Student Lab Projects at Kettering University

Independent Study Research Projects in Acoustics

As much as possible, our applied physics faculty try to get undergraduate students involved in serious research in the acoustics laboratory. Sometimes this means helping a professor on a research problem which will result in a paper publication or presentation. Sometimes it involves a student research project as part of the required research component towards the physics degree. Or, sometimes it just means a student is interested in a problem and wants to spend time in the lab learning how to do research on an interesting topic. Here's a summary of some of the student research projects Dr. Russell and Dr. Ludwigsen have supervised in the Acoustics Lab here at Kettering Univeristy.

Updated photos coming soon!

The Helmholtz Resonator as a Passive Acoustic Absorber (Winter 2006)

Matthew Juszczyk (ME '06) supervised by Dr. Ludwigsen
The Helmholtz resonator is often used as a passive noise control device to attenuate narrowband sound propagating in a duct or tube. Examples might include applications in HVAC ducts and intake manifolds in automobiles. Mr. Juszczyk constructed a 2-m duct with a 10-cm square cross-section, and fitted a variable Helmholtz resonator to interact with the sound in the duct at its midpoint. One end is driven with a loudspeaker and features a microphone that may be used as a reference for normalization. The other end may be closed or left open. Measurements may be made with microphones placed at the other end, or at several locations near the resonator. In addition to use as a demonstration of standing waves in a tube, this apparatus can be used to show how the Helmholtz resonator filters or effectively absorbs sound at its resonance frequency. Theoretically, this resonance frequency depends on the geometry of the resonator. Mr. Juszczyk has measured the resonance frequencies for several volumes of the variable cavity, and verified the dependence on the inverse square root of volume. In addition, results of a finite element model of the duct and resonator were verified, with a similar inverse square root dependence on volume. Mr. JuszczykÕs finite element models offer insight into the interaction of the resonator with the standing wave in the tube, which goes beyond the simple lumped element treatment of the resonator as an acoustic analog to the electrical LC circuit.

Acoustics of Bottles (Fall 2005)

Cayla Jewett (AP '08) supervised by Dr. Ludwigsen
The sound produced when one blows across an open bottle results from the lowest resonance of that acoustic system. Much has been written about the bottle as a Helmholtz resonator, especially when there is a well-defined cavity and neck. Bottles are also used as examples of pipes closed at one end and open at the other. In that context, a series of resonances can be understood by comparing with a closed-open cylinder. Bottles that have a less distinct neck and cavity often fit this model better. By combining measurements made in the Acoustics Lab with finite element results, Ms. Jewett is reaching a better understanding of all bottle resonances based on the eigenmodes of the particular bottle shape. She is surveying the actual pressure distributions at mode frequencies in the lab, and comparing them to the predicted mode shapes from the finite element models. The goal is to arrive at general principles about the mode shapesÕ dependence on the bottle geometry.

Assessment of a Psuedo-Anechoic Space (Fall 2005)

Scott Porter (AP/ME '06) supervised by Dr. Ludwigsen
The Acoustics Lab is undergoing some remodeling and expansion, and part of this involves a temporary Òpseudo-anechoicÓ chamber for testing and experiments. Intended as a space with minimal reflections from surfaces, an anechoic chamber provides a free-field environment where the only sound is directly from a source under study. Our temporary facility uses fiberglass wedges (roughly 60 cm deep) along the walls and the floor, to absorb sound instead of reflect it. Mr. Porter has applied techniques described in the literature to characterize this space. One way to judge the effectiveness of the absorbing surfaces is to measure the reverberation time (RT60) at frequencies across the spectrum. A very short RT60 indicates that sound is readily being absorbed, while a long RT60 Ð especially at particular low frequencies Ð can indicate standing waves in the room that are not absorbed well. Mr. Porter also measured the decrease in sound pressure with increasing distance from a source placed in a corner. Measured by a microphone mounted on a wire extending from the source, the pressure should be inversely proportional to the square of the distance if the room provides a free field. Finally, Mr. Porter recognized the need for an omnidirectional sound source for future testing, and built a dodecahedral source with twelve speakers. The next steps in this project that await another student: testing and refinement of the dodecahedral/omnidirectional source, and design work for the more permanent anechoic test chamber to be installed in expanded Acoustics Lab (remodeling slated for late 2006).

Mapping the Vector Sound Intensity Field around a Tuning Fork (Winter 2005)

Justin Junell (AP '05) supervised by Dr. Russell
A tuning fork provides a very nice example of a simple vibrating system with a well defined transition from a complicated near-field to a more simple far-field sound radiation pattern. Theoretical predictions show that in the near-field sound energy circulates in an interesting manner with sound energy propagating away from the fork in some regions and towards the fork in others. In the far-field the sound energy radiates away from the fork in all directions, but with varying magnitude depending on the angular position around the fork. Justin experimentally mapped out the vector sound intensity using two different intensity probes. One was a traditional p-p probe which uses two phase matched microphones to measure the pressure (average pressure from the two microphones) and an estimate of the particle velocity (difference in pressures divided by their separation distance). The magnitude and direction of the vector intensity were obtained from the cross-spectrum of the two pressure microhone signals. The other probe was a new p-u probe which uses a very small microphone to measure the pressure and a tiny dual hot-wire anemometer to measure the particle velocity directly. The vector intensity is obtained from the cross-spectrum of the pressure and particle velocity signals. The tuning fork was driven by an electromagnet coil and a small magnet attached to one of the fork tines, and the fork rotated on a turntable. Justin measured the radial and angular components of the intensity at a number distances from the fork. The measured vector intensity map agreed very well with the theoretical predictions.

Indirect methods for assessing baseball and softball bat performance (Summer 2005)

Jeremy Bemis (AP '06) supervised by Dr. Russell
: The current methods for assessing the performance of baseball and softball bats used by the governing bodies cannot be used in field applications, such as games and tournaments. In this study, I performed four indirect test methods; hoop frequency, compression, bat-swing pendulum contact time, and pendulum contact time. I performed each these indirect methods on 13 different softball bats and compared the result to the BPF data that was supplied with the bats. There were four objectives for this project:

Increase the understanding of the indirect methods.
Correlate the indirect methods to BPF.
Identify methods with greatest correlation.
Identify possible methods for field applications.
The comparisons showed that the hoop frequency and compression both have a strong correlation to the BPF. Both of the pendulum methods had a slight correlation to the BPF, but they also had significant scatter. For the bat-swing pendulum, this scatter could be attributed to vibrations and shifting of the test fixture. The pendulum method is going to need more testing to determine where the scattering is coming from. My conclusions are that the hoop frequency and compression tests both show great promise, but comparisons need to be made with more performance data. The bat-swing pendulum test fixture needs to be redesigned and retested. The pendulum method needs more testing to fully understand what is taking place during the test.

Acoustics of the Kalimba (Fall 2004)

William Rein (ME'05) supervised by Dr. Ludwigsen
The kalimba is a type of thumb piano developed by musicologist Hugh Tracey, and handcrafted in South Africa. These instruments are popular across Africa and beyond, and are a part of traditional religious ceremony of the Shona people. The plucked keys are thin strips of spring steel, mounted on a hardwood resonator box. The box has a large hole in the top plate, and two small holes in the back plate. These are used to create a vibrato effect by rapidly opening and closing them with the fingers. Mr. Rein made recordings of the instrument to quantify its timbre through spectral analysis. Then, to begin to understand how that timbre is created, he performed a modal analysis of the resonator box. While the tones of the instrument at playing frequencies (between 240 and 1100 Hz) dominate the spectra, the frequencies of the box modes (above 1300 Hz) are important to the click that forms the attack of kalimba notes. Mr. Rein also assisted with a separate study to understand the vibrato effect. The hollow box and its holes forms a Helmholtz resonator. Just as a bottle sounds a particular frequency when one blows across its neck, this box has a particular resonance that tends to emphasize one note above others. The frequency of this resonance is altered when the player opens and closes the holes in the back plate. The vibrato effect is the emphasis on a played note coming and going as the player opens and closes the holes.
This researcn was presented at a national meeting of the Acoustical Society of America. Abstract published as:
D. O. Ludwigsen, "Measurements of the acoustics of the kalimba," J. Acoust. Soc. Am., 116 2593 (2004). A more indepth paper is currently being written:
D. O. Ludwigsen and William Rein, "Acoustic and structural characteristics of the kalimba," to be submitted to J. Acoust. Soc. Am./manuscript in preparation.

Pulsed Magnetic Treatment of Work-hardened Aluminum Baseball Bats (Summer & Fall 2004)

Justin Junell (AP '05) and Scott Porter (AP/ME '06) supervised by Dr. Russell

Experimental Determination of End Correction for Flanged Pipes (Spring 2004)

Brandon Dilworth (ME '04) supervised by Dr. Ludwigsen
Standing waves in pipes are fundamental in courses in acoustics. Certain frequencies resonate in a pipe of a given length, giving tone color to organ pipes or causing high back pressure in an exhaust system. When the end of the pipe is open, the effective acoustic length is slightly longer because the air molecules at the open end form a mass load. The difference between the physical length and the effective acoustic length is called an end correction, and is proportional to the radius of the pipe. The constant of proportionality depends on the type of termination, ranging from a thin-walled pipe in isolation to a pipe that ends in an infinitely large flange. Mr. Dilworth measured the resonance frequencies for several pipes of different radius and varied flanges. In carrying out the experiment, he controlled many factors that affect the results, from issues of experimental precision in measurements of length and radius, to acoustic concerns such as precise temperature control and a need for an anechoic environment.
This researcn was presented at a national meeting of the Acoustical Society of America. Abstract published as:
D. O. Ludwigsen and B. J. Dilworth, "End correction for an open pipe from measured resonance frequencies," J. Acoust. Soc. Am., 116 2592 (2004).

Artificial Lips for Brass Instrument Research (Summer 2003)

Daniel Neill (AP/ME '04) supervised by Dr. Ludwigsen
The structural oscillation of the lips are coupled with the aerodynamics of a brass instrument in order to produce a musical tone. Current models of this interaction are nonlinear and represent the lips as an oscillator with two degrees of freedom, like a mass on springs that can move up and down as well as left and right. For greater understanding of the lip valve, an artificial apparatus can provide insight in addition to benefits of repeatability and a great deal of control. Mr. Neill investigated previous work done in France and Scotland, and designed and built an artificial player with wood, PVC, and water-filled latex tubing to be the lips. The embouchure can be controlled by the longitudinal tension in the latex tubes, as well as the transverse pressure applied to keep the tubes side by side. With a suitable air supply and some additional refinement, this apparatus will enable possible future study of the nonlinear propagation of sound in the instrument.

Input Impedance of Brass Instruments (Summer 2003 and Winter 2004)

Timothy Swieter (ME '04) and Chidi Uhiara (ME '04) supervised by Dr. Ludwigsen
The input impedance is a relationship that tells us how much acoustic pressure is required at the input of a system to yield a given acoustic flow into the system. Brass instruments play at frequencies where this input impedance is greatest. In other words, the instrument is best able to force the lips to oscillate in a regular pattern when the input pressure (in the mouthpiece) is largest. Thus, the input impedance function for a given instrument can indicate how easy it will be to sound notes and to play with good intonation. In order to measure the input impedance, a microphone may record the pressure, but the flow or volume velocity must also be taken into account.
Mr. Swieter initiated the construction of a device to measure input impedance. The key element was a piezoelectric disc. These are made from crystals that deform when a voltage is applied; small buzzers employ such discs to generate sound. With the disc mounted in a small tube that can be fixed to the rim of a trombone mouthpiece, he injected that sound into the instrument. A small microphone was also mounted in the tube to measure pressure. Mr. Uhiara assumed the task of calibration, critical to removing the frequency characteristics of the piezoelectric disc. The disc itself has a resonance, so some frequencies sound louder than others. Using an inverse filter to compensate for this effect, the disc would theoretically become a constant source of flow at all frequencies. Then the signal from the microphone would be proportional to input impedance. The results from the prototype were encouraging, showing the expected input impedance maxima and minima for a cylindrical calibration pipe. However, the resonance of the piezoelectric disc was difficult to remove entirely.

Binaural head transfer functions for virtual localization (Winter-Spring 2002)

Jason Kolenda (AP '03) and Kimberley Pregitzer (EE '03) supervised by Dr. Ludwigsen
Human hearing has evolved to recognize the effects of differences between sensations from the left and right ears. Cues based on different levels and timing help people to localize the direction of the sound source within the horizontal plane. Mr. Kolenda studied these (Interaural Level Difference, or ILD, and Interaural Time Difference, ITD) using the binaural head (HEAD Acoustics HMS I). A source was placed at the same level as the mannequinÕs ears at various locations. The signals from the ear microphones were analyzed for ILD and ITD. Results for the ITD fit well with the theory, which predicts the time difference between sound arriving at opposite sides of a spherical head.
The brain is able to localize sound above and below the horizontal plane, so the ILD and ITD are not sufficient to explain localization. Ms. Pregitzer continued the work by exploring the Anatomical Transfer Function (ATF). This relationship between the signal at an ear (left or right) and the incoming sound quantifies the effects of ear, head and torso shape, as well as subtle factors such as hairstyle. She recorded signals from the binaural head ear microphones for impulsive sources placed at locations in the horizontal plane at ten-degree intervals. A future project may fully map the entire 3-D space by placing sources above and below the horizontal plane. She convolved the recordings with a sample of speech. The result is a virtual soundscape, where a listener wearing headphones can hear the speech as if it comes from any location around the horizontal plane.

Acoustical Study of the Djembe African Hand Drum (Winter 2002)

Wes Haveman (AP/EE '02) supervised by Dr. Russell
The djembe is a hollow drum, originating from Africa, which is capable of producing three distinct sounds: bass, tone, and slap. Three djembe drums (two handmade and one factory made) were studied in this experiment in an attempt to determine how the vibrational modes of the shell, the resonances of the cavity, and the vibrational modes of the membrane individually contribute and couple together to produce the three distinct djembe sounds. Vibrational modes of the shell were determined via an experimental modal analysis experiment. Frequencies of the various shell modes are dependent on the tightness cables used to provide tension to the drum. There was no clear relationship between the shell modes and the drum sounds.
Membrane modes were determined by first identifying the resonance frequencies from a swept-sine analyasis (driving the membrane with a magnetic coil and a small magnet), and then mapping the modes with a microphone and watching the phase changes on a oscilloscope Lissajous pattern (see figure at top right). Frequencies were found to be dependent on membrane tension, but frequency ratios remained relatively constant and seem to be a characterstic of the drum.
Cavity modes were also identified - the most important cavity mode is the "Helmholtz" mode which creates the loud bass sound of the drum. Other higher order cavity modes were found to potentially couple with membrane modes.
At this point in time we are still trying to identify which membrane/cavity mode combinations are primarily responsible for the distinct tone and slap sounds. Our analysis was frustrated by the discovery that CD recordings by professional players reveal two separate slap sounds (a fifth and an octave above the tone) whereas most of what is written about the djembe (and our playing ability) suggests one slap sound which is approximately a fourth above the tone.
This research is continuing.
Preliminary results were presented:
D. A. Russell and W. Haveman, "Acoustic and modal analysis of an African djembe drum," 140th meeting of the Acoustical Society of America, Newport Beach, CA, December 4-8, 2000, J. Acoust. Soc. Am., 108(5) Pt 2., 2591, (2000)

The Helmholtz Resonator as a Passive Acoustic Absorber (Winter 2006)
Matthew Juszczyk (ME '06) supervised by Dr. Ludwigsen The Helmholtz resonator is often used as a passive noise control device to attenuate narrowband sound propagating in a duct or tube. Examples might include applications in HVAC ducts and intake manifolds in automobiles. Mr. Juszczyk constructed a 2-m duct with a 10-cm square cross-section, and fitted a variable Helmholtz resonator to interact with the sound in the duct at its midpoint. One end is driven with a loudspeaker and features a microphone that may be used as a reference for normalization. The other end may be closed or left open. Measurements may be made with microphones placed at the other end, or at several locations near the resonator. In addition to use as a demonstration of standing waves in a tube, this apparatus can be used to show how the Helmholtz resonator filters or effectively absorbs sound at its resonance frequency. Theoretically, this resonance frequency depends on the geometry of the resonator. Mr. Juszczyk has measured the resonance frequencies for several volumes of the variable cavity, and verified the dependence on the inverse square root of volume. In addition, results of a finite element model of the duct and resonator were verified, with a similar inverse square root dependence on volume. Mr. JuszczykÕs finite element models offer insight into the interaction of the resonator with the standing wave in the tube, which goes beyond the simple lumped element treatment of the resonator as an acoustic analog to the electrical LC circuit.
Acoustics of Bottles (Fall 2005)
Cayla Jewett (AP '08) supervised by Dr. Ludwigsen The sound produced when one blows across an open bottle results from the lowest resonance of that acoustic system. Much has been written about the bottle as a Helmholtz resonator, especially when there is a well-defined cavity and neck. Bottles are also used as examples of pipes closed at one end and open at the other. In that context, a series of resonances can be understood by comparing with a closed-open cylinder. Bottles that have a less distinct neck and cavity often fit this model better. By combining measurements made in the Acoustics Lab with finite element results, Ms. Jewett is reaching a better understanding of all bottle resonances based on the eigenmodes of the particular bottle shape. She is surveying the actual pressure distributions at mode frequencies in the lab, and comparing them to the predicted mode shapes from the finite element models. The goal is to arrive at general principles about the mode shapesÕ dependence on the bottle geometry.
Assessment of a Psuedo-Anechoic Space (Fall 2005)
Scott Porter (AP/ME '06) supervised by Dr. Ludwigsen The Acoustics Lab is undergoing some remodeling and expansion, and part of this involves a temporary Òpseudo-anechoicÓ chamber for testing and experiments. Intended as a space with minimal reflections from surfaces, an anechoic chamber provides a free-field environment where the only sound is directly from a source under study. Our temporary facility uses fiberglass wedges (roughly 60 cm deep) along the walls and the floor, to absorb sound instead of reflect it. Mr. Porter has applied techniques described in the literature to characterize this space. One way to judge the effectiveness of the absorbing surfaces is to measure the reverberation time (RT60) at frequencies across the spectrum. A very short RT60 indicates that sound is readily being absorbed, while a long RT60 Ð especially at particular low frequencies Ð can indicate standing waves in the room that are not absorbed well. Mr. Porter also measured the decrease in sound pressure with increasing distance from a source placed in a corner. Measured by a microphone mounted on a wire extending from the source, the pressure should be inversely proportional to the square of the distance if the room provides a free field. Finally, Mr. Porter recognized the need for an omnidirectional sound source for future testing, and built a dodecahedral source with twelve speakers. The next steps in this project that await another student: testing and refinement of the dodecahedral/omnidirectional source, and design work for the more permanent anechoic test chamber to be installed in expanded Acoustics Lab (remodeling slated for late 2006).
Mapping the Vector Sound Intensity Field around a Tuning Fork (Winter 2005)
Justin Junell (AP '05) supervised by Dr. Russell A tuning fork provides a very nice example of a simple vibrating system with a well defined transition from a complicated near-field to a more simple far-field sound radiation pattern. Theoretical predictions show that in the near-field sound energy circulates in an interesting manner with sound energy propagating away from the fork in some regions and towards the fork in others. In the far-field the sound energy radiates away from the fork in all directions, but with varying magnitude depending on the angular position around the fork. Justin experimentally mapped out the vector sound intensity using two different intensity probes. One was a traditional p-p probe which uses two phase matched microphones to measure the pressure (average pressure from the two microphones) and an estimate of the particle velocity (difference in pressures divided by their separation distance). The magnitude and direction of the vector intensity were obtained from the cross-spectrum of the two pressure microhone signals. The other probe was a new p-u probe which uses a very small microphone to measure the pressure and a tiny dual hot-wire anemometer to measure the particle velocity directly. The vector intensity is obtained from the cross-spectrum of the pressure and particle velocity signals. The tuning fork was driven by an electromagnet coil and a small magnet attached to one of the fork tines, and the fork rotated on a turntable. Justin measured the radial and angular components of the intensity at a number distances from the fork. The measured vector intensity map agreed very well with the theoretical predictions.
Indirect methods for assessing baseball and softball bat performance (Summer 2005)
Jeremy Bemis (AP '06) supervised by Dr. Russell : The current methods for assessing the performance of baseball and softball bats used by the governing bodies cannot be used in field applications, such as games and tournaments. In this study, I performed four indirect test methods; hoop frequency, compression, bat-swing pendulum contact time, and pendulum contact time. I performed each these indirect methods on 13 different softball bats and compared the result to the BPF data that was supplied with the bats. There were four objectives for this project: Increase the understanding of the indirect methods. Correlate the indirect methods to BPF. Identify methods with greatest correlation. Identify possible methods for field applications. The comparisons showed that the hoop frequency and compression both have a strong correlation to the BPF. Both of the pendulum methods had a slight correlation to the BPF, but they also had significant scatter. For the bat-swing pendulum, this scatter could be attributed to vibrations and shifting of the test fixture. The pendulum method is going to need more testing to determine where the scattering is coming from. My conclusions are that the hoop frequency and compression tests both show great promise, but comparisons need to be made with more performance data. The bat-swing pendulum test fixture needs to be redesigned and retested. The pendulum method needs more testing to fully understand what is taking place during the test.
Acoustics of the Kalimba (Fall 2004)
William Rein (ME'05) supervised by Dr. Ludwigsen The kalimba is a type of thumb piano developed by musicologist Hugh Tracey, and handcrafted in South Africa. These instruments are popular across Africa and beyond, and are a part of traditional religious ceremony of the Shona people. The plucked keys are thin strips of spring steel, mounted on a hardwood resonator box. The box has a large hole in the top plate, and two small holes in the back plate. These are used to create a vibrato effect by rapidly opening and closing them with the fingers. Mr. Rein made recordings of the instrument to quantify its timbre through spectral analysis. Then, to begin to understand how that timbre is created, he performed a modal analysis of the resonator box. While the tones of the instrument at playing frequencies (between 240 and 1100 Hz) dominate the spectra, the frequencies of the box modes (above 1300 Hz) are important to the click that forms the attack of kalimba notes. Mr. Rein also assisted with a separate study to understand the vibrato effect. The hollow box and its holes forms a Helmholtz resonator. Just as a bottle sounds a particular frequency when one blows across its neck, this box has a particular resonance that tends to emphasize one note above others. The frequency of this resonance is altered when the player opens and closes the holes in the back plate. The vibrato effect is the emphasis on a played note coming and going as the player opens and closes the holes. This researcn was presented at a national meeting of the Acoustical Society of America. Abstract published as: D. O. Ludwigsen, "Measurements of the acoustics of the kalimba," J. Acoust. Soc. Am., 116 2593 (2004). A more indepth paper is currently being written: D. O. Ludwigsen and William Rein, "Acoustic and structural characteristics of the kalimba," to be submitted to J. Acoust. Soc. Am./manuscript in preparation.
Pulsed Magnetic Treatment of Work-hardened Aluminum Baseball Bats (Summer & Fall 2004)
Justin Junell (AP '05) and Scott Porter (AP/ME '06) supervised by Dr. Russell
Experimental Determination of End Correction for Flanged Pipes (Spring 2004)
Brandon Dilworth (ME '04) supervised by Dr. Ludwigsen Standing waves in pipes are fundamental in courses in acoustics. Certain frequencies resonate in a pipe of a given length, giving tone color to organ pipes or causing high back pressure in an exhaust system. When the end of the pipe is open, the effective acoustic length is slightly longer because the air molecules at the open end form a mass load. The difference between the physical length and the effective acoustic length is called an end correction, and is proportional to the radius of the pipe. The constant of proportionality depends on the type of termination, ranging from a thin-walled pipe in isolation to a pipe that ends in an infinitely large flange. Mr. Dilworth measured the resonance frequencies for several pipes of different radius and varied flanges. In carrying out the experiment, he controlled many factors that affect the results, from issues of experimental precision in measurements of length and radius, to acoustic concerns such as precise temperature control and a need for an anechoic environment. This researcn was presented at a national meeting of the Acoustical Society of America. Abstract published as: D. O. Ludwigsen and B. J. Dilworth, "End correction for an open pipe from measured resonance frequencies," J. Acoust. Soc. Am., 116 2592 (2004).
Artificial Lips for Brass Instrument Research (Summer 2003)
Daniel Neill (AP/ME '04) supervised by Dr. Ludwigsen The structural oscillation of the lips are coupled with the aerodynamics of a brass instrument in order to produce a musical tone. Current models of this interaction are nonlinear and represent the lips as an oscillator with two degrees of freedom, like a mass on springs that can move up and down as well as left and right. For greater understanding of the lip valve, an artificial apparatus can provide insight in addition to benefits of repeatability and a great deal of control. Mr. Neill investigated previous work done in France and Scotland, and designed and built an artificial player with wood, PVC, and water-filled latex tubing to be the lips. The embouchure can be controlled by the longitudinal tension in the latex tubes, as well as the transverse pressure applied to keep the tubes side by side. With a suitable air supply and some additional refinement, this apparatus will enable possible future study of the nonlinear propagation of sound in the instrument.
Input Impedance of Brass Instruments (Summer 2003 and Winter 2004)
Timothy Swieter (ME '04) and Chidi Uhiara (ME '04) supervised by Dr. Ludwigsen The input impedance is a relationship that tells us how much acoustic pressure is required at the input of a system to yield a given acoustic flow into the system. Brass instruments play at frequencies where this input impedance is greatest. In other words, the instrument is best able to force the lips to oscillate in a regular pattern when the input pressure (in the mouthpiece) is largest. Thus, the input impedance function for a given instrument can indicate how easy it will be to sound notes and to play with good intonation. In order to measure the input impedance, a microphone may record the pressure, but the flow or volume velocity must also be taken into account. Mr. Swieter initiated the construction of a device to measure input impedance. The key element was a piezoelectric disc. These are made from crystals that deform when a voltage is applied; small buzzers employ such discs to generate sound. With the disc mounted in a small tube that can be fixed to the rim of a trombone mouthpiece, he injected that sound into the instrument. A small microphone was also mounted in the tube to measure pressure. Mr. Uhiara assumed the task of calibration, critical to removing the frequency characteristics of the piezoelectric disc. The disc itself has a resonance, so some frequencies sound louder than others. Using an inverse filter to compensate for this effect, the disc would theoretically become a constant source of flow at all frequencies. Then the signal from the microphone would be proportional to input impedance. The results from the prototype were encouraging, showing the expected input impedance maxima and minima for a cylindrical calibration pipe. However, the resonance of the piezoelectric disc was difficult to remove entirely.
Binaural head transfer functions for virtual localization (Winter-Spring 2002)
Jason Kolenda (AP '03) and Kimberley Pregitzer (EE '03) supervised by Dr. Ludwigsen Human hearing has evolved to recognize the effects of differences between sensations from the left and right ears. Cues based on different levels and timing help people to localize the direction of the sound source within the horizontal plane. Mr. Kolenda studied these (Interaural Level Difference, or ILD, and Interaural Time Difference, ITD) using the binaural head (HEAD Acoustics HMS I). A source was placed at the same level as the mannequinÕs ears at various locations. The signals from the ear microphones were analyzed for ILD and ITD. Results for the ITD fit well with the theory, which predicts the time difference between sound arriving at opposite sides of a spherical head. The brain is able to localize sound above and below the horizontal plane, so the ILD and ITD are not sufficient to explain localization. Ms. Pregitzer continued the work by exploring the Anatomical Transfer Function (ATF). This relationship between the signal at an ear (left or right) and the incoming sound quantifies the effects of ear, head and torso shape, as well as subtle factors such as hairstyle. She recorded signals from the binaural head ear microphones for impulsive sources placed at locations in the horizontal plane at ten-degree intervals. A future project may fully map the entire 3-D space by placing sources above and below the horizontal plane. She convolved the recordings with a sample of speech. The result is a virtual soundscape, where a listener wearing headphones can hear the speech as if it comes from any location around the horizontal plane.
Acoustical Study of the Djembe African Hand Drum (Winter 2002)
Wes Haveman (AP/EE '02) supervised by Dr. Russell The djembe is a hollow drum, originating from Africa, which is capable of producing three distinct sounds: bass, tone, and slap. Three djembe drums (two handmade and one factory made) were studied in this experiment in an attempt to determine how the vibrational modes of the shell, the resonances of the cavity, and the vibrational modes of the membrane individually contribute and couple together to produce the three distinct djembe sounds. Vibrational modes of the shell were determined via an experimental modal analysis experiment. Frequencies of the various shell modes are dependent on the tightness cables used to provide tension to the drum. There was no clear relationship between the shell modes and the drum sounds. Membrane modes were determined by first identifying the resonance frequencies from a swept-sine analyasis (driving the membrane with a magnetic coil and a small magnet), and then mapping the modes with a microphone and watching the phase changes on a oscilloscope Lissajous pattern (see figure at top right). Frequencies were found to be dependent on membrane tension, but frequency ratios remained relatively constant and seem to be a characterstic of the drum. Cavity modes were also identified - the most important cavity mode is the "Helmholtz" mode which creates the loud bass sound of the drum. Other higher order cavity modes were found to potentially couple with membrane modes. At this point in time we are still trying to identify which membrane/cavity mode combinations are primarily responsible for the distinct tone and slap sounds. Our analysis was frustrated by the discovery that CD recordings by professional players reveal two separate slap sounds (a fifth and an octave above the tone) whereas most of what is written about the djembe (and our playing ability) suggests one slap sound which is approximately a fourth above the tone. This research is continuing. Preliminary results were presented: D. A. Russell and W. Haveman, "Acoustic and modal analysis of an African djembe drum," 140th meeting of the Acoustical Society of America, Newport Beach, CA, December 4-8, 2000, J. Acoust. Soc. Am., 108(5) Pt 2., 2591, (2000)