Improving the Upright Piano Action: A Mechanical Engineering Project Follow Up.
The STEMinists is a group of senior mechanical engineering students with the task of solving a problem by applying engineering knowledge while providing evidence that validates their solution. After some consideration, we decided that redesigning the action of an upright piano, which is one of the two types of traditional pianos, was a worthy endeavor. To read a humorous explanation of the project and to learn about our initial thoughts please refer to our first blog:
On this blog, we will discuss about our plans to move ahead with project, the progress we have managed to achieve so far, and some of the challenges we have come across. In order to illustrate these topics, we have created a handy gantt chart for the first half of the year that reveals two concerning facts: We only have about a month to finalize our design, and perhaps more alarmingly, we only have 10 days for completing the first milestone (yikes). The good news is that we got spring break to catch up, so that’s exactly what we’ll do. Go back to the backseat personal life!
As indicated by the chart, about 60% of the work done on spring semester involves research, and besides an attempt at building a functioning upright piano action and a couple of sketches, that’s the work we have to show. We have learned a couple of relevant facts that we want to share in order to shine a light at what exactly is it that we are trying to achieve by redesigning a piano action.
The Problem at Hand:
We have discussed previously that the grand piano action is more “responsive” than the upright action due to it being assisted by gravity, but how does that work exactly? The figures below show an animation of how both actions operate, and one can clearly see the similarities between the two mechanisms. For example, both actions have dampers that rest on top of the strings to keep them from vibrating until the key is pressed. Both actions propel a felt hammer in order to hit the strings, and as soon as the hammer hits the string, it must be released back immediately to allow string vibration.
However, playing a grand piano feels different from playing an upright action, and even people who can’t play the piano (that’s three of us) can tell the difference. In order to reset, a grand piano simply relies on the weight of the hammer and damper, and gravity does the rest.
On the other hand, the upright piano relies on springs and the bouncing of the felt hammer after hitting the string to go back to its initial position. Other than creating an uneven feeling for pressing different keys due to springs deteriorating at different rates, this mechanism has another flaw: To play the key again, the key must completely reset back to its resting position. In contrast, the grand piano keys only have to move one-third of the way back in order to be pressed again and emit sound. If one tries to do the same on an upright piano, the key will go back down, but the hammer won’t be able to hit the string. This flaw is due to the L-shaped action part that moves the hammer forward called “the jack.” As shown in the animation, the jack’s top changes position during the action motion, and it doesn’t come back were it’s supposed to go unless the key is fully released.
Another feature on the grand piano that isn’t present on the upright counterpart is the presence of a repetition mechanism. The repetition mechanism allows the pianist to quickly repeat a note without having to rely on the piano to reset the action. Generally, the more expensive, higher echelon grand pianos have this feature. It’s generally accepted that pianists who don’t practice on a grand piano are not ready to learn difficult techniques that take advantage of this repetition mechanism. One of our potential designs utilizes the repetitive mechanism in conjunction with the traditional upright piano action.
To study these problems, we researched if there had been any attempts at solving this issue, and we came across the Takumi piano modification.
The Takumi modification consists of two additional springs that make the hammer bounce back to its original position at a faster rate. The idea seems promising at first, but the only team member who can actually play the piano was skeptical about the results, and apparently most people who know anything about pianos in the online community agree that the Takumi design didn’t prove to be a solution. They examined their success by simply showing that it can be played faster, but doesn’t mention whether or not the key must go back to complete resting position to be played again. We will continue to do research on this and other proposed solutions, and we plan to validate our results using other measurable experiments to test for responsiveness, evenness, and how far the key must reset before it can be played again. We also plan to become familiar with piano tuning to enhance the sound quality of the upright piano. More research on wood-working and piano crafting will be required.
Making a Model:
As far as research goes, that was a summary of what we understand so far about the problem. We gathered all this information and brainstormed ideas of how to improve the upright action mechanism, but we needed to physically work on something to verify that we understood the theory. We decided to order an upright action model from Ebay so that we can assemble it and then visualize the movement of the action. This was also meant to help us learn and determine a solution for the redesigned mechanism;moreover, having a functioning model would make explaining to other people what we are doing much easier. However, when we received the kit, there were some missing parts and the schematics/directions were not consistent with the rest of the kit (despite the steep price we paid).
Luckily, we had the most important parts so we were not totally doomed. In an effort to compensate for the missing information, we relied on our mechanical engineering instincts and learned from videos found online of a similar action piece that was already built. With that, we have decided to create our own alternative parts to add to the mechanism prototype so that it will function properly and will be ready for presentation. The following video shows what we currently have:
The model doesn’t work properly yet, as the damper isn’t released with pressing the key and the hammer would mute the string whenever it’s pressed all the way down. We will have to fabricate some parts for the reset action to work properly, and there needs to be a mechanism that releases the hammer whenever it hits the string called the regulating button. Fixing these issues will get us a lot of experience, and once we are done, we will be able to say that we have achieved our first milestone between March 9th and March 17th.
Proposed Designs:
So far the team has conceived two ideas that are still a work in progress and another design that has been discarded. The discarded design was featured in the second blog update, where the redesigned action would have the linear motion of the pressing the key transformed to rotational motion using a wheel mechanism that would then allow the hammer to strike the string. We realized that this would not work when assembled together because the wheel would not actually be able to rotate, so we stopped developing the idea any further.
We will attempt to explain the two surviving designs as best as we can, but keep in mind that neither has been completed and that one is the product of a fever dream caused by a tonsil infection (we won’t tell you which one it is though). In order to explain the ideas behind the proposed designs, we feel like it’s necessary for the readers to have a basic understanding of how complex the action mechanism of a piano actually is; therefore, we have done our best to describe the device.
How it works:
As shown in the above figure, an upright piano action can be divided into six major components — key, wippen, jack, back check, hammer butt and felt hammer. In a nutshell, a downward force applied on the right end of the key would propel the left end of the same key in the upwards direction which, in–turn, would impart an upward momentum to wippen and jack. The jack would then push the hammer butt at an angle between 90° and 180° which would impart the horizontal force required to push the hammer in the horizontal direction to hit the string. The biggest gripe pianists have with this mechanism is that they must wait for the hammer to return to the hammer rail before another note can be played. This leads to lag which is especially evident during a complex repertoire recital, and thus, grand pianos are preferred by most pianists.
Repetition Mechanism Design:
We designed our own version of the repetitive mechanism. The whole
concept of this design is to enable the pianist to play another note without letting the hammer regain its original position. As shown in the figure above, we designed a two — fold link capable of folding onto itself. This link connects the left end of the “key” (point e) to the left end of the “wippen” (point h). The link at the left end of “wippen” (point h) is connected to a ball — bearing
which can rotate freely. The advantage of including ball — bearings is that when the “key” is pressed down, the link, which also experiences an upward motion, would be able to fold onto itself. Thus, it would not be able to interfere with counterclockwise rotation of “wippen” by applying a clockwise moment. When the key is pressed down, this new action works like a
traditional upright action, the workings of which have been explained in an earlier paragraph. However, the repetitive mechanism really comes into play when the applied force is removed. The weight at the left end of the “key” (point e) will push the “key” downwards if there is no force acting at the
right end. The two — fold link that got fold in the previous motion will unfold itself and apply a downward force at the left end of the “wippen” (point h). Now, it is important to realize that the “hammer” will try to get back to its original position once the force is removed. Also, the “wippen” can rotate freely around “pulley” (point p) and therefore the downward force at left end of the “wippen” (point h) would impart a counter — clockwise moment which would be propel the “back check” (point c) in the direction of the returning “hammer”. The resulting “hammer — back check” collision would reverse the direction of momentum of the “hammer”. The “hammer” would then be propelled in the direction of string before both “hammers” return to their respective position. Thus, this version of repetitive mechanism would enable a pianist to play two notes before the “hammer” return to its original position. This should drastically reduce the lag associated with modern upright actions.
Gear and Movable Rack Design:
The design shown above represents a hammer mechanism that is propelled by a gear and rack combination. The idea is based on our previous studies of design components in mechanical engineering, which are widely used on many other industries. Whenever we first saw the traditional upright action, one of our first instincts was to ask ourselves whether or not all those parts were really necessary, and the more we learned about the device, the more we realized how monumental the task ahead of us really is; however, we feel like our question still stands. Does the action of a piano really need a jack or a wippen? Perhaps it is possible that instead of modifying the piano by adding some parts, we could come up with a completely new mechanism that is analogous to the upright action; however, for it to be worth the effort, we must also include the better features of a gran piano mechanism of course.
On paper, this design is supposed to transfer the linear motion of pushing the key upward into rotational motion that makes the hammer hit the strings. The moving rack is connected to the back end of the key, and it is mounted on a rail that controls its movement. The purpose of this rail is to separate the rack and gear whenever the hammer hits the string; by doing so, the hammer will bounce back and let the strings vibrate. At the same time, the damper will be moved apart from the strings from the same motion and brought back on top of them whenever key is released. The reset and repetition mechanisms haven’t been included for this design, and the idea is still in a very early stage.
Unfortunately, there we are certain that challenges will arise. The two proposed designs might not work or survive scrutiny, the model of the upright action might not give the team an exact representation of how the upright works, or not having a model of the grand piano key to physically analyze will hinder our analysis; however, at least we are able to foresee these challenges, as I’m certain we will have to face unforeseen issues that will sow up until it’s time for validating our designs. Hopefully, the STEMinists will succeed.
