1 - that is the plan. Everyone interested to pipe up on a desired shot size. Frogs were brought up as something to use when we are cleaning out the unit just before purging!
2 - everything is on the table now. But we’d really prefer NOT to burn the MF down!
As long as the heater wire is affixed to aluminum with high temp RTV and has proper PWM control, I think burning stuff down is far less of a risk than the aforementioned molten plastic vs pressure vessel 
Oh also - have you considered flex sensors? Torque wrenches drift out of calibration and are expensive, in my experience. I also assume you’ll want to press at a constant (known) pressure for a certain length of time, instead of a steadily increasing pressure until the torque wrench pops
All things are on the table. So far I think we are modelling after the couple of examples that we’ve seen online.
We are also considering the space issue that we have at VHS so something modular that is easy to break down and store is also being considered in the design. So with using an arbour press, it wouldn’t be tied permanently to this device.
Hi guys - I’ve been away on vacation so my apologies for disappearing for a while. Excited to see this thing moving along. I apologize if the points below have already been discussed and considered, but here are my thoughts:
*re: the PDF design to use a threaded conrod to adjust the shot size: I’m not sure if it’s necessary, as from what I can tell the amount of plastic that gets injected is simply self limiting as the mold is filled to capacity. The same effect can be achieved simply by filling fewer pellets into the chamber.
(2) Heater cartridges are super cheap on eBay - if I remember correctly there are 300 watt, 110V ones for like $3 shipped. The ones made for 3D printers run at much lower wattages and would be underpowered for this application, and require an additional power supply to convert from line voltage to 12V.
(5) The feeder system in the designs I’ve seen have all simply been gravity fed. The most elegant one I’ve seen is simply a conical hopper on top of the chamber, and the pellets fall into the chamber when the piston is raised up.
(6) Using a K type thermocouple would be more accurate and appropriate for this application (and more readily paired to off the shelf PID controllers), and I have some on hand. They’re commonly used for DIY sous vide setups and cost less than $30 shipped for the controller, thermocouple and an SSR for switching the heating element. For safety, I think a thermal fuse attached to the heater block that physically cuts power to heater would offer better redundancy against thermal runaway caused by sensor and electronic failures (and is a proven, common design in espresso machine boilers). I’m not sure if 4 sensors are necessary - the biggest temperature differential will not be different spots in the metal of the heat chamber itself, but between the heaters and the molten plastic centre of the chamber, and it’s difficult or impractical to position a sensor there. I think if we position the thermocouple near/at the nozzle, that will give us a good idea of whether the plastic is ready to inject or not, but more sensors may help in the design phase to help optimize heater placement. I also have a thermal cam we can use to get an idea of the heat distribution.
I hope I haven’t stepped on anyone’s toes with regards to design ideas etc, as I wasn’t able to make it to the meeting where these discussions took place. I feel like we should err on the side of simpler, proven designs as a starting point as it would minimize the amount of R&D we’d have to do on our own. With some more experience in hand, we can then better identify the shortcomings to improve on the designs.
Thanks Ryan,
Some of our thoughts were that if we want to add details to the hardware now, it’s cheaper and easier to put them in, and not use them, then to add them later. It doesn’t cost much to add a hole when it’s in the early stages as opposed to tearing everything apart later and have to set it up. By Cost I’m meaning time.
For the thermocouple/Thermistor positioning, I think a lot will depend on the shape of the block housing the chamber.
A lot of our discussion was on the shape of the chamber specifically at the bottom that leads to the nozzle.
Some other stuff that we discussed was feeder. Some of our concern was around gravity systems getting too hot and the pellets getting hot and starting to gum up the entrance to the chamber. All things to consider, and we all left there planning on meeting again before me start putting things together.
Laser printer fusers have a high wattage heater with integrated over-heat cut-off and thermal probes. It’s also wrapped in a nice heat resistant cylinder with a non-stick coating. We can probably get some cheap from discarded laser printers.
I know where to get some printers!
But if the rollers are heated and not rolling, they will melt to surfaces
they are touching. Ask me how I know.
I personally feel that since I don’t know what I don’t know, it’s not obvious to me whether smaller details like optimizing heater placement or chamber geometry matters in practical terms (i.e. ease of use or part quality), or we are just complicating the design and making more work for ourselves. After seeing the diversity of existing, proven to be working designs (some of which are obviously sub optimal), it seems like there’s a good amount of fudge factor in designing a working machine - the Gingery design doesn’t even have accurate temperature control, relying on an electric bang bang controller and literally an oven thermometer.
For me, it seems like the challenge to making this thing exist is in more basic, fundamental design decisions like how specifically the piston and heater chamber attaches to the arbour press (assuming we’re using one), and whether the heater chamber even needs to be a machined block at all, when a steel pipe with heaters running the entire length could serve the same purpose.
I do agree that chamber geometry or hopper design is worth considering (for example, E3D claims that their V6 3d printer nozzle’s interrnal conical taper reduces pressure needed for extrusion), but I feel that we may not be getting the most bang for our buck spending too much time thinking about / prioritizing them, when Gingery’s simple flat bottomed heat chamber with a brazed in nozzle and the EPM Model 150 with the simple conical hopper are both designs that are proven to work.
With a PID controller paired with a smartly positioned sensor (i.e. if overheating the plastic is a concern, then we’d position the sensor closer to the heater and simply wait longer for everything to reach equilibrium temperature), we can minimize scenarios where the system gets too hot, as temperature overshoots can be minimized by tuning the PID constants.
That being said, I recognize that it’s simply a difference in approach and priorities so I don’t want to impose; I do get that the earlier in the design process mistakes/shortcomings are identified the cheaper they are to fix; the tl;dr version of what I’m trying to say is that because none of us have experience using such a machine, we might be better equipped to identify design deficiencies after we have a working one to gain some experience with.
the short of why we are leaning to a conical or bullet shaped end to the plunger and chamber is the proven design of an injection molding machines screw. http://www.engineeringplastics.in/images/c3.jpg since they pretty much ALL go through the point of this detail, it is a geometry worth considering early on.
The designs we are looking at, are fine. But there is nothing to say that we can’t make it better with the collective skills base and resources that we have on hand. Especially when in the meeting we had 2 machinists. One with Milling expertise and another whose primary area of focus is Lathe.
I don’t see the metal work as being an issue with what ever design we decide on.
As I think I stated earlier, it’s best not to get too worked up over details now. We’ll have to get together with a few more sessions before we start putting concepts into CAD.
Speaking of which, band heaters of various diameters are also readily available for cheap on eBay (<$10), which is another reason a pipe rather than machined block w/ cartridge heaters could be worth considering.
I think the conical bottom to the heat chamber is a sensible idea, which reminds me that a protruding conical nozzle (i.e. not flat like filabot’s) may better centre the mold and nozzle openings.
Anyway, it’s clear to me now that there has been way more thought put into this than what I was able to glean off the meeting summary post (my fault for not being there), and I respect that we’re looking to build something that’s perhaps more ambitious than what I originally had in mind.
For the heating, I’m currently building a 1000W induction heater with modular elements that could be used. I intend to use it for hardening toolbits and other small metal parts, but making it multi-purpose would be awesome.
Doesn’t induction heating require that the object/material be electrically conductive? I guess you could use inductive heating to heat up a metallic core in the center of the chamber.
P.S. you can buy inductive counter-top stoves for about $60-70, all you would need is to replace the coil with one of the right shape for your purposes.
Induction can heat the melt chamber and the material through that. Also
came up.
You guys really missed out! Ha ha!
So should we set a date for a second brainstorm session at VHS?
Would be nice to get everyone together, I think at the last meeting we were missing some key members who were very interested in this project.
Additionally we can settle on the shot size which is the critical parameter for figuring out the rest of the dimensions of this project.
A Friday night or weekend would be my preference.
A Friday would have to be after 8. But it’s prime time for some beer
fuelled brainstorming.
So that’s what happened to the son of John Rhys-Davies
This machine is the best I’ve seen for desktop injection. The video describes some of the design features. http://www.mediummachinery.com/medium_machinery_llc.html.
One thing I don’t see discussed in this thread (although maybe it was at the meetup) is the clamp. Most of the desktop injectors on the market have, in my opinion, very insufficient clamps. If it can’t resist the injection pressure, the mould will be forced open and you will get more flash and less crisp detail. Even the 12 ton clamp on the Medium Machinery injector is a little on the light side, giving you only 1.7" surface area in the clamping plane before the clamp starts to fail at 13,500 PSI.
I don’t think we talked about clamps for the molds in too much detail. We were focusing on the injection side of it.
The depending on how intricate we make the machine the clamping would have to be integrated into the mold design, or we could come up with a mold carrier that would have the clamp design integrated, and then we would all have to stick to standard shaped mold blanks.
SCORE!!!
https://preciousplastic.com/en/
I think they have everything we need to know including plans and some great videos. I’m watching them now
