Atomic Energy for Military Purposes (The Smyth Report)

The Official Report on the Development of the Atomic Bomb Under the Auspices of the United States Government


As was pointed out in Chapter VI, the control of a chain reacting pile is greatly facilitated by the fact that some of the neutrons resulting from uranium fission are not emitted until more than a second after fission occurs. It was therefore important to study this effect experimentally. Such experiments were described by Snell, Nedzel and Ibser in a report dated May 15, 1942 from which we quote as follows:

"The present experiment consists of two interrelated parts - one concerned with the decay curve, and one concerned with the intensity of the delayed neutrons measured in terms of that of the 'instantaneous' fission neutrons.


"The neutron source was the beryllium target of the University of Chicago cyclotron struck by a beam of up to 20 microA of 8 MeV deuterons. Near the target was placed a hollow shell made of tinned iron and containing 106 lbs. of U3O8. This was surrounded by about 2" of paraffin. The interior of the shell was filled with paraffin, except for an axial hole which accommodated a BF3- filled proportional counter. The counter was connected through an amplifier to a scaling circuit ('scale of 64') equipped with interpolating lights and a Cenco impulse counter. A tenth-second timer, driven by a synchronous motor, and hundredth-second stop watch were mounted on the panel of the scaler, close to the interpolating lights and impulse counter. This group of dials and lights was photographed at an appropriately varying rate by a Sept camera which was actuated by hand. The result was a record on movie film of times and counts, from which the decay curves were plotted.

"The actual procedure was as follows: During bombardment the stop watch was started and the timer was running continuously; the counter and amplifier were on, but the pulses leaving the amplifier were grounded. The scaler was set at zero. After a warning signal the cyclotron was shut off by one operator, while another operator switched the output of the amplifier from ground into the scaler, and started taking photographs. It was easy to take the first photograph within half a second of turning off the cyclotron. Sixty to a hundred photographs were taken during a typical run. The necessity of using both a stop watch and a timer arose from the fact that the hundredth-second precision of the stop watch was needed for the small time intervals between photographs during the initial part of the run, but the watch ran down and stopped before the counting was complete. The timer then gave sufficient precision for the later time intervals.

"Some forty runs were taken under varying experimental conditions. Short activations of one or two seconds were given for best resolution of the short periods. Long, intense bombardments lasting 15-20 minutes, as close as possible to the target, were made to make the long period activities show up with a maximum intensity. Some 5-minute bombardments were made, keeping the cyclotron beam as steady as possible, to study the relative saturation intensities of the various activities; in these activations the cyclotron beam was reduced to 1 or 2 &mul; A to prevent the initial counting rate from becoming too high for a counter (300 per sec. was taken as a reasonable upper limit for reliable counting). Two BF3 counters were available, one having a thermal neutron cross section of 2.66 sq. cm., and the other 0.43 sq. cm. After a strong activation, we could follow the decay of the delayed neutrons for some 13 minutes. Background counts (presumably chiefly due to spontaneous fission neutrons) were taken and were subtracted from the readings. They amounted to about 0.4 counts per sec. for the large counter.

"A study of all the decay curves gives the following as a general picture of the neutron-emitting activities present:

57 ± 3 sec. 0.135
24 ± 2 sec. 1.0
7 sec. 1.2
2.5 sec. 1.2

"Any activity of period longer than 57 sec. failed to appear even after the most intense bombardment we could give, lasting 20 minutes. The relative initial intensities given are the average values obtained from three curves.

"These results give the following equation for the decay curve, of the delayed neutrons after activation to saturation:

Activity = constant (1.2e-0.28t + 1.2e-0.099t + 1.0e-0.029t + 0.135e-0.012t)

where t is in seconds."

The second part of the experiment measured the total number of neutrons emitted in the time interval 0.01 sec. to 2.0 min. after the cyclotron was turned off. Assuming that all the delayed neutrons observed were in the four groups measured in the first part of the experiment, this second result indicated that 1.0 ± 0.2 per cent of the neutrons emitted in uranium fission are delayed by at least 0.01 sec. and that about 0.07 per cent are delayed by as much as a minute. By designing the effective value of k, the multiplication factor, for a typical operating pile to be only 1.01 with all the controls removed and the total variation in k from one control rod to be 0.002, the number of delayed neutrons is sufficient to allow easy control.