First Rotations

With all the superstructure built up, there’s much less visual progress being made on the mount. We’re ordering lots of parts and waiting for their delivery so that we can begin work on other components like the drive system and the routing of gas hoses. In the meantime, we got a few videos of the mount in action.

The first video shows the elevation stage moving. As the post about its installation described, this stage moves in a nodding-like motion with a full range of 55 degrees (including 10 degrees past vertical).

The second video shows one of our rotary unions in action. This union is attached between our stationary structure and the Azimuth ring (rotating in the above video). As intended, the bottom portion of the joint remains fixed during this rotation which gives us a stationary interface to which we can attach our pressurized hoses. The upper half of the joint then rotates along with the Azimuth structure, allowing the hoses to move without getting twisted up.

Progress on the mount is slowing down significantly now that we have the steel superstructure assembled, however there’s plenty more happening at Minnesota that will soon make its way online.


Installation of the Theta stage and Rotary Unions

After we installed the Elevation stage early in the week we were ready to install the final ring which we refer to as “Theta”. Unlike the Azimuth and Elevation stages which change the direction that the entire array points, the Theta stage rotates all four receivers about their common center known as the “bore sight” (the red dashed line in the below image).BA_Assembly_IsoSection_20180417.JPG

Bicep Array observes polarized light, and its detectors are sensitive to either light that is polarized “vertically” or “horizontally”.  Rotating about the bore sight changes the angle of our detectors (redefining “horizontal” and “vertical”) without changing where they are pointing on the sky. Observing at a few select rotations of this third axis helps suppress systematic errors due to the shape of our beams or differences in response from detectors.

Unlike Azimuth and Elevation the Theta stage didn’t need to be built up at all before it was installed, so it went up pretty quickly. As evident in the right hand image above, this ring actually slides downwards into the elevation stage, so we had to align things very carefully. The middle flange of this ring attaches to the other half of the second ring bearing and will be driven by the two gears visible on the top of the elevation stage in that same image.

After the ring was installed, we attached the top frame. The central piece of this frame has four openings which provide holes for the receivers to sit in, and hoist points where we can mount the pulleys that will hoist the receivers up into the mount. In the right hand image above, extensions to this frame are being attached. The outside of these extensions provide the second attachment point for the environmental boot (transparent teal in the first image of this post) and a base on which to mount the top panel that closes out the internal mount space.


With just about all the large pieces attached the mount is impressively tall, though it still looks a bit bare without everything that goes inside it. Next up was installing two of these components, the rotary unions.

As mentioned in a previous post, these rotary unions provide channels for pressurized Helium and Nitrogen, along with electrical and data connections. Since the upper and lower halves rotate separately, we can use these connect our rotating stages. Using these eliminates the need for a cable wrap that winds and unwinds as the mount rotates and gives us continuous rotation about the Azimuth and Theta axes. In the above image, the red caps indicate the locations of the channels for pressurized gas, while the electrical and data connections are indicated by wires poking out the top and bottom ends.

With the rotary unions installed we began testing the motion of the mount, driving it a bit in each axis and adjusting the gears as needed. This doesn’t produce much of a noticeable change when rotating in Azimuth or Theta but everything looks a bit different when we rotate in Elevation. The entire top of the mount tips forwards and it gets a bit taller. The internal structure changes pretty significantly as well, and the rotary unions move fairly far apart as seen in the right hand image above.

With the steel structure assembled we can now turn our attention to the other components of the mount: like the drive assembly and controllers, cable and hose routing, and receiver and electronics installation.

Installing the Elevation Stage

The end of last week was mostly preparation for the beginning of this week, so I decided to wait until this week for the next post. After getting the first ring bearing and the Azimuth stage installed, we went to work getting the Elevation stage ready to attach. This stage gives us our second axis of rotation, nodding the entire telescope up and down.

The first thing we did was to unpack and attach our second ring bearing. This bearing attaches to the top of the elevation stage and provides the mounting point for the final ring which we simply refer to as “Theta”. As with the first bearing, we inspected the ledges which needed to fit together and did a little grinding away at the paint to make everything match up nicely, then picked the bearing up. Installation was a bit harder since the elevation ring it installed onto was still sitting on it’s shipping support at a 45 degree angle. IMG_20180928_102218

Next we needed to prepare a cage-like structure which sits underneath the elevation ring. This frame helps support the elevation ring and keep it from flexing. But because it sits on the bottom side of the elevation ring, we needed to attach it to the ring before we attached the ring to the mount.

Once we had the frame built up, we lifted off the elevation ring, set it down on top of the frame, and bolted the two together.  We were then ready to lift this new substructure upwards and attach it to the rest of the assembled mount.

The pictures above show a few different views of the elevation stage. On the left, one of the two gear reducers is visible and the large gear can be seen meshing with the outer half of the second ring bearing. Although the bearing is identical to the first one, this bearing operates slightly differently. Rather than the toothed half being fixed in place, for this bearing it is the free half. In this orientation then, the gears drive the outer half of the bearing and will cause the Theta stage (not pictured) to rotate. The right hand picture shows one of the two bearings that will enable the telescope to rotate in elevation (think “nodding” motions”). These are extremely resilient bearings, supporting over 10 Tons between the two of them.

But one might ask how this addition integrates with the parts we’ve already built up. There are two main interfaces between the elevation stage and the Azimuth stage. You might remember from the previous post that the Azimuth stage has two large triangular sections sticking up off the main ring. The tops of these triangles are flattened and provide an attachment place for the two bearings. The alignment of these bearings needs to be pretty precise because their orientation affects the second interface, the elevation gear (seen at left). This huge toothed gear connects with the two gear reducers mounted on either side of one of the triangular Azimuth sections (at right still in their protective covering). This huge gear is somewhat like the Elevation stage’s counterpart to the toothed half of the ring bearing. As the motors turn the two gear reducers, they will apply torque on the large gear which will then rotate the entire elevation stage by way of the two bearings on either side.

The next step of course, was installation. We lifted the entire Elevation assembly upwards and brought it over to the rest of the assembled mount structure. We had to go quite high in order for the stiffening frame to clear the rest of the structure. With some careful maneuvering, we then moved the stage downwards and inwards in very small steps in order to align it properly with the Azimuth stage. This was a slow process since we needed to be very careful not to smash against any of the gear teeth that were exposed as we tried to get everything to align.

Once we had everything in nearly the right position, we were able to use the bearings on either side to guide us the rest of the way. Then after a slight bit of maneuvering left and right, we had everything aligned correctly and the gear teeth all mated nicely.

With the elevation stage installed, the mount seems even more like a jungle gym. The cross braces provide an easy way to climb up inside the mount where ladders can’t really reach. Although it seems very open right now, much of this space will be taken up by receivers, electronics, and Helium lines when the mount is fully populated. In the meantime though, it’s a nicely accessible space. As long as you watch your step.

With the elevation stage attached we’ve realized we had to adjust the webcam in order to continue capturing our progress. So if you’ve noticed the webcam shift over the past day, that’s the reason. The top-down shot shows a high bay which is significantly less cluttered than when we first unpacked everything. Next up is our final rotation stage “Theta”. Once that’s installed we’re only a short time away from our first motion tests.

Installing the Azimuth Stage

With the first Rotek bearing unpacked and inspected, the middle of this week was devoted to the installation of the Azimuth stage. First we installed the bearing, raising it up and lowering it down on top of the stationary ring. There is a small ledge on the stationary part of the bearing that mates with a similar ledge machined into the large ring. These two ledges need to be nearly identical in size or the bearing won’t mate correctly. One issue here is the bright blue protective paint. The thickness of the paint layer varies enough to potentially impact the interface, fortunately a little grinding and sanding is enough to make things right.

After the bearing was installed and we’d torqued all the bolts to spec, it was time to assemble the Azimuth ring. This structure breaks down into two halves so that it can fit within the allowed envelope on the LC-130 planes which will deliver everything to the South Pole. The triangles on either side of the ring are the two pivot points for the elevation stage which provides the second axis of rotation for the array. Once the two halves were attached to each other, we lifted it up to measure the ledge on its underside, destined to mate to the other half of the ring bearing. Once that checked out we were ready to attach it to the rest of the assembled structure.  Also visible from the underside is one of the two gears that connect to the ring bearing we installed on the stationary structure.

Lifting this up with the crane was a little intimidating, but exciting. With two gears poking out the bottom side, we had to be very careful in our alignment and in the motion of the crane so that we didn’t accidentally smash any of the gear teeth. After some careful coordination, keeping close eyes on the gears, we set the stage down and it interfaced perfectly. It’s always amazing to see assemblies happen on this scale, the bearing and the ring structure were made separately by different companies, but still went together without problems.

From the last picture above, one of the gears can be seen interfacing with the outer ring of the bearing. This half of the bearing remains stationary while the inner half (which supports the weight of the Azimuth stage) is movable. The gears then drive the stage around the outer ring. Above the gears but out of sight in the above picture are the two gear reducer boxes. As the name suggests these boxes contain a series of gears that together provide a torque multiplier for our driving motors, allowing us to turn the entire stage with smaller motors. The Keck Array mount (which this mount will replace) can actually be turned easily by hand using a standard Allen wrench because of a similar gear reducer setup. The two boxes can be seen from the topside of the ring (see below).


Once the Azimuth ring was attached, we also attached a small spaceframe to its outside. This frame will provide the lower attachment point for what we refer to as the “environmental boot” which helps shield the inside of the mount and all it’s electrical components. When deployed, the center of the telescope mount will be kept open to the tower on which it sits. By enclosing the space with a fabric shield (the boot) we effectively extend the building atmosphere (and temperature) into the inside of the mount where many of our electronics crates and the telescopes themselves will be located.

On the left hand side, the upper image shows the existing Keck Array, and it’s environmental boot (green on the side with a white panel in front). The see-through teal-ish accordion structure in the lower image is the Bicep Array environmental boot. As evidenced by the scale of the person in the upper image, the enclosed volume is pretty big, though much of it is taken up by electronics crates, cabling, steel beams, and other components. On the right, the bottom frame of the boot has been attached to the Azimuth stage of the Bicep Array mount.

One more small piece needed to be attached to the Azimuth stage before we moved on to the elevation stage. The first of two “X” shaped pieces. This brace has a hole in the center surrounded by a small bolt hole pattern. It sits just above a corresponding hole attached to the stationary ring.

These two central fixtures will provide the two end mounts for one of two rotary unions  (made by DSTI) that will be integrated into the mount. These rotary unions pass pressurized gasses (Helium and Nitrogen) along with electronic signals and data channels between the rotating stages without the need for any kind of cable wrap. The lower end of the union remains fixed in place while the upper end rotates along with the stage it is attached to. Using two of these unions will allow this mount to rotate continuously, no wrapping / unwrapping of cables required!


Above is a picture of one of these rotary unions with the red caps showing the Helium channels and the wired showing the electric and data channels. It’s going to be very exciting when we finally get to install one of these and watch it in action.

Tower setup and arrival of the ring bearings

After the rest of the steel structure was delivered last week, we began assembly. The first thing to come together was the triangular base section. This integrates into the top of the building at the south pole and serves as the anchor point for the rest of the telescope mount. At the South Pole, this triangle structure will replace the existing one that supports the Keck Array telescope mount.



The reddish blocks visible on the top side of this triangular structure are adjustable feet we can use for leveling.  Now since this is the bottom part of the telescope mount, we need to raise it up so that we can access the underside and load receivers into it. Next up then was setting up the three pillars that we call the “tower simulator”. These pillars won’t be shipped down to the South Pole, but they replicate the building structure to which the rest of the mount attaches. Once we had them laid out, the triangle was lifted up (about 12 feet) and attached to the top.



After we got the triangle attached, it was time to attach the first ring structure. This first ring serves mostly as a support, giving us a stable and adjustable structure to which we can attach the moving parts of the mount. Since the rings sit at a 45 degree angle on their shipping structures, we had to use a lever hoist in combination with the gantry crane. This let us lift up the ring at an angle then rotate it down to horizontal before we lifted it up onto the tower proper. Then adjusting the six blocks underneath the ring let us level it out. The shiny metal surfaces visible on the top side of the ring are precision machined surfaces coated with an anti-corrosion coating. One of these surfaces will hold the first of two ring bearings that allow the telescope to rotate. The second surface gives us a place to mount an encoder strip which gives us high precision location information for the telescope control systems.  We then had to move the whole assembly backwards in the high bay to make room for the third and final delivery on Monday.


Monday morning saw the arrival of the final truck with the two large diameter ring bearings made by Rotek.  Just as we thought, they were loaded onto the front end of the flatbed truck which meant the truck had to back in quite a ways in order to get the bearings underneath the gantry crane. Even though we moved the structure as far back as we could, we only had about 2 feet of extra room by the time the crates were far enough into the high bay for access.


Two axes of the telescope will rotate on these large diameter ring bearings, which means that they have to support a large amount of weight, multiple tons each. Along with supporting that weight they need to remain very rigid so that the mount doesn’t flex too much during operations. Considering the requirements placed on them they are astonishingly skinny in cross section. Using two hands you can just about grab the entire bearing. The teeth on the outside of the rings will interface with the motors and gears, allowing the outer half of the bearing to rotate while the inner half remains fixed.



For the first part of the week we’ll be working on getting the first bearing installed and attaching our azimuth structure on top of it.