curved towards the end of the cutting- 
tool life. 
The flat chips are cut into pieces of 
2-cm length. Five pieces are placed 
side by side to form a square on a 
Perspex disk. This is centered under 
an end-window G-M tube 
The curved chips are broken into 
small pieces of a few millimeters length. 
These pieces are built up into a mosaic 
of approximately the same area as the 
groups of flat chips. 
The exact width, total length, and 
thickness of each group of chips is 
measured. The activity of one of the 
groups was followed for a few days, 
giving a half-life of 24 hr. 
Activity Measurement 
To calculate the absolute amount of 
material transferred to the chips, one 
of two methods can be used: 
1. The specific activity of the car- 
bide tips can be calculated from irradi- 
ation and nuclear data, if the counter 
efficiency is known. 
2. A weighed piece of the same 
material as the tip, used as a reference 
source, is irradiated together with the 
tip. The piece is dissolved in a suit- 
able chemical agent. A known part 
of the solution is evaporated, and the 
activity is measured under identical 
geometrical arrangements as for the 
measurements of the chips. Dividing 
all activity figures for the turnings with 
this figure gives directly the cor- 
responding amount of radioactive 
material. 
In the present work, the second 
method is used. The first method 
gives a check of the results. 
Reference sources. As the activity 
of the small reference pieces is too high 
= 
E 
S 
o 
= 
FIG. 4. Rate of material transfer from rake side to turnings 
to be measured without dilution, these 
are dissolved in molten sodium nitrite, 
the tungstate dissolved in water, and 
1/1,000 of the solution evaporated on 
an iron disk with the same area and 
bottom thickness as the turnings, thus 
making backscattering corrections 
unnecessary. 
The activity of the reference sources 
is measured under the same geometrical 
arrangements as for the chips. These 
operations are performed two days 
after the other work. Because of this, 
the reference-source activities are fol- 
lowed for a long period to permit an 
exact extrapolation. 
The curves show that a _ small 
amount of a long-lived activity was 
present in the sources. After complete 
decay of the W187, this remaining ac- 
tivity is determined and subtracted 
from the total source activity. 
A linear semi-logarithmic relation 
between activity and time results. 
The half-life values for W187 calculated 
from these curves are 23.9 and 24.2 hr, 
respectively. These are very close to 
the accepted figure of 24.1 hr (4). 
Calculations and Results 
For each group of turnings, a total of 
about 3,000 counts was made, giving a 
statistical error of 2%. All radio- 
activity figures are corrected for the 
deadtime of the counting equipment, 
for the background, and for the half- 
life of W187 (24.1 hr). 
A separate experiment was made to 
find the relation between area and 
counting rate (to correct for the differ- 
ences in area of the chips). This 
experiment showed that, within the 
range of areas used (230-350 mm?), this 
relation is linear. No corrections for 
Quality iS 
the backscattering of the beta particles 
are needed since the thickness of the 
chips gave saturation backscattering 
[backscattering factor = 1.45 (3)]. 
The beginning of each run is fixed as 
“zero time.” All radioactivity figures 
are corrected back to that time. 
The corrected counting-rate figures 
for the chips are divided by the count- 
ing rate of the corresponding reference 
source, giving the weight of the carbide- 
tip material transferred during a cer- 
tain period. The length of this period 
can be calculated from the cutting data 
(including the cutting ratio, i.e., the 
ratio chip thickness before:after re- 
moval), and thus the material trans- 
ferred in gm/min can be determined. 
Figure 4 shows that the material 
transferred per minute from the tip of 
quality no. 2 is almost constant. 
Quality no. 1 shows a large transfer 
during an initial period of 3 min, 
followed by an almost constant wear. 
The total amount transferred, found 
by integrating the curves of Fig. 4, is 
given in Fig. 5. As may be expected, 
the curve for quality no. 1 is a straight 
line; for quality no. 2, the relation is 
linear after the initial period. 
All the above figures of counting rate 
correspond to material transferred 
from the rake side of the carbide tip. 
Only’ for one group of turnings was the 
transfer from the flank also measured. 
For this group of chips, the rake-to- 
flank ratio is 3.5. 
The chips corresponding to 14.5 min 
turning with quality no. 2 gave an 
extraordinarily high counting rate 
(9,000 cpm, the ordinary being 400 
cpm). This was due to one piece. A 
small piece of the carbide tip had 
broken off and cut a groove in the chip, 
Quality 4 
‘Quality 2 
FIG. 5. Total wear of rake side as function of time 
153 
