By A. L. NO RBURY, D.Sc.
Stages in F orm ation
Fig. 1 shows in a diagram m atic m anner the gas bubbles form ing and increasing in size and bursting from the enam el surface. It will be noted th at between the enamel A and the cast iron C there is a layer of discoloured enam el B which is form ed by reaction between the enamel and the iron. W hen a bubble bursts, this dis
coloured layer (or dark ground coat, if present) is forced to the surface of the enam el and p ro duces discoloration. If gas is still being evolved when the enam el sets, the burst bubble
bubbles is greater than the expansion due to gas still being evolved, the bubble will contract and cause a depression on the enamel surface.
Sources of G as Bubbles
In Fig. 1 it has been assumed that the gas is form ed from a reaction between an oxidised defect and carbon in the metal, with the fo rm a
tion of C O z and CO. This is, however, only one type of gas evolution. A second type arises from a reaction between graphite or carbide in the m etal and oxide in the enamel, again with
f?AS B u b b le O x id is e d
\ D e f e c t
— ./
Fi g. 1.— St a g e s i n Fo r m a t i o n o f Bl i s t e r s, Pi n h o l e s a n d Bl a c k Sp e c k s. ( A = En a m e l. B = Di s c o l o u r e d En a m e l. C = Ca s t Ir o n.)
Fi g. 2 . — Se c t i o n s o f 3 -i n. d i a. Si l a l Ba r s En a m e l l e d Wi t h o u t An n e a l i n g. Ch i l l- Ca s t Ba r s ( Le f t) c o n t a i n e d En t r a p p e d Ga s e s w h i c h p r o d u c e d Bu b b l e De f e c t s. Sa n d- Ca s t Ba r s ( Ri g h t) d i d n o t. x jV
will not heal up and a pinhole will result. If, however, the gas evolution ceases soon after the bubble bursts the enamel Will tend to heal up and a black speck will result. (Black specks may, of course, also be caused by particles of iron or dirt in the enamel.) The diagram only shows the form ation of the bubbles, and during the cooling o f the enamelled casting it m ust be imagined that if the therm al contraction of the
* T his P a p e r was p rep ared for p resen tatio n to a jo in t m eetin g of th e I n s titu t e of B ritish F o u n d ry men a n d 'th e I n s titu te o f V itreous E n am ellers. I t w as originally p riv a tely p resen ted a t th e B ritish C ast Iro n R esearch A ssociation’s T h ird S tu d y Course.
the form ation of CO and C 0 2. C ertain enam el constituents such as tin oxide, which is reduced to metallic tin, react rapidly, while others, such as silica, are, of course, not reactive at the enamelling tem perature. A third type o f gas bubbling arises from the liberation o f gases which have been entrapped in the solid m etal during solidification. (Rusting after spraying may also produce gas from the oxide-carbon reaction.)
In w hat follows the various different ways in which the above defects can arise are discussed.
133
Pi n h o l e
Bl a c k Sp e c k
In practice, it m ay happen th a t defects on a single casting are n o t always due to a single cause, bu t to two o r m ore causes which may operate separately, each producing its own defects, or they m ay operate together and only produce bad defects when they are operating together.
Bubbles due to Entrapped Gases Escaping Considering the “ entrapped gas ” type o f gas th at the rapidly solidified chill-cast sections evolved gas, while the m ore slowly solidified annealed. T he difference between sand-cast m etal and chill-cast m etal is only one o f degree,
then cut vertically into i in. thick sections.
W hen one of these sections was enamelled (with
out previous annealing) the parts near the chill, whose m icrostructures are shown in Figs. 4 a
Fi g. 4b. x 50.
T h e fine g raphite in Figs. 4 a an d 4 b did n ot give blisters on enam elling, since the gas bubbles fo rm ed by the re ac tio n betw een the g rap h ite an d the enam el w ere to o sm all to coalesce.
and 4 b , did no t show any bubbles on enam el
ling. The parts farth er away from the chill, however, w ith the structures shown in Figs. 4c and 4 d , showed increased am ount of boiling.
Fi g. 4 d . x 50.
T h e coarse g raphite structures in Figs. 4c and 4 d gave blisters o n enam elling.
135
A pparently the gas bubbles form ed by the small graphite flakes were too small to coalesce, while those form ed by the larger flakes were able to do so and produce sufficiently large bubbles to produce visible blisters.
T he boiling on the coarser p arts of such sec
tions was considerable if they w ere shot-blasted before enamelling, but was alm ost absent if they were em ery-papered or honed before enamelling. O ther conditions being equal, the boiling was less if the section was annealed be
fore enamelling, presum ably owing to graphite being b u rn t aw ay during annealing. T he ex
planation of the shot-blasting differences appears to be that shot-blasting burnished over the surface graphite and delayed it reacting w ith the enam el until the latter had fused, after w hich the attack o f the m olten enam el on the surface o f the m etal and the annealing effect of the enam elling tem perature each had the effect o f uncovering the graphite flakes and allow ing them to react w ith th e enam el and p roduce gas at a later stage in the enam elling operation.
D efects due to coarse graphite tend to be fo u n d over a thick boss or along a thick section of the casting, o r opposite a runner, th a t is to say, in the m ost slowly cooled parts o f the casting. A nother effect o f coarse graphite is th a t it tends to produce a higher com bined ca r
b on in the iro n th a n is produced w hen the graphite is fine, and com bined carbon, as dis
cussed below, is another source of gas fo rm a
tion.
Bubbles due to Reaction between Enamel and Combined Carbon
C om bined carbon in cast iron reacts with the enam el in a sim ilar way to graphite, bu t if n o t present in too large am ounts, m ay no t cause sufficient gas evolution to produce defects. Its presence is, however, undesirable, since it is a potential source o f gas evolution. If the com bined carbon is considerable in am ount, due, fo r exam ple, to the m anganese content being insufficient to neutralise the sulphur present, the resulting high com bined carbon will probably produce boiling all over.
In this connection the silicon content is also of great im portance, since the lower the silicon content the lower the “ critical tem perature ” of the iron, and if the silicon content is below a certain figure, o f ab o u t 1.5 per cent., the iron at the enam elling tem perature will not be below its critical tem perature. This critical tem pera
ture is the tem perature above which graphite can dissolve and diffuse in the solid iron to form m ore com bined carbon. C onsequently, if during enam elling the iron is above its criti
cal tem perature, m ore com bined carbon will be
form ed and can diffuse to replace th a t which has reacted a t the surface of th e m etal w ith the enamel, so th a t instead o f the gas evolution ceasing, it will proceed continuously and cause considerable general boiling. If, how ever, the silicon content is sufficiently high, fo r exam ple, a b o u t 2.5 per cent., the iron will not be above its critical tem perature during enamelling, and the reaction betw een the enam el and the surface com bined carbon will cease very early in the enam elling process, owing to lack o f diffusion, and the gas evolution will no t cause defects as in the previous case, when the reaction went on continuously.
Bubbles due to Reaction between Oxidised Metal and Carbon in the Metal A nother type o f gas evolution m ay arise from a reaction between carbon in the m etal and
Fi g. 5.—Se c t i o n o f Ph o s p h o r i c Pi g- Ir o n An n e a l e d a t 800 d e g. C. f o r 20 m i n., Sc r a t c h- Br u s h e d a n d . Cr o s s- Gr o u n d t o a De p t h o f A - i n., t h e n En a m e l l e d. Th e Re a c t i o n b e t w e e n Ox i d i s e d Su r f a c e a n d Ca r b o n i n Me t a l h a s p r o d u c e d Ga s De f e c t s, x
oxidised defects below the surface. Oxidised defects right on the surface are rem oved by efficient shot-blasting. If, how ever, shot-blast
ing is n o t effective in rem oving this surface oxide— and this is m ost likely to occur a t re
entrants on the surface 'o f the casting— the above oxide-carbon reaction will cause defects.
T he im portance o f efficient shot-blasting needs no emphasis.
As an exam ple, Fig. 5 shows a section of phosphoric pig-iron w hich was annealed in a muffle so th at oxidation penetrated a short dis
tance below the surface, dow n the graphite flakes. It was then cooled, w ire-brushed, then the St. A ndrew ’s cross ground on it w ith an em ery wheel to a depth o f ab o u t -¡V in., then 136
sprayed and enamelled in the ordinary way.
It will be seen that no bubbling has occurred on the ground, oxide-free cross, whereas vigor
ous bubbling has occurred elsewhere, due to surface and sub-surface oxide reacting with graphite in the metal.
Formation of Sub-Surface Oxidised Defects Sub-surface oxide is probably one of the m ost com m on sources of defects in actual practice, and m ay form , as shown in the above case, down coarse graphite flakes. It m ay also form , as shown in Fig. 6, in w hat is probably a gas- hole defect near the surface of the casting. T he defect was found beneath a crop of pinholes in the enamel, and had become oxidised, probably
Fi g. 6 .— Cr o s s- Se c t i o n o f Ed g e o f Ca s t i n g b e l o w Pi n h o l e s i n En a m e l, s h o w i n g Ox i d i s e d Po r o u s Ar e a w h i c h r e a c t e d w i t h Ca r b o n i n Me t a l t o p r o d u c e Ga s.
x 100.
chiefly at high tem peratures, while the casting was cooling in the m ould and, to a lesser extent, while it was being annealed and while it was being heated up for enamelling.
Such gas defects in the m etal tend to form in the last part of the m icrostructure to solidify
— i.e., round the phosphide—and are m ore likely to be produced if the m etal is poured too cold or if the mould is insufficiently permeable to allow the escape o f mould gases. They m ay be due to sand particles coated with coal dust and clay becom ing detached from the m ould and giving off gas in the solidifying metal, o r they m ay be due to oxidising slag particles in the metal, such as might arise from the use of a dirty ladle, that is to say, one with an oxidised
skull whose oxides react with the m etal right up to the m om ent of pouring.
Sub-Surface Defects due to Included Sand Grains
T h at sand grains can become detached from the m ould and cause surface and sub-surface defects is shown in Fig. 7, which shows sand grains which have become entangled in the sur
face o f a Silal casting. W hen blistering o f the enam el is due to such defects it will usually be found th at if the enamel is rem oved by shot- blasting, under each enam el defect there will be found a small hole in the casting containing sand grains, as shown in Fig. 7. It will also
Fi g. 7 . — Cr o s s- Se c t i o n o f Ed g e o f Ca s t i n g s h o w i n g En t r a p p e d Sa n d Gr a i n s, x 1 0 0 .
be found that the holes w hich com m unicate freely with the outside of the casting do not give such a bad defect as deeper-seated holes which have only a small aperture connecting them to the outside and which are consequently less effectively cleaned out by shot-blasting.
Defects due to entrapped sand grains are often localised in certain parts of the casting—such as the rim of a bath—where the sand is most likely to collect.
Incidentally, it will be noted in Fig. 7 that the graphite flakes are coarsest round the sand grains and at the outside o f the casting. Fig. 7 does not illustrate this particularly well, but it has been repeatedly found in castings having supercooled graphite structures that the outside
edge, in contact with the sand m ould, is coarsened, just as if the sand had an “ inoculat
ing ” effect, giving coarse graphite w here it is least w anted from the enam elling point o f view.
Effect of C om position on Enam el T he nature o f the enam el will also obviously have a considerable influence on the size o f the above gas bubble defects, according to the tem perature necessary to fuse it; to w hether it will frit at a low tem perature an d entrap a greater am o u n t o f gas than if it fritted a t a higher
tem perature; to its fluidity at the highest tem pera
ture an d the ease w ith w hich it will allow gas to escape and be able to heal up gas holes w ithout form ing visible defects. T he reactivity o f its constituents will also affect the rate at w hich gas is form ed by their reaction with carbon in the metal. T he com position o f the enam el also, o f course, affects its adhesion, liability to crazing, colour, lustre, opacity, cor
rosion, heat-resistance, etc., b u t consideration of these properties is outside the present scope.
138