A Rack for Shoes¶

I wanted to make a shoestand for my home. I started the project without taking any measurements in my home, and just eyeballed everything, thinking that it would be at least one meter wide in the end.

I needed to change my plans a bit along the line.

Shoe stand¶

I started the design like I like to start all parametric Fusion design: by defining all variables that I think I will need for it. For this, I started with these variables: cabinetWidth, cabinetDepth, cabinetHeight, materialThickness, toothWidth, toothLength, toothClearance,

Variable	Description
cabinetWidth	The width of the whole contructions once it it finished
cabinetDepth	The depth of the whole construction.
cabinetHeight	Intended as the full height of the final object, but additional features meant that it did not end that way.
materialThickness	The thickness of the ODF board. 11mm.
toothWidth	The width of the joint teeth.
toothLength	The length of the joint teeth.
toothSideMargin	Distance from the tooths are from the edges of the walls.
toothClearance	The amount of additional material removed from the female joint for the teeth.
toothSlotGap	Slots are described later. The thickness of the slots.
toothSlotWidth	The width of the slots.
higherMargin	Distance from the top to the top shelf.
lowerMargin	Distance from the bottom to the lower face of the bottom shelf.

Additional variables would come later, as I figured out what the design needed.

The box¶

Drew two sketch rectangles, one that will be the shelf and one that will be the side wall. This shelf is not intented to have any other walls. They gain their proper dimension from variables, and by removing the teethLegnts from them. The teeth lengths are removed so that the final dimensions the same as the cabinet* varibles are.

Added teeth using the proper variables and distances. And slots.

Slots are there to fasten the pieces together. I would not trust friction to keep this thing in one place, so the slots are cut to the teeth in order to me to place a wedge trough them. That way the wall will be fastened between the shelf and the wedge.

Also added small feet for the thing to look pretty. That needed a couple of new variables like legLength and legWidth.

But what about the requirements¶

But at this point I made the measurements for the place that I intended to place this shoe stand, and I noticed that the dimension for that place are 65cmx45cm. That is less than a meter, which is required by the assignment. To go around this problem, I get the marvelous idea of creating a stackable shoe self system. If I stack three of them together, the combined height will be over one meter.

But vertical stacking needs to be stable. So I needed to add joints for the stacking. For front and back movement I could leverage the legs that I had done, by making an opposite shaped outcropping to the top of the shelves, using the same variables as I used with legs.

For left and right movement I needed to be a bit more creative. That needed a new backwall, with similar leg like joints. So I made one, cutting a tooth to the back of the shelf, and creating a small plank with the width of a new variable backSupportWidth, height of the cabinet and tooth gaps the same distance apart as with the walls. As an additional bonus, the back support reduces the sideways wobble of the construct.

Create holes¶

Rectangular pattern: use variables holeCount in y axis and -(cabinetDepth - 2*holeDepthMargin - holeGap) for the dimension in y axis. Cloned them using a Create -> Rectangular Pattern tool with the above settings.

Problems with copying¶

Copyed sketches clearly do not copy all values, especially patterns. The results were weird. Decided to do the copying in the solid phase of design, and not during sketching.

In solid design, I used the Create -> Rectangular Patter command to copy the extruded shapes into their place. I adjusted the distances, so that the milling bit would fit. I had to do this for all objects separately.

Back support¶

In order to make it more stackable, and to account for sideways wobble, I added a back support. I needed to make two new holes

(backSupportWidth - toothWidth - toothClearance) / 2

Because new slots made the holes too tight, I adjusted the variables to make it appear better.

Exporting¶

I opened the Fusion in Manufacture mode. There I created a new setup, where I selected the correct machine and set the stock point to the bottom left corner, toughing the stock bottom.

alt text

The other settings were set according to what we learned during the group assignment, and can be seen from the following screenshots:

After that, I needed to define the routes.

Then I selected 2D Contour, and selected each outline for the objects. I had some problems with getting the milling depth set to proper height. At first it seemed to only want to mill at the surface height, which would not have made a scratch to the material. After a while, I noticed that I had accidentally selected the upper edges as the contour lines, and not the bottom lines, so to quickly fix it, I set the Heights -> Bottom Height -> From to Stock bottom.

For all the holes, I used 2D Pocket. This was most likely a mistake, as most of the holes, especially the bigger ones, could have been cut as contours, and would have been faster. I created three different pocket routes in the program, one for each type of hole I had: The big dirt holes, the construction holes and the wedge holes. All of them used the same settings:

After that, I pressed Actions -> Post Process in the header ribbon, and exported the post processing file.

Milling¶

Milling took about two hours to complete. It would have been faster, if I would have not carved the holes as pockets, but as contours. That would have about halved the time it took me to mill the board. But you live and learn.

And I probably should have done a couple of test pieces first, before milling the whole thing.

But before I could even start milling, I ran into a multitude of problems while trying to setup the milling machnine for the design.

Problem #1 - WCS Warning¶

WCS Warning

When I tried to export the toolpath using Autodesk post processing command, it gave me an unclear error. The error said “Work offset has not been specified. Using G54 as WCS.” Really unclear, and after googling a bit with instrutors, we found that the error would go away if I go to Setup -> Post Processing -> WCS offset and set it from 0 to 1. No idea what that setting does, and why it fixes the problem, but the warning went away.

Now to the milling machine!

Setting up the milling machine¶

According to the label on the side, the milling machine in our university is ES-1325. To operate it, we used a NCStudio.

alt text

Took our 1200mmx2600mm OSB board, placed it on the bed, and started the suction with above buttons.

annotated ui

The user interface for the NCStudio was quite clear. I basically pressed the buttons in the indicated order.

1 is for manual control of the machine arm. Here I first selected the Manual tab, after which I also needed to select the HW, i.e. hand wheel. After that I could move the machine with the hand wheel nect to the device.
2 is used to zero the machine to its current position in X, Y, and Z axes. X and Y are setup using these, but Z is zeroed using Mobile Calibrator.
Here I opened the correct file
3 starts the simulation of the file.
4 shows the simulation. It can be controlled by dragging (which updates the view only after you let go), or through right click menu.
5 starts the milling machine proper.

I zeroed the z-axis using the automatic z-setter tool in Operation -> Mobile Calibrator. For that I needed to use a mobile calibrator, a conductive piece of metal of specific size, that the tooltip will touch, and measure the intended depth of the base.

I set the x and y origin manually using the hand wheel and clicking the coordinates in the UI. As I intended to use the almost the full area of the board for my milling, I needed to be as precise as my eyeballs could be to set the zero coordinate to the bottom left corner of the bed.

alt text

The I needed to place the bit. (Image does not contain the correct bit.)

alt text

Placed the protectice panel in front to protect myself from debris. Also added safety glasses and ear protection.

Problem #2 - Milling body as a pocket¶

After everything was I opened the toolpath file, and simulated it in NcStudio. Everything seemed OK, except two topmost pieces, which the milling machine wanted to pocket out completely.

After looking through the file in Autodesk fusion, I noticed that I had accidentally clicked to objects as two Geometries for pocket cutting in addition of the pockets that were in them. I removed them, and created a new post processing file, and ran it into the machine.

Problem #3 - Design was bigger than the bed¶

After that, I noticed that the milling goes outside the marked working area in the simulator. I had decided to use the full dimensions of my board, which were 1200mmx2600mm, but had not taken into account that the machine bed is only about 1200mmx2300mm. To fix that issue, I removed the toolpaths from the topmost objects from the design, and recreated the toolpath file.

Those three additional pieces could be cut later, by moving the unused parts of the board to proper position, and cutting just those three pieces.

Play button

Now, we could press the play button to start cutting. And cut it did.

Problem #4 - It did not cut through¶

Well, the cutting didn’t last long. While the machine was cutting the first pocket, the instructor became suspicious. At first the instructor noticed that the z position of the tool did not seem right, it seemed to be too high. And once the pocket was finished, and the machine moved to the new one, even I noticed that the machine had not cut the material completely, and left half a millimeter thick base where it should have cut completely through.

After analysing the issue, we paused the machine, and looked at the design file. In the all pocket milling setting, I had accidentally left Stock to leave option on, and left 0.5mm thickness to the bottom. In the contour cutting I did not have that setting on, probably because by default it is not on there. So the contours would have been cut propoerly.

So I toggled those settings on, recreated the toolpath file, and the milling recommenced from the beginning. The machine cut the first two holes again, but that was only a couple of minutes of additional wait, so it was not that huge of an issue.

The machine decided to optimize tha paths weirdly. It cut some shapes only halfway throught before moving to other parts, only to come back to them later. In the end, it cut everything.