DE102010032216B4 - Pulsed spallation neutron source - Google Patents

Abstract

The invention relates to a device for generating a time-smoothed neutron flux, in particular for a subcritically operated nuclear reactor. The device comprises a particle accelerator which generates a pulsed particle beam with a very high pulse repetition rate of, for example, 3 kHz. The particle beam hits a spallation target, from which it knocks out a likewise pulsed neutron flux, which in turn reaches a neutron moderator. The neutron moderator has a known distribution of dwell times. A temporal smoothing of the neutron flux leaving the moderator is achieved in that the dwell time and the pulse repetition rate are adapted to one another, either by choosing a moderator who has a suitable distribution of dwell times, or by setting a suitable pulse repetition rate fpuls for a specific, fixed moderator becomes. E.g. 1 / fpuls = TV, 0, where TV, 0 indicates the most likely dwell time of the distribution.

Description

The invention relates to a neutron source for use in a subcritical nuclear reactor.

In a subcritical nuclear reactor, the nuclear reactions take place without the reactor going into the critical state. Instead of maintaining a self-contained chain reaction, an external neutron source is used to provide the neutrons needed for the nuclear reactions. This purpose is fulfilled, for example, by a particle accelerator whose particle beam is directed towards a spallation target from which neutrons are subsequently ejected, which finally reach the subcritical nuclear reactor in order to effect the nuclear reactions there. This technology is known per se, for. B. from the WO 01/03142 A2 , Such subcritical nuclear reactors that utilize a particle accelerator are referred to as ADS (Accelerator Driven System) reactors, or ADSRs for short.

Subcritical reactors have a number of advantages. For example. they can be used to destroy heavy isotopes contained in the spent fuel of conventional nuclear reactors, while at the same time generating energy. Theoretically, the transuranium contained in nuclear waste can be split, which in turn energy can be obtained, while the resulting cleavage products are less durable than the transuranic. Such a process is z. B. from the US 2006/0215799 A1 known. Further development of the subcritical reactors and their driving accelerator driven neutron sources is therefore of great interest.

• – Die normalleitenden Beschleuniger müssen in der Praxis gepulst betrieben werden, da die geforderten Baugrößen zu elektrischen Feldern in den Beschleunigerzellen (Resonatoren bzw. Kavitäten) führen, die in den umgebenden, feldumschließenden Leiterstrukturen, bspw. in den Resonatorwänden, eine exzessive Verlustleistung verursachen. Ein gepulster Beschleunigerbetrieb und in der Folge ein gepulster Neutronenfluss aus dem Spallationstarget hat jedoch zur Folge, dass der Reaktorkern in der Energieproduktion stark schwankt und damit thermisch und mechanisch stark belastet wird.
• – Dieses Problem könnte unter Verwendung eines supraleitenden Beschleunigers gelöst werden. Dieser könnte zwar kontinuierlich betrieben werden und damit das genannte Problem des normalleitenden Beschleunigers vermeiden, jedoch wäre eine kompakte Baugröße nicht realisierbar. Bspw. für die Europäische Spallations-Neutronenquelle ESS ergäbe sich eine Länge von 630 m. Aufgrund dieses Nachteils ist die Verwendung eines supraleitenden Beschleunigers für eine weitestgehend kommerzielle Anwendung nahezu ausgeschlossen.

For the accelerator structures suitable for this purpose, a distinction must be made between normal-conducting and superconducting accelerators:

• - The normal-conducting accelerator must be operated pulsed in practice, since the required sizes lead to electric fields in the accelerator cells (resonators or cavities), which cause in the surrounding field enclosing conductor structures, for example. In the resonator walls, an excessive power loss. However, a pulsed accelerator operation and, as a consequence, a pulsed neutron flux from the spallation target has the result that the reactor core fluctuates greatly in energy production and is thus heavily loaded thermally and mechanically.
• This problem could be solved using a superconducting accelerator. Although this could be operated continuously and thus avoid the mentioned problem of normal-conducting accelerator, but a compact size would not be feasible. For example. for the European spallation neutron source ESS a length of 630 m would result. Due to this disadvantage, the use of a superconducting accelerator for commercial use is almost impossible.

It is therefore an object of the present invention to provide a compact device for generating a time-smoothed neutron flux whose use, for example. With a subcritical nuclear reactor does not have the disadvantages mentioned.

This object is achieved by the features specified in the independent claims. Advantageous embodiments emerge from the dependent claims.

The solution according to the invention starts from a normal-conducting accelerator operated in pulsed fashion with the aid of a correspondingly established control device, whose pulsed particle beam is directed onto a spallation target from which neutrons are knocked out in a known manner. In this case, an extremely high pulse repetition rate is selected, for example of the order of ≥3 kHz. Downstream of the spallation target, d. H. between the target and the reactor core, a neutron moderator with a known neutron residence time is arranged, with which it is to be achieved that the reaching into the subcritical reactor core neutron flux from the target is temporally smooth, so that the passing of the moderator into the reactor core neutron flux does not vary significantly. For example. would be an order of magnitude of up to about 20% acceptable. In principle, however, this of course depends on the respective intended application.

This is achieved by either using a neutron moderator whose neutron residence time, that is to say a size dependent on the material and dimensions of the moderator, approximately corresponds to the selected pulse repetition rate, or the pulse repetition rate, for example, with the aid of a corresponding control device for the particle accelerator the known Neutronenverweilzeit is adjusted. Furthermore, of course, the residence time is also dependent on the speed of the neutrons and thus to a certain extent from the target. An accurate residence time curve can be obtained, for example, by simulating the overall system.

For example. a pulse repetition rate or, accordingly, a time interval between two successive pulses of the pulse sequence is selected, which corresponds approximately to the neutron dwell time. Ie. at a repetition rate of, for example, 3 kHz, the dwell time in the moderator should be approximately 1 / (3 kHz) ≈ 0.33 ms or more. This has the consequence that the neutron flux from the moderator, despite the pulsed particle beam and despite the pulsed neutron flux leaving the target, does not vary significantly in time and, accordingly, the reactor core driven by the neutron flux does not perform excessive power fluctuations.

Due to a possibly more complex geometry of the moderator and / or the entire arrangement, the paths of the individual neutrons may differ more or less strongly by the moderator, so that it is no longer possible to speak of a single neutron residence time but of a distribution of the neutron residence times between one Minimum value and a maximum value. In this case, there are several possibilities for adjusting dwell time and repetition rate, as explained in connection with the figures.

• – einen Teilchenbeschleuniger mit einer Steuervorrichtung, wobei die Steuervorrichtung derart einstellbar ist, dass mit dem Teilchenbeschleuniger ein gepulster Teilchenstrahl S umfassend einen Pulszug mit einer Vielzahl von Pulsen mit einer Pulsrepetitionsrate f_puls erzeugbar ist,
• – ein Spallationstarget, das im Teilchenstrahl S positionierbar ist und aus dem durch den Teilchenstrahl S ein eine Vielzahl von Neutronen umfassender erster Neutronenfluss N_tar herausschlagbar ist,
• – einen Neutronenmoderator, der eine bekannte Verteilung X_n(T_V) von Neutronenverweilzeiten T_V aufweist, wobei das Target und der Moderator derart zueinander angeordnet sind, dass zumindest ein Teil des ersten Neutronenflusses N_tar in den Moderator gelangen und diesen durchqueren kann, so dass der zu erzeugende Neutronenfluss N_mod den Moderator verlässt.

The device according to the invention for generating a neutron flux N _mod has:

• A particle accelerator having a control device, wherein the control device is adjustable such that with the particle accelerator a pulsed particle beam S comprising a pulse train with a multiplicity of pulses with a pulse repetition rate f _{puls can be} generated,
• A spallation target which can be positioned in the particle beam S and from which a first neutron flux N _tar comprising a multiplicity of neutrons can be _knocked out by the particle beam S,
• A neutron moderator having a known distribution X _n (T _V ) of neutron residence times T _V , the target and the moderator being arranged relative to one another such that at least part of the first neutron flux N _tar can reach the moderator and pass through it that the neutron flux N _mod to be generated leaves the moderator.

The pulse repetition rate f _puls and the distribution X _n (T _V ) of the neutron residence times T _{V of} the moderator are now matched to one another in such a way that the neutron flux N _mod leaving the moderator is smoothed over time.

For this purpose, the control device is adjustable in such a way that the pulse repetition rate f _pulse corresponds to the neutron residence time T _V.

Specifically, the control device is adjustable so that for the time interval ΔT _pulse = 1 / f _pulse between two successive pulses of the particle beam S:
ΔT _pulse = 1 / f _pulse ≥ T _{V, min} , where T _{V, min is} the minimum residence time of the distribution X _n (T _V ) of the neutron residence times T _V.

In this case, the following applies for the time interval ΔT _pulse = 1 / f _pulse between two successive pulses of the particle beam S: ΔT _pulse = 1 / f _pulse ≦ T _{V, max} , where T _{V, max is} the maximum dwell time of the distribution X _n (T _V ) of the neutron residence times T _V.

This relation between ΔT _pulse , T _{V, min} and T _{V, max} ensures that the neutron flux N _mod triggered by a first pulse of the particle beam has not yet decayed when a further neutron flux N _{mod is} generated by the subsequent pulse of the particle beam ,

The control device may also be adjustable in such a way that for the time interval ΔT _pulse = 1 / f _pulse between two successive pulses of the particle beam S:
ΔT _puls = 1 / f _pulse ≈ T _{V, 0} , where T _{V, 0 is} the most probable residence time of the distribution X _n (T _V ) of the neutron residence times T _V.

It is chosen a pulse repetition rate f _pulse of at least 3 kHz.

The moderator is arranged in such a way that the neutron flux N _mod leaving the moderator passes into a reactor core of a nuclear reactor, which operates in a subcritical manner in particular.

Due to the smoothing of the neutron flux, the nuclear reactor or its core is neither thermally / mechanically heavily loaded despite the use of a pulsed particle beam nor fluctuates the energy production beyond a tolerable level.

The particle beam is advantageously a proton beam.

In the method according to the invention for generating a neutron flux N _mod , a particle accelerator generates a pulsed particle beam S comprising a pulse train having a multiplicity of pulses with a pulse repetition rate f _puls . The pulsed particle beam is directed onto a spallation target, wherein a first neutron flux N _tar comprising a multiplicity of neutrons is _knocked out of the target by the particle beam. At least part of the first neutron flux N _tar passes into a neutron moderator and traverses it, so that the neutron flux N _mod to be generated leaves the moderator. The moderator has a known distribution X _n (T _V ) of neutron residence times T _V. The neutron flux N _mod leaving the moderator is now time-smoothed by the pulse repetition rate f _puls and the distribution X _n (T _V ) of the Neutron dwell times T _{V of} the moderator are matched.

The adaptation is carried out in such a way that for the time interval (ΔT _pulse = 1 / f _pulse ) between two successive pulses of the particle beam (S): ΔT _pulse = 1 / f _pulse ≥ T _{V, min} , where T _{V, min} the minimum residence time of the distribution (X _n (T _V )) of the neutron residence times (T _V ).

Furthermore, for the time interval (ΔT _pulse = 1 / f _pulse ) between two successive pulses of the particle beam (S): ΔT _pulse = 1 / f _pulse ≦ T _{V, max} , where T _{V, max is} the maximum residence time of the distribution (X _n (T _V )) of the neutron residence times (T _V ).

In particular, the adaptation is carried out in such a way that the following applies for the time interval (ΔT _pulse = 1 / f _pulse ) between two successive pulses of the particle beam (S):
ΔT _puls = 1 / f _pulse ≈ T _{V, 0} where T _{V, 0 is} the most likely residence time of the distribution (X _n (T _V )) of the neutron residence times (T _V ).

In summary, the problem is solved by the use of a particle accelerator, which is operated with extremely high pulse repetition rate, in combination with a suitable neutron moderator, which temporally smoothes the neutron flux leaving the target. The neutron moderator or the neutron residence time in the moderator and the pulse repetition rate are adapted to each other.

In order to achieve the high pulse repetition rate, an RF transmission arrangement is necessary, which allows correspondingly short filling times of the RF resonators of the accelerator path (maximum RF power with varying load). A solution to this problem is not the subject of the present invention, but is found in the DE 10 2009 053 624 A1 , There, RF cavities or resonators as well as an accelerator equipped with these RF cavities are indicated, whereby in particular the problem of the coupling of RF power into the RF cavity is in the foreground. It is therefore with regard to the problem of low filling times of the resonators, which is solvable with a suitable device for coupling the RF power in the resonators, on the DE 10 2009 053 624 A1 directed.

Further advantages, features and details of the invention will become apparent from the embodiment described below and with reference to the drawings.

Showing:

1 the device according to the invention for producing a time-smoothed neutron flux for a subcritical nuclear reactor,

2 the temporal behavior of the pulsed particle beam and the different neutron fluxes and

3 an exemplary distribution of neutron residence times.

The 1 shows a device for generating a temporally smoothed neutron flux N. Shown are the core 110 a subcritical nuclear reactor 100 as well as a spallation target 120 and a neutron moderator 130 , The target 120 becomes with a proton beam S from a particle accelerator 200 irradiated, wherein the particle accelerator 200 an ion source 220 , an accelerator route 230 with at least two RF resonators 231 . 232 and a control device 210 for driving the RF resonators 231 . 232 having.

The RF resonators 231 . 232 the accelerator route 230 are arranged one behind the other in the beam direction and cause the acceleration of the ion source 220 removable particles, for example. Protons.

By the control device 210 be in the RF resonators 213 . 232 Generates RF fields which serve to accelerate the particles taken from the ion source. This basic mode of operation of such accelerators, including the manner of generating the RF fields in the RF resonators, for example using a klystron, is known and will not be described further here. The wording that "the control device 210 generates the RF fields ", this should include per se known mode of action, ie the control device 210 includes both the components needed to generate the RF fields in the resonators, including, for example, the klystron and the possibly used waveguide, etc., as well as an electronic assembly that performs the various necessary calculations as well as the actual control of the components.

With the help of the control device 210 can the RF resonators 231 . 232 as well as in DE 10 2009 053 624 A1 be described separately, ie the RF field of the first RF resonator 231 can be independent of the RF field of the second resonator 232 be generated. Due to the achievable RF decoupling, it is possible, the individual RF cavities 231 . 232 with the help of the control device 210 independently of each other, causing the accelerator 200 more flexible and more flexible to each reaching, desired acceleration can be adjusted. The adaptation is more flexible than with an accelerator, in which the RF cavities in the RF region are coupled together so that the control of one RF cavity simultaneously influences the RF fields in the adjacent RF cavities. Nevertheless, the invention set forth here can also be realized with an accelerator in which the individual RF resonators 231 . 232 can not be controlled separately.

Regardless of whether the RF resonators 231 . 232 be controlled separately, is the control device 210 set up such that the particle accelerator 200 is operated pulsed, that is the the particle accelerator 200 removable particle beam S as a pulse train consisting of successive particle packets or pulses, so-called bunches.

In the 2A is a pulse train P (t) of the particle beam or the particle flux density of the particle beam purely exemplary and shown in arbitrary units as a function of time t.

As described in the introduction, the particle beam S causes the spallation target 120 of the nuclear reactor 100 that from the target 120 Neutrons are knocked out. Due to the pulsed particle beam S is the target 120 leaving neutron flux N _tar also pulsed ( 2 B ), ie the neutron flux N _tar has a similar temporal behavior as the pulsed particle beam S.

This pulsed neutron flux N _tar first enters a neutron moderator 130 , in which the neutrons are slowed down. The moderator 130 leaving neutron flux N _mod ( 2C ) gets into the nucleus 110 of the nuclear reactor 100 where finally the desired nuclear reactions take place. N _mod gives the number of the moderator 130 neutron leaving per unit time. The temporal behavior of N _mod is the distribution of the residence times T _V in the moderator 130 dependent. The residence time describes the period of time that the neutrons need to be the moderator 130 to cross. The residence time is therefore dependent on the initial speed of the target 120 leaving or in the moderator 130 penetrating neutrons, from the braking power of the moderator 130 and by the length of the path that the neutrons pass through the moderator 130 cover, ie its geometry and dimensions.

The 3 shows by way of example and in a greatly simplified manner the distribution of the residence times T _V by a pulse of the particle beam S from the target 120 knocked out neutrons in the moderator 130 , Plotted is the number of neutrons X _n as a function of the residence time T _V. The curve is therefore to be interpreted such that, for example, a certain number X _{n, 1} remains in the moderator for a period T _{V, 1} before it enters the reactor core. Accordingly, a number X _{n, 2} remains for a period T _{V, 2} in the moderator, etc. Thus, the area under the curve corresponds to the total number of neutrons knocked out of the target by a pulse of the particle beam (provided that none of these neutrons are lost goes). In the 3 it was assumed that the trajectory X _n (T _V ) shows a quadratic behavior or is parabolic, ie in principle
X _n (T _V ) = -C 1 * (T _V - T _{V, 0} ) ² + C ² , where C 1, C ² and T _{V, 0 are} constants that depend on the choice of moderator.

Of course, in practice, the shape of the trajectory depends essentially on the geometry and material of the moderator 130 , Geometry and material of the target 120 and on the time length and shape of the pulses of the particle beam S from. T _{V, min} and T _{V, max} denote the minimum and the maximum residence time of the neutrons in the moderator. T _{V, 0} generally designates the residence time that the majority of neutrons spend in the moderator or the residence time for which the curve X _{n has} a maximum. T _{V, 0} therefore can be described as the most likely dwell time.

The neutron flux N _mod (t) caused by a single pulse has a qualitatively substantially similar temporal behavior to that in the 3 illustrated curve. Assuming that the pulse is the target 120 At the time t = 0, the first neutrons leave the moderator at a time t _{V, min} 130 , At a time t _{V, 0} , the neutron flux leaving the moderator is maximum, while the neutron flux has decayed again at the time t _{V, max} .

By a suitable choice of the moderator 130 , in particular with regard to material, geometry and arrangement relative to the target 120 , the trajectory according to 3 or the distribution of the residence times T _{V are} influenced. Conversely, a particular moderator stands out 130 by a certain Neutronenverweilzeit T _V or by a certain distribution of residence times T _V out. In an extreme case, an arrangement would be conceivable, with the distribution of the residence times not as in the 3 is relatively wide, but is essentially reduced to a comparatively narrow single peak at T _V = T _{V, 0} . For example. could be the moderator 130 seen in the beam direction behind the target 120 may be arranged so that only the neutrons leaving the target straight ahead in the beam direction strike the moderator, so that these neutrons essentially require the same time period T _{V, 0} to pass through the moderator.

For the course of the N _mod curve in 2C It has now been assumed that the distribution of the residence times T _V or the chronological behavior of the pulse-triggered neutron flux N _{mod of} the curve of the 3 equivalent.

In the 2C is based on the behavior according to 3 the neutron flux N _mod resulting from a plurality of pulses. It is indicated by the dashed curves for each pulse of separately triggered by this pulse neutron flux. The entire neutron flux N _mod results from a simple summation of the individual neutron fluxes.

The entire superimposed neutron flux N _mod from the moderator 130 according to 2C has a completely different time behavior than the neutron flux triggered by a single pulse 3 After a certain number of pulses or after the transient phase, the neutron flux N _{mod is} temporally smoothed and fluctuates between a minimum value N _{mod, min} and a maximum value N _{mod, max} with ΔN _mod = N _{mod, max} -N _{mod, min} .

The difference between the minimum and the maximum value of the neutron flux N _mod , ie the fluctuation ΔN _{mod of} the neutron flux, now depends on several factors. The interaction between the time interval ΔT _{pulse of} the pulses of the particle beam on the one hand and the distribution of the residence times T _V for the neutrons of a pulse according to 3 on the other hand, the temporal behavior of the neutron flux N _mod ultimately determines.

If, in simple terms, the maximum residence time T _{V, max is} smaller than the time interval ΔT _{pulse of} the pulses of the particle beam, the minimum value N _{mod, min of} the neutron flux N _mod becomes comparatively small, ie N _{mod, min} ≈ 0, since the By a first pulse triggered neutron flux has already decayed when the triggered by the subsequent pulse neutron flux occurs. So there is no accumulation of neutron fluxes and no settling. Accordingly, the fluctuation ΔN _mod of the moderator 130 leaving neutron flux N _mod very large, which should be avoided as explained in the introduction. Conversely, that can be in the 2C reach behaved when T _{V, max} > .DELTA.T _pulse applies. In the exemplary example there, the ratio T _{V, max} : ΔT _{pulse is} approximately 3: 1.

According to the invention, the moderator and the pulse repetition rate are matched to one another. At a fixed predetermined moderator with known properties, especially regarding. The distribution of X _n Neutronenverweilzeiten T _V, with the aid of the control device 210 the pulse repetition rate f _{pulse can} be adjusted accordingly.

Conversely, the selection and interpretation of the moderator may also change 130 after a desired pulse repetition rate f _pulse = 1 / ΔT _{pulse of} the particle beam S judge, ie the type or material and the dimensions and the geometry of the moderator 130 are selected such that the distribution X _{n of} the residence times T _{V of} the neutrons of the neutron flux N _tar from the target 120 in the moderator 130 adapted to the pulse repetition rate.

In both said cases of adaptation, it could be used as an adaptation criterion such that the minimum residence time T _{V, min of} the distribution X _n (T _V ) corresponds at least to the time interval ΔT _pulse between two successive pulses of the particle beam, ie ΔT _pulse = 1 / f _pulse ≥ T _{V, min} . A further limiting criterion, which is also used to create the diagram of the 2C T _{V, min} ≦ ΔT _pulse ≦ T _{V, max} or concrete ΔT _pulse ≈ T _{V, 0} would be used, for example. Depending on the width of the distribution of the residence times, in this case ΔT _pulse must correspond more or less exactly to T _{V, 0} .

The control device 220 is now set up so that it generates an extremely high pulse repetition rate of, for example, f _pulse ≥ 3 kHz.

Of course, the high pulse repetition frequency of ≥ 3 kHz requires the shortest possible filling times of the RF energy in the RF cavities of the accelerator, ie a maximum transmission power. This is a worthwhile goal, especially with high-performance accelerators. The power absorbed by the particle beam in the cavity should, as is known, also dominate the power loss in the accelerator in the interest of high efficiency. Thus, during the filling or transient phase, ie without particle beam, the entire available transmission power can be used to fill the cavity, so that short filling times are made possible. The RF concept of the accelerator must be designed so that as much RF power as possible is available during the filling phase, ie with variable HF load. A suitable device including the appropriate RF cavity is, for example. In DE 10 2009 053 624 A1 described.

Claims (13)

Device for generating a neutron flux (N _mod ), comprising: - a particle accelerator ( 200 ) with a control device ( 210 ), wherein the control device ( 210 ) is adjustable so that with the particle accelerator ( 200 ) a pulsed particle beam (S) comprising a pulse train with a plurality of pulses with a pulse repetition rate f _{pulse can be} generated, - a spallation target ( 120 ), which can be positioned in the particle beam (S) and from which a first neutron flux (N _tar ) comprising a plurality of neutrons can be knocked out by the particle beam (S), - a neutron moderator ( 130 ) having a known distribution (X _n (T _V )) of neutron residence times (T _V ), the target ( 120 ) and the moderator ( 130 ) are arranged to each other such that at least a part of the first neutron flux (N _tar ) in the moderator ( 130 ) and can traverse this, so that the neutron flux (N _mod ) to be generated leaves the moderator, characterized in that the pulse repetition rate (f _puls ) and the distribution (X _n (T _V )) of the neutron residence times (T _V ) of the moderator ( 130 ) are so coordinated that the moderator ( 130 ) leaving neutron flux (N _mod ) is temporally smoothed. Device according to claim 1, characterized in that the control device ( 210 ) is adjustable such that the pulse repetition rate (f _pulse ) corresponds to the neutron dwell time (T _V ). Device according to claim 1, characterized in that the control device ( 210 ) is adjustable such that for the time interval (ΔT _pulse = 1 / f _pulse ) between two successive pulses of the particle beam (S): ΔT _pulse = 1 / f _pulse ≥ T _{V, min} , where T _{V, min is} the minimum Residence time of the distribution (X _n (T _V )) of the neutron residence times (T _V ). Device according to claim 3, characterized in that the control device ( 210 ) is adjustable such that for the time interval (ΔT _pulse = 1 / f _pulse ) between two successive pulses of the particle beam (S): ΔT _pulse = 1 / f _pulse ≦ T _{V, max ,} wherein T _{V, max} the maximum residence time of the distribution (X _n (T _V )) of the neutron residence times (T _V ). Device according to one of the preceding claims, characterized in that the control device ( 210 ) is adjustable such that for the time interval (ΔT _pulse = 1 / f _pulse ) between two successive pulses of the particle beam (S) applies: ΔT _pulse = 1 / f _pulse ≈ T _{V, 0} , where T _{V, 0} the most likely Residence time of the distribution (X _n (T _V )) of the neutron residence times (T _V ). Device according to one of the preceding claims, characterized in that the pulse repetition rate (f _pulse ) is at least 3 kHz. Device according to one of the preceding claims, characterized in that a nuclear reactor ( 100 ), in particular a subcritical nuclear reactor ( 100 ), with a reactor core ( 110 ), the moderator ( 130 ) is arranged such that the moderator ( 130 ) leaving neutron flux (N _mod ) in the reactor core ( 110 ). Device according to one of the preceding claims, characterized in that the particle beam is a proton beam. Device according to one of the preceding claims, characterized in that the particle accelerator ( 200 ) at least two in the beam direction successively arranged RF resonators ( 231 . 232 ), by means of which the pulse train comprising a plurality of particle pulses can be accelerated. Method for generating a neutron flux (N _mod ), in which - a particle accelerator ( 200 ) generates a pulsed particle beam (S) comprising a pulse train with a plurality of pulses at a pulse repetition rate f _puls , - the pulsed particle beam (S) onto a spallation target ( 120 ), wherein from the target ( 120 ) a first neutron flux (N _tar ) comprising a multiplicity of neutrons is ejected by the particle beam, - at least part of the first neutron flux (N _tar ) is converted into a neutron moderator ( 130 ) and passes through it so that the neutron flux (N _mod ) to be generated leaves the moderator, the moderator ( 120 ) has a known distribution (X _n (T _V )) of neutron residence times (T _V ), characterized in that the moderator ( 130 ) Leaving neutron flux (N _mod) is smoothed in time by the pulse repetition rate (f _pulse) and the distribution (X _n (T _V)) of the Neutronenverweilzeiten (T _V) of the moderator ( 130 ) are coordinated. A method according to claim 10, characterized in that the adaptation is carried out such that for the time interval (ΔT _pulse = 1 / f _pulse ) between two successive pulses of the particle beam (S) applies: ΔT _pulse = 1 / f _pulse ≥ T _{V , min} , where T _{V, min is} the minimum residence time of the distribution (X _n (T _V )) of the neutron residence times (T _V ). A method according to claim 11, characterized in that the adaptation is executed such that for the time interval (ΔT _pulse = 1 / f _pulse ) between two successive pulses of the particle beam (S) applies: ΔT _pulse = 1 / f _pulse ≤ T _{V, max} where T _{V, max,} the maximum residence time of Distribution (X _n (T _V )) of the neutron residence times (T _V ). Method according to one of claims 10 to 12, characterized in that the adaptation is carried out such that for the time interval (ΔT _pulse = 1 / f _pulse ) between two successive pulses of the particle beam (S) applies: ΔT _pulse = 1 / f _pulse ≈ T _{V, 0} , where T _{V, 0 is} the most likely residence time of the distribution (X _n (T _V )) of the neutron residence times (T _V ).

Images