First Operational Experience with a Cryogenic Permanent Magnet Undulator at the Esrf

A cryogenically cooled in-vacuum undulator was installed in the ID6 test beamline of the ESRF in January 2008. This 2 metre long hybrid undulator has a period of 18 mm. The magnetic assembly is based on NdFeB permanent magnets cooled at a temperature close to 150 K. A liquid nitrogen closed loop is used for the cooling of the undulator. This cooling system is well adapted to achieve a uniform temperature along the magnetic assembly. A large part of the study was focused on the heat budget of the undulator under beam in the different filling modes delivered at the ESRF. The impact of the undulator on the ultra high vacuum of the ring was investigated with several warming/cooling cycles. This paper presents the main outcomes from this first experience. INTRODUCTION A Cryogenic Permanent Magnet Undulator (CPMU) is an in-vacuum undulator which makes use of the enhanced magnetic properties of NdFeB material at low temperature [1]. It therefore allows the extension of high brilliance undulator radiation to higher photon energy than is usually achieved with conventional in-vacuum undulators. The maximum performance can be reached using NdFeB grades having the highest remanence (Br > 1.4 T) at room temperature but also the lowest coercivity (Hcj <1000 kA/m). Such coercivity is insufficient to operate an in-vacuum undulator safely at room temperature due to the possible demagnetization of permanent magnets under electron beam exposure with small gaps. This weak point is eliminated when the magnet material is cooled at cryogenic temperature. Typically, the coercivity is increased by a factor of 2.5 when the NdFeB material is cooled from 293 K to 150 K. The temperature of 150 K corresponds to an optimum in the dependence of the undulator field on the temperature [2]. It is related to a specific magnetic property of the NdFeB material called Spin Reorientation Transition (SRT). Below 150 K the undulator peak field decreases. This unusual magnetic behaviour has required some studies on NdFeB samples at low temperature [2],[4]. An accurate magnetic model of NdFeB material at low temperature requires a non linear representation. A linear model can still be used but requires knowledge of the material magnetic permeability versus temperature [5] in order to correctly predict the field performances of a CPMU . At the ESRF, all in-vacuum undulators are baked at 120 deg C after their installation in the ring in order to limit the Bremstrahlung sent to the beamline .This baking of the undulators is not compatible with the use of low coercivity NdFeB material. Consequently, the use of high remanence NdFeB material for the CPMU requires the suppression of the baking phase; this aspect is discussed further in this paper. In order to retain the baking option, the first CPMU was based on high coercivity NdFeB ( Hcj=2400 kA/m) and therefore with reduced field performances (Br=1.16 T). The undulator is a hybrid structure with a period of 18 mm and a total magnetic length of 2 m. The primary target of this prototype was to develop the required magnetic measuring systems and cryogenic cooling method as well as to obtain some experience in the operation of such a device in a storage ring. The two first topics have already been discussed [2],[3]. This paper focuses mainly on the operation of this first prototype installed in the ID6 straight section in January 2008 (Fig. 1). The undulator is connected to a liquid nitrogen cryo-pump system located in the technical gallery with a cryogenic line totalling 50 m in length. Figure 1: First CPMU prototype installed in the ESRF ID6 straight section. HEAT BUDGET WITH BEAM Average Temperature The undulator was cooled to a temperature of 138 K before the accelerator re-start in January 2008. Due to the high thermal constant of the undulator (12.5 hours) it was not possible to use machine dedicated time for the thermal study. Instead, the device was monitored during use by the ID6 beamline. The temperature measurements of the undulator rely on 20 thermocouples distributed along the magnetic assembly. The power extracted from the cryogenic loop is derived from the measurement of the differential pressure and temperature of liquid nitrogen between the input and output of the cryogenic pump. WE5RFP067 Proceedings of PAC09, Vancouver, BC, Canada 2414 Light Sources and FELs T15 Undulators and Wigglers Figure 2 shows the average temperature of the undulator (green) and power extracted from the cryogenic loop (blue) in 7/8 filling mode with 200mA (Fig. 2 a) or in 16 bunch mode with 90 mA (Fig. 2 b) versus time. It reveals a strong dependence of the power on the undulator gap. The maximum extracted power and temperature are reached for a fully opened gap. This effect is observed in all beam filling patterns. The origin of this effect is still not clearly established but it is suspected that it originates from Higher Orbit Modes (HOM) of the radio frequency resonating in the undulator tank. Table 1 presents the measured average temperature of the undulator and power extracted from the cryogenic loop for different filling mode and gap settings. The last column of Table 1 corresponds to the net power due to the electron beam. The highest power of 106 W is observed in the hybrid filling pattern (24x8+1 bunch, 200 mA).