• John
  • Felde
  • University of Maryland
  • USA

Latest Posts

  • USA

  • James
  • Doherty
  • Open University
  • United Kingdom

Latest Posts

  • Andrea
  • Signori
  • Nikhef
  • Netherlands

Latest Posts

  • CERN
  • Geneva
  • Switzerland

Latest Posts

  • Aidan
  • Randle-Conde
  • Université Libre de Bruxelles
  • Belgium

Latest Posts

  • Vancouver, BC
  • Canada

Latest Posts

  • Laura
  • Gladstone
  • MIT
  • USA

Latest Posts

  • Steven
  • Goldfarb
  • University of Michigan

Latest Posts

  • Fermilab
  • Batavia, IL
  • USA

Latest Posts

  • Seth
  • Zenz
  • Imperial College London
  • UK

Latest Posts

  • Nhan
  • Tran
  • Fermilab
  • USA

Latest Posts

  • Alex
  • Millar
  • University of Melbourne
  • Australia

Latest Posts

  • Ken
  • Bloom
  • USA

Latest Posts

Warning: file_put_contents(/srv/bindings/215f6720ac674a2d94a96e55caf4a892/code/wp-content/uploads/cache.dat): failed to open stream: No such file or directory in /home/customer/www/quantumdiaries.org/releases/3/web/wp-content/plugins/quantum_diaries_user_pics_header/quantum_diaries_user_pics_header.php on line 170

Monica Dunford | USLHC | USA

Read Bio

Blooming Dipoles

Tuesday, March 25th, 2008

The Geneva area is really quite ideal in terms of climate. During the winter the many mountains in the area get tons of snow but there is rarely snow in the city or surrounding towns. As a result you have the best of both worlds: great skiing very nearby without the drudgery of constantly digging out your car. But occasionally we do get snow storms in the lower altitudes, such as this past weekend. For me this serves as a reminder of why I don’t live in places like Buffalo, NY, for example. Although, I do rather enjoy the occasional digging out of my car from a foot or so of soft, fluffy snow. It is quite therapeutic. It is not, however, therapeutic to dig out my car from 4 feet of very compact snow, dumped by the snowplow directly in front of the car. This being the situation I found myself in this morning. Honestly, snowplow person? Did you not see my car there?

Those frustrations aside, as a native Californian I was raised with the belief that the end of March is Spring (go ahead, laugh. But ask any Californian when Spring is and you will get a similar response. This is because Californians deep down believe that seasons are really just fictional, made up by Northerners and East-coasters to discourage us from vacationing there). So as April looms, I expect to wake up to my garden flowers blooming, not to my front-door stairs becoming a ramp of snow.

But apparently the LHC magnets are responding to the call of Spring. Over the past few weeks, magnets have been popping up everywhere. In the center of round-a-bouts, outside supermarkets, and several places around CERN, such as this superconducting dipole magnet which is just outside my office building.


All of these magnets are being displayed in anticipation of CERN’s ‘open days’, which take place during the first weekend of April. During the open days, all access points including the beam tunnel and all experiments are open to the public for tours. If you are in town, go! It is a great opportunity to see the guts of the LHC and the detectors.

This dipole magnet shown in the picture is what one of the main dipole magnets used in the accelerator ring looks like. Of which there are 1232 in total. Actually there are almost 9600 different magnets used in the LHC. This fact guide (linked at the bottom of the page) gives a description of the purpose of the many different magnet types.

And of course if you are entering a round-a-bout and happen to see one of these huge magnets in the center, don’t worry about your car being sucked into it, these are just shells. But anyone can see the real ones during the open day.


Particle Interactions

Thursday, March 20th, 2008

Why is this detector so complicated? I often find myself asking that question. It is usually coupled with the exasperated ‘Why isn’t this code compiling?’ or ‘Why is the DAQ not configuring?’ or ‘Why has the front-end electronics stopped sending data?’

As I was meandering through interactions.org (which is a great resource for fancy physics pictures), I came across this nice picture of particle interactions in ATLAS. Which does in a simple way start to address the question of why ATLAS is as complicated as it is.

ATLAS Particle Interactions

What is shown here is a pie-slice of ATLAS from the perspective of looking down the beam-pipe. The white circle at the bottom is the beam pipe, in the center of which the proton-proton collisions occur. One of ATLAS’s design goals is to detect new particles such as the Higgs or Supersymmetric particles. But that is a bit misleading because we don’t really detect these particles themselves, we detect their decay products. By measuring those decay products, we can reconstruct any new particle’s properties, such as its mass.

So the particles we actually observe in the detectors are mostly just the ordinary things like electrons, photons, muons, protons, pions and the like. Different detector types are better at measuring say, an electron, than a muon, therefore in ATLAS we use many different detector technologies so that we can be sure we don’t miss anything.

This picture nicely shows which sub-detectors within ATLAS are better at measuring what. The closest detectors to the interaction are the pixel detector, the semiconducting tracker (SCT) detector and the Transition Radiation Tracker (TRT). Collectively known as the ‘trackers’ or ‘inner detector’, these detectors aim to track the trajectory of charged particles. The charge particles are bent by the magnetic field provided by a solenoid magnet. From the direction and magnitude of the curvature, we can determine the charge and momentum of the particle.

The next layer, the calorimeters, measure the particles’ energies. The first layer of calorimetry, the electromagnetic calorimeter measures the energy from photons and electrons whereas proton and neutron energies are largely measured in the second layer of calorimetry, the hadronic calorimeter. AKA Tile Cal. Muons are hard to stop and generally exit the detector completely. Similar to the inner detector, the Muon system is a series of tracking chambers to measure the trajectory of the muons. Here there is a second magnetic field (not shown in the figure), the toroid magnetic which again is used to bend the muon’s path (and is where the `T’ in ATLAS comes from). Particles like neutrinos are completely invisible to ATLAS. We can only infer their existence by measuring the `missing energy’—the energy that the neutrino takes with it as it leaves the interaction and the detector.

In that light, if you have ever wondered, ‘are all those sub-systems really necessary’, the answer is definitely, ‘yes!’.


Cosmics vs Beam

Friday, March 14th, 2008

In response to my last posting about the ‘6th Milestone week’, the following questions were posed by Jacques.

Once you are satisfied with the results of this test, or any subsequent test that might be decided using cosmic rays, will you “just” have to wait for the LHC to start beaming, at which time you can immediately start gathering and exploiting the data from proton collisions, or does another heavy test campaign begin then? For how long?

Outside the difference in the data frequency/volume(is this not a 1 to 1 million ratio?), are there limitations in the current ATLAS tests due to the nature of cosmic rays? Or will the most volatile particles created by the collision decay so quickly that you can only track the results of such decays for which cosmic rays are a good subrogate to calibrate the various detectors and check they deliver consistent “tracks” whenever a given particle crosses them?
How about the calibration of the Level One Trigger in this context?

For starters, ‘satisfied with the results of this test’ is a constantly changing criterion. A year ago, satisfaction was getting just two sub-systems to run together. Today satisfaction is running all sub-systems with a level-one trigger rate of 10kHz. Next month, satisfaction will be running at 100kHz (which is the level-one rate we want to have during beam running). At the start of each ‘M-week’, we have a whole list of problems that we experienced in the previous ‘M-week’. At the end of the week, we have an entirely new list of problems which are generally more complicated and therefore more difficult to solve. So I suppose forward progress is defined as finding harder and harder problems.

We will never be in a position where we ‘just wait’ for the beam. Nor when the beam is running will we just be waiting for the data to roll in. It is a cultural trait of high energy particle physics to push the system. If we are stable with a level-one trigger rate of 100kHz, someone will suggest an idea to push that rate to 120kHz. There is a saying, ‘If it was easy, it would have been discovered already’. Thus in order to make the big discovery, we have to be willing to take risks and push the detector to its design limits and if possible beyond. And pushing the limits is all the fun!

In the period before the beam, cosmic rays aren’t a great way of testing ATLAS’ limitations. But it is all we have. Using the muon trigger chambers, the cosmic ray rate is about 100Hz. The beam will be 40 MHz which is roughly a factor of a million greater. We try to push the rate during cosmic running by using a high rate ‘random’ trigger but there are no physics events associated with these triggers. Additionally most cosmic events tend to be a single muon slicing through the detector. Whereas with the beam running there will be hundreds to thousands of particles in the detector.

Cosmic muons are very useful to study tracking in the inner detector and muon chambers. For the calorimeters, we can use them as a preliminary cross-check of our energy calibration. They are also helpful to establish the relative timing between sub-systems (which needs to be known on the nanosecond level). While cosmic muons are helpful for calibrating the part of the level-one trigger that looks at muons, it doesn’t help us much with calorimeter-based level-one trigger. The reason is that the calorimeters measure energies that are typically associated with “jets” of many particles, not a single track like those from cosmic muons.

The bottom line is cosmics are all good and fun. But if given the choice of cosmics vs beam. Give us beam!


The Sixth Milestone

Tuesday, March 4th, 2008

Another ‘Milestone week’ is here. As with the previous ‘Milestone week’ the purpose is to run as many subsystems of ATLAS together, all taking cosmic ray data. In the usual day-to-day routine most subsystems do their commissioning independently. It is a special event to combine multiple subsystems and a milestone to combine them all.

This week is the 6th and supposedly final ‘M’ week. Or M6. Some people say it is the final M-week probably because there is as of today no M7 scheduled. But in the fine tradition of schedule-making, the end of the Milestone weeks probably just means moving to a new letter. ‘N’ being the next logical choice in this case.
But the M-weeks are fun (as well as frustrating and exhausting). The control room is a hub of activity. The air is simultaneously filled with tension and excitement. With each M-week, the goal has been to add in more and more subsystems. For M6, we have all but the pixels. So written out in full acronym glory, this is

MDT, RPC, TGC, LAr, Tile, L1Calo, SCT, TRT, CTP, HLT, DAQ, DCS, Tier0, DQ, Offline

This can be generally translated as ‘no available chairs in the control room’.

With five M-weeks behind us, it seems the subsystems have finally started to play nice. In the past, one of the most time-consuming things was simply to combine the data acquisition systems for each subsystem. So far the combination of multiple systems has been very straightforward and rather painless.

But then again I shouldn’t speak so soon. It is after all only tuesday.


Tower of Babel

Wednesday, February 27th, 2008

I am trying to think of a word that adequately describes my ability to learn languages. Abysmal, for example. That would be a good description. Actually even my English can be quite ‘creative’ at times. So much so, it once led my advisor, Gene, to ask, ‘Are you sure that English is your native language?’. Thus, my personal inability to learn languages makes me envy and admire many of my colleagues who have no problem conversing in whatever language necessary.

When it comes to communication, English is certainly this field’s lingua franca. Talks at conferences, meetings, analysis papers for example are all in English. But at CERN, native English speaks are a minority. So it is not uncommon to be in a meeting (in English) and yet be the only native speaker. Nor is it uncommon to be speaking with someone whose English is very, very good, essentially perfect and then learn that this person has never actually been to an English-speaking country. That they have learned English in the presence of non-native speakers.

One hears complaints that the English spoken here has become ‘twisted’. In the sense that it has become its own dialect. We have of course both American and British dialects, but now maybe we have a generic non-native dialect which is something of a melting pot of the 100 or so languages spoken here at CERN. I for one rather like it. It is interesting to see how the Greeks will phrase a sentence, compared to the Italians or Russians or Chinese. It is a little glimpse into their own language through the English.

Although English is generally the one language nearly everyone uses, it is interesting to see this web of communication that people create. Most of my European colleagues, for example, tend to speak three languages: their native language, English and then a third. It is very common therefore to have two people who not only both speak English but also both speak German. And why speak only English, when you can speak both? Some things are better said in one language then another. And I often see people switching from one language to the next where it best fits the conversation. The two languages are simply combined. And why not: isn’t two always better than one?


Small Wheels Descend

Tuesday, February 19th, 2008

I fully and completely admit that I am a total sucker for heavy machinery. It is something I have in common with most four-year-olds. Take us to a construction site, feed us peanut butter and jelly sandwiches and we will be happy all day.

On that note, I couldn’t resist showing some pictures of the ‘small wheels’ being lowered. The small wheels are part of the Muon system. They are located between the end of the calorimeter and the toroid magnet end-caps. And they are the last big piece of the detector to be installed.

As the small wheels were assembled on surface, the first order of business is transporting the wheels to the ATLAS surface building. Fortunately CERN has many custom trucks for this kind of transport. Seen here one of the wheels is moved (very slowly) into the surface building on this massive, two-lane wide flat-bed truck. And this truck is just one of many in CERN’s armada of very cool transportation vehicles. These vehicles do contribute to frustrating local drivers, though. It is one of the interesting ‘features’ of the area. It is not uncommon, for example, to get stuck behind either some slow-moving tractor carrying hay or some slow-moving truck carrying a huge super-conducting dipole magnet.

Moving the Small Wheel

From the surface building, the wheel is lowered by the crane down one of the access shafts to the ATLAS cavern. In the picture, one small wheel is attached to the crane and suspended over the access shaft, while the second (in the foreground) waits its turn.

Small wheels at surface

So why do we call them small wheels? In this picture as the small wheel is lowered into the cavern, you can see one of its siblings, the ‘Big Wheel’ on the right. Small really is just relative.

small and big wheels


ATLAS Real Language

Tuesday, February 12th, 2008

The main language within ATLAS is not English or French or Italian. It is the language of acronyms. Ask any new person what the most difficult thing about ATLAS is and they will say understanding what people are saying. Oh yes, they can understand all the words and syllables but not the encoded meaning of the acronyms.

The number of acronyms is a little out of control. We have webpages devoted to our acronym definitions. ATLAS itself is a bit of a kluge of an acronym.

ATLAS: A Toroidal LHC ApparatuS

I think this is a great example of deciding on the acronym first and then settling on the meaning.

Additionally with so many acronyms, you end up having acronyms that are phonetically the same. Such as

DAQ: Data Acquisition System
DAC: Digital to Analog Converter

This leads to conversations like

Person one: I am trying to measure the exact setting of the DAC
Person two: But we know what value the DAQ is setting. It is written out to the log file.
Person one: What are you talking about?
Person two: What are you talking about?
Person one: The DAC setting. D-A-C!
Person two: Oh. Right.

And then it is inevitable that you get to acronyms of acronyms. Like


which fully stands for Read-Out Driver Crate Data Acquisition System

And then there are acronyms like this….

OTSMOU (pronounced Ots-moo) meaning the Operation Task Sharing and Maintenance and Operation Update.

This is my personal favorite acronym of all time. This acronym is genius for so many reasons. a) Seeing the full meaning of the acronym offers the reader zero insight about what exactly this is. ‘Maintenance and Operation Update’? Is that a group of people or a software package? It is unclear. b) The word ‘Operation’ is used twice. c) The acronym just exudes management. It is difficult to create an acronym that once spoken instantly conveys the image of management to the listener. Yet, this acronym accomplishes just that. And d) the person inventing this name clearly must still be laughing.

My only comfort is that if people (management excepted) continue to stick to the TLA (Three Letter Acronym) rule, we will eventually run out of combinations.


Down and Up Again

Tuesday, February 5th, 2008

The looming nature of ‘the closing’ has injected an increased sense of urgency into all those still working on detector installation. Since TileCal has been installed in the pit for several years now, you would think we would be exempt from such urgency. But no. In the past year or so, TileCal has undergone a campaign to repair some less-than-optimal components in our electronics. The front-end electronics for the TileCal are organized in long ‘drawers’ which can be pulled out from the ends of the calorimeter. We are replacing things like the power connectors which have been causing some problems over time. In the past year, we had the time and detector access to make these repairs so we decided not to wait for the problems to worsen.

There are 256 electronics drawers in TileCal so upgrading every single one is more than a day’s work. The drawer itself is almost nine feet long. It has to be removed from the calorimeter, lowered down to the floor of the detector cavern. On the floor, three electronics tables are set up where technicians can make the modifications. Once done, the drawer is tested, raised back to the calorimeter, re-inserted and re-tested.

One of the most difficult parts of this procedure is just getting the drawer from the calorimeter to the technician’s tables. There is no space to make the modifications right at the calorimeter so moving the drawer is the only solution. Furthermore, the scaffolding surrounding the calorimeter is accessible by ladders so we have to invent some creative ways to get the electronics drawers up and down the scaffolding.

One technique is to lower and raise the drawers through the access areas in the scaffolding. As seen here. The blue boxes at the top of the picture is the part of the calorimeter, where the electronic drawers are inserted. In the center of the picture, part of one of the drawers is being raised between two access ladders. This is delicate work. You don’t want to go banging your newly repaired electronic drawer against the sides of the scaffolding. And these are all custom-built electronics. It is not like you can go get a replacement at Radio Shack. On my former experiment, SNO, the electronics racks had signs reading, ‘Careful! These electronics cost more than your house!’.

raising a drawer

It is times like these where I really admire a technician’s patience. The pressure of ‘the closing’ is increasing, everyone knows that soon the scaffolding will come down, the collective heart rate has gone up several beats. But the technicians are never fazed and continue to raise the drawers with the same patience and precision as ever. They know better than anyone: you raise the drawer too fast, you will break it. We are in good hands with those guys.


The Closing

Tuesday, January 29th, 2008

Despite its size, ATLAS is not a completely stationary detector. It has a stationary infrastructure but it does have some parts that can move around, allowing for access to the internal elements of the detector. (By contrast CMS is very mobile–it is like a gigantic Yule log where each section can be separated from each other.) On each side of ATLAS, there are three big moveable parts as seen in this picture. The muon wheels shown by the turquoise arrow can be moved back. This allows for the end-cap magnets indicated by the red arrow to be pulled out. Once the end-cap is moved, we can gain access to the extended barrel calorimeters (such as TileCal). The extended barrels shown by the blue arrow can also move forward, allowing for access to the calorimeter barrel as well as the inner detector. But space is limited so it is not possible to access all places at once. And so in typical ATLAS fashion, there are committees in place to organize who gets access when.


But the end of April is ‘the closing’. Or perhaps I should say THE CLOSING. Or maybe even, THE CLOSING.

Where all the moveable parts are put into their final location. And the installation of the beam pipe and the beam pipe shielding begins.

The bad news is that we will lose access to the detector. The good news is that it means the beam is coming.

The closing is one of those things that can simultaneously fill you with ecstatic excitement and absolute panic. And it really is a simultaneous feeling. On one hand it is like sitting in a Ferrari waiting for the keys. Feeling the pressure of the seat against your back, the feel of the wheel under your hands, anticipating the roar of the engine when you hit the gas and thinking, ‘Alright. Show me what 0 to 60 really means.’ And yet at absolutely the same moment, it is like the night before your last final exam. Staring intently at a pile of books, willing the information to leap out of the pages and into your brain, lamenting, ‘why didn’t I get to this sooner’, but knowing that it was impossible in any case because you had other exams which took precedent.

But heedless of anyone’s feelings on the subject, the closing approaches. And I will bet that as it continues to approach the two most common phrases in the halls will be either ‘We are closing soon, the beam will be here!’ or ‘We are closing soon, and it is the end of January already. How did it get to be the end of January already?!’.

The end of April, it is so far and yet so close.


In the Maze

Wednesday, January 23rd, 2008

Ah, the ATLAS counting rooms. The destination of many of the cables coming from the detector. It is quite a nice series of rooms. All with raised flooring so that cables can be run underneath. There are a large number of racks with water cooling for each sub-detector. Most of my time underground is spent in the counting room also known as USA15 (which does not stand for United States of America). I can be found usually near the Tile racks on the first level or the Level-one trigger racks on the second level (which is where Tile’s trigger cables are).

But I am not the only person working in USA15. And sometimes the work patterns of the other sub-systems can turn the counting room into a human maze. Suddenly just getting from point A to point B becomes an amusement-park challenge.

Let’s take this particular day last week for example. It went something like this…

Finish tests on the upper level and start lugging heavy oscilloscope back downstairs to the Tile racks. Lo and behold, people are running cable and they have removed a piece of the floor near the door.


No problem. Will use other door on this side. Work self through TRT racks to other door. More people working. Naturally.


Will take back route. Go back through TRT racks, past stairs with intention to go behind elevator wall. Hm. See many level-one people frantically cabling on both sides of the racks. Do not look like they wish to be disturbed.


Go back upstairs. Can’t we get a lighter oscilloscope? Cross the upper level, take the outside stairs down to hallway on the lower level. People on other side of door have one section of the floor removed. Is this a joke? Observe that people behind door are in the same cabling group as those on the right-hand side door. Obviously they are running cable under the hallway. Should have seen this when I first tried to exit on the right-hand side door.


Retrace steps. Back upstairs, then downstairs to level-one cabling people. Apologize profusely as I interrupt them to pass. Stagger to Tile racks to put down scope. Right arm is now completely dead. Will be one-hand typing for the rest of day.


Now must get to elevator. Head back to the level-one racks, interrupt a different set of level-one cablers. Again apologize profusely. Back upstairs, across the upper level, and downstairs. Notice that hallway cablers are done. Both doors now clear. Typical. Head to elevator. Success! Have conquered the maze again!