A New Continuous 4π-Frozen-Spin-Target for the Crystal Barrel Experiment

Abstract

The CBELSA/TAPS experiment has produced a large amount of data on photo-production reactions in double polarization reactions. For this a frozen-spin target has been used, which consists of a horizontal dilution refrigerator with an internal holding coil for a 4π acceptance. For further experiments with CBELSA/TAPS, a new horizontal dilution refrigerator has been designed as a continuous polarised solid target to increase the FOM. This can be realized by an internal polarization magnet for a continuous polarisation of the target material during the data taking of the experiment. This refrigerator was designed to:<br /> • Support an internal superconducting magnet for continuous polarisation during the measurement<br /> • Improve the performance to reach lower temperatures for the frozen-spin mode (30 mK)<br /> • Optimise the loading procedure to lower the maximum temperature during the change of the target material <br /> First, the different heat sources to the mixing chamber were calculated. To minimise the influence of these heat leaks a vacuum shell with two radiation shields at 40 K and 3 K was designed to minimize the heat input by radiation. Also the holding structure and the connections between the different heat exchangers were built from stainless steel for a low thermal conduction along the structure. For thermal isolation, a critical part of the refrigerator is the beam-tube, due to the necessity of a good isolation vacuum in this part of the refrigerator. The beam-tube is sealed by an indium O-ring and the tightening of this seal is important for the vacuum in the beam-tube. Taking all heat sources into account gives a heat of 40 μW to the mixing chamber.<br /> In order to improve the knowledge about the fluid characteristics of the precooling stages, CFD simulations were performed. Thus, with a FVE-method, it is possible to calculate a vector field for the fluid pressure, velocity, density and the temperature of the different heat exchangers and the different streams flowing through these heat exchangers. These simulations were performed for a ³He→He mixture flowrate of 1 - 20 mmol s^-1. It was possible to predict the necessary ³He flow rate for the heat exchangers, HE1 and HE2, to precool the circulating dilute of the refrigerator. Nevertheless, for HE3 only a coarse model was used, due to the fact that one of the 4 streams in this heat exchanger has to be super fluid and only the thermal conditions of this state were implemented in this calculation. This means that the predicted flow rate of ³He is only a rough assumption for the performance of this heat exchanger. It is the first time CFD-simulations have been used to predict the performance of a dilution refrigerator for a polarised target. Now these simulations can be used to optimise the performance of the condenser unit of other dilution refrigerators. It is possible to draw a specific geometry and use the input parameter of these simulations to test new heat exchanger concepts before building a prototype.<br /> The dilution unit has to cool the target material, under the usage of the dilution effect, to temperatures around 30 mK for the frozen spin mode and to 200 mK in the case of a continuous operating target. Additionally, the geometry of the target front is limited through the geometry of the detector and the external polarisation magnet. Thus, the outer diameter of the front of the refrigerator has to be smaller than 94 mm including the vacuum shell, the radiation shields, the internal magnet and the dilution heat exchanger. Also, these parts surrounds the beam tube which is necessary to guide the particle beam to the target material. For a good performance of the dilution unit a copper sintered heat exchanger has been designed. It is composed of 10 stages to prevent heat conduction along the axis of the heat exchanger. The calculation shows, that, including the heat leaks, it should be possible to reach temperatures around 30 mK in dilution mode. Due to the available computer resources the copper sintered heat exchanger was not simulated. In principle, by programming a specific solver it should be possible to use <em>openFOAM</em> for the simulation of the copper sintered dilution heat exchanger if the necessary computer power is available.<br /> A test facility was built to measure the performance of the refrigerator. The facility consists of a vacuum system, including several probes to measure the pressure and the volume flow in the different cooling streams, a NMR- and microwave-system, special electronics for the determination of the temperature and a mass-spectrometer to determine the ³He-⁴He ratio.<br /> First test measurements with ⁴He, instead of a ³He-⁴He mixture, circulating in the closed cooling circuit were performed and the simulated data for the first two heat exchangers are in good agreement with these measurements. Unfortunately, due to a cold leak between the dilution unit and the isolation vacuum it was not possible to operate the refrigerator in the dilution mode. It was not possible to reach the necessary temperature of 1 K in the evaporator due to this leak. This also increased the temperatures of HE3, which reduces the performance of the refrigerator. In addition, to test if a continuous mode of the refrigerator is possible, the prototype of an internal magnet was used to test the cooling of the magnet. It was possible to reach temperatures below 1.5 K and to generate a magnetic field of 2.2 T. Thus with a 50 GHz microwave-source, it should be possible to use the refrigerator as a continuous mode polarised target, if the leak of the dilution unit is fixed.<br /> The tests to locate the leak are ongoing. Unfortunately, the cold leak is only detectable at temperatures below 80 K. If the leak has been located and fixed, the refrigerator will be first tested in ⁴He mode for further leaks and afterwards the dilution heat exchanger can be mounted. Then, the refrigerator has to be tested in dilution mode and after optimising the operation parameter it can be used for the measurements at CBELSA/TAPS.