Stir the Cavity To Improve Averaging

Measuring large-diameter or long radius of curvature optics frequently requires that the laser interferometer be located a significant distance from the test piece, sometimes tens of meters away. Air movement, convective currents, or turbulence over these long cavities will invariably cause variations in the density and refractive index of the intervening air. These variations will contribute additional phase structure in the measurement data and can be impossible to distinguishable from the shape of the part itself.

An example of the negative effect of convection currents in a measurement can be seen in this data from a two meter cavity measurement when a warm surface is placed beneath the test path.

Averaging several measurements together can significantly reduce the effect of air turbulence in measurement data. Over the course of many averaged measurements, zones of poorly mixed air will pass across the aperture or will dissipate, thereby only appearing at any location in the data for a small percentage of the measurements. The contribution of the turbulent air is therefore greatly reduced relative to the shape of the part, which remains constant across all frames.

Nevertheless, averaging may not always sufficiently remove turbulence from the data. The problem is particularly acute when the air is slow-moving and poorly mixed.

Given that interferometers have traditionally been tucked away in carefully controlled environmental labs it may seem sensible in these cases to try to reduce the turbulence, by shutting off air handlers, by measuring late at night when thermal currents are weakest, or by adding tents and baffles to block air flow.

In fact, it is often more helpful to do the exact opposite and increase air flow, typically by blowing a fan across the measurement cavity. The fan helps to mix the air in the cavity, effectively reducing the size of turbulent pockets and moving air more quickly across the frame.

Real-world testing: slow-moving air in this lab setup is stirred with a fan to reduce the effect of turbulence.

For shorter measurement path lengths, blowing a fan across the cavity has the greatest effect. Longer paths will require several fans to stir the entire cavity. Positioning the fan to blow the length of the cavity, as in the image below, may be the most effective solution in these cases. The generated air flow must influence the entire cavity or its effectiveness will be reduced.

It may be helpful to take a number of measurements with a fan in various configurations and examine the measurement precision to determine the most effective stirring method for your application. A Quality Control (QC) measurement in 4D's 4Sight analysis software will provide statistics about the measurement repeatability and precision to help you determine the optimal test setup.

Real-world testing: a fan stirs air over the length of this 10-meter cavity.