FRIB Operators

Welcome! If you haven't already please check out What is an Operator for a better idea of what your role as a new Operator here is. I also introduce some of the basics of the machine there and will reference those portions.

Operators come from a wide variety of backgrounds, some of us have Master's in Physics and some have Bachelor's in totally different fields. All this is to say that nobody joined as an expert in accelerators and even those who had some foreknowledge had to become accustomed to our particular lab. These are pretty complicated machines and by their nature are hard to understand in their totality, a lot of concepts intersect, overlap, and interact is some confusing ways.

Accelerators take plasma, and by repeatedly exposing it to sinusoidal electric fields as it progresses down the beamline. accelerate it to higher and higher velocities. If that's confusing, and it should be at least a little confusing, then it can be simplified to ions going through the accelerator, being accelerated by cavities. You'll hear a lot about cavities as they are as vital as they are temperamental, often causing many of MPS Faults mentioned in What is an Operator You don't need to know how to disassemble a cavity blindfolded in under 60 seconds in the middle of warzone, but knowing the name will help you familiarize yourself with them as they pop up in many Control Room conversations.

This page is intended more as a long term project to read through or a reference you can check. I tried to organize the What is an Operator page from most basic to most complex primarily and most common to rarest secondarily, but here I'll summarize in a fair amount of detail each component from the source of the beam to its furthest possible destination. That being said, the intention is still to help along newer Operators, so I'll try to remain approachable.

Sources are where things begin. Source Physicists start by obtaining a sample of an element, this can be a variety of isotopes that are either gaseous or solid. Solid sources are placed into an oven until material sublimates off, this is unnecessary for gaseous samples. From here they enter into the Plasma Chamber.

There are actually a variety of isotopes in the Plasma Chamber, this is referred to as the cocktail. This allows us to change the beam we're sending to another isotope fairly quickly by adjusting devices.

These isotopes fly around the Plasma Chamber and get excited by Radio Frequency Waves. This strips off some of their electrons and imparts them with more energy. You can already imagine that we don't want random air particles in there, so much like the rest of the beamline, the Plasma Chamber is under vacuum.

To ensure that the ions stay in the Plasma Chamber for long enough, there are two solenoids on either end that create two magnetic fields, one by the injection side where the samples enter, and one by the extraction side where they leave. These two fields are not equal in strength, the extraction side is weaker so that some ions may escape out into the rest of the beamline.

A common graph that Operators use will look at the following measurements that will then be explained.

What is ECR Drain Current? Just realized I don't know

What is Average Pressure? The Pressure within the Plasma Chamber This normally hovers around a few nTorr (as of writing, it's 16 nTorr, about 2e-6 Pa or 2e-11 atm.) Good to check to see if the source is being unstable or to compare to the Drain Current.

What is Microwave Power? The power of the RF that's exciting the ions, I don't check this too much as it's a fairly reliable system that isn't dependent on what's going on inside the Plasma Chamber. This can be adjusted if instructed by the Source Physicists to help beam stability or power.

What is Bias Disk Current? The bias disk is weird, To quote the best source document I've read “All agree that it has clear benefits, but similarly agree that the mechanisms of how those benefits are achieved are less clear.” It is a negatively biased disk in the Plasma Chamber that will be stand among the Drain Current and Pressure as things to check when you think the source may be acting strangely.

The last twist I'll throw at you is that for FRIB there are two Sources, and that's just FRIB! They are Artemis B and HPECR (formerly Venus). We only send from one source at a time, but I'll explain that more in the next section.

The front end is the section of the facility where the FRIB sources live. These are called ISRC1 and 2, and the beamlines attached to them are called SCS1 and 2. Hopefully you get the vague gist of how we allow beam to go down the beamline, but first let's focus on how we stop beam from continuing down.

First are the Source Slits, these are are devices that can move to create a smaller or larger area for beam to travel through. There are 4 source slits per source, two vertical and two horizontal. We do not touch the horizontal source slits. If you imagine the fresh beam that exits the sources, you can think of it as being largely homogeneous vertically, that is any ion should look similar to any other ion if you took a vertical slice. Horizontally it is heterogeneous, with the ionization or charge state varying from one side to the other. Accelerator Physicists use the horizontal slits to exclude unwanted charge states from the beam, Operators and others use the vertical slits to make small adjustments to the amount of beam being sent. Note: Opening the vertical slits too far can cause beam loss and lead to trips so this should be done with a fair deal of consideration.

Also in SCS1 and 2 are the Source Cups. Source Cups are Faraday Cups that are very close to both of the sources, the ideal “safe state” the machine can be in without the lengthy process of turning the sources off and on. Anytime there is expected time where the beam will not be sent, it's common to insert the source cups. This also allows us to do a variety of measurements on the source and the beam that's immediately exiting it. A Faraday Cup is a device that will block beam from going further down the beam line. They're useful for making sure a beam doesn't go further than intended or for measuring the strength of the beam. Based off of the current read from a Faraday Cup we can estimate beam power without having to send it or see how much beam is present at a particular point in the beam line. Another use is that only Operators interact with the Chopper, but many physicists can control the Faraday Cups, so it is a common practice to leave beam on Faraday Cup 1102 and allow the physicists to either retract or insert it as they want beam.

You'll also find magnets in the Front End, any diagram you see will have bends in the beamline, they're vital for excluding unwanted beam. The beam doesn't naturally bend, but we can make it bend with Dipoles these are magnets that create a magnetic field either pointed up or down, using basic vector cross multiplication (right hand rule!) you'll see why this causes beam to turn. All Dipoles exist to turn beam. Note that due to math I really don't want to have to type out that these magnets do not change the speed of the ions, or the magnitude of their velocity. Think of a dipole as only moving the steering wheel in your car but leaving the pedals alone. By either turning on or off dipoles in the Front End, along with inserting one of the source cups, we can ensure we're only sending beam from one source.

Another kind of magnet is the Quadrupole, which is like a combination of two dipoles. Quadrupoles exist to focus the beam and should not redirect it, though they can if the beam isn't centered. You'll find both horizontal and vertical quadrupoles, these work in tandem to first smush the beam in one direction, then the other. Their field is a bit more complex to imagine, but the configuration is the elementary consequence of having four magnets in each corner of a square, alternating polarity so the bottom left and top right magnets are both North and the top left and bottom right are both South.

After SCS1 and 2 join together we get our ULEBT, or Upper Low Energy Beam Transport. Remember, we've created a plasma but haven't really accelerated it so this is very low energy. The “Upper” refers to the architectural location of this portion, it is literally above the rest of LEBT. The first major thing here is the Chopper, which is probably the most familiar bit of the machine for you at this point. It's crucial that the Chopper is in LEBT as the lower energy beam is much easier to redirect.

The Chopper is great, but has more utility than I've previously let on. This comes from the Repetition Rate and Pulse Width Parameters. Pulse Width tells the Chopper to spend this much time allowing beam to continue down the beamline. So if the Pulse Width is 100us it will stay off for that period of time and then turn on until it repeats the cycle. How long the cycle lasts is determined by the Rep Rate, which tells the Chopper how many cycles there should be every second. You can then find the period of the cycle by finding inverse of the Rep Rate.

So let's return to our example, let's say we have a Pulse Width of 100us and a Rep Rate of 100Hz. We can say the period of each cycle is then 1/100 s just as it would be 1/50 s if the Rep Rate were 50. Let's convert 1/100s into microseconds to match the units of our Pulse Width and we get a cycle period of 10,000us. So the Chopper will allow beam to pass by for 100us, then stop beam from the remaining 9,900us. A little more math reveals that we're only allowing beam to pass by 1% of the time. This is an important calculation that we refer to as the Duty Factor.

By Default, we like to send CW Beam, or Constant Wave. This means a Rep Rate of 100Hz and a Pulse Width of 9950us. You might have realized that those numbers don't give us a 100% Duty Factor, and you'd be right! We use the 50us gap to help with machine timing, so it's our best approximation of allowing 100% of beam to pass through. We don't change these values too often, but they are helpful for Beam Development.

From here we have our most famous Dipoles, those being the E-Bends. E-Bends are very important to our PPS System. With the E-Bends powered off there is no way for beam to go into the LINAC, note that beam should be stopped on the Chopper or ideally a source cup before turning off the E-Bends. There are 4 E-Bends in total I DON'T ACTUALLY KNOW WHY and they first divert the beam into traveling downwards through VLEBT or Vertical Low Energy Beam Transport for obvious reasons, then it redirects them into moving horizontally through the LLEBT or Lower Low Energy Beam Transport.

Getting Bored? That's alright, don't forget to bother the other Operators about what's going on and also we're on to a new kind of device!

That's right! More Beam power. You might remember that we can make slight adjustments to beam power using the Source Slits, but for larger changes a common device we use are the Attenuators. I like to think of them as industrial sieves that you strain beam through. There are 4 different attenuators on their separate drives. Each one of those is actually two smaller drives, so each attenuator can be in a total of four different states. The first attenuator has the following states

The others are different, but we'll use this one as an example. If both drives are not extended, then there is no part of the attenuator that is in the path of the beam. If the first drive is extended then the 2x attenuator enters the path of the beam, blocking about half of it though attenuators are notoriously inaccurate and weird. NEVER trust what the attenuator says. Only extending the second drive would make the 5x attenuator enter the path of the beam, leaving you with about 20% of whatever it was initially. Extending both inserts the 100x attenuator. You'll notice if you try changing the attenuator configuration that it will shoot up sometimes or return odd numbers at others. This is a consequence of this drive system. You can see the Attenuation Factor on one of the TV Displays in teh Control Room, ask one of the other Operators where it is.

Generally because of how they work, the different attenuators will interact in pretty weird ways that normally kills any beam output you were hoping to get through. Best practice is to skip one attenuator when possible, so should you need to use 2 attenuators it is better to use the first and third than the first and second. Attenuation also muddies up the beam in an odd way, which mostly isn't a big deal, but the Accelerator Physics sometimes wish to not use them for this reason.

I call it the RFQ mostly because you'll almost never hear the full name of the Radio Frequency Quadrupole, I like the RFQ because it's the first time we really accelerate beam in the direction of the beamline and also it is normally pretty well behaved. CAN“T RESEARCH IT ON THIS MACHINE HAHAHA

The MEBT or Medium Energy Beam Transport gets its name from the slight boost to energy that we see after the RFQ. Before we get to the cavities and everything, I'd like to shout out Faraday Cup D1102. This is a particular Cup that is designated by it's “D Number”, this is the number of decimeters from some point, this changes at points through the beamline so for simplicity's sake just remember that that number INCREASES as you go down the beamline. So if we compare Faraday Cups D0998 and D1102, you can tell that D1102 comes after D0998. This is very important for Faraday Cups as if you have both D0998 and D1102 inserted it's important that you know where the beam is stopping.

D1102 is a very important Faraday Cup. It is used often when working with Physicists or checking beam before sending it along. Good sources for this are the pages for Primary Beam Development (PDT), Secondary Beam Development (SDT), and Beam Power.

This is the meat of the accelerator. There are 45 cryomodules and inside each there are generally 8 cavities. These cryomodules are split up into different groupings to help clarity when dealing with such a large number. The vaguest distinction we can make is where the cryomodule is in LS1, LS2, or LS3.

These are “Linear Segments” of the beamline. The beam first enters LS1, which is the middle of any paperclip-esque diagram you'll see of the accelerator, then it will be turned 180 degrees by dipoles through the first “Folding Segment” until it enters LS2. The same process occurs through FS2 (Folding Segment 2) and the beam then goes through LS3.

From here, the cryomodules are further divided based on their name. Cryomodules aren't referred to by their D Numbers like Faraday Cups are, you can imagine they're quite a bit bigger so it's a less useful designation. Instead They're referred to as “C-XYZ” where X is a letter, and YZ is a number. If YZ is less than 10, Y will be 0. For example, the first cryomodule is CA01, the next is CA02, then CA03, then CB01. Notice how we changed the lagging number back to “01”. There are more CB cryomodules than CA cryomodules. Last thing of note is that cryomodule CC12 is called as such, so if the number is greater than 10, the 0 disappears. THIS IS EXPLAINED BADLY

The cavities within the cryomodules do have D numbers, but this can be a confusing topic. Operators will almost always refer to cavities by their D number, you'll hear of “cavity 1195”, “1372”, and “1711” quite a bit and those names are just their D number (with the D excluded for brevity). However, many physicists and cavity experts will refer to them by their cryomodule. Here, the aforementioned cavities would be “CA03 cavity 1”, “the third cavity in CB03”, or “the third from last cavity in CB08”.

So what ARE the cryomodules and cavities? Cryomodules are collections of cavities that rely on the same cryogenics system for cooling, cavities within the same cryomodule can affect each other but it's much more rare for cavities from different cryomodules to affect each other as each cryomodule is a big metal box.

Cavities create electric fields that quickly vary their strength sinusoidal, or like a sine wave. The idea being that any electrical force is conservative, you can't gain energy by approaching and then going away from an electric force. Let's say the force is repulsive, you would lose the same energy approaching the cavity as you would moving away from it, so if you're only considering this electric force, you haven't accelerated at all.

Cavities then diminish the field as ions approach so as to not repel them and then ramp up as the ions fly away. The process is a bit more complicated, but know this was well considered enough to get a large portion of beam up to one consistent energy.

The period of this ramping sine wave is not something to worry about, two measurements that are worth understanding are the “Amplitude” and “Phase”, You can probably guess, but the amplitude is a way of measuring the vertical span of the sine wave which is tied to the electric field strength we want the cavity to output. The phase is the offset of this sine wave in comparison to other cavities.

We talked MPS Faults before, and cavities are a main perpetrator of these. There are two main kinds, first is a Phase Wobble, this was a brief instability in the phase that fell out of tolerance. The cavity itself doesn't shut off and MPS sees it as being good to go, but the Interlock still exists and needs to be acknowledged. Cavity Trips are some error that “turned off the cavity”. These need to be investigated and the cavity restarted.

The scare quotes in the last sentence are probably worrying, but there's two modes of “off” for a cavity. There is the Low Level Radio Frequency and the Amplifier. I don't fully understand these so I'll spare you from my hypotheses, though I do have a hunch. With a Cavity Trip the LLRF is most likely what turned off and the Amplifier is on. In general, anything that causes an amplifier to turn off is a pretty big deal and scary.

Speaking of scary, remember that these cavities are all cooled within the cryomodules, that means one of the measurements we take is the Fast Thermometry of the cavities, which is normally just the variance in temperature compared to what we think it should be. I say this is scary because any cryo related thing scares me, but realistically this is more often a sign of beam loss.

I'll start off by saying that I really like everyone who works on the Lithium Stripper. They're smart and friendly people that I have a lot of respect for, that being said the Lithium Stripper is a challenging piece of machinery. The Stripper gets its name from stripping the incoming beam of even more electrons, it does this by sustaining a thin film of liquid lithium that travels super fast. With such a dynamic process it should follow that there are occasional instabilities that can change what kind of ions come out of it.

One frequent example are minute holes rapidly appearing and disappearing in the film. This will trip MPS and stop beam, though it is very easy to clear. Even within stability it can change though, there are a few different stable configurations that the film can swap between irregularly. Some of these configurations trip MPS more than others, but any change can cause excess energy loss that, you guessed it, trips MPS.

We use the Lithium Stripper Stage Feedback Tool in order to counteract this, the stripper stage eludes me somewhat admittedly, but a summary of how we interact with it can be found here.

I talked quite a bit about tripping MPS there, as far as the Lithium Stripper goes this will likely be because of the Beam Current Monitors or BCMs. They measure the amount of beam at one point in the beamline and gives units of uA. The ones that trip MPS the most are the Differential Beam Current Monitors, which are just two BCMs that compare readings. If a significant amount of beam is lost somehow between those two points then beam trips off. We have poked holes in the beamline from “losing” beam on a bellow so it's something we need to be very careful about. This is covered under MPS as this beam loss really only endangers the machine.

Different Differential BCMs are normally named after the region they're monitoring. LS1TRANS is the transmission through LS1, once you're comfortable with the LINAC and its components most of them will make sense, the only exception possibly being the LINACTGT, which measures from the start of the LINAC to the target.

“Lost Beam” being solely a machine hazard may sound worrisome, but