Download presentation
Presentation is loading. Please wait.
Published bySuzan Barker Modified over 8 years ago
1
MICE Magnetic Modelling CM 36 IIT 17 th – 19 th June 2013 P. J. Smith
2
MICE Updates to the Hall Model 17/06/2013P J Smith - University of Sheffield2 There have been no major structural changes to the hall model since the last CM. However a lot of time and effort has been spent on refining the improving the model. A number of issues with the model have been investigated and resolved. Mesh refinements to tighten the meshing in areas where there was a higher error in the mesh and extension of the model boundary. The removal of small geometric errors in the model that were causing meshing complications. An investigation into the inexplicable hanging of the model solver being traced to the use of BH curve with a couple of points that gave a non-monotonic relative magnetic permeability. (This problem took a lot of effort to find.) The model typically takes 1-2 days (down from 3-4 days not so long ago!) to solve if all goes well, so feedback from changes to the model have sometimes been slow and incremental. I am at the stage where I’m not making any further changes to the Hall model – awaiting a pending report from Vector Fields.
3
MICE Model Checks with Vector Fields 17/06/2013P J Smith - University of Sheffield3 To give us some faith in using the results from the Hall model we have instigated an external review of the undertaken by Vector Fields. The purpose of the review is to examine the hall model itself and to view it in context of doing sanity of checks on the model, checks on the structures/meshing and general technical validation. They will provide us with a report on the model’s validity and the likely level of error in the model. We visited VF last week to see how this was progressing. In brief they feel that the model is in good shape, they see no outstanding technical issues with the hall model. A full report from Vector Fields will be published shortly. Whilst this does not replace the need for some measurement exercise, having an external body who is experienced with FEA review the model is a useful stepping stone towards ensuring we have a model that we can trust. It is clear that checking the hall model with real measurements is going to difficult. Some successful cross correlation of a model with measurements has been completed successfully – I refer to Melissa’s R9 –see later slides -but each model stands on its own merit,so at some level we have to trust the hall model.
4
MICE 17/06/2013P J Smith - University of Sheffield4 (Random) Image illustrates extent of model boundaries - Step IV 240Mev/c. 5 gauss scale. You need to check which images are relevant to you! Many, thousands of images across different planes can be found at http://www.hep.shef.ac.uk/research/mice/opera_models/autogen_plots.html
5
MICE MICE/ISIS Meeting 17/06/2013P J Smith - University of Sheffield5 We held the first meeting with ISIS on 28/03/2013. At this meeting we were able to present some output from the Hall model to give ISIS some feel for what field level the model was then predicting for Step IV in areas of interest to ISIS. My impression was that this was well received. There are many slides (too many to present today) but this presentation can be viewed at: http://www.hep.shef.ac.uk/research/mice/opera_models/presentations/MICE_ISIS_2 013_03_28/ David Findlay gave MICE a list of items that ISIS were particularly worried about. These items have been added to the ‘Fry List’; this is a list of items that could potentially be affected by the MICE magnets. ISIS have appointed a representative to report back to them on the progress of developments on the MICE magnetic field mitigation.
6
MICE Sub Models 17/06/2013P J Smith - University of Sheffield6 We have made some progress with sub-models. I have so far looked at two sub models but these both need more work. Quad Sub Model- This was a model of just the quads Q7-Q9 and the local structures, intended to look at the field in the region around Q9. This was based on the hall model code but with most of the geometry from the hall model switched off. From examining the mesh in this sub-model around the quads made us realise that the mesh needed Improving in the hall model, which then lead to a number of problems that took some time to resolve. This model is ready to be re-run with the improved meshing. Sub Station – This is a relatively new sub model that aims to model the panels in the substation more accurately to get some feel for what field the instrumentation that is set in the panels might see. I have a basic model of the substation built and it solves, I now need to tighten the model up to get it to the point where it’s producing results that are useful.
7
MICE Sub Models 17/06/2013P J Smith - University of Sheffield7 Quad Sub model to look at the region around Q9. This could be done with a full hall model but we wanted to test using components of hall model for separate model and to test how much effect removing iron from model had on results. Meshing resolution can also be increased if necessary with smaller model. Sub-station. First pass of model to show direction of applied external field and geometry of sub-station. Quite simple model but very high resolution.
8
MICE Tackling the Fry List 17/06/2013P J Smith - University of Sheffield8 The ‘Fry List’ is a list of items in the MICE Hall and surrounding areas that may be susceptible to increased magnetic fields. Very comprehensive list with many items. The hall model is now sufficiently developed that I think we are in a position to use it as a first pass at estimating the likely field level in many areas in and around the MICE Hall. With this in mind we have now started to try to take the items in the Fry list and put a likely air-field value on those items. We are only part of the way through this list as it has become clear that we didn’t have enough positional information on many of these items to match them up well enough with the Hall model, a second iteration of this process is now underway. It is clear that we need to mark each object both by a field level and a confidence level based upon field homogeneity, proximity to other iron that may or may not be in the model. Areas looked at so far (test region) was North Wall above and below the North Mezzanine.
9
MICE 17/06/2013P J Smith - University of Sheffield9 Map components from TD to Hall Model: - Ascertain Field / Gradient & what’s missing in the area of each component - Repeat Ad-nauseum
10
MICE 10 Overview of R9 Model Slides 10-14 provided by Melissa A George.
11
MICE 11 -x z y R9 Coordinate System
12
MICE 12 R9 Fields XY plane through Bore Bz in Tesla
13
MICE 13 Comparison of Data, Calculation and Opera Model R9 Fields Field at bore height along z, x = 0, y =0
14
MICE 14 Comparing Data to Opera Model of R9 80 cm above bore, along z, x = 0 Effect of metal wall
15
MICE Plan & Scheduling 17/06/2013P J Smith - University of Sheffield15 Trying to put field values on the items on the Fry list is a high priority as it is only from this that we can get some idea as to what further mitigation work will be necessary. This could include: Mechanical shielding, Moving of objects, Further modelling work The results from this task is a key schedule driver as it is hard to define a schedule without knowing what further mitigation needs to be undertaken. Now that we know that we have a good hall model (from a modelling perspective) we now need to now what is missing from this model to ascertain the likely effect. Continuing effort on sub-models. The scheduling is being driven by two dates ; an external review in August followed by a decision as to whether to pursue the localised shielding or partial return yoke solution in September.
16
MICE Conclusions 17/06/2013P J Smith - University of Sheffield16 Work on the hall model as it stands is pretty much complete. Unless I have good reason I don’t intend to add anything more to this model. Now that we have confirmation that it is a good model I do think we need to understand its limitations a little more. If you have equipment in the MICE hall ensure you check out the on-line images to get a feel for what field your equipment may see. I do need to rerun these images with a slightly more accurate algorithm over the next week or so if you have already looked then keep an eye out for the new images (They’ll be called NODAL_INTEGRAL) Work is now focusing on: Fry List. Is there anything on the fry list that we are really worried about? Sub Modelling. Starting to model areas/equipment with known or suspected problems. I think we to have a clearer picture of where the problems areas are likely to be by the review but my suspicion is that the work will not be complete by then.
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.