Apparently a minimum of about 4 percent of haze for

Apparently a minimum of about 4 percent of haze for sheets up to
100 mils thi ck is the best that can be done and for thicker sheets this
would be somewhat higher. The fog or cloudiness in sheets with
haze values of this order is quite easily detected by the naked eye and
is a much more serious defect than the lo wering of white-light trans-
mission by blue dye. Th e va lu es recorded in table 1 for the materials
oth er than cellulose nitrate, cellulose acetate, and acrylate resin, were
obtained on experimental samples only, such products not being
available in sheet form for aircraft windshields at the present time.

AxilrOd]
[({ine
•
Transparent Plastics jor Aircrajt
III. INDEX OF REFRACTION
373
The index of refraction is of inter e st in considering the light-trans-
mission characteristics of plastics. This property was measured by
means of an Abbe refractometer on the same specimens used in the
light-transmission and haze tests. Two contact liquids, alpha-
bromonaphthalene and mercuric iodide, were employed in order to
guard against incorrect values resulting from reaction between the
sample and the contact liquid. The data obtained are presented in
table 1. The refractive indices for the materials studied range from
1.47 to 1.58.
IV. VISIBILITY DISTORTION TESTS ON PLASTICS
Visibility distortion tests were made in accordance with the proce-
dure suggested by the American Standards Association for laminated
safety glass using sheets 12 inches square. In this test, a line is pro-
jected with a lantern so as to fall midway between two parallel lines
1
inch apart marked on a screen. The test specimen is 25 feet from
the screen and the lantern is adjusted so that the center 10- by 10 -inch
square fills the beam. The sample in a suitable frame is moved across
the beam and the movement of the projected line observed; the safety-
glass specification requires that the projected line shall not deviate
more than
}6
inch as the sheet is moved across the beam. Several
materials were tested but none failed to meet the specification.
Measurements were then made on each sample of the deviation of the
projected line as the square was moved an inch at a time across the
beam. This measurement was made at the top, center, and bottom
of the screen for the sample normal to the incident light; this was
repeated just at the center of the screen for angles of 75, 60, 45, and
30°. The data did not prove suitable for giving an estimate of the
visibility because a specimen with many relatively small ripples
simply causes the projected line to move back and forth rapidly as
well as causing it to get out of focus; a sample which is slightly wedge-
shaped or one with a large wave in its surface might cause the line to
deviate as much or more, yet it would be far superior as regards
distortion.