One common type of maker project is custom tools. When we are working on a project, whether a new build or a repair, we use a wide variety of tools. Sometimes the right tool isn’t available, so we improvise: we modify an existing tool or build our own. In the end, we have a new tool that can continue to make our future work easier, better, and more productive. Sometimes a side tool project results in a tool that can continue to be useful for years, even if the utility is somewhat niche. This is one such project.
A few years ago at work, I was trying to apply some hot melt glue to a prototype I was working on. The part that I was trying to glue was recessed in a cabinet with a narrow opening, and the hot glue gun would not reach the point of work. I was trying to hold the gun above the part and use gravity to drip the glue onto the work area. This was a frustrating and messy task, and it didn’t work very well.
Recently I ran into a similar situation with a home project. It was very difficult to get the glue where I needed it. At the time I had to make do with some dripped glue that I spread a bit with a narrow stick, but this was far from ideal.
I realized that since I had a Lulzbot Taz 3D printer, I could do something about this problem. I disassembled a small low-cost glue gun and I saw that the internal components are quite simple. Inside the two-piece injection molded case is a simple heater assembly that has the glue nozzle at the front end and accepts a glue stick at the back end. Behind that is a trigger-operated feed mechanism. The power wires are connected directly from the cord to the heater, so that the heater is on whenever the gun is plugged in to live power.
I quickly made a preliminary plan to design a long-reach gun housing that would re-use all of the internal parts of the existing gun. The heater assembly does not actually touch the case. The rear portion of the heater has a silicone tube over it as the interface between heater body and plastic housing. (The glue stick feeds into this silicone tube.) The front of the heater, right behind the tip, uses a plastic washer that actually contacts the tip area and suspends the front of the heater in the opening of the housing. Because of the greater length of the glue gun, I will use 10-inch (254mm) long glue sticks.
Then I went to work with OpenSCAD, a free 3D CAD program that uses a programming language to define 3D objects. I designed the main body of the gun based on a scaled cylinder, while the handle is based on a cube and uses the minkowski transform for rounded corners. Inside the main body I implemented several features for holding the heating element and the glue feed mechanism, as well as bosses for the screw fasteners, and a strain relief for the power cord.
This is a good time to consider some differences between injection molding, as with the stock glue gun housing, and additive manufacturing using Fused Filament Fabrication 3D printing on the Taz. In the injection molded housing, all of the internal features have similar thicknesses. This includes the mounts for the heating element, the glue feed area, and the bosses for the screw fasteners. They have a few reasons for using consistent thicknesses:
- Using thin cross sections can minimize material usage, which can result in significant savings for high-volume manufacturing.
- Avoiding thick areas results in consistent and short times needed for the liquid plastic to solidify before the part is extracted, which allows for shorter molding cycle times and therefore greater manufacturing throughput.
- Avoiding wide variations in material thickness also reduces internal stresses that would result if some areas cooled and solidified before other areas. These stresses, if excessive, can result in warped parts after they are extracted from the molds.
For parts that are being designed for low-volume 3D printing, we can ignore these constraints. (We can also ignore the draft angles that are necessary to assure that molded parts can be extracted from their molds.) This freed me to use thicker cross sections for some of the internal features. For example, the three small support ribs that hold the back of the silicone tube in the injection molded housings are not desirable features in a part that will be 3D printed. 3D printed parts are relatively weak in the Z-axis direction, so implementing such thin features would result in parts that are prone to breaking. Therefore, free of injection molding constraints, I made the rear mounts one solid piece in my design.
The long reach of this glue gun means that the glue stick needs to be guided through the extra distance internally. I used a similar solid piece here, internally tapered, to guide the glue stick from the feeder to the input of the heater assembly. Other areas that use thicker cross sections are the cord strain relief and the bosses for the screws.
The main body of the gun is 250mm (just under 10 inches) long, a good match for the generous build platform of the Taz. I used Slic3r to create the G-code, which includes support material to hold up the outer shell as it is being printed.
As part of my planning, I measured the temperature of the heating element. This particular glue gun comes in two variations; a standard gun, which is blue, and a low-temperature version, which is orange. Using a thermocouple, I measured 166°C inside the melt chamber of the standard gun, and 156°C for the low-temperature gun. I used the low temperature heater in my initial prototype. These temperatures are above the glass transition temperature of ABS, which is about 105°C. However, I believe ABS is OK for use as a prototype, and for short-duration use if I can avoid excessively heating the housing. This has proven to be true in my testing so far.
ABS is amorphous and therefore has no abrupt melting temperature. It begins to soften at the glass transition temperature and becomes progressively softer as the temperature rises. It flows well enough for 3D printing at about 225°C. I would like to try printing the glue gun with higher temperature plastics, such as polycarbonate, but I don’t yet have a high temperature extruder for my Taz. I would also like to test my prototype glue gun further to find out exactly whether and how it fails at high temperature. I’ll wait until I have another one built, as I don’t want to destructively test my only prototype.
I set up the print files to print the two halves of the housing with the seam against the print bed. This results in smooth mating surfaces that let the two pieces fit snugly together. Slic3r takes 30 minutes to create the G-code for each half of the housing. I believe most of this time is spent generating code for support. Each half took nearly 14 hours to print on the Taz.
After printing, I was faced with the task of removing the support material. It took me a few hours to carefully remove the support. During this time I was wishing I had dual extruders so I could use a soluble support material. That would allow me to soak and rinse the parts, avoiding the tedious process of manually removing support material.
I didn’t bother to carefully clean all of the support from the inside walls. The support is tightly bound, and removal is very difficult. Inside the housing it doesn’t matter, as this remnant of support will not be visible and does not cause any problems.
Next it was time to install the internal components and adjust for final fit. On this prototype I had to do a bit of cutting with Exacto knives and some coarse grinding with a rotary tool to make everything fit well. I also used drill bits to clean the holes for the mounting screws, the trigger pivot, and the trigger return spring anchor.
The trigger return spring anchor points out another difference between injection molding and 3D printing. The spring has a 5mm loop at each end, and the molded case has a small boss, 4mm in diameter, to anchor the loop at one end of the spring. 4mm is too small to make a reliable feature with 3D printing, because it would extend in the Z direction, which has limited tensile strength because of low layer adhesion. So instead I designed a pair of larger opposing bosses, both set in from the parting line, and with 4mm holes. I have some wooden kebab skewers that are 4mm in diameter, so I cut a short dowel from one of these and inserted it as the spring anchor. I installed the rest of the internal components and used a bit of hot glue to hold the power cord in place internally. Then I was ready to fasten the two halves together.
Next I carefully installed the other half of the housing, making sure all of the components lined up. I screwed the housings together, and it was ready to test.
The long-reach glue gun works well, but as mentioned earlier, because of the ABS material, I keep it powered on for only a few minutes to perform a specific job. I look forward to trying some new materials and also maybe some minor modifications to the design to fine-tune the internal fit, to improve strength in some areas, and to make the body and tip thinner. I don’t think it will win any industrial design awards, but it is a functional tool for some specific situations, meeting my original goal.