Replacing a Ta In Shutter in a Glovebag Journal
Replacing In Shutter in a Glovebag Journal
Written by NTS from email 5/31/2016.
This is a serious procedure for which we must thoroughly plan and prepare.
This procedure was done for the In cell, and we will speak in terms of both the In shutter and a Ta Shutter.
Timeline
Timeline of Successful Ta In shutter replacement action
Day 15 (4/27): Prep for trimming new Ta In shutter, and Trim new Ta In shutter using tin snips and small, full cuts.
Day 16 (4/28): Acetone and isopropanol sonication treatment on shutter blade (standing on blade in Treatment beakers), put shutter on Bravo trolley and pump down to UHV in LC (secure with Ta wire on pins), pump purge LC and RM, LC lamp bake on LC (only), review plan.
Day 17 (4/29): Attempt Ta In shutter replacement number two. We succeeded!
Prep
How to replace the In shutter without taking off the SF has been discussed. Taking off the SF is a major maintenance that requires baking of the Growth chamber and would add a lot of time to this maintenance cycle. The first thing we have to do is finish scraping the In pocket. Installing a new shutter and then scraping more would be a risky, bad idea.
Plan for Replacing In Shutter
This work will likely require some special ultraclean tools and techniques. The new shutter will have to be trimmed and cleaned before install.
Remove the old shutter
Currently, the shutter rod is attached to the adjustable shutter bellows with a set screw, and attached to the remaining pBN shutter. We have to slightly remove the bellows, grab onto the shutter rod through either the bellows port or In port, and undo the set screw on the SF side (while holding the bellows and shutter). We can then extract the shutter out of the GC (through the In port).
Install the new shutter
Put the new shutter into the GC through the In cell port and get the shutter rod in the GC then out of the bellows port (use two ultraclean long forceps). Hold the shutter bellows next to its port at an angle, put a gasket onto the bellows, simultaneously tilt up and insert the shutter rod into the bellows rod holder/female end, then tighten the set screw. We can then raise the gasket with an ultraclean fingertip, contact the bellows to its port, and make sure the shutter is oriented correctly. Put in top and bottom bolt loose fingertight. Check alignment with viewports and slow manual actuation. Do more minor adjustments, the shutter bellows flange is rotatable, which simplifies rotational adjustments. When good, finger-tighten top and bottom bolts, and put in the rest fingertight, while watching the shutter constantly. We then wrench tighten down the bellows, check alignment, and have a new shutter for the In pocket.
In Shutter Replacement Attempt One
Glovebag SF - P1: Kyle, P2/P3: Nate, Emily, Dan, Stephen
Glovebag RHEED - P1: Siff, P2: AK
Parts and Tools: new 3" adjustable pBN shutter, ultraclean Tyvek sleeves and ultraclean nitrile gloves, ultraclean long forceps, ultraclean allen keys.
Outcome
Not successful, by millimeters.
Negatives
Shutter won't fit in through SF water cooling panel from RHEED port side. The cryoshoud blocks part of the panel toward the outside, where it is wider.
CAR is kind of in the way.
Shutter won't fit into In source port.
Dropped the shutter, likely into the sump.
In bellows are damaged at an edge weld.
Positives
Much was learned with an eye toward successful replacement next time.
Caught damage to In shutter while already in maintenance.
The camera idea worked pretty well.
Pictures were taken for posterity.
The Story
We started by taking off the In cell and transit casing it, then removed the RHEED screen viewport. Siff could then ultraclean tyvek sleeve and glove up, and get his arm into the GC. Kyle removed the shutter bellows slightly and loosened the shutter rod set screw. Siff attempted to reach his fingers into the SF water cooling panel pocket and grab the shutter for extraction, but the resulting gap between the cryoshroud and SF panel was too small to get more than a finger or two in. The contortion of his arm to get it into the chamber and avoid other things such as the CAR was noted. The CAR was in the 380-410 degree range for best camera shot and arm movement. Using ultraclean tools to grab the shutter rod from the SF side (and passing off the bellows for holding by P2), Kyle was able to extract the remaining one-third of shutter through the bellows port.
The install of the new shutter was attempted. Siff put it into the growth chamber through the RHEED port, rod first, and fed that into the SF panel. Kyle could grab the rod with the long forceps. The two of them tried to push, pull, wiggle, and rotate it into the pocket, but it would not go in. The cryoshroud to SF panel gap was too small. Amazingly, there were no issues with the shutter itself. During this time is when the In shutter bellows damage was discovered. The next idea was to try to get the shutter into the chamber through the In cell port. While extracting the new shutter from the GC, it was unfortunately dropped. A rescue was attempted, but it couldn't be found and likely fell into the sump. We will look again Monday with the beam flux IG.
Soldiering on, another suitable shutter and bellows pair was found. Inserting it through the In port was attempted, but the shutter rod prevented this. The blade was a bit too big to go in head on, and inserting it at a suitable angle wasn't possible due to the shutter rod. Without further ideas and prepared parts, we started to close up the system. The old RHEED screen was removed while we had it out. We did not put in the new one, so as to minimize the chance of any damage occurring during the remaining maintenance tasks. We pumped down, formulated ideas, and leak checked. The chamber went down very slowly, which isn't surprising for the amount of GC junk and As was kicked up, and the nature of the procedure we attempted (there was an arm in the GC).
New Idea
Moving forward, our ideas are either to get a tantalum (Ta) shutter and trim it a bit so it will fit in through the In source port, or modify (grind down) a pBN shutter so it will fit through the In source port. Trying either is desirable compared to pulling the SF, in both time and money. Likely in contamination risk as well. We will figure out the theoretical dimensions needed to succeed, and have a spare SF panel somewhere to practice. We will also be replacing the In shutter, as it looks like the bellows are deformed at an edge weld. Information to complete the current task will be gathered over the weekend and on Monday. Monday has no planned glovebag time, as we work to resolve the In shutter situation. Maintenance is on hold for the time being, but if the wait is not short, we may move forward with removing the Si cell (TBD). The time frame for (acquiring and) modifying a shutter is likely our current limiting factor. A Ta shutter was found in the back cabinets.
Shutter Trimming
Today, the shutter was cut to a size that will fit into the source port. Kyle and Siff got the cleaned System Foxtrot Source Flange from the shed, and evaluated how to cut the tantalum shutter we have on hand. Tin snips were chosen. The wide side corner across from the shutter rod fixture was the first part to be cut.
An edge of the shutter was slightly cut (a tiny bit at a time) and rounded. After each cut or two, Siff would check to see if it would go into the source flange via a source port. This was repeated a few times. After a few attempts, the shutter could be maneuvered into the source port. Repeatability was tested, because we want this to work the first time, and will be in a glovebag while attempting the install. Another tiny cut was made to ensure this.
Overall, less was cut off than expected. Also, a bit of the shutter near the rod fixture was snipped off. We will need to clean the shutter blade with the Treatment tomorrow, and then either bake or desiccate it. We will also run through the install process. In two days, we will attempt the real install on the GC.
Shutter Cleaning
Today, we cleaned the shutter blade with a partial Treatment (acetone and isopropanol sonication), and then put it in Bravo's load chamber to be pumped on. We put it in the LC by using Ta wire to attach it to two pins of Bravo's trolley. The shutter kind of naturally rested on two pins, but would fall off easily. Using ultraclean pliers, like the Al crucible, a few small twists were made to loosely hold the shutter in place, but this was not enough to secure the shutter while moving the trolley. Then the twists were twisted such that they would choke up on the shutter rod and trolley pins, but not so tight that anything would bend. Now the wire was snug enough to hold the shutter in place.
The trolley was put into the LC and it was pumped down, all the way to UHV, under 1e-7. Pumping on it like this could quickly draw out any residual moisture from the solvent clean. We then vented the LC, put the shutter in an ultraclean foil pocket, and placed it in the front desiccator. The trolley was returned to the LC. Because some moisture may have been sucked into the RM (used to vent the LC), we pump-purged the LC-RM three times to clear it out. The RM was left at rough vacuum and isolated from the LC. We need to run a LC lamp bake on the LC. Turn the LC lamp eurotherm to auto for running the LC lamp bake recipe. Run the LC lamp bake.
Tomorrow, we will attempt the In shutter install on the GC. We will be starting at 9 AM. It will be good to see the beginning prep part and GC venting (remember to save the RGA scan and turn off the box properly). We will only need the SF glovebag, and won't be using the RHEED screen port. The RM is at rough vacuum and the LC is at high vacuum, so we will have to move the trolley to the BC, turn off the LC turbo, connect the RM to the LC, then pump both down to UHV. There are two things we need to add to the SF glovebag: the shutter (front desiccator, maybe not labeled), and 3" adjustable bellows.
In Shutter Replacement Attempt Two
Glovebag SF - P1: Kyle/Siff, P2: Siff/Dan, P3: Stephen
Parts and Tools: new 3" adjustable Ta shutter (custom trimmed to fit into source port), new 3" adjustable shutter bellows, ultraclean nitrile gloves, ultraclean long forceps x2, ultraclean allen key.
Outcome
Successful, though with a learning curve.
Negatives:
The shutter rod was a bit hard to get into the shutter bellows female end.
The bellows locking clamp was spinning around.
The clamp was hard to tell when tight because the shutter bellows has a rotatable flange. When holding the flange or side hinges, the bellows and rod can rotate, making the shutter rod clamp look loose.
Rotating the shutter was hard at first, then we realized the bellows has a tail you can use to rotate it.
Positives:
Much was learned about shutter replacement, and the shutter bellows assembly.
The bellows tail and rotatable flange made the replacement a little easier.
The camera idea worked very well for alignment of the shutter rotation.
Pictures were taken for posterity.
The Story
We started by taking off the shutter bellows port blank, wiping knife edges, and foiling the port. Then we removed the In cell and transit cased it. To protect the shutter rod while wiggling the shutter into the In source port, Kyle put some ultraclean foil around the port flange, such that it was covering the non UHV side-- think of this as a reverse foiling of the port. The next step is to get the shutter into the pocket, through the In port.
Kyle inserted the new shutter into the In port using ultraclean gloves and carefully but firmly wiggled it into the pocket, success #1. To extend it further into the pocket, he used ultraclean long forceps to put it in. He used a second pair, inserted into the (freshly unfoiled by P2) bellows port, to grab the shutter rod and pull it out of the bellows port, success #2. The shutter is now in the pocket, and the rod is coming out of the source port. It could be safely rested there, so continuous holding was not needed. The next step is to get the shutter rod into the shutter bellows female connector (its home).
The new shutter bellows knife edge was wiped, and a clean gasket was put onto the bellows, with careful attention paid to the rotatable nature of the flange, meaning the flange can drop away from the knife edge, so gaskets can fall unexpectedly. The new bellows were put next to the bellows port and tilted over while holding the shutter rod out of the port with forceps, so the rod could contact its home in the bellows, and the gasket would be semi-secure (it's not in place yet, but it can't fall). With some effort, the rod was put in its home, success #3. The next step is to tighten the rod in its bellows home using the little clamp.
We thought this would be straightforward, but the clamp that would tighten it was very unruly, spinning around the bellows female connector, upon which it must be tightened, and going onto the rod. With teamwork, Kyle and Siff got the clamp to behave and tightened down the set screw with an ultraclean allen. However, due to the rotatable flange on the shutter bellows assembly, the shutter kept rotating and it looked like the set screw was not locked down. Even with Kyle really tightening it, the shutter would move. This is because where you hold the shutter bellows assembly steady (the side hinges) is connected to the flange, not the bellows. And the set screw Kyle was tightening and the shutter rod & shutter are connected to the bellows not the flange. In this way, the rotatable flange made this part more difficult. The next step is to align the shutter in the pocket and bolt in the bellows.
Siff swapped in for Kyle, and it was after this swap that the above was figured out, success #4. With the bellows and rod secured to each other, we needed to align the shutter. The gasket had to be lifted a bit with an ultraclean fingertip just prior to putting the flange in place, of course. At first, the rotatable flange made this hard, because the shutter and bellows flange would be good, but then something would shift. However, it soon made it easier. The bellows flange could be lightly put in place, with loose top and bottom bolts. Then the bellows could be rotated a bit to align the shutter rotation. With the flange in place, the shutter rotation was aligned. When it was good, the top and bottom bolts were fully finger-tightened, and other bolts were put in, also to fingertight. The whole while, the shutter was watched to make sure it didn't shift. Once everything was finger tight and aligned, the bolts were tightened down while still watching the shutter, success #5.
With the new shutter installed and new bellows tightened, the next step is to return the In cell to its port, and pump down the system, which we did. We pumped down and thoroughly leak checked all of the bellows and bellow flanges, the HVP and nipple, and the In cell. The chamber pressure went down quickly and is currently in the low -8s. We put a glovebag on the BOS for the Al cell, and prepped the glovebags most of the way for Monday. Overall, today was a big success: we accomplished in-glovebag Ta In shutter replacement.