3D Printed Emergency Whistle (Over 115 DB!)

While going on a hike, it is essential that you always have an emergency whistle at your disposal. However ordering one online or going to a store can take a while. That is why I designed a quick emergency whistle that can be printed in under 20 minutes. It is over 115 db loud and has a dual pitch of 2.6 kHz and 2.3 kHz .

For those of you who just want to print the whistle it can be found here.

I'm a 9th grade student at the Maria von Linden Gymnasium

Supplies

1. 3D Printing Filament (PLA/ABS/PETG)
2. (Optional) Keychain/String

Tools

1. Fusion 360
2. 3D Printer

To design a whistle I first had to know how a whistle functions. This video by Steve Mould really helped me get started with designing my whistle. Whistles can be made very differently, but I chose to go with a classic wedge design, because most emergency whistles that I saw online had this kind of design.

In this kind of whistle design, a stream of concentrated air goes past a wedge. Because of vortex shedding and flow instability, the stream of air starts to oscillate up and down. This causes waves of varying pressure, to exit out both sides of the wedge. This alone can create sound, but usually the resulting tone is too high-pitched and quiet to hear. To fix this a resonating chamber is added to lower the pitch and amplify the sound.

I knew I would be going through several revisions of the whistle so creating a good, adjustable base was key. Fusion makes it really easy to alter previously created constraints. In addition, the timeline and parameters made Fusion ideal for this project.

• To start off I created two rectangles which would form the resonating chamber
• After that I added a collinear constraint to the right lines of both rectangles
• Next I constrained these rectangles so that the wall thickness is 4 x 0.4mm (The extrusion width) = 1.6mm
• Finally I constrained the height of these rectangles to be 5mm (I did not decide on a width yet)

• To create the mouthpiece I extended the upper wall out using a rectangle
• After that I created 3 lines in the shape of a parallelogram onto the lower wall
• I added parallel constraints to the opposite sides of the parallelogram, added a collinear constraint to both lines on the far right and added a dimension of 3 mm between both lines
• Finally I added an angle dimension of 157.6 degrees to the bend in the lower wall

• To create the wedge I added a horizontal line on the top wall and dimensioned it 10 mm from the mouthpiece
• Next I created a diagonal line through the upper wall and set the angle of this line to be 15 degrees
• Finally I dimensioned the distance between the two created lines at 3 mm

1. First I extruded the entire wall up 10mm
2. Then I added ends to both sides using 2 extrude features
3. Finally I just filleted all sharp corners (excluding the wedge) for a cleaner look

I printed this version on it's side with a 0.1mm layer height, however it didn't function. I looked at the data from a frequency analyzer, which disproved my assumption that the sound produced was too high pitched or low pitched to hear.

My next assumption was that the air was not concentrated to a fast laminar stream yet, so for my second iteration I edited the original sketch and changed the opening size dimension from 3 mm to 1mm. When I printed this version though, it still didn't produce any sound.

For version 3 I further concentrated the air, by adding a channel for the air to flow through. I did this by adding a ledge on the bottom of the wall in the original sketch. In addition, I changed the size of the opening near the wedge from 3mm to 1mm. I quickly started printing it, but it failed to produce a sound yet again. However I felt like I was on the verge of success and quickly came out with another revision.

This time, I sized the resonating chamber down quite a bit, and added a slight upwards tilt to the ledge that was in V3. I hoped this would concentrate the air to a much finer stream and that the smaller resonating chamber would help bring the created tone into the audible range. To get a cleaner and more precise wedge and air channel, I move the seam to the outside near the mouthpiece. Then, I put my printer to work and, because of the smaller resonating chamber, the print was finished in no time even for a 0.1mm layer height. I anxiously pulled the print off the printbed and blew into the whistle. To my surprise an extremely loud tone was produced that nearly blew my ears off. I measured the loudness with an online website and got a maximum loudness of 110 db! It was very high pitched at a measured pitch of 4.5 kilohertz! I reprinted the design using a higher quality filament and got a maximum loudness of 128 db! Hearing protection definitely advised! The online website could be a bit off, but even 110 db is really loud.

I wasn't really satisfied by my previous success, because although I had produced an insanely loud whistle, I hadn't learnt much and my approach was very unscientific. Because of that I decided to test small variations of the functioning whistle to learn a bit more and hopefully use that information to create a dual tone variation. This whistle would produce two differently pitched tones, which would interfere with each other and create a shrill sound which would be more attention grabbing. I decided to test varying resonance chamber sizes and wedge angles.

Designing many variations of the same whistle can seem tedious, but Fusion 360 comes to the rescue once again.

1. I created 2 parameters (ResCham: 12.4 mm and WedgeAngle = 15˚)
2. Next I edited a previous sketch and set the wedge angle to be equal to the parameter "WedgeAngle" and the resonating chamber length to the parameter (ResCham)
3. If you have a paid or student license you can use the "Configure" function in Fusion to set a bunch of parameter values and have it output the corresponding models. However I have the personal license (I don't know why I don't have the Student license) and had to change the parameters and save each model individually.

This is the measuring tool I used to measure the loudness in decibels: https://youlean.co/online-loudness-meter/

And this is the website I used for measuring the frequency in hertz: https://www.compadre.org/osp/pwa/soundanalyzer/

I used one continuous blast of air and recorded the loudest produced tone and the average frequency from a meter away.

My results are in the picture above. These show several things:

1. The greater the wedge angle the quieter and more low pitched the sound is
2. The greater the chamber size, the quieter and more low pitched the sound is
3. A mm of extra chamber size equates to a frequency change of roughly 0.28 kHz
4. Wedge angle has a minimal effect on pitch and loudness with little correlation

I wanted a dual tone whistle with a frequency around the 3kHz range because I read that this frequency can be most easily picked up by humans. To do this I calculated the required resonating chamber size by first looking at how much the frequency is changed by changing the chamber size by 1 mm. Using my data from the test before, a 1mm larger resonating chamber equated in a frequency drop of 0.288 kHz. That told me that I needed a resonating chamber size of roughly 17.4 mm. I still wasn't so sure how accurate my calculation was, because I used a linear function to approximate the frequency. In the real world however, a pitch of -4 kHz makes no sense, but my function says that this would be the case with a chamber size of 42.4 mm. Nevertheless, I proceeded with printing the whistle and was surprised, when the frequency (3.32 kHz) closely matched the expected value (3.21 kHz).

A two tone whistle has, as the name implies, two tones. For this I had to combine two whistles into one. I could just make two whistles and join them using the combine feature, but the nozzle opening would be too large to comfortably blow into. So I had to make the whistle thiner. To do this I just edited the extrude feature and scaled the extrude length down by 2/3. Then I increased the resonating chamber size by 50% to get a similarly sized chamber.

1. To add the second tone, I scrolled back in the timeline to before I did all the filets and created a sketch on the side of the first whistle
2. Next, I projected the previous sketch and made sure that the projected sketch would not change once the original sketch changed.
3. This meant, however that I had to constrain everything once again
4. After that, I created another parameter called ResCham2 and used this to set the length of the second resonating chamber
5. Finally I extruded the resulting sketch up 6.667 mm and added a upper cap using another extrude feature

1. To add the keychain, I created a sketch on the side of the whistle and drew a isosceles right triangle with the base on the side of the whistle
2. Next I created a circle tangent to all three lines of the triangle and used this midpoint to draw a 3mm circle around it
3. After that, I extruded this sketch all the way to the other side of the whistle
4. To remove a bit more material I created a triangle like sketch on the top and used it to cut off some excess material.
5. Finally I filleted all sharp edges around the keychain

Previously I undid a lot of the filets using the timeline and now I had to go through all these filets and fix all the errors, by redefining where the fillets should go.

Finally I used a sketch on the front to add the text "Emergency Whistle"

One thing I noticed on my previous versions, was that the whistles slipped out of my mouth while holding my ears shut and blowing as hard as I could. To fix this I added little tubes using the pipe feature in Fusion. Adding them was as simple as creating a sketch with a single line following the tube path, and using the pipe feature to create a 1mm wide pipe along the line. Finally I filleted the edges and repeated this on the other side.

Print as is with regular settings:

1. 0.2 mm layer height
2. 3 walls
3. 10% infill
4. No Support
5. Position your seam AWAY from the wedge using paint on seams (Prusaslicer) or z-seam alignment (Cura). Position it near the mouthpiece.

The STL file for the final whistle I designed can be found here: Whistles

Although I didn't get to print my latest version before a trip (due to my printer being broken), I did manage to test V4 and V5 out in nature. They both performed extremely well and could be heard in the woods from at least 350 meters away. My mother reported hearing the higher pitched sound (from V4) better than the lower pitched on in V5. This matches my observations and tests (V5 is indeed quieter than V4). However lower pitched sounds should penetrate objects better. My final emergency whistle combines both of these whistles to get the best of both worlds.

After the trip I fixed my printer and printed the latest design. This one worked as expected and the created dual tone design made it difficult to ignore. I measured the loudness and pitch and got 126dB up close and 120dB from one meter away. (Take this measurement with a pinch of salt because I'm not sure the website accurately records loudness) The pitch was around 2.3kHz and 2.6kHz combined, each with their respective overtones and timbre.

All in all I learnt a lot from this project. From learning the theoretical aspects of whistles to actually implementing them into a easily 3D printable version I really enjoyed every step of the process. I learned how to properly use parameters in Fusion and learned about all the intricacies of whistle design. Plus I now have a whistle I can actually use in the future and can do you by 3D printing your own.