Progress in the Development of the Nb 3 Sn MQXFB Quadrupole for the HiLumi Upgrade of the LHC

TLDR: In this paper, the status of the development activities at CERN is described, in particular on the cold powering test of the first MQFXB prototype and on the construction of the second full scale prototype.

Abstract: The high-luminosity upgrade of the Large Hadron Collider (HL-LHC) requires new high field and large-aperture quadrupole magnets for the low-beta inner triplets (MQXF). With a nominal operating gradient of 132.2 T/m in a 150 mm aperture and a conductor peak field of 11.3 T, the new quadrupole magnets are based on Nb3Sn superconducting technology. After a series of short models constructed in close collaboration by LARP (LHC Accelerator Research Program) and CERN, the development program is entering in the series production phase with CERN on one side and the US Accelerator Upgrade Project (US-AUP) on the other side assembling and testing full-length magnets. This paper describes the status of the development activities at CERN, in particular on the cold powering test of the first MQFXB prototype and on the construction of the second full scale prototype. Critical operations such as reaction heat treatment, coil impregnation and magnet assembly are discussed. Finally, the plan towards the series production is described.

Figures

Fig. 1. Cross-section of the Nb3Sn low-β quadrupole magnet MQXF

TABLE I. INTEGRAL TRANSFER FUNCTION (T/KA)

Fig. 13. Integral transfer function versus current computed with ROXIE and measured in MQXFBP1.

Fig. 4. Average (and std) gap between segments before winding, after relaxing the winding tension and after reaction for RRP coils.

Fig. 2. Average critical current density degradation due to cabling. Fig. 3. MQXFB coil production dashboard.

Fig. 5. Estimated coil RRR based on samples reacted with the coil. The vertical line indicates the change of duration of the last plateau during reaction.

Frequently asked questions

Q1. What have the authors contributed in "Progress in the development of the nb3sn mqxfb quadrupole for the hilumi upgrade of the lhc" ?

After a series of short models constructed in close collaboration by LARP ( LHC Accelerator Research Program ) and CERN, the development program is entering in the series production phase with CERN on one side and the US Accelerator Upgrade Project ( US-AUP ) on the other side assembling and testing full-length magnets. This paper describes the status of the development activities at CERN, in particular on the cold powering test of the first MQFXB prototype and on the construction of the second full scale prototype. Critical operations such as reaction heat treatment, coil impregnation and magnet assembly are discussed. 

Q2. How many layers are used to wrap the coils?

8. One 125 μm polyimide layer is wrapped around the coil, and two around the collars, with a total thickness of 375 μm which constitute the ground plane insulation. 

Q3. What is the maximum voltage drop between the pole and the insulation coil?

The pole is floating, and the maximum voltage drop during quench in the pole turns is 10 V meaning that 10 MΩ resistance between coil and pole is considered sufficient to assure safe operation. 

Q4. What is the longitudinal prestress at warm?

The longitudinal prestress at warm in MQXFBP1 was equal 0.6 MN, aiming at a total force of 1.2 MN after cooldown which corresponds to the longitudinal electromagnetic forces at nominal current. 

Q5. How much pressure is in the SS shell?

The target pre-stress in the SS shell is 75 MPa to assure that shell is still in contact with the magnet aluminum shell after cool-down. 

Q6. How much is the b6 in the MQXFBP2?

Thanks to the iteration on the cross section, increasing the pole insulation thickness and re-ducing the mid-plane shim thickness by 0.125 mm, b6 is close to the target in MQXFBP2. 

Q7. What was the first coil produced in 2016?

The production of MQXFB coils started in 2016 with two copper coils followed by two low performance Nb3Sn RRP conductor (CR001, CR002, CR101 and CR102) [14]. 

Q8. What is the expected impact of replacing the 13.8 mm key?

The expected impact of replacing the 13.8 mm key by the conical key is an increase of the stress in the shell of 18 MPa in the magnet extremities whereas the stress in the center remains constant. 

Q9. What is the process of transferring the coil in the mold?

Before transferring the reacted coil in the impregnation mold, a flexible circuit which includes quench protection heaters is installed in the coil outer radius. 

Q10. How many kA is the current in the LHC?

The first prototype MQXFBP1 was limited in performance reaching a current of 15.1 kA, equivalent to 6.5 TeV operation in the LHC. 

Q11. Why is the SS shell used in the shrinkage cylinders?

Due to the impossibility to access the reverse side of the longitudinal weld of the shrinkage cylinders, a permanent backing strip has been chosen, in order to provide support for a fully penetrated root pass [31]. 

Q12. How many conformed coils have been produced?

From September 2019, ten conformedcoils have been produced, one is on-hold and one was rejected due to a critical nonconformity during winding. 

Q13. How many coils are being produced per month?

Conductor procurement is close to completion and coil fabrication resumed after and in-depth review of the manufacturing process with a production rate of one coil per month. 

Q14. What is the corresponding value of the quench velocity at 15 kA?

Quench propagation velocity at 15 kA is 15 m/s, similar to the typical quench propagation at this current level measured in short models and cable samples ([34],[35]), pointing to a local defect and not a global degradation of the coil. 

Q15. What is the stress in the shell?

Figure 11 shows that the measured stress in the shell is 100 to 175 MPa, corresponding to an increase of azimuthal stress in the coil of 15 to 20 MPa which was considered adequate for MQXFB magnets. 

Q16. What is the preferred solution for the coils?

The azimuthal solution is in general preferred, as it allows to bring the four coils inner surface at the same radius, with beneficial effects on the field homogeneity [19]. 

Q17. How many tons of MQXF coils are expected to be delivered by summer 2021?

The remaining quantity to fulfill the production, 1.7WDOI: 10.1109/TASC.2021.3061352 FERMILAB-PUB-21-126-TD (accepted)This manuscript has been authored by Fermi Research Alliance, LLC under Contract No. DE-AC02-07CH11359 with the U.S. Department of Energy, Office of Science, Office of High Energy Physics.tons of RRP 108/127, are expected to be delivered by summer 2021 [9]. 

Q18. What is the difference in RRR between the connection and non-connection side?

Thanks to an improvement on the leak-tightness of the reaction fixture, the difference in RRR among the connection and non-connection side decreased significantly in the last coils.