FIG. 1. 
irradiation, are arranged to make most 
efficient use of electrons as tray passes 
under beam 
SUTURES ON TRAY, ready for 
objects to be sterilized are not homoge- 
neous and have fixed dimensions. 
Therefore penetration has to be set 
according to the densest part of the 
object, resulting generally in reduced 
beam utilization. 
Consideration of these points in rela- 
tion to the product we proposed to 
sterilize indicated the need for an 
energy of 4-7 Mev and an output of at 
least 2.5-3.0 kw. 
Radiation Source 
Gamma emitters like Co® and waste 
fission products provide penetration 
considerably greater than our needs. 
However, their low dose rates necessi- 
tate very long dwell times for adequate 
dose. This creates some complex engi- 
neering problems to provide accurate 
and controllable dose. High-power 
X-ray machines also offer adequate 
penetration, but their power output is 
low and they pose many design prob- 
lems. Van de Graaff accelerators 
above 3 Mev are of tremendous size and 
comparatively low output. 
The source that appeared to meet our 
needs of adequate energy with high 
power output was the linac (2). 
Operating requirements for a com- 
mercially used linear accelerator are 
vastly different from those of universi- 
ties and research laboratories. Com- 
mercial sterilization demands continu- 
ous uninterrupted performance with 
unvarying beam characteristics. Low 
maintenance costs and ease of oper- 
ation are highly desirable, of course. 
Throughput. The production capac- 
ity of a machine can be given in several 
ways: (a) in kilowatts, (b) in mass proc- 
essed per hour at a certain dose (1 mega- 
rad equals 4.50 kw-see of radiation 
totally absorbed in 1 lb of product; at 
2.5 megarads a 1-kw device is capable 
146 
of irradiating 320 lb in an hour), (ec) in 
area coverage per minute for a certain 
dose (l-ma beam current delivers 1 
megarad to the useful-penetration 
thickness of 1,676 in?/min). 
The 100%-utilization capacities of 
our three machines are as given in 
the table on this page. 
Using irradiation from one side only, 
the maximum utilization of ionizing 
energy cannot be more than about 60%. 
This is the area under the equal-en- 
trance-and-exit-dose line in Fig. 2. 
The area above this line represents 
about 24% of the total ionizing energy. 
This part is really wasted as overdose 
in the sterilization process. The tail 
end beyond the useful penetration, 
representing about 16%, is also com- 
pletely lost because it leaves the 
product. 
Another important consideration is 
the efficiency with which an electron 
beam can be used. It is convenient 
to define three efficiencies, each corre- 
sponding to one of the dimensions of 
the product. Thus the scan efficiency 
is the fraction of the scan in which the 
beam hits the product. The loading 
Zo 
a 
@ 
on 
° 
i=) 
FIG. 2. PENETRATION CURVE meas- 
ured with stack of slides shows depth of 
“useful penetration” (A) and area of 
useful dose, area of overdose (B) and 
area of wasted dose (C) 
efficiency is the fraction of time in 
which the conveyor motion places a 
part of the load under the beam. The 
penetration efficiency is that portion of 
the beam (measured in the direction of 
electron motion) that is usefully em- 
ployed as indicated by Fig. 2. Our 
2-Mev Van de Graaff and our linac 
both have 80% scan efficiency and 95% 
loading efficiency. Penetration effi- 
ciency of the Van de Graaff is 15-25%, 
and that of the linac is 20-40%. 
Facility Design 
We decided to house both the 2-Mev 
Van de Graaff and the linac in the same 
area and place them close to our older 
Capabilities of Three Accelerators 
Van de Graaffs Linac 
Beam energy 
(Mev) 2 3 5 
Beam power (kw) 0.5 3.0 Pia 
Capacity 
lb/hr at 
3 megarads 160 960 672 
in?/min at 
3 megarads 167 = 667 280 
sterilization facilities. This provides a 
logical flow of materials through our 
manufacturing operation and permits 
the greatest flexibility for any later 
mechanization we want to undertake. 
The accelerator building is separate 
from the main manufacturing building 
but connected to it by an enclosed pas- 
sageway. This permits eventual ex- 
pansion of the manufacturing area in 
the direction of irradiation facilities 
and at the same time maintains our 
normal process flow. 
Split-level building. Consultations 
among our engineers, the architect and 
the accelerator builder led to design of 
a split-level accelerator building shown 
in the figure on p. 87. On the lowest 
level (partly underground) are two 
irradiation rooms each surrounded by 
7 ft of concrete. This thickness was 
based on calculations of probable hard 
X-rays generated by electrons striking 
the metal conveyor belt and other 
objects. 
Half a story higher and in front of 
the target room is the production area 
for loading and unloading conveyor 
belts. We have only recently extended 
the conveyor system to lead into the 
manufacturing area. Operating con- 
trols and monitoring devices are also 
on this floor. About 3 ft higher, above 
the target rooms, are the accelerator 
rooms. 
The highest level is half a story above 
the production area. This is a mezza- 
nine that houses the high-frequency 
amplifiers and the power supply for the 
linac. Maintenance workshop and 
testing instruments are also on the 
mezzanine. 
Access to both accelerator rooms and 
target rooms is through a passage with 
two right-angle turns. These turns 
effectively block all stray radiation. 
Installation of our linac began in late 
1957, and we started the checkout 
early in 1958. 
