Developing #185
Upgrade of the decayer: target of an arbitrary shape.
| Status: | Закрыта | Start date: | 05/12/2018 | |
|---|---|---|---|---|
| Priority: | Высокий | Due date: | ||
| Assignee: | Mikhail Kozlov | % Done: | 100% | |
| Category: | Software | |||
| Target version: | - |
Description
The shape of the target is not always planar. The target can be rotated around an axis (e.g vertical one) in order to face a telescope thus reducing the amount of material crossed by a reaction product. It is necessary to upgrade the algorithm which defines the interaction point inside the target.
In the picture below
The target is black, its boundingbox is grey, the beam axis is lilac, the diagonal of the boundingbox is green, the path of an ion is orange. The path inside the target is thick orange, it has the length l. X- is the interaction point. X0 and x0+l - the points where the trajectory of an ion crosses the target boundary.
The interaction probability should depend on l. Convoluted with the beam profile and angular distributiion it will give the correct distribution of the interaction poins in space. If the nuclear interaction length is Lambda. The probability of the interaction can be normalized to its maxilum possible value i.e. (1-exp(-A/Lambda))
It is necessary to study how Vitaly uses the ROOT geometry navigation in reconstruction and generalize the definition of the interaction point for the complex (but convex) shape of the target. The case of thin target when the distribution of the interaction points along l is uniform, can be implemented first.
Comments and questions are welcome!
Subtasks
History
#1 Updated by Sergey Belogurov over 6 years ago
- Description updated (diff)
#2 Updated by Sergey Belogurov over 6 years ago
- File deleted (
target.png)
#3 Updated by Sergey Belogurov over 6 years ago
- Description updated (diff)
#4 Updated by Mikhail Kozlov over 6 years ago
I've three questions:
1) Is it true that the nuclear interaction length is https://en.wikipedia.org/wiki/Nuclear_interaction_length ?
2) Is Lambda the parameter that defined by user through interface or we recieve it from the GEANT methods?
3) What is the probability of interaction: parameter that we've not used yet? How to normilize it with respect to (1-exp(-A/Lambda)?
#5 Updated by Sergey Belogurov over 6 years ago
- Description updated (diff)
1) Roughly speaking –yes 2) Lambda is the parameter defined by user through interface 3) In reality many ions pass throug the target without nuclear interactions. For planar targets perpendicular to the ion beam, the probability of the nuclear interaction is qual for all the ions. For this reason we normally neglected the ions passed throiug and assumed that each incident nucleus interacts, we only define the position of the interaction, However in the case of complex shape of the target the interaction probability depends on the thickness of the target material crossed by the ion. The maximum possible interaction probability is (1-exp(-A/Lambda)). To normalize the interaction probability for every particular trajectory meanse calculate the following: (1-exp(-l/Lambda)) /(1-exp(-A/Lambda)). We neglect the fact that many ions go through without nuclear interaction. We still let some (less number than in reality) ions to go throug, but we reproduce correct distribution in space of the interacion points.
Is it getting more clear?
#6 Updated by Vitaliy Schetinin over 6 years ago
Статус по этой задаче?
#7 Updated by Vitaliy Schetinin over 6 years ago
Ветку влил - задачу закрыл
#8 Updated by Mikhail Kozlov over 6 years ago
- Status changed from Открыта to Закрыта