Project Logs

Collapse

• hand sanitizer bottle dispenser

  John Opsahl • 04/11/2020 at 17:22 • 0 comments

  In response to the Coronavirus pandemic distilleries have begun to offer hand sanitizer in one liter plastic liquor bottles. This log describes a hack to add a self sanitizing dispensing mechanism to these bottles for less than $5. Simply mount the bottle upside-down and tap to the bottom of the mechanism to dispense hand sanitizer.

  I designed a 3D printed part to fit around the existing bottle cap. It provides a surface that the spring can be loaded against and clearance from the bottle cap for the rubber stopper to travel within during the dispensing motion. Sanitizer flows through the slotted holes of the part.

  A hole is drilled in the bottle cap the size of the tapered rubber stopper. The spring pulls the rubber stopper against the bottle cap (via the the bolted connection) to create a water tight seal. The seal is broken and hand sanitizer flows through the hole in the bottle cap when the user presses on the head of the bolt and compresses the spring.

  The build was successful but I found a few things I would consider on the next revision. The rubber stopper absorbs the alcohol and swells so I had to cut it down a little to get it to seal against the bottle cap properly. It took maybe two or three tries to get the right rubber stopper shape. I will consider extending the clearance from the cap to the 3D printed part and different ways to cut the stopper to make it work the first time. Additionally, after a dispense, the alcohol continues to drip for three or four drops from the head of the screw. Not a big issue since alcohol evaporates and doesn't create a wet spot below the dispenser, but it's also not ideal. I will consider extending the slotted holes in the 3D printed part towards the tip so there is a more direct flow path to the head of the screw and less opportunity for the sanitizer to pool behind the tip and flow slowly through the bolt hole. 

• foot pedal toilet flush

  John Opsahl • 04/04/2020 at 19:44 • 0 comments

  A commercially available kit for converting your home toilet to hands-free flushing will cost you $25 or more. If you have a 3d printer, the hack I detail in this log will cost you less than $10. 

  The mechanical connection to the toilet handle is a 3D printed hook plate attached with a wall hook adhesive strip. I got a little overzealous with 3D printing here. You could probably just use a wall hook with an adhesive strip. It took less than twenty minutes to 3D print at 30% infill and 0.6mm wall thickness.

  The foot pedal is simply two 3D printed plates mounted on a hinge. Adhesive picture frame mounting strips are used to secure the bottom plate to the floor. A 3D printed loop (similar to the loop of the handle hook plate above) on the backside of the top pedal plate is used to connect the rope to the pedal. The pedal plates printed in less than two hours at 30% infill and 0.6mm wall thickness.

  The only thing I may consider for the next revision is reducing the number of locations were dirt can accumulate so it's easier to clean and disinfect. 

• forearm deadbolt turn

  John Opsahl • 03/30/2020 at 17:56 • 0 comments

  Deadbolt lock levers are another door feature that require hands to operate. This project log discusses a 3D printed hack that allows the user to turn the deadbolt lever with their forearm. The video below demonstrates the right handed and left handed forearm motion required to unlock and lock the deadbolt with this device.

  The greatest challenge of this hack was accommodating the large variation in deadbolt geometries. Deadbolt levers have different widths, heights, and lengths. Some are located in the middle of the deadbolt cover plate while others are offset. Because of all the geometry variations I wasn't able to develop a universal forearm deadbolt turn design. Instead, the dimensional parameters may have to be tailored to each deadbolt design. 

  Perhaps the most difficult geometry to design around is the taper on the sides of the levers. The tapered sides makes it very difficult to make a mechanical connection to the lever without gluing something to it. To get around a permanent glued connection, the design makes a connection to the door using adhesive strips and is raised off the door just enough to mechanically engage the deadbolt lever. 

  Three parts are required for the assembly - the rotating disk and two ring halves. There are two cylindrical cutouts at 90 degrees from each other on the face of the rotating disk to fit the user's forearm for right handed and left handed motions. M4 bolts are used to capture the rotating disk and connect the two ring halves. The deadbolt lever is made visible through the assembly so the user can see it's orientation and determine if the door is locked. 

  It took five and a half hours to 3D print all three parts at 30% infill and 0.6mm wall thickness. I used drawer slide lubricant on the rotating contact surfaces. My landlord probably doesn't want me to drill holes in the door so I used adhesive picture frame strips to secure the assembly to the door.

  I would consider this a successful build. The only things I might change on the next revision are removing the sharp corners on the face of the rotating disk and raising the front face of the rotating disk a little farther from the door face (at the current offset, the user's hand may occasionally hit the door frame).

  At this point, I am starting to run out of door features to hack for hands-free operation.

• foot pedal door unlatch

  John Opsahl • 03/29/2020 at 15:53 • 0 comments

  This Saturday I hacked together a foot pedal device that allows the user to turn a door knob or door handle without touching it with their hands. I decided at the beginning of the design process that this hands-free door unlatching device should not require mechanical modifications to the door (i.e. no drilling holes in the door) and should be easily adaptable to any door knob and handle geometry. If you have a 3D printer, this device can be made and installed on any door for less than $5. 

  A hose clamp is used a the door knob to adjust to any size of door knob diameter. If connecting to a door handle, a smaller diameter hose clamp could be used. I put the clamp strap in a 1/2in vinyl tube to prevent the strap from scratching the door handle and to protect the user from sharp edges in the event that they need to turn the handle to open the door. The rope is tied around the hose clamp screw using a square knot.

  The foot pedal assembly consists of two 3D printed parts. The foot pedal slides up and down the rail plate. Three adhesive picture frame strips are used to attach the rail plate to the door. The rope is tied to the hook feature of the foot pedal. A petroleum jelly lubricant was applied to the contact surfaces between the foot pedal and rail plate to reduce friction and prevent binding.
  

  I included a 0.25mm clearance between the foot pedal and rail plate slide feature in the 3D model. This ensured that the foot pedal would slide on the rail plate without much friction. The 3D printed parts still had more friction than I would have liked. On the next revision, I will use a 0.5mm clearance. 

  The parts printed in less than six hours at 30% infill and 0.6mm wall thickness. No support material is required. 

  By itself, the foot pedal door unlatch device is not hands free on the pull side of the door without the custom forearm door pull. Unfortunately, adhesive picture frame strips are not strong enough to secure the forearm door pull to the door; a screw or bolted connection to the door is required.

  I am pleasantly surprised with the simplicity and flexibility of the final foot pedal door unlatch design. The adhesive picture frame strips enable it to be mounted on any type of flat surface (e.g. wood, glass, metal). The string length can be easily adjusted for any type of door knob or handle geometry. Using the spring of the door knob or handle to return the foot pedal to the top position reduces the foot pedal assembly complexity and part count.

  The next unsolved hands-free door challenge is a forearm door hook for the push side of the door. For doors that do not automatically close when you go through them, a pull hook is needed on the push side of the door to pull the door shut after passing through from the pull side. At the moment, I am imagining a modified version of the forearm door pull that I modeled previously (https://hackaday.io/project/170412-hands-free/log/174912-3d-printed-forearm-door-pull). 

• 3D printed custom forearm door pull, part 2

  John Opsahl • 03/26/2020 at 05:04 • 0 comments

  This next revision of the forearm door pull is the most promising design of this project. It incorporates many of the successful design features and avoids many of the pitfalls of previous designs.

  The forearm door pull is installed at the edge of the door as shown in the image below. A low volume, high stiffness (and 3D printable without supports) triangular profile is used along the curve of the hook. The hook is inset from the edge of the door so as not to create a catch hazard when the user enters the door. A small straight section at the base of the hook spaces the user's hand away from face and edge of the door as the user swings the door open and releases their forearm from the hook. It has a wide face on the front of the hook so if someone entering the door swings it into someone trying to leave through the door there is less likelihood of causing injury.

  To achieve a total print time of less than six hours, I reduced the height from 60mm to 50mm, decreased the nominal wall thickness from 8mm to 7.5mm, and decreased wall thickness from 0.8mm to 0.6mm. Even with these reductions, the strength of the part is more than sufficient. 

  Stress and deformation checks showed no red flags. The simulation used fixed supports at the bolted connection to the door and a 900N (200lb) force at the tip of the hook. 

  It ended up printing in five hours and fifty minutes at 30% infill and 0.6mm wall thickness. The hook deflects more than the previous design if I pull on it with my hands, but I suspect the additional deflection will not be an issue.

  You know it's a good build when you solved all the previous issues and struggle to find any new issues. I am pleased with how this design turned out.

  The door pulls I have investigated so far assume that there is not a latching mechanism on the door. Bathroom doors and entrance doors to buildings are examples of doors without latching mechanisms. Residential front doors and internal building doors are examples of doors that use latching mechanisms. The next great design challenge of this "hands free" project is to develop a foot or forearm operated door latching/unlatching mechanism. 

• 3D printed custom forearm door pull, part 1

  John Opsahl • 03/24/2020 at 00:30 • 0 comments

  The previous forearm pull design I developed was a 3D printable version of a commercially available product. The custom forearm door pull detailed in this post in my own design. It is mounted at the edge of the door and allows the user to pull the door open with their forearm in a more natural motion. The previous forearm door pull (bottom hook in image below) is mounted away from the edge of the door. It requires that the user stand outside the path of the door and quickly pull their arm from the hook near the end of the door pull motion. If the user doesn't pull their arm from the door quick enough, their hand and forearm may hit the face or side surface of the door. If their hand hits the door, it's no longer an effective hands-free door pull design! Instead, the forearm door pull that I have devised (top hook in image below) will swing away from the user's forearm with the natural motion of the door. One downside of this new design is that it has potential catch a user's arm or clothing when entering from the "push" side of the door or closing from the "pull" is of the door. I will address these challenges in the next revision of this design.

  I use a triangular profile along the crossection of the hook to add stiffness and still allow the part to be 3D printed without support material.

  A quick deformation and stress check shows that highest stress occurs where the hook connects to the base (so I increased the fillet radius there) and deformation is less than 4mm. I used a 900N (200lb) force at the tip of the hook. 

  It print in six hours and fifteen minutes at 30% infill and 0.8mm wall thickness. It deflects a little when I hold the base and pull on the tip of the hook, but overall I think it is strong enough for the application.

  I am happy with this build. Writing this log has got me excited about some changes I am going to make to the next revision.

• 3D printed toe door pull

  John Opsahl • 03/22/2020 at 04:20 • 0 comments

  The toe door pull is the next style of hands free door pull device I have decided to design for 3D printing. It is essentially a toe cup that you hook your foot under to pull the door open (see example image below). Commercially available versions can run up to $100 each. The first revision of this design is similar the commercially available designs but after a round of testing I believe this 3D printed design can be made much simpler.  

  An example of a commercially available toe door pull.

  Initial target for these 3D printed designs has been be able to print them in six hours or less. One strategy I use to achieve quicker prints is to design them in a such a way that support material is not required. This requires that no undercut surfaces are at angles larger than 45 degrees relative to the build plate surface and that undercut surfaces are continuous. One of my favorite parts about this project is imagining the geometries that both achieve design intent and meet "no support material" requirement.

  I ran a stress analysis on the part to understand how it might realistically deform and break during intended operation. I don't have high confidence in the numbers generated by the simulation and use it primarily as a guide on where to add additional stiffness/strength to the part. For example, after noticing that the highest stress was near the side screw mount, I added a larger fillet between the screw hole and toe cup to increase strength at that location. A 900N (200lb) load was applied to the inside face of the toe cup.

  The design printed in six hours and ten minutes. I used 30% infill and 0.8mm wall thickness. I am a little worried that the angle of the toe plate is too shallow and user's toe will slip out of the cup. Hope to get it mounted on a door and start testing soon.

• 3D printed forearm door pull

  John Opsahl • 03/20/2020 at 23:03 • 0 comments

  A forearm door pull is another alternative that enables someone to pull open a door without touching it with their hands. They typically sell for about $90 for two of them. It's a simple enough geometry that its no problem to 3D print on the cheap so lets get designing.

  The concept is simple. Place your forearm between the hook and the door and pull.

  The image below is an example of a commercially available forearm door pull installed on a bathroom door.

  The only unique feature of the design that I have contributed is that it can be 3D printed without any supports. Otherwise, it is essentially just a wall mounted hook.

  For a quick stress and deformation check I applied fixed supports at the mounting holes and applied a 900N (200lb) force at the inside face. No major concerns from these checks.

  The 3D print time is about 5 hours 45 minutes. I used PLA with 30% infill and 0.8mm wall thickness. 

  It is probably stronger than it needs to be. On the next revision, I will reduce thickness in some areas to arrive at a faster print time.

  The only issue I can anticipate is that your forearm might slip off the inside face of the hook when pulling the door open. To address this I will either lengthen the inside hook face, curve the end of the inside face towards the door, or add a high friction surface on the inside face.

  I bought an old public bathroom door from a resale store this weekend so I plan to do some actual testing in the next week or so.

• 3D printed foot door pull

  John Opsahl • 03/17/2020 at 04:42 • 1 comment

  Doors are probably the biggest opportunity for reducing the number of shared surfaces in public spaces. It's simply not possible to open some doors without a firm grip around the handle or knob. An alternative solution has been to pull the door open with your foot rather than your hands. Several designs are commercially available at around $30 a piece so lets make a 3D printed one on the cheap.

  The video below shows how to use a foot door pull. 

  I imagine one major disadvantage with foot door pulls. If someone pushes through door from the other side at the same time that you are approaching the door, they might throw the foot door pull into your foot. I certainly wouldn't want to be wearing sandals when that happens.
  

  The commercially available designs are made of aluminum and steel. I knew that I would need to add rib features to the 3D printed design to make up for the reduced material stiffness of PLA. Also, the "toe grab" features are a little larger for the same reason. 

  Next as a quick check to ensure the ribs and sides were thick enough, I simulated a 450 lb downward force on the front "toe catch" section of the part. The simulation shows high stress that approaches the the tensile strength of PLA near the bottom inside corner of the ribs. I think my interpretation of the simulation results was conservative. The final print was much stronger that I anticipated. 

  It took almost 10 hours to 3D print with 60% infill. I haven't installed in on a door yet but I did jump up and down on it a few times. I think there is opportunity to reduce the amount of material and still maintain all the functionality. The 60% infill is was overkill from a structural standpoint. A reduction in infill percentage will greatly reduce print time as well.

View all 9 project logs