week 8, thursday: 16 customized and printed mouthpieces

This post is quite a long story, in which we will tell you about the progress of this week, after reflecting on the first stages of this project. We will also give you an idea of how much time the steps in this project will take you. This week, we had two solidworks-models that were ready to be adjusted! See the two screenshots below:

Screenshot drake 2 mouthpiece (click to enlarge):

Schermafdruk 2014-10-24 13.43.52

Screenshot mystery alt saxophone player mouthpiece (click to enlarge):

Schermafdruk 2014-10-24 13.42.44

The past weeks, we discovered that it takes a lot of time to reverse-engineer a mouthpiece. To go from the mouthpiece to an X-ray is not such a big step, but it takes about half a day. The next step is to ‘compile’ the slices we obtained into a file that can be used in Meshlab – the software used to get from the slices to Meshlab is called Avizo, and it takes quite some time to get this setup and to have a basic understanding of this. Once you have gone through these steps, a 3D-image can be imported into solidworks and the true modelling will begin. This is the part that takes lots and lots of time, and you will need a high level of solidworks-skills. To give you an impression of the amount of time this takes: our supervisor Zjenja is experienced with this kind of work and it takes him about a day to get this done. It took us a lot more time: if you are new to Solidworks, it will take you all five weeks of the project (!). If you are a third-year Industrial Design student, you should count on ten full days. To go from zero to a printed and working design. Once you have a model that is similar to the scan (a few days later), be quick to get the first prints out! You will need to test, hear & feel how they sound. This is where the fun part starts!

After all of this, it took us about an hour to model 12 different variations of the alt saxophone mouthpiece. To get the right settings off the printer and to get some nice prints out is a story of its own, so keep reading..

12 variations have been modelled, and are being printed as I type.  One variation used a ramp, the other a baffle. They are located in the spot that has the highest turbulence – as determined by Aerospace Engineering PhD Valerio. A shot of his research, where red-orange-yellow marks the part with the highest turbulence:


The ramp looks like this:

Schermafdruk 2014-10-24 13.46.23

The baffle looks like this:

Schermafdruk 2014-10-24 13.49.59

The ramp has been made in the following variations:

  • postion: ‘reference position’ and a position closer to the tip of the mouthpiece
  • size: small, medium, large
  • this gives 6 mouthpieces

In that same location, a baffle has been made (a small gap) in the following variations:

  • position: ‘reference position’ and a position more inwards
  • size: small, medium, large,
  • this gives 6 mouthpieces

 The above mentioned variations have all been made to the mouthpiece of our mystery alt saxophone player.

About the bariton sax mouthpiece, drake 2: we are printing 5 variations at the moment. The variations are as follows:

  • baffle size: small, large
  • baffle position: ‘reference’ vs inwards

We were able to obtain the solidworks files of this bariton mouthpiece from last years project group! This has saved us a lot of time, plus it enabled us to follow up on their product. Last year, their mouthpieces had been printed with an object printer as found in the PMB (workshop) of the Faculty of Industrial Design Engineering. The printer we had acces to this year, has a higher resolution. Also, the material it prints in  is much nicer to use for  saxophone player!

This is a nice breakthrough in this project, especially regarding the feedback of last year’s tester and professional saxophone player, Peter Broekhuizen. The main reason to discontinue the use  of his 3D-printed mouthpiece, was the material it was printed in.

Another adjustment we made to last years model was the fitting to the neck of the saxophone. It should now fit better and be easier for the player to ‘tune’ the mouthpiece.

some failed prints and a day later..

3D printing is a pretty cool technique, but it proves time and time again to not be as straightforward as it sounds. Results can be really nice and reproducible, but they can also be a disappointment..

  • Printing a batch 6 mouthpieces using the form1 printer takes about 20 hours.
  • Printing a batch of 10 mouthpieces using the objet printer takes a little less time.

A few pictures of last night’s failed prints:







After we restarted the prints, the printer failed us again – we are slightly changing our solidworks models right now. Our third print will be started in about hour from now..


some mouthpieces in the exposition of the Dutch Design Week

It was really nice to see that last year’s mouthpieces were part of the TUDelft exposition in the #DDW Dutch Design Week! A picture below of the exposition, next to a short movie explaining the ‘how and why‘ of the project:






Be Sociable, Share!

Comments are closed.