Wednesday 10 July 2013

Version 2: Planning and Building the Y-Axis

When designing the reworked version of my gantry warehouse system I decided that starting from the Y-axis was the best idea since the design of the rest of the robot which depend on how the Y-axis turned out. As a staple building material, I used mostly angled aluminium and some flat aluminium. Its light, strong, easy to work with and machine, and it can be bought in a number of pre-made dimensions at any bunnings warehouse. Aside from this, a range of M3 screws and nuts were the other staple supply i used in this.

The first stage was simply going down to bunnings warehouse and looking around for what dimensions were available whilst also looking into possible options for a roller design to replace the rod design for the X and Y movement. I then managed to get hold of a company which were giving away off cuts of aluminium from which I managed to get the bulk of my materials. However, the use of slider door rollers was out of the question due to the incredibly high prices.

In the end I chose to use, 20x20x1.5mm angle, 35x25x1.5mm angle and 12x3mm flat. From this I produced a rough sketch of the design shown below.



Upon making this sketch, I then had to build it beginning with the movable component of the Y-axis. This was essentially a tiny car upside down, its wheels running across a length of 12x3mm flat bolted to 20x20 angle. This would replace the rod mechanism and whilst it is much more complex and in need of accurate measurement, it produces little friction and will not bend, a problem with the old design. The car like mover was made of two pieces of 20x20 bolted together. These where drilled with the vertical mill then drilled by hand. The opposite sides of the car where drilled on the vertical mill 80mm in diameter to allow an M8 screw to pass through for each wheel and nut. The alternative used rather than door rollers, was ball bearing wheels. These are stronger, cheaper, easy to bolt, with the only downside being their heavy weight. On the bottom of this contraption, another piece of 20x20 was bolted as a mount for the servo. The servo was then positioned in a hole made by the nibbler tool and then bolted in place. 

Next, the structure for the runners and motor was made. The motor was cable tied to a piece of 35x20x1.5mm angle. A hole was nibbled out for the encoder wheel and gear to fit through. The two long pieces of 20x20x1.5mm and 12x3mm were bolted together, spaced to allow just enough room for the car to hang and run freely without any resistance. M3 screws were used primarily however at one point I ran out and had to switch to M4 screws. 

The issue of tension in the last design was due to the lego having a set distance, however, with aluminium I could custom fit. To ensure the axis was always in tight tension, I did not seal up the other side of the Y axis. It was left free for a movable end to be screwed in. This way, the opposite end could be stretched out to a comfortable tension and then bolted in if ever maintenance was needed rather than cutting a whole new piece of aluminium. This also allows the timing belt to be taken off easily whenever needed. This extendable end was made from 2 lengths of 20x20x1.5mm angle. These were faced opposite to eachother and drilled through to fit an M6 screw. This holds the idle wheel which the timing belt goes around. As spacers nuts and washers were used in between. Finally, the two pieces of 20x20x1.5mm were held together by another piece of 20x20x1.5mm by M3 screws.

Another issue with the last design was the timing belt not catching the gear hub of the motor. This was due to the 3D print not coming out accurate enough. After a lot of testing and reprinting, a very close fit has been produced so that the teeth catch with no problem. On top of this, a new encoder wheel has been made. This was laser cut, the new wheel with 20 slots. This will provide extremely high resolution for the new vesion. 

At this point, the mechanics of the Y-axis are done aside from some final touches that need to be made. For one, the motor is held by cable ties. This is not strong enough and the motor has movement. The reason screws into a mount were not used because the size needed was M2.5, a difficult size to find. However, once I find it I will be able to make a mount, bolt it in properly, then add the screws to hold it in tensiono. Right now it is slack  as I do not know how much tension it can hold as the motor is moving. After this is done, the limit switches need to be mounted and the photo interrupter has to be mounted. After this the process can be repeated on a larger scale for the X-axis.