An optical engineer is sometimes required to specify surface roughness for optical components. There are several standards that can be used for reference, and each is appropriate in a given industry. However, all of the specifications ultimately refer back to ISO 4287 Parts 1 and 2 for definitions of terms and measurement methods.
The length of a measurement is a critical part of specificying surface roughness. You can measure one part over 0.1mm, 1mm and 10mm with the same instrument and get completely different answers for the same surface roughness measurement. To deal with this, surface measurement is typically divided into three regions: roughness, waviness and form. These correspond to measurements over short, medium and long distances. You get to decide the boundaries between these regions, but I typically specify waviness to cover the range from 80µm to 1mm. Roughness is then measured over a range from the minimum the instrument can measure up to 80µm and form covers any feature that extends more than 1mm. The following paragraphs cover how to specify roughness and waviness on an engineering drawing.
The ANSI/ASME B46.1 standard in the US and ISO 1302 in the rest of the world are the standards used by mechanical engineers and mold makers for specification of surface roughness. Recent editions of these standards have come a long way toward harmonizing the specifications, so they can normally be used interchangeably. In a picture, the specification looks like this:
Interpreting this figure is easier than it might look at first. The triangle/long division symbol (roughness symbol) is the basis for the callout; it does not change.
The first row of numbers and letters within the roughness symbol is a specification for surface roughness on a fine scale. The first number (0.0025) is the minimum feature size that need be measured, given in millimeters. The second number (0.8) is the sampling length – the distance over which these small features need to be measured, again in mm. The letters (Rz) specify the way the roughness data is to be interpreted. The most common ways of interpreting surface roughness are Ra (average roughness) and Rq (RMS average roughness). Mechanical engineers and mold makers typically specify Ra, while optical engineers prefer Rq for reasons that are discussed on the Scatter versus Surface Roughness page. The final number is the maximum allowable value for surface roughness, measured in nanometers. See the Scatter versus Surface Roughness page for reasonable numbers.
The second line of numbers and letters within the surface roughness symbol specify the “waviness” or “ripple” of the surface. This line is optional, but can be useful. The first number (0.8) is the minimum feature size for waviness. It should always be the same as the sampling length for surface roughness, which is why they are both blue in the figure. The second number on the waviness line is the sampling length – the maximum distance over which waviness is considered. This value should be appropriate for the instrument being used for the measurement. Surface roughness instruments are commonly capable of at least 2 mm sampling lengths, and some of them are capable of a couple cm. “Wt” is the standard symbol for waviness, and ideally the specification should be based on what will actually work for the optical surface.
For specifying surface roughness to a company that specializes in optics, the appropriate standard is ISO 10110 Part 8 “Surface Texture”. This standard allows for simplification of the above standard. For instance, just putting a “G” above the basic symbol and nothing within it specifies a ground surface. A “P”, optionally followed by a number between 1 and 4 specifies a polished surface. The number calls out the allowable number of microdefects per 10 mm length, with “1” specifying less than 400 and “4” specifying less than 3.
For plastic optics, it is possible to use the simple specifications published by the Society for the Plastics Industry. Of these, SPI A-1, which specifies polishing with 6000 grit diamond, is the only one worth considering for optically smooth surfaces. See the bottom of the Scatter versus Surface Roughness page for why you might not want to use this spec.
If you need surface roughness measured, calibrated surface roughness samples or the instruments to do it yourself, contact or call us. Schedule a consultation with our expert engineers to help figure out the best course of action.