Choose timezone
Your profile timezone:
Update #1
→
Europe/Zurich
Description
Including Davide and Christian
Attendees: Emma Gautheron, Ariel Haziot, Marco Masci, Roland Piccin,
Christian Scheuerlein, Davide Tommasini, Gianluca Vernassa
General Discussion
- What is the issue?
- In magnets, we observe a larger loss of pre-stress when cooling down than what we are able to simulate.
- After a powering cycle coils are smaller than their shape in virgin state
- In some magnets degradation occurs although the computed stresses are below the max stress limit, and some magnets are working at computed stresses much above stress limits.
- Defining our objectives
- Model properly what is happening in the conductor during the life of a magnet (assembly, cool down, powering)
- Understand how the coil stack behaves thermo-mechanically at the scale of the conductor block (which we may call from now on the "macroscale").
- Understand the role of the various components (Copper, Nb3Sn, Epoxy and Glass fiber) impact on the "macro"-behavior (to do so, we step into the "meso-scale")
- How much does this impact the SC properties
- Understand how macro-stress states can impact the SC properties (i.e., from macro to micro-scale)
- Comprehensive understanding of the coil pack
Models
- MQXF approach: Linear elastic material behavior for the whole coil bloc. Defined by a single parameter {E: Young's modulus.}
- 11 T V1. Multilinear elastoplastic with isotropic hardening. (in the 11T case defined by three parameters {E, σy : yielding stress, ET: tangent modulus of curve σ-εtot) Ok but did not account for peak stress due to lack of interference in the model
- 11T V2 (Bertarelli's approach), introduced intermittences but also changed the way to describe the material. Division in the FE model of conductors and insulation, both modeled with Multilinear elastoplastic with isotropic hardening. In the 11T V2 case many parameters were used for both materials, each one with its own E, and then a series of {σyi, ETi}.
- 12T today uses the same model as 11T V2. In order to validate this type of model we need:
- Wolf's type of measurements in 3 directions
- CTE
- Issues while going from 3D to 2D
Considerations
- All the previous models are able to capture (more or less) a 1D behavior (σ-εtot) in the transversal direction.
- Little (or no) attention has ever been paid to model the behavior in the other directions.
Conclusions
- This work-group, initiated for 10 stacks measurements is actually a more general brainstorming on mechanical design models and appropriate material characterization attached to it.
- We need a new name for this topic : "Nb3Sn coil packs: mechanical characterization and modelling" ???
- What we need i short term is a Wolf type measurement on a 12T stack
- Next meeting, we invite M. Guinchard to present him our status on this topic. Ask him:
- E modulus difference depending on measuring techniques
- Possibility to use extensometer as a measuring techniques
Actions list
- launch meeting with M. Guinchard to have his opinion
There are minutes attached to this event.
Show them.
