Six species of commercial papers (60 lbs. simili papers) A, B, C, D, E and F were examined to ascertain the length and length distribution of fibers composing them. Five species, A, B, C, D, and E were determined to be composed mainly of coniferous wood fibers, and F only of hardwood. The mean fiber length observed were as follows : A=1.24 mm B= 1.24 mm C= 1.29mm D=1.26mm E =1.14mm F =0.70mm
From these fiber length measurements we can say that the average length of coniferous fibers in paper is reduced to about 1/3 of its original length in wood, and the length of the hardwood fibers to about 1/2. This is attributed to the fact that longer fibers must have many chances to be cutted than shorter fibers in beating.
From the length distribution curve of paper fibers, we found out that each frequency curve of any paper fibers always resembles in shape to that of Poisson distribution, and differs from that of normal distribution which will be applied to the raw wood fiber distribution.
As regards to the relation between the strength of papar and fiber length, we came to the conclusion that longer fibers does not always make stronger papers, which is admitted by some authors as Jayme.
In the second part of experiments we tried if we can find the actual paper fiber length and its distribution only by the use of Tappi standard classifier.
Defibrated paper fibers were passed through four screens with running water, and each residues (Fraction I=+24 mesh; Fraction II=-24, +42; Fraction III=-42, +80; Fraction IV=-80, +150) were gathered, weighed and examined about their fiber length and fiber length distribution.From this examination of five species of papers (A, B, C, D and E) composed of coniferous fibers, the mean value of fiber length of these papers which comes in each fractions were obtained, and also the mean distribution of fiber length which will be named “model distribution” fiber length were determined. From this model distribution of each fractions and the weight % of each franctions, we tried to super-pose the length distribution curve of paper fibers.
We have compared thus super-posed distribution curve with the microscopically determined length distribution curve, and also with the theoretical Poisson distribution curve. At least on five papers A, B, C, D and E, from the chi-square test, the super-posed distribution curve agrees with the microscopically determined curve. On the other hand the super-pose distribution curve agreed with the theoretical Poisson distribution curve only in the case ef A, D and E. and was ascertained to be significant in B and C.
抄録全体を表示