.
i
Section 2.
M
b4,
-
r-
40-
S / Ranges for samples
S .22
303
4 0
5 0
._Ampere -miriutes/ ft2 (x100)
SFig. 3: Increase in Sheet Thickness/Surface v Total! mpr
Rne finutes at Various Current Densities.
2Section
2.
04 /
2. 4 6 8 -01 41
I
I
I
I
Fig. 4: zua F! nc" :
1
ftor 12 11!at12' Rinicix 1 pet*
Procol now 1 7 iV!) 5%S 0 p! nt Sc!v 1. 2 on 1. 5

I .jc
- 21 -
Fi g. 5: 7T11 .l anr,' Aftter 1.2 Months Iu,.i-!dity Test.
Procol L)nw 17 t Surface SealDin + ITI) 55S0 Paint Scheme
Section 41.3.
- -Cos rtw s2C
low
I.I t,7 A tr 1
Fig. 7: /7W]. Fl :
1
-*:l*: AV topr 1-2 Iiobnt ~"-
Procol iA) uf' 01 ~DD5- i ~
______1_
- 0
:>1
00
I-__
__
LLa
: :1-
61
1 0
TS~ ~ ~ I ILS X~
a:
IJ
4
L
I-I
0 - 0
I'- zk SS8L -
F z 34
S I 0 i
10, A!PEDICES
10.1 APPENDIX
I - DETAILS OFSURFACE
21RATEN1'S
(1) C1IRATE
ThEAThNT
TO DM 911C
(a) Fluoride
Anodisini
Path Composition
25% r-/v Ammonium
Bifluoride
in water
Anodising
Conditions
I0 mins. at 120 volts
(unless othe'-wise
stated)
(Ambient
temperature)
(b) ppin%
Bath Composition
10% w/v Chromium
Trioxide
in water
Stripping
Procedcre
10 mins. immersion
in
boiling solution
(a) Activation
Dath Composition
15% HF in water
Pr ceduare
5 minutes immersion
at
ambient temperature
(Ii) Chrome ,Manganese
Bath Composition
10% w/v Sodium Dichromate
Crystals
5% w/v Manganese Sulphate
4 2
5% w/v Magnesium
Sulphate
(mgSo
4
7-%0)
Procedure
2 hours immersion
at
aobient temperature
(2) DOW 17 TREATMENT
Bath Composition
24. w/v Arnmoniu-zz
Biflttoride
10% Sodium Dichromate
Crystals
9% v/v 85% Ortbo-Phosphoric
-
Acid
Current Denaity
30 am;s/sq. ft. (unless
S~otherwise
s tat-W)
Temperattre
70 .- 80C
Time
Dependent
on thickness
required
(See Fig. 1)
- 35 -
(3) H.A.E. TREAIMEN
Bath Composition 12% Potassium Hydroxide
1. 04e Aluminium
3.% Trisodium Phosphate
(Na
3
PO
4
. 12H20)
3.-% Anhydrous Potassium Fluoride
2.2% Potassium Manganate
Current Density 60 amps/sq. ft. (85 volts max)
Temperature Ambient
Time 90 minutes
(4) SURFAC SEALIWG
Material CIBA Araldite 985E
Procedure Sampl~s preheated to 220
0
C - cooled
to 60 and dipped. Air dried for
10 minutes, then stoved for 20 mins.
at 220C. Two further coats applied
by dipping (inverting sample each
time) and stoved for 20 and 45
minutes respectively at 2200C
Total thickness of 3 coats: 1.5 mils
(5) J. IUS CE AE PT71TEED
PRIME
Mrterial J. Halls Chromate pigmented stovin9
epoxy primer Ref: 588/0066,
jApplicarion 2 spraged coats steved for 19 rains.
at 12,5 C and 60 mins. at 200 C
respectively.
Total thickness approximately 1.0 mi3
(6) OVERPAINTIKI 'M DT 555:-%
Material Primer - Cellon chromate pigmented
cold curing epoxy primer
2 parts SL5539: 1 part SL5538 +
10% Thinners TSL 5373
Finish - Cellon pigmented cold
curing epoxy finish (White)
1 part SL5459: 1 part SL5538 +
S!;: :hinners TSL 5373
Applicatio 1 Spray coat off each of primer
Sand finish to give 1.0 mail and
1.5 roil coatings respeotively.
(7) OVYPAINTING TO DTD 3580
Material Primer - Pinchin Jobnson's chromate
pigmented epoxy primer SL6362 4-
Catalyst CSH 6331.
Finish - Pinchin Johnson's WRfito
Pivgiented Polyurethane Finish
S.3054 + Catalyst CSL3055
10.2 APPENDIX II - TESTING PROCEDURES
(a) PHYSICAL TEST;
(1) Bond Test
A 4" x 2" x 16 swg. ZW3 panel was bent slowly round a " diameter
steel mandrel until frnacture of the metal ozcurred. The sample
was then removed carefully to maintain the 'hinge' formed by the
coating system on the inside of the bend. The two fractured
sections of the metal .yere then gently p:iled apart. The extent
of detachment of the coating system from the point of fracture
indicated the adhesion. The performance of the coating on the
inside and outside of the bend was noted.
(2)
shot
Test
A 3" x 3" x 6 swg. ZW3 panel was used for the test. A 0.22 long
rifle bullet was fired at the coated sample from a range of
25 yards. The extent of spalling of the coating system from
the point of impact on the front and bi:k faces of the panel
was noted.
(b) CORROSION TESTS
(I) Seawater Spray-humidity Test
This test was usually carried out on the panel from the shot test
above. The panel was scribed with two crossed lines, 2 4riches
long, penetrating to the metal. Panels were then sprayed with a
fire mist of n-tural seawater then exposed in a cabinet at 98 -
IGO% humidity Zor 6 months. Samples were resprayed after each
intermediate examination.
The test was similarly conducted on flange assemblies, although
no scribed damage was included.
(2) R.A.E. Seawater Spray Test
Panels were scribed with two, 2 inch long, crossed lines to
expose the metal. They were then exposed in a shelter, open to
tae atmosphere on one side, and sprayed three times per working
day with natural seawater. at magnesium Elektroll Ltd., Manchester.
The test was similarly conducted on flange assemblies, although
no scribed damage was included, unless stated.
(3) Atmospheric Ex-posure Test
Planels or flange assemblies were exposed outdoors in the grounds
of M.E.L. at Clifton Junction, Swinton for a period of 12 months
or longer. depending on the degree of deterioration.
(4) Marine Atmospheric Exposure (Beaumaris)
Panels or flange assemblies were exposed outdoors in the grounds
of Laird (Anglesey) Ltd., 200 yards from Mean Sea Level and
25 ft. above it. The site was used by the courtesy of Birmidal
Developments Ltd., and Laird (Angles,y) Ltd.
-37-
(5) Distilled Water Column Test
A 5 ins. long coluwr. of distilled water contained in a glass tube
was located on a flat area of the panel 1 ins. in diameter by means
of rubber washers and clamps. The column was maintained in that
position for up to 15 months with periodic testing and examining.
Testing was e.arried out fortnightly by applying an e.m.f. of 12 v.
bet*een the base metal of the test panel and the top of the distil!.ed
water column, and noting any deflection of the microammqter in circuit.
The deflection indicated the extent of conduction (water permeability)
of the coating system.
38I

You might also like