Electrical pulse counters – pulse dividers – or shift registers: c – Applications – Counting animate or inanimate entities
Reexamination Certificate
2000-05-01
2001-12-25
Wambach, Margaret R. (Department: 2816)
Electrical pulse counters, pulse dividers, or shift registers: c
Applications
Counting animate or inanimate entities
C377S006000, C377S019000, C250S390010, C376S153000, C376S154000, C376S155000
Reexamination Certificate
active
06333958
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to a method and apparatus for improving the precision of neutron coincidence and multiplicity assays of plutonium and uranium.
BACKGROUND
Neutron-coincidence counting (“NCC”) is used routinely around the world for the non-destructive mass assay of uranium and plutonium in many forms. Passive neutron multiplicity counting (“PNMC”) is used routinely for the nondestructive assay (“NDA”) of plutonium scrap and waste. During the fission process multiple neutrons are emitted within a very short time frame, that is, in coincidence. The number of neutrons emitted in coincidence determines the multiplicity of the event.
For nondestructive analysis the sample in question is assayed by the detection of coincident fission neutrons from the spontaneous fission of, even-mass plutonium isotopes in the presence of a random neutron background (e.g., (&agr;,n) reactions).
As is well known,
3
He neutron detectors are much more efficient for detecting slow or thermal neutrons, compared to fast neutrons. When a thermal neutron collides with a
3
He molecule, a voltage pulse is produced.
FIG. 1
is a schematic illustrating this process, in which
11
is a gas tube having a casing
13
and an anode wire
15
. Casing
13
is filled with
3
He. Anode wire
15
is connected to source of high voltage
17
(e.g., 1700 volts), a capacitor
19
and a resistance
21
, as is well known in the art. The ionization resulting from a thermal neutron colliding with a
3
He molecule produces a voltage pulse, as is also illustrated in FIG.
1
.
Neutrons originating from (&agr;, n) reactions in the sample, from external sources, or different fissions are uncorrelated in time (i.e., random), whereas neutrons emitted by the same fissioning nucleus are time correlated. Typically, to distinguish correlated neutron events from random events (including neutrons from different fissions), two equal time periods are sampled by a coincidence circuit after a neutron has been detected. These circuits are also known as shift-register circuits. Los Alamos publication LA-UR-96-2462, “A 2-Fold Reduction in Measurement Time for Neutron Assay: Initial Tests of a Dual-Gated Shift Register (DGSR)” discloses unequal time periods, in which the R+A gate is 64 &mgr;s while the A gate is 1024 &mgr;s.
FIG. 2
is a neutron detection probability vs. time diagram depicting the operation of conventional shift register circuits. For a fission at time zero, the probability of detecting a fission neutron at time t decreases exponentially with time, namely:
P
(
t
)=(1/&tgr;) exp (−
t
/&tgr;) (Eq. 1)
Where &tgr; is the neutron “die-away time.” After a long delay, &Dgr;, the probability of detecting a neutron from a fission at t=0, is negligible. Therefore, upon detecting a neutron at time t, conventional shift register circuits count real coincidences R (neutron pulses from the same fission) plus accidental coincidences A (neutron pulses from other fissions plus time-random neutron pulses, e.g., from (&agr;,n) reactions), in the time interval t+p to t+p+G, where G is the gate length, and p is the predelay. The predelay p removes bias due to electronic deadtime effects. Upon detecting a neutron at time t, conventional shift register circuits also count accidental coincidences A in the interval t+p+G+&Dgr; to t+p+2G+&Dgr; where &Dgr; represents a long delay (e.g. 1 ms). At the end of the counting interval, one quantity of interest is the number of real coincidence pairs, or doubles (D), one NDA signature for fissile material mass. For doubles, the unfolding of R from R+A is simple:
D
=(
D+A
D
)−(
A
D
) (Eq. 2)
Where D is the real doubles and A
D
is the accidental doubles. The statistical error in D (precision) is given approximately by:
&sgr;
D
=[(
D+A
D
)+
A
D
]
½
(Eq. 3)
In most actual cases, D<<A
D
, and the D error is approximately
&sgr;
D
=(2
A
D
)
½
(Eq. 4)
The error model in Equation 3 is based on the assumptions of independent errors in (D+A
D
) and A
D
as well as Poisson statistics. Neither of these assumptions is valid for coincidence counting. However, this simple error model agrees with doubles sample-standard-deviation measurements to within a few tenths of a percent, depending on the item measured.
A conventional shift register circuit
111
, which is illustrated in
FIG. 3
, includes a predelay
113
, a shift register
115
, an up-down counter
117
, R+A accumulator (a/k/a sum)
119
, A accumulator
121
, R+A multiplicity accumulator
123
, A multiplicity accumulator
125
, and a strobe
127
. In operation, a pulse entering shift register
115
increments (+1) up-down counter
117
, while a pulse leaving shift register
115
decrements (−1) up-down counter
117
. Thus, the number of pulses in shift register
115
is just the count in up-down counter
117
. When a digital pulse
131
(a trigger pulse) crosses trigger point
133
, strobe
127
is triggered. The contents of up-down counter
117
are added to R+A accumulator
119
and A accumulator
121
, as well as R+A multiplicity accumulator
123
and A multiplicity accumulator
125
, as indicated by strobe arrows
135
,
137
,
139
and
141
. As those skilled in the art will appreciate, the strobe for accumulators
119
and
123
is simultaneous and occurs immediately upon a pulse crossing trigger
133
, whereas the strobe for accumulators
121
and
125
which is also simultaneous is delayed by long delay
143
(e.g. 1 ms). The total number of trigger pulses is accumulated in totals register
145
. Because a neutron pulse which enters the predelay
113
is produced at a later time than those neutron pulses already in shift register
115
, the R+A accumulator actually tallies events which precede the neutron pulses which strobe the accumulators. This is functionally equivalent to the conceptual timing diagram of FIG.
2
.
Precisions of neutron coincidence counting and neutron multiplicity counting are largely determined by the level of accidental coincidences pulses A. The higher A, the worse the precision. Thus, many neutron coincidence counting and neutron multiplicity counting assays are precision-limited, and require long count times for acceptable results. In the past, attempts to improve precision have been focused on detector design.
The article A New System for Analyzing Neutron Multiplicities: Characterization and Some Specific Applications, G. S. Brunson and G. J. Arnone, Los Alamos National Laboratories, LA-11701-MS (November, 1989), discusses decoupling the R+A and A accumulators. However, no purpose is mentioned for decoupling, and the disclosed circuit is not capable of the high sampling rates (e.g., 4 MHz) necessary for precision improvement. The circuit is limited to a pulse rate of 12.5 KHz and an A accumulator sampling rate of 125 KHz.
Accordingly, it is an object of the present invention to significantly improve neutron coincidence and neutron multiplicity counting precision by decoupling the sampling of the R+A gate from the A gate.
It is another object of the present invention to measure the R+A gate at the pulse rate, while measuring the A gate at a clock rate of 4 MHz, much faster (e.g., a factor of 5 to 10) than the pulse rate, thereby increasing the measurement precision of accidental coincidences. The greater the difference between the pulse rate and the clock rate, the greater the gain in precision of A. This, in turn, improves the precision of R, because R is obtained by unfolding R from R+A. Here, R, R+A, and A, can be either simple sums, in the case of conventional neutron coincidence counting, or pulse multiplicity distributions (0s, 1s, 2s, etc.) for neutron multiplicity counting.
It is a further object of the present invention to improve measurement precision, which permits, for a fixed precision, a significant reduction in measurement time.
It is a f
Bourret Steven C.
Krick Merlyn S.
Stewart James E.
Sweet Martin R.
Morgan DeWitt M.
Wambach Margaret R.
LandOfFree
Advanced electronics for faster time-correlation analysis of... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Advanced electronics for faster time-correlation analysis of..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Advanced electronics for faster time-correlation analysis of... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2589627