Monday, 16 May 2011

FMP - Eleven and way overdue (24/03/11 - 16/05/11)

There's a fair bit to get through, so let's get started.

Before Easter (24/03/11 - 08/04/11):

The middle part as seen in the previous post has come and gone. It was originally designed to have the mirrors fixed in place (as they sit behind the cladding and provide additional support) while the rest was filled and finished, having the mirrors masked up:



But when unmasking, and to my annoyance, the mirrors were scratched. It is near impossible to get them out of mirrors too, so I rebuilt the entire thing but this time leaving a slot in the top so that I could slide the mirrors into place when the rest was finished. The two together:



By itself with the slots on top visible:


Along with this I finished the rotating pieces, having both cladded and primed, ready to be sprayed. When I assembled what I had, I noticed the bottom rotating piece wasn't rotating as smoothly as I'd like, so I built a thrust bearing (using marbles) which sits underneath it. The bearing itself is based off the ones seen in the previous post, just with a much larger diameter in order for it to sit on the base board:


What the entire model looked like before I left for Easter:


The biggest change to my project was that during my mock interview towards the end of term, the interviewer mentioned the lack of rapid prototyping models in my portfolio. This is the area of modelmaking I'd hope to get into, and if I have nothing to show for it other than my aspirations, I'm not going to get too far. As a result I changed my brief somewhat, from having the top façade of the carousel sculpted to building it digitally and then 3D printed/CNC'd.

Easter holidays (09/04/11 - 25/04/11):

With the workshop shut, there wasn't much I could do to my existing pieces other than get them ready (as spraying outside does not work too well) for finishing. So I got on with the façade and after altering my initial design (which was going to be sculpted) I began the digital model, starting with the legs. To keep with the organic Seuss design of things, I made everything using curves, which were drawn onto surfaces (which matched that of the object):


To make it solid I added a circle to the top and bottom and used the vertical curves to split it, so that I could use the 'Sweep 2 rails' tool:




This way of going about it was used extensively for the rest, but smaller pieces such as the ribs and base of the legs were made using the 'Revolve' surface tool and then had their control points pulled around to make it asymmetric:



In order for the legs to be 3D printed they needed to be hollow, so all intersecting surfaces were trimmed:


Once one leg was done, another 2 were made using the same base curves, just with their command points altered so that they are not similar, these were then copied to bring the total number to 5. The actual roof was done in a somewhat similar manner, outline curves were drawn (one smaller next to one larger to create a dip) and then arrayed around the overall diameter. These curves where then in turn moved slightly, re-scaled so that once again they'll look asymmetric:


The intersecting horizontal curves were put in place to give a more organic shape that highlighted the dip between higher neighbouring curves. Just like the legs, the sweep tool was used intensively:


Once completed the interior was designed to house a mechanism (not 3D printed, but lasercut) which would counter-rotate the above façade using the main shaft. This would then allow this additional piece to sit on top of the roof you see above and below and rotate counter to the carousel. Sitting on this piece, which was going to be designed as a bent down Circus tent top was to be a sculpted Horton (an elephant, whose weight had caused the top to bend back on itself) and on his back, a Sneetch playing a three-horned trumpet. But as this was going to be sculpted, it wasn't made digitally, but the rest of the façade was so I made it one gigantic piece:


The company whom I was going to get it printed with were Ogle Models (who were my chosen company for the mock interview), and after a tour of their workplace, we got to my model. After initial praise of doing the first STL. made using Rhino which needed no fixing, came the hard truth that the price would easily be over £2,000, and that was for materials alone. I was willing to go up to £1,500, but I was completely unrealistic, as the Nylon resin used (of which 70% is wasted) costs £70/Kg. And being that my model would be printed alone, there was no one else to share the cost with, as other pieces could've been put around it.

Needless to say I couldn't afford it, so I had to opt for CNC'ing it once I got back to Uni, so I re-designed it once again to match its requirements (maximum height of 85mm), plus with time running out, simplified it by removing the rotating top:


Start of term up till now (past the deadline and all; 26/04/11 - 17/04/11):

Well a lot has happened, the biggest of which; work re-started on the characters; which I messed around with before the holidays but having to sculpt directly onto the acrylic poles I needed to use milliput as it needs no heat to harden, and I hate milliput. As I tried to sculpt, the figures kept sliding down the poles, so I scanned in the character outlines and lasercut them with offset outlines either side which were then glued to the poles:


These were then sculpted onto in layers:


With additional detail coming last:


Come deadline week I had a good streak of bad luck, for one the leg (just one leg from the above digital model) that I had 3D printed and planned to cast, broke too many times to be fixed. Next, a piece of wood in the spray room fell straight onto the bottom rotating piece breaking a leg clean off. Wouldn't be much of a big deal, but building a new one and re-attaching it means I need to spray the entire thing again as I've run out of the paint. Anyways, here's what it looked like on the day:


Not proud of it, but it works. In the meantime (and assuming I haven't been referred) I'm quietly getting on with it, such as getting the roof CNC'd (it's done in three layers, each one under 85mm):



As well as getting a new leg 3D printed:



To prevent the same trouble happening again, I literally flooded this one in resin, its previous brother I went a bit too light with and is the reason it broke so many times:


Along with this I'm currently sanding down the characters to get them casted, I'm aiming to be done by the end of next week if I can get on the CNC again.

Wednesday, 23 March 2011

FMP - Ten (17/03/11 - 23/03/11)

First off, the letters are finished; they had their big night yesterday and I was working right up to the deadline. I had all the letters made by Friday afternoon so the start of this week was focused on finishing them. There was a bit of trouble with the finishing as I was originally going to spray them, but with their large size and there being seven of them I'd have used far too much thinners and paint to justify it when they're going to be drawn over anyway. As a substitute they were just going to be painted with white emulsion, but with time against me I didn't have all of them painted by yesterday afternoon when the workshop closed. So I had to take the letters to the event, on the way picking up some paint as well, and paint them in the back with a few of the organisers as well as several of the charity members. Needless to say they were done on time and put on the tables during the event and within seconds people were drawing all over them. I didn't take any photos last night but I made sure the event photographer (Jade Evans) took some. All I had on my phone was the finished letters un-painted:


With the letters done and just over two weeks remaining I jumped straight back into the carousel. I spent the better part of the day finishing the base off as well as trying out the bearings (which are three part ones). The bearing sits in the middle part, which will be cladded with mirror styrene, on top of the acrylic rod as seen in the photos:


It sits about 2mm above the middle section so the actual rotating parts have nothing to collide with. I was originally going to use felt between the acrylic layers, to lessen friction between them but after taking advice from a fellow course mate I tried out bearings. The bearing now takes all the weight of the rotating parts and is completely out of contact with any static piece. 

Tomorrow I plan on finishing the base, giving it its final coat as well as the front cladding. Along with that getting the rotating pieces fully assembled, so that I can start building the above structure. The biggest task, and no doubt the most important is the sculpting of the characters, with time ticking down I've been thinking about CNC machining them, but with there still being a large queue for it I may just have to stick with my hands.

Wednesday, 16 March 2011

FMP - Nine (16/03/11)

Along with my carousel I'm also building an exhibition model for the 'This is an Adventure' event next Tuesday (22/03/11). For the last event back in February, titled 'Up in the Air', I made a Zeppelin which was hung from the ceiling, with the event title engraved into the bottom side of the cabin. These two projects, which are both live and showcased in a public scenario  make up the other half of my brief, which was originally going to be an architectural model.

Moving onto the model, while it has nothing to do with the event itself, a local charity against human trafficking approached the organisers at the last event and asked if they could have something to show this time round. They wanted the word "FREEDOM" in large letters, with each letter being individual mobile so that they can be placed on different tables and wrote on by the public

The font is Nevis and was chosen by the organisers of the event who also came to the agreement of each letter being 450mm high with a depth of 250mm. Once I had the font, I just scaled it up in Rhino and exported it so that it could be lasered.

I managed to get three letters done today, F, one E and M. Each started off with the front and back panels threaded with dowel cut to 250mm:


They were originally going to be cladded with styrene but that would use up a lot, so I opted for 3mm MDF and 1.5mm plywood for curves. This of course boosted the overall height to 456mm, but that isn't a problem, considering the wooden cladding makes it overall stronger than styrene would have:


For letters, such as the F, which are either top heavy or will have a problem standing up, weighted metal rod was threaded parallel with the dowel:


Here's what they look like so far, of course they still need sanding, filling and spraying:


Here's a link to the Thisis site where more information is available about the event:

Tuesday, 15 March 2011

FMP - Eight (10/03/11 - 15/03/11)

To be honest, progress has slowed down a bit over the past 2 weeks in getting the base finished, but at least I can proudly say that all the work and time has paid off, as the CAM works! It's been sanded down to create a smooth track, so of course I assembled the bits I have to give it a test run. I didn't attach the motor, so it was rotated by hand, but nonetheless it worked. The masking tape represents the different height changes as it's rotated:



As you can probably tell, the rod going round isn't one of the ones I've made. I tried the ones I made on it last week (before the CAM was probably sanded down) and it snapped the ends off, probably due to catching on some protruding filler:

As I mentioned this was last weeks attempt, since having sanded down the CAM, the ones that still have the ends work fine. But this problem has highlighted the issue for me, so even if I stick with the tyre idea there's a good chance of them snapping and there's nothing I can do to fix it. So to prevent this I'm going to sand the ends down into half a sphere, just like on my prototyping. This instantly strengthens the end tenfold, and is a lot less fiddly than the tyres. It's worth noting as well that I was only using the tyres when the CAM was only 5mm thick with guides either side so that it couldn't rotate, but now it's almost tripled in width and there are no more guides this idea wont work.

The new design to prevent them from rotating is to simply thread a bar through the poles and have two pieces attached to the base, so that they prevent the pole from rotating, this picture should clear that up a bit:



To end on a good note, here's a video of it working (keep your eye on the masking tape):

video

Wednesday, 9 March 2011

FMP - Seven (9/03/11)

As promised here are the photos from what I mentioned in yesterdays update. To create a fillet with lasercut sandwiched pieces, the top and bottom layers have to be slightly offset inwards from the outer edge. This is what they start as, with the two thinner pieces sandwiching the larger piece:


Once lasercut and attached to each other, they are covered in filler:


And then finally sanded back to create the fillet and give them an overall more circular appearance:


The larger piece is what attaches the lower moving base to the top, they slot into each part:


Moving onto the new CAM, here it is:



As you can see it still needs to be sanded down and filled, but it has fixed the alignment problem thankfully.

Tuesday, 8 March 2011

FMP - Six (4/03/11 - 8/03/11)

A lot has happened since the last update, on the one hand I got the carousel working at the end of last week, but then on the other there's been a serious alignment problem.

To get the bad news out of the way, somehow, even though I used templates straight from my digital model, the centre MDF piece is lopsided, and as a result this has had a knock on effect of everything which surrounds it, most importantly the CAM. The now realised off centre CAMs and MDF piece:



As it is lopsided, the above slot for the vertical poles to slide down doesn't accurately follow the CAM, and in other places falls off it. Being only 6mm wide, there isn't enough room for this leeway, so I've had to extend it's width by sandwiching the original CAM with two 5mm lasercut and then heatbent acrylic pieces, I've yet to get a photo of this, but I will for the next update. 

The problem doesn't end there, as the original CAM was only 6mm wide and sandwiched between two higher matching styrene profiles, to create a trench, I could use circular vertical poles, as the wheel they'd be attached to would prevent it from going on its side (a massive problem with my mock up). But now the width has been extended to 16mm, this prevention is useless, so I've had to redesign them so that they are now more square thus unable to rotate. This does have its upsides though as due to the cartoonish nature of the carousel, these poles were going to be heat bent, and for the wheels to be attached they'd have to be milled at one edge. But now these poles are made up of a 3mm profile sandwiched between 1mm thinner profiles, which will be filled so that it has somewhat rounded corners. Its cutout in the base is slightly larger than this, but due to the square sides and edges it prevents it from rotating. Photos to better describe this fix will be uploaded tomorrow.

Moving onto the good news, getting it to work. This started with lathing the aluminium bar to create a hole in one edge for the rod from the motor to slide up to. A perpendicular hole was then drilled to penetrate this hole, and this hole was threaded in turn so that a screw could be fitted. The rod from the motor has one flat side (which doesn't show up too well in the photos) which is meant for an external screw to go up against. All in all this prevents the bar from freely rotating while attached.

The bar attached to the motor:


The hole in the edge where the rod slots up:



The blurry flat edge of this rod:



In celebration of it working, and the top piece successfully rotating the bottom one I recorded a short video. Even though this was the moment that highlighted the alignment issue, it is still worth noting.

video