The NEXT 100 geometry - next-exp/nexus GitHub Wiki

General overview

The geometry of NEXT-100 is complex and this page has the aim of explaining the assumptions and the simplifications that have been made w.r.t. the technical drawings, to be able to run simulations efficiently.

The 0 of the z axis (axis of the TPC) in the global reference frame is placed at the beginning of the drift volume and is usually referred to as gate. There are three fundamental z dimensions that are used in more than one class in the code to place the different volumes correctly relatively to each other, i.e.:

• The distance between the gate and the surface of the tracking plane (TP).
• The distance between the gate and the surface of the sapphire windows.
• The length of the step that is cut out in the bars of the inner copper shielding, in the energy plane (EP) side.

Additionally, there is a fourth fundamental dimension, which does not appear in the code, but is used to adjust volume positions in several places, i.e., the distance between the surface of the EP copper plate and the teflon reflectors (24.8 mm).

In the current version of the code, the meshes are simulated as dielectric of arbitrary thickness; this thickness is also propagated throughout the code to place all the inner volumes correctly.

The field cage volumes are displaced from the TPC axis. In order to keep the active volume in the center of the global reference frame, the volume that contains all the other volumes (the lead castle or Hall A, depending on the type of simulation) is moved by this displacement.

In this picture, a global view of the main elements of the geometry is shown.

Vessel

The vessel is a union of several volumes, as explained in this document. The flanges are simulated simply as external volumes; however, the EP flange has a part that goes inside the vessel, which holds the EP plate. To simulate this we do the following:

1. We build a solid cylinder made of stainless steel.
2. We build a smaller cylinder made of xenon.
3. We cut out two thin cylinders from the xenon volume, in the place where the inner part of the EP flange goes.
4. We place the xenon volume inside the stainless steel one.

This way, the inner part of the EP flange emerges as the part of the inner volume of the vessel which is not occupied by xenon.

The PORT_Nx regions are points on the inner bottom of the corresponding ports, if no source is instantiated. If the Th source (or any other in the future) is placed in the detector, the regions correspond to the active volume of the source, in the corresponding port, placed as inside as possible.

Energy plane

The following simplifications have been done:

• The flanges of the copper huts behind the PMTs haven't been simulated, therefore their mass must be added to the activity of the huts.
• Only the central hole for the gas flow has been simulated. The other three ones (close to the borders) are not straight, but they bend at roughly the middle of the copper thickness, and we have decided not to simulate them.
• The PMT bases are made of two different pieces: only the internal part is being simulated, which contains pins and resistors. The external part (dirtier, which contains the capacitors) is supposed to be placed on top of the huts, which are also cut diagonally. However, in the simulation the huts are not cut, and the external part of the bases is not simulated. A generator behind the copper huts is used to simulate their rejection factor.
• A small overlap appeared between the shortest PMT huts and their mother volume (xenon); we have reduced the length of said huts by 1 mm (passing from 60 mm to 59 mm). Given the simplifications of the previous bullet, this reduction should not be relevant.
• The screws of the sapphire windows are not being simulated. Their activity must be added either to the windows (worst case) or to the copper (best case).

Field cage

• The volumes which make up the FC (active, buffer, cathode ring, field rings, ring holders, teflon panels, hdpe) are displaced a few mm towards negative x/y, w.r.t. the TPC axis, due to the presence of a double hdpe semicylinder covering only less than half of the circumference, as shown in the following picture.

We implement this displacement in the simulation, moving all these volumes rigidly away from the center of their mother volume (the gas of the vessel). However, due to the simplifications of the geometry (i.e., the ICS is a hollow cylinder, rather than made of bars with slightly different thicknesses), the distances between each stave and the ICS do not reproduce exactly those measured in the lab (the red numbers in the picture). We think that the approximation is the best we can do, given the limitation of the simulation.

We simulate only one hdpe cylinder, while the second semicylinder is not simulated, for simplicity, since both the density and the thickness are small and should not affect much the path of background gammas.

Warning: the anode and gate rings and the EL gap are not displaced!

• On top of being displaced from the vessel axis, the field cage volumes have their relative positions not placed in a perfect cylinder shape. A consequence of this is that the distance between to opposite teflon panels (therefore the active volume diameter) is not the same for all pairs of panels. We don't know the exact distances of all pairs, therefore we have decided to stick to the dimensions of the technical drawings.

• The frame that holds the anode mesh is simulated with a smaller dimension in z to avoid overlapping with the SiPM teflon masks. In the real detector, the masks on the borders of the tracking plane are cut to fit within the anode frame. Simulating this would add significant complication to the geometry; since the SiPMs which are not covered by the masks are very unlikely to detect any optical photons, we have left the masks uncut and reduced the thickness of the frame to fit them.

Vertex generation

The vertex generators are defined w.r.t. the mother volume of their geometry. For instance, all the vertex generators of the tracking and energy plane and the field cage are defined w.r.t. the gas inside the vessel. Since the gas is displaced to obtain an active volume centred in the TPC axis, one needs to translate the vertices, too, to the global reference frame. This is done in the Next100 class, in the GenerateVertex() function, before returning the vertex.