技術資料72 views

Air Turbulence in Autocollimator Measurement: Sources, Magnitude and Field Countermeasures

A 0.1 °C gradient over a 1 m path deflects the beam by about 0.02″, and turbulence noise often reaches 0.2″-1″, the top disturbance in long-path work.

华旋光电技术部
Air Turbulence in Autocollimator Measurement: Sources, Magnitude and Field Countermeasures

Physical sources

The autocollimator beam travels back and forth through air whose refractive index varies with temperature, dn/dT about -1.0×10⁻⁶ per °C at standard pressure. A transverse temperature gradient bends the beam toward the warmer side: 0.1 °C across a 1 m path deflects it by about 0.1 μrad, roughly 0.02 arc-seconds. Worse than static gradients is turbulence: convection from heat sources — operators, lamps, electrical cabinets, a sunlit machine bed — makes the refractive-index field fluctuate randomly, and spot jitter over a 3 to 5 m path commonly reaches 0.2 to 1 arc-seconds, completely masking the resolution of a 0.1 arc-second instrument.

Order of magnitude

Practical figures: desktop measurements within a 1 m path show turbulence noise of about 0.05 to 0.1 arc-seconds; an unshielded 5 m path in a workshop exceeds 0.5 arc-seconds; a beam running within 50 mm of a warm machine bed multiplies the noise three to five times. In dual-axis work the vertical axis is more sensitive to rising thermal plumes, and vertical noise is often two to three times the horizontal noise.

Field countermeasures

  • Shielding: enclose the beam path in PVC pipe or a draft hood to isolate it from convection, cutting noise by an order of magnitude;
  • Averaging: turbulence is approximately zero-mean and random, so average at least 100 samples per point and the noise falls as 1/√N;
  • Shorten the path: measure long guideways in 2 to 3 m segments and stitch them through common points;
  • Avoid heat sources: let the machine reach thermal equilibrium before measuring, keep people away from the beam, and prefer early-morning stable conditions when possible;
  • Raise the beam: keep it at least 150 mm above warm surfaces, clear of the floor-level exchange layer.

Verification

If turbulence is suspected, run a static noise test: fix the reflector and record readings for five minutes; the peak-to-peak value is the environmental noise floor. When the floor exceeds five times the instrument resolution, fix the environment before discussing accuracy. The noise value should also enter the uncertainty budget as a Type A component.

⇄0比較仕様比較