Report compiled by R. le Maître during December, 1986.

Results of three groups of test are included, the mass and size tests the impact or bounce tests and strike tests performed with a mallet in a frame similar to the Izod test for Shear but modified to measure Kinetic energy absorption and loss.

MASS AND SIZE TESTS

COMMENTS ON MASS AND SIZE

1. Jaques balls are milled by turning. There are dissimilarities between balls which rule moulding out unless several moulds are used. They appear to be subsequently coated and cured; see enlargement of new ball with defective coating.
2. The milled grooves in Jaques balls are sloped at ±300 to the diametral plane, so that the total subtended angle from groove side to groove side is about 60'. Average groove depth 1.lmm. Width of grooves from ±0.5mm to 1.5mm. Approximate volume removed by the total of 100 cuts per ball is ±16 cm³ or ±18gm of material.
3. The Jaques balls show a mass variation averaging 2.69 grams overweight per ball, but the individual mass variation is very considerable, from -49.9gm to +32.8gm on the 454gm required. Only one out of 11 Jaques balls passed the mass specification. 6 out of 11 were more than 20gms. out from the specified average.
4. The Jaques balls were good for size when new; there were none oversize and 4 out of 11 marginally undersized. When of medium age the balls show a marked decrease in diameter of about 1.6mm below average specified size. This makes for easy hooping, and does not seem to be so much due to wear, as a combination of 'flattening' of the milling ridges with some wear/micro-fracturing of the surface.
5. Substantial flattening occurs on a new Jaques ball upon a single impact during elasticity testing. A reduction of up to 0.5mm has been noted on the 8 or so 'peaks' or nodules between milling grooves which are affected by such an impact.
6. The Walker balls, of which only 2 were tested, showed remarkably good identity although both were marginally overmass. The grooves do not appear to have been machined, but rather moulded. Release from the mould has been obtained by using an even shallower groove wall angle than Jaques, some 50° to the vertical giving 100° to the whole groove instead of the Jaques 60°. This release from the mould would also be facilitated by the shallowness of the groove, some 0.2mm deep as compared to Jaques 1.1mm.
7. Using a value of a for steel (mould) of 1.139 x 10^-5/° C and of 7 to 10 x 10^-5/°C for nylon (66) therefore b for steel volume = 3.417 x 10^-5/° C and b for nylon volume ±25.5 x 10^-5/°C. Cooling nylon from 200° to 25°C would give a differential shrinkage of 1.3mm on diameter between mould and ball, sufficient to release the Walker ball shallow grooves easily.
8. The epoxy or polyester resin based compound of the Barlow balls gives with fillers, a value for a as low as 18 x 10^-6/° C which is substantially less than the expansion coefficient for nylon, and should therefore result in a product closely following the dimensions of a steel mould, or precisely following the polyester mould used. In this regard it is interesting to note that the Jaques Balls showed a variation in new size from 90.2mm for the smallest (red) ball to 92.6mm for the largest (pink) ball, a variation of 1.4.nm, whereas the Barlow balls exhibit a variation of only 0.35mm on diameter from largest to smallest.
9. The Barlow balls exhibit much better constancy in mass than 'the Jaques balls, varying by a maximum of +22.0 and +6.2 or 15.8 grams as against the untenably large variation of the Jaques from -49.9gm to +32.8gm, a total of 82.7 gram. It is obvious that the low value for b is also important for mass constancy.
10. What is apparent is that the specification is too tight for the Jaques bipartite method of manufacture without extreme care being taken in the process of manufacture, and if players were to demand specification balls it would be very costly considering that none of the brand new Jaques balls nor the two older ones passed specification for both mass and size. Bounce is dealt with in a separate report.
11. The Barlow balls are irregularly and unevenly milled in comparison with either Jaques or Walker balls. Several points arise here. Firstly the balls need more accurate milling from the point of depth and evenness.

     A jig for perfect tool/ball centralising or a contoured cutter or both.

    Secondly, the number of grooves varied from a low of 68 to a high of 87 between balls, as compared with the exact 100 of both Jaques and Walker. Depth averaged 0.4mm, but ranged from 0.2 to 0.5mm.

    Thirdly, the slight overmass of the Barlow balls could be adequately reduced to the specified limits with the milling of angled, more frequent grooves. The present overmass average above maximum limit is only 8.09gm, easily removed in the improved milling envisaged. See Article 2.

    Lastly the grooves should have round edged crowns and radiused bases, to prevent surface micro-fracture evident on some of the balls now in use. This will prevent stress concentration. 

12. Consideration should be given to the cooling of the Barlow mould. At a thickness of 100mm sufficient heat may be generated by the component reaction to cause temperatures in excess of 150°C to be reached in the mould centre, detrimental to physical properties. Best results are obtained if a temperature of ± 150°C can be attained and maintained, followed by slow cooling, say in boiling water allowed to cool.
13. While the Barlow Balls average 0.94mm undersize, this should be considered much more constant than the Jaques, which exhibits a drop to 1.625 below average after partial use, because of the soft nodules as mentioned in Article 4
14. It is probable that the Barlow balls, although marginally undersized by 0.14mm on minimum and by 0.94mm on the mean of 92.0mm, will retain their size for a long time in play. On the other hand the Jaques balls, initially size either side of 92.0mm, quite rapidly became permanently undersized in play. If specification is to be maintained, then the Barlow balls must be increased by 0.2mm to bring them to a point just above minimum, involving an increase in mould radius of 0.lmm or 3.9 thousandths of an inch. The corresponding increase in mass would be 3grams removable by improved milling.

IMPACT TEST (5 Bounces per Ball)

Official specification Page 3 of The Laws, balls must be dropped from 60 inches (1524mm) onto a piece of 1" M.S. Plate in concrete, rigidly set. They must rebound to 30 inches min, 45 max.

IMPACT TEST COMMENTS

1. None of the Jaques balls passed the bounce test. The variation between all the Jaques balls ranged from 22 to 29 inches. The average height of rebound was 24.9 inches. The variation of 22 to 29 inches or 7 inches is thus 28%.
    
2. All the Barlow balls passed the specification. The average height-- of rebound was 36.1 inches, with a variation from 32 to 39 or 7 inches, equal to 19% of the average.
    
3. The two Walker balls passed specification, and it is to be expected that with proper pigment/filler control Walker balls should have close identity. Maximum variation 1 inch in 41.5 or 2.4% of the average.
    
4. The two worn Jaques balls showed the expected increase in bounce due to flattening of the nodules so that the elasticity is less inhibited. The average rebound height was 26.9 inches or 2 inches up on the new balls. Variation was 3 inches in 26.9 or 11.2%.
    
5. Taking the specification range of 30 to 45 inches rebound height, the indicated average rebound should be 37.5 inches. The Jaques new balls are short oil this figure by 12.6 inches or 33.6%. The Barlow balls at 36. 1 inches are 1. 4 inches or 3. 7% short, the Jaques worn balls are 10. 6 inches or 28.3% short while the Walker balls average 42.5 inches, an excess of 4 inches or 10.7% over. Thus the Barlow balls are not only closest to the specification, but almost exactly on it.
    
6. Going back in time, it is noted that according to D.M.C. Prichard the following events regarding ball specifications are recorded:

   Early balls were made of ivory, willow, beech, coconut fibre and rubber, boxwood with a diameter of 3 - 4 inches, coloured in stripes.
   1870, boxwood was mandatory.
   1871, size 3 5/8" diameter, weight greater than14 1/4 oz.
   1897, 13 3/4 - 14.25 oz.
   1898, composition balls obtainable.
   1903, weight increased to 15 -16 1/2 oz. at Jaques request "To use better boxwood" although all balls were composition by then.
   1906, composition balls mandatory for all tournaments.
   1909, weight increased to 15 3/4 - 16 1/2 oz.
   
   Over the years the major manufacturers were Slazenger with the 'Stadium', 'Brento' and 'Boxite' while Jaques produced the 'Glis-Glis' and 'Eclipse' , Eclectic made the 'Excelitel and Ayres made the 'Championship'. The Excelite came with either coarse milling similar to today's or fine and almost smooth milling. One of the most interesting historical ane6dotes'in Prichard's book reads as follows:
   
   "Over the years opinion hardened in favour of the Ayres ball which by 1939 had become the almost universal choice for tournaments........... "The Council pleaded with Ayres to transfer their existing stock of balls to Jaques, together with the know-how for making the Ayres Ball. This was done, but the result was a disappointment. Somehow in the process of the hand-over the technique seemed to have been lost. Certainly the balls provided by Jaques were different and in the opinion of most players lacked the resilience and reliability of the Ayres Championship Ball."  

7. Mr. D.G. Cunningham confirmed by telephone on 21.12.1986 that he had likewise never found a Jaques ball, new or old, to exceed 28 inches on rebound, and that all Jaques balls failed this specification. In the light of the Prichard comments in Art. 6 above, it seems possible that the 30 - 45 inch specified rebound came in with earlier balls such as the Ayres Championship. Mr. Cunningham offered to contact Prof. Bernard Neal for information regarding this possibility and the reason for the original specification.
    
8. It is notable that the nitro-cellulose covering of Jaques balls is the same as is used for Billiard and Snooker balls, perhaps not so hard. The density of this covering is probably of the order of 1.84 gm/c.c. as against the 1.1135 for the whole ball. It may be that this was the reason that the mass specification changed from 14.5 oz. to 15 - 16 1/2 in 1903 and finally to 15 3/4 - 16 1/4 oz. by 1909. to accommodate the heavier covering. The covering of the Jaques ball varies (on the same ball) from 1.8 to 2.8mm inclusive of the milled groove, i.e. to the other surface of the ball. The difference in specific gravity would require a weight increase of ±1 1/4oz. Which could account for the increase historically applied at Jaques' request, that is to say, the original ball of boxwood or other had a mass of 13 3/4 - 14 1/4.oz, later made 14.5 oz. in composite balls, and 15 3/4 - 16 1/4oz for the final N.C. coated compound. The core material of the Jaques ball looks like a mixture both in colour and size of wood particles possibly interspersed with particles of what might be more resilient and more translucent material, but with no pigment in the core. The Core is covered with a bonding layer and then a second coat plus a top 'lacquer' coat. All these coats are pigmented, but only the outer coat correctly represents the main colour. The core itself may be bonded by a low percentage addition of nitro-cellulose, which should cross-link to the outer casing when cured. A cross-sectional photograph of a brown Jaques ball is appended.

R. le Maître ©1986