All Posts By

BHudson

3D Printed Springs

By | Robox User Blog | 3 Comments

3D printing has become a critical tool in product development and the increased availability of high-strength, dependable materials has enabled the use of 3D printing in areas that have not previously been possible. The open-minded thinking of the maker community has helped drive this advancement. Companies like Taulman and ColorFabb have made materials readily available for any fused filament fabrication (FFF) 3D printer that have strength, durability, and stiffness far above the basic PLA and ABS materials commonly used. One of these new materials is PET, a polyester based thermoplastic commonly used in water bottles for its chemical stability, strength, and durability. Taulman released T-Glase, a PETT variation of the basic material, and ColorFabb followed with XT, using a proprietary Eastman-Kodak formulation called Amphora. Both materials offer better strength and rigidity than PLA, hydrophobic properties, and ease of printing. I started using T-Glase to get away from ABS due to the fumes and the curling and warping that occurred in many of my prints, rendering some of them completely unusable. I added XT because of the opaque colors and found that it is the easiest printing material in my inventory.
Several months ago I was contracted to help a local company solve a problem with one of their products. They needed to release a latch to open a spring-loaded drawer. It seems like a simple problem on the surface, but they also needed the system to run on a 3VDC battery with very low current draw. The solution they were using was a solenoid and due to the low voltage and current draw the solenoid did not generate enough mechanical force to overcome the friction between the latch and its mating catch. Compounding the problem was the available volume; the entire device is about the size of a smartphone and is abou5/8 of an inch thick. The area reserved for the mechanism, battery, and controller board is about 2.5 inches long and an inch wide. The battery and controller took up close to 2/3rds of that space, leaving only about an inch by inch by half-inch volume to work with. My initial solution was to use a miniature gear motor purchased from Pololu for a robotics project. This was coupled with a steel leaf spring with a formed latch feature. A cam on the motor pushed on a brass pin, in turn pushing the spring to release the catch. For a single prototype this worked very well, but when more than one was needed several problems became apparent; the most challenging being the difficulty of forming the integrated spring and latch. The hardened and tempered spring steel worked great for one unit but the cost and complexity of developing more than one was considerable.

3D Printed Springs

On a whim I started looking at using T-Glase as a spring. I had noticed that when printed in thin, solid sections the material was flexible without breaking. A few trials later I had what would become the final prototype design. The latch, spring, and cam follower were integrated into a single 3D printable part.

3D Printed Springs2

The cam follower pin became a pivot pin and the Pololu motor was replaced with a smaller, cheaper planetary gearhead unit with lower current draw. The motor came with a 6-point spline interface, so a cam head was designed and printed to slip over the existing interface unit. The cam head provided another challenge; the first set of prints were made with PrintedSolid PET+ and have not been repeatable. The PET+ material provides many of the properties of ColorFabb XT but with a more rubbery, flexible characteristic. It was chosen only because it was what was in the printer at the time. Subsequent attempts with T-Glase and XT have failed to print without voids due to the thin areas on the part, or have not been able to hold the tolerances required to mate with the motor interface, requiring extensive hand finishing. In production, this will be an injection molded part; for now it isn’t a problem as I printed more than enough to cover the needs of the project.

3D Printed Springs3

Extensive testing with this design showed the longevity of the spring – over 20,000 latching cycles were recorded without a failure of the spring and with only a slight loss in stiffness. The first spring was printed with clear T-Glase and after about 12,000 cycles fine cracks were visible at the root of the leaf spring arm. Subsequent parts were made from black XT and show no such cracking. This may also be due to a slight lengthening of the leaf spring arm as well. The gray PET+ did not have enough stiffness to properly latch.

The success of the design encouraged me to look further into replacing torsion and compression springs with 3D printed leaf-style springs. The same company came to me with another project; they wanted to use the same technology to make a firearm trigger lock. The idea was to allow the user to bypass the key lock with a control system using Bluetooth Low Energy (BLE), enabling the lock to be triggered between the locked and unlocked states based on proximity or a command signal from the user’s smartphone. As with the previous product, only a small volume was available. My customer wanted to modify an off-the-shelf lock and just integrate the BLE system while keeping the keyed operation so the user could choose between unlocking methods. The trigger lock presented a larger challenge as the force required to keep the lock dependably closed is relatively large – on the order of 4-5 lbf. In the key-only design this was accomplished with a compression spring. Due to the force limitation of the motor, a smaller force had to be used for prototyping with the assumption that the design would be changed to ensure proper retention of the lock. The prototype for this product involved replacing much of the volume occupied by the compression spring with the controller board and adding additional volume by extending the height of the unit. To fit the space and to provide optimal use of the available motor force, the linear spring and latch was replaced with a pivoting latch and an inward-curved leaf spring, as opposed to the outward curve used on the previous design. The inward curve provided a large increase in force, matching or exceeding that of the compression spring, in about the same space as the compression spring. This was too much for the motor to handle, so the force was reduced by changing the radius of the spring curve
and the amount of deflection. Several trials were made until a balance between the available force from the motor equaled the force required to keep the lock latched during product development and initial user testing.

3D Printed Springs5

A similar curved spring was used on a subsequent project. This third project was to develop a cabinet lock. The lock mechanism required the use of a serrated rod mounted to the cabinet door that interfaced with a spring-loaded locking unit mounted on the cabinet frame. When the door was closed, the rod slid into the locking unit and an internal latch caught one of the serrations on the probe. The desired size of the latching unit was cuboid and 1.25 inches on a side. As with the previous two projects, the restricted volume prompted the use of a small stroke spring. By integrating the spring into the latch slider, the part count was reduced and the assembly of this complex mechanism was simplified.

3D Printed Springs6

The key to utilizing 3D printed springs in production prototypes is to maintain the design to allow the printed spring to be replaced with a metallic spring with a minimum of redesign once the unit goes to production. Some products may be able to maintain the use of a plastic spring when the parts are being injection molded, but it will depend on the material used and the expected life of the part. For prototyping, 3D printed springs can enable rapid manufacturing of customized spring units and can integrate what may have been several separate and sometimes tiny parts into a single larger part. The technique can enable rapid testing and trials to enable the proper operational force to be determined; data that can be used later in the development cycle to properly size replacement metallic springs.
For more information on the products described here, please visit www.smartarmortech.com

CAD Model of the Alarm Clock Design

Prototyping with Robox: Chime Alarm Clock

By | Robox User Blog, Stuff and Things | 4 Comments

There is a joke around about how engineers operate under the idea that “If it isn’t broken, it doesn’t have enough features or do enough.” Humorous as that is, it sometimes comes close to the truth. In the pursuit of making it better, I have started a project that, like many of my projects, is targeted toward making something better for someone else.

Basic battery powered alarm clock

Basic battery powered alarm clock

My girlfriend uses a cheap, battery powered alarm clock that with a rather potent buzzer. She buys them for about six dollars each and they last about a year. I asked her one day why she didn’t get a nicer one and she gave me a couple of reasons, among them that the alarm clock she really wanted was an expensive unit that used a chime bar instead of the buzzer. These units sell for about a hundred dollars and consist of a bamboo box with the chime bar, a small striker, and a flip-up lid with the clock display. I looked at one of them in the store for about five minutes and was not impressed with the build quality or the value for the money. The clock didn’t even seem to be backlit. Based on her requirements and my own desire to do something creative, I started the design for my own version of the chime alarm clock.

The design centers around an aluminum bar chime tuned to A# or about 1121 MHz. I used an app on my smartphone to do the tuning so it isn’t exact. The bar is suspended on woven aramid fiber strings so that I can tighten it very tightly and so that they don’t stretch significantly over time. An aluminum frame will support the mounting strings and provides the strength needed to keep the strings tight. The outer housing will be made of hardwood, probably a combination of different colors and grains to give the finished unit character. Right now I am thinking about maple for the sides and walnut or mahogany for the end caps. One of the parts that I disliked about the commercial version is that the chime sits inside the box and is somewhat muted by the box. To better expose the chime, the aluminum frame will slide upward out of the wooden housing when the user does some action to open the clock.

A large backlit 2-line LCD will be framed into the front of the wooden housing and will display time, day of the week, date, and alarm status. The housing will conceal and mount the Ardiuno Nano and time chip that will run the clock and the battery pack that will provide power. The chime is activated with a pull-type solenoid mounted to the aluminum frame. The solenoid pulls a lever to which a striker is mounted. The striker impact on the bar causes the bar to chime and a spring retracts the solenoid and striker bar. I started thinking about the whole arrangement of the solenoid and striker bar with the idea that making the striker movable along the striker arm would enable a rudimentary volume control.

Test setup for the alarm clock

Test setup for the alarm clock

To test this idea, I used the Robox to fabricate a frame, the striker, and the striker bar. The frame was made from ABS, printed on the default fine settings. It turned out pretty well, though it did require some deburring. I printed it to allow the solenoid to be pressed into the housing with a little force to keep from having to secure the solenoid with a fastener. The chime bar and striker were printed with Taulman T-Glase PETT. This material has become my new favorite for printing on the Robox because it is strong, durable, doesn’t break, and prints rather easily. It also seems to give better surface finish results than ABS and it doesn’t smell nasty. Since the striker would be a forced fit onto the striker bar, I wanted the flexibility of the T-Glase for the striker. I ended up needing to print four strikers since it has an overhang and the first one fused to the support material too well. Two of the replacement parts also had issues but the fourth turned out ok. I drilled a couple holes for the aramid string and an anchor screw and assembled everything. Tapping the chime bar with a hex key yields a clear note at about A# and flicking the solenoid plunger with a finger results in a slightly flatter tone, mostly due to the dwell time of the striker against the chime bar. The Arduino controlled motion showed significantly better performance and, other than some clacking from plastic parts, performed very well: (AlarmClockVideo) The test of the volume concept showed that there was a volume change by moving the striker closer to the striker bar pivot point but that it wasn’t really significant. I am going to look into adjusting the position of the solenoid to see if that will help adjust the volume.

The next step with this project will be to tune the chime bar to the exact note she wants using the test frame. Since the chime bar is supported on the aramid strings, I am not foreseeing a difference in tone between the ABS mount and the final aluminum mount. I am considering replacing the PETT striker with an aluminum or acetal resin striker just to give a little clearer note.

The Robox will most likely be used to prototype the remainder of the frame parts and the extension/retraction mechanism to make sure they work as intended. One of the reasons I bought a 3D printer is to be able to test these complex mechanisms before I spend the time and money making the final parts. I like to work with hardwood, the exotic varieties of which can be rather expensive. This project will likely be the first one to use the 3D printer heavily and I foresee several iterations to get everything set in the housing so I have room for the batteries, all the electronics, and the chime frame.

CAD Model of the Alarm Clock Design

CAD Model of the Alarm Clock Design

A few more details I plan to add on the alarm clock will, I hope, blur the line between form and function to make the final result a balance of both. These include a programmable display color, adjustable backlight brightness (a requirement from my girlfriend), capacitive sensors for the sleep bar and possibly for the control panel. Silver and contrasting hardwood inlays  will add a decorative touch and will give visual indications of the control locations.  I chose Arduino for the brains because of the price and flexibility of the platform. With a little programming and a few parts, I will be able to give this alarm clock functionality that the cheap alarm clocks don’t have, including multiple alarms, alarms by day, different display brightness depending on the day, or with an added light sensor, on ambient light. A battery status is easy to add and will minimize the chance of getting caught with a low battery when it is inconvenient.

This is the design state after a year of off-and-on work. Hopefully with the Robox I will be able to run the prototypes of the frames and get things moving a little faster. Maybe in time for Christmas this year.

Robox bridges personal and professional 3D printing

Robox in the Workplace

By | Robox User Blog | 2 Comments

Professional 3D printing usually brings to mind large, expensive machines and expensive media. Prints are of good quality, materials are strong and often production quality with the right materials. The price, though, makes it unreasonable for every engineer to have a 3D printer or even to have access to one if the company isn’t large enough to afford them.

Personal 3D printers often conjure images of Yoda prints and plywood frames holding bare electronics. Only recently have personal 3D printers become polished enough to be part of the toolkit available to anyone who has the need for one. Each user also needs to be able to know the software to generate the 3D files and to tweak the output to the printer to optimize the results.

The Robox has the potential to change all of this. With the material profiles and print profiles being easily transferable and shareable, with the AutoMaker software able to import most stl files, it isn’t hard to see a personal 3D printer as part of an engineer’s standard equipment. For the cost of a second workstation, every engineer in a company has the ability to rapidly and cheaply print a prototype. The profiles that have been shown to work well are shared with everyone and the same materials can be used throughout the company. Because the material profiles are loaded on the smart reels, material changes are as easy as swapping reels. The AutoMaker software cuts the learning curve providing fast results with a minimum of training.

Couple the out-of-the-box simplicity of Robox with a local support representative and you have a 3D printer solution that is very difficult to beat. A new machine can be set up and ready to print onsite in less than two hours, faster if the initial calibration is performed by the support representative before delivery. The AutoMaker software plays a large part in that speed due to its simplicity and user-friendliness. A new user who is trained in CAD will be able to start importing profiles and parts to be printed in a matter of minutes and will be ready to print shortly thereafter.

Robox and the shared common framework that accompanies the product represent the next step in day to day access to 3D printing, both for the home user and for industry. Improvements slated for Robox will build on that strength. The dual material head, for instance, will allow support material or dual color prints with the same Robox simplicity that the current single material capability provides now. There is room to grow and Robox has the potential to continue to be a contender in a rapidly growing market.

I chose to support the Robox project for the qualities outlined. As an aerospace professional working on complex, cutting edge technology and also as an inventor and maker, I see the need to be able to make a physical model very soon in the development cycle. My use of the Robox will be to create prototypes, to make test units so that I can confirm geometry and operation of a design, and to make parts for small production runs of products. In ten years of experience in my industry and in twice that in personal projects, the uses for a 3D printer have been uncountable. The Robox brought simplicity, ease of use, and a high level of adaptability, both in material types that it can use and in the potential to change out the head to add capability. Even in the beta period, the Robox has been valuable in confirming that a design works and has potential in the application it was intended to fill. The prototype I printed helped the customer feel comfortable with a technology even though it hasn’t been fully tested and confirmed. That alone makes the project worth the money and time I have spent this far and in the future I fully expect it to pay for itself in ease of prototyping and in reduced production cost as I move products to market though small-scale production.