Sunday, 20 February 2011

FMP - Three (18/02/11 - 20/02/11)

As promised I made a similar working model to that in the last update, but with a bar instead of a rod. The same principle applies, but to highlight the problem of only using one CAM I had one of each design along side each other (one CAM and then two CAMS):

As I was using a bar instead of a rod, to achieve the maximum grip between parts, I had to lasercut my own gears. This bar was slotted through acrylic rod which was in turn threaded and glued to the vertical dividing pieces, to strengthen and maintain the bar at an equal height along its length:
Back to the CAMs; the two different CAMs next to each other:

As you can see, the CAM device on the left is comprised of two CAMs; a secondary one fitted between the outer CAMs which freely rotates on a rod between these two outer ones. The vertical pole which is attached to the horse is also attached to the other end of this inner CAM and is again, allowed to freely rotate. So that when the outer CAMs are rotated up the inner one is also brought up, but as it's allowed to rotate freely, the end where the vertical pole is attached can maintain its vertical stance (due to it remaining over its centre point) while being lifted up:

And vice versa when it comes down:

The same can not be said for the other CAM device (the one on the right), which angles the vertical pole while going up and down, and as a result, needs a larger hole to be slotted through. Going up:

And coming down:

This angle at which this vertical pole becomes when completing its journey is obvious when compared to that of the outer CAM device. And as I mentioned before, the idea that the horse goes back on itself when brought up is clear in the following comparison photos. The nearer horse is attached to the two CAM device, with the further horse being attached to the one CAM device:

That is them going up, and it's clear to see the idea of the horse going back on itself.

And coming back down, notice the angle at which the vertical pole is at compared to the nearer one. Now the same comparison but from a different angle:

This has lead me to believe that even though I don't have concrete proof that in actual carousels they use this secondary CAM device, they must incorporate something incredibly similar as to prevent the horse going back on itself, which could possibly slide a rider off. It's also interesting to note that the secondary CAM horse doesn't go nearly as high as the other one.

This design is also titled as the most reliable method of getting a carousel to work and provides the smoothest ride. But from my perspective, via the making of these working models, it's also the most complex and fragile, requiring more engineering than lasercut acrylic pieces. For this reason I'm going with the below CAM design which I used on my first working model.

In other news, my gearbox motor finally arrived:

It's smaller than I thought, so I plan on attaching it upright instead of using bevel gears. Also from my working models the idea of using a bar is far more reliable than that of a rod, so I'm going to mill the rod coming from the motor square, and do the same with a larger vertical rod which will be the main centre point which everything rotates from. To attach the two I plan on milling out a square indent in one end of this centre point rod so that rod from the motor can slot up into it.

Along with the motor, and in preparation of using the below CAM design I've brought a small tyre (I had to order in 11 more as this is all they had, they should be available by Wednesday) and some springs:

Thursday, 17 February 2011

FMP - Two (15/02/11 - 17/02/11)

First things first, I fixed the first working model:

I had trouble again trying to get the length of the heatbent acrylic right, but then I realised what I was doing wrong. The circumference I was getting was found by using the diameter of the inner circle that I was bending it around, I hadn't taken into consideration the width of the acrylic, which then knocks the diameter up by 10mm, having a great effect on the circumference. However this new circumference came up too large and struggling to think what was wrong I tried using the diamater from the middle of the acrylic, so increasing the diameter by 5mm this time (2.5mm both sides), so that the radius of the circle I'm using to get the circumference would slice the acrylic in half. This worked a treat and came up perfect but due to warping from the heat, I had to fill the tiny gap as seen in the photo below:

In regards to the trouble of the poles becoming stuck, I swapped them for thinner ones with rounded off edges to help them glide the CAM. Also pieces of rod were glued in the middle which the above base now sits on and can easily rotate. All in all there is no trouble with it, but it has highlighted a few issues. The major one being that these poles, as they hole they sit in was cut for a larger diamater, spin a lot. It's hard to keep them from doing, but I've thought about attaching small tires onto the ends which together with guides either side should prevent them from spinning.

I've also thought of placing springs into the top acrylic houses which the poles go up into, so that they're constantly forced down, thus hopefully eliminating them getting caught mid air. This is the acrlyic housing:

Moving onto a second possible design of getting to work, is the above CAM design that I mentioned in my before post. This works by having horizontal poles coming from the centre point (which is turning) which have a bevel gear attached which rotate due to a horizontal bevel gear. So that when the centre point rotates, the poles rotate along with it as well as rotating along their length. Attached to their length is an offset CAM device which, while rotated, creates a greater radius than that of the pole, this is where the vertical pole along with the horse are hung. If that sounds too confusing the following photos should clear it up:

So as the structure is rotated around the wooden base the vertical gear is rotated too:

This in turn rotates the CAM device:

I was initally having trouble making the rod and gear one piece as the hole in the gear was too large (it was store brought). To fix this I attached a copper arm which is soldered to the rod as well as slotted through the gear:

A problem I raised in my research was that the CAM device would cause the pole to go back on itself when it was rotated. The vertical pole should just move up and down with as little forward and back motion as possible. So to keep it centred but also allow it to raise, I added a secondary CAM (the inner parts in the photo below) which when rotated by the outside ones would go up or down, but being loose on one end it allowed the pole to stay somewhat centre:

The idea works fine but it's the model that is fragile and prone to these bits falling off. Just like the vertical gear, attaching these CAMs to a rod isn't as easy as I intially thought as there is little to grip onto from anything circular. For this reason I'm going to try using a sqaure bar tomorrow, which will turn anything slotted over it regardless of how tight the joint is.

Monday, 14 February 2011

FMP - One (14/02/11)

Well today marked the first day that I physically start my FMP, after having spent too much time bogged down with my External Brief referral.

For my FMP I'm building a working carousel based on the works of Dr Seuss, this entails that the overall design will be asymmetrical, wonky and colourful. Having become bored of the perfect finish needed for product models, this freedom to allow certain angles or corners be off is a welcome addition. After putting some serious thought into the chosen mechanism I'll be using I opted for a curved CAM, as you'll see in the photos. Choosing this over the other options which include a CAM above the figures, and a similar curved CAM at the top, this design involves less, which of course means there is less to go wrong.

Having only drawn up these designs, I needed to put it to physical test so today I focused on a working mock up. The initial design is incredibly straightforward, a curved/hilly CAM sits underneath the rotating base which has poles running through it which meet the CAM. As the base rotates these poles are in turn rotated and move up and down according to the height of the CAM underneath it at that time, as seen in the photos: the curved CAM was lasercut and then heatbent around a former of a similar circumference (a mistake in finding the circumference of the circle resulted in the piece coming up short, this will be fixed tomorrow):

The rotating structure on top which houses the poles and figures:

The extra rings at the base keep the poles on the CAM so that they don't slide off either side, a new design will be looked into to fix this as these rings sometimes lipped the pole resulting in it not following the CAM as seen in the two following photos, firstly it working how it should:

The pole becoming stuck at a height:

One way to prevent the pole from sliding off either side is to attach a secondary base disc which fits tightly around the centre point (the larger pole), as you can see in the above photos the the middle gap in the base disc is a lot larger. The new disc is the black one:

Currently this mock up relies on me turning it for it to work, of course on the final one it will be powered by a motor. To get the ideal rotations per minute I've had to buy a gearbox motor, sadly the one I brought isn't powerful enough and I'm waiting for the larger model (which RMP is dependent on voltage) to be delivered, For now though this motor will do and allows me to design ways of how to connect it to the centre pole which rotates the entire thing. The circuit design is final, just a straightforward power source, switch and motor: