Hey all.
Since I am about to finish my 3DP machine that I named “Empiric 3000”, I thought I will start sharing my progress with the printer. I did not share process of the making as it was quite long and had significant pauses due to Kungflu virus. Printer itself was finished before Kungflu virus, but due to the lack of time in Makerspace (and lockdown), I was not really able to continue to work with it. But now, slowly, I start doing something.
Firstly, my main idea and goal is to print metals. Although I am interested in metals, polymer 3DP printing was my second goal. So all my work is now dedicated to reaching knowledge in these techniques. However, I will start with regular plaster/gypsum materials, because there is more info about it and it will help me to learn more about the process. So let’s start with the machine.
Empiric 3000 was inspired by Oasis and its design is great, but I immediately had my own improvements based on my experience building 3D printers and other things. Also I wanted to add as much flexibility to the machine as possible. So eventually everything ended up in a complete redesign. SolidWorks initial design took around 3 months to make.
Oasis used aluminum profiles, I decided not to use as budget with profiles increased so significantly, that it made no sense for me. I used my own developed platform of 40x25x2 aluminum profiles and FDM printed corner-joints.
Of course it might lack the overall rigidity compared to engineering profiles, but the cost of the whole setup for me was around 95€, whereas extrusions would have ended up with more than 300€ incl. all nuts etc. Also, I thought of 3D printing as much parts as possible. I know, 3D printer for metals printed out of plastic seems doomed, but I am expecting at least something…
Empiric 3000 3DP printer
Re: Empiric 3000 3DP printer
Gantry and especially carriage part was quite similar to Oasis as there are so many parts for HP45 that I did not want to redesign. I only made some minor improvements to coater mechanism to add more flexibility and more adjusting possibilities.
Actually the whole build started from gantry part and after it was made and spreader was working (mechanically), I moved to the whole frame. Spreader roller was also quite pricey, so I machined that myself from calibrated steel.
Actually the whole build started from gantry part and after it was made and spreader was working (mechanically), I moved to the whole frame. Spreader roller was also quite pricey, so I machined that myself from calibrated steel.
Re: Empiric 3000 3DP printer
After all of this, the most difficult part of the project started: feeder and build pistons. As I intended to print metals, I had to come up with a completely new design. The main requirement was, perhaps, that it should have been made out of metal components only or some components out of PTFE or graphite at worst. The reason for this is that I wanted to have an ability to run initial thermal treatment of printed parts while they are still in the build piston. I was not sure about all of the details, but having an ability to heat it up to at least 200-230C would be good. So all in all I needed a metal a cylinder and a piston. Who has such parts? I thought about a car, yes, internal combustion engine! I started to look at this and found out that there are such things like cylinder liners that are used when motor and cylinders are refurbished. However, it was difficult to find quite large-volume (and diameter) cylinders as nowadays usually there are more cylinders of lower volume than fewer with a large volume. So I started to look at older cars or SUVs.
I managed to get some very old cylinder liners as well as pistons. Do not ask me why there is a dick on the top surface of this piston. I ran some initial tests with this setup (with rings on the piston installed). I tried to move the piston by hand and it seemed that there might be some potential in this solution. But at this point I made a conclusion that I will leave this option as my plan B. There are a lot of complex reasons for this decision and I think I will not go into many details yet unless someone is interested in this.
I ended up buying polished AISI 304 tube with D=90mm. Walls of the tube were quite thin to reduce mass.
Thin walls also resulted in the fact that such tubes, especially when cut, are far from perfectly round. They are more of elliptical shape rather than circular. So I had to find a solution to make it round.
I managed to get some very old cylinder liners as well as pistons. Do not ask me why there is a dick on the top surface of this piston. I ran some initial tests with this setup (with rings on the piston installed). I tried to move the piston by hand and it seemed that there might be some potential in this solution. But at this point I made a conclusion that I will leave this option as my plan B. There are a lot of complex reasons for this decision and I think I will not go into many details yet unless someone is interested in this.
I ended up buying polished AISI 304 tube with D=90mm. Walls of the tube were quite thin to reduce mass.
Thin walls also resulted in the fact that such tubes, especially when cut, are far from perfectly round. They are more of elliptical shape rather than circular. So I had to find a solution to make it round.
Re: Empiric 3000 3DP printer
So I machined these ring-shaped parts with lathe and pressed them tightly on the tube. Surprisingly, this solution worked really well. At this point I had perfectly circular tube.
The piston itself was a huge headache, but I ended up making a sandwich of PTFE/Graphite packing.
I know, there could have been shorter ways of doing this, but I tried a lot of solutions in the process and most of them I did not like. So I just ended up with all the legacy of previous versions. If I had to do this from scratch again, I would do it a bit differently.
Now comes the another tricky part: motion system of the piston. Since I wanted all-metal thing, I decided to do everything from aluminum. CNC was the main tool here.
The piston itself was a huge headache, but I ended up making a sandwich of PTFE/Graphite packing.
I know, there could have been shorter ways of doing this, but I tried a lot of solutions in the process and most of them I did not like. So I just ended up with all the legacy of previous versions. If I had to do this from scratch again, I would do it a bit differently.
Now comes the another tricky part: motion system of the piston. Since I wanted all-metal thing, I decided to do everything from aluminum. CNC was the main tool here.
Re: Empiric 3000 3DP printer
Next big part was electronics. I am not electronics engineer, so a lot of things were challenge to me. I decided to put everything into sealed plywood housing to avoid contamination (I did not wish metal powder to get inside electronics…).
Re: Empiric 3000 3DP printer
Also this control box has a UI dashboard to see and monitor things.
All electronics is basically ATX power supply with RPI3 as host computer. There are other components that are required for Oasis. Additionally, there is a PCB to handle ATX power supply shut-down via button (as with PCs, i.e. when shutting down RPI, it also switches off power supply). RPI holds control software and I connect to it via remote desktop (VNC). There is a ton of voltage and current meters everywhere.
Build surface area has some flexibility.
That red thing is 6mm steel plate. It can be taken off together with build and feed cylinders. This gives an ability to put larger (or perhaps even smaller) cylinders if necessary with a single move. This was the main idea if I needed smaller cylinders for material testing, although current ones are pretty small already. Since metal printing will cause severe increase in weight in this area, the whole top surface is supported by 40x25 aluminum profiles across the entire area.
All electronics is basically ATX power supply with RPI3 as host computer. There are other components that are required for Oasis. Additionally, there is a PCB to handle ATX power supply shut-down via button (as with PCs, i.e. when shutting down RPI, it also switches off power supply). RPI holds control software and I connect to it via remote desktop (VNC). There is a ton of voltage and current meters everywhere.
Build surface area has some flexibility.
That red thing is 6mm steel plate. It can be taken off together with build and feed cylinders. This gives an ability to put larger (or perhaps even smaller) cylinders if necessary with a single move. This was the main idea if I needed smaller cylinders for material testing, although current ones are pretty small already. Since metal printing will cause severe increase in weight in this area, the whole top surface is supported by 40x25 aluminum profiles across the entire area.
Re: Empiric 3000 3DP printer
So after all of this, machine looks like this.
Next step is to work with materials, which is the most interesting part, probably, to me. I will post updates as there are some already. I have not printed with the printer yet, so I am simply hoping for the best... Now I prepare material prototypes and testing before pouring to printer... By the way, that box on the printer surface is powder box that will hang below the main build surface. Spreader will dump powder into that bin.
Next step is to work with materials, which is the most interesting part, probably, to me. I will post updates as there are some already. I have not printed with the printer yet, so I am simply hoping for the best... Now I prepare material prototypes and testing before pouring to printer... By the way, that box on the printer surface is powder box that will hang below the main build surface. Spreader will dump powder into that bin.
- dragonator
- Site Admin
- Posts: 597
- Joined: Fri Aug 14, 2015 4:48 pm
- Location: The Nethelands
- Contact:
Re: Empiric 3000 3DP printer
That was an amazing write-up. You can certainly see the Oasis legacy with a lot of improvements. You can see that you put a lot of work in designing and building this. I like how robust the pistons look. This is a weak spot on my current Oasis.
You have a resin printer? I see FDM parts, but the printhead carrier looks way too clean for FDM. I thought it was metal for a while.
I would love to know how you got the Oasis controller software to run well on the Pi (what Pi is it?). My attempts on a Pi 3 were not that great, and I would love to include instructions on how to do so.
I am looking forward to seeing this machine print
You have a resin printer? I see FDM parts, but the printhead carrier looks way too clean for FDM. I thought it was metal for a while.
I would love to know how you got the Oasis controller software to run well on the Pi (what Pi is it?). My attempts on a Pi 3 were not that great, and I would love to include instructions on how to do so.
I am looking forward to seeing this machine print
Re: Empiric 3000 3DP printer
Yes, they are SLA printed. I thought I will give SLA a try with this one. Some parts were good, some were still made with FDM due to toughness reasons. I am now planning to finalize motor connections to build pistons with GX12 connectors and that would be it. Also I bought quite large plastic bin that would work as de-powdering chamber, so I will have to make this before printing as well.
I am also testing mixers for powder (for small scale batches). I will get back, when they start working reasonably. Here is the snippet.
P.S. Drone motors are usually the only things I have not used by their intended purpose ever...
I am also testing mixers for powder (for small scale batches). I will get back, when they start working reasonably. Here is the snippet.
P.S. Drone motors are usually the only things I have not used by their intended purpose ever...
- dragonator
- Site Admin
- Posts: 597
- Joined: Fri Aug 14, 2015 4:48 pm
- Location: The Nethelands
- Contact:
Re: Empiric 3000 3DP printer
I first thought it was a ballmill. It looks like a great mixer. I like the usage of random parts and controllers to make it.