For me the madness of first collisions started on Monday morning because of this plot by CMS.
I have no idea how this CMS person was able to post this on the USLHC blog.
We would NEVER be able to do that in Atlas.
It is a plot of the position of the interaction points of protons in the LHC beams colliding with particles of gas in the beampipe. The LHC beam pipe is under a good but imperfect vacuum, so every now and then one of the protons collides a stray molecule. We refer to these events as beam gas events. They are normally a nuisance, however on Monday the LHC wanted to make sure that the positioning of the two beams was good enough so that they would actually hit each other when they attempt collisions. One method of doing this is to look at the position of these beam gas events and make sure they are coming from the same place for both beams.
CMS said they could measure the position of each beam with beam gas events to pretty good accuracy and presented the plots above as evidence. A few weeks ago the ATLAS experts took a look at doing this and determined that it was hard and would require a fair amount of dedicated work to be possible. However, Monday morning ATLAS management said that we needed to give the plot to the LHC as soon as possible. Most of the experts were away at conferences or otherwise unavailable so I was asked to help Juerg, an LBL staff scientist in charge of the beam spot group, find these beam gas events. At noon we holed up in an office and gave it a shot.
What makes finding beam gas events difficult is the fact that you have to separate them from beam halo events.
Beam Halo is caused by the beam interacting with gas or other material upstream of ATLAS. It produces a lot of junk in the detector and is not correlated with the position of the beam. The challenge is to separate these events from the beam gas events. To make a long story short we worked all night on this. Juerg went back to where he was staying and slept for a couple of hours, but I, having youth on my side, pulled the all nigher I never had in college and eventually found a way to separate these events in a run we were testing. It was a small moment of triumph where I felt that I had made a significant contribution to the experiment. Then we tried the method on the current data and found that it didn't really work because the events were much less contaminated by beam halo than the test sample and so the discriminator that I had found didn't discriminate anymore. Still, in the end we were able to give the LHC a beam position that was not as precise as CMS's but precise enough to say that "we could not rule out that the LHC beams would collide".
We came to this conclusion at 6am, just in time to prepare for the collisions! The LHC then tried twice to inject the two beams and bring them to full energy and failed. Since it takes several hours between attempts, I took the opportunity to sleep for an hour after the second failure.
When I came back from my restorative power nap and shower, I headed back up to the satellite control room where I had spent the nighttime hours I should've been sleeping. The satellite control room is a smaller, less impressive space two floors above the main control room where people look at the data after it comes off the detector.
The LHC takes almost an hour to go from injection energy to collision energy so it gave us ample time to bite our nails and stare at the monitors which showed it's progress. Once the monitors read "collapsing separation bumps", a sign that the were bringing the beams into collision position, the excitement and tension in the room was palpable.
When the collisions finally came we watched the monitors around us and saw event after gorgeous event flash by. It was amazing and thrilling and reminded me why I like physics so much. I was staring at collisions of protons that were turning their mass and energy into different particles and those particles were being observed by our detectors! I was looking at the data that would go into my thesis! The words that kept coming to my lips were gorgeous and beautiful and aesthetic appeal to the events was almost overwhelming. Not long after we saw the first events the leader of the scanning team started yelling out event numbers and the room turned to chaos. Luckily for me, the data I needed to run my processing scripts had a half an hour delay between when the events were taken and when they were made available to me. I took that opportunity to go down to the main control room and watch the festivities. It was packed and ebullient and the champagne was flowing. I got there too late to partake in the bubbly drinking, but several people had the foresight to bring extra bottles for the satellite control room so we got ours a bit later.
I spent the rest of the day marveling as the events kept on coming and continuing to update my plots for the collaboration. In a few hours we took as much data as we had in all of last year's lower energy runs. Eventually, around 6pm the LHC lost the beam and we left. I attended a meeting, ate dinner with my group, drank some celebratory wine and slept for a glorious 11 hrs.
An example of a beam gas event from CMS.
CMS said they could measure the position of each beam with beam gas events to pretty good accuracy and presented the plots above as evidence. A few weeks ago the ATLAS experts took a look at doing this and determined that it was hard and would require a fair amount of dedicated work to be possible. However, Monday morning ATLAS management said that we needed to give the plot to the LHC as soon as possible. Most of the experts were away at conferences or otherwise unavailable so I was asked to help Juerg, an LBL staff scientist in charge of the beam spot group, find these beam gas events. At noon we holed up in an office and gave it a shot.
What makes finding beam gas events difficult is the fact that you have to separate them from beam halo events.
Beam Halo crap in ATLAS.
We came to this conclusion at 6am, just in time to prepare for the collisions! The LHC then tried twice to inject the two beams and bring them to full energy and failed. Since it takes several hours between attempts, I took the opportunity to sleep for an hour after the second failure.
When I came back from my restorative power nap and shower, I headed back up to the satellite control room where I had spent the nighttime hours I should've been sleeping. The satellite control room is a smaller, less impressive space two floors above the main control room where people look at the data after it comes off the detector.
The satellite control room on Tuesday Morning.
I was there as an extension to the event scanning team, a group of people whose job it was to find the prettiest events and make them public as soon as possible. All the event pictures I've shown in the previous posts were made in this room. I was not working on the event displays but on a webpage which took the data and made general plots available to the collaboration soon after the run started so people could get a feel for the events.
The LHC takes almost an hour to go from injection energy to collision energy so it gave us ample time to bite our nails and stare at the monitors which showed it's progress. Once the monitors read "collapsing separation bumps", a sign that the were bringing the beams into collision position, the excitement and tension in the room was palpable.
When the collisions finally came we watched the monitors around us and saw event after gorgeous event flash by. It was amazing and thrilling and reminded me why I like physics so much. I was staring at collisions of protons that were turning their mass and energy into different particles and those particles were being observed by our detectors! I was looking at the data that would go into my thesis! The words that kept coming to my lips were gorgeous and beautiful and aesthetic appeal to the events was almost overwhelming. Not long after we saw the first events the leader of the scanning team started yelling out event numbers and the room turned to chaos. Luckily for me, the data I needed to run my processing scripts had a half an hour delay between when the events were taken and when they were made available to me. I took that opportunity to go down to the main control room and watch the festivities. It was packed and ebullient and the champagne was flowing. I got there too late to partake in the bubbly drinking, but several people had the foresight to bring extra bottles for the satellite control room so we got ours a bit later.
I spent the rest of the day marveling as the events kept on coming and continuing to update my plots for the collaboration. In a few hours we took as much data as we had in all of last year's lower energy runs. Eventually, around 6pm the LHC lost the beam and we left. I attended a meeting, ate dinner with my group, drank some celebratory wine and slept for a glorious 11 hrs.
No comments:
Post a Comment