My Dream
I guess I am obsessed … Since reading a book called “The Stone and the Flute” by Hans Bemmann, I am dreaming of building a flute made of silver (Sterling Silver). In this blog I will post my progress until I will get there.
I have worked with wood my whole life which made it easier to build my first native american style flutes. Meanwhile I am pretty happy with my design and the flutes sound very nice.
But I have never really worked with metal and need to start from scratch. This journey will be exciting but probably also frustrating from time to time. So please be patient with me and let’s learn everything necessary to end up with a great sounding flute made of pure silver.
I will share every step of my way, so you can learn with me and possibly skip some of my mistakes …
My target is to make a flute in Gm4 pentatonic. This way the flute won’t be too big/heavy and I already have some experience with G flutes using a 20 mm bore diameter. The flute will be something around 500-550mm in length, at least if I look at my wooden examples. I don’t know yet if something will require to make it longer … but I guess not.
My goal is to have the final silver flute in my hands latest at Christmas 2023. Let’s see …
The Copper Plan
The material for my silver flute will be pretty expensive and I am planning to build only one. Therefore I am practicing with cheaper copper. Copper is a lot less expensive but should have similar properties. This will allow me to figure out:
1. The overall design
2. The hole positions
I have ordered 4 tubes with 20mm diameter and 1 mm wall size which equals 22mm outer diameter of the tube. The tubes measure 500mm in length and cost around $5 each.
The difference to silver will be hardness and weight. On “Mohs Hardness Scale” silver is at 2.5 – 3 and copper at 3 which should be relatively similar. The weight is also relatively close at 8.96 g/cm^3 for copper and 10.49 g/cm^3 for silver.
I can already tell that the flute will be pretty heavy when holding a copper tube. I guess more than 1mm wall size will not be feasible if I want to play the flute comfortably.
The tube seems very straight forward to be used for the bore. The only hard part will be the mouth piece. But I will deal with this later. The bigger issue is that I have no good idea how to make the sound/nest area. I could cast it or use a block of material and saw/file it to the desired shape. I decided for the “casting” method for now and ordered a 5mm thick copper plate with the dimension 100mm width and 100mm length which should be big enough no matter how the nest area will look like. I got 4 of them because I am sure I will make mistakes and the number of 2-3 iterations has proven to be a good number for me.
The Plan
My plan is to make the flute out of three parts:
1. Mouth piece
The mouth piece will be modelled in wax and a mold created. I will then cast the silver. The piece is not too big and therefore smaller equipment should be needed.
2. Sound Area / Nest
Same idea as the mouth piece. I will create a mold and cast it.
It might make sense to cast the complete mouth piece / nest area in one bigger piece to simplify the soldering to only the bore. I will think about …
3. Tube / Bore
I found suppliers that offer silver tubes with 20mm diameter and 1mm wall thickness (22mm outer diameter). This should work perfectly for the bore.
Soldering:
The three parts need to be soldered together. I hope the seems will look good … I guess I will have to practice a lot.
To practice the molding / casting part, I am planning to use tin to practice the process. Tin melts at a much lower temperature and is much cheaper than the silver. Regarding the seems and soldering I will practice on small silver parts that I can reuse when casting.
Casting Process & Tools
Here are my current ideas on the casting process and all needed tools. I have no idea if this process will work and if the tools are right but I am sure I will learn …
1. “Wax” for “Lost Wax Casting”
To create the actual shape of the mouth piece and nest area, I will use a wax based material to create the mold and cast the pieces using the “lost wax method”. One option would be to carve/file a wax block to the desired shape. But because I love high tech, I decided to use a 3D printer to print the models. I found a certain type of filament that can be used with most regular 3D printers. The wax material is supposed to “disappear” when I burn the ceramic mold around the “positive” shape to create the “negative” form.
The brand of the filament is called “PolyCast” and the maker is “Polymaker”. Here is the link: click here
I am using the software Fusion 360 (click here) to model all the parts in 3D. To print the model with my 3D printer I am using the free “UltiMaker Cura” software (click here).
Here is the 3D model file (only mouth piece and a long bore) of the flute: click here
2. Investment
The powder to create the mold is called “investment”. This powder is mixed with water to create a “thick paint” that is used to cover the “positive” wax model that I have created in the last step. After drying the covered wax model it is burned in an oven, the wax is burning away and what’s left after the burning process is the negative form / mold that can be used to pour in the liquid metal. Unfortunately the “paint” will contain a lot of air after mixing with water which should ideally be corrected in a small vacuum chamber to create a “paint” that is as smooth as possible. I decided NOT to invest in a vacuum chamber in the first step and see how far I can get regarding casting detail if I just live with it. I suppose I won’t be able to create very fine surface details (rings, coins, etc.) but the two pieces (mouth piece, nest) are not very detailed and I might get around the cost of the vacuum chamber.
I have chosen the following investment powder because it is available in Germany: click here
3. Burnout
After the investment material has dried around the wax model it needs to be burned. First to get rid of the wax inside and second to harden the investment (ceramic) mold.
I found a decent priced oven that should be able to do the job It is rather small but the everything I want to make incl. some other projects like rings etc. should fit. The important part is that it can be programmed to heat in certain temperature cycles which is what the investment supplier specifies for the burnout process.
Here is the link to the oven: click here
I have ordered it a few days ago and it should arrive today. I am very excited to try it out!
4. Melting the metal/silver
In order to melt the silver I need a temperature of around 1100 degrees celsius. I guess this should work with the burnout oven but there a special small melting ovens that make it easier and that are made for this task. I chose a cheap one and hope it will do the job.
Here is the link to the melting equipment: click here
5. Casting
If everything works out the way I imagine it, the last step will be to pour the liquid silver into the mold, cool it off and that’s it! In theory it would be good to use vacuum once again to make sure there are no air pockets but like I said I will try to avoid this additional equipment. I guess if the model / mold is designed in a smart way, the level of detail will be ok for my purpose. I might be totally wrong … We will see!
6. Soldering
All pieces need to be soldered together. This step is crucial for functionality but also for the looks because of the seams that I want to be “invisible”. I suppose this step needs the most practice. I am using a gas based jeweler torch for this step. Its a very cool little device that I am also using for my jewelry work.
Here is the link to the torch: click here
Beside the torch I need “silver solder”, “soldering flux” and some “sodium bisulfate” to get rid of the oxidation after casting.
All the missing stuff is now ordered (08/11/2023) and should arrive during the next days. I will keep you updated …
Burnout Oven
In order to burnout the investment mold later on in the process I need a burnout oven. I decided to buy a small oven that is big enough to be used also for jewelry projects. These ovens are very expensive and big if their purpose is ceramics but I found a reasonable priced one that seems to be made for “my” purpose.
My requirements in choosing an oven were:
1. At least 11oo degrees celcius (2000 degrees fahrenheit).
2. Programmable cycles because it seems like the investment needs a certain heating procedure.
3. Big enough for the flute mouth piece, flute nest and for my other jewelry projects but not too big for my garage 😉
4. Not too expensive.
It came yesterday and today I want to set it up. It need to go through an initial burnout to prepare the material inside the oven. This initial step is important to make the oven last longer. One issue that I had was the fact that the product was only available in 110V (US standard). BTW can’t we change to ONE standard? Anyway … I needed to buy a step-down transformer that has enough power (I hope) to run the oven with 220V in Germany.
I found a video on youtube that guided me through the steps of programming the burnout cycles (click here). As I am writing it is heating up and smoking a lot. Scary … It will take 70 minutes until it is done … I am doing all this in the garden and have a fire extinguisher close by!
Meanwhile the initial burning cycle is done and my house is till standing … It smoked but did everything as expected. It looks a little burned inside but I guess that’s what to expect. I will now store the oven until I am ready to use it for the mold burnout at some point. I am very excited that the first step is done and can’t wait for the next ones!
Melting the Metal
I guess I could use the burnout oven to melt the metal … But too late! I had already ordered an electrical melting pot that is made for this purpose and will hold a bigger flask to melt the silver (or copper). I found an affordable Chinese made pot that was reviewed positively online and is supposed to do the job. Here is a link to the maker with the misleading and name “ToAuto Tool” (click here).
I turned it on and it heats up! Cool! We’ll see how long it will l(c)ast …
The more Youtube videos I am watching, the more I am also getting excited to do other projects with the equipment. I will definitely cast some silver/gold rings at some point. It will be fun to design them in a 3D software, print and cast them Exciting!
Casting the Mouth Piece – 1
I have printed the first piece in PolyCast today. I have used my standard PLA printing settings but it came out great! The measurements are like expected with a 20mm inner diameter, 1mm wall size and 8mm blow hole. I have no idea how well this thin wall thickness can be casted but I guess if detailed rings are regularly casted, this should work too!? I will create a mold in the next days, melt some pewter – which has a much lower melting temperature – and will hopefully get a first feeling for the process. Stay tuned …
Making the Casting Tree
I am learning … In order to cast the right way I need to build a casting tree. The pieces to be cast will be attached a kind of wax “stem” that acts as a channel for the liquid metal during the casting process. I couldn’t find a ready made wax stem to order. Only molds for the wax the create them myself. But they were around $100 which I find very expensive for the purpose. So I decided to design and print my own. Here is a picture of one side of the mold. If it works out I will be able to create lots of “stems” (sprues) and also use them for jewelry making. I will print it today and see if it works … Printing time will be around 10 hours for the top and bottom!!!! Luckily I can use it for a long time … Here is the Fusion 360 model in case you want to print it yourself: click here
Making the Wax Tree
I feel like I have no idea what I am doing!!!!!! I used my stem-mold to create some stems and I guess they came out ok. I have used one and some thinner wax wire (3mm) to create my first funky looking tree. I added two pouring support channels but I have no idea if that is even necessary. I will create a second tree without the extra channels to compare the result. Here are some pictures of the process:
And now? I understand that I need to pour the investment around my tree to create the mold. To do that I need a base for the tree that fits perfectly on the flask. I decided to print one …
Creating the Mold
My goal today is to create a mold to practice with pewter. I will mix investment, pour it around the wax model-tree, let it dry and burn it.
The first step is to mix the investment. I got X-Vest investment powder which is asking for 320ml of water for 1KG of investment powder. I guess this means a factor of 0.32. I used around 300g of powder and therefore 96ml water. It is recommended to use a mixer to thoroughly mix the powder and water but I don’t have one right now. So I guess my first try will be suboptimal but at least it will allow me to learn the benefits of each refined step later. For now I will also not vacuum the investment neither before or after pouring into the mold. I will see how well the piece turns out and potentially vacuum the next time. If I am making finely detailed rings it will probably be necessary. I took around 10 minutes to measure the powder / water, mix it and pour it into the flask that contains the wax tree. The instructions say the investment needs to dry at least for 2 hours … I am excited to take off the plastic base and burn this thing! Here are some pictures of the steps:
One big lesson learned today is that I will have a lot of wasted metal because of how I designed the tree and especially the stem diameter and length. I will have to optimize this or the flute will get even more expensive …
The flask is now in the burnout oven. I programmed two steps: 60 minutes from ambient temperature to 800 degrees celsius (1475 degrees fahrenheit) and then 90 minutes at that constant temperature (800°C / 1450 °F). Following the BlueCast instructions (supplier of the investment) this is the “fast burnout” mode. I can see the glow in the oven through a small hole on top and the display says it reached the target temperature. Cool!
The oven worked perfectly and automatically cooled down after the 150 minute burnout process. It took around 30 minutes until I was able to touch the flask and check it out. The wax seems to be completely gone which is good news. Unfortunately there are some hairline cracks that are visible. I don’t think they will influence the casting too bad but they shouldn’t be there. Next time I will use a mixer, vacuum the mixture and use the slower burnout process. Anyway … tomorrow is the pewter casting day!
Proof of Concept – Casting a Pewter Mouthpiece
The target today is to have a first cast mouth piece made of pewter. To get there I will heat up pewter to a melting point of around 170–230 °C (338–446 °F). At the same time I will heat up the mold to the same temperature with the tabletop furnice to reduce the temperature difference when casting. The rest should be simple … Pour the liquid pewter into the mold, cool it off in cold water, break the mold and discover the result … Stay tuned!
Eureka! The phase is done!
As you see in the above image the plan worked out better than I thought for the first attempt (I think). The mold internally reflected the shape, the walls are 1mm thick, the overall surface looks good and you can even see the horizontal structure (layer height) from 3D printing the model and the material reached every part of the model. I could now saw the piece off the support structure, file and sand it and polish the part. But that wouldn’t make sense with a prototype pewter model … I will keep it as a memory of “The Beginning”.
Phase 1 – Lessons Learned
This first phase was fun and I have learned a lot:
1. The wax model needs to be as perfect as possible. I will rework details and also “polish” it to make surfaces as smooth as possible. I got a “Polysher” made by PolyMaker that uses alcohol fog to smooth the surfaces of the model. When I consider making rings I guess my standard 3D printer will not be good enough to print fine details. I might have to invest in a resin printer. But for the flute I should be fine.
2. I have to think more about the structure of the wax tree. The objective should be to have as little support structures touching the piece but still make sure the liquid metal gets everywhere. On the other hand I also need to reduce the “waste” especially when using the expensive silver later on.
3. I will put more time into creating the investment. I got a cheap electric mixer and a basic vacuum pot. This will help in getting the investment fluid very smooth. I will still try to live without vacuum when pouring the metal into the mold. The vacuum pot will not work for that step and an appropriate machine is pretty expensive. Another change will be how I burnout the mold. I will use the slower burnout cycles as recommended by the supplier (see image below).
4. I have to be more patient 😉 I will use the slow burnout cycle when burning the mold. It will take more than 6 hours but might be worth it to make sure the mold does not have any cracks because of fast heat-up.
I guess that’s it for now. Overall I am very happy with the first try but stay tuned for Phase 1 – Second Iteration!
Polishing the Model
In order to get a very smooth surface for my wax model I have tried to use a machine called Polysher by PolyMaker (click here). This little machine is simply using alcohol-dust / fog to evenly distribute a thin layer of alcohol on the model to treat the surface and make it more smooth. You can see it in action in the following short video. I have used pure (99.9%) isopropanol for 30 minutes. The result is surprisingly good! I can’t feel any printing / layer structure on the model anymore when touching it. This should improve the quality of the cast very much. I am not sure yet how much it changes the model when it comes to details but in case of the mouthpiece it should not matter too much. For delicate rings that might be different.
Tabletop Vacuum Machine
It seems like the ideal solution for casting is a tabletop vacuum machine that allows to vacuum the investment on top of the machine and also support sliding in a perforated flask to vacuum the mold from underneath and sides during the pouring of the liquid metal. Here is a picture of a model that seems to be the standard at least for most hobbyist. It is made by Arbe USA.
I am sure its a great machine and I would love to have one BUT it is relatively expensive (> $1000), very heavy and I couldn’t find one in Germany. Shipping from the US would cost around $400 and customs would be a nightmare.
So … I had to find an alternative plan. My goal was to have the EXACT functionality, no steel soldering because I have never done this and last but not least cheaper to get.
Here is my solution. At least in my head:
1. Use a vacuum pot incl. vacuum pump like the following which is readily available for < $200.
2. Replace the glass lid with a steel lid that has an opening the size of the the flask diameter just like the Kaya Cast machine does.
This would allow to the the same vacuum pot for investment treatment when using the original glass lid and for pouring using the steel lid. I am making a mistake here? This should be 100% equal to the expensive machine regarding functionality assuming the same quality vacuum pump.
The only problem could be that the standard vacuum pump is not made for super hot air. But I will try and potentially get another pump.
The biggest issue was the perfect steel lid but I found a service that MAKES them after your specification for <$50 !!!!!! I ordered one with an outer diameter of 250mm which resembles the original size of my vacuum pot-lid and an inner diameter of 65m which is 5mm bigger than the flask I have started using for all my castings. The key was to find a pot that has the compressor connection on the side and not in integrated into the lid.
I can’t wait to get the metal disc and find out if my thinking was wrong . Here are two images showing the idea:
Improving Casting – The Mouthpiece
After the above learnings I a planning to cast 4 more mouthpieces to practice more. I have printed and polyshed the models and I am preparing 4 different ways of attaching the piece to the sprue to compare how many wax channels are best and where to attach them.
Another new experience will be to use different material. I ordered some bronze pieces as seen in the picture. Bronze needs a much higher temperature to melt (996° celcius / 1825° fahrenheit) and therefore the process should be closer to the final silver casting process than my pewter experiments (see melting points table). Each of the bronce bars weighs a 100g and should be enough for 2 mouthpieces. The material was even cheaper than the pewter and might be the best metal to practice with. I guess I should build a complete prototype flute completely made of bronze.
And if I completely mess up and never make a complete flute, I will at least cast a bronze sculpture of someone. Maybe myself?
I have tries two different ways to get the metal into the mouthpiece. One channel from the sprue and two on opposite sides of the piece. I have no idea if that even makes a difference. I will see how they turn out and draw my conclusions. The first tree had channels on different heights and turned out pretty well but this time I find out if that is really necessary.
Casting Bronze
Today is the day!!! I will cast a mouthpiece made of bronze. The mold is currently in the burnout oven and should be done in 2-3 hours. I will start to melt the bronze in the melting pot about 30 minutes before the burnout cycle is done. This will ensure that the mold is still hot when I pour the bronze.
I am a little scared of the superhot metal (> 1000° celcius) … so hold your thumb!
Meanwhile I have watched a lot of videos about jewelry making. I am getting more and more interested in casting some rings and set stones. BUT first my dream flute!
It’s the next day and I am still alive!
The filled mold has been sitting in a water bath since last afternoon to soften the investment to scrape it off in the next step. I am excited to see how it looks but I feel that the bronze was not hot enough when I poured it. It was at 1000° and looked ok but next time I will add a 100° to allow it to slightly cool off when I get it out of the melting pot and move it a few seconds. It also loses some temperature during pouring. It might have worked but since my model has very delicate structures (thin walls) the metal should be very liquid during casting. I will post a picture of my current attempt, even if it is messed up. This is all about learning!
It feels a lot like opening an “Übrraschungsei” when I was a kid … I guess I am still a kid, just with different “Überraschungs-eggs”.
What a surprise! It worked out very well!!!!!!!!!!!!!!!!!!!! In the picture you can see the model right after getting rid of the investment (which took quite some time). I am very happy with the result and the only thing that I have learnt is that one channel from the sprue is enough and will make post-processing much easier that multiple metal channels to get rid off. I don’t know if the color pattern looks good or if any professional can see some flaws!? Anyway I am happy.
Sometimes it feels good not to have to learn too much 😉
I will not saw off the pieces because it wouldn’t be worth the effort but I will polish it just for the fun/looks of it.
Tired of Mouthpieces 😉 – Hello Sound Area
Now that I feel I understood the mouth piece, I will turn my focus on the sound area. The most important part of a flute when it comes to sound creation and very delicate when it comes to sizes/angles/etc. Let’s see how it goes … Considering the current learning speed, Christmas still seems possible and Santa will be proud of me. He might do some Native American Dance for me on Christmas Day when I play the flute for him? Stay tuned …
The Sound Area – Tree
After printing 2 sound areas, I will cast them in the next days to find out if the more delicate areas like SAC / TSH will turn out well. If they do, I will focus on the final design of especially the SAC interior which is not really yet designed. I need a closed room between mouth piece and SAC exit that will be different from my wooden flutes. But for now … let’s cast a preliminary sound area. The following pictures depicts the final wax tree for my trial.
A Big Mess
Back again … with a big mess. I think I have seriously learnt yesterday, that failure seems to be unavoidable if I want to learn! I poured bronze into the prepared and burned-out mold and everything seemed fine. But when I tried to uncover my hidden bronze beauty 😉 I couldn’t find anything. What an interesting experience … What happened? The mold broke in the bottom and all the bronze simply “fell” through and collected in the bottom of the flask. I didn’t even realize it until I scraped off the investment, never thoroughly checking the bottom. One of the mouth pieces was completely empty, the other a little recognizable but no detailed texture (like ramps, etc.) at all.
What do I think I have learned:
1. I let the investment dry for 2 days before I realized that a little moisture is good when you start the burnout, as I learned online. Next time I will let it site for 2 hours as the manufacturer suggests and burnout right after.
2. Too little investment below (above) the model to be casted. I will put more thought into the sprue size! Shorter channels and further away from the bottom (in casting direction).
3. The details in the one sound area were not detailed at all. I guess that had two reasons: 1. The melting temperature was not high enough and the bronze was not thin enough when poured. Next time I will make sure the bronze is at 1100 degrees celcius (2100 degrees fahrenheit)) 2. The temperature of the flask/investment/mold was too low while pouring and the metal cooled off too quickly when flowing. Next time it should be at around 650 degrees celcius (1200 degree fahrenheit)
4. Because I am currently not casting under vacuum, I need to pay more attention to the metal flow. I had some delicate areas in the mold that need the metal to flow up to fill them. This will be hard and depend on a lot of factors that are hard to control. Next time I will “visualize” the metal flow better and make sure there is only downflow necessary.
A little frustrating … but I will turn this feeling into motivation to get better! So stay tuned.
Sound Area – I will never give up …
Avoiding all of yesterdays mistakes I had another try today and it worked! The sound area ramps and openings are pretty precise and you can even see the small lines on the ramp from printing. I did not even care to file them away in the wax model, which I definitively will when I cast the final flute.
I will skip the filing/sanding/polishing – I am sure it would like beautiful – and will now focus on finalizing the concept / 3D-design for the flute.
Current progress:
1. Mouth piece is understood and ready to be casted in silver. I will solder an internal 8mm tube to the blow hole.
2. The SAC (internal shape) is still not clear to me yet. I will have to find a design in Fusion 360 that makes sense. It should be different than my wooden flutes in order not to waste too much silver. probably something like a tube which narrows down to the blow hole tube on the mouth piece side and guides the air towards the SAC exit on the other. I will see…
3. The sound area is also not 100% set. I will either just cast the top part as I did in the last steps, create an opening in the long tube and solder it OR I will cast part of the 20mm tube below the sound area with it. No idea … I will see …
3. The rest should be easy. A long silver tube with 20mm diameter and finger holes at the right place …
I feel like I am about 30% of the way and Christmas is still some time away 😉
Printing Castable Resin
I am sure my family will soon kick me out because of all my new machines 😉
A small detour! I am am seriously starting to get interested in casting and this could become a base activity for several hobbies. When my flute is done my next projects will be (jewelry-) finger rings. After printing the sound area several times I realized that the resolution of my FDM printer is not good enough. Fine details are missing and the surface is not smooth enough. Isopropanol washing is helping but I lose some texture details to get the model smooth.
I decided to buy a resin printer! These type of printers are basically hardening liquid resin in very thin layers and are being used in various industries incl. jewelry making. The videos I watched were very impressive and it seems like this could ease my projects very much. The better the model the less headache I expect in the subsequent (metal) cleaning and post-processing of the cast.
I chose the ANYCUBIC M5s (link) which seems to be a very nice printer. I have already set it up and getting excited to print my first mouthpiece and sound area test pieces with castable resin.
Wow! This is cool. I used castable resin to print the following models. They have a perfect surface and I am now confident, that this is the right equipment and process to go on.
Big step – New Investment Powder
As stated I have been using X-Vest investment powder for all of my learning journey. I am sure it is great for some people. But not for me anymore. I ordered a new investment to try out and I can’t believe how much time I have wasted!!!!
The problem with X-Vest was that it is super hard after the burnout and does not come off the casted object easily. I had to scrape for 30 minutes using all kinds of tools to get the cast out of the flask and clean the model.
This is now history. The new investment powder is made by: Certus and called “Optima Prestige” (link). It mixes easily, the casting surface is very detailed and smooth and to get the model out of the investment after casting, all you have to do is quench it in water after a cool-off period of 15 minutes. When quenching the investment is magically flying off the model and there is very little and easy cleaning left. What a relief!
