Structure


 * You are also provided with an eating plan, and calculated serving sizes that will ensure you are eating the right proportions, and getting the right amount of carbs and calories. [|the diet solution program]Structure**

Welcome to the Structure page! This page is dedicated to the advancements of our PTM device's structure. The Structure group is led by Kevin Dickey, with the help of Jarrett Almand, Dashiell Kopp, and Carl Stahoviak.

Welcome to Phase I of the Imaging Project! With the Winter Break over, the Structure group has hit the ground running in Week 4. Prior to the break, we ordered the Phase I dome - an acrylic hemisphere measuring 24 inches in diameter. Here are the specifications of our new dome, named "Sunny Side" (for its strange appearance).
 * Progress (1/4/11)**


 * Diameter: 24 inches
 * Distance from apex to base: 12 inches
 * Material: acrylic
 * Weight: approximately 15 pounds
 * Thickness at base: 3/8"
 * Thickness at apex: 1/8"
 * Flange length: 4"



One of the more unique aspects of this structure is its thick flanged base. While other hemispheres are designed to be merely hemispherical, this custom hemisphere has a thick (3/8") base that extends 4" from the outside of the hemisphere, parallel to the surface that the hemisphere is resting on. This allows for easier positioning and mounting of camera supports, creating a single, static structure that incorporates both the imaging device and lights.

Full Structure CAD Drawing: (Link) Acrylic Dome
 * Structure Process CAD Drawing:** (Link) FULL PHASE Drawing

Additionally, we have been requested by our advisor to defend our decision to buy this hardware. Arguments are listed below:
 * This dome, a custom piece, is especially thick (3/8" at the base). Some other companies produce stock hemispheres with a set thickness of 1/4" at the base, but those can be as thin as 0.07" at the apex. This decreased thickness could lead to problems during drilling and hole-cutting - mainly, an increased risk of cracking or breaking the hemisphere. With a thickness of greater than 1/8" (0.125") at the apex, this dome should reduce that risk significantly.
 * While determining a supplier for this dome, I asked manufacturers if their structures would be sufficient to support 15 pounds of mounting materials, as well as being strong enough to drill without bending or cracking. Suppliers generally responded that the stock thickness of 1/4" was too thin to support the camera and drill without worries of warping and cracking, leading me towards a thicker dome.
 * As a custom piece, this dome was manufactured with a large (4") flange. This flange is, like the base of the dome, 3/8" thick. This flange is significantly thicker than some other hemispheres, and will serve as a stable mounting area for the dome's imaging system. With its greater thickness, this dome is less prone to warping, and it should prove to be a stable base on which to mount camera mounting brackets later in the project.
 * Using this dome as a structure for mounting both the lights and camera, we can eliminate any errors introduced by camera misalignment in relation to the dome's lighting system. A sturdy base structure for introducing the imaging and lighting systems will ensure that these errors are kept to a minimum.

With these arguments in mind, we feel confident that this structure is the right one for our intended application. It is highly durable, providing a sturdy and flat surface for securing camera mounting brackets. The increased thickness of the hemisphere will allow us to drill holes for lights and our imaging device, without raising concerns of cracking or damaging the structure.



**Update: March 21, 2011**
Time has certainly flown by! The last few months have been busy ones for the Structure group. Due to a variety of challenges, we've been a little slow to complete work on the dome, but we are well underway now. An aluminum frame has been fabricated, using some stock aluminum from 80/20 Inc. This structure is extremely strong and durable, and will serve to mount the hemisphere and camera.
 * = [[image:ife2010/DSC_0040.JPG width="800" height="531"]] ||
 * = //The new aluminum structure and black dome, with markings for lights.// ||

On the left side of the picture is the new structure. The hemisphere will be mounted to the first row of horizontal rails on the aluminum structure, and a moving tray will allow users to position objects underneath the dome. The tray is still being constructed - linear motion pieces have arrived, and we're just waiting for a tray piece to connect all four corners. This tray must be very rigid and not too heavy, so we may use a sheet of thicker aluminum. As previously noted, we are still working on our tray piece.

On the right side of the picture is the dome! We have not drilled our holes for the lights yet - we plan on doing that in Week 3 of Spring Quarter - but we have given it a fantastic coat of black paint. While it looks reflective on the outside, it is very dull and non-reflective on the inside - this is because we did not paint the outside of the dome. The interior was sprayed with a matte black acrylic spray paint; the shine on the outside surface is from the shiny clear acrylic on the outside of the dome.

In addition to painting the dome, we have begun preparing it to be cut and drilled. Using a few tablespoons of water, we found the center point of the dome. By turning the hemisphere upside-down and keeping it level, a few drops of water settled at the lowest point - the center - and we drilled an approximation of center there.

We also marked the positions of our lights. A variety of outside sources suggest that the best PTM lighting angles are between 15 and 60 degrees, so calculations were done based on that. We split our 24 lights into three "rows" - one at 20 degrees, one at 35 degrees, and one at 55 degrees. Since the most important illumination for PTMs is raking illumination (low-angle light), and also because the diameter at lower angles is larger, we biased the lights towards low angles. The first tier (20 degrees) had 10 lights, with 8 lights on the second tier (35 degrees) and 6 lights on the third tier (55 degrees). Calculations were done using simple trigonometry and arc-length calculations - we first determined the circumference of the hemisphere at a given point on the outside arc, then divided that by the number of lights at that tier. That gave us an arc-length approximation of spacing between lights at each tier.

Since the dome is black, points were marked using a dot of white-out. A compass was then traced around the circumference of the light at each dot to mark the hole for the light. This will make drilling much easier later in the week. Drilling of the dome presents some interesting challenges. The most prominent challenge of drilling this particular setup is that all holes must be drilled normal to the surface of the dome. On a flat surface, that would be easy - but this is a continually curving surface. Our best solution is to use the concept of a "dome on a dome". Using small hemispheres acquired from the Munsell Color Science labs in building 18, we could drill accurate holes. If we drill directly through the center of the apex of the small dome, we can use that as a guide for drilling all of our holes in the large dome. A hemisphere on a hemisphere will always align such that the apex of the small hemisphere is parallel to the surface tangent at any given point.
 * [[image:ife2010/DSC_0041.JPG width="800" height="531" align="center"]] ||
 * = //A white-out mark denoting the center of a light. A compass was used to mark the circumference of the hole, and was traced with a Sharpie to make the circumference more visible. This will make drilling much easier.// ||

Once we have taken care of that, we have to tackle one final issue - camera mounting. This should be dealt with fairly easily. We have ordered materials to help mount a smaller piece of aluminum as a cross-member at the top of the structure. A special v-shaped piece that fits into the slotted aluminum piece is the key to this setup. It can be tightened in place at any point on the aluminum cross member using a simple hex wrench, and it has a threaded hole for a screw that will also fit into the camera on the other side. To avoid any issues with mounting, we will be looking for a quick-release camera mount so that the camera can be quickly removed from the system. Thus, we can create a highly stable and durable mount for the camera that is also extremely adaptable to the dimensions of different cameras and lenses.

Update: April 3, 2011
All twenty-four holes have been drilled in the hemisphere, which is now ready to accept our lights (see the Illumination team's page for more details). A system for securing these lights is under development, though we are leaning towards a simple friction seal between the light and its corresponding hole. A wrap of duct tape around the base of any given light is enough to securely place it in any hole in the hemisphere. More details on light mounting will come later.

We have been working towards a solution for placing objects in our PTM dome. Our initial concept was to use a sliding metal tray that users could raise, lower, and lock in place for creating PTMs. We picked up a set of linear motion bearings from 80/20 Inc, which have been assembled and installed at the base of our aluminum structure. Attached to our linear motion pieces are small (2 inches in length) vertical frame members, which will support the outer edges of a large, rigid tray. These smaller frame members were tapped with 5/16-18 holes last week and can now be easily bolted to our tray piece.

The tray brings us to an interesting crossroads of ideas. Our original concept was to use aluminum plate for our tray. It is strong, relatively light, and doesn't bend much. Torsional rigidity is key here because the tray must be large (roughly 27" by 27") and all four corners must move at the same time - otherwise the linear bearings will catch and bind. Aluminum plate seemed like an excellent solution to start with, until we considered its weight. A .25-inch thick plate of T6061 aluminum large enough to cover the necessary area would weigh //seventeen pounds//. This excessive weight made it unsuitable in this application. We were thus forced to reconsider our options. It is possible to create an aluminum frame to support a lightweight tray at the base of the structure, but this too will come at a weight penalty. One particularly interesting solution was to use marine plywood. This wood is treated to be resistant to changes in humidity and ambient conditions, while also being strong and fairly light. A 1/2-inch slab of marine plywood will weigh only 6 pounds, whilst also having high torsional rigidity. A decision about this tray material will be complete by the end of the week (before Friday, April 8) so that materials can be shipped for next week.

This week we were able to finish our camera mount! We started with a Manfrotto 323 quick-release plate (seen here), modifying it so that we could fit it into our system. After running into a variety of problems with the dimensions of the bolts and threads of the structural implements (i.e. the bolt thread extended too far towards the bolt head, and the bolt head was too large to fit flush with the base of the quick-release plate), we machined a custom bolt to fit properly into the system. With this new camera mount in place, we can now quickly and easily mount our camera to our structure. This will ensure a proper camera/lights alignment every time we need to remove and re-mount the camera.


 * = [[image:ife2010/DSC_0117L.jpg width="800" height="531" align="center"]] ||
 * = //Sparks fly as the Structure group machines a bolt to fit the exacting specifications of the Manfrotto quick-release adapter.// ||

With the camera mount finally finished, the Structure group's work with the Camera group is done for the Phase I device. Now it's time to focus on that tray piece, with more updates to come about that sub-project as the week progresses.
 * = [[image:ife2010/p_00230.jpg width="800" height="600"]] ||
 * = //The new camera mount, supporting our D50 over the drilled hemisphere.// ||

Update: April 10, 2011
This week has been eventful, with the marking of lights, as stated above. We also drilled mounting holes in the flange at the base of our hemisphere, so that we can bolt the hemisphere directly to our aluminum structure. This will not only ensure that the lights will not move during shooting sessions, but also that the system can be easily disassembled and reassembled for easy transportation. Our linear motion pieces from 80/20 have arrived, and we're putting together a test system right now to test the pieces in a tray mechanism which can be raised and lowered under the hemisphere. More on these bearings later.

Update: April 18, 2011
The lights have now been mounted in the hemisphere. Since the hole saw used to drill the holes was a near-perfect fit for the lights, all we had to do was widen the base of the lights to secure them in the holes. This was done simply by wrapping a small strip of masking tape around the outside of each light, which was enough to secure the light tightly in place. With these lights in place, the Electrical team was able to take care of wiring the lights into our controller board.

Speaking of the controller, that has been mounted, too. The original mounting holes on the controller were designed for 1/4-20 screws; we widened them to fit the 5/16-18 screws that slotted into the top of the frame. With three points of contact, the controller is sturdy and secure, and should be able to easily withstand continuous daily use. At the top of the frame, it is out easy to access if any electrical troubles arise.

We've also been working with the tray piece this week - more specifically, the linear motion pieces that the tray rests on. Initially, these pieces did not move smoothly along their tracks, catching and binding occasionally, even when lifted from their exact center. This was a problem, as the tray needed to be able to slide up and down smoothly and easily. To counter this, we applied a silicone-based dry lubricant to the linear motion pieces, as well as the tracks that they moved on. This solved our problems, and the linear motion pieces moved smoothly through their full range of motion.

We've decided that the best solution for the tray itself is to keep it as simple as possible - use a mildly thick piece of regular plywood. We intend to build an I-shaped support system that holds the plywood tray from underneath, preventing it from bending or warping over time. A crossbar under the tray would keep rigidity up whilst ensuring low weight, one of the goals of this system. This piece will be cut and installed next week.

One of the biggest issues still facing us is how to raise and lower the tray assembly. While the crossbar and plywood setup will decrease the overall weight of the assembly in comparison to an aluminum plate, it will still be difficult for a single person to operate. There are a variety of plausible options for the movement of this device - right now we're looking at scissor jacks, bottle jacks, pulleys, and lab jacks. While scissor jacks and bottle jacks are good options, they are designed around lifting much heavier things - cars, for instance. Some bottle jacks can lift locomotives. If they are properly implemented though, these would be good options. They are durable, reliable, and easy to use by one person. Their only downside is that we may risk damaging the system if they are implemented poorly. Pulleys may also work, though it would require two sets of pulleys to lift the tray assembly - something which seems overly complicated to us. Lab jacks seem like the best solutions right now - we can find one with a large range of motion - perhaps 12 inches - that should fit our needs well. They are designed to handle much smaller loads - usually a maximum of 100lbs - and are easy to operate. Motorization is also an option here, using a modified drill to turn the screw that operates the jack.

Update: April 25, 2011.
The system has been progressing nicely! Our lab jack has arrived, and additional aluminum pieces have also arrived to complete the tray mechanism. We were quick to cut our aluminum and mount it to the tray assembly, so we can have at least the skeleton of our tray mechanism operational.

With the other members of the Structure team working on other projects - Phase II structural implements, Phase I tray - I (Kevin) set out to work on motorizing the tray mechanism. I have seen videos and articles from others who have modified drills to operate scissor jacks without too much difficulty, so I tried the process for myself. I started with a used cordless drill and a socket extension for a wrench. In order to be able to operate on its own power all day, the drill had to be converted from cordless to corded - something which caused me a little trouble. I drained the battery, popped the drill housing open, and cut the wires to the battery. After connecting the leads and grounds between the supplied charger and the drill, it was soon apparent that the AC adapter was merely a trickle charger, and could not supply near enough power for the drill to function. In fact, it couldn't even get the drill to //turn//. Our local Goodwill had a wide variety of AC adapters; I picked the highest amperage one I could find (1500mA), cut the wires on the end, and re-soldered it to the lead and ground wires of the drill. Success! The drill ran off the power from the new adapter.

Two of us focused on modifying the socket extension into a drill bit - we bored out the socket hole in the front to make it circular, then tapped that hole to fit the size of the screw on the lab jack. After cutting of the excess 7 inches of the extension rod, we had a functioning custom adapter for our lab jack. From there, we had to mount the drill to the jack, a task which turned out to be much more difficult than it sounded. The screw that operates the lab jack travels in two directions - x and y - over the course of its travel. In order to operate that with the drill, the drill must also be able to move in the x and y directions. A mount was devised using a 1/4-inch piece of flat metal that was attached to the same supports in the jack that the screw ran through. This way, it would move in the y direction with the screw. To compensate for x-direction movement, we devised a simple linear motion system using leftover aluminum from the rest of the structure, which could move in and out in alignment with the main screw.

With the motorization of the tray mechanism complete, we focused back on our tray piece. We sized and cut the plywood for the tray, and proceeded to drill a set of twelve holes on the outer edges of the plywood for simplified mounting of the tray to the support system. To provide a soft, diffuse surface for imaged objects to rest on, we employed a novel approach - mouse pads. With an abundance of these objects here at RIT, we can use the bottom of the mouse pad to create a diffuse surface that is forgiving to fragile objects being imaged.

As of today, the tray mechanism is complete, as well as the motorization mechanism. The tray is ready to be mounted, a project which will have to be saved for later in the week.

This is the email I received from Bill Ambrisco, referred to me by Tom Malzbender. Bill has a lot of practical experience with designing and building PTM structures.
Hi Tom, hi Dashiell I've included some of pics of PTM/RTI lighting structures that I've designed and built or modeled. On my experimental models I focused on portability and flexibility, with forensics and field work of various sorts in mind. Of note is the open frame design rather than the fully closed hemisphere. This was a result of the designs I did for Mark Mudge at Cultural Heritage Imaging ( www.c-h-i.org ). While I started out with radial arms cut from carbon fiber/foamcore composit material, we found that bending the arms from 1/8" x 3/4" flat aluminum extrusion was __**much**__ less expensive and complex, providing we tied them together at the equator for rigidity. We also had them black anodized after fabbing them.

The BIG ARM setup had a single arm with 12 flash units that rotated around the camera opening at the north pole. Although I had it counterbalanced, it was too wobbly for good registration.

The large tripod in some of the pics is from a telescope I had.

I haven't built any of the design you describe, but it should be simple enough to build if it gives you the info you're after. You're welcome to use anything of interest in the pics and I'd be glad to answer any questions you might have.

Bill Ambrisco





=__**Phase II**__=


 * As I hope you all know, we are nearing the end of our project on a timeline perspective. However, the end still seems so far away.**

Phase II must be complete within the next seven weeks of class. I can only speak for myself regarding this matter, but I feel that is not a lot of time. Just some food for thought, similar o what Joe had mentioned the one day, I think we should possibly consider more of a proof-of-concept approach to phase II as opposed to our hopefully fully functioning phase I. I'm not sure how realistic it is to finish phase I, start and finish a phase II, and prepare our set-up and plans for Imagine RIT in seven weeks. Hopefully, we can discuss this in class and get some feedback from everyone.

In the meantime, here are two designs I drew up depicting two options we have for phase II based on some quick brainstorming in class. The way I see it, we have three main paths from which we can choose to go down with our next phase. The first, is an aerial mounted, rotating single-arm, counterbalanced approach. This offers great accessibility but possibly some major engineering hurdles, especially with the camera needing to stay fixed while the arm rotate 360 degrees. The second is a rotating arch which rotates 180 degrees and back. This also provides good accessibility but may pose problems with the loss of lighting angles when the arch passes in front of the camera. A variation from this that was suggested was using a printer scanner instead of fixed lights. The third option is to simply stay with the dome and adapt it to house new light sources or other manipulations. I'm sure there are plenty of ideas and variations to these, but these are kinda what I was thinking for now. No idea is a bad one so try to come up with something for class when we brainstorm.

by: Jarrett Almand
 * Phase II Requirements:**

Illumination Group: -LED's with In-fared Light -No use in using UV light due the camera not being able to pick up most of the light reflecting off.

Camera Group: -Camera needs to be 264mm away from the ground and needs the proper mount on a rigid structure.

Structure Group: -Lights must be equidistant from the subject. -The structure needs to be quick and flexible to other requirements.

Programming Group: -Camera needs to be compatible with Lab-view.


 * Phase II Proposal Designs:**
 * by Jarrett Almand**

Powerpoint info on the Choices: Phase II Proposal Sketches: -Arm and Half Arc: