1. TQ Program and Review Charge
BNL - FNAL - LBNL - SLAC
TQ Program and Review Charge
LARP Technology Quadrupole Review
LBNL, November 29, 2006
Gian Luca Sabbi

2. LARP Magnet Program Goals
Investigate viability of Nb3Sn technology for the LHC luminosity upgrade
1. Capability to deliver predictable and reproducible performance:
TQ (Technology Quads, ) D = 90 mm, L = 1 m, Gnom > 200 T/m
2. Capability to scale-up the magnet length:
LQ (Long Quadrupoles, ) D = 90 mm, L = 4 m, Gnom > 200 T/m
3. Capability to reach high gradients in large apertures:
HQ (High Gradient Quads, ) D = 90 mm, L = 1 m, Gnom > 250 T/m

3. TQ Goals, Implementation & Parameters
Objective: develop the technology base, in preparation for LQ & HQ:
evaluate conductor and cable performance
develop and select coil fabrication procedures
compare mechanical design concepts and support structures
optimized models: achieve 200 T/m after training & thermal cycle
Two series of models, same coil design, different mechanical support:
TQC models: collar & stainless steel shell; low axial pre-load
TQS models: aluminum shell over iron yoke; high axial pre-load
Magnet parameters:
1 m length, 90 mm aperture, T coil peak field
Maximum gradient T/m

4. TQC and TQS Design Concepts
Yoke
Pad
Shell
Key
Axial rod
Filler
TQC
TQS
Stainless steel collars and skin
Control spacers to limit pre-load
End support plates, no pre-load
Aluminum shell over iron yoke
Assembly with bladders and keys
Aluminum rods for axial pre-load

5. TQ Performance References & Range
Magnet
Top [K]
Gss [T/m]
Bss(body) [T]
Iss [kA]
TQS
4.2
222
11.4
12.5
1.9
239
12.3
13.6
TQC
215
11.2
13.0
233
12.1
14.1
Jc = 2 kA/mm2
(12 T, 4.2 K)
MJR strand
Conservative HT
“First models”
Magnet
Top [K]
Gss [T/m]
Bss(body) [T]
Iss [kA]
TQS
4.2
245
12.6
13.9
1.9
264
13.5
15.1
TQC
239
12.4
14.4
255
13.2
15.5
Jc = 3 kA/mm2
(12 T, 4.2 K)
RRP strand
Aggressive HT
“Final models”
Ic data from extracted strands determine common performance reference for TQS/TQC
Issue: relatively wide range of extracted strand Ic (TQ01: K)
Reference magnet performance limits for a given test run are adjusted for measured Tbath
Actual conductor-limited quench levels may be lower due to other degradation effects

6. Coil Design and Fabrication
Design features:
Double-layer, shell-type
One wedge/octant (inner layer)
TQ01: OST-MJR strand, 0.7 mm
TQ02: OST-RRP strand, 0.7 mm
27-strand, mm width
Insulation: S-2 glass sleeve
Winding & curing (FNAL - all coils)
Reaction & potting (LBNL - all coils)

7. 2D Coil Stresses (Baseline Designs)
TQC Layer 1 stress - sq
TQS Layer 1 stress - sq
Main differences: warm pre-load, cool-down effect, stress uniformity (pole to mid-plane)
Peak stresses are high & no consensus on degradation limits  cable testing needed
Peak stress ~20 MPa difference: stress-relief slot, different Gss & pole stress range at Gss
Detailed FEA shows that 3D effects have a significant impact on actual coil stresses

8. TQC01 and TQS01 Quench Training
SSL 4.5K
SSL 3.2K
TQS01
TQC: limited to 70% of short sample at 4.5K, but achieves 85% at 1.9K
TQS01: start training at 80% of 4.5 K short sample, limited to 87% in one coil
TQS01b: start training at 75% of 4.5 K short sample, limited to 82% in one coil
Maximum quench gradient was close to 200 T/m in TQC01 and TQS01

9. TQ Plan & Schedule – Updated 11/28/2006
Q Q Q Q Q Q Q Q4-07
TQ01
TQ01R
TQ02
TQ03

10. 2/06 Review Findings (1)
The TQ model program is already well advanced, the review appears a bit late in the process. It would have been beneficial to the program to carry out a first mechanical review during the design phase to better coordinate goals and plans for instrumentation and mechanical models.
The Review Committee would like to underline the strong pioneering effort which is being done in refining the FE mechanical models in particular in 3D. The convergence between 2D and 3D models is very good.
In both cases, for TQC and for TQS, the mechanical models were not sufficiently exploited. In particular the TQS model was reduced in scope due to schedule constraints, and the TQC model should have been assembled in different variants (this is however under way) and iterated with the FE model.
The TQC models, FE mechanical and experimental, do not yet fully match, probably also due to the more complexity of the TQC model when compared to the TQS model.

11. 2/06 Review Findings (2)
The Review committee also remarked that important differences between coil pre-load conditions exist between the two models.
In particular in central region of the magnet the target azimuthal pre-stress conditions at short sample seem the same, but this is not the case for the magnet extremities, neither is the case for the axial pre-load.
Furthermore, due to the absence of the pole cut in the inner layer coil, the first version of the TQS model coils will be submitted to a “29” MPa higher peak-stress at cold than the TQC coils : it is not clear how this difference will evolve during magnet excitation and an experimental verification is certainly recommended.

12. 2/06 Review Findings (3)
Another important difference between the two models, which however shall be considered as part of the interest of such a program, is the different way the two magnets achieve the target coil pre-stress conditions at cold.
The TQS builds-up the required coil pre-stress by relatively large movements between ambient temperature and cold, the TQC on the contrary builds its final condition already at assembly and the transition warm-cold is a relatively uniform shrink at constant pressure.
In spite of an evident effort done by the two laboratories to converge into a common table of properties of materials and components, still differences are present in the input data for material properties considered in the FE models.

13. 2/06 Recommended Actions (1)
The Review Committee would like to remind and stress that the TQC/S designs are not the final step in the process. They are proof of principles and therefore should be used to learn such that the final LQ will succeed.
A summary of the recommended actions is presented below :
Establish a bi-monthly assembly and test meeting between the two laboratories.
Agree on a common form in presenting and comparing the models and the results.
The two mechanical designs are different in many aspects: no matter the initial results, we encourage pursuing the program of each design exploring different assembly variants.

14. 2/06 Recommended Actions (2)
Maximize learning from each model (do not dismount before having completed all measurements).
Establish a test plan allowing extracting information even if one coil does not perform as expected.
Make sure the traveller is sufficient and duly filled.
Make a TQC mechanical model with 0.05 mm thicker shims collars/yoke.
Reconsider new iterations of the mechanical models, with better instrumentation and refining of the FE models.
Make next TQSs with the same inner layer coils as TQCs (with pole cut), no matter the results/performance obtained on the first TQS without pole cut.

15. 11/06 Review Charge (1)
Evaluation of the TQ model design, modeling, fabrication and test results to date:
Is the quality of the fabricated coils adequate for achieving the magnet performance objectives? Are there aspects of the coil design, tooling, or fabrication procedures where improvements are needed?
Are the TQC/TQS mechanical designs adequate to achieve the magnet performance objectives?
Is the finite element modeling carried out to a sufficient level to guide the magnet assembly and predict/understand the test results?

16. 11/06 Review Charge (2)
Are the magnets sufficiently instrumented to allow (a) understanding the performance; and (b) checking and further refining the design and the finite element modeling predictions?
Are the test plans and their implementations adequate to characterize the magnet performance? Are the tests providing the needed data to guide the magnet design optimization and structure selection?
Did the analysis of the fabrication and test results (including post-test disassembly and inspection) provide satisfactory explanations for the observed magnet behavior? Are there additional tests and/or analysis that should be carried out to better characterize and understand the magnet performance?

17. 11/06 Review Charge (3) Plans for future TQ models:
Evaluate the proposed test variants and design changes in the context of the fabrication and test results to date, and the findings/recommendations from the previous TQ review. Does the committee recommend any additional test variants or design changes that were not included in the proposed plans?
Does the committee recommend additional tasks that should be pursued in support of the TQ model magnets development? (conductor/material testing, design studies, sub-scale models etc.)
The TQS and TQC models are intended to generate and select the best design features in support of the development of long quadrupoles (LQ) and high gradient quadrupoles (HQ). Do the TQ results obtained so far provide opportunities for converging on particular aspects of the mechanical designs in the near term?

18. Review Agenda (1)

19. Review Agenda (2)

20. Review Agenda (3)

21. Summary
First TQ prototypes fabricated and tested, achieved 200 T/m gradient
Quench levels are limited below short sample in both magnets
Established performance reference, developed optimization strategy
First revised TQS model tested – results are similar to TQS01
TQC02 and TQS01 coils are being fabricated
Updated R&D plan takes into account TQ01 magnet feedback
TQ will provide a basis for the LQ coil and structure design
TQ will also provide input for HQ design and fabrication