ind trams 
ice. The 
we found 
previous 
Ame after 
ry to wait 
setup be- 
y. 
sics argon 
or 1 week 
Switzer- 
he green 
ats of the 
letric test 
2 m and 
n our ho- 
along the 
leflected 
er which 
weighing 
e holder. 
"a meter 
weighed 
sed were 
quipped 
1y finish, 
ünted or 
ose parts 
'ot much 
was un- 
Night be- 
ywer and 
forming 
he screw 
  
yas studied 
nts in the 
wer left. 
which pressed against the shaft of the machine. The 
screw was rotated in a threaded hole in a steel spring 
which was clamped in a vice fixed to the table of the 
machine. Between the two exposures I simply went to 
the machine and rotated the spanner fixed to the screw. 
The double exposed spanner is seen in, e.g., Fig. 12 
where its positions before and after the application of 
the force produced a bright inclined vee which is seen 
in front of the horizontal shaft of the machine. 
The sandwich holograms were made the same way as 
described earlier. Each hologram plate was exposed 
with its emulsion forward in combination with a com- 
pensating plate in front or behind. No extra glass plate 
was used to increase the separation of the plates. The 
distance between the machine and the plate holder was 
about 4 m. The exposure was 5 sec on Agfa Gevaerts 
Scientia 10E 75 plates. 
To study the stability and find the suitable defor- 
mation range, real time fringes! were first studied. 
These fringes were never stable. During daytime they 
were almost impossible to see, and even during the night 
they never stopped moving. To decrease the influence 
of air movements we did not build, this time, any tent 
around the machine, but we tried to stop the cold airflow 
from the window and the hot air from the radiator by 
insulating with foam plastic. 
As usual in sandwich holography a reference surface 
in the form of a heavy piece of steel was placed close to 
the machine (lower left corner of Fig. 12). After a 
waiting time of a few hours we made an ordinary double 
exposure hologram and found that the reference surface 
was useful also for this type of holography. As seen in 
Fig. 12 it was covered by two fringes that had apparently 
been caused by air turbulence. The electric cable to the 
laser which passed in front of the machine was under- 
rated and was so hot that it could not be held by hand. 
After that the cable was changed and when moved be- 
hind the machine the fringes disappeared. 
A sandwich hologram of the machine is seen in Fig. 
13, and as the fringes are very similar to those of the 
double exposure and there are no fringes on the refer- 
ence surface we can conclude that the hologram presents 
a correct image of the deformation. The fringes in the 
lower right part of the machine have changed very 
slightly compared to Fig. 12, probably because of the 
removed cable. There is still a small variation in 
brightness on the reference surface indicating an error 
of a fraction of a fringe probably from air movement. 
All the sandwich holograms photographed here were 
reconstructed with the green light from an argon laser 
that had correct divergence. When we tried to recon- 
struct with the red light from a He-Ne laser, fringes of 
high contrast were formed but these fringes were erro- 
neous because of the difference in wavelength. Thus 
it was impossible to make the reference surface fringe 
free. About two fringes in the form of concentric circles 
could never be removed. The error was so small that 
it could be neglected when single details were studied 
which were smaller than half the reference surface, but 
if we wanted to compare the relative displacements of 
parts over the whole machine surface the errors had to 
be taken into account. However, using the argon laser 
  
Fig. 13. A sandwich hologram was made recording the same situation 
as in Fig. 12, but a heat source in front of the machine had been re- 
moved. The deformation and motion of the left support of the shaft 
were difficult to calculate because the whole over arm of the machine 
had bent and made a torsional motion. 
  
Fig. 14. The sandwich hologram of Fig. 13 was tilted so that the over 
arm became fringe free around where the left support is fixed. The 
number of fringes on the left support then became a measure of its 
motion in relation to the over arm. 
for reconstruction, these errors were totally eliminat- 
ed. 
Figure 13 shows the milling machine with its main 
body to the right, on top of which the over arm rests with 
its head at the upper left. From the over arm two 
supports stretch downward toward the shaft onto which 
the cutter is fixed. The vice with the loading screw is 
clamped to the table which rests on a saddle that can 
slide on the knee which finally is supported by a heavy 
September 1977 / Vol. 16, No. 9 / APPLIED OPTICS 2529