Sebastiaan FabAcademy

Week 14 Composites

website version1.0

This weeks assignment is a what I want to more often in the near future. To be able to build larger scale objects, like velomobiles. But I also like to work with ecological materials. Material that are not hazared for my health and are recycleble or reusable. I startet looking for hemp. But I did not had so much time to get hemp, so I decided to go for burlab and there was some pieces left at my home. Then I looked for Bio Epoxy. I found bio resin from Voss Chemicals in Belgium . I ordered 1 kg of it and they where able to send it to my place on this Tuesday. The Datasheet (sorry only in Dutch avialable) can be downloaded here .

I decided to design the cover for the leafBot in SolidWorks and try that one to laminate to a composite. In Week2 I made my first LeafBot Cover design. I used the basics of the design of that cover to design this new cover, because the first one doesn't fit on the current dimension of the LeafBot. I designed this new cover as one part.

This is the cover without the Leafbot. I used draftangles of 5 degree on the side walls, to get the cover more easy out of the mold. Except for the wheelcover. Otherwise I had to make the cover even wider. As a test I decided to try non drafted area also, to look if the model will be easy to remove from the mold. Then I designed a additional edge on the bottom-edge to reinforce this bottom edge. At the front a cut out section is added to fit the sonar sensor in it.

The rear is open because of the tilting storage to emty this leaf storage.

The Cover will be able to turn open over the axle of the tilt mechanisme of the leaf storage. I have to laminate a joint in the rear corners so the axle will be able to fit in the joints.

I made joint in SolidWorks. I created some extra ribs on the top and bottom to get a better connection to the burlab laminate. I added a draft to the hole, so that the vacuum foil is able to follow the outside shape better than if the sidewalls of the joint have a sidewall angle of 90 degrees. The vacuumbag I used from the Blokker . They are not the best quality but my big mold fits in it.

The ribs have also a chamfer applied to all the edges so the burlab is able to follow the shape better then with sharp corners. The hole has an offset towards the ribs so the burlab, that is put on the insert, will not effecting the axle.

In week 12 during the moldmaking I discovered some mistakes in my mold because, due to time reasons, I had designed the negatif mold in my head. It didn't want to do that this time. I followed a online tutorial of the mold making tool in SolidWorks. This was very helpfull for me. I followed the steps of the Solidworks tutorial and was able to make the negatif mold.

This is the inside of the negatif mold.

Now it is time to think how to mill this big mold on the Shopbot, because it is to big to mill from one block of material. Also then a lot of material is wasted. The way to go for is slicing the mold in different parts. My idea is to devide the mold in 2 side walls, top side and a front and backside. Screw them together to be able to get the laminated product easier out of the mold. An other limit is the length of the mill. In the FabLabEnschede we have several mills. I have decided to use the Onsud straight cut 1/2". This has an cutting lenght of 25 mm, but can cut about 40 mm in total. The maximum depth to mill is on one spot 36 mm (2x 18mm plywood). Since I have a large object to mill, outside dimension 1050 x 300 mm, I decided to use this bigger mill to speed up the milling proces. Due to the 2 days of closure of the FabLabEnschede Paul and I had to work ahead. So my plan was to mill my parts in week 13. Due to the Spindle driver problems I could not mill on full speed. The new 380V Spindle driver is still not arrived. Therefor I only can mill with a maximum of 15000 RPM and a feedrate of 8 m/min and a depth of 2 mm. I tried to do more mm in the depth at ones, but the spindle driver didn't managed it and stopped working. For the finishing job I decided to use the 1/8 inch onsrud ball nose mill. The length of this mill is needed for the same reasons as for the roughing milling proces.

I decided to mill first 1 side of the complete mold. It has a thinkness of 4 x 18 mm plywood. But that is to thick to mill it at ones. First I looked for plywood that was left over from previous milling processes and glued 2 layers of 18 mm plywood on top of each other.

I fixed the plywood on to the shopbot. Had a good look where the machine would not be milling. On these spots I could apply the 4 screws.

The finishing job of the first 2 layers of plywood finished. This finishing job took 4 hours of milling. Now I could glue on some more left over pieces of plywood on top of the first 2 layers of the mold.

First the roughing and then the finishing again. Since the second layer has a kind of S-Shape, I hoped that finishing job would be milling only the material away that is in the CAD file itself, but it milled the same outside dimension as in the first layer. A lot of time was wasted milling in the air. I have to check this in Part Works to change the setting so it then follows the outside shape instead of the square of the outside dimensions.

Milling in the air

Milling in the air. Good to see it mills a square.

To speedup the milling proces and give Paul also some time to mill, I reduced the milling overlap to 50 %. I did this for the roughing and for the finishing job. The half of the mill diameter is then milled, but with a ball nose shape, the milled surface looks like this.

But as mentioned before, the things I could do at home was, sanding the mold. With my festool Delta sanding machine and vacuumcleaner attached to it, I sanded the surface.

But some spots I had to do it by hand.

For the sharp corners I have a Makita belt sander machine. I also used this one for the corners and small radius in the mold.

I used putty that is been used by car painters. It is a two component system. It hardens very quikly and it can be sanded very nicely. Also the putty doesn't stick to the sanding paper.

Because the mold has a lot of surfaces that run in to each other, I used two flexible spatulas to apply the putty in the mold.

Quickly mixed the 2% hardener in the putty. Now I had to rush, because I only have a few minutes to apply it to the mold.

The putty smoothend in to the mold

Sanding the putty with the Festool machine and by hand again.

Finished the sanding.

I have sawed the sides off and glued on a additional reinforcement on the bottomside of the mold, because it was the wood was bit bended and only 6 mm thick.

All the edges where chamfered to prevent that the vacuumbag will tear due to the sharp edges during the vacuum proces.

I took a used karton box to cut out an piece to laminated to get a stiffer side wall.

Made a cutting mold from some 2 x A4 papersheets.

Cutted the cardboard.

Next is cutting the first layer of burlab. I used the pencil lines later also to check if the burlab placed well in the mold when I apply the resin. I cutted the burlab 2 cm bigger then the lines. I did this to have a bit extra metrial if the burlab would be moved a bit during the dipping of the resin and vacuum proces. The burlab can then be moved a bit and with this extra material without the problem that the edges are not applied with enough burlab.

I had different types of burlab. I used the one with the finest weave at the outside and the more open one the inside, so the resin can then move better through the laminate.

I carefully cutted the extra pieces of burlab to fixate the 3D printed PLA insert.

I had burlab avialable for 3 layers

To get to know how my clothing vacuumbag works and how strong the vacuum cleaner should suck, I tested it dry.

Applied the vacuum a bit to fast

So the bag could not follow the mold in the corners. I had to do it again and apply the vacuum slower so I could pull the vacuumbag in to the corners.

Now it did work well, The bag was pressing the burlab in all the corners.

I cutted a piece of LDPE foil to prevent that the vacuum bag is glueing to the resin.

In my lunch brake I purchased vaseline and some stiff Block brushes at the toolshop. I took these brushes because the hairs of the brush are stiffer. In this case the resin can be dipped better in to the burlab.

Applied the vaseline to the complete mold. Also to the topside of the mold.

Preparing the laminating. Covered the table with a plaster foil (is been used by contractors to protect the flooring and is very strong). I had extra handgloves ready, so I could make pictures of the lamination proces.

The waste container next to me. Additional papers for cleaning. The pieces of burlab in the right order placed behind the mold. I was wearing handgloves, mask with filter, protecting clothing.

Lets start

Applied the first layer of burlab and dipped the bio resin in to it.

Next layer is the carboard.

3e layer is the more open burlab.

I realised I hadn't so much resin. I had to dip the resin very good in to the burlab. With this open burlab if was more difficult to devide the resin.

The joint is applied to the mold. I used some extra resin to glue the joint to the burlab.

All the 3 layers of burlab applied to the mold. I had no resin left. Used it all.

Put the LPDE foil on top.

Put the mold in the vacuum bag and vacuumed it slowly so I could press the burlab nicely in to the corners of the mold. Then I used a rolled to press the resin trough the burlab, during the vacuum.

But I used up the 30 minutes of the open time of the resin, and to squeeze the resin through the burlab was not to easy any more.

Applied more vacuum. The plastic pushed the burlab, and the the shape of the mold could be seen very well. I didn't feld spots where the burlab was not pressed well to the mold.

After a while the first resin come upwards.

On this pictures it is better to see.

After 1 hours even more resin was pulled out the burlab

This Wednesday early in the morning, before 6 am, I took the foil of and placed the mold in my car to drive to the Waag in Amsterdam.

I was not able to find the breathable foil, and therefor the resin that came out of the burlab could not be removed from the burlab. This white surface is the resin that is been sucked out of the burlab.

Carefully I was able to release the side wall of the LeafBot cover out of the mold.

The shape looks good

But some putty is attached to the resin. Perhaps I have not cleaned the mold after sanding well enough and applied not enough vaseline to the mold.

On some spots the resin isn't applied well trough the burlab.


vCarve files Roughing files

vCarve files Finishing files

STL files of the 2 milled layers of the mold