THE EFFECT OF CHANGES IN CHEMICAL
COMPOSITION ON THE GRINDING
ACTION OF ALUMINA-BASE ABRASIVES
R.H. SCHLEIFER, Abrasive Research & Development Engineer
Simonds Abrasive Company
The work reported herein was undertaken primarily as a survey to determine the type of information of
use in designing a research and development program for the evaluation of additives to abrasive crude. This is
strictly a literature survey; existing reports of previously run grinding tests were used as data sources, and no
tests were run specifically for this work. Since these previously run grinding tests were designed to solve an
entirely different problem, the amount of useful information which could
be gleaned from the reports was very limited. However, data on the grinding characteristics of two
series of abrasive were collected. One series consists of heavy-duty’ abrasives used for snagging, and the
other of ‘friable” abrasives used for surface grinding. This is particularly fortunate since it illustrates the two
basic types of materials which can be used to modify alumina-base abrasives.
TYPES OF MODIFIERS:
Basically, there are two classes of)substituents. Class I is composed of those substances which will fit
Al.)0 or corundum lattice, statistically replacing the aluminum atoms. The oxides of those substances
which will not fit into the corundum lattice, and therefore crystallize out as a second phase, are designated
Class II. The only Class I modifier currently used in any amount is chromia. Chromic oxide occurs in the same
crystal class and system as corundum, and the trivalent chromium ion is of such a size that it fits quite easily
into the corundum lattice. Zirconia. which has recently become popular as a modifier for alumina-base abra-
sives is an excellent example of a Class II additive. It is far too large to fit into the corundum lattice, in addition
to having the wrong valence; nor does it form any compounds with alumina. It therefore, crystallizes out of the
solidifying alumina as a discrete phase.
FIELDS OF USE OF EACH CLASS:
Class I modifiers would normally be used only in white alumina abrasives, which are essentially single-
phase materials. However, the converse is not necessarily true; it may well be worth-while to try Class II
additives in white abrasives. In fact, a fused, cast, white alumina which contained zirconia was tested in a
snagging wheel several years ago. It gave a better ratio than did a wheel made with standard heavy-duty
grain. However, the cost was so high that this material did not appear practical as an abrasive.
But in actual use one would expect Class II modifiers to be added to the regular alumina-base abrasives,
ASA NATIONAL TECHNICAL SESSION 1963 Paper 2 PAGE 1 of 3
the brown or grayish-brown materials. These abrasives already have an exsolved, intercrystalline phase. In
effect the zirconia modifies, not the corundum crystallites, but this intercrystalline slag”. The effect of a Class I
additive on the grinding characteristics of a regular abrasive would generally be negligible compared to the
effect of this intercrystalline material on these same grinding characteristics.
TYPES OF DATA TO BONSIDERED:
Customarily, new abrasives are screened by some form of resistance-to-breakdown, or toughness test,
and then evaluated by grinding tests. However, grinding tests are expensive, and it is highly desirable to have a
screening test which will enable one to make optimum use of them, and the effectiveness of toughness testing
for such screening is questionable. A possible reason for this will be seen in the discussion on the series of
crudes modified by zirconia. Since this was to be a basic study, it was necessary to decide what were the
appropriate data to work with. Obviously, it was necessary to treat the two series of abrasives separately,
since different parameters are important in snagging and in surface grinding.
ZIRCONIA MODIFIED ABRASIVES:
The basic parameters measured in our snagging tests are as follows:
Cut Rate - reported as pounds of metal remover per hour.
Grinding Ratio-reported as pounds of metal removed per pound of wheel lost.
Wheel Wear-reported in pounds per hour.
Within reasonable limits, the wheel wear is not as important as the cut rate; a wheel with a high cut rate
is better than another wheel with the same ratio but a long life and a low cut rate. Therefore, grinding ratio and
cut rate are considered in the zirconia modified series of crudes. Figure # 1 shows grinding ratio plotted against
weight per cent zirconia. The correlation is linear to a surprising degree; the point marked ‘“TZ”” is added for
information only, since this material contains no alumina. It is felt that the 20% zirconia material was somehow
defectively manufactured. It was out of line in every test run on it; in fact, in Figure #2, in which cut rate is
plotted against weight per cent zirconia, the point for the 20% zirconia crude has dropped completely off the
graph. Even its specific gravity failed to fall into line. In this second figure, the scatter is more pronounced, but
the correlation is still reasonably good. Again, the point “TZ” is added only for information.
In Figure #3 a resistance-to-breakdown index is plotted against the zirconia content. Here it appears that
a maximum toughness is reached, and that resistance to breakdown decreases with continued addition of
zirconia. This may explain why the zirconia-base abrasives, which showed up well in laboratory grinding tests,
failed to perform well under more severe field conditions. However, the actual shape of this curve is largely
conjectural, since these crudes were not produced under controlled conditions. Rather, they were obtained as
they became available from various manufacturers, and in many cases were obtained as finished grain. Non-
standard grain shape will in itself affect both grinding tests and toughness tests. It is hoped to engineer a repeat
of this work from scratch, sometime in the future. Crudes containing systematically varied amounts of zirconia
will be crushed and graded in a standard way. By doing this, scatter due to manufacturing variations will be
ASA NATIONAL TECHNICAL SESSION 1963 Paper 2 PAGE 2 of 3
eliminated, and more data will be available.
In both the 1 .5% zirconia and the 10% zirconia there seemed to be a reaction between the zirconia and
the titania. Where the TiO.,: ZrO ratio was high there was ‘a strong tendency for the cutting rate to be
higher and the grinding ratio lower than in material with the same amount of zirconia but less titania. The
crude containing 1.5% ZrO was manufactured in two modifications, having the same chemical composition,
but differing crystal sizes. Here the modification with the large crystal size showed a very high cutting rate
and a comparatively poor grinding ratio, while with the smaller crystal size the converse was true. This
material is probably anomalous in that the amount of zirconia present is actually lower than that of the titania.
Several crudes containing 10% ZrO were tested, and in addition to having a high cutting rate and a relatively
poor ratio, the material with a high TiO :
ZrO ratio was quite sensitive to heat treatment. The grinding ratio of this abrasive was directly related,
and the cutting rate inversely related, to the calcining temperature. The resistance’ to breakdown was also
raised by increasing the calcining temperature.
While the corundum crystallites are quite pure chemically, the intercrystalline phase is very complex,
with many possible compounds whose melting points are in the range of temperatures used for calcining.
When the crude is poured it is cooled very rapidly and non-equilibrium conditions probably exist. During the
calcium operation, as carried out, the grain is at a higher temperature for an extended period of time, and
some remelting of the intercrystalline material, along with deuteric reactions, would bring about a closer
approach to chemical equilibrium.
In summary, both the cutting rate and the grinding ratio are increased by the Class II modifier zirconia
in direct proportion to the amount added, up to about 40 or 50% ZrO . Above this point, the effect begins to
fall off, the ratio more so than the cutting rate. The toughness also begins to drop after going through a
maximum. The titania to zirconia ratio has a strong effect on the zirconia modified abrasives; this effect is
also influenced by the thermal history.
CHROMIA MODIFIED ABRASIVES:
There are three abrasives available in this series; white
alumina with zero chromia, a ruby grain with 2.2% to 2.4% chromia, and a pink grain intermediate
between the two in composition. Unfortunately, there were only two analyses available for the pink grain, and
the reliability of both is open to question. However, it will be shown that the chromia content could be uncer-
tain over a very large range without basically affecting the results.
In evaluating these grains it was felt that grinding ratio was probably not as important as surface finish
and heat damage. Therefore, the grinding ratios were varied by changing the wheel specifications and depth
of eat and the power drawn during grinding was recorded. It was found that for a given grinding ratio, the
power consumed was inversely proportional to the chromia content. Plotting peak power against per cent
Cr O in Figure #4, one would expect the slope to be much steeper at the 2.5% chrornia level than at the
This is because the behavior of chromia dissolved in alumina changes at about 5 to 7% Cr O . Here the
density and hardness of the ruby crystal begins to decrease, and the material becomes a less efficient
abrasive; the power consumption curve should go through a minimum and then increase. Since the curve is
non-linear but continuous, the actual analysis of the pink grain would determine the slope of the curve at that
point, hut not its shape.
This lowered peak power consumption with increased chromia content basically means that the grain
becomes more friable. This in turn, means that a harder wheel may be used to increase efficiency and/or
deeper cuts may be made without causing heat damage. As an alternative, a better finish may be obtained
with standard depths of cut.
Summarizing, friability of pure alumina abrasives increases with increasing chromia, giving cooler
grinding action. This action is limited by the fact that above a certain chromia content phase changes cause
the abrasive to become less efficient.
ASA NATIONAL TECHNICAL SESSION 1963 Paper 2 PAGE 3 of 3