Tutorial on Test Plates and their use in Optical Testing

OPTI 515 Tutorial

Melanie Saayman
November 5, 2009
Introduction

Test plate testing is the oldest, simplest and most common method used to verify the flatness
or sphericity of an optical surface during fabrication. It uses the interference effect known as
Newtons rings or Fizeau fringes. When a convex test surface is placed on top of a reference
flat, circular interference fringes called Newtons rings (Figure 1) will result. Straight fringes,
called Fizeau fringes, are formed in an air wedge between two flat, transparent surfaces as in
Figure 3.
Overview of Test Plate Testing

For a curved optical surface with a large radius of curvature R, the classical Fizeau
interferometer (or Newton interferometer as it is referred to in [1]), can be used to determine
R, by placing the optic on top of a reference flat, as in Figure 1.

Figure 1: Newtons Rings [4]

The radius R of the test optic is found by counting the number of bright fringes to a radial
distance r
m
, where the first bright fringe is m = 0, and applying the equation:
=
r
m
2
z( +

)
or br rns
The equation for dark fringes is given by:
=
r
m
2
z
or or rns
Concave or convex surfaces with smaller radii can be measured by placing them on top of a
spherical test plate with approximately the same radius of curvature (but opposite inflection) as
in Figure 2.

Figure 2: Radius of curvature measurement using a classic Fizeau interferometer and a spherical test plate [5]


The radius deviation is then given by:
A =
z
2

2

Where
R = radius deviation of test surface
R = test plate radius of curvature
m = number of fringe spacing the fringes depart from straightness
d = diameter of test part
= wavelength of light

If a test surface has a defect (like a hole or a bump), the fringes will curve around it as in Figure
3.

Figure 3: Fizeau Fringes [5]
The height of the surface defect h is given by:
=
z

.
A
s

Where s is the fringe spacing and is the fringe displacement, as defined in Figure 3.
The direction in which curved fringes move when pressure is applied shows whether the test
surface is concave or convex. Pressing near the edge will cause the bulls eye (center of rings)
to move toward the point of application of pressure if the surface is convex, while the bulls eye
will move away from the point of application of pressure if the surface is concave (Figure 4a).
Pressing near the center of the ring system will cause the rings to expand for a convex surface,
and contract for a concave surface (Figure 4b).

Figure 4: Two methods to determine whether the surface under test is convex or concave with respect to the surface: (a) by
pressing near the edge and (b) by pressing near the center of the top plate. [1]

Table 1 gives a summary of the appearance of Newton fringes for different types of surfaces
with reference to a standard flat.

Table 1: Nature of Newton fringes for different surfaces with reference to a standard flat. [1]

Light Sources

Test plates can be used with white light, and the sequence of color fringes can provide an
experienced optician with more precise surface data than those obtainable with standard
monochromatic illumination.
If fringe definition is more important, an extended monochromatic source should be used.
Common monochromatic light sources for test plate testing are listed in Table 2.
Table 2: Common Extended Monochromatic Light Sources
Source Color Wavelength [nm]
Sodium D Bright orange 589.3
Mercury Dull green 546.1
Helium Pale orange 587.6
Matching Radii to Vendor Test Plates

Professional optical shops fabricate standard test plates from master test plates. Master test
plates are made very accurate in terms of radius accuracy and surface figure, and never used
for in-process testing. Working plates are then made from the standard plates. Working plates
are used by the optician in the shop at the machine, while standard plates are used in the
quality assurance lab. [2]
Because of the high cost of making test plates (typically several hundred dollars or more each),
and the long time to manufacture them (perhaps several weeks), lens designers should strive to
match all radii to existing vendor test plates. Vendor test plate lists can be found in most lens
design software packages, and these packages often also have automatic test plate fitting
routines.
Test Plate Materials

The choice of test plate material is influenced by the optical properties, wear resistance,
coefficient of expansion, required accuracy, availability and cost. Typical test plate materials
are listed here.
BK7: Rarely used anymore because it is relatively soft with a high coefficient of
expansion. Advantages include low cost, excellent optical properties and easy
availability. Easy to polish, but precise figure is difficult to control. Good for working
plates for noncritical applications.
Pyrex: Good for general quality optical flats and test plates. Most popular material for
standard or working test plates today. Commonly available, low expansion material.
Optical transparency and homogeneity not very good. Excellent for large plano test
plates.
Fused silica: Generally the best material for making optical flats. Very hard, wear-
resistant and very low coefficient of expansion. Optical properties depend on the grade
selected. Commercial quality is used for standard test plates, while optical quality is
reserved for master plates. Polishes well and holds a precise figure over a wide
temperature range. Expensive.
Cervit: Low-expansion glass ceramic. Good wear resistance. Undesirable optical
properties (yellow coloration and material inhomogeneities), but it is an excellent choice
for large reference quality flats.
Zerodur and ULE: High relative cost and large diameters may not be readily available.
Performance and Accuracy

Despite the availability of advanced shop interferometers which can test spherical surfaces to a
high degree of accuracy without the use of test plates, most optical shops still depend on test
plate testing as the only practical in-process testing method. The reasons for this are the easy
availability of the test plate at the machine and the speed with which the test can be
performed. On-the-block testing in an interferometer is very difficult. [2]
The accuracy of test plate testing can be limited by a variety of factors, for example:
Partial fringes that cannot be accounted for with accuracy.
Dirt particles trapped between surfaces.
Distortions caused by thermal effects.
Optical effects caused by viewing the fringes at oblique angles.
Distortions caused by material inhomogeneities.
When the test specification permits a few fringes, these effects can often be ignored. These
effects can also be minimized be following accepted test plate testing practice.
The accuracy of test plate testing is limited to about a fringe (1/8 wave), since this difference
can still be judged with reasonable accuracy. If no curvature is detected in the fringes, then we
can say that the test is at least 1/10 wave, and an interferometer test must be performed to get
a more accurate measurement. [2]
Test plate testing should be restricted to the shop floor as a fast and convenient in-process
check, while interferometers should be used for final quality control.
References

1. M. V. Mantravadi and D. Malacara, "Newton, Fizeau, and Haidinger Interferometers," in
Optical Shop Testing, 3
rd
Edition, D. Malacara, Ed. (Wiley, 2007) pp. 1-44.
2. H. H. Karow, Optical Shop Testing Methods and Instruments, in Fabrication Methods for
Precision Optics, (Wiley, 2004) pp. 584-643.
3. R. E. Fischer et al., Optical System Design, 2
nd
Edition, (SPIE Press, 2008) pp. 348-353.
4. E. P. Goodwin and J. C. Wyant, Field Guide to Interferometric Optical Testing, (SPIE Press,
2006).
5. OPTI 513R Class Notes.