Behavior of the magnets during low b squeeze

1. Behavior of the magnets during low b squeeze W. Venturini Delsolaro Acknolowgements M. Giovannozzi, S.Sanfilippo 28 February 2007 LHCCWG meeting

2. Position of the problem • From a magnet standpoint, the low b squeeze is a sequence of current ramps, eventually with changes of sign and stopsfor beam measurements and corrections • There are two possible implications, i.e. • Hysteresis crossing • Decay and snapback • Both change the actual field produced for a given current (transfer function)

3. Magnets • All the DS and MS quadrupoles: MQM, MQY but also MQTL and MQT • Working at 4.4 K (from Q4 to Q6) and at 1.9 K (from Q7 to Q13) • All currents in the -5390,5390 A range • Magnetization effects ~ Jc(T,B)

4. Assumptions and approach • Squeeze from injection optics scaledat 7 TeV (plus change from injection to collision tunes). In some older measurements squeeze followed a cycle to nominal current • Squeezing cycles from MADX (β* = 11 m in IP5/IP1) • Linear TF used to generate currents • Measure fields on (some) squeezing cycles • Evaluate deviations due to hysteresis • Measure decay on last steps (where stops are more likely )

5. Example, MQM in Q7L5, applying squeeze cycle after cycle to nominal current

6. Same MQM in Q5L5

7. Still another example, in Q9R8 2.3 units

8. A recent measurement: Q6R5B2 squeeze cyclewith stops on the last 3 steps (to measure decay)and a more realistic pre cycle (starts from injection optics scaled at 7 TeV)Measurements from injection current up to nominal

9. Small loop No decay, or if any, below measurement noise (pretty high in this particular case)

10. MQTL test in SM18 on 12/12/2006, Q11R8B2 squeeze cycle

11. MQTL, Q11R8, B2

13. More examples of hysteresis crossing at low current… 4%

16. MQT (all cycles for Q12, Q13 in IP5 measured)

17. in terms of field strength…

18. Max setting errors, from hysteresis loops(without modeling hysteresis crossing) • MQM~ 30 units at 320 A, 10 units at 1000 A, 5 units at 2000 A • MQY~ 25 units at 200 A, ~10 units at 300 A • MQTL ~ 90 units at 17 A, ~ 25 units at 34 A, etc.. • MQT ( same as MQTL) …diverging at zero

19. Decay of MQY, MQM for the reference cycle MQY MQM 16 apertures 6 apertures from S. Sanfilippo, FQWG meeting on 30.1.2007

20. Do MQT and MQTL Decay ? Not really…

21. Concluding remarks (1) • Δk = f(k) can be extracted from hysteresis loops • These errors due to hysteresis add to global uncertainty on gradients, with the present FIDEL model • FIDEL modeling of hysteresis crossing should bring errors in the range of few units, but this gets harder at low currents • Very low settings for MQTL and MQT, difficult to manage: transfer functions diverge, it is difficult to get the desired field • No decay in MQT and MQTL • On MQM and MQY full decay characterization needs more data (measurement foreseen in 2007) Data are available at the median injection current, which is only indicative.

22. (2) Squeeze on the ramp, from a magnet standpoint Would possibly reduce the number of hysteresis crossings Ramp rates could be chosen such that there would be no need to “stop and wait for the arc”, decay would probably be reduced as measurements on the last steps (lowest β*) would take place with the insertion quads already sitting on their final values Benefit in terms of optics errors remains to be evaluated Would simplify magnetic model if all ramps are kept monotonous