Engineer S. BALAKSHIN

STUDY

OF FLOUR PARTICLE SIZE

WITH A FARINOGRANOMETER

Tomsk—1929

Engineer S. Balakshin.

Study of Flour Particle Size Using a Farinogranometer.

Introduction.

The study of millstones undertaken by the author in 1922 raised the question of determining the particle size of flour. It became clear that the methods of determining particle size that had been used up to that time were unsuitable for the intended purpose, and the author first had to undertake a study of mill sieves¹ used in flour milling, and then the development of the question of studying particle size in general. For this purpose the author had to develop his own design of a special apparatus—the Granometer². Only after this was the author able to approach closely the question of studying the particle size of flour. And only now, after several years of work, is it possible to publish some of the data obtained by the author on flour particle size.

It must be noted that first of all the author had to deal with the question of studying the most common grades of flour, determining for them the curves and coefficients of particle size; the data on this are presented in the present work.

I. Method of studying the particle size of flour with a farinogranometer.

Strange as it may seem, the concept of flour particle size in everyday life is rather indefinite, elastic, and conditional. The simplest method of determining the particle size of flour consists in rubbing a small quantity of it, taken in the palm of the hand, between the fingers and determining the particle size by feel (an organoleptic test).

Whereas one and the same flour in one place and in certain years is considered coarse, in another place the very same flour at the same time is found to be medium. For a more precise determination of the particle size of flour, laboratories have up to now used the method of sifting it on 2–3 sieves, coarse and fine; either the condition is set that on each sieve a certain percentage quantity of the flour being tested should remain unsifted, or the condition is set that all the flour must pass through a certain sieve (particle-size limit).

Thus, for example, sometimes the condition is set that flour of ordinary milling, when sifted through metal sieves

No. 12 or 12 threads per 1 English dm.
No. 20 ” 20 ” per 1 ” ”

must leave no residue on sieve No. 12, but a residue on sieve No. 20 of not less than 1½% is desirable (otherwise there is excessive grinding).

For bolted flour:

Sieve No. 12—no residue;
” No. 20 residue normally 0.5% and up to 1.5%.

For semolina:

On sieve No. 52—55 (or 55 threads per 1″) there must be no residue.

This gravimetric method is both inaccurate and not entirely definite. Thus, here the number of threads per 1″ is specified for the sieves; but in view of the fact that wire cloth may have wires of different thickness, the size of the openings of sieves with one and the same number of threads per 1″ may be different (by 2–6 times), and this method may give very different results depending on the thickness of the cloth. On the other hand, what is of purely interest is the very structure of fineness, the composition of the flour, i.e., of what particles, heterogeneous in size, it consists. All this led the author to undertake an investigation of the sieves used in milling and to work out in greater detail the question of their openings, as well as the question of fineness in general. As a result of many years’ work, the author succeeded in constructing a special apparatus, a “granometer,” which, as applied to determining the fineness of flour, he named the “farinogranometer.”

Without going here into a detailed description of the design of the farinogranometer, which we have given in our work on the study of the fineness of products¹, we shall briefly indicate here that, in general, the farinogranometer consists of a set of sieves, beginning with a coarser one and proceeding to a finer one (see below), on which a definite quantity of flour (10 g), whose fineness is to be determined, is sifted. The residues obtained on each sieve are each poured into a separate compartment of the second part of the apparatus, the “curvimeter.” The quantity of each residue gives the ordinate of the “fineness curve,” which, by means of a special device, is mechanically recorded on paper.

In addition, by means of the next part of the apparatus, the “granograph,” from the fineness curve obtained there is also mechanically obtained a total, or cumulative, fineness curve of constant, definite height, which is likewise mechanically recorded on paper and which easily makes it possible not only to compare different flour samples with one another in terms of fineness, but even to express the fineness by a single coefficient in the form of a fraction (the fineness coefficient).

The use of the farinogranometer, introducing complete definiteness, clarity, and objectivity into the question of fineness, gave the author the possibility of carrying out a comparative study of flour.

II. The set of sieves of the Farinogranometer for studying the fineness of flour.

It was indicated above that the principal part of the apparatus constructed by the author for the mechanical determination and recording of fineness (the Farinogranometer) is a set of sieves.

It is clear that the fineness curves obtained with its aid also depend on this set; in view of this, the author had to devote special attention to this question in order to establish this basic set.

The sieves in a set for determining fineness are usually selected in such a way that the openings of their apertures vary according to a certain regularity. The ratio of the larger opening to the smaller opening of the neighboring sieve in the set is called the modulus of the sieve set¹.

The modulus of the sieve set may be either a constant or a variable quantity.

Depending on the coarseness of which products or materials are to be studied (e.g., flour, cement, gunpowder, grain, etc.), the corresponding sieve set may also be used. Since our task included the study of the coarseness of flour, below we give the data for the set accepted by us for determining the coarseness of this product, i.e., for the farinogranometer.

But the same sieve set can, of course, also be successfully used for determining the coarseness of other loose materials, such as, for example, cement, gunpowder, molding sand, tooth powder, face powder, etc.

In selecting the sieve set for the farinogranometer, we were guided by the following considerations:

1) That in it the limiting coarse sieve should be one through which the coarsest flour would pass, and the limiting fine sieve one on which almost no fine flour dust (the finest flour) would remain;

2) That the set should include all the most commonly used sieves both for flour and for bran;

3) That the modulus of the sieves should vary according to a known regularity. It should be noted that at groat mills the products are usually characterized as follows:

In the so-called seyaniki, flour is obtained by passage through sieves Nos. 7 to 9 (apertures 210—163 thousandths of a mm).

As regards bran, in household use for bolting flour of simple milling, sieves of yellow copper are usually employed; moreover, the author was able to ascertain that in good harvest years and in productive districts people buy from dealers finer sieves, namely 100 threads per 1 vershok (56 threads per 1 English inch, aperture 380 thousandths of a mm), while in poor harvest years and districts they buy from dealers coarser sieves, namely 80 threads per 1 vershok (44 threads per 1 inch, aperture 477 thousandths of a mm). And sometimes even hair sieves of 24—28 threads per English inch are used (aperture 840—700 thousandths of a mm).

In seyaniki, bran is obtained by the tailings of sieves of 21—28 threads per 1 inch (aperture 934—700 thousandths of a mm).

At groat mills, in grinding on rolls, coarse bran is obtained as tailings from No. 26 (aperture 794 mm), and fine bran by passage through sieve No. 26 and, in grinding on millstones, as tailings from No. 40—48 (aperture 760—397 thousandths of a mm).

Taking all the foregoing into account, and especially the wide prevalence of simple milling, the author considers hereafter as bran everything obtained by bolting flour as tailings down to a sieve of 56 threads per 1″ (100 threads per vershok) with apertures of 380 thousandths of a mm (i.e., all particles larger than 380 thousandths of a mm).

After a fairly prolonged selection of sieve sets and a series of experiments, we settled for the farinogranometer on the following sieve set, as most fully satisfying the principal requirements.

The sieve set consists of 15 sieves, of which the coarsest is a metal one (12 threads per 1″), with an opening of 1620 thousandths of a mm, and the finest is flour (silk) sieve No. 25, with an opening of 29 thousandths of a mm.

Taking the modulus \(M\)—const—constant, i.e., the change in the size of the openings along a logarithmic curve, we have:

\[ \lg 1620 = 3.20952 \]

\[ \lg 29 = 1.46240 \]

\[ \text{Difference . . . } 1.74712 \]

Since, with a set of 15 sieves, there are 14 intervals, we obtain:

\[ \lg M = 1.74712 : 14 = 0.124794,\quad \text{whence} \]

\[ \text{modulus } M = 1.3627. \]

The value \(b_n\) of each opening of the sieve set may be determined by \(\lg\) from the equation (where \(n\) is the number of the sieve in the set)

\[ \lg b_n = \lg 29 + (15 - n)\lg M \]

Thus, for example, the opening of sieve No. 14 of the set is determined as

\[ \lg b_{14} = \lg 29 + (15 - 14)\lg M \quad \text{or} \]

\[ \lg b_{14} = 1.46240 + 1.0,124794 = 1.5879,\quad \text{whence} \]

\[ b_{14} = 39 \text{ thousandths of a mm.} \]

Table No. 1.

Sieve set of the farinogranometer
of S. Balakshin

Table No. 1 gives the results of calculations of \(\lg\) of the passage widths \(b\), the sizes of the passages, and the nearest most suitable sieve numbers that we have adopted for the sieve set of the farinogranometer.

III. Study of the Coarseness of Whole-Ground Flour

Whole-grinding remains, to this day, the most widespread among our population.

The rural population usually grinds its grain at local mills into flour with bran, sifting it at home (before setting the dough) and using the bran as feed for livestock, etc.

Taking this into account, it was of primary interest to study flour with bran from simple stone grinding on millstones.

We set ourselves the following tasks:

1) To determine what flour the population considers coarse, medium, and fine (by establishing for it cumulative coarseness curves and their coefficients);

2) To reveal the influence of the dullness of the millstone on the coarseness of the grind.

Flour samples for investigation

In order to obtain homogeneous results, we first took, from one stone mill located not far from the city of Tomsk (70 versts), average samples of stone-ground flour of grinds of various coarseness and with different degrees of dullness of the millstones.

Having established the character and constancy of the curves, we tested stone-ground flour from other mills and from other districts.

Our main attention was directed to the study of wheat flour, as the most widespread in Siberia, but several investigations were also carried out on rye flour and on flour from other kinds of grain.

The average sample of flour taken from the mill was carefully mixed; the required amount (10 g) was separated out and investigated on the farinogranometer in order to obtain the necessary coarseness curves.

It should be noted here that repeated experiments under different conditions usually gave coinciding curves, which we accepted as the basic ones.

A. Wheat flour of whole grinding

Coarseness of flour with different degrees of dullness of the millstones.

Carrying out experiments with whole-ground flour of different coarseness, as it is characterized at the mill (coarse, medium, and fine grinding), with different degrees of dullness of the millstones, we obtained a series of characteristic cumulative curves and coarseness coefficients for this flour.

In Fig. 1 are shown the cumulative coarseness curves obtained with a sharp millstone; in Fig. 2, with a medium one; and in Fig. 3, with a dull one.

Comparing these curves with one another, it is clearly seen that the difference between coarse and fine grinding with a sharp millstone is considerably greater than with a dulled one (a greater distance between the curves).

The same is also evident from the coarseness coefficients, by which we mean the ratio of the area bounded by the curve \(abc\) and the axes \(ad\) and \(cd\), see Fig. 1, to the entire area of the rectangle \(abcd\)\(^1\).

\(^1\) See S. Balakshin—study of coarseness.

The values of these coarseness coefficients are given in Table No. 2.

Table No. 2.

Coarseness coefficients of straight-run flour at different degrees of dullness of the millstones.

From Table No. 2 it is evident that the coarseness coefficients fluctuate:

for coarse grinding from $\eta_1=0{,}473—0{,}477$
for medium grinding from $\eta_2=0{,}383—0{,}413$
for fine grinding from $\eta_3=0{,}263—0{,}342$

The coarseness coefficient of straight-run flour for medium coarseness of grinding fluctuates very little with different degrees of dullness of the millstones ($\eta=0{,}383—0{,}413$), i.e., the coarseness of medium-ground flour changes very little.

In Fig. 4 we have presented ordinary coarseness curves for wheat flour of straight-run grinding.

From these curves it is evident how the product, as it approaches fine flour, shifts the ordinates of the coarseness curve to the right, so that, owing to the intersections, it is difficult to compare the coarseness of products by the curves.

For greater clarity, we have also shown side by side the cumulative coarseness curves in Fig. No. 4 for coarse flour at different degrees of dullness of the millstones, in Fig. No. 5 for flour of medium coarseness, and in Fig. No. 6 for fine straight-run flour.

From the composition of the cumulative curves of medium-coarseness flours for different degrees of dullness of the millstones, Fig. 6, it is evident that they almost coincide.

Thus, the cumulative curve of medium coarseness with moderately dull millstones may be regarded as the standard cumulative curve of average peasant wheat straight-run grinding with a coarseness coefficient $\eta=0{,}403$.

The cumulative curves for coarse and fine grinding at different degrees of dullness of the millstones differ more.

Having established, with the aid of the farinogranometer, the cumulative coarseness curves characterizing coarse, medium, and fine flour, it is easy to determine where a sample of some flour should be assigned, if the cumulative curve obtained from it on the farinogranometer is drawn on transparent paper and placed over the established standard curves, since the character of the curves for homogeneous grades of flour remains constant in this case.

The cumulative coarseness curves also give us the possibility of determining the quantity of bran at different degrees of dullness of the millstones and different grindings.

Above, we agreed to regard as bran all those flour particles which, during sifting, remain on the sieve with openings of 380 thousandths of a mm. This will correspond to the fifth ordinate of the cumulative curves. By plot-

sieve on the farinogranometer makes it possible to determine more easily the residue on this sieve for various cases of grinding and of dulled millstones.

Table No. 3.

Quantity of bran in plain-ground wheat flour at different degrees of coarseness and different dullness of the millstones.

From Table No. 3 it is evident that the quantity of bran with medium grinding ranges from 12.5 to 16%, reaching as much as 35% with coarse grinding and, with the finest grinding (on sharp millstones), falling to only 3%.

B) Plain-ground rye flour.

In Fig. No. 8 cumulative curves are given for plain-ground rye flour.

From comparison of these curves with the corresponding curves for wheat grinding it became clear that rye flour is, on average, somewhat coarser than wheat flour.

Thus, for example, the coarseness coefficient of fine rye flour is \(\eta_3=-0.358\), whereas for wheat flour \(\eta_3=0.263—0.342\). For rye flour of medium coarseness \(\eta_2=0.404\). The quantity of bran in rye flour, as is evident from the cumulative curve (ordinate 5), is from 13% to 17.5%.

IV. Study of the coarseness of re-ground flour.

A. Seyanka flour.

Even before the war (1914), in Siberia there began to spread widely mills of a type borrowed from Samara Province, the so-called “seyanki.” In these mills, from grain brought in by a peasant, by regrinding and bolting in two or three operations, flour without bran is prepared, the so-called “seyanka” (the peasant receives the bran separately). The same type of flour is also produced by some commercial mills.

In Fig. No. 9 are shown the cumulative curves obtained by us from various districts of Siberia, namely:

1) Semipalatinsk;
2) Yurga district (Tomsk);
3) From Vladivostok.

As is evident from these cumulative curves, the coarseness of “seyanka” is [[unclear: line appears incomplete in the visible image]], since bran is absent here (up to sieve No. 7—mesh opening 380 thousandths mm, the fifth ordinate here has no residue).

The coarseness coefficient of the flour ranges from \(\eta=0.153\) to \(\eta=0.263\).

B) Semolina flour.

Whereas before the war the semolina mills of Siberia produced 8–10 and up to 15 grades, in recent years the standard for commercial mills has been reduced to three, and now even to a single grade.

Figures 10–12 show the cumulative granularity curves of grades I, II, and III of semolina flour from various regions.

From the cumulative curves for the first grade, Fig. 10, it is evident that its granularity varies rather considerably depending on the region, changing from $\eta = 0.109$ to $\eta = 0.100$ and only up to $\eta = 0.198$. Here the difference in granularity of the flour of individual regions is smaller; finally, grade III semolina flour has a smaller difference in granularity, $\eta = 0.114$ to $0.170$. From the curves it is seen that the semolina flour from the mill in Vladivostok is considerably coarser in all grades. Especially fine, as the experiments showed, is foreign semolina flour (for example, English).

V. Granularity of flour and other kinds of grain and various products.

Figure 13 gives the cumulative granularity curve obtained by us for an average sample of buckwheat flour.

From the granularity curve in Fig. 13 the granularity coefficient for buckwheat flour is seen.

In Figure 14 there is a granularity curve for potato flour (starch); from it it is evident that this product is very fine, having $\eta = 0.110$.

Cumulative granularity curves of various products.

In order to have a comparative idea of the granularity of various products and of the nature of its change, in Fig. 14 we give the cumulative granularity curves of various products¹ and, below, a table of their granularity coefficients.

Conclusion.

The above work carried out by the author on the investigation of flour granularity, using the apparatus constructed by him for this purpose—the “farinogranometer”—and the methods developed by him, undoubtedly places the study of granularity on a scientific basis.

By conducting experiments with the farinogranometer on flour of any granularity and obtaining granularity curves, it is possible not only to determine quite accurately the granularity of the flour, but also to form a quite definite idea of its structure.

Once definite standard curves and granularity coefficients have been established, it becomes possible quite objectively, definitely, and indisputably to assign products to one or another category of granularity; and, undoubtedly, the farinogranometer will in the very near future find the widest application in mills of both simple and semolina milling, where it will be very useful for regulating and controlling the operation of the grinding, sifting, and middlings-purifying and other machines of this branch of industry.

Investigation of granularity by the farinogranometer will undoubtedly be useful in the production of powder, tooth powder, gunpowder, and other fine—

products; with other dimensions and a different set of sieves, the granometer can also be used for the study of coarser products (grain, groats, potatoes, etc.).

The author’s work on studying the fineness of flour by the scientific method he has applied is undoubtedly only the first step in this direction. The author has carried out only the first experiments in determining fineness, but it is undoubtedly very important in the future to establish the influence of flour fineness on its bread-baking qualities, its keeping capacity, etc.; this, in turn, will provide concrete indications for the rational conduct of its milling and other investigations, and for changing milling schemes toward better flour quality, and so on.

All these are questions of the near future.

In conclusion, the author cannot but wish that the granometer and farinogranometer may come into as wide a use for determining the fineness of products as the chondrometer has for determining their test weight; that standard curves and coefficients of fineness be developed for various products, and that its determination in the future rest always on objective measurements by an instrument, and not on an individual, often deceptive, method.

Fig. 1.

Flours from sharp-grained wheat.

From the article by Eng. S. A. Borodin.

[Graph with vertical scale from 0 to 100% and horizontal scale from 1 to 15; axes marked \(a\), \(b\), \(c\), \(d\), \(e\). Curves labeled \(\eta = 48.5\%\), \(\eta_2 = 38.8\%\), \(\eta_3 = 26.3\%\).]

Legend:

— coarse grade
--- medium “
········ fine

Typolithography “Krasnoe Znamya”

Fig. 2.

Flour at medium density of filling.

For the article by Eng. S. A. Boranshchina.

![Graph: flour at medium density of filling. Vertical scale 0–100%; horizontal scale 1–15. Three curves are shown: solid, dashed, and dotted, labeled \(\eta_1 = 47.3\%\), \(\eta_2 = 40.3\%\), and \(\eta_3 = 34.2\%\).]

Legend:

— coarse grind
-- medium “
···· fine “

Fig. 3.

Grinding by laboratory millstones.

For the article by Eng. S. A. Dobrokhotov.

— coarse grinding
--- medium “
······ fine “

Fig. 4.

Curves of coarseness for
plain-milling flour.

To the article by engineer S. Ya. Budakshina.

[Graph]

Legend:

— coarse
-- medium
···· fine

Fig. 5.

Coarse milling with different degrees of grain hardness.

For the article by Eng. S. A. Balakshina

![Graph: Coarse milling with different degrees of grain hardness. The vertical scale is marked from 10 to 100% on both sides; the horizontal scale is numbered 1–15. Three curves are shown: solid, dashed, and dotted.]

— hard grain

• • • grain of medium hardness

······ soft grain

Fig. 6.

Mean particle size at different degrees of grain dullness.

From the article by Engineer S. A. Bogatyrev.

Legend:

— hard grain

• • • grain of medium dullness

·········· dull grain.

Fig. 7.
Fine grinding with different degrees of grain dullness.

For the article by engineer S. A. Balakshina.

Legend:

— sound grains

• • • grains of medium dullness

······ dull grains

Fig. 8.

Rye flour.

To the article by Eng. S. A. Budrikhin.

[Graph with vertical scales from 10 to 100% on both sides and horizontal scale from 1 to 15. Two curves are shown, labeled on the plot \( \eta = 48.4\% \) and \( \eta = 35.8\% \).]

— medium grinding

--- fine grinding

Fig. 9

“Sift flour”

From the article by Eng. S. A. Balakshina

Graph annotations: \(\eta = 26.3\%\); \(\eta = 15.3\%\).

Legend:

Fig. 10.

Coarseness of first-grade flour.

Milling by engineer S. A. Burakshin.

![Graph of flour coarseness curves with percentage scale from 0 to 100 and horizontal scale from 1 to 15. The plot includes handwritten annotations \(\eta = 42.6\%\) near the upper curve and \(\eta = 10.9\%\) near the lower dotted curve.]

Legend:

— from Vladivostok
- - - - - Tomsk mill No. 209, 1927
— · — · — same, 1925
— · · — · · — same, No. 208, 1926
·········· Omsk mill No. 30

Fig. 11.

Coarse flour of the II grade variety.

For the article by Engineer S. D. Budnitsyn.

Legend:

— from Vladivostok
--- “ Tomsk, mill No. 209, 1927
······ “ “ “ 1925
-·-·- “ “ mill No. 208, 1926

Fig. 12.

Coarse flour, grade III

For the article by Eng. S. A. Boratinov.

Legend:

— from Vladivostok

– – – – — “Tomsk Mill No. 209, 1925”

⋯⋯⋯ — “ ” “ ” 1926

— · — · — — “ ” “ ” 1927

Typo-lithography of the Publishing House “Krasnoe Znamya”

Fig. 14.

Cumulative particle-size curves obtained with the granometer for various products.
For the article by Eng. S. A. Balakshin.

Legend:

— sugar sand
— coarse wheat flour
— medium wheat flour
— fine wheat flour
— tooth powder
— semolina
— medium rye flour
— potato flour
— cement
— rice powder

1. See also S. Balakshin, “Investigation of the Granularity of Products.” ↩↩↩↩

2. See S. Balakshin, “Study of the Particle Size of Products.” ↩