Making the silicon mould

Yesterday we edited our last test piece from the cast-able resin. We printed a version which was not digitally restored and then tried to restore it manually with wax to see if this method works better than restoring it digitally. The restoring was done with a soldering iron to melt the wax and then dab it onto the surface. We also tried to scrape away material at the edges with some sculpting tools to make the ornaments on the crown more detailed. The cast-able resin has a soft wax layer at the surface, but is a lot harder beneath the surface layer. It turned out to be harder than we expected to edit the piece.

Restoring the cast-able resin piece with wax and a soldering iron

Dripping the wax onto the model

So now we have four different types of test pieces which will be casted.

Four types of test pieces

After the restoration we started with the production of the wax moulds. To make a piece of the crown from wax, we need a silicone mould where we can pour the wax into. We plan to make the silicone mould by pouring liquid silicone together with harderer in a wooden box. In the box we will stick our four crown pieces with some double sided tape on the bottom of the box. Then we will pour the silicon in and let it harden for twentyfour hours. Then we can break the walls of the box away and we have our silicon mould!

For the box we used a piece of mdf board. To assure that the silicone would not stick too much to the box we used clear tape to cover the inside of the wooden box. We also used sculpting clay to cover all the edged of the box so that there are no sharp corners in the silicon mould.

We found out that the cast-able resin pieces did not stick too well to the double sided tape, probably because of the thin wax layer, so we decided to stick them to the bottom of the box with the sculpting clay. We are not so sure if the pieces are well enough attached to the bottom. If that is the case, then the silicone can slip between the edges and fill up the cavity beneath the test pieces. Hopefully this will not be the case.

Making the wooden box for the silicone mould


The finished silicon mould

We mixed the silicon with the harderer and poured in into the wooden mould.

Pouring the silicon in the mould

Then we put the mould in the vacuum oven. Just to be sure we put the temperature a low as possible. The silicon rose a lot in the mould so it would have been better if we made the sides of the box a bit higher. To prevent the silicn from pooling over we let a little air in the oven bit by bit. This worked fine.

Removing air bubbles with the vacuum oven

There were quite a lot of large air bubbles so maybe some of the silicon slipped beneath the edged of the test pieces.

Inside the vacuum oven

This is the end result. It seems that most of the bubbles in the mould are gone so we are positive that the wax moulds will turn out fine. We let the silicon set over the weekend and on Monday we will start with the wax pouring.



First look at the crown and scanning the model

Today we could finally see the crown in real life! The crown is 14 by 14 by 13 cm large, and quite heavy! It was pretty damaged, just as we expected. A lot of the paint was chipped of and you could see some cracks on the surface. 

Click on the image to get a 360 degrees view!

Top view of the crown


Apparently there was a gold layer beneath the paint, which you can still see at some parts. Sadly we do not have enough budget to have gold layers on our reconstructed crowns, so they will be painted.

Bits of the gold paint showing through

When we looked at the crown from up close, we saw that there were a lot of split lines running horizontally from the yellow parts of the crown.

Spit line where two different parts meet

After a quick inspection, it was clear that the crown was casted from six (!) different pieces. We expected it to be one solid piece, so we were surprised about it.

Layout of the different parts

We are going to try to cast it out of one piece, because we would need to make 48 wax or resin moulds otherwise.

One of the problems of the original crowns that they got stolen easily. The crowns had three legs which fit into three holes in the poles. The were glued together with cement. This was not strong enough. We wanted to try a different approach by screwing the crown onto the pole. We will use the knob at the top to attach the screw thread on. We drill a hole with screw thread through the whole length of the crown as well. To prevent someone from unscrewing the knob we will pour glue through the hole of the crown before screwing the knob on.

Possible approach to attach the crown onto the pole

After we discussed some things about the crown we started with the scan. We scanned it in two parts with the Spider.  In the picture you can see the setup.


Below is a picture of the raw data and the model. We were very happy with the results. The quality of the scan is really good, you can see the little details like the cracks quite clearly and the paint did not turn out to be a problem.

Raw data of the top part

3D model of the top part of the crown

The underside was scanned as well and the final model can be seen here:

Final 3D model with top and bottom part combined

We did it all in one go so the whole process went extremely smoothly. The 3D model can be saved as an stl file, so we will try to retouch the crown as best as we can. We will try to retouch it with Rhino, if that does not work we will try out other programs.

Playing around with the 3D scanner

Before we can scan the crown and make a 3D model we need to practise with the scanners. Maaike arranged for us a meeting with Bertus Naagen, a staff member at the TU Delft who has a lot of experience with 3D scanning. He introduced us to the two scanners they used at the faculty: the Artec Spider and the Artec Eva. We learned that the Artec Spider is used for larger objects which can be scanned from a larger distance. The Artec Eva is better used for smaller objects to scan in a closer range because it can scan more details.

Artec spider

Artec Eva

With help from Bertus we started with a scan of a statue of a penguin to try out the program and the machine. Both Artecs can be operated by hand but you can also put them on a tripod to reduce shaking. Because we wanted to scan every side of the penguin we used a rotating platform with a stool on top where the penguin could stand on.

Set up with the tripod and rotating platform

It took some practice to get used to the scanning, but at the end of the morning we learned a lot and were able to do a scan on our own. We learned that you have to move the tripod with the machine on it slowly. You have to be careful that all the parts are scanned, but the scan should not be too long otherwise the file is too large to process. If the shape is too complex to scan in one go, you can make two scans and merge them together with the software. We did this with the penguin as well, one scan from the top and one from the bottom and afterwards we aligned the two. We will need to do this for our crown as well if we want all the details to be there on the 3D model. After we made our first scan Bertus showed us how to make a model out of the raw data. We also learned some tricks such as erasing the ground from the scan and putting texture on the model. Once you know the basics, the program is quite easy to use.

After the basic instruction we could play around so we made some more scans, for example this one:

Example of a test 3D scan

3D model of Vivians face

Rohan and his 3D model

We also scanned our own faces which was really awesome. We first tried it with the Artec Spider but that didn’t really work so we switched to the Eva scanner.

Hair has a very fine texture and shines a lot so that is why the scanner could not scan our hair properly. Form Bertus we heard that he had some trouble with scanning shiny objects in the past. Sometimes he sprayed dry shampoo on the objects to make it more matte. We wonder if the crown will be too shiny for the scanner because of the paint. Hopefully it will work out.

First day of the project

Today we met Maaike de Vries, our coach for this project. She gave us an introduction on what we are going to do during these four weeks. She told us that we will work together with Paleis het Loo to reconstruct royal monument the Naald in Apeldoorn. It was built in 1901 as a gift for the marriage of queen Wilhelmina and Prince Hendrik. In 2009 the monument was mentioned a lot in the news because an attack on the royal family was attempted there during the national Kings day. The monument is obelisk shaped and has a round gate with eight poles surrounding it. Below you can see some original pictures of the monument from the opening in 1901.

Opening in 1901

Detail of the gate

This is what the monument looks like now.

Monument in 2009

If you look closely you can see that the crowns from the original picture are missing. Around 1950 so many of the crowns were stolen or damaged that Paleis het Loo decided to remove the remaining crowns and include them in their collection. This year Het Loo wanted to give the monument back its crowns by making eight brand new pieces which are reconstructed from the original crowns and they asked us and Maaike to do so! We are all very excited to start working on this project.

The original crowns were made of cast iron. We are going to cast the reconstructed crowns as well by a company in Groningen called gieterij Borcherts. They do investment casting, a process where you have a replica of the object that you want to cast made out of material that will melt. The mold is created by putting sand around the replica and then heating it up so that the replica melts away. Iron is poured into the created cavity. When the object is cooled down the sand will be brushed away.


Explanation of investment casting

Because casting eight crowns will take a lot of time, we will not be able to finish them before the project deadline. However, we can still cast some test pieces to try out which casting method works best. For now we want to try out three moulds:

  1. A wax mould made by pouring wax in a silicon mould. The silicon mould is created with a 3D printed crown made out of PLA

      2. A 3D printed wax mould made out of wax filament which will be printed in the Ultimaker 3

      3. A castable resin mould which is printed on a Form2 machine

To make the moulds we will make a 3D model of the crown and retouch it digitally.

We heard a lot of information today and we are excited to start working on the project. For now we will figure out how to make a 3D model of the original crown so stay tuned for another blog post!



Hello and welcome to our blog! We are Tom Salden, Vivian Vriend and Rohan McEvoy and we do the minor Advanced Prototyping at the TU Delft. In this minor we get the opportunity to work on a four week long project about advanced production techniques. Our project is called: Reconstructing a royal monument. In our project we will explore the possibilities of 3D scanning, modelling and printing combined with traditional, historical cast-iron techniques to restore a monument in the Netherlands. In this blog we will keep you updated on our progress, and hopefully make you more enthusiastic.