GA Field Maps

This pages presumes that you are familiar with the overview of how Mu2e deals with magnetic fields.

There are a few details that are pending verification from Mau. These parts are marked with the comment "Fixme".

Table of Contents

• Introduction
• Nomenclature
• Summary Table
• Details of Maps
• GA00
• GA01
• GA02
• GA03
• GA04
• GA05

Introduction

The table below gives a summary of the magnetic field maps that we have received from GA; the table is precede by a description of some of the terms used in the table and is followed by a more complete description of each map.

On the mu2egpvm* machines, on detsim and on the GP Fermigrid worker nodes the files that make up each map can be found in the subdirectories of:

   /grid/fermiapp/mu2e/DataFiles/BFieldMaps/

   /grid/fermiapp/mu2e/DataFiles/BFieldMaps/GA00

   /cvmfs/mu2e.opensciencegrid.org/DataFiles/BFieldMaps/

Nomenclature

This section defines some nomenclature that will be used below.

GA Name

In the summary table, the column labeled "GA Name" contains the name of the field map as delivered from GA. We felt that a numbering system that had 0.1 and 1.0, sometimes written as 01 and 10, might be too confusing. So we switched to a simple increasing number system. The GA numbers come from the Mu2e docdb document.

Usable

In the summary table the column labeled "Usable" says whether or not this map is usable for physics studies.

Map GA00 is usable for physics studies but it is not interesting to do so because, in the fiducial region, it is identical to Mau9.

The maps GA01, GA02 and GA03 use non-standard positions for the TSd coils; the fringe field of the TSd is significant inside the fiducial volume of the DS so these maps are not interesting to use for physics. In addition we believe that these maps do not contain fringe field from either the TSu or the PS. This is probably not important inside the DS but it is important near the upstream end of the TSd.

Format

The summary table has a column labeled "Format"; this can take one of two values, "Half" or "Full". If the column is labeled "Half" then the grid points are only specified for y>=0 and the field for y<=0 is given by the symmetry:

  B_x(x,-y,z) =  B_x(x,y,z)
  B_y(x,-y,z) = -B_y(x,y,z)
  B_z(x,-y,z) =  B_z(x,y,z)

If the column is labeled "Full" then the grid points cover the interval: ymin <= y <= ymax, where ymax is the same as for the upper half maps and where ymin=-ymax. Maps with this format are sometimes referred to as "full azimuth maps", "complete azimuth maps" or as simply "full maps".

Planar Coils vs Helical Winding

In the Mau9 field maps, and in all earlier field maps, the solenoids were modeled as "ideal solenoids". This means that they were modeled as a collection of planar current loops with a pitch of zero (here pitch is used in the sense of pitch, roll and yaw, with the forward direction being the nominal symmetry axis of the solenoids ). It is because all coils have a pitch of zero that these maps have the up-down symmetry discussed in the previous section. The current loops were modeled as ideal line currents; ie there is no transverse extent to the current density and no notion of boundary conditions at the surface of the conductor.

Fixme: check the above para with Mau.

One of the innovations introduced by GA is to model each coil as helical winding. This is a more realistic model of the magnetic field but it breaks the up-down symmetry discussed in the previous section. Any map modeled with helical windings needs to be available in the Full format.

Some of the maps delivered by GA use the planar coil model and some use the helical winding model. The maps with a planar coil model were delivered for comparison purposes to help us test whether or not we understand what they are doing. The maps with a helical winding model are the maps that are intended to represent their actual desgin.

GA DS Coil Positions

The Mau9 DS design delivered to GA was not a 100% complete mechanical design. When their engineers completed the design, mechanical constraints required that they make small changes to some of the coil positions. There are two versions of the GA DS coil positions:

• v1: used in GA03, GA04
• v2: used in GA05

Some of the early maps delivered by GA use the orginal Mau9 DS coil positions while the other maps use their positions calculated by their engineers. The maps with the Mau9 coil positions were delivered for comparison purposes to help us test whether or not we understand what they are doing.

GA is responsible for the construction of the PS and at some later date will also give us field maps in which they have updated PS coil positions. For the time being (September 2015) this only affects the DS coil positions. GA is not responsible for construction of the TS and will always use TS coil positions that we give them.

Non-standard Positions for the TSd Coils

Early in our interactions with GA we gave them a set of coil positions for the DS and TSd. The coil positions for the DS were the same as those for Mau9 but the coil positions for the TSd were not the same as for Mau9; nor were they the same as for Mau10, which is still under development as of mid-september 2015 and has not yet been released. The coil positions for the TSd captured some point in the development between Mau9 and Mau10 and do not correspond to any official Mu2e field.

These non-standard TSd coil positions were used by GA to compute the maps GA01, GA02 and GA03.

Summary table

Fixme: check that GA05 uses Mau9 versions of PS and TSu coils.

Details

GA00

This field map is GA's calculation of the Mau9 field starting from coil positions and currents; in ths model all coils are modeled as planar current loops. The reason for asking for this map is to verify that both Mau and GA get the same answer when starting from the same input.

Why might they get different answers with the same input?

Both Mau and GA use the program OPERA to convert coil positions and currents to field values on a grid. In this caculation OPERA needs to do numerical integrals and any numerical integral has accuracy vs speed tradeoffs. By comparing the GA00 maps to Mau9 we can check if Mau9 and GA are making compatible choices for these tradeoffs.

We refer to the comparison of GA00 to Mau9 as the "null test".

Fixme: Reference to document discussing the null test.

GA01

This map was the original GA attempt at producing a map to use for the null test; all DS coils were modeled as planar current loops. However it has several issues that make it unsuitable for physics: it was produced using non-standard placement of TSd coils; it does not include fringe field from either the PS or the TSu.

The output format is the upper half map.

Fixme: are the DS coil positions are the Mau9 positions or the GA updated positions?

Fixme: add links to docdb documents that discuss this map.

GA02

This map was produced using a coil configuration identical to that used for GA01. The only difference is the output format, which is the full map, not the upper half map.

As for GA01, this map is not suitable for physics use. The one use that it does have is to compare with GA01 to verify that the modifications to our code that read the full format maps works correctly.

Fixme: are the DS coil positions are the Mau9 positions or the GA updated positions?

Fixme: add links to docdb documents that discuss this map.

GA03

This was the first attempt at GA to provide a map that has their best model of the DS, using the GA DS coil positions and the helical winding model for the DS coils.

This map is not suitable for physics because it was produced using non-standard placement of TSd coils.

Fixme: if GA02 used the same coil positions then a comparison of GA02 and GA03 shows the difference between the planar coil and helical winding models. If not, then there is no use for this map.

Fixme: references to docdb documents that discuss this map.

GA04

This is the first map that was released for doing physics calculations. It uses the Mau9 coil positions for the PS, TSu and TSd. It uses the GA DS coil positions and the helical winding model for the DS coils.

It was later discovered that, in the DS, at values of z near the calorimeter, the field has a positive gradient at radii as low as 300 mm; see pages 15 and 16 of Mu2e-doc-6025, "Gradient Comparison on a Zoomed Scale".

Fixme: Reference to comparisons with other maps.

GA05

This map can be used for physics calculations. The map starts from the model of GA04; the placement of some coils are large z was modified in order to push the region of negative gradient to larger values of radius.

Security, Privacy, Legal