Design and validation of the LEBT for the project LINAC 7, a low-current low-energy compact LINAC

DESIGN AND VALIDATION OF THE LEBT FOR THE PROJECT LINAC 7, A LOW-CURRENT LOW-ENERGY COMPACT LINAC I. Arredondo∗, J. Feuchtwanger, I. Badillo, J. Jugo, J. Portilla, V. Etxebarria University of the Basque Country (UPV/EHU), Leioa, Spain I. Ariz, R. Enparantza, J. M. Seara,Tekniker, Eibar, Spain Abstract In this paper we present the design and validation of a compact LEBT for the LINAC 7 project. Specifically, the LINAC 7 project focuses on building a new generation, lowenergy, low-current compact accelerator. The core idea is to achieve an energy of 7 MeV in less than 12 m while maintaining enough current to generate isotopes for medical uses. Through this work we explain the procedure we followed for the design, including the tests that we carried out to reach the final result. This includes the iterations we needed to overcome various problems such as how to keep the LEBT compact and deal with cooling, when is the best time for packing factor calculation, how to solve cooling problems,... Although this LEBT is intended to be used with protons, further simulations have been carried out to show that it could be used for other species as well. 1 INTRODUCTION The IL7 project aims to build a compact 7 MeV low intensity accelerator for medical purposes [1]. Due to its application, it is necessary to achieve a device of reduced dimensions and cost, with low maintenance and easily transportable from one place to another. This means that tuning must be carried out frequently and it must be as simple as possible. In this article the design of the LEBT for this linear accelerator is presented and taking into account the above mentioned characteristics, a two-solenoid system is chosen. The reasons are that it is relatively compact and cost-effective, allows easy inclusion of diagnostics and is suitable for a wide range of beams [2]. Figure 1: IL7 injector scheme. 2 DESIGN An iterative process will be followed in which changes are made according to the results and possibilities. The following is a summary of the steps to be followed: Base Design some design minima are set on which to iterate and some simple calculations are made to determine feasibility: physical limits, approximate geometry, magnetic field confinement structure,.... ∗inigo.ar[email protected] Mechanical Design According to the mechanical limits of the previous step the solenoids are designed. Height and width, wire type (material, section, insulation, hollow/solid) and packing factor are chosen. The casing is also studied to confine the magnetic field so that the material is not saturated (assuming an approximate magnetic field). It is intended to be manageable, easy to assemble and disassemble, compatible with an alignment system and of adequate cost. Magnetic Design (Focusing) The beam trajectory is simulated for different currents in the solenoids of the previous step so that the necessary currents are set to make the beam converge. The beam loss and whether the generated magnetic field does not saturate the confining casing material must also be taken into account. In addition, the winding configuration is chosen so that the current sources to be used are within manageable current and voltage parameters. If this is not the case, the mechanical design should be revisited. Thermal Design Based on the currents and the mechanical design, the operating temperature of each solenoid is calculated and the appropriate cooling is designed if necessary. In this way, the conductivity changes in the materials, both in the winding and in the casing, are simulated. With this, the voltage required to achieve the current provided by the field is recalculated and the current and voltage parameters are rechecked to ensure that they are manageable. Validation Depending on how many iterations have been performed the validation process will be different. For example in a first iteration, a low cost prototype is made to analyze if it is in specification. This will be as similar as possible but taking advantage of the available material. In addition, the packing factor that can be obtained in the winding is checked. In successive iterations the prototype will converge to the final design and more exhaustive tests will be carried out. If the validation concludes that it is not possible to reach the specification, the first step is started again. It is worth bearing in mind that the cooling design is mainly conditioned by the possibilities offered by the mechanical design. 2.1 Base Design 2.1.1 Mechanical Our mechanical design is limited by the cross section necessary to include vacuum and system diagnostics (Fig. 1). In our case and for the sake of standardization it is convenient to use a DN100 tube to install the vacuum system. On the other hand, the pipe through which 14th International Particle Accelerator Conference,Venice, Italy JACoW Publishing ISBN: 978-3-95450-231-8 ISSN: 2673-5490 doi: 10.18429/JACoW-IPAC2023-WEPL082 MC5.D01: Beam Optics Lattices, Correction Schemes, Transport 3295 WEPL: Wednesday Poster Session: WEPL WEPL082 Content from this work may be used under the terms of the CC BY 4.0 licence (© 2022). Any distribution of this work must maintain attribution to the author(s), title of the work, publisher, and DOI. the particle beam is transported has to be optimized so that it is neither too wide to require too high a magnetic field, nor too narrow to avoid aberrations. For this reason, a DN40 tube was chosen. Note that the minimum distance between the solenoids will be determined by the DN100 tube of the arms perpendicular to the beam direction. 2.1.2 Magnetic Design (Focusing) By means of various simulations carried out with SIMION (Fig. 2), the parameters in the Table 1 are obtained. Figure 2: Focusing simulation. Table 1: Base Design Parameters. 𝑁: Turns Parameter Value Ion Source Voltage 30 kV Einzel lens central electrode Voltage 15 kV Solenoid 1 Total Current 20000 A·N Solenoid 2 Total Current 29500 A·N 2.1.3 Thermal Design In order to determine the cooling requirement, the power used to generate the required field simulated in the previous section must be calculated. To do so, the following steps are followed: • Calculation of the number of turns ( 𝑁 ): Ideally, it depends on the cable type and the space. • Calculation of the cable resistance ( 𝑅 ): It is related to the length and section of the cable and the conductivity of the material. The last one also depends on the temperature of the wire. In the case of copper: 𝜌𝐶𝑢𝑇 =𝜌𝐶𝑢20 1+3.9,9−3(𝑇−20) where 𝜌𝐶𝑢𝑇 is the conductivity of the copper at a given temperature T, 𝜌𝐶𝑢20 is the conductivity of the copper at 20 º𝐶. • Calculation of the current through the wire ( 𝐼 ): In practice there are many factors that affect the effective current flowing through the cable (type of cable, experience of the wirewinder.,...) to take them into account the packing factor ( 𝑝𝑎𝑐𝑘 ∈ [0,1) ) is defined. It can only be determined experimentally. 𝐼= 1 𝑝𝑎𝑐𝑘 ∗𝑁𝐼𝑖𝑑𝑒𝑎𝑙 𝑁 • Calculation of the power (P): Once the resistance and the current have been obtained, the calculation of the power is straight: 𝑃=𝑅𝐼2 With all this, the following results are obtained for the unreal case of T = 20 º𝐶 which will give us a very optimistic value, concretely P = 841 W. That power is too high for an uncooled solenoid and cooling will need to be included in the design. 2.2 First Iteration For this iteration it is worth building a prototype to measure some parameters that only can be estimated experimentally such as the packing factor and, thus, the actual current needed to generate the magnetic field. 2.2.1 Mechanical Based on the base design the first iteration design is shown in Fig. 3 Figure 3: First design. 2.2.2 Magnetic Design (focusing) By means of similar simulations as in the section 2.1.2 we reach to the parameters in Table 2. Table 2: First Iteration Design Parameters 𝑁: Turns Parameter Value Ion Source Voltage 30 kV Einzel lens central electrode Voltage 15 kV Solenoid 1 Total Current 23000 A·N Solenoid 2 Total Current 36000 A·N 2.2.3 Thermal Design Thanks to the prototype built, the need for cooling can be accurately measured. In our case, it is experimentally observed that even in the best case, using standard cooling, the required current cannot be reached. So it is decided to divide the solenoid into separate subcoils (including a non-magnetic good temperature conductor separators), which, in simulation, achieves the desired current (Fig. 4). Figure 4: Thermal simulation with subcoils (COSMOSFloworks). 14th International Particle Accelerator Conference,Venice, Italy JACoW Publishing ISBN: 978-3-95450-231-8 ISSN: 2673-5490 doi: 10.18429/JACoW-IPAC2023-WEPL082 3296 MC5.D01: Beam Optics Lattices, Correction Schemes, Transport WEPL082 WEPL: Wednesday Poster Session: WEPL Content from this work may be used under the terms of the CC BY 4.0 licence (© 2022). Any distribution of this work must maintain attribution to the author(s), title of the work, publisher, and DOI. 2.3 Final Design The final design is shown in Fig. 5. Figure 5: Final design scheme. Experimental tests have been carried out (Figs. 6 and 7) that prove that the dimensions, cooling and magnetic field fits with the expected values. Figure 6: Experimental cooling test: Several thermocouplers installed inside the solenoid give the temperature for different input currents and flow rate. Figure 7: Magnetic field comparison. 2.3.1 Beam Dynamics Simulation Results The beam dynamics are simulated (Fig. 8 and 9 ) to see if the specifications will be met at the next acceleration structure input. It can be stated that the beam goes from diverge to converge, that can be constrained to a circumference with radius less than 2 mm (our goal) and with an acceptable emittance. Figure 8: Density distribution of the beam particles an emittance at the output from the ion source. Figure 9: Density distribution of the beam particles an emittance at the output from the LEBT. Figure 10: Density distribution of the 𝐻𝑒+ beam particles and emittance at the output from the ion source. 2.3.2 Multispecies Simulations Simulations for 𝐻𝑒+ (Fig. 10) and 7𝐿𝑖3+ (Fig. 11) show that the injector could handle these species too. For the experiments the maximum currents are set in order to know if the focus of the beam is less or equal the distance to the next acceleration structure. Figure 11: Density distribution of the 7 Li 3+ beam particles and emittance at the output from the LEBT. 3 CONCLUSION In this paper we have presented the successful process of design and construction of a low current and low energy LEBT valid for a multispecies injector. REFERENCES [1] J. Feuchtwanger et al., “New Generation Compact Linear Accelerator for Low-Current, Low-Energy Multiple Applications”, Appl. Sci., vol. 12, no. 9, p. 4118, Apr. 2022. doi:10.3390/app12094118 [2] K. Hanke et al., “Comparison of Simulation Codes for the Beam Dynamics of Low–Energy Ions”, Rev. Sci. Instrum, vol. 73, no. 2, pp 783–785, Feb. 2002. doi:10.1063/1.1427349 14th International Particle Accelerator Conference,Venice, Italy JACoW Publishing ISBN: 978-3-95450-231-8 ISSN: 2673-5490 doi: 10.18429/JACoW-IPAC2023-WEPL082 MC5.D01: Beam Optics Lattices, Correction Schemes, Transport 3297 WEPL: Wednesday Poster Session: WEPL WEPL082 Content from this work may be used under the terms of the CC BY 4.0 licence (© 2022). Any distribution of this work must maintain attribution to the author(s), title of the work, publisher, and DOI.