A CANS is composed of the following elements:
A proton or deuton source producing a particle beam at energies on the order of 100 keV with a peak intensity up to 100mA
A RFQ stage (Radio-Frequency Quadrupole) whose role is to shape the continuous ion beam and ensure a first acceleration up to an energy of a few MeV
Transport lines to the target.
A target made of a material generating neutrons when interacting with protons
A moderator and a reflector whose role is to slow down the neutrons down to the energy requested by the final users (typ. 2 - 100 meV).
Several neutron beam lines bringing the neutrons to the spectrometers.
While on a reactor the number of produced neutrons is in the range of 3x1014 to 1015 n/s, the actual neutron flux on a sample is on the order of 107 n/s. That is only a fraction on the order of 10-7 of the produced neutrons is actually used. This leads to side effects such as massive shielding requirements.
In CANS, the term “Compact” refers to the Target Moderator Assembly (TMR) which can be made very small (a few liters) compared to reactor reflector vessels whose volume is in the 1m3 range. Hence while the raw number of neutron produced on a CANS can be small, a high brilliance of the source can be maintained in the small TMR volume. Only useful neutrons are produced. The whole philosophy of compact neutron source designs is “produce what you need”. The whole source is also physically compact (10-30m long) compared to spallation facilities (600m long) operating at very high proton energies (~1 GeV).
In Europe several institutes are considering high end CANS facilities using the latest available technologies to achieve High Brillance using low energy accelerators.
The CEA is considering a reference design SONATE with the following parameters:
Ep = 20MeV, Ipeak = 100mA, duty cycle = 4%, P = 80kW, fixed Be target.
These parameters were chosen partly because they correspond to the first 20m of the ESS Linac (out of 600m). Hence components (Source, RFQ and DTL) are available with no R&D developments.
For the SONATE design, a simple Target-Moderator-Reflector (TMR) was considered and a brilliance of 1.2x1011 n/cm²/s/sr was calculated using MCNP and GEANT4 Monte-Carlo simulations. This brillance value was used as an input in instrument Monte-Carlo simulations using McSTAS.
The table below compares the performances of different scattering techniques in terms of flux at the sample position. (n/cm²/s). These calculations suggest that a High Brillance CANS can provide performances on-par with medium flux reactors.