Sebastiaan FabAcademy

Week 7 Computer-Controlled Machining

website version3.0

Presentation of FabAcademy of the Waag students

During the weekly lesson with Niel, Niel ask us to demonstrate our Shopbot milled objects of this week assignment. This is de video

I have looked forward to this weeks assignment. In the fablab Enschede, where I work, we have the shopbot (#10162) PRSalpha 96-60 running since a few month but we haven't made much with it due to troubles with the machine. Special the spindle driver is causing of problems. I like the machine and for this week my challenge was to be able to learn how to operate the machine and how design products for it. I startet by following the Shopbot online tutorials. The youtube movies are very helpfull.
My personal goal was to learn the machine and to play with the different joining technics made out of wood without the use of screws and or glue. So the 3 products I have designed for this weeks assignment are connected with only the use of clamping force and wedges.
I have startet this week to change the leafbot and prepare it to be milled on the shopbot milling machine. The connections of the wooden plates have been changed. The connections between the plates are holded together by clamping. I used clamping force between the wood it self and also I used wedges to clamp the different parts.

This frame structure is joined together by wooden springs. The hooks acts like springs, so there is constant clamping force. The corners are made "square" by adding an extra rounded corner by the mill. Next time I will try to drill the corners, so the mill can run faster trough the corners without overheating the mill. Also the radius of the corners can made smaller. The minimal radius is now 3.25 mm because of the mill diameter.

The rear side is closed and has the same joints

The tilting storage is designed for the laser. It is also clamped. In SolidWorks I have designed the tabs nominal, so with no play. I saved the seperate files as DXF and opened them in Adobe Illustrator. In Illustrator I joined the line and added an offset of 0,2 mm. The laserbeam of trotec speedy 300 has a width of 0,1 mm. With the offset of 0,2 mm I have created a clamping of 0.15 mm, because the laserbeam cuts over the outside line. The half of the laserbeam width is been cut inside the cad model, so by appling an offset of 0,20 mm the clamping will be 0,15 mm. The laser cutting setting are: power 100% , speed 0,45 , frequency 1000. This are standard setting the FabLabEnschede uses for 6 mm plywood

At the backside a 18 mm plywood milled part has been created, with is line with the frame stucture above it. This part is milled and clamped over the laser cutted parts.

The side panels of the main leafbot structure and the front suspension are clamped by wedges.

All the leafbot SolidWorks files have been saved as DXF

In VCarve the leafbot files have been imported. The files are packed close to each other so less material is wasted.

The internal shapes have been selected. The toolpath is created with the toolpath tool "pocket". The plywood thickness is 18mm, a ramp of 5 mm is added. This toolpath is saved. The gabs, where the bearings are been pressed in, have a depth of 10 mm. For this a special toolpath is created also with the toolpath tool "pocket". After this all the outside shapes where selected. But what I did was select each product separatly and make for each part a special toolpath, not knowing that all outside lines could be selected in one outside toolpath. The selected tool is an Onsrud Single Flute - Carbide Tipped Straight 1/4 inch.
Each outside toolpath I added several tabs with a height of 3 mm and lenght of 20 mm. Below the settings of the mill.

After saving the toolpath I simulated the toolpath in VCarve to check it for faults.

Now it was time to vacuum clean the Shopbot

Fixate a plywood plate with a size of 1250 * 2500 mm. The plywood is special water resistance variant multiplex plate for the formwork also called concreteplex plate. It has a toplayer of fenolcoating. Because of the water resistance I use this for my leafbot. The plate is fixated with 3,5 mm wood screw. Later someone advised me to use aluminum screw. If by msitake the mill millss trough the screw, then the mill is not so badly damaged, because the aluminum screw is much softer then a steel screw.

Ready to mill but safety first

First the machine is turned on. Then the X and Y axis are been zerood.

The Shopbot software runs an automatic program to zero the X and Y axis.

After the machine has run the program it gives a warning that the X and y axis are zerood.

Now the Z axis must be zerood. First at a distance from the table so a dry run can be performed to check if the toolpath is correct without damaging the mill and wooden plate.

The software warns the operator that the Z axis starts after hitting the ok button. Now the machine moves the spindle in the Z axis twice and hits the aluminum plate. After hitting the plate twice, the software recalculates the height by adding the aluminum wall thickness to the height of the tip of the mill.

When the Z axis is zerood the software gives a massages that the process is finished.

The shopbot starting to mill the first leafbot parts

A problem ocured. The spindle stopt turning and a warning popped up. It stopt somewhere in the toolpath. I seached for I on google but it seams to search for a addition Z axis with is not there and is not programmed.

I looked into the code, but nothing strange to see.

Then a drive fault popped up when the X and Y axis was zerood.

Other fault massages popped up.

Since the fablab Enschede has the shopbot there is often a issue with the spindle driver.

I restarted the shopbot and the computer and opened the toolpath file with Cntr open so the toolpath can begin there where it stopped. Strange is that is moved on where the fault appeared. But more often in the toolpath it stopped again

And again

I restarted the toolpath and it milled more of the objects

Again the same problem

After the restart of the toolpath it suddenly had moved the zero axis of the X and y axis. It ran into the plate fixation screw.

And instead of milling the wegde next to the other wedges it milled somewhere in the middle of the plate the other wedge.

I zerood al the axis and started again.

Final result looks ok

No burn marks of the mill. This is a good signal because the mill was not overheating and therefor did not wear very fast.

Detailed view of the milling result. The finishing could be a little bit smoother.

The round pocket looked not 100% round. This is the joined for the tilting storage.

The plate was take off the shopbot

With a Fein MultiMaster the parts where cutted free from the original plywood plate. This is perfect tool for this job. It works very fast and is very flexible.

The tabs where sanded away.

First look how the leafbot looks like in reality. From 3D CAD to the real product it is always suprising how it looks in real. Does it look like as intend to be? It does!!

Because the fenolcoating was breaking of I decided to chamfer al the outside corners to prefent the chipping.

With an hand milling machine I chamfered all the edges.

The bearing 15 x 32 x 9 mm did fitted in the pocket.

The wegde construction worked very good. With a hammer I pressed the wegde in the gab, resulting is a very stiff connection.

Also the frame stucture for the tilting fitted very strong. I had to sand a little bit because the wood was 18,5 mm thick instead of 18mm. In Solidworks I worked without any play. I had allready drawn the leafbot in SolidWorks before I had time to by the plywood. I don't do that again in this sequence. I discovered a mistake in the left structure part. The pocket for the joint had moved backwards. It must have been caused by a transform action in VCarve, because in SolidWorks this is ok. So next time I must be look to it that by a movement action I have to select all the lines.

After the first Leafbot parts I milled more parts. Now I changed the toolpath. I discovered that I could select the inside pockets all at ones and also the outside path also. So I ended up with only 3 different toolpath instead of more than 20. During the milling it give only one fault instead of more then 10 times with the first milling run.

In SolidWorks I designed an Abs trainer out of wood with only wooden joints.

The wooden parts are held togehter by wedges. The abs trainer should be stored compact and therefor the front part is made with rotating joints.

Next object I like to make on the shopbot is a trolley to move the big wooden plywood plates. I build a prototype at home out of scrap materials. The weight distribution was not ok, a bit to much to the angled side. But for the rest it moved very good.

The design was made in SolidWorks. Now I measured first the wall tickness of the plywood with is 17,7 mm. I constructed the next two object with this wall thickness. The complete trolley is joined with wedges, the same as in the abs trainer.

The holes are there not for weight reduction but for fixation objects with a wire to prevent it falls off the trolly.

Picture of the handle bar construction

The horizontal plate must be mounted first, then the handle bar and the wedge fixates the handle bar.

The bottom section is also clamped by these wedges.

I did a interference check. Only the wedges where interfering

Both the files of the abs trainer and trolley where saved as DXF and all the DXF files where imported in Illustrator. In Illustrator I moved the parts so that they fitted all at ones in the plywood dimension.
First I selected the pockets and the mill and parameters where the same as with the other milling runs.

Next thing was the selction of the outside lines and make a toolpath with the outline toolpath tool. Tabs where added and a ramp of 40 mm was added.

Again I had troubles with the unexpected output fault somewhere in the toolpath. Again I had to restart the program.

First the pockets are milled

Again a fault massages

I had to start again from this point in the toolpath.

Some parts where laying loose So I have removed them before the mill get stuck in it.

The round joint did not fit. i constructed it with a play of 0,3 mm, but it looks again not round, but oval. The joint is also not closing enough. During the rotation movement the joints comes loose. This is not ok and must be changed.

Al the parts must be sanded.

The first parts of the trolley clammed togehter.

The wedge hammered in the gab. The wegde clammed a littlebit to much so I had to sand it a bit.

The next wedge to prevent the firts wedge will fail out.

It looks good, but is not usable.

I forgot one vertical pocket to fit the handle bar to mill. It didn't select it in VCarve.

I disassembled the trolley and milled the missing pocket in. It looked ok.

The horizontal bar clammed

The wheels fitted

The horizontal bar assembled. But here I made a design mistake and the pocket was not milled on the right spot so the handlebar does not fit as desiged. I have to make a new side frame structure.

But it looks ok

And it works very good and it is allready in use at the FabLab Enschede.

Because the abs trainer design was not ok, I redesign it and milled the redesigned parts

Sanded them. The milled showed signs of wear, because the edge where very rough

The assemly of the first trainer abs trainer parts. It fits perfect.

The reinforcement was assembled in the middel front of the trainer app.

The rear grond support added.

The top support added

The wedges hammered in.

The back added. I had to sand here the pockets to make it fit more easely.

The axis pressed in.

The first part of the front part added

The front added and the abs trainer can stand on his feet

The feet support added.

It can fold.

demonstration of the abs trainer

Presentation of FabAcademy of the Waag students

During the weekly lesson with Niel, Niel ask us to demonstrate our Shopbot milled objects of this week assignment. This is de video

29-03-2016: I updated the trolley because there was mistake in the design of the trolley and an miskake happend during the milling. So I redesigned the trolley and milled the grip and side walls again.

The mill starting to get rough, so the edges where splintered

The complete trolley was disassembled and to assemble the new milled side walls

The new handlebar is assembled. The trolley parts fitting perfect together. Trolley is ready to be used.

Files of week 7

SolidWorks files

Shopbot V-Carve files

DXF files

Laser cutted parts

Illustrator files prepared for milling