Home » NUSTAR » NUSTAR PRESPEC » AGATA Crystal positions (Look-Up Table)
AGATA Crystal positions (Look-Up Table) [message #18008] |
Mon, 16 March 2015 11:46 |
thuyuk
Messages: 68 Registered: July 2014
|
continuous participant |
From: *ific.uv.es
|
|
Hi everybody,
I have some problems with the correction for the Doppler effect in my experiment. The target position was selected forward, and this makes the system very sensible for very small changes in the determination of the particle positions or the gamma positions.
I am comparing whether the crystal positions is calculated properly, because we know that there is a small rotation in the phi axis between the Legnaro and the GSI configurations of the AGATA detectors. The rotation from the original positions is 7.8 degrees in the Legnaro case, and 10 degrees in the GSI cases. I have two look-up tables in my hand, one is from the GammaWare software and the other is from the prespec analysis package. I see that the coordinates in the look-up table of the Gammware software are in agreement with the coordinates inside the femul software. But, the coordinates that the prespec software package uses are different.
I don't know if femul uses the original coordinates, or the coordinates that are already rotated 7.8 degrees for the Legnaro case. But, a rough calculation shows that the rotation is approximately 10 degrees between the two different look-up tables I possess.
Therefore, I was wondering if the rotation introduced in the coordinates of the prespec package is considered from the original positions of the detectors, or the Legnaro configuration. If, the person who performed the rotation considering that the coordinates are already for the original positions, but in fact they are for the Legnaro case, then the coordinates inside the prespec package should be corrected.
Maybe this issue addresses Michael. So, I would be grateful if you could inform me, Michael, regarding to the transformation you have done.
Thank you very much!
Tayfun
[Updated on: Mon, 16 March 2015 15:31] Report message to a moderator
|
|
|
|
Re: AGATA Crystal positions (Look-Up Table) [message #18016 is a reply to message #18010] |
Wed, 18 March 2015 14:22 |
Damian Ralet
Messages: 35 Registered: July 2014 Location: Darmstadt
|
continuous participant |
From: *gsi.de
|
|
Dear All,
The CrystalLookupTable allow to pass from the local referential (detector level) to the global referential (laboratory frame). The transformation need the crystal Id to associate the crystal position in the frame.
The table was generated in the Geant4 simulation code for the LNL campaign. I was not aware that they were two tables, but I hope we got the proper one.
The issue with the table between the LNL and the GSI campaign comes from the crystal labeling in the frame. The convention is I think different. The reason might be the position of the AGATA frame with respect to the beam axis, meaning forward at GSI, and on the side at LNL.
I think there is a rotation of 180Degree to performed between the LNL lookup table, and the GSI one. In addition, the axis need to be swap and opposite:
//x,y,z->-y,-x,z
I hope this help,
Cheers,
Damian
|
|
|
|
|
Re: AGATA Crystal positions (Look-Up Table) [message #18019 is a reply to message #18008] |
Thu, 19 March 2015 14:46 |
thuyuk
Messages: 68 Registered: July 2014
|
continuous participant |
From: 84.79.220*
|
|
Dear all,
If we compare the coordinates of the first several crystals in both cases:
Lookup table inside the GW:
0 0 48.08390 31.04885 -227.95092
1 0.45352 0.86888 0.19842
2 0.86889 -0.48060 0.11858
3 0.19839 0.11862 -0.97292
1 0 40.86042 88.71803 -213.76361
1 0.64447 -0.74500 0.17214
2 -0.73906 -0.54919 0.39011
3 -0.19610 -0.37863 -0.90454
2 0 96.65970 58.36943 -206.11951
1 -0.90585 -0.01692 0.42326
2 0.09899 0.96308 0.25036
3 -0.41187 0.26869 -0.87073
3 0 -14.67047 55.32513 -227.95092
1 -0.68622 0.72558 -0.05146
2 0.69982 0.67784 0.22535
3 0.19839 0.11862 -0.97292
4 0 -71.74929 66.27595 -213.76361
1 0.90204 0.29209 -0.31782
2 0.38454 -0.87825 0.28426
3 -0.19610 -0.37863 -0.90454
5 0 -25.64313 109.96598 -206.11951
1 -0.37407 -0.92117 -0.10731
2 -0.83092 0.28151 0.47991
3 -0.41187 0.26869 -0.87073
6 0 -57.15075 3.14396 -227.95092
1 -0.87762 -0.42045 -0.23022
2 -0.43637 0.89953 0.02069
3 0.19839 0.11862 -0.97292
Lookup table insdie the prespec an. pack.:
0 0 41.34692 39.57936 227.95091
1 0.95033 -0.26615 0.16133
2 0.23981 0.95660 0.16553
3 -0.19839 -0.11862 0.97292
1 0 95.77207 19.18782 213.76361
1 -0.56958 -0.70689 0.41938
2 0.79820 -0.59745 0.07704
3 0.19610 0.37863 0.90454
2 0 79.18105 80.50124 206.11951
1 -0.11363 0.93294 0.34164
2 -0.90413 -0.23965 0.35372
3 0.41187 -0.26869 0.87073
3 0 -24.86530 51.55395 227.95091
1 0.06560 -0.99203 -0.10758
2 0.97793 0.04248 0.20459
3 -0.19839 -0.11862 0.97292
4 0 11.34650 97.01402 213.76361
1 -0.93514 0.34976 0.05632
2 -0.29505 -0.85692 0.42266
3 0.19610 0.37863 0.90454
5 0 -52.09293 100.18191 206.11951
1 0.82476 0.51622 -0.23084
2 -0.38746 0.81322 0.43422
3 0.41187 -0.26869 0.87073
6 0 -56.71452 -7.71727 227.95091
1 -0.90979 -0.34695 -0.22782
2 0.36458 -0.93035 -0.03909
3 -0.19839 -0.11862 0.97292
First of all, it is not clear for me what would the "small" coordinates mean, in other words, the coordinates with the id "0" make sense, but the ids "1, 2 and 3" are not obvious for me.
Well, if we compare the coordinates with id 0, then we immediately see that the z coordinates are swapped from -z to +z, from the GW to the prespec code, respectively. But, the transformation on x and y seem totally confusing. Damian, did you find out by yourself the transformation, or you were told by somebody else?
Does anybody know if the crystal numbering was changed from Legnaro to GSI? This information is required to focus on comparing crystal by crystal without worrying if also the crystals were scrambled.
edit:
sorry for the bad view of the coordinates, the tabs are not shown properly in the html, I guess.
I uploaded two file, so you can compare them with your favorite viewer.
So, the one with suffix __gw is from the GW, and the __orig is the one inside the prespec code.
[Updated on: Thu, 19 March 2015 14:51] Report message to a moderator
|
|
|
|
Re: AGATA Crystal positions (Look-Up Table) [message #18021 is a reply to message #18020] |
Thu, 19 March 2015 15:33 |
wieland@mi.infn.it
Messages: 13 Registered: July 2014
|
occasional visitor |
From: *mi.infn.it
|
|
Dear All,
so if i understood correctly,
the coordinates of the crystals are taken from the LNL table and then
rotated to match with the prespec setup.
Why we do not take the precise measurements done from the mechanics group in 2012 determining the
crystal front face positions.
In my point of view, these measurements should be better than adapt old tables,
or is this already done ?
We need at least a precision of mm to do good measurements at high v/c.
oliver
In case it is needed
Please see the latest word file from the 2012 campain i have with the coordinates in the beamline coordinate system.
Who knows how to convert them into the AGATA coordinate system
[Updated on: Thu, 19 March 2015 21:56] Report message to a moderator
|
|
|
Re: AGATA Crystal positions (Look-Up Table) [message #18023 is a reply to message #18021] |
Fri, 20 March 2015 08:58 |
Damian Ralet
Messages: 35 Registered: July 2014 Location: Darmstadt
|
continuous participant |
From: *gsi.de
|
|
Hi Oliver,
The coordinate measured by the mechanic group might be more precise, but I am not sure if we can use these only.
The CrystalLookupTable allow to orientate the crystal and to position it in the global referential. The position of the crystal center, as given in the table might miss the rotation of the crystal along its z axis.
I think, the table from the mechanic is really useful to make sure that the positions determined by the CrystalLookupTable are good. We just need orientate properly the two coordinate system, and check the position center of each crystal. If this agreed in the 2012 campaign, they should be good as well for 2014.
Cheers,
Damian
|
|
|
|
Re: AGATA Crystal positions (Look-Up Table) [message #18026 is a reply to message #18008] |
Fri, 20 March 2015 11:22 |
thuyuk
Messages: 68 Registered: July 2014
|
continuous participant |
From: *eduroam.uv.es
|
|
Hi all,
Michael, thank you very much for posting the related part of the code where the coordinate transformation is defined. But, these transformations are applied to the coordinates inside the Lookup table, if I'm not wrong. Then, one has to be sure here if the transformation is applied twice. Because, if the Lookup table coordinates are the ones that already transformed, and we make another transformation as we process them, then there will be a problem. Do we know who did provide these coordinates?
Oliver, thanks a lot for posting the document where the measurement results could be found. This might be a very good reference to understand if the crystal positions in the global frame are as same as with the ones after the proper transformation and/or/of the positions inside the Lookup table.
I think there is a good question asked by Oliver: How could we be sure whether we are using the right coordinates?
If we draw the hit points in x,y,z, from which perspective we should see them?
Since there is a 2.2 degrees of rotation difference in the phi axis due to the traslation of the array from Legnaro to GSI, how could we "see" this small angle difference? Let me remind you that the experiments in which the target position is selected forward and with high beta, small changes in positions and angles affect the analysis very much.
Cheers,
Tayfun
|
|
|
Re: AGATA Crystal positions (Look-Up Table) [message #18027 is a reply to message #18026] |
Fri, 20 March 2015 11:43 |
wieland@mi.infn.it
Messages: 13 Registered: July 2014
|
occasional visitor |
From: *mi.infn.it
|
|
Dear all,
i did a simple test
(tell me if i am wrong)
i took the look up prespec table x,y,z and calculated the r distance
and as you see they look very much "theoretical" as they are all the same !
Then i took the table from Strachen Laser measuremnts at GSI,
and the cristal front face have different r, as ecpected from "real life".
(see list at end).
Frontface means if i understood correctly, front of cristal housing.
Written in the word document.
A second test i did now is to take serious the measured target position offset that wrote Strachen,...
and the 2plus state shows up better.
If this is true, that the measured positions are different from the tabled one, i think we have to reevaluate all lookuptable and we have a problem.
maybee a serious problem, if we continue using the old tables.
Who is able to take care of this problem ?
radius distanze
prespec lookup
235.0270515
235.0270487
235.0270439
235.0270429
235.0270428
235.0270419
235.0270416
235.0220086
235.0220077
235.0220057
235.0220053
235.0220049
235.0220048
235.0220035
235.0220026
235.0220014
235.0220008
235.0220005
235.0220005
235.0219985
235.0219984
Strachen 2012 measures
235.8678514
235.4930124
235.1198427
235.0128345
234.7970693
234.5157763
234.1244172
233.8492784
233.707791
233.5889812
233.366668
233.0576531
232.6759456
232.4498048
232.3274615
232.2338872
231.6427568
231.361688
231.2412362
231.1137296
231.0875362
230.9375023
230.4040869
230.1122638
230.086363
230.0175034
229.7865397
229.4736394
229.3584166
229.3414726
229.2368009
229.2167074
229.0517248
228.7686377
228.7648408
228.7026844
228.4926037
224.1008758
223.224044
222.7145377
[Updated on: Fri, 20 March 2015 12:53] Report message to a moderator
|
|
|
Re: AGATA Crystal positions (Look-Up Table) [message #18028 is a reply to message #18027] |
Fri, 20 March 2015 13:02 |
Damian Ralet
Messages: 35 Registered: July 2014 Location: Darmstadt
|
continuous participant |
From: *gsi.de
|
|
Hi Oliver,
The test is good, but in my opinion highlights that the two coordinate systems do not use the same center, which I noticed already with the Uranium Xrays.
I think I will try to give a look a bit more carefully at the relatives positions between the theory, and measured... but I see how much time I have in the next days, and try to come with some answer.
Cheers,
Damian
|
|
|
Re: AGATA Crystal positions (Look-Up Table) [message #18029 is a reply to message #18028] |
Fri, 20 March 2015 14:05 |
wieland@mi.infn.it
Messages: 13 Registered: July 2014
|
occasional visitor |
From: *mi.infn.it
|
|
Dear Damian,
the point is not that they have a differnt distance,
(this can be corrected easily, the measured target position of Strachen and the table distance give an offset of approx 1.1 cm in beam direction and 3.6mm in x, which works good for our 2plus)
BUT the point is that they have ALL exactely the same distance !
This looks very much like a CAD or GEANT table, and not like a measurement.
The measured distances are different between each others of up to some cm.
This is only the distance ! So also (surely) the positions may vary, which is worse !
This means, in my opinion, that the table is NOT respecting the real setup !
This can have very big effects.
For this reason i think it is very important to have the correct table.
oliver
[Updated on: Fri, 20 March 2015 14:15] Report message to a moderator
|
|
|
Re: AGATA Crystal positions (Look-Up Table) [message #18030 is a reply to message #18029] |
Fri, 20 March 2015 14:56 |
Damian Ralet
Messages: 35 Registered: July 2014 Location: Darmstadt
|
continuous participant |
From: *gsi.de
|
|
Dear Oliver,
Sorry if i am not clear, but I completely agree with you that we need to find the a way to match reality (and not simulation).
The lookup table is a CAD design, not a measurement. I also agree that the positions are different if radial distances are.
Cheers,
Damian
|
|
|
|
|
Re: AGATA Crystal positions (Look-Up Table) [message #18035 is a reply to message #18008] |
Sat, 21 March 2015 11:25 |
Damian Ralet
Messages: 35 Registered: July 2014 Location: Darmstadt
|
continuous participant |
From: *dynamic.qsc.de
|
|
Dear All,
Just a little update concerning the positions of the detectors. If you have any questions, please let me know.
Convention
I use the same convention as given in the mechanical report. THe beam is along the Y axis. The X is horizontal. The referential is direct.
Radial distance
I started to check the same thing as Oliver, i.e. the radial distances of the detector. Previously Oliver mention that the distance were different for the crystal. If we look at the systematic as a function of the crystal position in the X-Z plane (perpendicular to Y axis, the beam). The histogram is in attachement, the color scal highlight the radial disance:
If I now shift the target center with the distance given in the report, we get a uniform radial distance, the difference in the color scale is in the order of 10^-4 mm. THe bin weights are taken here as the ratio of the corrected radial distance over the mean of the radial distance (after target repositioning).
Angles
If we know consider the spherical angle theta and phi. we can see a phi rotation of about 7 degree. The following figure highlights this rotation in the X-Z plane. The background histogram is the detector positions given in the crystalLookupTable, and the '+' dots are the measured positions. The 'x' are the measured positions after a rotation of 7 degree.
Conclusion
There is a rotation, and we need to adjust the frame positione properly in order to adjust our crystallookuptable to the measurement.
TODO
The angle of 7 degree is a visual estimation, I need to get sth more precise, even if a rotation of phi does not influence the Doppler correction.
The theta angle seem a bit off for the triple, while it look quite good for the double. There might be an other adjustement to perform.
Cheers,
Damian
|
|
|
|
|
|
|
|
|
Re: AGATA Crystal positions (Look-Up Table) [message #18053 is a reply to message #18008] |
Tue, 24 March 2015 11:17 |
thuyuk
Messages: 68 Registered: July 2014
|
continuous participant |
From: *ific.uv.es
|
|
Dear all,
After looking at the drawings that Damian did, and compare them with what I get with the current CAD-lookup table, I see that the whole array suppose to be swapped in the horizontal axis.
These are the positions in the measurement report that Oliver sent:
These are the positions of the crystals reported by -probably- Plamen before:
Sorry for the poor quality of the second image, I didn't have it in soft copy and had to take a photo of the hard copy. What I get with the CAD-lookup table is pretty much look like this. I intentionally didn't put what I get, but put this image, since the cryostat numbers are given there.
My point is that the cryostat number 1 in the second image does not correspond to the cryostat number 5 in the measurement report, but it seems to me that it corresponds to number 2.
My idea is supported by looking at the number 4 and number 12 in the second image that they have a touching edge, but number 3 and 8 don't have a touching edge in the image that is inside the measurement report. Is it true what I'm saying?
I think we should be careful with this too, since the array has no symmetry according to the vertical axis. The detected angles would be different than the real angles, and this is something that we cannot survive with.
cheers,
tayfun
P.S.
I just found another drawing that Damian sent to me in our private communication. The positioning of the AGATA detectors there also suggest that we should swap the horizontal axis. I didn't get the permission from Damian, but I think he will be OK that I put the image below:
|
|
|
Re: AGATA Crystal positions (Look-Up Table) [message #18055 is a reply to message #18008] |
Tue, 24 March 2015 12:51 |
thuyuk
Messages: 68 Registered: July 2014
|
continuous participant |
From: *ific.uv.es
|
|
Hi, again,
After what Michael wrote regarding to the unknown position of the Target DSSSD in the measurements report, I remember that the hit points suggest a small shift in the -x axis in my experiment:
The centroid seems to have an offset of ~-3 mm in the x axis. The y axis seems fine.
Maybe the detector positions were not certain and there were "small" offsets in reality with respect to the theoretical positions. Is there any way that we can confirm the exact positions of the detectors?
I concern about another thing: I don't understand what these numbers suppose to mean inside the lookup table:
0 0 41.34692 39.57936 227.95091
1 0.95033 -0.26615 0.16133
2 0.23981 0.95660 0.16553
3 -0.19839 -0.11862 0.97292
I don't understand what would mean the numbers that the absolute values are smaller than 1.
I also want to make another point clear regarding to what Michael wrote:
Thanks a lot for including the rotation part, but, since we have the exact measures in the lab frame, does it really necessary to make transformations inside the code? Would it be better to include only the offsets with respect to the "theoretical positions"?
If I'm following the whole conversation properly, one needs to create a new lookup table with the measured values, and take the x,y,z values given inside the report as -x,z,y; could someone confirm this?
Another point that I think I didn't get: Oliver wrote that the AGATA target position has an offset of 1.1 mm in the beam downstream and 3.6 mm in the horizontal axis. I didn't see this offset inside the report. Only thing that I can see is this:
Agata array target 3.6 -2.9 -2.7
So, the 3.6 mm offset is clear to see, but why 1 mm offset in the beam axis?
In addition, we have been taking the offset for the target forward position as 150 mm, the measurements suggest that this is 142.7 mm, is this correct?
cheers,
tayfun
|
|
|
Re: AGATA Crystal positions (Look-Up Table) [message #18056 is a reply to message #18053] |
Tue, 24 March 2015 12:52 |
wieland@mi.infn.it
Messages: 13 Registered: July 2014
|
occasional visitor |
From: *mi.infn.it
|
|
Dear Tayfun,
this is indeed all very confuseing,
but the cad image i think is seen from lycca whereas the plots are seen from frs side i guess.
The 1.1cm shift was the mean difference in distance between measurement and table.
But as i think now, it is enough to rotate the agata array and apply the measured x,y,z shifts.
Does anybody know the differences between 2012 and 2014 setup ?
Target positions have been already estimated, maybe Michael or Damian did some work about target position ?
oliver
[Updated on: Tue, 24 March 2015 13:00] Report message to a moderator
|
|
|
|
|
|
Re: AGATA Crystal positions (Look-Up Table) [message #18067 is a reply to message #18008] |
Wed, 25 March 2015 16:25 |
thuyuk
Messages: 68 Registered: July 2014
|
continuous participant |
From: *ific.uv.es
|
|
Dear all,
I made a table to convert different naming conventions each other for myself, and thought that it might be useful also for somebody else.
So, as far as I understand, the crystal mapping between three different conventions is as the following:
cr# AGATA GSI
-----------------
1 -> 3B -> 00C
2 -> 3G -> 00B
4 -> 2B -> 01C
5 -> 2G -> 01B
13 -> 4B -> 04C
14 -> 4G -> 04B
18 -> 9R -> 06A
19 -> 9G -> 06B
22 -> 8G -> 07B
23 -> 8B -> 07C
36 -> 13R -> 12A
37 -> 13G -> 12B
38 -> 13B -> 12C
39 -> 12R -> 13A
40 -> 12G -> 13B
41 -> 12B -> 13C
42 -> 11R -> 14A
43 -> 11G -> 14B
44 -> 11B -> 14C
Please let me know if you see anything wrong with it.
cheers,
tayfun
edit: this is for the 2012 campaign. I don't know if it is modified for the 2014 campaign, but for sure more crystals were added. So, if you let me know them, I could modify the table to include the two added crystals.
[Updated on: Wed, 25 March 2015 16:27] Report message to a moderator
|
|
|
|
|
|
|
|
|
|
|
|
Goto Forum:
Current Time: Tue Nov 05 23:09:56 CET 2024
Total time taken to generate the page: 0.00919 seconds
|