Build Blitz Weekend 1

Posted by thingsimdoing

With the deadline for complete rapidly approaching, progress has stepped up. I spent most of my Saturday machining, welding, and cutting.

The bottom of the main valve on top of the injection cylinder has pipe threads to accept a compression fitting. One note about pipe tap, you need to be really careful about your tap depth. If you tap too deep with a tapered pipe tap you can end up in a situation where your fitting bottoms out before the threads tighten up, leaving you with a leaky pipe connection. So pay attention to that thread call out and test with a fitting if you are unsure.

Here are all the parts for the injector valve assembly. This regulates the flow of plastic and air into the mold cavity via the shuttle valve sliding back and forth.

Starting hole layout for the melt tank. I kind of wish I had used steel for the tank instead of stainless steel. Stainless is a real pain to work with. It’s hard to cut, hard to drill, hard to bend, hard to weld. Just really unfriendly in general. The original mold-a-ramas used an aluminum tank, but I got a really good deal on the stainless tank I started with to make this part.

Another area where I’m differing from the original mold-a-ramas is the piston design. The original injection cylinder sealed agains the rod of the hydraulic cylinder used to inject plastic for the melt tank to the mold. In my design I have the seal on the piston. I’m very ~~nervous~~ excited to see how well this works. There is another disk, not shown, that will retain the seal against this piston.

A few more from this weekend:

I really need to hire a hand model.

There are a few more bits and pieces that need to be fabricated (frame for cover, water tank, brackets for a few items), but the bulk of it is done. Next up is programming, wiring, and test/debug.

Mold Machine, Meet Welder

Posted by thingsimdoing

Metal met welder this week in the shop. I finished machining a few brackets and started tack welding some assemblies together.

Welding aluminum is tricky, the coefficient of thermal expansion is almost twice that of steel, which means it moves from the heat of the weld, much more so than steel. Here’s some tips when welding together plates of aluminum:

1. Aluminum moves during welding. A lot. Start slowly and clamp together anything that may warp or shift during welding.

2. Check for square and check your dimensions, early and often. Since the material moves so much during welding, what started out nice and square will end up looking like a pretzel. So weld slowly and move around, alternately tacking on opposites sides (the heat from the weld has a tendency to pull the plates together on the welded side and apart on the opposite side. Recheck for square/level after very few tacks.

3. I always find faces in the stuff I work on:

I was able to tack together the complete injection cylinder, minus a small plate on top which I later added to give me a spot to bolt the valve block to.

I also tacked together the swinging mold cylinder brackets:

Having the mold cylinders swing up allows me to service the parts inside the melt tank without disturbing the alignment of the molds.

And the sub-frame that a majority of the machine gets attached to.

That’s it for now. Next up is cutting the 1×1 angle for the top cover and finishing some parts on the mill and lathe.

Mold-A-Rama Build Begins

Posted by thingsimdoing

I have spent the last 5 months designing, buying parts, redesigning, and buying more parts. I am finally at the point of cutting metal and bolting stuff together (the fun part). I have to say, it feels pretty good. Below you can see the 80/20 frame (mostly 1010 series with a lower frame made from 1020 series extrusion) , it is still missing a few crossbars near the top, as I hadn’t yet tapped the ends of the extrusion to accept a corner fitting.

I was initially concerned about how rigid the 1010 uprights would be given that the they are only attached to the bottom 1020 frame with two plates each. I found that it holds up surprisingly well to small amounts of force (hand applied), and this is even without the aluminum weldment that fills the space between the two angled bars. You can also just barely make out the waterjet cut 120 degree brackets on the front angle.

No longer just shapes on a computer screen.

I’m also very close to some initial tests with the pneumatics. I have new sensors for air cylinders that interface better with a micro controller, and have started breadboarding the transistors to drive the solenoids valves.

DIY Mold-A-Rama Update 3

Posted by thingsimdoing

A lot of progress has been made in the last few months on my version of the 1960’s classic Mold-a-Rama machine.

The design has been further refined

The frame where the mold cylinders attach has been changed from 80/20 to 1″ square aluminum tubing.
- This change was made because I was having a hard time mounting and aligning the various components to the slots in the aluminum extrusion.
- The mold halves will press against each other with several hundred pounds of force; the friction fit nature of t-slot construction would likely have failed under this load. The new frame is a one piece welded structure.
- The new frame, while not re-configurable, will be much stronger and allow for easier alignment of the plastic tank and mold halves.

The plastic pump has been completely redesigned
- I’ve switched from a right angle gear motor to an air cylinder driven piston pump (think giant aluminum syringe)
- This change eliminated a complex machined part ($$$) and replaced it with a much simpler welded tube and plate design ($)

The actual mold cylinders have been selected
- So much of the design decisions and components selections have been driven by what I can get on the surplus market Case in point: the mold cylinders. I’ve seen that the actual mold-a-ramas have an enormous amount of play in their mold cylinders and mounts. I thought using a twin piston cylinder would help with the side loading on the pistons (due to the weight of the molds want to slide down on the angled frame. I found a good price for two twin piston SMC cylinder on eBay (with sensors and flow control fittings!)
- I designed a robust mounting system using aluminum tube, bronze bearings, and 5/8″ steel shafting.

Parts have been bought

I’ve also been spending quality time on eBay, at surplus stores, and throwing money at various other online retailers. Here’s where some of the money has gone:

Left to right: Air compressor, air tank, compressor switch

The air compressor is a Thomas & betts unit I got at C&H surplus (super cool store, check it out if you’re in SoCal). It has twin cylinders and puts out more CFM at a lower decibel than most compact air compressors.

Left to right: water pump, auger for plastic hopper, cartridge heaters

The top two cylinders move the mold halves together, the bottom one moves the piston in the plastic injector, and the solenoid manifold on the right controls it all.

50lbs of hard to find plastic

The ability to buy the plastic pellets was a make or break moment for this project. There were not a whole lot of suitable replacements for this particular plastic (more specifically, polyethylene wax). Fortunately the west coast distributor for this happened to be close and had several bags it was willing to sell to me (normally this product only sold in 1000kg pallets, which is about 950kg more than I need). Getting a hold of this was a major load off my mind.

It’s like the warehouse in Indiana Jones, but instead of ancient relics there’s plastic resin.

I’ve begun cutting metal for the frame and plastic melt pot, hopefully welding will start this week!

Cardboard Mockup: The Original Rapid Prototyping

Posted by thingsimdoing

I often find it hard to get a sense of scale when designing things using CAD software. Staring at a model on a 24″ screen can sometimes make small objects look massive, and large objects tiny. Placing references (such as a person) next to your model helps some, and there are many free human models for pretty much all CAD platforms. However nothing beats a physical prototype. If your part is small enough 3D printers are perfect. For larger parts you have to get creative. The final frame will be made of 80/20 aluminum framing, but I did not want to commit to cutting it up just yet, as I was still playing with the frame dimensions. I needed a cheap material that was easily formable, yet sturdy enough to hold its own weight, next to my door was the answer: double wall cardboard.

Just as architects use cardboard to construct scaled down buildings, I built a 1:1 model of my mold-a-rama replica. Using only packing tape and cardboard, I was able to quickly (and cheaply) build a model that’s accurate to about 0.25″.

I started by measuring the outside dimensions of my model in solidworks, and transferring those to cardboard. Here’s some tips I learned from doing it:

The boxes I had on hand were medium to small-sized, with fold lines all over. If you can’t cut around the lines, take another piece of cardboard and tape it over the fold line to reinforce it.
Reinforce the corners by making a long L-bracket and tape it to the inside.
Cardboard tabs can be used to prop up unsupported spans of cardboard.
Get some good blades for your utility knife, I like Irwin bimetal blades. They make cutting through thick double wall cardboard a breeze.

And here’s the end result:

Those 3rd grade arts and craft skills are finally paying off.

As I suspected it was bigger in full-scale than what I thought it would have been.

Another benefit is that the interior volume is very close to the usable space inside the actual machine. I was able to place most of the bulkier components inside, allowing me to play with layout:

From left to right: air compressor, injection cylinder, plastic melt tank, water pump (blue thing peeking out), water chiller.

Moving components around inside the cardboard model was so much faster than doing it in CAD. It also gave me a better idea of how much space I need between components.

The layout above is mostly complete. I have since gotten an air tank that sits in the back left corner. The stainless box you see is the starting point for the plastic melt tank (I’ll go into more detail in the next post where I’ll show off some of the components that have already been bought)

This method really only works if what you are building is mostly flat panels that meet at right angles. If you’ve got a contoured model and you are set on using cardboard, make an STL file of your part with layers that match your cardboard thickness and print out patterns. This can be done using the free software AutoCad 123D (http://www.123dapp.com/make)

DIY Mold-a-Rama Update 2

Posted by thingsimdoing

I started back up on the design of the mold machine this week. A few major changes have been made:

Instead of using a piston pump to move the molten plastic from the melt tank to the mold I am using an external gear oil pump. The plastic I’m using has a melt viscosity similar to that of cold oil, so the pump should be capable of transferring the plastic.
I have also fleshed out the melt pot mounting details and devised a new valve body that incorporates the oil pump mount. The pump is driven by a right angle gear motor. This will make the shot size easily adjustable.
A refill hopper, auger, and pot stirrer have been added to the rear of the machine.
Some of the larger subassemblies (water chiller, air tank, compressor, power supply) have been added to gauge enclosure size. I’m still shooting for coffee table sized, but finding a refrigerant based chiller the size of a shoe box is difficult. A much smaller option would be a thermoelectric (peltier) based chiller, but the heat input into the system from cooling the plastic (about 250W) and the ~2 gal water capacity is a little out of the range of most thermoelectric chillers. Not to mention they are very inefficient.

I’m going to continue to refine mounting details for a few weeks or so. In the mean time I am tracking down parts: 100rpm right angle gear motor, <1/4hp aquarium chiller, air solenoid manifold, oilless compressor, and air cylinders. Some of the parts above may change depending on what I can find at surplus shops/eBay.

DIY Mold-A-Rama Update

Posted by thingsimdoing

Here’s where I’m at:

Still many missing components and systems, most notably: mold cooling system, pneumatics controls, part ejection cylinder, microcontroller, sensors, power supply, mounting locations for most components, and many many more.

The pump that transfer the molten plastic is a simple piston pump using two check valves:

http://www.egr.msu.edu/~lira/supp/steam

In my first design I created my own check valves by integrating a ball bearing and spring into the actual pump body. While this made the pump more compact, it also added several machining steps which would have increase the price. So instead I’m using off the shelf check valves that screw into the pump body. Much cheaper, much simpler, and likely to work better.

In another bit of “don’t-redesign-the-wheel” thinking I’m using an electric deep fryer as the melting pot. Candles makers already are using this as a melting vessel. This already has a heating element and thermostat built-in, and it only costs ~$30. Compare that to several hundred dollars had I made my own vessel from stainless, used silicone band heaters, and made my own temperature regulation. The latter would have definitely been cooler looking, but for this project I’m trying to keep costs low.

Not as cool as stainless steel and band heaters.

I am at a bit of a crossroads design wise. I’m deciding between using pneumatics or electric actuators for motion control. Pneumatics are more impressive visually and give me a chance to learn about pneumatic systems (plus they make a neat pssst pssst sound), however they require an air compressor (loud) and are more expensive. Using electric motors would make the system smaller, as I don’t need to house a compressor and air tank in the enclosure.

I’m leaning towards pneumatics provided I can find a quiet air compressor. I’ve read of people re-purposing a refrigerator compressor, but I’m concerned about their CFM rating and live span.

I’m also looking very seriously at buying a small mini mill. I should recover the cost of the machine and then some verses paying to have it done. Ideally I could find a small Chinese mini-mill on Craigslist, but I haven’t seen much out there. Maybe I’ll just break down and get this guy:

http://littlemachineshop.com/3900

DIY Mold-A-Rama: How it Works

Posted by thingsimdoing

One of the first steps in reverse engineering a product is understanding how it works. Ideally I would have access to the actual mold-a-rama (MAR), however they are rare and expensive. I could find an operating installation, but the closest operating MAR is at a zoo several hours away. I also have been unable to find pictures or video of the mechanics below the cabinet. The only parts visible from the images I’ve seen are the molds, mold rams, and ejector. All the complicated bits are hidden from view in the cabinet. The last avenue is products documentation in the form of manuals and patents. Fortunately I am an awesome google-er and have found both. The operation and repair manual can be found here:

http://www.scribd.com/doc/37301298?secret_password=1ehb8rphnat7ykzt6zu3

View this document on Scribd

Continue reading →