Phase-shift shadow-moire topography is a noncontact optical technique for m
easuring the shapes of surfaces. Artifactual bands resembling isoheight sur
face contours are observed during measurement of small changes in shape by
use of this technique. The shape-reconstruction algorithm used in shadow-mo
ire topography is based on a mathematical model of the fringe patterns gene
rated on the surface to be measured. We hypothesize that the observed bands
reflect systematic errors caused by ignoring height-dependent terms in the
mathematical model of the fringe patterns. We test the assumption by simul
ating the fringe patterns for a virtual test surface by using a model that
contains height-dependent terms and one term that is idealized by ignoring
these terms. Small systematic errors in shape are observed only when the su
rface is reconstructed from fringe patterns simulated with a model containi
ng the height-dependent terms. Shape-error curves are computed as a functio
n of the surface height by the subtraction of the reconstructed shape from
the known shape. Simulated shape-error curves agree with experimental measu
rements in that they show an increase in error with surface height, and bot
h the experimental and the simulated shape-error curves contain ripples. Al
though the errors are small in comparison with the dimensions of the surfac
e and are negligible in shape measurements and in most deformation measurem
ents, they may show up as noticeable bands in images of small deformations.
(C) 2000 Optical Society of America. OCIS codes: 120.4120, 000.4430, 050.5
080.