Sebastiaan FabAcademy

Week 12 Molding and casting

website version1.0

Today it is kingsday in the Netherlands. So the Netherlands is celebrating this day. In Amsterdam it is very busy and we have no class at the Waag.

This week I had in mind to cast the sole of a shoe I like to make. In Week5 I already scanned my cycling shoes. My plan is to design my own cycling shoes. They have to have a stiffer sole, to efficiently transfer the power to the pedals. I was inspired by the Nike Eames shoe design. This is again inspired by the design of the Eames loungh chair.

It is I think only a 2D design proposal, because I have only found pictures of this side of the Nike Eames shoe. But I like the idea to use a laminated wooden sole to reinforce the sole. I like to design products that are build from eco friendly materials and constructons. So my idea was to design my own shoe and use the inspiration of the Nike Eames shoe. Build the shoe with the use of my 3D scans. For this week my first idea was to make the rubber sole and clamb it to the wooden sole and not glue it.

But I didn't had so much time to engineer the complete shoe, and on thursday at the end of the day I had to decide to change this weeks assignment and make something "easier" so that I was able to fullfill this weeks assignment. I will continu working on my shoe design. I already have bought thin fineer beech plywood for the laminating week.
On thursday afternoon Paul and I had to rush to the FormX store. We came just before closing time but we where able to by some additional resins to cast our products, but still not knowing what to cast. We have bought Smooth-on Mold Star 30 silicone to make the molds. I have bought Smooth-on PMC770 to cast with. This is a shore A 70 Polyurethane with should be a bit flexible. I also have bought Smooth-on FormCast Rhino premium PU casting resin. This is a strong Polyurethane material. So I was able to make something stiff or flexible.
I know that in my FabLabEnschede we didn't had small mills to use for the Shopbot. On Fryday early in the morning, I was able to buy 3 mm mills in my home village Enter at Slijptechniek the Netherlands. Its a super modern factory where they make mills and other tools. So with 4 3mm mills I drove to Enschede to my FabLabEnschede. My planning was to mill the molds this fryday on the shopbot at FabLabEnschede. But meetings and regular work kept me from designing an 3D model to be milled. In the afternoon I got an idea to make hooks to replace the Ikea hooks I have at home. These Ikea's constantly falling down to the ground. I designed a complete mold, and I added draft angles to all the walls to prevent that the small finishing mill would tough the side of the mold. I had bought mills with an clamping diameter of 4 mm, so I had more options to chose from, the mills where longer and the mill is stiffer.

The finishing mill on the left, the roughing on the right. Good to see that the mills changes in diameter. So that is something to keep in mind.

I designed the complete mold, with allingment pins. Also with a pooring hole and a hole to get the air out of the mold. I designed a left and right mold. I designed the mold so the mill could pass all around the objects, like the pins that are sticking out. So the disdtances between the objects where more than the mill diameter. Also I made the silicone volume smaller to save on material. And design the mold so when it would stand upright the product inside was tilted so the air can been pushed out.

Saved as STL is was imported in to the Shopbot Vcarve 3D. This was my first time working with a 3D milled part and this software. I opened the STL file and only a big block was displayed. Time was quickly passing by and there was only 1 hour left, before the school building was closing. I could not find the problem fast enough so I decided to remove the mold from the hook and imported it again in Vcarve.

I quickly fixed the concrete plywood on to the shopbot, so I will be able to collect the wax chips more easier

We will use blue machineable wax from ferris

Applied double side tape at the back of the machinable wax.

Placed the wax somewhere in the middle of the concrete plywood, so the chips will be fall on the plate and not a the side of the plate.

Imported the halve of the hook in Vcarve. I made the separtion in this was to be able to get it out of 1 wax block. Also because I had then to use longer mills which I didn't had. The lenght of the mills in this case a bit of a limitation.

A block around the hooke was defined

The roughing parameters where applied. I used a 3 mm mill.
Spindle RPM: 12000
Feed rate: 80 mm/sec
Plunge rate: 20 mm/sec

The finishing parameters where applied. I used a 3 mm ball nose mill.
Spindle RPM: 12000
Feed rate: 80 mm/sec
Plunge rate: 20 mm/sec

The cut out path was not selected

Running the simulation. It looked ok

Starting the Shopbot. First do a dry run.

The model was not placed in the same way as the wax was placed on the shopbot. So the toolpath had to be recalulated and the model had to be turned 90 degrees.

And change the toolpath again, because the zero-pionts where still in the middle of the wax block. But the shopbot was set to the machine zero piont. So in Vcarve I had to change this also.

The milling could start. It looked ok.

During the finishing the mill was toughing the inner side walls of the mold and pushed them. So the wax mold released it self from the concrete plywood and was moving

For safety reason I stopped. The finishing was half way when I stopped the shopbot

Time was up for this week and I had to clean the machine

Collected all the wax chips

On Saterday morning I made myself at home a mold out of concrete plywood.

Made it a bit higher so the chips would fit in. the chips had much more volume then the wax block.

My first idea was to use this wooden mold to meld the wax. but I looked to the datasheet of the wax to look for the melting temperture of the wax. I discovered that this was 250 degrees Celsius. I also looked to the self ignition temparture of the concrete plywood. This was between 250 and 350 degrees Celcius. So that was not good. I surged in my barn and found a metal box to meld the wax.

The temp of the ceramic oven was set to 250 degrees Celcius for 1 hour.

Not knowing how fast the wax would meld I stayed nearby the oven.

Checked it several times

The wax melted

Casting it in the wooden mold.

Have it cooled very slowly in the warm oven.

Safety first. I was wearing safety glasses. Thick leather apron, safety shoes, and leather gloves.

To be sure I had enough wax, I only had 2 Ferris wax blocks with me, I bought 200 candle lights and big LDPE plastic bags.

Removed the aluminum and candle wicks

Cutting the plastic bag in smaller chips

From the website I found a recepy. 200 candle lights and 5 LDPE bags. So that mix ratio is used.

It was running late in the evening and I stopped at 10pm.

It was Sunday morning when I removed the wax out of the oven. I tried to get it out of the mold and sadly the bottom was still sticky and soft so the wax had to be casted again.

I had allready my doubts about the use of wood for the mold. The wood seemed to cool down and heat up strangely.

In the afternoon Paul came to my home because I had to help him casting bigger sized wax blocks. Because Paul needed more blue wax I give him my blue wax block. So I took the risk to fully rely on my own wax mixture and melting proces. It took all afternoon but it looked ok. We have cooled the blocks in a old microwave, without turning it on. Just to keep the heat, that is comming form the warm wax, inside the chamber.

At 6pm it was time for me start making my own 3D model and mold design. Paul showed his mold design. A mold with inserts. I thought I also can try to make a mold a bit more difficult then the Hooke. Also I had to design a new product, because my laptop had run out of battery power and my charger still in the fabLabenschede. So I had no acces to any of my files. On my home CAD station I had to start from scretch. I realised I still had to design a connection part to connect the front wheels of my LeafBot. I got the idea to make a flexible part, and to use the flexibility of the PU. Also I had to fix the distance sensor to the robot. This could be intergrated.

So I designed an flexible joint. I made it without any fancy design. Because I was needing al the time that was left and I had still to meld the candle lights.

The joint is clamped over the wood

The joint in detail

Then I started to design the mold it self. I created spaces in the wax mold to fit in the insert. I made bigger sizes inserts so they would be better to handle and better to fix in the silicone mold. So I created bigger spaces where no silicone would be casted. Otherwise the inserts would not fit. The pooring and air outlet are also added. Because on Fryday I already discovered that if I designed the mold to be milled with a small 3 mm mill it would take very long to mill this mold. So this mold is design to be milled with an 1/4 inch (6,35 mm) up cut mill. The mould will be have more milling marks, but I have not more time then this.

It was 9pm when I finished the 3D CAD work. I startet to make a wooden mold. I could do this now because I had the dimensions of the end product.

Try to guess how much of the candle needs be to melted. I used 138 candle lights.

In a old cake tin the half of the candle light fitted in and I spreaded the LPDE chips, so it would melt nicely mixed.

At 10pm I put the cake tin in the oven and turned on the oven and heated it up to 250 degree Celcius.

Checking what would happen. It started to meld slowly.

Almost meltet to a liquid mixture.

Mixing the wax with LDPE. It was fully transparent, but also funny to look at when I took the mixing wood out of the material.

I put the casted wax in the oven again to cool it down. Then I had to heat up the other part of the candle lights. At 11pm I could also cast the second mixture in the mold. I discoverd that the first casted wax was partly melting. So I allready was thinking about creaks and air bubbles inside the wax.

Monday morning at 7 am I took the mold out of the oven, still feeling warm. All the castings I did. I also made small molds for the insert (right side of the picture)

Because my white wax was still a bit warm, I was affraid to take it out of the mold and face the same problem with as Saterday, that wax would break of and stick to the wood. So paul could mill his molds first. I had to help him operation the Shopbot. Paul also had to give a lesson So I stayed at the shopbot. Also later in the day Paul went to the deparment party and I finished the milling job. But I took very long to mill his parts. Paul tried several milling parameters. One was with a bigger overlay for the finishing milling. It looked nice but it took 2 hours to finish this finishing job. At 7pm finaly it was my time to mill.

My idea was to combine the two molds to one stl. When I checked the file the side wall of the molds where not there, ending with a hollow part. Only sheets could be seen. This can not be milled. So I had to mill the two molds separtly.

Result of the casting of the candle lights. In the middle a shrinkage.

It released very well from the mold

Back side looking not so bad.

First I had to fix the wax to the concrete plywood. I added extra ductape because the double sided tape was not sticking very well to the wax.

The flattning milling was looking ok. it was a bit of an relief to me that it seemed to work.

A nice flat surface. I milled to a depth so I had just enough height, to fit the mold in. Due to the amount of shrinkage the height was a bit of a problem to fit the mold in. After some measurement and change to the 3D CAD file of the mold I was able to fit it in.

Finished the flattning surface.

Time to load the mold file. Imported as STL. I changed the position of the mold so it would fit in the right position in the wax.

Adjusted the block size parameters to 46 mm thickness of the wax. No silhouette and no milling margins selected.

The roughing parameters where applied. I used a 1/4 inch mm mill.
Spindle RPM: 12000
Feed rate: 80 mm/sec
Plunge rate: 20 mm/sec

The finishing parameters where applied. I used a 3 mm ball nose mill.
Spindle RPM: 12000
Feed rate: 80 mm/sec
Plunge rate: 20 mm/sec

No cut out toolpath selected

Checking the toolpath and saved it. I imported the roughing toolpath in the shopbot

A new error appeared. Because I saw the spindle moving very slowy downwards, I linked it to that. The setting was on 1 inch/min. Strange because I changed it to 20 mm/sec.

The milling could start. But first I did several dry runs to be sure that I had not made any mistake, because there was not time left to make mistakes and to the melt the wax all over again. I also have to do this at home and I live not nearby my FabLabEnschede.

It was 8:15pm when the first mold was ready. Yes looking good. My fabLabEnschede team leader Ruben was so kindly to bring me some sandwiches, since I was not able to join the department party. I was tired and very hungry. So this help of Ruben was very helpfull to me. Many tanks to Ruben.

I programmed the shopbot very nicely. I only had to move the Y-axle by 110 mm and zero it again to start with the second Mold. But I had to programm this mirrored Mold first. But is was a copy of the 1 first mold.

Then problems again. I wanted to use the pressed air to blow away the chips. But instead of pressed air a fluid came out of the air gun.

This was very frustrating end I was pretty pissed of, because I needed the pressed air to remove the chips and cool down the wax a bit. The building officer could also not help me fixing this problem. A mechanic had to be called but the resonsable for that was not there anymore.

What I was affraid for was that the chips would melt and stick to the mold surface. This was the case. But also I discovered cracks and holes in the mold. It was 9:30pm when I stopped the shopbot went home very tired.

Early in the mornimg after a short night I was back behind the shopbot, now milling the 3 other inserts molds out of the blue wax I still had. But again t was tricky. I had no spare and the 3 insert just fitted in the blue wax block I melded on Sunday. But I had allready dicovered several creak in the wax. But I had no alternatif, so I started milling. But during the milling parts broke out of the wax.

All the 3 wax molds where broken.

With a hot gun I tried to meld the wax and glue it together. This was a very frustating work and did not work very well.

The damage of the molds was so big that I had to plan the things that had still to be done. On Wednesday the FabLabenschede and the Waag in Amsterdam are closed due to the national Kingsday we celebrate. So If I would repair the molds in the FabLabEnschede and cast the silicone I had to wait again untill in the evening. So at 2pm I decided to move to my home and repair the molds. During the ride home I was thinking about my repair options. At home I decided to try hotmeld to fix the white wax.

First I cleaned the mold very properly with a small brush.

Then appliing the hotmeld and heated a spatula with the heated air blower to melt the hotmeld and wax to flaten the surface

The screwdriver worked also very fine to rework the wax and hotmeld.

After 1,5 hours of work I had repaired all the spots in the mold.

The insert looking very rough. I had finished it with a big 1/4 inch mill to reduce the milling time. Also the depth that had be cutted was to big for the small 3 mm mill.

I glued the the milled molds ontop of each other.

The molds where ready.

It was 6pm when I was ready to cast the silicone in the molds. I calculated the amount of silicone I had to mix. It was 0,845 liters. The 2 containers where applied with 0,9 kg of silicone. So this was tide.

I mixed slowly the silicone and scrabed the side walls clean.

Safety first, so safety glasses and handgloves. And my dad to support me.

Temperature a bit to cold for the casting. It was about 5 degrees in my barn. I don't have heating yet in it. But I did started to poor the silicone in the molds of the inserts first. This had to be done first because I needed them fully filled.

In the bigger molds I have some play to reduces the toplayer thickness (5mm) of the mold.

Slowly filling it from one side tom the other

I had just enough silicone to fill the molds. My calculations where matching.

This silicone ws able to cure faster at 60 gedrees Celcius. So I placed the molds in the oven. But again I noticed that I had troubles with the 18 mm wood. I could not get the heat in the siliconen. But I was facing problems that the silicone was pooring out of the gabs in the molds.

In the right insert the amount of silicone reduced alot. This will give problems with the final casting.

Here maybe the wood is the reason that the silicone has blown up. Again I think I comes from the thicker wood. Later I analysed this in comparison to the melting of the Blue wax blocks for Paul that he had no problems with cracks. But he used thinner 9 mm plywood. Also in my white wax I had no problems getting the heat in the silicone.

With the air gun I was able to get the silicone out of the mold.

Chips of the milling had come loose. I had melted the wax with the air gun and cleaned the mold with hot water and with an brush, but as you see ther where still chips there.

Same in the other mold. Also a lot of air bubbles. I know I have poored the silicone from a close distance. later in my review with Luis Carvao I learned not to do that, but increase the distance to pull the resin so the air bubbles can be taken out of th resin during the pooring.

Cutting the flush away

The inserts placed in the mold.

This gab at the sides was not intended. Mistake by myself. Something I had not designed well when building up this negatif mold in my head during the construction of the positif mold.

At the right side of the upper insert there is a gab that should not be there. Again a small mistake. In the final product I have to cut this rib, that will appear, away. Next time it is better to build the negatif mold in 3D CAD, instead of constructing a positif mold and design the negatif mold in my head.

Detail view of the insert to fit the sensor.

The silicone mold ready for the PU resin.

Reading the instructions. I can not cast the PU resin below 18 degrees Celcius.

But it is only 5.5 degree Celcius. So I have moved to my other barn with has heating. But to heat up a big barn takes a while.

I put the mold above the radiator to head up the mold a bit.

It is now 7pm and the barn temperture still not above the 12 degree Celcius. So I can not cast the PU resin today. I will heat up the barn all night and hopefully tomorrow I can cast.

Thursday in the Morning the temperature in the barn was still not ok. Is was even less then in the evening before. So I turned down the boiler and took the stuff to FabLabEnschede and cast the PU resin there. I did the casting in the chemical chamber to work in a more save enviroment. I read the instruction sheet of the PU resin and it turned out to be very bad for my health.

The resin was mixed in 2 component A and 1 component B. I made something more then I had calculated, so I would not be in shortage of resin, when I would spill some resin. I mixed 2 drops of green colour in the component 1 first before adding the component 1.

I learned from Luis Carvao the keep the bucked, with the resin, very high during the casting, so the resin is then stretched so the air bubbles can escape. But the resin was thicker then espected. The resin didn't flow that easy in the mold. It took its time to flow in the mold and therefor I constantly had to stop pouring. I had not a squirt with me to inject the resin in the mold. So I quickly made a funnel out of paper. Now I could pour the resin in the funnel at a higher position and it had the time to flow in the mold. The mold was angled a bit to make sure the air could escape. The moment that the resin came out the upper air vent is displayed on this picture.

The resin had to cure for 16 hours. So after the weekend I was able to take the model out of the mold.

I released very easy. But a lot of flush is on the final product. Also the bad wax surface is very good visible. There are also the air bubbles visible from the silicone casting.

Getting the insert out of the product was a bit difficult. The reason for that is the bad side of the insert. The reason for that is discribed above, due to the fact the resin poured between the wax molds.

The round gabs for the sonar sensor are missing. So the fit of the insert in the mold was not ok.

I opened the round gabs with a special cutting tool

Assembled the part in the frontsuspension of the leafBot

I used shore A 70 but the PU rubber is to flexible.

I will need a big design change to fix this problem. The PU is not stiff enough and the design of the part must be changed also to avoid this kind of flexibility.

Not everything went well this week, but I learned a lot of it.