CHEESE-CRATE TESTS.

New Zealand Journal of Agriculture, Volume XXXIV, Issue 1, 20 January 1927, Page 19

CHEESE-CRATE TESTS.

DEVELOPMENT OF BALANCED CONSTRUCTION TO MINIMIZE BREAKAGE DURING TRANSIT.

A. R. Entrican,

Engineer in Forest Products, and W. C. Ward, Forest

Assistant, New Zealand State Forest Service.

Complaints have been received from time to time relative to the damage to cheese which takes place in export service, due to the failure of the containers used. This article presents the results of tests made by the Forest Service on behalf of the Dairy Division of the Department of Agriculture, with the object of improving the serviceability of cheese-crates at present used in the export trade. The work carried out is similar to that described in the article entitled “ Butter-box Tests,” published in the issue of this Journal for May last. The laboratory studies here described combine practical experience—which is a knowledge of the designs in use, of what timber is available, and of crate-manufacturing practice —with actual scientific tests made on the package itself, packed as in actual service and subjected to strains that approximate actual transportation conditions. The main purpose of, the study was to develop a balanced and economical construction —that is, a crate which has enough strength in each part for the purpose for which it is intended, and no more strength in any part than is necessary to balance the average strength in any other part. The essential qualifications of an export cheese package are (i) that it be strong enough to stand up under exceptionally rough handling ; (2) that it be able to resist punctures from the corners or edges of other containers ; (3) that it occupy a minimum of space ; and (4) that it be difficult to open or close without special tools —a preventive of concealed pilfering. It is necessary to secure these four qualifications without a burdensome cost. Although the immediate purpose was to formulate a specification for a standard crate for the export trade, the scope of the work was extended to include a study of the various types of containers now in common use, and to provide data for the general instruction of crate manufacturers and users regarding certain fundamentals of crate design. The study may be still further extended at a later date to the investigation of other types of cheese-containers which appear to promise improvement upon existing packages. Material tested. STANDARD CRATE. The standard cheese-crate now in common use for the export trade carries two cheeses weighing approximately 80 lb. each. The crate is an approximately cylindrical package, having two compartments, as shown in Fig. 1. It is constructed of two ends and one centre-piece, with these twelve-sided single pieces connected together by twelve battens approximately 27|in. in length.

Most of the .tests described in this report were made upon packagesapproximating to the foregoing standard, the length of all the battens used being kept constant at 27 in. , One small series of tests was made upon a crate designed to carry only one 80 lb. cheese. The battens, of this crate' are I3|in. in length.

The distance between parallel sides of ends and centres was kept constant at ,15 in. for both standard and single cheese-crates. In assembling the standard crates the . directions of the grain in the two ends are placed at right angles, the grain of the centre running diagonally between that of the ends. Each batten in the standard crate is attached to each end and centre by two smooth wire nails, if in. long. The whole crate is reinforced by three 14 B.W.G. soft iron wires, fastened around the centre and each end by fin. staples driven into the ends and centre between all battens.

Many failures or breakages in cheese-crates having been attributed to poor nailing and .reinforcing, tests were accordingly made to determine the effect of substituting cement-coated hails for smooth wire nails, and flat nailed strapping and Acme flat nailless strapping in tension for the reinforcing-wire. “ MAT ” CRATE. As the standard crate is assembled and opened only with considerable difficulty, tests were made of other types of construction which appeared to minimize this disadvantage. One of these —the “ mat ,r crateis shown in Figs. 2 and 3. Fig 3 shows a section of a “ mat ,r delivered from a stitching or fabricating machine and ready to be nailed to the ends and centres. The battens are held to the wooden hoops by means of staples. The crate is assembled by folding the mat around the ends and centres, and placing another wooden hoop exactly

over the hoop to which the battens are stitched. Nailing through the hoops is commenced in the centre of the mat, and each succeeding batten nailed, working round to half-way, when the cheese is placed in position and the nailing completed. Fig. 2 shows the assembled ■crate.

“ CLEATED ” CRATE. Standard-sized and single cheese-crates similar to those shown in 'Figs. 4,5, and 6 were also tested. These crates are constructed in two halves, which are assembled around the cheese or cheeses and fastened together by cleats at the ends and reinforced by Acme flat metal strapping. GENERAL. Five crates were tested to study each variable. The results in all •cases were consistent enough to give a reliable average based on this

number of tests. The recommendations.of the Madison Forest Products Laboratory, U.S.A., were followed in the nailing and strapping of crates. Only 5D and 7D cement-coated nails were procurable from local stock. It was therefore' necessary to use these in some cases where 6D would have proved more suitable. Flat iron nailed box-strapping, fin. wide by 0-015 in. in thickness, and No. 14 B.W.G. iron wire, both of soft (f.g., annealed) metal, were used on crates bound with nailed or stapled bindings. - Acme flat metal strapping, in. by 0-018 in. and fin. by 0-015 in., both of hard (i.e., unannealed) metal, were used on crates bound with nailless bindings. Note. —The size of cement-coated nails is expressed in pennies and designated throughout this article by the letter " D.” The dimensions are as follows: 4 penny (4D), if in. long, 14 A.W. gauge 5 penny (SD), if in. long, 131 A.W. gauge; 6 penny (6D), if in. long, 13 A.W. gauge; 7 penny (7D), 2| in. long, 12J A.W. gauge. Cement-coated nails are designated by letters “ c.c.” ■ Methods of Test. TESTS OF CRATES. The most practical method yet devised for testing wooden packages is the revolving-drum test. For this purpose the machine shown in Fig. 7 was installed by the Forest Service at its timber-testing station maintained at the School of Engineering, Canterbury College, Christchurch. The drum is a hexagon-sided machine, and revolves slowly at a rate of if revolutions per minute. The crate to be tested is packed with cheeses of ordinary weight, as in commercial service, and

placed in the drum. In the drum are arranged a series of hazards which cause a crate to follow a regular cycle of drops, falling on battens, ends, and edges, and flatwise upon a projection similar to the corner of a box. These drops simulate the usual hazards of transportation. Each face of the drum is counted as one drop. To facilitate the recording of locations and character of failures the battens and ends of the crates were numbered. As the crate moves on from one drop to the next the observer notes the beginning of the failure of the weakest point of the construction, and follows its development and that of other weaknesses until the crate fails and spills its contents. The weak features of the crate may be too few or too short nails, thin battens or ends, or insufficient reinforcement in certain parts. Such weaknesses are studied by the observer until he is able to build up a crate having equal strength in every feature and capable of delivering its contents in the same condition as hen first shipped. TESTS OF REINFORCEMENTS. The mechanical tests of the bindings and fastenings were made at the Forest Service timber-testing station maintained at the School of Engineering, Auckland University College. . Six tests of each type of reinforcement were made to determine the maximum tensile strength and percentage elongation of the metal used. As only the maximum tensile strength of the. stapled-wire and nailed strapping was required, specimens were limited in length to io in. between grips. In tension strapping a low percentage elongation is necessary, and for this determination specimens 60 in. long between grips were tested. The speed of testing for all specimens was 0-024 in. per minute. Self-aligning grips were used throughout the work. . * Results of Tests. GENERAL. The results of the tests confirm the practical experience of shippers both in New Zealand and abroad. They, prove conclusively that the standard export crate as at present assembled is an unbalanced construction. . The outstanding features of the tests were the weakness of the nailing and the low margin of safety given by the battens for the safe carriage of the contents of the crate. Shippers and graders commonly remark upon both of these features. In practically every export shipment nails draw and protrude from crates, and where any laxity is displayed in the use of battens of any. wood except beech, either below j in. in thickness or of an inferior quality, excessive splitting and breakage occur. This is easily understood. In handling, if slips occur, the tendency is. for crates to fall diagonally, and thus strike the ground or other surface where battens are nailed to the ends. Nails and battens are therefore called upon to absorb very high stresses. The use of a wire reinforcement round the ends and centres undoubtedly assists to 'absorb these stresses and to reduce the drawing of nails. The battens, however, tend to split, even with the best of . reinforcement, and, strictly speaking, are too thin. The extent of these and other weaknesses, and the measures necessary to minimize them, are clearly set out in the analysis of results of tests.

ANALYSIS OF RESULTS. The results of the tests are shown in a series of tables. A detailed description of the crates tested accompanies each table, which is confined as far as possible to the study of a single variable in design. For convenience in making analysis, and in comparing other designs, the white-pine crate, requiring 330 drops’to cause failure and conforming to the following specification, is taken as 100 per cent. : Battens Twelve battens, each 27 in. long by 3 J in. wide by. f in. thick. Ends and centreOne piece, each twelve-sided, 15 in. between opposite parallel sides, by J in. thick. —Two 5D cement-coated nails into each end and centre of battens. Reinforcement Three 14 B.W.G. wires fastened around ends and centre with one J in. staple between battens. • ' ' This is the standard export crate with cement-coated nails substituted for smooth wire nails. EFFECT OF SUBSTITUTING CEMENT-COATED FOR SMOOTH WIRE NAILS Details of Crates tested. Number of Nails. — Two nails into the centre and ends of each batten. Reinforcement. — 14 B.W.G; wires fastened around the centre and each end with one f in. staple between battens.

Table 1 represents the results of tests upon four groups of crates nailed with ordinary and cement-coated nails. Crates manufactured from both Swedish spruce and New Zealand white-pine were used, The results were remarkably consistent, the cement-coated nailed crates of both species of wood showing a -per-cent, superiority over those fastened with smooth wire nails. The method of failure of crates nailed with ordinary nails was typical of that which occurs in the export handling of cheese-crates. In all cases the nails tended to loosen early in the test, ultimately drawing where driven into the end grain and thus causing the ends to fail rapidly by splitting. This type of failure is illustrated in Fig. 8. It was sensibly reduced by the use of the cement-coated nails, which are recommended to shippers for both export and domestic service. The drawing of nails, in addition to weakening crates, also causes considerable damage to the clothes and hands of labourers. This is a serious disadvantage, largely eliminated by the use of cementcoated nails. Cement-coated nails are used almost universally by box-manufacturers in Canada and the United States of America, as they have a much higher

resistance to withdrawal than plain uncoated nails. The cement coating of the nail consists of various resinous gums mixed by a secret formula and put on the nails by a baking process. Though the makers do not claim that the nails are absolutely rust-proof, they do claim that nails thus treated will resist the effects of moisture from 20 to 50 per cent, better than the uncoated wire nail. But it is when in use that the non-rusting quality is most evident. There is more coating on the nails than is actually necessary for holding-power. The heat caused by the friction in driving the nail softens the coating, and the surplus is forced towards the head, completely closing any opening ; this prevents the admission of moisture between the wood and the nail. Under similar conditions of use the life of a cement-coated nail will be about twice as long as that of an uncoated one. They are claimed to require less force to drive, as the softened coating forms a lubricant. The advantages to be gained by the use of cement-coated

nails are so great that it is very desirable that -they should be universally adopted for boxing and crating work throughout New Zealand. Whereas a 2 in. cement-coated nail driven if in. into the side grain of a piece of American pine required a force of 226 lb. to withdraw it, a common nail under the same conditions was withdrawn with a force of only 106 lb. Complaints are sometimes heard regarding their flow in the nailing-machines, but no real difficulties will be experienced if powdered soapstone is mixed with the nails and the pans of the machines are not filled too full. That cheese-producers have been alive to the poor holding-quality of the ordinary smooth wire nail is clear from the increasing use of barbed and twisted nails. While no tests were made of crates fastened with these nails, there is no doubt that they are superior to the ordinary smooth wire nail, although inferior to the cement-coated nail.

EFFECT OF VARYING SPECIES OF TIMBER USED. Details of Crates tested. Nailing. Two 5D cement-coated nails into the centre and ends of each batten. Reinforcement. — 14 B.W.G. wires fastened around the centre and each end with one in. staple between battens.

Increasing the holding-power of the nails, however, caused them to shear through the ends of battens, allowing the end to again fail by splitting. Resistance to this type of failure by various woods in common use for cheese-crates is shown in Table 2. The crates manufactured from silver-beech, the wood of greatest density or specific gravity, gave the best results. On the other hand, although the four woods —white-pine, spruce, hemlock, and insignis —are of approximately the same density, the serviceability of the white-pine crates was almost three times that of the other crates. This is probably accounted for by the fact that in white-pine there is little difference between the density of the summer wood and spring wood. In the other three species this difference is more marked. At the same time it must be conceded that the quality of many white-pine crates commonly used is much below that of the specimens

tested. A group of such crates, tested to demonstrate the effect of •quality of timber, in particular the effect of knots and diagonal and •spiral grain, reduced the number of drops required to cause loss of ■contents to 128, thus showing a superiority of only 30 per cent, over the •crates manufactured from the three other woods. In appearance the hemlock and spruce crates were superior to the white-pine crates, and especially to the . insignis-pine crates. The insignis-pine crate, indeed, was of inferior quality. Improved manufacture would produce a more serviceable package. EFFECT OF VARYING SIZE OF NAILS USED. * Details of Crates tested. Timber.— spruce. Nailing. —-Two nails into the centre and ends of each batten. Reinforcement. — 14 B.W.G. wires fastened around the centre and each end with one f in. staple between battens.

Increased holding-power in nails also varies with their size. This is clearly indicated by Table 3, which again shows remarkable but not unexpected consistency in the gain in holding-power with length in both smooth and cement-coated nails of corresponding sizes. Increasing the length of if in. nails (by | in., or 30 per cent.) to 2-f in., according to these tests, increases the strength of spruce crates by over 100 per cent. In hemlock and insignis - pine crates a similar increase in strength would be attained, but in white-pine and beech packages the percentage increase would be smaller owing to battens failing before The nails could develop their full strength. EFFECT OF REINFORCING CRATES WITH WIRE BINDING NOT IN TENSION. Details of Crates tested. Timber. — Norway spruce. Nailing.— 7D c.c. nails into the centre and ends of each batten. Reinforcement. — l4 B.W.G. wire fastened to crate with one fin. staple between battens.

. Wire bindings applied over battens and fastened by staples driven, into the ends and centre between battens absorb a large part of the stresses which would otherwise be borne by the nails. , While the wire' is placed on by hand without the use of any stretching-machines, the method of attachment results in the binding being placed under considerable tension. The wire is placed around the crate and the ends firmly fixed to a batten. A staple is then driven over the wire between the battens, thus drawing the wire between the battens down so as to touch the ends. Fig. 8 illustrates this, and shows the grooving of the battens where the wire has bitten into them. The results of the tests set out in Table 4 show that two bindings placed over the ends almost double the strength of the package, while three bindings placed over both the ends and the centre increase the strength of the package six times.

The unwired crate failed, by nails pulling through the ends of battens, allowing the ends to split. Applying two wires at the ends delayed this type of failure. The first weakness to develop was the drawing of nails at the centre, since a crate falling diagonally on an edge causes battens to act as slender columns eccentrically loaded at the ends. These, therefore, tend to bulge and break at the centre of their length. Immediately the centre became loose the . crate skewed, the ends of battens again pulling through the nails. This is well illustrated in Fig. io. A third wire applied at the centre effectively eliminated the weakness of the two - wired crate, the final failure occurring by battens breaking across the grain. The efficient stapling of the wire bindings is essential if the full strength of this type of construction is to be developed. The superiority of the three-wired over the two-wired crate has long been known to the industry in various districts. In some, however,

there has been objection voiced to the third wire. The foregoing tests do not support the objections. An alternative' has been proposed in one district to substitute a longer nail in the centre for the short nail together with the wire binding. With smooth wire nails this alternative would give almost as good results, but both constructions are unsatisfactory, as the smooth wire nails pull easily, causing the crate to fail rapidly and also causing damage to the clothes and hands of labourers. In the cement-coated nailed crates the superiority of the three-wired over the two-wired packages is very marked and clearly economical. EFFECT OF REINFORCING CRATES WITH FLAT METAL NAILED STRAPPING NOT IN TENSION. ’ Details of Crate's tested. Timber. New Zealand white-pine' ’ Nailing. — sD cement-coated nails. Reinforcement. — f in. flat metal nailed straps fastened around centre and each end . of crate.

. The poor finished appearance of the wire reinforcement, together with the damage caused to the clothes and hands of labourers, warranted the study of promising substitutes. The results of a series of tests upon crates reinforced with flat-nailed strapping are shown in Table 5. The finished appearance of this reinforcement is certainly good, as depicted in Fig. 1. It has the added merit of being more easily applied, but the crate is somewhat more difficult to open for weighing and inspection. The strapping, too, is more likely to tear or break between battens, causing damage to the clothes and hands of labourers, as in the case of wired crates. In nearly all crates in these groups one or more straps broke in this manner. Efforts were made to secure the flat-nailed strapping in a manner similar to the wire binding—that is, by nails driven into the ends , and centre between battens. This method of attachment proved a failure, due to the strap splitting away from the nails while driving, thus causing the strapping to break soon after the commencement of the test. The method of attachment finally developed consisted of applying the flat strapping without tension, merely driving the nails through the strapping and the battens into the ends. Fastening the battens to the ends and centre by 'one nail and the strapping to the battens by one nail gave a construction equivalent to that of the three-wired crate, in which the battens are attached by two nails and a wire binding applied over the battens. It has the added merit of eliminating the use of over forty staples per crate. The results of the tests, as would be expected, were the same as for the three-wired crates.

The strongest package, using this type of reinforcement, was assembled by using two nails per batten and affixing the strap with an additional nail through the strap into each batten. The extra strength was due solely to the extra nail. On the other hand, more than one nail per batten driven through the strap caused it to break ■early in the test, due to the reduction in the effective cross-section of the metal. COMPARISON OF DIFFERENT TYPES OF REINFORCEMENT. : Details of Crates tested. Timber. Norway spruce. Nailing.— 5D c.c. nails into centre and ends of each batten. Reinforcement. — Wire without tension. 14 B.W.G. wires fastened around the centre and each end with one | in. staple between battens. Strapping without tension : | in. flat nailed strapping , fixed around the centre and each end with one 5D cement-coated nail per batten. * . Strapping with tension : | in. by 0-018 in. Acme flat strapping placed around the centre and each end and kept in position with four staples per strap.

The tension type of nailless flat metal strapping was also studied as a suitable substitute for wire bindings (Table 6). Acme unannealed metal strapping was applied under tension over the nail-heads at the ■ends and centre, and kept in position by means of staples driven over the strapping into the battens. It proved the most effective type of ■end and centre reinforcement, crates fastened in this manner being' ■considerably stronger than both the wired and nailed strapped packages. The failure in the three types of crates was substantially the same, the rate at which the various weaknesses developed being retarded in the case of the stronger containers. The ultimate failure was in all cases due to the breaking of one or more wire bindings, nailed strapping, or Acme nailless strapping. Although applied under tension the nailless strapping requires to be held in position by staples, &c., otherwise the skewing of the crate and the tendency of the strapping between battens to catch and tear upon projections is apt to loosen the reinforcement, allowing it to slip off over the ends. Any size less than that used in the tests would give much poorer results. Just as the nailed strapped crates were more difficult to open and ■close for weighing and inspection than the wired crates, so those reinforced with Acme nail less strapping were at a still greater disadvantage in this respect. Whereas the wire binding and nailed strapping may be used again after the opening of the crates, the Acme strapping must be replaced by new strappings.

COMPARISON OF DIFFERENT TYPES OF CRATE-CONSTRUCTION. Details of Crates . tested. . 1 Standard. Mat Type. Cleated Type. Timber .. . . N.Z. white-pine Spruce . . N.Z. white-pine. Nailing per batten . . Two 5D c.c. . . One 7D c.c. Three 5D c.c. Number and thickness of battens . . Twelve f in. . . Sixteen | in. Ten j in. Reinforcement . . 14 B.W.G. wires Gum hoops | in. by 0-018 in. straps. Ends . . .. . ’ Single piece . . Single piece Two piece fastened with cleats.

Reference has already been made to the difficulty of opening and closing standard cheese-crates for weighing and inspection. Four per cent, of all crates are opened and closed, both in New Zealand and abroad, for the purpose of verifying weights, &c. This represents the opening and closing of one crate in every twelve ; and since each crate so dealt with occupies a man’s time on the average for fifteen minutes the difficulty is a very real one. Improving this feature of crate design will further reduce the original cost of assembly at the factory and enhance the value of the used crate. ; Two types of crate thought to offer a solution of'the difficulty'were' tested (Table 7). The mat type shown in Figs. 2 and 11 gave poor

results. These were not unexpected. The gum wooden hoops reinforcing the ends and centre were of insufficient strength, though they effectively protected the battens before they themselves split to pieces. The substitution of flat metal strapping for the wooden hoops would probably improve the package. It certainly has, some promise for the export trade. The cleated type of crate shown in Figs. 4,5, and 12 gave results slightly lower than those displayed by the standard wired crate with cement-coated nails. It is, however, easier to assemble, and to open and close for inspection, and has, in addition, a higher salvage, value than the wired crate, While it requires less timber for its manufacture, more nails are necessary. This disadvantage, however, is offset by the fact that no staples are used, and that only two Acme flat straps are used in place of the three wire bindings. This type of binding bends the battens in towards the cheese, and to prevent any possibility of their touching the produce the diameter of this crate is increased by J in.

The method of failure, was much the same as in the wired crates. It was noticeable, however, that the ends of the cleats absorbed a large portion of the stresses caused by the various drops in the machine. A study of cheese-crates in actual service was made in order to study a similar effect on the wired crates. A number of factories assemble their crates, having the ends of the battens about | in. away from the edge of the ends, as shown in Fig. 13 a, instead of the usual practice as in Fig. 13 b. Accordingly, when a crate falls diagonally on to an edge, the end is often the first member of the crate to take the shock. In such cases there is a tendency for the nails in the end of the batten to compress the wood of the - battens towards the centre, as in Fig. 13 a'. This weakness, however, is not as serious as that which develops in the crate of ordinary construction, in which the end of

(a) End of batten kept away from end of crate ; (6) end of batten flush with end of crate ; (ft 1 ) after diagonal fall on to edge of crate the end is hit first, tending merely to loosen nail in batten ; (& 1 ) after diagonal fall batten hit first, tending to pull nail right through end of batten. the batten is the first member of the crate to take the shock in a fall diagonally on to an edge. Here the ends tend to pull the nail through the end of the batten, as in Fig. 13 b l . Other things being equal, attaching the ends of battens away from the ends would appear to be the better practice. An examination

of several hundred crates, however, indicated that with a smaller thickness of the end to nail into the nailing was decidedly poorer, there being a greater tendency to split battens and ends. Some of the field studies of crates in actual service also suggested that the weight of the standard crates holding two cheeses, approximating to 175 lb., was too great for ease of handling. This did not obtain where mechanical handling equipment was in use. A series of tests was accordingly made, using a cleated type of crate holding only one 80 lb. cheese, as in Figs. 6 and 14. This proved to -be the most serviceable and balanced crate tested, being approximately two and a half times as strong as the standard wired crate with cement-coated nails. It is admittedly slightly more expensive in comparison with the two-cheese size. This will probably militate against its adoption. Whether the cleated type of crate is adaptable to export conditions is questionable. Crates awaiting export are at present piled on their ends. With cleated crates greater care would require to be taken in stacking. Piling on their sides, as in shipment by boat, would effectively remove this difficulty. At various ports, too, crates are moved on their ends along gravity conveyers. These would again require greater care in the handling of the cleated crate. The branding of the ends, although presenting some difficulty, is not by any means an insurmountable obstacle to the general adoption of the package. EFFECT OF VARYING THICKNESS OF BATTENS. Details of Crates tested. Size of Nails. — J in. and: f in. battens : 5D cement-coated nails. in. and | in. battens: ?D cement-coated nails. Number of Nails.— Ten-sided crates : Three nails into centre and ends ,of each batten. Twelve-sided crates : Two nails into centre and ends of each batten. Reinforcement. Standard crate : 14 B.W.G. wires fastened to centre and each end with one f in. staple between battens. Cleated crate : Two | in. Acme straps placed around crate 5 in. each side of centre. Ends and Centre. — Standard crate: Single piece, -Jin. thick. Cleated crate •’ Ends two-piece, f in. thick, fastened together with two cleats, 2 in. by f in., affixed with twelve 5D c.c. nails per cleat; centres fastened together with one dowel.

" Throughout the foregoing series of tests constant reference has been made to the questionable thickness of battens used. As shown in Table 8 a comprehensive study was made of this feature of crate construction, forty-five crates of different woods ■ and different types being tested to destruction. The tests indicate that unless battens are supplied to a strict specification, in., material is very preferable to the f in. material in common use. This applies only to white-pine, hemlock, spruce, and insignis pine. The silver-beech battens fin. thick give a crate stronger than in. white-pine. Increasing the thickness of battens from fin. to |in., or 33.| per cent., consistently resulted in a 100-per-cent. increase in the strength of the crates. ‘ EFFECT OF ALTERING NUMBER OF BATTENS. Details of Crates tested. Timber. —-New Zealand white-pine. Size of Nails. — in. battens: 5D cement-coated nails. -J in. battens : 7D cement-coated nails. Number of Nails. Ten-sided crates : Three nails into centre and ends of each batten; Twelve-sided crates : Two nails into centre and ends of each batten. Reinforcement.— Standard crate : 14 B.W.G. wires fastened to centre and each end with one f in. staple between battens. Cleated crate : Two | in. Acme straps placed around crate 5 in. each side of the centre. Ends and Centres.— Standard crate : Single piece, f in. thick. Cleated crate : Ends two-piece, f in. thick, fastened together with two 2 in. by f in. cleats, affixed with twelve 5D cement-coated nails per cleat; centres fastened together with one dowel. /

In developing the cleated cheese-crates it was considered that greater strength might be obtained by decreasing the number of battens, thus reducing their tendency to break across the grain. The work was further extended to include the standard wired crates. In all, three sets comprising six different groups of crates were tested. In each set a reduction of battens from twelve to ten was made, the width of the battens in the latter case being increased by an amount sufficient to make the total air-space between battens the same in both the ten- and twelve-sided crates. The diameters of the ends and centres were kept constant in all cases. The results are shown in Table 9. It was found with the standard wired crate that decreasing the number of battens as stated resulted in a 14-per-cent. decrease in the strength of the crate, due to the more frequent splitting of wider battens. With the cleated - style

crates, however, the opposite result occurred, the reduction in the number of battens resulting in a 48-per-cent. increase, in the cratestrength. This was probably due to the method of reinforcement. The Acme straps, which are drawn very tightly around the crate, reduce the splitting tendency of the wide battens. The tendency of the battens to break across the grain, too, is reduced by increasing the width of the battens. A stronger crate is thus developed. EFFECT OF INCREASING NUMBER OF NAILS. Details of Crates tested. Types of Crate. Standard two-cheese crate and cleated one-cheese crate. Reinforcement. type : 14 B.W.G. wires fastened to the centre and each end with one in. staple between battens.' Cleated type : One J-in. Acme strap placed around centre of crate. Ends. — Standard type: Single piece, % in. thick. Cleated type: Two-piece, in. thick, fastened together with two 2 in. by f in. cleats, affixed with twelve 5D cement-coated nails per cleat. Number of Battens. Standard type : Twelve. Cleated type : Ten.

During the course of the study various sets of tests were carried out on similarly constructed, crates, with varying numbers of nails holding the battens to the centres and ends of the crates (Table 10). In all, comparative figures are available for four sets of tests, comprising in all eight groups of crates. As already indicated, nailing has a decided influence on the strength of cheese-crates. With one set of spruce crates too small nails were employed, and in consequence the increased strength due to increased nailing was not marked. Using nails of a more suitable size, although of the smooth wire type, a further, set of spruce crates developed a 100-per-cent. increase in strength for the use of two nails in place of one. Similarly, a 50-per-cent. increase in the nailing of a white-pine crate resulted in a 48-per-cent. increase in the strength of the crate. . The most marked increase occurred in the cleated-type one-cheese crate. With this type , a 50-per-cent. increase in nailing resulted in an increase of over 200 per cent, in the strength of the crate. This was due to the lower weight per nail ratio and the more balanced construction of this package. ' 1 ■

7 - Conclusions. The results of the foregoing studies may be summarized as follows : (i.-) The export crate as at present designed is an unbalanced container. (2.) The use of cement-coated nails is essential if an economical and balanced package is to be designed. (3.) A suitable-sized nail should be used for each species of wood—--4D for silver - beech, 5D for white-pine, and 6D for insignis pine, hemlock, and spruce. r (4.) Silver-beech, white-pine, spruce, insignis pine, and hemlock rank in., this order in suitability for cheese-crates where carryingqualities, are considered. . (5.) Flat' strapping and wire binding applied with or without tension are both of great value as a reinforcement for crates. (6.) Resistance to loss of contents increases with the number of bindings used. (7.) Tension-applied nailless strapping, flat-nailed strapping, and wire binding rank in this order in strength as crate reinforcements. (8.) The cleated and mat-construction crates offer possibilities for export service; and further experimental shipments should be forwarded abroad for study and comment. (9.) Battens, j in. thick, of white-pine, insignis pine, hemlock, and spruce are dangerously thin unless supplied under a rigid specification ; otherwise a thickness of in. is recommended. Battens, fin. thick, of silver-beech make a very strong" crate. (10.) A crate that could be opened and closed more readily for inspection purposes would be a decided advantage, and it' is hoped later to evolve a design embodying this feature. The subject generally is by no means finalized at the present stage. Recommendations for Standard Crate. ■Having consideration to the various factors involved, the Forest Service recommends the use of one standard type- of crate for the export ■trade. It can be manufactured from any of the timbers in usebeech, white-pine, insignis pine, hemlock, or spruce. It consists essentially of twelve-sided one-piece ends and centre, fin. thick and 15 in. between parallel sides; battens, 27|in. long by 3 in. to 3jin. wide and not less than in. thick if constructed of white-pine, insignis pine, hemlock, or spruce, and not less than r 5 in. thick if manufactured of silver-beech ; two cement-coated nails through ends and centres of battens ; and an approved type of metal. binding applied around each end and centre. This crate is much stronger than the present standard export package, and, further, is a more attractive container. A . detailed specification for this crate follows : SPECIFICATION FOR' STANDARD METAL-BOUND CHEESE-CRATE FOR EXPORT. ’ . ' \ ■ Section A : General. (i.) Definition : The crate as herein specified shall be known as the Standard Metal-bound Cheese-crate —Export Type.” •

Section B : Timber.

(2.) Woods used : The following timbers shall be admitted under, this specification : White-pine {Podocarpus dacrydioides), silver-beech {N othofagus Menziesii), insignis pine (Pinus radiata), western hemlock {Tsuga heterophylla}, spruce (Picea excelsa), and other timbers approved by the Forest Service. k (3.) Material: '(a.) The battens and . ends shall be well manufactured, and shall be cut true to size. All defects in the timber which materially lessen the strength of the part, or expose contents to damage, or interfere with proper nailing, shall be prohibited. (A) The wood shall be thoroughly seasoned, and shall have a moisture content of not less than 12 per cent, nor more than 18 per cent., based on the weight of the wood after oven-drying to a constant weight. - (4.) Dimensions -(a.) Battens shall be 27-J- in. long, not less than 3 in. nor more than. 3-J- in. wide, and shall be not less than in. thick for white-pine, insignis pine, hemlock, and spruce boards, and not less than T s s in. thick for silver-beech boards ; the ends shall be twelve-sided, 15 in. wide between opposite sides, and -J in. thick. (6.) The variation in thickness of the boards above or below the thickness specified shall be not more than in., and this variation below the specified thickness shall not extend over more than 10 per cent, of the face of that particular board. (5. Width of parts : (a.) Battens, ends, and centres shall be of single-piece material. (b.) Matched and glued or -jointed boards shall be regarded as single pieces. No end or centre shall consist of more than three boards so joined. , (6.) Jointing : (a.) Matched and glued ends or centres shall in addition be fastened with not less than two galvanized corrugated fasteners, 1 in. by f in. per joint. (&.) The edges of all battens shall be rounded or chamfered along their entire length on one side. (7.) Surfacing : The outside surface of the battens and tops may be finesawn or veneered finish ; otherwise they shall be smooth-planed. • . (8.) The grain of the two ends shall be at right angles to one another, and the grain of the centre midway between the two. Section C : Nailing. ■ (9.) Nailing schedule: (a.) if in. cement-coated nails shall be used when driving into white-pine, if. in. cement-coated nails when driving into insignis pine, hemlock, and spruce, and if in. cement-coated nails when driving into beech ends. (&.) Nails shall be driven flush. (c.) Each batten shall be attached to each end and centre by two nails. Section D : Metal Binding. (10.) Metal (a.) Flat nailed strapping or stapled wire binding shall -be of •soft metal, and shall have a maximum tensile strength of approximately 56,000 lb. per square , inch. (&.) Tension-applied nailless metal binding shall be of hard unannealed metal, and shall have a maximum tensile strength of approximately 84,000 lb. to the square inch, (c.) The binding shall be galvanized or otherwise treated to protect against rust. ; (11.) The ends of' tension-applied nailless bindings shall be fastened in such a manner that the joint shall have a breaking-strength of not less than 75 per cent, of the ultimate strength of the binding. (12.) Size of binding : The, metal binding shall be not less than J in. in width by 0-018 in. thickness or of equivalent cross-sectional area. ' - (13.) Application : (a.) Three bindings shall be used per crate, placed around the ends and centre, and covering the nails driven through the battens into the ends and centre. (6.) Each nailed strapping shall be fastened to the crate with twelve nails, one nail being driven centrally through each batten. (c.) Each wire binding applied without tension shall be fastened to the crate with twelve staples, one staple being driven into the end or centre between all battens:. {d.) Each tension-applied binding shall be kept in position on the crate. with four staples per binding driven into the battens over the binding. • .•>•:

Acknowledgments. - The following organizations have co-operated with the Forest Service in the work here described: Dairy Division, Department of Agriculturegeneral; School of Engineering, Canterbury University College crate tests ; School of Engineering, Auckland University —binding tests ; Messrs. J. F. Hargreaves, and Co. (Limited), Wellington, N.Z., Acme strapping and spruce crates ; Messrs. Johnson, Clapman, and Morris (Limited), Wellington, N.Z., and United States Steel Products. Company, New York—cement-coated nails ; Messrs. J. B. Mac Ewan and Co., Wellington— crates ; Hawera Cooperative Dairy Company — crates ; Messrs. P. Carey and Co., Auckland— crates ; Mr. B. Hughes, Temuka—insignis-pine crates. Special acknowledgment is due to the Madison Forest Products Laboratory of the United States Forest Service for its many reports upon box and crate , construction. These have enabled the present . work to be . carried to. a conclusion without the laborious investigation of many features of design already studied by the American laboratory. ' A large number of the crate tests were carried out by Mr. E. H. A. Englebretch, of the School of Engineering, Canterbury University College. To him special acknowledgment is due for his untiring work in this section of the study. COMMON AND BOTANICAL NAMES OF TIMBERS MENTIONED IN . THIS ARTICLE. : '■ ' Common Name. Botanical Name. - Country of Growth.' ’••• White-pine ’ .. Podocarpus dacrydioides .. • New Zealand. Insignis pine .. Pinus radiata .. . . . ~ Silver-beech .. Nothofagus Menziesii .. ~ Norway spruce .. Picea excelsa . . . . Scandinavia. Sitka spruce .. Picea sitchensis "• ■ .. Pacific Coast of North America. Hemlock .. < .. Tsuga heterophylla .. Pacific Coast of North America. Gum .. .. Liquidambar stryaciflua . . North America..

Paper Method of Weed-control. — In tree nurseries of the State Forest Service last year trials were initiated with Pabco-Thermogen paper mulch as a weed-reducer. Results'were, on the' whole, unsatisfactory (states the annual report of the Service) ; in most cases the efficacy of the mulch in smothering weeds was lost by the impossibility of fixing the paper so as.to adhere closely to the ground. The effect of the mulch on plant-growth varied—-in one ; case seedlings touching the paper appeared to be “ burnt,” while better plant-growth was observed in some lines treated with mulch. Generally speaking, however,. no great difference in growth was found. Soil analysis made showed that no “ souring ”of the soil was produced by the application of the mulch in the lines. v •'

Pasteurizing Milk for Cheesemaking. — During the dairying season of 1925-26 the quantity of cheese made in factories equipped with pasteurizing plants equalled 76 per cent, of the total output of the Dominion, as against 69 per cent, for the preceding season. < • •' Correction. —In the C.O.R. list , published in the Journal ' for November last the mature Jersey cow Woodlands Gipsy, with a record of 651-41 lb. butterfat, was inadvertently entered as tested by J. G. Morgan, Ngawapurua. This should have been S. G. Morgan, Woodville. : \ . ■ ... ■. ' . j.j .. . j

Timber. Type of Nails. Size of ■ Nails. Number of Drops required to spill Contents. Relative Strength to Smooth-wire--nailed Crate. White-pine . . Smooth wire 50 190 1-00 >> Cement-coated 50 330 330 174 1’74 Norway spruce Smooth wire 7 D 119 1-00 Cement-coated 7 d 210 i- 7 6

Table I.

Timber. Number of Drops required to spill Contents. Relative Strength to White-pine Crate. White-pine 33° ■ 1-00 Beech .. < .. 54 6 i-66 Spruce . . IOI 0-31 Hemlock 98 0-30 Insignis pine . . ■ 98 . 0-30

Table 2.

Size of Nails. Type of Nails. Number of Drops required to spill Contents. Relative Strength to Crates nailed with 5D Nails. 50 Cement-coated IOI 1-00 70 ,, 210 2-08 50 Smooth wire 58 1-00 ■ 70 ,» • • 119 119 2; 05 2; 05

Table 3.

Number of Reinforcements. j Number of Drops required to spill Contents. Relative Strength to Unreinforced Crate. None . . . . 36 1'00 2 . . . . 69 1-92 3 ..... 210 5-83 '

Table 4.

Nails per Batten. Additional Nails holding Strapping to Battens. Number of Drops required to spill Contents. Relative Strength to Unreinforced Crate. Two 5D Unreinforced . . 57 1-00 One 5D Two 5D 291 5-n One 5D . . One 5D 328 5-7 6 Two 5D One 5D . 423 7'43

Table 5.

Type of Reinforcement. Number of Drops required to spill Contents. Relative Strength to Wired Crate. Wire IOI 1-00 Flat nailed strapping . 125 . 1-28 Acme strapping . . 182 i-8o

Table 6.

Type of Crate. Number of Drops required to spill Contents. Relative Strength to Standard Crate. Standard . 330 1-00 Cleated one-cheese 835 2-53 Cleated two-cheese . . 286 0-87 ■ Mat-construction 93 0-28

Table. 7.

Thickness of Battens. Timber. Style of Crate. Number of Battens. Number of Drops required to spill Contents. Relative Strength to Thin Battens. t in White-pine' . . Standard . .. 12 330 1-00 to m. >} J, • • 12 . 12 5i3 5i3 1-56 1-56 i m. 12 623 1-89 -1 in. .. • • .. • • IO 280 1-00 i in. IO ,, 545 >> i-95 IO 545 i-95 i in. .. • • Cleated 12 193 1-00 I in. ,, ,, 12 >> 430 12 ■2'22 430 ■2-22 . i in. Beech Standard 12 265 1-00 i in. , J • • 12 548 ' ' 2-06

Table 8.

Style of Crate. Number of Battens. Thickness of Battens. Number of Drops required to spill Contents. Relative Strength to Twelve-sided Crate. Standard .. .. 12 ' ■ t in. 330 1-00 „ . . . • . . IO in. ■ 280 0-85 . . . . .. 12 12 ■ i in. 1 in. 623 623 1-00 1-00 ,, . . . . IO | in. 545 0-87 Cleated . . .. " 12 i-in. 430 1-00 IO 2 in ■ 635 1-48 ,, . . . . IO 2 in. 635 1-48

Table 9.

Nails per Batten. Type of Crate. Timber. Thickness of Battens. Number of Drops required to spill Contents. Relative Strength to Crates with fewer Nails. One 5D c.c. .. Standard.. Norway spruce fin96 / 1-00 Two 5D c.c.. . ■• • >• f IOI 1-05 ,, f in. IOI 1-05 One 7D s.w... fin. 56 1-00 ' Two 7D s.w. . . fin. 119 2-12 Two 7D c.c... - "White-pine in. 623 1-00 Three 7D c.c. ,, • > fin. | in. 920 920 1-48 1-48 Two 5D c.c. . . Cleated . .. f in. 270 1-00 Three 5D c.c. »» • • T ,, fin. f in- 8.35 835 3‘°9 3-09

Table 10.