The year 2020 turned out to be a year of COVID. My flights to back home in Dec 2020 got cancelled. So spend a part of the holiday to take my art of casting a bit further. In the end, I could have a ring made out of brass. Carrying forward the experience from last year, this time, my lost wax process was smoother, but still I came across quite a lot of unexpected subtle things. I am documenting some things that I learned in each of the parts of the process. Hopefully, I continue with this hobby and hopefully get better at it.
Here is my final result for this year:
- Designing the model in CAD
- 3d Printing
- Silicone Mold Making
- Wax Model making
- High temperature mold making
- Metal melting
- The pour
Designing the Model in CAD
The first step in this process is designing the model. I used OnShape (a general purpose CAD system for mechanical parts). I had picked up a bit of how to use a CAD system last year. That skill came in handy. This year, I spend least amount of time designing the model. In the future, I wish to spend more time designing and possibly also to make use of a more jewellery oriented CAD system, like RhinoGold or Gemvision’s MatrixGold or 3Design.
In Onshape, I do have a handful of designs. However, for this year I make use of a really simple design. The idea is to setup my tooling and realize a brass ring.
Once we have a CAD model designed, the next step is to print a 3d model of it. I purchased the Anycubic Zero resin printing system. Unlike a 3d printing which prints with plastic spools, this one prints with a liquid resin. Compared to a FDM 3d printer (the which prints from plastic spool), the resin printer is a lot harder and nastier to operate. Thankfully there are several Facebook user groups with members very willing to help.
In my previous attempt, I did use an FDM 3d printer to print the model. The major advantage of the resin printer is that it can print very fine details in the model, atleast 5 times intricate models than the best of the FDM printers. Ofcourse the FDM printer can be used for jewellery making, but the resin printer is a better match. Disadvantage of the resin printer is the smaller build volume and nastier chemicals involved.
Initially I tried to use the slicer from anycubic. Although it is usable, but it is not good enough. Particularly its auto-support features are very bad. I hardly could get a successful print from the default settings. Once I watched a few youtube videos on, ‘how to manually design supports for resin printing’, I was on a roll. This youtube-channel was particularly helpful to understand the fundamental concepts of designing supports. Also, notable is the Anycubic photon usergroup : https://github.com/Photonsters. I found this group has a better documentation on the printers than anycubic itself. PhotonValidator was a particularly helpful software to detect potential supporting issues. After a bit of trying and failing and more trying and more failing, I was able to get successful prints.
In the end I used the Lychee Slicer (an advertise supported software) and PhotonValidator. Both these tools can be used on Ubuntu. With this, I usually get a successful prints on my photon zero. These tools generate a .pw0 file (A file that the printer can execute). Once the printing was done, I used the wash and cure station to wash the model with Isopropyl Alcohol (99%), followed by UV curing. Both processes for 10min each.
I used the basic resin from anycubic. This resin is not suitable for burnout. However, there are wax like resins available (a bit expensive, atleast 50 USD for 250 ml) which are suitable for burnout process. In the future, I am interested to buy one of the castable resins and do a burnout directly on the 3d printed model.
Silicone Mold Making
In my earlier attempt, I had made a home made silicone model making compound by using the silicone sealent and corn starch. This time, I am making use of more professional silicone mold making products from Smooth-on. There are several kinds of silicone products for mould making, however I prefer the kind which involve two liquids and upon mixing these liquids solidify.
I used Smooth-on’s mold star 30 (a platinum cured silicone rubber) to make a cut out mold of my ring model. This youtube channel demonstates several of the nuances of mold making with silicone. In retrospect, I feel I should have used the tin-cured silicone rubber. The reson being, that there is a cure inhibition (the silicone does not set) with the resin models. Although using an clear acrylic laquer (the same one used by artists to protect their paintings) alleviate the problem.
In the end, I was successful to make molds (after coating my resin model with the laquer) of my 3d printed model with platinum cured silicone rubber. In the future, I am interested in trying out other products from smooth-on particularly,
– Mold max NV 29 (tin cured, low viscosity)
– Mold max 27T (tin cured, transparent, but needs vaccum degassing).
My molds still had some air bubbles. For the future, a vaccum degassing unit is on my wishlist. Also note that the silicone rubbers are sensitive to room temperature and humidity. I have a room thermometer and hygrometer setup. Usually 23 deg C and 50 % humidity works well. Colder rooms below 18 deg C are bad for silicone curing process.
Making a Wax Model
Last time, I had created by own wax suitable for wax model making (using parafin wax, sterin and glue sticks). This time, I am using a freeman injection wax. In my observation, the difference between the home made wax and the real injection wax is that, the injection wax absolutely does not stick, is very flowy and solidifies very very quickly. Just pouring the wax is not sufficient. A pressure mechanism is needed for the injection wax to go into all the intricate corners of the mould. Also the injection waxes are micro-crystalline wax (and not parafin wax), with added release agent, flow enhancing agents etc. It would be very very hard to replicate the recipie for a DIYer and the injection wax prices are not too bad (costs about 10 USD for 500 gm).
This time I am using a metal springe to inject the wax into the mould. I purchased this springe from aliexpress, under the name of BBQ marinade injector. This was suprizingly hard to purchase locally for me. I would say, this more or less worked for me. I made additional sprues using bees-wax.
In the future, I am interested to give a shot at building my very own wax injector. It could also make for a good youtube video as fewer people have attempted anything like this. Also there are several wax injectors available from China in 100-150 USD price range. I could also test those out at some point in the future.
The core principle on which a wax injector operates is the pascal’s law and incompressible liquids. The idea is to have an air tight container (similar to a bike tire but made of metal). This container will contain molten wax and air under pressure. Can use a simple bike pump and a gauge to monitor pressure. At another hole in the container will be a nozzle from which wax will jet out. A push nozzle is available for purchase from China (on aliexpress), however it is a bit expensive for my comfort.
High Temperature Mold Making
Next is to make a plaster mold which is suitable for pouring liquid metals. Last time I used plaster-of-paris (construction plaster) to make these molds. The issue with the construction plaster is that, it has a larger grain size (so not suitable for even moderately intricate patterns), cracks at higher temperature (I dare not to pour things like Brass into it, low temperature metals like solder metal or Pewter are still ok for it), gives bad surface finish, flashing (internal cracks in which some metal flows into).
This time, I switched to a proper jewellery grade investment powder. These powders act similar to plaster-of-paris in the sense that it is initially a white powder, once water is added you have about 10 minutes and then it solidifies. The difference is that, these powders contain other refractory materials to withstand the high temperatures and have a much smaller grain sizes. These powders are suitable for non-ferrous metals (eg. brass, bronze, gold, silver etc.). There are other powder formulations available for steel and for platinum. These powders are not too much more expensive than plaster-of-paris. Mine cost me USD 75 for 22.5 kg. However it is possible to get it cheaper than that. A 22.5 kg bag of plaster could cost about 15-30 USD.
I used the Eurovest Investment Powder. It needs 100gm of powder for 40 ml of water. The website of the manufacturer (SRS Ltd.) has the technical datasheet which gives more details on this powder. There are also other powders available from SRS Ltd.
I was extremely pleased to see excellent surface finish of the final metal casting with this powder. However this powder contains crystalline silica which causes respiratory allergies, so a good dust mask (FFP2 or FFP3?) is needed to safely use this powder. I would definately want to find out the most suitable mask for this kind of powders. Air bubbles is yet another problem. Currently tap the flask to remove some of the air bubbles. In the future, I am looking to get a vaccum degassing unit.
I think, my burnout process was the weakest link in the chain. At the end it caused a lot of issues. I did not improve much from last time in this part. Only difference is that, last year I did not understand the whole point of steam de-waxing. I thought then, it was simply not needed. However the thing is with steam de-waxing you can get rid of say 95% of the wax. This will cause significantly less weird smells in the burnout process.
My burnout process this year was>
– Steam de-waxing for 1 hour
– Cooking oven burnout, raising the temperature from 110 deg C to 230 deg C in about 1.5 hrs. Followed by hold 230 for 2 hours and then 275 deg for 1.5 hrs.
To be 100% honest, with my burnout setup I got a lot of failed castings. Of the 10 castings I attempted, only 1 was success, 3 were partial success, rest failed pretty badly. The prime reason is that the melting temperature of Brass is about 950 deg C and the casting temperature is atleast 600-750 deg C. So, as soon as the brass hits my molds, it solidifies before reaching all the corners. My setup is however sufficient for low melting point alloys like solder metal, pewter etc. Also, higher temperatures are needed to fully harden the high temperature mold.
For steam de-waxing I used a hot plate, ikea steamer insert and a steam pot (covered). This actually works great. It usually can get rid of 95% (a guess-estimate) of the wax in an hour of steam de-waxing. The wax goes into the water in the pot, but does not smoke, which is great for a small home setup. If you try to put this directly in an oven, you get weird smell of wax burning, which is not very desirable. Be sure to put the flasks inverted on the steamer plate so that the wax flows out into the water underneath.
The investment powder I used suggests the following burnout schedule. Unfortunately I did not have the means to do it this time. Building a small electric kiln with firebricks, ceramic wool and kanthal wire is definately on my list. There are several youtubers who have built an electric kiln. I hope to build my kiln soon and hopefully improve my castings.
There are two ways to go about melting a metals (like Brass, Bronze, Silver and Gold), viz a gas torch or a kiln (arc furnances is another way, but is way beyond the scope of a hobbist). All these metal alloys have liquidus temperature of 1000-1200 deg C. Kilns are better suited for larger quantities (atleast 200 gm+), while gas torches are better for smaller quantities (say 10-150 gms).
Last time I had used a handheld propane torch (even tried with 2 torches together) and was able to melt Aluminium (about 30-40gm, melting point 660 deg C). However, I had failed to melt Brass in sufficient quantities. The marketing on a lot of hand held torches report the flame temperatures (2000 deg C), however real number of relevance is the working temperature. Propane torches usually have a working temperatures of 600-800 deg C.
Having an oxy+fuel torch will raise the working temperature significantly. There are several oxy+fuel torches on the market. There are mini oxy-fuel torches and professional oxy+fuel torches (like the smith little torch, swiss torch, sievert torch etc.). The mini oxy torches are marketed as brazing and repair torches. Compared to professional torches these are more beginner friendly and inexpensive. If you use them a lot, the mini-torches are expensive to operate. However they are a good starting point.
I used a mini-oxy-fuel gas torch kit (my set cost be about 150 USD). My torch has a 1 liter oxygen tank and 1 liter fuel tank (mixture of butane, propylene and propane). The flame temperature is 3100 deg C and working temperature is rated as 1250 deg C. With my torch, I was able to melt about 20-30gm of brass on a crucible. I believe it can also melt bronze, silver, gold and copper. However, don’t expect it to melt steel.
It is important to throughly read the safety and operating instructions before using such torches. These can do a lot of damage (and potentially kill you), if incorrectly operated. However, if correctly used they are pretty safe. Here is my check list when I use these torches:
– Make sure all the valves are in off position
– Connect the fuel tank. As you connect there is some momentary hissing. However it should not hiss continously.
– Connect the oxygen tank. My torch has a left-handed threads for the oxygen tank.
– Check for leaks. I used soapy water and a brush to apply this water at the joins. Make sure there are no leaks before proceeding
– Open the fuel valve a bit and ignite it.
– If the flame is good, open oxygen value a bit. There are 3 kinds of flames in oxy-fuel torches. Neural, Reducing and Oxidising. You want a neural flame.
I used Borax to glaze my crucible (so that the melted metal does not stick to the crucible) and also as a melting flux. It helps to avoid oxide formation. Beware that borax is a somewhat toxic chemical.
I am generally quite satisfied with my torch setup for now. However, I wish to acquire the swiss torch sometime in the future (costs 600-700 USD for the entire system). The advantage of swiss torch is that I will be able to work with platinum with it (still a very distant dream). However for larger quantities of metal, I will be better off with a kiln.
The Metal Pour
Once you have the burnout process done and the metal melted, the next step is to pour the metal into high temperature mould and then quench it in water to get your metal part. I did a simple gravity pour. This is generally produces good result for parts containing not too fine details (assuming a good burnout). However, to get the metal into intricate parts, people generally use either a) centrifugal casting or b) the vaccum casting technique.
I am keen to purchase a neycraft centrifugal system (a standard in jewellery industry) in the future.
In this project, I did not attempt polishing. However, in my precious attempt, I had attempted polishing and got pretty decent results. Next time I attempt this, I want to get a mirror finish on the metal.
Once you have the casted part ready, one would usually cut the sprues and sand it a bit. However, with my investment powder the surface finish was pretty smooth. I did not see a need to sand it.
For polishing you need
a) some kind of rotary tool (like the dremel) and ideally a flex-shaft,
b) a coarse brush (like the goal hair brush)
c) a buffing brush (like a cotton-felt)
d) a super soft brush (like a cotton mop)
e) polishing compounds (like the set from Dialux).
After every step of the polishing process, people wash the metal part in an ultra-sonic cleaner. People also report to get good result by using simple soap, water and a tooth brush.
Professionally people also use a polishing wheel to get the job done. But for lower volume it is totally unnecessary. Although I would love to have a polishing wheel, it is quite at the bottom of the priority list. I would rather invest in a good quality flex shaft first.
I really only have bookish (or rather youtube knowledge) on stone setting. Never yet attempted it. Once I master the craft of metal casting I would spend most of my attention to stone setting. Not happening anytime soon.
There are lots and lots of types of stone setting. The simplest one being the bazel setting (also called flush setting). It needs a few flex-shaft burs (atleast drilling burr, ball burr and setting burr) and ofcourse the stones. It is possible to buy cubic zirconia (CZ) in transparent and in color from Aliexpress for pretty cheap. Once I have some success at stone setting, buying a stone setting microscope is on the horizon.
Something New Things I Learnt from this Attempt
- Operating a resin 3d printer. resins, FEP cleaning and safety
- Cure inhibition in platinum cure silicone
- Injection waxes are micro-crystalline
- Investment powder chemistry
- How to build a DIY kiln
- Brass and zinc fever.
Cure Inhibition in Platinum Cure silicone: These kind of silicone are very prone to cure inhibition. Some of the common materials which cause cure inhibition: modelling clay, sulphur containing materials, latex, polyster, epoxy, UV resin etc. Usually one should try out a very small quantity on the material before the project. One option is to use InhibitX which is a product from Smooth-on which is similar to laquer. However, in most cases using acrylic laquer works just as fine. Another option is to switch to tin cured silicone rubber. Although this kind of silicone is not so prone to cure inhibition, it has a smaller useful life.
How to build a DIY Kiln: Several people have built their own kiln on youtube and successfully melted metals, heat treated steel parts, made pottery etc with those. A good idea to adapt one of those designs. Although a basic kiln can be purchased for about 500 USD, the ones with ramp controller will cost upward of 1000 USD. In general it seems a good idea to DIY a kiln.
Supplies needed might be: a) kanthal wire (heating element, should be resistant to atleast 1300 deg C, b) k type thermocouple (rated to measure atleast 1200 deg C, c) ceramic terminal blocks, d) triac relay (should be able to handle 15A @ 240V), e) firebricks (rated for 1200 deg C), f) ceramic wool g) micro switch (to turnoff the kiln when the door is open). A basic Temperature controller module is a good 1st step (is cheap but has not ramp functionality). If a basic kiln works, it would be great to build a DIY controller using arduino with ramp functionality.
I especially liked the video and a short technical discussion on Kiln by Robert Murray-Smith (https://www.youtube.com/watch?v=lCQBsPCCqcY )
Brass and Zinc Fever: Brass is an alloy of Copper and Zinc. Brass is not the best metal for casting. Ideally you want to cast Bronze, Sterling silver, or Gold. As we try to melt brass, we get very near the boiling point of Zinc (904 deg C). So as we melt brass, we quite often see a green flame. This is actually the zinc evapourizing. This vapour/smoke is toxic to humans and known to cause Zinc fever. Although it is not fatal, it will cause sickness for a few days. Thus, it is recommented to atleast switch to Bronze (Copper + Tin).
Wish List of Equipment
- Castable resin
- Neycraft Centrifugal Casting system
- Stone setting burrs and tools
- Wax injector
- Vaccum chamber
- Better FEP for the 3d printer. Like the non-FEP.
- Swiss torch set