Classification And Testing Of Glues
( Originally Published 1906 )
Preliminary Considerations. It is obvious that the manufacturer, in assigning a price to the finished glue, must reckon with the cost of stock, labor, etc. Theoretically, he might base the price upon these considerations alone; but, since the most expensive stock may, through faulty treatment, yield a weak or inferior finished product, he cannot, in practice, be altogether governed by the cost of production. In addition to the latter, the quality of the product must be considered, and the measurement of quality necessitates the formulation of a system of test. Glue is sold on "test"; that is, the price of the product is governed by its strength and certain other properties, for the measurement of which many divergent tests have been formulated. Some call for the determination of the exact content of GELATIN, since, as a rule, the greater this content, the purer and, in some instances, the stronger the glue. The methods formulated for this assay are faulty and inexact; but, apart from this, the content of GELATIN has absolutely no bearing upon the commercial fitness of the product. Some take note of the amount of cold water absorbed by the glue in a given time. Others, again, are concerned with the breaking strain of the dry glue. It is to be observed that such test methods as are limited to the determination of only one constant or property are deficient in scope. Only by examining for several well defined constants, can any useful estimate of the general fitness of the glue be formed.
Accordingly, the dry glue must be examined as to color, finish and odor, and, if in the form of flake, sheet, ribbon or strip, as to the nature of the fracture when a piece is broken between the fingers or by other means. The solution must be examined as to odor, acidity or alkalinity, grease, foam and viscosity. The jelly resultant from the solution is tested for strength. The viscosity and jelly strength are compared with those of accepted standards.
Standards. The constants or measurements of quality in glue testing are arbitrary and of value only when compared with the corresponding constants of a standard glue. The consumer is frequently at a loss to select the proper standards of comparison. He is usually content, having secured a glue that fully answers his requirements, to compare each succeeding delivery with this. While this doubtless enables him to check uniformity of delivery, it in no wise answers the question with which he is chiefly concerned; namely, was the original purchase which he now uses for a standard, the best quality obtainable for the price paid? To possess this degree of protection, he must adopt the identical standards employed by the manufacturer in assigning price.
These grades were for many years considered the best made; and competing manufacturers sought to produce glues corresponding with them in all respects. Hence they remain the authentic standards of comparison. It must not be inferred that they are the only reliable make of glue. Many manufacturers produce glues far stronger than the A Extra, and the other grades are regularly duplicated.
It will be readily understood that glues are produced which, in strength, fall between the various grades. No manufacturer is so expert that he can proceed with the definite intention of producing a glue of A Extra or 1 3/8 or No. 2 strength. He can control the strength of the product only by the careful selection of stock and in the treatment throughout the course of manufacture, as outlined in the previous chapter. Once finished, the glue must be compared with the above standard grades in order that it may be classified and a price assigned to it. Hence the value of their employment by the consumer, if he desires to be enlightened upon the all-important factor of price. Furthermore, the peculiar designations of these grades have become the accepted vernacular of the trade. While defying definition, the term 11/2 for example, conveys to the initiated not only the idea of definite strength, but also enables him to form a definite estimate of the limits of price. In addition, the acceptance of the Cooper grades as standards is justified by their unfailing uniformity. 'While the manufacturers are subject to the same limitations as others in attempting to predetermine the strength of their glues, with the result that they produce many that fall between the various grades, these latter do not come into the market under the manufacturer's name, only those that are up to standard in all particulars receiving their special brand. The result is, that the 1+ of today is identical with that of five years ago.
Test Methods. The dry glue should first be carefully examined, for, while the appearance of the glue, save in isolated instances, tells nothing in regard to its strength, such examination is often sufficient to warrant the rejection of the glue, without the necessity of further tests. The following points are worthy of consideration :
1. The surface of the piece of glue should be free from bubbles. These may be merely the result of the incorporation of some air with the glue; but, if sizeable and irregular in outline, they are evidence of the fact that the stock from which the glue was made, and hence the glue itself, is in an advanced stage of decomposition. Of such a glue, the jelly will rapidly putrefy and the solution will have a putrid odor, which, in some instances, may be detected in the glue itself, particularly if the sample be slightly moistened.
2. The piece of glue should break with an even fracture. If the fractured edges are splintery, the glue has not been properly boiled. On the other hand, if a glue breaks under a very slight strain, even though the fracture be even, the glue is either very weak or unduly dry.
3. A good glue should present a velvety or glossy surface, although the absence of this does not always condemn the glue. A "dead" surface may be due merely to the fact that dust settled upon the glue while it was drying upon the nets.
4. The cut of the glue is at times an indication of its strength. If, for example, the glue is cut very thin and is sufficiently flexible to permit of bending almost double without fracture, it is a glue of great strength.
5. If the sample is a flake glue, it is well to compare several pieces in a good light, for even flake glues may be mixed. They are somewhat difficult of recognition inasmuch as the dealer is at great pains in selecting the constituents of such mixtures. Mixed ground glues are, of course, frequently met with. Such as are a mixture of foreign and domestic grades may readily be recognized with the aid of a magnifying glass; the particles of the foreign glue being usually lighter in color and more transparent than the domestic. Mixtures of two or more foreign grades may be detected by noting the differences in color exhibited by the particles. In examining ground glues to detect mixture, the observer must not be led astray by slight differences in color, as in all ground glues some particles are finer than others, and hence lighter in shade.
Sampling. The sampling of the glue preliminary to test is of the greatest importance. The small sample of a few ounces which is usually submitted to the prospective purchaser should be the subject of the above preliminary examination. If it is a flake glue and is seen to be a mixture, a little of each constituent must be taken for test. Where the dealer goes to the extreme of mixing flake glues, he uses a strong glue as a base, and adds enough of a weaker and hence cheaper glue to make the price right. Hence, unless the above precaution is observed in connection with the sampling of a mixed flake, the result of the test is misleading, as the operator may unwittingly select pieces of the stronger glue only, or portions of the weaker alone. It is rare to find more than two constituents in a flake mixture, whereas, with ground glues, as many as four different grades may be mixed together. If the initial sample is a ground glue, even though the preliminary examination reveal no mixture, it is well to shake the package thoroughly before withdrawing the quantity required for test. In the case of mixed ground glues, this precaution must invariably be observed.
Many consumers are content with an examination of the sample and take for granted that the delivery will correspond with it. Such render themselves an easy prey to the unscrupulous dealer. It is manifest that if the consumer is to achieve the full measure of protection, he must test the delivery and note wherein this differs, if at all, from the original sample. Having, furthermore, secured a delivery to his entire satisfaction, he should reserve a generous portion to be used in conjunction with the other standards in future tests.
Whether the delivery consist of a single barrel or a car load, the sampling should be done systematically., Glue is sometimes stored in a damp place that additional weight may accrue to the barrel. In such case, those portions contiguous to the top and the bottom of the barrel contain more moisture than that at the center. Hence, if the glue is ground, even though the preliminary examination does not show it to be a mixture, samples should be withdrawn from several parts of the barrel, including the center, and should then be mixed well together and the requisite test quantity withdrawn from this. This applies equally to flake glues, when the several samples should be comminuted in a small mill and, after mixing the ground resultant, the test sample may be weighed off.
Where the test is concerned with a bundle of gelatine, it does not suffice to select only one sheet from the bundle, but small pieces of each sheet should be broken off ; and where there are a number of bundles, samples selected in this way should be taken from each.
It may be objected that this sampling consumes too much time. Nevertheless, if the test is to be representative of the delivery, the above precautions must be observed in detail. The writer has been frequently called upon to test a car of glue (sixty barrels), of which no two tested exactly alike. The operator is, of course, at liberty to adopt such methods of sampling as will expedite the test; but he must be sure that the sample he tests is an average one, else the whole test is a waste of time. It will save time if, when the glue to be tested is in the flake, the sample be ground prior to soaking. The writer successfully employs for this purpose a small coffee-mill in which the sample may be ground to varying degrees of fineness. This preliminary grinding of flakes often aids in the development of hidden odors. Inferior grades of glue may betray no undue odor when in the dry piece, but, when freshly ground, these become manifest and are subsequently verified in the solution of the glue.
Before describing the test system as a whole, we shall review the methods applicable to the determination ςf the individual factors.
1. Odor of the Glue Solution. The glue having been dissolved to a solution of definite strength (see below), any odors emanating from the hot solution should be noted. The solution should be "sweet." This does not mean that it should be free from any odor whatsoever, as even the solution of the purest gelatine will have a characteristic smell. There should be no trace of putrefaction. If the glue is very greasy, this may impart an odor to the solution, and yet the glue be rated sweet. The odor of the solution will vary according to the stock from which the glue is made. Thus, solutions of common bone glues have a marked "beefy" or "soupy" odor, and yet may be excellent glues of their class. Goat and sheep stocks impart a characteristic odor to the glues prepared from them, while glues from skin or epidermis are more or less odoriferous.
It is difficult to describe an "off" odor in glue, but this requires no extended experience to recognize. Frequently a glue that exhibits a disagreeable odor upon going into solution loses this completely as the solution becomes hotter. Care must be taken not to condemn the glue because of smell, without first considering whether this be not due only to the stock. Odors due to partial putrefaction and other radical defects are permanent, and heating the solution does not dissipate them. Acid-treated glues, particularly those from junk bone, are, as a rule, very sweet, though some smell strongly of acid.
2. Alkalinity and Acidity. Few glues are strictly neutral when in solution, being either acid or alkaline, depending upon the treatment which the stock has received. As previously stated, some manufacturers, after liming the stock, kill the excess of lime by means of a mineral acid. Glues boiled from such stock will be more or less acid in reaction, as will also those prepared by leaching. Where the bulk of the lime has been removed from the stock by simple washing, enough remains to impart to the solution of the finished glue an alkaline reaction.
Sufficient acid or alkali is present to react with litmus, and hence either condition is conveniently determined by dipping, first, a strip of blue litmus paper into the hot solution, and subsequently a strip of red litmus paper. If the glue is acid, the blue litmus strip will be more or less reddened and the red strip unaffected; conversely, if alkali is present, the blue strip will be unaffected while the red strip acquires a blue coloration varying in intensity with the amount of alkali. The test is best carried out by immersing the dry strips in the hot glue solution and, after keeping about one half the strip submerged for thirty seconds, withdrawing and draining the excess, of glue solution by wiping both sides of the strip on the edge of the vessel containing the solution. If too much extraneous solution be permitted to adhere to the strip, it is difficult to observe the change of color.
Excess of acid or alkali in glue is of disadvantage chiefly where the glue is to be employed in sizing delicate colors, or insoluble lake pigments ("pulp colors"), as in the wall paper and surface-coated paper industries. Many of these pulp colors, particularly the cheaper grades, are poorly fabricated, the excess of material employed to precipitate the aniline or coal-tar color upon the insoluble base having been only partially washed out. As a result, trouble frequently arises in the wall paper factory, owing to the inter-reaction of the acid or alkali in the glue and the excess of precipitant in the colors. In the same industry, difficulty is at times experienced owing to lack of smoothness in the mixture of glue and clay used for the initial grounds. Excess of acid or alkali in the glue employed for sizing the clay often has a disintegrating effect upon it, the mixed clay and glue solution becoming lumpy and leaving the rollers of the machine in irregular patches instead of with a uniform flow.
Unless acid or alkali is present in large excess, it may be disregarded save as indicating, to a certain extent, the treatment which the stock has undergone. It has already been noted that acid glues keep longer than alkaline ones, as the bacteria of decomposition reproduce far quicker in alkaline media than in acid. If the glue has a rather earthy smell, and the solution turns red litmus deep blue, the stock has been over limed. Such a glue should be rejected as the excess of alkali in this case will gradually weaken the glue, causing it to rot rapidly. Although the appearance of over-limed glue may frequently give indication of strength, upon test this will be found to be lower than was anticipated, owing to the fact that the lime has already radically weakened the fiber of the glue.
3. Grease. The presence of grease in glue has always met with strenuous objection, more or less justified, upon the part of the consumer. As a matter of fact, its presence is of material advantage for some purposes and of equal disadvantage for others. In any one make of glue, the amount of grease will vary considerably in the course of, say, a year. As has been pointed out, there are times when the manufacturer relies upon the recovery of the bulk of grease in order to reduce the cost of production. At others, when the market value of the recovered and rendered grease is so low as not to warrant the additional labor-cost attendant upon its complete removal, he permits a goodly proportion to remain in the liquors, and this is subsequently detected in the finished product.
A simple and accurate test for grease is effected by placing upon the corner of a sheet of white paper a few grains of some aniline color such as methyl violet, magenta, or any other readily soluble in water. A clean flat brush, one or two inches in width, the edge evenly trimmed, is dipped in the glue solution, which must be hot (170180° F.) so as to insure a good flow of the solution, which is apt to cool rapidly upon the surface of the paper. The brush is withdrawn from the solution, hastily drained, and then applied to the color with a stirring motion so as to dissolve all. Without lifting the brush from the surface of the color, it is drawn with a rapid sweep across the paper. Any grease will manifest itself in the form of "eyes" or elliptical spots which soon disappear. Bubbles of air,arising from faulty manipulation of the brush, must not be mistaken for grease spots. The more color used, the better the grease will show. Excellent results may be obtained by keeping the brush always firmly pressed against the surface of the paper in dissolving the color as well as in the subsequent "painting out." The best results of all can be had by using lampblack in place of aniline color; but this has the disadvantage of being insoluble in water and hence it is difficult to , clean the brush properly for the next test.
Glue at times contains free fatty acids resulting from the destruction of lime soaps by mineral acids. These impart a more or less acryline odor to the hot solution. Their presence is sufficient to condemn the glue, as they are markedly hygroscopic, absorbing sufficient atmospheric moisture to radically retard drying of the solution.
Of the various glues, that made from sheep stock is the greasiest. Next in order comes glue made from goat; and glues from ox or cow hides are more or less greasy, depending upon the care exercised in the removal of fat. Acid-treated glues seldom exhibit more than a trace of grease.
In the manufacture of "glazed" or surface-coated papers, greasy glue is undesirable, for reasons that are manifest. Since the glue is incorporated as a size or binding medium with the colors, any grease would show upon the surface of the paper just as in the test. For similar reasons, greasy glue is regarded unfavorably by the manufacturer of wall papers. It is claimed that the grease is apt to penetrate the paper, making typical grease spots. The objection is not a good one; as the formation of such spots would presuppose the presence of an enormous excess of grease to penetrate both paper and clay ground. On the contrary, the presence of even a moderate amount of grease is of advantage. It not only prevents the glue from foaming in connection with the clay or colors, but serves to materially brighten the latter even though it may somewhat retard the drying of the printed paper. In the majority of instances where the glue solution is applied to the material to be sized by means of a machine roller, the presence of a small amount of grease assists, rather than hampers, the operation, preventing foam and insuring smoothness of flow from the roller.
4. Viscosity. As applied to solutions of glue and gelatine, the term viscosity comprehends their rate of flow as compared with that of a standard fluid, viz., water. Just as two solids, two blocks of wood, for example, moving against each other, exhibit more or less friction or resistance to movement, according as their surfaces are smooth or rough, so do fluids manifest friction or resistance to movement, not only towards the solid surfaces over which they flow, but also between their constituent molecules. Viscosity has its origin in the ease, greater or less, with which these molecules pass or roll over one another. It is exhibited by all liquids to a varying degree. If a vessel of water be given a slightly rotary motion, so as to cause its contents to oscillate, it will be observed that with each swing the amplitude of the oscillation becomes smaller, the oscillation gradually ceasing and the surface of the liquid coming to rest. If oil or glycerine be substituted for the water, and the experiment repeated, a much greater resistance to movement will be observed, due to the viscosity or internal friction of the fluids.
The measure of the viscosity of the glue solution is important. It has been adopted by several prominent Western manufactories as the sole means of determining glue strength. Of itself, it is insufficient for this purpose, but, considered in conjunction with jelly strength, it is of great value in grading the glue. It is an arbitrary constant and is of value only when compared with the corresponding measure of a standard fluid. Thus, to say that the viscosity of a glue solution is 35 seconds conveys absolutely no meaning; but from the statement that the viscosity is 35 seconds, when, under the identical conditions, the viscosity of water is 15 seconds, a definite estimate of the "body" or rate of flow of the glue solution, can be formed. Water is advisedly the standard of comparison for viscosity, since it is the medium for extracting the glue stock, and the strength of the glue is in large measure dependent upon the amount of gluey material that has combined chemically with the water. It must not be inferred from this that the higher the viscosity, the stronger the glue. This is true only of certain kinds of glue.
Owing to the importance of determining the viscosities of lubricating oils, several elaborate forms of viscosimeter have been devised, such as the Engler, Sayboldt, Hurst, Coleman Archbutt, Kunkler and Red-wood. These may be employed in the determination of the viscosities of glue solutions; but they supply figures that are inconveniently high and have, further more, been standardized by means of oil and not water.
Satisfactory results can be obtained with much simpler apparatus.
The simplest form of viscosimeter consists of a volumetric pipette, of about fifty cubic centimeters capacity. This is shown at Fig. 1. By means of a a file, the delivery end is cut off at a and the aperture carefully reduced in the Bunsen flame, the object being to procure an instrument that will deliver water from the mark b to the aperture c in. exactly 15 seconds, measured by a stop watch. Any reliable dealer in chemical apparatus will, at low cost, make such a pipette. Should the operator himself desire to standardize the instrument, he proceeds as follows. Having shortened the delivery tube as directed, the FIG. aperture is first slightly reduced by holding it in the flame, giving it a rotary motion so as to preserve the circular perimeter. A trial had best first be made with a piece of glass tubing, in order that the necessary precautions may be studied. The aperture is at first only slightly reduced, when, having cooled, the pipette is filled and adjusted to the mark b. In filling a pipette, it is held by the upper tube bd between the thumb and middle finger of the right hand, leaving the index finger free to be placed subsequently at d. The point c is immersed in the liquid and the lips applied at d, the liquid being drawn by gentle suction to some point between b and d. As the lips are withdrawn, the index finger is instantly placed at d and pressed firmly against the tube end so as to maintain the level of the liquid. By gently releasing the pressure of the finger, thι liquid will flow to b and is held at this point by firm pressure of the finger. The operator now takes the stop watch in his left hand, using the thumb to release the movement of the watch. With the pressing of the spring, he simultaneously releases the contents of the pipette by removing his finger, and as the last drop of the contents passes c he closes the movement of the watch. The time required for the contents (water, free from sediment, at about 180° F., is used for the standardization) to flow from b to c is registered by the watch. If the operator has proceeded carefully, this will be 12 or 13 seconds on the first trial.
It is now necessary to close the aperture c still more. The pipette is thoroughly dried (carelessness in drying will inevitably result in the breaking of the instrument) and the point c is held in the flame with the usual rotary motion until it just reddens. Care must be taken not to close the aperture too far as otherwise the point must be broken and the entire operation begun over again. Having cooled, the pipette is filled and adjusted as before, and the time of delivery again noted. On this trial it will be found to be nearly correct about 14 seconds. The greatest care must now be exercised in so adjusting the aperture that the time of delivery will be exactly 15 seconds. The pipette is again dried and the point c held in the flame a few seconds only at a time, repeated trials of delivery being made between heatings. When the time of delivery between the points b and c is exactly 15 seconds, the instrument is ready for use.
Much labor may be saved by the inexperienced operator, if he will first study the principles of manipulation by using a piece of glass tubing. In this way he will be able to see just what is necessary in properly reducing the aperture of the pipette. The selection of a pipette of 50 c. c. capacity supplies figures that are most convenient in use. The time of delivery is purely arbitrary, and may be any larger figure than 15 seconds. This, however, is the most convenient. Pipettes of larger capacity, supplying larger figures for viscosity, are awkward to handle and the contents cool. perceptibly during efflux, the resulting figures being inexact. By the system recommended, it will be found that the lowest viscosity exhibited by a 25 per cent glue solution is 151 seconds, or one half second greater than that of water under the same conditions; while the highest will rarely exceed 70.
An improved form of viscosimeter is shown at Fig. 2. The instrument is of 50 c. c . capacity and is provided with a glass stop-cock s, which obviates the necessity of controlling the contents with the finger as in the above form. The points of flow are a, the lower surface of the stop-cock, and the aperture c. The instrument is standardized in the same way as the plain pipette. After filling with the glue solution and adjusting to the mark, the upper tube is firmly clamped to a support, leaving the operator free to release the watch movement and open the cock simultaneously.
Where great accuracy in the determination of the viscosity is desired, the following form of viscosimeter, devised by the author and used by him throughout his tests, will be found to give satisfactory results. The instrument is shown at Figs. 3 and 3a. It consists of the viscosity pipette, A, and the bath B for maintaining the contents of the pipette at constant temperature. The pipette is provided with the stop-cock S for controlling the contents, and the time required for the contents to flow from x to c is exactly 15 seconds.
The bath B is made of seamless copper and consists of an outer compartment D which serves as a water-jacket, and an inner cylinder in which the pipette A snugly fits. This cylinder projects some distance below the jacket D, and is provided with a window, F, made of thin sheet mica, so that the aperture, of the pipette may be clearly seen.
At Fig. 3a the apparatus is seen set up and ready for use. The bath B is supported by the ring G of the ring-stand. Some distance below is adjusted the ring-burner E, by which the water in D is brought to the desired temperature recorded by the thermometer T. The pipette is shown snugly fitting in the inner cylinder, ready to discharge the glue solution, y. The aperture c of the pipette is visible through the window F. The contents of the pipette are returned,to the glass, H, in which the glue was originally dissolved.
The pipette having been filled and the contents adjusted to the mark, it is placed in the cylinder and the FERNBACHIS VISCOSIMETER FIG. 3 a water in D brought to about 185° F., making the temperature of the contents of A practically 180° F. Simultaneous with the release of the contents, effected by giving the stop-cock a sharp turn, the watch, held in the left hand, is started, and the moment the last drop has passed c the watch is snapped and the time of efflux between x and c noted. This is the viscosity of the glue solution, as compared with water, standard = 15. The viscosity may also be expressed in terms of water as unity, by dividing the number of seconds for the glue solution by 15. Thus, if the glue solution has a viscosity of 27 seconds, this may also be expressed as 1.80 referred to water as unity.
Certain precautions are to be observed in taking the viscosity. All the glues to be compared must be at exactly the same temperature (180° F.). Furthermore, they must be of exactly the same strength. It is obvious that two solutions of the same glue, one 10 and the other 25 per cent, will exhibit widely divergent viscosities. During the test, the temperature of the glue solutions should not exceed 180° F., as the protracted heating is apt to decrease the jelly strength. Again, the operator must practise until he is able to release the contents of the pipette and start the stop-watch simultaneously. He must keep his eye constantly directed to the point c that he may be prepared to snap the watch the instant the last drop passes that point. Practice for a short time will enable him to execute the test with absolute accuracy.
Value of Viscosity as a Test Factor. By some, notably Western, producers, viscosity is taken as the ultimate measure of glue strength. This is a grave error. Viscosity is of importance in the glue test as it indicates the treatment the stock has undergone; and is, further, a reliable index of the value of the glue as a sizing medium. It must, however, be considered side by side, with other factors. The assumption that the strength of the glue is in direct ratio to the viscosity of its solution is true only of certain glues. Those prepared from the same stock, under identical conditions, may exhibit viscosities varying directly as their strength; but these constitute the exception rather than the rule. The fallacy of rating glues according to viscosity alone, will be readily seen from consideration of the following. It is assumed throughout that the glues are in 25 per cent solution, and that their viscosities at 180° F. are referred to water = 15 seconds.
It will be observed that the higher grades exhibit greater divergence in viscosity than the lower. Such is the case with the majority of glues.
Assume, now, that the above are being employed as standards, and that the operator relies solely upon the viscosity as the ultimate measure of the glues he is testing. Of four glues on test, he finds, let us say, the viscosities to be 35, 26, 22 and 19 seconds, respectively. His inference is that the first tests No. 1, the second 11/4, the third 1 1/2, and the fourth 11. If the glues he is testing were made from the same stock and under the same conditions as those in the above table, his inference would be correct; but unless such were the case, his report would be entirely wrong.
Acid-treated bone glues, as well as those from acid-treated hide stock, exhibit viscosities remarkably low in proportion to their jelly strength. Such glues, of A Extra strength, may have as low a viscosity as 30 seconds, and acid-treated glues of 1 3/8 strength, with viscosity as low as 18 seconds, are common. Hence, had any of the above glues on test been derived from acid-treated stock, for example, the one with viscosity of 19 seconds, its jelly strength might be 1 X and not 1 3/4. Now, suppose this glue had been offered at 10 cents per pound. The operator, rating it as a 1 3/4 glue, would at once have decided that the price was ridiculously high; whereas, had he rated it properly as 1 X, he would have realized that he was offered a bargain.
The viscosities of opaque and colored glues are always slightly higher than those of clear glues of corresponding strength (compare 1 X molding with 1 X clear in the above table of standards). Hence, if viscosity is to be taken as the ultimate measure of strength, it is necessary to provide separate standards for comparing opaque or colored glues, separate standards for clear glues, for hide glues, for bone glues, and for acid-treated glues. Even if this were done, the results would not always be accurate, .as will be understood from the following considerations.
If the glue stock has been incompletely washed, permitting mucin to be present, or if it has been overlimed, with the result that soaps are present in the finished article, the viscosity will be disproportionately high. The author has seen glues which, while testing only 1 3/8, have had as high a viscosity as 30 seconds. Others, testing A Extra, had viscosities ranging from 59 to 69 seconds. These were hide glues from improperly washed or over limed stock.
The majority of foreign glues, and certain grades of domestic, having been clarified, are low in viscosity. This applies to both bone and hide glues. If the clarification has been effected by means of alum, the viscosity is very high, as alum renders the glue solution stringy. This is to be observed in the case of rabbit glues, the viscosities of which would mislead the observer as to their strength. Gelatines, because of the special stock from which they are made, and the special treatment which this undergoes, are comparatively low in viscosity.
Table I, compiled from the author's laboratory note-book, will convey some idea of the variation in viscosity shown by gelatine, hide glue, acid-treated and regular bone glues, of the same jelly strength.
Table II emphasizes the fallacy of relying solely upon viscosity in rating the glue. It is seen that the gelatine with a viscosity of 49 seconds is two grades stronger than the hide glue with the same viscosity. Similar discrepancies are to be noted throughout.
The exact value of viscosity as a test factor will be better understood after the reader has studied the discussion of the test system as a whole, in the succeeding pages.
5. Foam. Foam, in glue, arises from the incorporation of minute bubbles of air with the solution, when this is beaten rapidly. There may be present in the solution, substances which render the emulsion more or less permanent; or the emulsion may be only temporary, the foam receding and disappearing in a few moments. The same defects which fictitiously increase the viscosity, contribute materially to the foam in glue. Thus, over-limed or poorly washed stock will retain mucin or soaps which make the solution foam badly. Glues from alum-clarified liquors foam, as do also cheap bone glues, because of the impurities they contain. If, however, there is grease present to any extent, it will prevent the foam emulsion from forming. This applies to all makes of glue.
Foam, although undesirable in glues, is often demanded in gelatine, particularly that used for the manufacture of confectionery. In making marshmallows, the solution of gelatine, combined with other ingredients, is beaten rapidly in a special machine.
The test for foam is executed as follows: The hot solution previously employed for the odor, grease and viscosity tests is transferred to a stout glass vessel, wide enough to accommodate the usual size of rotary egg-beater. This vessel should be neither so. thick that the heat of the solution will crack it, nor yet so thin that it is apt to be broken by the egg heater. On the side, a mark should be placed at the level of the glue solution, and above this three others, a half inch apart. The egg-beater is slowly rotated in the solution for half a minute and the quality as well as the quantity of foam noted. If this is only slight and disappears in a moment, the glue is non-foaming; if somewhat more permanent and reaching the second line, the glue is slightly foamy. If the column of foam reaches the third mark and takes some time to recede, it is moderately foamy; and so on. The operator should agitate the solution neither too rapidly nor too long, as, under these conditions, any glue solution will emulsify. The test conditions should be approximated to those obtaining in actual work. Two revolutions of the egg-beater, per second, will suffice for the test.
6. Jelly Strength. This is the most important of all the test factors and, unless considered in conjunction with this, the others are of only limited value. The stronger a glue, the greater the resistance offered by its jelly to outside pressure. The statement that a glue tests 1 1/2 means that its jelly offers the same degree of resistance to pressure as the jelly of the standard 1 1/2 glue. Test methods seek to measure this resistance in definite terms.
So many divergent methods for determining glue strength have been formulated, that lack of space forbids their discussion here. Those dependent upon chemical operations are absolutely worthless, as has already been noted. Those involving the use of special apparatus supply results that in no way correspond with the commercial value of the glue. Of these, the test with the so-called "shot machine" is worthy of some notice, although the method is open to considerable criticism.
The shot test was formulated by Lipowitz, whose apparatus, has since received considerable modification. The test has for its object the determination of the weight sustained by the glue jelly. Upon the surface of the jelly rests a saucer-shaped piece of tin. The iron rod connecting this with the funnel designed to hold the shot passes through and is supported by the tin cap which rests upon the edges of the glass containing the jelly. The combined weight of the saucer, rod and funnel does not exceed half an ounce. Into the funnel, shot is poured until the saucer just penetrates the jelly. The shot is then weighed.
It will be seen that no account is taken of the friction between the parts of the apparatus, as well as that between the saucer and the jelly. This objection holds better with the more modern forms of shot apparatus in which the rod is permitted to sink through the entire jelly and rest upon the bottom of the glass. Again, the formation of a skin on the surface of the jelly frequently invalidates the test, as the weight determined is really that required to pierce this skin and the jelly, plus that required to overcome the friction between the penetrating rod and the skin and jelly. Lipowitz's apparatus is undoubtedly the best form as it is required just to penetrate the surface of the jelly. For certain classes of glues it is a fairly accurate test. In practice, however, the comparison, by its means of a clear glue and an opaque of identical strength, yields divergent results, as the skin on the opaque glue is slightly tougher than that on the clear.
For practical commercial purposes, there is no better method of measuring the resistance of the glue jelly than by means of the finger. The fourth finger of the left hand is used, as it is the most sensitive of all. In operation, the test is as follows: The glues, having jelled, are arranged in a single line on the table before the operator. Let us assume that these are nine in number, five unknown glues and four standards, the latter being 1 1/4, 1 1/2, 1 1/4 and No. 2. The glues are in numbered glasses, the unknown being numbers 1, 2, 3, 4 and 5; and the standards, numbers 6, 7, 8 and 9. They are arranged in numerical order, running from left to right. The operator first assigns a preliminary rating. To effect this, he first presses the finger upon the jelly of No. 1 and then on that of No. 2. Any difference in the resistance of the jellies to the pressure of the finger is at once noted. If No. 2 offers less resistance than No. 1, the glasses remain as they are; but if it is seen that No. 2 is the stronger, it is shifted to the left of No. 1. If the order of the glasses remains unchanged, the operator next compares Nos. 2 and 3; but, if 1 and 2 have changed places, he compares 1 and 3, the stronger jelly always being placed to the left of the weaker, and so down the line. In this way the glasses are finally so arranged that the strongest jelly of all is at the operator's extreme left and the weakest at his right, the intermediate glasses representing successive steps in strength, calculating from right to left.
Now, since by careful rating the operator found that the strength of No. 5 is equal to that of No. 6, sample No. 5 must test 1 1/4 Since No. 4 was found to be halfway between 6 and 7, that is to say,- halfway between 1 1/4 and 11/2, sample No. 4 must test 1 3/8. No. 3 was found to be halfway between 1 1/2 and 1 3/4. Hence it tests 1 5/8. Nos. 1 and 9 being equal in strength, the former must test No. 2 grade; and as No. 2 was found equal to No. 8, it is of 1 3/4 test.
Doubtless the reader is now able to appreciate more fully the importance of standards in glue testing. If these be incorrect the whole test is worthless, no matter how carefully conducted.
We cannot here enunciate any. set rule governing the price of the various grades. Much depends upon cost of stock, labor, etc.; but if we assume that the above glues were offered for sale at a time when market conditions were normal, several valuable inferences may be drawn from the above hypothesis. In the first place, a 1 1/4 glue at 14 cents per pound is a luxury; all the more, a 11 glue (four grades weaker) at 12 cents. The prices assumed for the other grades in the table are about .right. It is evident that the manufacturer is either endeavoring to impose upon the consumer in the cases of No. 5 and No. 3, or else he is compelled to ask the suggested prices for those glues, owing to the fact that his yield was lower than anticipated while the cost of production was unusually large. We may simplify the entire hypothesis by assuming that the consumer applied to the manufacturer for samples of glues at 12 cents per pound, this being the price he was accustomed to pay for his glue, and that the manufacturer accordingly submitted the above five samples, all quoted at this price. Because of the test, the consumer is speedily enabled to decide upon a choice.
Simple as is this finger test, it may be made, by dint of practice and experience, remarkably accurate. The author has carefully compared the results obtained by it with those furnished by the use of the so-called testing machines, and the verdict is decidedly in favor of the finger test. To insure its success, the operator must, of course, be entirely free from bias. The glasses should all be numbered and the numbers identified only after the glues have been carefully rated. In this way, the test is a fair one. Its accuracy, of course, presupposes that all the glues tested have been treated exactly alike. This is essential to the success of any system.
THE TEST SYSTEM AS A WHOLE
A. Apparatus Required.
1. Glasses. The bar glasses commonly known as "sours" are best adapted to glue testing. These may be had of any dealer in hotel or bar supplies, and are much more durable than beakers. Only those having fairly thin bottoms should be selected, and to minimize danger of cracking while in the melting bath, the glasses should first be tempered. This is done by filling them with cold water, placing them in cold water in the bath,and slowly raising the temperature of the bath to boiling. The glasses are then removed and the contents permitted to cool to room temperature. They are then freshly filled with cold water and the operation , repeated, save that this time the hot contents are immediately thrown out and the empty glasses held in a stream of cold water from the tap. Some will break, but those that survive the treatment will seldom crack through alteration of the temperature of the bath, and will be of service for a long time.
2. Glue Melting Bath. This consists of a trough, preferably of galvanized iron, provided with a false bottom resting about two inches above the true, and perforated with one-inch holes. Upon this the glasses rest, and in this way cracking from unevenness of flame is avoided. The sides of the bath should be of such height as to correspond with that of the glasses as they rest upon the false bottom. The bath should accommodate thirty-six glasses.
3. Scales. A fairly accurate balance is required for weighing out the samples. The pharmacist's prescription balance serves nicely, although less expensive forms of apparatus may be employed.
4. Viscosimeter. The different forms have already been described.
5. Egg-beater. This is required for the foam-test. It should be of the rotary form provided with a side wheel. For this test, there is also required a fairly stout glass jar, just wide enough to accommodate the egg-beater.
6. Brush. For the grease test, a flat brush, 1 1/2 inches wide and evenly trimmed, is required. In addition, some aniline color should be on hand.
1. Selection of Standards. Extended experience in glue testing will enable the operator to form some preliminary estimate of the strength of the glue to be tested. He cannot say with any degree of certainty what grade it tests, but he can roughly gage the limits within which it lies. In this way, the experienced tester has only to use the minimum number of standards. As a rule, the eleven standards should be placed on test with the unknown samples, be the latter few or many. If the glues submitted be high-priced, it is to be inferred that they are high-grade, and accordingly, such standards as A Extra, No. 1, 1 1/4 should be selected for the comparison. Where any doubt exists, however, it is safest to employ all eleven standards.
2: Weighing out the Glues. The samples and standards having been numbered, the glasses are numbered to correspond and placed in single file in numerical order, beginning at the left. Twenty-five grams (or one ounce) of each of the samples and standards are then weighed off and transferred to the glasses. If the weighing is done carelessly, the entire test is invalidated. Exactly the same weight of each glue must be taken.
3. Softening, or Soaking the Glues, No hard and fast rule can be formulated as to the time required for this, much depending upon the form of the dry glue. If in the flake, strip, or piece, and very thick, from six to eight hours is required. If ground, one half an hour may suffice. The best way is to let the glues soak over night and melt them for test the following morning.
As the glues are transferred from the scale-pan to the glasses, each glass is provided with, a stirring-rod.
One hundred cubic centimeters (or four ounces) of water are measured out with the utmost care and poured over the glue, and the latter stirred with the rod so as to insure uniform immersion. Exactly the same amount of water must be poured on each glue not a drop , more or less.
4. Melting the Glues. When the glues are properly softened, the glasses are placed in the bath. This is then heated, beginning with a small flame. Simultaneously, the flame is started under the viscosimeter bath. Lest any skins form on the solutions while melting, these must be frequently stirred. The formation of a skin will seriously interfere with the test. Further, the solutions must be uniform, else the figures obtained for viscosity will be inaccurate. The temperature necessary to the determination of viscosity has already been noted.
In seeking the various factors, the following order is most convenient :
(a) As the glues are melting the odor of the solution is noted.
(b) When all are in solution, Acidity or Alkalinity is determined.
(c) Next, the grease test is applied to each.
(d) By this time the glues are sufficiently hot for the viscosity test.
From now on, care must be taken to prevent skin formation. Where a plain pipette is used for determining viscosity, this must be thoroughly rinsed in boiling water between each test.
(e) The viscosity determined, the foam may now be estimated.
(f) The glues are now removed from the bath and permitted to jell, when they are rated.
5. Jelling or Cooling the Glues. This requires especial consideration, lest faulty practice invalidate the test. In the winter time, it suffices to let the glues remain a while in a cool place until jelled. They must never be allowed to freeze. The greatest difficulty in jelling the glues is experienced in hot weather, when a small refrigerator may be used. In this, the temperature is about 40° F., which is just right for jelling the glues without interfering with the rating. When the glues have been removed from the bath, they should be stirred slowly for a time, and in this way partially cooled. If this is done, the final jelly will have a flawless surface which will aid the operator in the rating.
It were superfluous to state that, throughout the entire test, the glues must be treated exactly alike. Unless this is done, it is useless to go to the trouble of testing them.
C. Recording the Test.
Some record of the test must be kept for future reference. Such records are apt to be cumbersome. In the following system the record is concise and contains full data concerning all points of interest. It serves, among other things, to describe the appearance of the dry glue, long after the original sample has been thrown out. To effect economy of space, the description of all test factors is reduced to a system of mnemonics, to wit: In describing or defining odor, grease, foam, alkali or acid, 1 = None; 2 = Slight; 3 = Moderate; 4 = Considerable. In describing the appearance of the glue jelly,1 = Clear; 2 = Translucent; 3 = Dense.In describing the form of the dry glue,1 = Flake; 2 = Ground; 3 = Strip; 4 = Sheet; 5 = Powdered;6 = Shred; 7 = Ribbon.
In describing the color of the glue,1 = Light Yellow; 2 = Medium Yellow; 3 = Dark Yellow; 4 = Light Brown; 5 = Medium Brown; 6 = Dark Brown; 7 = Opaque; 8 = Colored; 9 = Colorless.
In describing the cut of the glue, 1 Very Thin; 2 = Thin; 3 = Medium; 4 = Thick; 5 = Very Thick. The following is a specimen test record used by the author. The various columns record the name of the sample, viscosity, odor, grease, foam, alkali, acid, appearance of the jelly, and strength. In the last column, the litmus strip which has been used to determine the reaction is pasted by means of the tested glue adhering to it. In the upper left-hand corner of the space for recording the name of the sample are placed small numbers, describing the color, cut, and form of the dry glue. Thus, No. 1, KA, is a dark yellow, ground glue; No. 2, II, is a medium yellow, thin flake. Samples Nos. 6 and 7, which are gelatines, are described as medium yellow, medium thick sheets.
The value of keeping test records in this form can hardly be overestimated. A year may elapse from the time the operator has tested a certain glue when he is again called upon to test it. By referring to the original test, he may readily compare the two samples. In a separate index, the author notes the number of the test sheet upon which a given glue appears. For example, against the entry J.B.O. is the number 47, which is that of the above test sheet.
Some slight modification of the above test system is to be made in testing gelatines. Here the observer is concerned with the amount of water absorbed by the gelatine, its odor, reaction, grease, and foam. He may omit the determination of jelly strength. On the other hand, this jelly strength is directly proportionate to the amount of water absorbed, and if the sample of gelatine is tested in the same way as glue, the factors are easily determined.
A study of the viscosity column of the above test record, will confirm what has already been said relative to the inadequacy of viscosity as the ultimate measure of glue strength. Compare the viscosities of Nos. 6 and 7, which are high-grade gelatines and which test two grades stronger than A Extra, with the viscosity of No. 15, the 1 Extra standard. Had we relied solely upon their viscosities to rate these two, we should have placed their strength at only 1 Extra.
The author has tested more than thirty thousand samples of glue and gelatine, using no other test method than the one here described and relying invariably upon the finger test for the ultimate rating of the glue, in preference to using any test machines. The results have always been concordant. Proceeding carefully, two operators will differ only by two points within a grade, this difference being attributable to personal equation. So long as all the glues are treated alike and the operator is careful to eliminate personal bias, the test is an accurate one, the results agreeing closely with variations in price of the glues.