1g a dis- 
Thus the 
ulated by 
e crossed 
ne to the 
n the di- 
machine, 
ight par- 
rmation. 
It can be 
angle can 
‘es across 
' the ma- 
been de- 
d by the 
tool and 
nation of 
inge pat- 
the main 
en tilted 
| became 
tudy the 
1ces from 
gles that 
ange the 
tilt of the 
t to mag- 
ponds to 
the head 
n the ref- 
e for the 
that only 
re we ex- 
ifference 
made an- 
  
came three 
h hologram 
uld hardly 
nee of the 
caused by 
  
Fig. 8. The sandwich hologram of Fig. 7 was tilted so that the de- 
formation of the knee could easily be studied. The closely spaced 
fringes that were moved from the knee to the reference surface could 
no longer be resolved because of the large speckles. However, the 
magnitude, direction, and sign of knee tilt could be evaluated from 
the tilt of the hologram. 
  
Fig. 9. The sandwich hologram of Fig. 7 was tilted to make possible 
a study of the deformation of the machine head. 
other change too because during our earlier experiment 
we found that the sandwich tilt needed to eliminate the 
fringes on the table was uncomfortably large. By 
making the separation larger between the two emulsions 
of the sandwich hologram (d of Fig. 1, Ref. 5) the tilt 
angle could be kept smaller. The two holograms were 
placed with their emulsions outward instead of forward, 
and moreover we added a third glass plate (hologram 
plate with the emulsion removed) between the two 
plates to provide extra separation. 
  
Fig. 10. The load was finally increased still more so that the fringes 
on the knee could not be resolved at all. 
Fig. 11. After some tilt of the sandwich hologram of Fig. 10, however, 
fringes appeared on the knee, and further tilt revealed its 
deformation. 
Figure 7 shows the result when no tilt has been in- 
troduced to this new sandwich hologram. The tilt of 
the knee is now so large that the fringes could not quite 
be resolved. However, after a tilt of the sandwich ho- 
logram during reconstruction fringes appeared on the 
knee, and after some further tilt the number of fringes 
crossing it was set at a minimum (Fig. 8). Now it is 
possible to measure the deformation of the knee of the 
milling machine in spite of this deformation being so 
small that it only caused about eight of the about 125 
September 1977 / Vol. 16, No. 9 / APPLIED OPTICS 2527